Choroidal Neovascularization

Choroidal
Neovascularization
Jay Chhablani
Editor
123
Choroidal Neovascularization
Jay Chhablani
Editor
Choroidal
Neovascularization
Editor
Jay Chhablani
Department of Ophthalmology
University of Pittsburgh Medical Center
Pittsburgh, PA
USA
ISBN 978-981-15-2212-3 ISBN 978-981-15-2213-0 (eBook)

https://doi.org/10.1007/978-981-15-2213-0
© Springer Nature Singapore Pte Ltd. 2020

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or
part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way,
and transmission or information storage and retrieval, electronic adaptation, computer software,
or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in
this book are believed to be true and accurate at the date of publication. Neither the publisher nor
the authors or the editors give a warranty, expressed or implied, with respect to the material
contained herein or for any errors or omissions that may have been made. The publisher remains
neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore
189721, Singapore
To my family, for standing by me, through everything
To my mentors, for showing me the right path
To my fellows, for asking the right questions
Foreword
Nearly two decades ago, I read a fascinating and illustrated book The Hole in
My Vision (Penfield Press, 2000) authored by the University of Iowa ophthal-
mic artist and photographer Lee Allen. This book is an illustrative documen-
tation of how a patient sees the world with progression of age-related macular
degeneration (AMD), from the stage of drusen to formation of the choroidal
neovascular membrane (CNVM), and the response to repeated laser treat-
ment, including the entopic phenomenon.
CNVM is classically described in AMD, but there are other causes, and
some of the important ones include the polypoid choroidal vasculopathy
(PCV; also called Asian AMD) and pathological myopia. Today, CNVM
associated with any retinal disease has several treatment options other than
laser suggested by the Macular Photocoagulation Study (MPS) in the 1980s.
Our understanding of the disease pathology and natural history have expanded
with new knowledge of underlying biological mechanism (vascular endothe-
lial growth factor, VEGF, and several complement factors), improved imag-
ing modalities (indocyanine angiography, optical coherence tomography, and
OCT angiography), and newer treatment options directed from mere preser-
vation of residual vision to possible improvement of vision (photodynamic
therapy, intravitreal anti-VEGF therapy, and vitreoretinal surgery).
Pharmaceutical companies and regulatory authority sponsored clinical trials
have increased the opportunity and confidence of the treating physicians to
practice evidence-based medicine.
The book, Choroidal neovascularization, captures it all from history to
basics, from clinical entities to lessons from clinical trials, and from treat-
ment to rehabilitation. Have we reached the last mile in understanding and
care of CNVMs?
Not yet.
One of the hurdles with the current standard of care with the intravitreal
anti-VEGF is the need for repeat injection, nearly every month, due to short
half-life of the currently used molecules. This reduces patient compliance,
increases both direct and indirect costs of care, and exposes the patient to
some of the injection-related adverse events, such as endophthalmitis. Thus,
there is a need for a longer duration anti-VEGF molecule or a surgically
placed depot device with slow and pulsed release of the anti-VEGF molecule.
But would blocking of VEGF-A, as done by ranibizumab, suffice? Scientists
have already identified the role of VEGF-B and placental growth factor, and
there are molecules, such as aflibercept, that block these substances. Current
vii
viii Foreword
research is also directed to block the complement factors, and lampalizumab

(blocks Complement Factor D) is one such monoclonal antibody currently
under trial. Unfortunately, there is a ceiling effect of all these pharmacologi-
cal treatments. In such a situation, visual rehabilitation with a variety of mag-
nifying system helps. While most times the magnifying systems are external
devices, hand-held or table-top magnifiers, an implantable miniature tele-
scope (IMT) in the capsular bag after a standard cataract extraction is also a
distinct possibility to improving the near vision.
Randomized control trial (RCT) is the gold standard to create evidence-
based practice guidelines. But it is expensive and does not necessarily adhere
to the real-world patient criteria. Hence clubbing the RCT information with
the patient’s preferences and the clinician’s expertise creates a new standard
of care that is closer to the real-world medicine and meets patient’s medical
and social need. The pro re nata (PNR) strategy of intravitreal anti-VEGF in
CNVM is a step close to the real-world patient care. The scientific way of
reaching a consensus of treatment decision that is between the RCT and the
clinical expertise is to answer PICO (Patient, Intervention, Comparison,
Outcome) questions of any specific disease. This is more than necessary in
care of CNVMs.
Two important chapters in the book are the “Future Therapy” and
“Rehabilitation.” The future of care for CNVM is promising with newer
imaging technology, with newer drugs to block choroidal bleeding, and with
the possibility of stem cell therapy to reverse the disease process. And where
these medical strategies stop, visual rehabilitation begins so that the quality
of life is not decreased to depression.
This book, Choroidal Neovascularization, would be an excellent compre-
hensive treatise on the subject and would fulfill a long-felt need of ophthal-
mologists at all levels of training and practice.
Taraprasad Das
L V Prasad Eye Institute, Hyderabad, India
Foreword
The most exciting advances in ophthalmology over the past two decades are
those that have advanced the diagnosis and management of chorioretinal vas-
cular diseases. These have been driven by the invention of laser-based imag-
ing technology, particularly optical coherence tomography (OCT), and the
concomitant development of retinal pharmacotherapy, especially those drugs
that inhibit the actions of vascular endothelial growth factor (VEGF). Together
these technologies have revolutionized the management of a variety of cho-
rioretinal vascular diseases, but most importantly they enable us to success-
fully treat choroidal neovascularization. This breakthrough comes none too
soon since neovascular age-related macular degeneration (nAMD) is the
leading cause of blindness in the industrialized world, and the post-World
War 2 baby boomers of North America and Europe continue to swell the at-
risk populations.
Laser photocoagulation was introduced in the late 1960s to treat choroidal
neovascular membranes (CNVM), but after 30 years of disappointing results
physicians clamored for more effective therapy. Photodynamic therapy was
greeted with great fanfare (2000), but it only served to bridge the gap between
laser and the soon-to-be introduced, much more effective, pharmacotherapy.
The pathway to anti-VEGF therapy began (1983) with the little heralded
discovery of vasopermeability factor (VPF) and the subsequent (1989) iden-
tification and sequencing of vascular endothelial growth factor, which turned
out to be the previously described VPF. After 15 additional years of prodi-
gious basic science discoveries and elegant drug development, the anti-VEGF
drugs pegaptanib and bevacizumab were approved (2004). The subsequent 15
years have featured successful pivotal trials for the treatment of CNVM due
to both nAMD and myopia, parallel advances in OCT, and approval of new
anti-VEGF drugs.
Into this rapidly changing field enters an ambitious text by Dr. Chhablani
and colleagues. The authors have jumped head first into one of the most
intensively researched areas of ophthalmology with a comprehensive work,
the likes of which we have not previously seen. Other texts have discussed
CNVM within the broader context of conditions such as nAMD, but none
have been written with a focus on CNVM itself.
Senior ophthalmologists will enjoy revisiting the history of CNVM, while
their younger, less seasoned colleagues will benefit from learning what
patients and physicians faced in the past few decades. Appreciating those past
medical successes and failures is one of the best ways to both understand the
ix
x Foreword
present and prepare oneself for the future. I am pleased that the authors have
devoted considerable space to the pathophysiology of CNVM since both
basic scientists and research-oriented medical retina specialists will appreci-
ate the detailed discussion of CNVM biology.
The main body of this book discusses CNVM within the context of several
important retinal diseases. Readers will be treated to state-of-the-art discus-
sions of AMD and a variety of other macular conditions, with neovasculariza-
tion from the choriocapillaris serving as their common thread.
The release of this book is particularly well timed since we are entering a
period of accelerated drug development for the treatment of CNVM. For the
first time, investigational drugs promise extended duration anti-VEGF ther-
apy and true combination therapy for CNVM. The latter resembles the strate-
gies commonly employed by oncologists, an approach that has been
anticipated by retina specialists for several years.
The book will appeal to a broad group of medical professionals since it
contains useful information for everyone. Medical students taking ophthal-
mology rotations will find important clinical information upon which to build
their knowledge bases; ophthalmology residents will acquire principles of
patient management and understand the needs for referral; retina fellows will
use this to manage a broad spectrum of medical retina conditions; general
ophthalmologists will familiarize themselves with the latest research; and
retina specialists will enjoy the detailed coverage of pivotal clinical trials and
important pathophysiology.
My only concern about this volume, which has nothing to do with the
quality and scope of the work, is its durability within the rapidly changing
CNVM landscape. If this book receives the accolades it deserves, the authors
may need to publish frequent updates to keep us up to date.
Finally, I have known Dr. Chhablani for the past seven years and I am
honored to write this foreword. Dr. Chhablani has established himself as one
of the rising stars in ophthalmology, and this book adds to the volume of qual-
ity work he has produced in his brief academic career. I look forward to the
success of this book and I eagerly await future publications by this talented,
industrious, young physician.
Michael W. Stewart
Mayo Clinic Florida, Jacksonville, FL, USA
Preface
In the last three decades, with advanced imaging, the understanding and man-
agement of choroidal neovascularization (CNV) has been transformed.
Through extensive basic research, several pathomechanisms have been estab-
lished over the years, and subsequently various therapeutic approaches have
been developed toward these molecules and/or biological steps. A recent
surge in newer therapeutic approaches to prevent, stop, or slow the CNV
progression has brought hope for the patients. Advancement in monitoring,
rehabilitation, and use of peripheral vision is a new dimension in this field.
CNV has been part of many comprehensive books on the retina; however,
there is no book which focuses on this specific entity.
Book titled Choroidal neovascularization includes sections on basics,
clinical conditions associated with choroidal neovascularization (CNV), clin-
ical trials related to CNV in various conditions, future directions, and reha-
bilitation. The section on “Basics” includes chapters on pathogenesis,
proposed mechanisms, disease models, histopathology, and electron micros-
copy. The section on “Clinical conditions” includes CNV secondary to vari-
ous clinical conditions including common conditions such as age-related
macular degeneration, myopia, and uncommon conditions such as choroidal
osteoma. This section includes various clinical features, imaging characteris-
tics, treatment approach, and prognosis. The section on “Clinical trials”
describes clinical trials in various CNV conditions with recent updates. The
section on “Future directions” includes stem cell therapy, gene therapy, newer
molecules, and lasers. Rehabilitation is an important aspect of the manage-
ment of this disease; a focused section on “Rehabilitation” includes home
monitoring and low vision aids.
This is a comprehensive unique book on “CNV.” Unlike other books, this
book comprises all aspects of “CNV” from bench to bedside including future
directions and rehabilitation, with updated information on clinical trials. All
clinical chapters cover the latest evidence-based approach in the diagnosis
and management of CNV and supplemented with many illustrations and
figures.
I would like to thank all the authors who are experts in this field and have
contributed their best work to make this book very comprehensive, up-to-
date, and state of the art. I hope the readers would enjoy reading this book and
this helps in better patient management.
Pittsburgh, PA Jay Chhablani
xi
Contents
1 Choroidal Neovascular Membrane: Historical Perspectives �� 1

Aniruddha Agarwal and Krinjeela Bazgain
Part I Basic Science
2 Pathogenesis of Choroidal Neovascularization �� 7

Mayss Al-Sheikh and Daniel Barthelmes
3 Histopathology of Choroidal Neovascularization �� 15
Evangelina Esposito, Julio A. Urrets-Zavalia,
and Pablo Zoroquiain
4 Choroidal Neovascularization Animal Models�� 37
Takayuki Baba
5 Electron Microscopy of the Choroidal Neovascularization�� 47
Tapas C. Nag and Sneha Gupta
6 Genomics in Choroidal Neovascularization�� 57
Kenji Yamashiro
Part II Clinical Entities
7 Neovascular AMD: Clinical Features and Imaging�� 73

Cláudia Farinha and Rufino Silva
8 Management Strategies for Neovascular AMD�� 99
Irmela Mantel
9 Myopic Choroidal Neovascularization�� 109
Seung-Young Yu and Kiyoung Kim
10 Pachychoroid-Related Choroidal Neovascularization�� 117
Apoorva Ayachit and Jay Chhablani
11 Inflammatory Choroidal Neovascularization�� 129
Alvaro Olate-Perez, Carolina Bernal-Morales,
Aina Moll-Udina, Alfredo Adan, and Javier Zarranz-Ventura
xiii
xiv Contents
12 Dystrophy-Related Choroidal Neovascularization �� 139

Pierluigi Iacono, Stefano Da Pozzo, Alessandro Papayannis,
Francesco Romano, Alessandro Arrigo,
and Maurizio Battaglia Parodi
13 Choroidal Neovascularization Associated with
Angioid Streaks�� 151
Christof Hänsli and Sandrine A. Zweifel
14 Idiopathic Choroidal Neovascularization�� 167
Faisal A. Almarek and Sulaiman M. Alsulaiman
15 Subretinal Neovascularization Associated with Idiopathic
Juxtafoveal Telangiectasia �� 179
Matthew R. Starr and Sophie J. Bakri
16 Peripapillary Choroidal Neovascularization �� 187
Sumit Randhir Singh and Jay Chhablani
17 Choroidal Neovascularization in Pediatric Population�� 203
Şengül Özdek and Hatice Tuba Atalay
18 Polypoidal Choroidal Vasculopathy: Diagnostic
and Therapeutic Considerations�� 217
Adrian H. C. Koh
19 Choroidal Neovascularization Associated with Rare Entities�� 237
Pasha Anvari, Masood Naseripour,
and Khalil Ghasemi Falavarjani
Part III Clinical Trials
20 Clinical Trials Related to Choroidal Neovascularization

Secondary to Age-Related Macular Degeneration�� 259
Paisan Ruamviboonsuk, Peranut Chotcomwonse,
Variya Nganthavee, Warissara Pattanapongpaiboon,
and Kornwipa Hemarat
21 Clinical Trials Related to Myopic Choroidal
Neovascularization �� 283
Christine P. S. Ho and Timothy Y. Y. Lai
22 Clinical Trials Related to Non-AMD Choroidal
Neovascularization �� 297
Tariq Alzahem, Nayef Alswaina, and Marwan A. Abouammoh
23 Reading Centers for Clinical Trials of Choroidal
Neovascularization (CNV): Present Role and Future
Opportunities�� 317
René Rückert, Lala Ceklic, and Marion R. Munk
Contents xv
Part IV Therapy and Rehabilitation
24 Anti-Vascular Endothelial Growth Molecules�� 331

Eduardo Tomazoni and Eduardo Buchelle Rodrigues
25 Surgical Interventions�� 343
Elizabeth D. Marlow and Tamer H. Mahmoud
26 Home Monitoring for Age-Related Macular Degeneration �� 363
Voraporn Chaikitmongkol
27 Rehabilitation �� 375
Yogeshwari Bansal
28 Gene Therapy for Choroidal Neovascularization �� 381
Joo Yong Lee and Joon Hyung Yeo
29 Radiotherapy for Choroidal Neovascularization�� 391
David Pérez González, Matias Iglicki, and Dinah Zur
30 Laser for Prevention of Choroidal Neovascularization�� 401
Jeffrey K. Luttrull and David Kent
31 Choroidal Neovascularization: Newer Molecules �� 425
Aamir A. Aziz, Ibrahim Khanani, Fawwaz A. Siddiqui,
Ryan N. Constantine, and Arshad M. Khanani
About the Editor
Jay Chhablani is a Vitreo-Retina Specialist at

the University of Pittsburgh Eye Center. He
completed a clinical vitreo-retina fellowship at
Sankara Nethralaya, Chennai, India, and was an
International Council of Ophthalmology (ICO)
fellow at Jules Gonin Eye Hospital, Switzerland,
in 2009. He was a Clinical Instructor at the
Jacobs Retina Center at Shiley Eye Center,
University of California, San Diego, USA
(2010–2012), before joining the faculty at L V
Prasad Eye Institute, Hyderabad, India (2012–
2019). His areas of interest are macular disor-
ders and recent imaging techniques. He has
published more than 300 articles in peer-
reviewed journals with a focus on the field of
the choroid. He is the editor of the books
Choroidal Disorders and Central Serous
Chorioretinopathy, and he is on the review
boards of all the high-impact ophthalmology
journals. He is also on the editorial board of
several specialty journals, including the
American Journal of Ophthalmology. He is a
member of the American Academy of
Ophthalmology’s Global ONE network com-
mittee. He has won several national and inter-
national awards and delivered the inaugural Ian
Constable lecture at the Asia-Pacific Vitreo-
Retina Society in 2016. He received the
Inaugural Namperumalsamy Young Researcher
Award in 2018, presented by the Vitreo-Retina
Society of India.
xvii
Choroidal Neovascular Membrane:
Historical Perspectives 1
Aniruddha Agarwal and Krinjeela Bazgain
1.1 Introduction 1.2 Historical Perspectives

of Choroidal
Choroidal neovascular membranes (CNV) repre- Neovascularization
sent the pathological growth of blood vessels that
can result in loss of vision [1]. Age-related macu- In the year 1876, Sattler first noted blood vessels
lar degeneration (AMD) is the leading cause of in between the RPE and the Bruch’s membrane in
central visual loss and legal blindness in patients the fundus periphery [2]. Reichling and KIemans,
over the age of 65 years. As many as 30% of Friedman et al., and Daicker noted similar pat-
adults over the age of 75 develop signs of senile terns of neovascularization in their studies of the
retinal degeneration. The prevalence of AMD is fundus [3–5]. Very few of these vessels have
on the rise worldwide due to an aging population. shown continuity with the choroid, leading to
The exudative or neovascular form of AMD, speculation of their origin [6]. Spitnaz suggested
which is characterized by choroidal neovascular that an additional vascular layer is present in
membrane (CNV) growth and/or serous retinal between the Bruch’s membrane and the RPE,
pigment epithelial (RPE) detachments, accounts anterior to the equator that results in these patho-
for over 90% of the cases with severe visual loss. logical changes [7]. He suggested that this is a
Since the initial descriptions of AMD-related physiologic process rather than an aging pathol-
CNV more than 100 years ago, there has been a ogy [7].
complete paradigm shift in the diagnosis and Oeller [8] in 1905 coined the term “disciform
management of this condition. From a near- degeneration of the macula.” Junius and Khunt
certain blindness in these patients in the past, the [9] in 1928 described disc-shaped lesions in the
current management strategies ensure stabiliza- macula associated with significant vision loss.
tion and improvement of vision over a long term Holloway and Verhoeff [10] reported eight cases
period. In this chapter, historical perspectives of similar appearance in the macula with disci-
that led to the discovery of CNV have been high- form lesions. In one of the eyes they noted few
lighted. In addition, early epidemiological trials blood vessels with a small amount of connective
that established the worldwide significance of tissue extending from choroid through breaks in
this condition have been briefly described. Bruch’s membrane [10]. Verhoeff and Grossman
A. Agarwal (*) · K. Bazgain

Department of Ophthalmology, Advanced Eye
Centre, Postgraduate Institute of Medical Education
and Research (PGIMER), Chandigarh, India
© Springer Nature Singapore Pte Ltd. 2020 1

J. Chhablani (ed.), Choroidal Neovascularization, https://doi.org/10.1007/978-981-15-2213-0_1
2 A. Agarwal and K. Bazgain
[11] said that the disciform lesion was secondary 1.3 Early Epidemiological
to a reparative process and also termed the lesion Studies of Choroidal
“juvenile disciform macular degeneration” as Neovascularization
they found it in a 29-year-old patient. Sorsby and
Mason postulated that a breach in the elastic lam- The large body of literature on AMD-related
ina of Bruch’s membrane results in capillary her- CNV is due to the strong foundations laid by
niation with subsequent subretinal hemorrhage large studies focusing on the epidemiology and
[12, 13]. They also suggested that wandering his- natural history of the disease. Two of the most
tiocytes and fibroblasts from the choroid form significant studies include Beaver Dam Eye
connective tissue in the subepithelial coagulated Study and Blue Mountain Eye Study.
transudate [12, 13].
Gass in 1967 said that loss of normal adhesion
of the RPE to Bruch’s membrane, breach in 1.4 Beaver Dam Eye Study
Bruch’s membrane, and neovascular invasion of
the sub-pigment epithelial space from the cho- The Beaver Dam Eye Study was conducted in
roid predisposes the eye to hemorrhagic disci- 1987 [21]. The study consisted of more than
form detachment [14]. He also said that fundus 5000 patients from the Beaver Dam area of
fluorescein angiography helps in differentiating Wisconsin. Follow-up data of these patients
the disciform stage of the disease from intraocu- were also included. Fundus photography and
lar neoplasms [15]. Sarks noted new vessel pro- standardized macular grading were performed
liferation through Bruch’s membrane in the for all the eyes. This study was significant
vicinity of the macula on histological examina- because it provided the first ever evidence of
tion [16]. She also described basal linear deposits high prevalence of AMD and CNV in the
between the plasma infoldings and the basement elderly Caucasian population. In addition, the
membrane of RPE [17]. These deposits were sug- study suggested genetic linkage to the develop-
gested to be secondary to RPE failure. They con- ment of CNV as well as a potential link between
sist of banded fibers embedded in granular cigarette smoking and advanced forms of
material [17]. AMD. Other risk factors identified included
Various experimental models have employed sunlight exposure and cardiovascular risk fac-
various methods in producing choroidal neovas- tors [21].
cularization. However, Heriot et al. [18] in 1984
proposed that phototoxicity damages the RPE
cells and promotes choriocapillaris budding. The 1.5 Blue Mountain Eye Study
adjacent healthy RPE slides forming a bridge
over the vessels [18]. Hence, an antecedent break The Blue Mountain Eye Study was performed in
in the RPE is not needed for the formation of Australia in 1992 [22]. The study evaluated base-
CNV [18]. The budding capillary forms a lytic line and follow-up clinical data of more than
hole secondary to RPE damage [18]. In the same 3500 individuals aged 49 years and above. The
year, Penfold et al. suggested that lymphocytes, investigators used fundus photographs and stan-
monocytes, mast cells, and fibroblasts may have dardized macular grading protocols similar to the
a role to play in the formation of a hole in Bruch’s Beaver Dam Study and observed a strong corre-
membrane and subsequent choroidal neovascu- lation between age and AMD. Patients with pig-
larization [19, 20]. In the past three decades, the mentary fundus changes and drusen were linked
knowledge of the patho-anatomy of CNV and its to the progression of AMD. Other risk factors
natural history is still evolving. However, the sci- identified by the Blue Mountain Eye Study were
entific contributions by various researchers in the smoking, plasma fibrinogen levels, and family
past century is truly extraordinary. history [22].
1 Choroidal Neovascular Membrane: Historical Perspectives 3
1.6 Summary 9. Junius P, Kuhnt H. Die scheibenfo¨rmige Entargung

der Netzhautmitte (Degeneratio maculae luteae disci-
formis). Berlin: Karger; 1926. p. 132.
Since the initial descriptions of CNV, there has 10. Holloway TB, Verhoeff FH. Disc·like degeneration
been a significant advancement of knowledge in of the macula with microscopic report concerning
the field of choroidal imaging, and diagnosis and a tumor·like mass in the macular region. Trans Am
Ophthalmol Soc. 1928;26:206–28.
management of CNV. CNV has been associated 11. Verhoeff FH. Histological findings in a case
with a number of conditions, AMD and myopia of angioid streaks. Br J Ophthalmol. 1948;32:
being the leading causes, ocular inflammation 531–44.
being the next most frequently implicated etiol- 12. Ashton N, Sorsby A. Fundus dystrophy with unusual
features: a histological study. Br J Ophthalmol.
ogy of development of CNV. Irrespective of the 1951;35:751–64.
cause, the novel treatments available for this con- 13. Sorsby A, Mason MEJ. A fundus dystrophy with
dition have greatly impacted the visual and ana- unusual features (late onset and dominant inheritance
tomical outcomes of patients, and contributed in of a central retinal lesion showing oedema, haemor-
rhage and exudates developing into generalised cho-
reducing the blinding complications of CNV. roidal atrophy with massive pigment proliferation). Br
J Ophthalmol. 1949;33:67–97.
14. Gass JDM. Pathogenesis of disciform detachment of
References the neuroepithelium. III. Senile disciform macular
degeneration. Am J Ophthalmol. 1967;63:617–44.
1. Agarwal A, Invernizzi A, Singh RB, et al. An update 15. Gass JDM. Pathogenesis of disciform detachment of
on inflammatory choroidal neovascularization: epi- the neuroepithelium. IV. Fluorescein angiographic
demiology, multimodal imaging, and management. J study of senile disciform mac·ular degeneration. Am J
Ophthalmic Inflamm Infect. 2018;8:13. Ophthalmol. 1967;63:645–59.
2. Sattler H. Ueber den feineren Bau der Choroidea des 16. Sarks SH. New vessel formation beneath the retinal
Menschen nebst Beitra¨gen zur pathologischen und pigment epithelium in senile eyes. Br J Ophthalmol.
vergleichenden Anatomie der Aderhaut. Albrecht von 1973;57:951–65.
Graefes Arch Ophthalmol. 1976;22(Pt 2):1–100. 17. Sarks SH. Ageing and degeneration in the macular
3. Reichling W, Klemens F. Uber eine gefassfilhrende region: a clinicopathological study. Br J Ophthalmol.
Bindegew·ebsschicht zwischen dem Pigmentepithel 1976;60:324–41.
der Retina und der Lamina vitrea. Albrecht von 18. Heriot WJ, Henkind P, Bellhorn RW, Burns
Graefes Arch Ophthalmol. 1940;141:500–12. MS. Choroidal neovascularization can digest
4. Friedman E, Smith TR, Kuwabara T. Senile choroi- Bruch’s membrane: a prior break is not essential.
dal vascular patterns and drusen. Arch Ophthalmol. Ophthalmology. 1984;91:1603–8.
1963;69:220–30. 19. Penfold P, Killingsworth M, Sarks S. An ultrastruc-
5. Daicker B. Lineare Degenerationen des tural study of the role of leucocytes and fibroblasts
peripheren retinalen Pig· mentepithels; eine in the breakdown of Bruch’s membrane. Aust J
pathologisch·anatomische Studie. Albrecht von Ophthalmol. 1984;12:23–31.
Graefes Arch Klin Exp Ophthalmol. 1973;186:1–12. 20. Penfold PL, Killingsworth MC, Sarks SH. Senile
6. Bec P, Secheyron P, Arne JL, et al. La neovascularisa- macular degeneration: the involvement of immuno-
tion sous·retinienne peripherique et ses consequences competent cells. Graefes Arch Clin Exp Ophthalmol.
pathologiques. J Fr Ophtalmol. 1979;2:329–36. 1985;223:69–76.
7. Spitznas M, Bomfeld N. Development and ultrastruc- 21. Klein R, Klein BE, Linton KL. Prevalence of age-
ture of peripheral subretinal neovascularizations. related maculopathy. The Beaver Dam Eye Study.
Albrecht von Graefes Arch Klin Exp Ophthalmol. Ophthalmology. 1992;99(6):933–43.
1978;208:125–33. 22. Mitchell P, Smith W, Attebo K, Wang JJ. Prevalence
8. Oeller JN. Atlas Seltener Ophthalmoskopischer of age-related maculopathy in Australia. The
Befunde. Zugleich Evga¨nzungstateln zu dem Atlas Blue Mountains Eye Study. Ophthalmology.
der Ophthalmoskopie. Wiesbaden: JF Bergeman; 1995;102(10):1450–60.
1900–1905.
4 A. Agarwal and K. Bazgain
Krinjeela Bazgain is currently pursuing her M.Ch. in

Aniruddha Agarwal is currently working as an
Vitreoretinal surgery in the Postgraduate Institute of
Assistant Professor in Vitreoretina and Uveitis in the
Medical Education and Research (PGIMER),
Department of Ophthalmology, Postgraduate Institute of
Chandigarh, India. She did her Ophthalmology resi-
Medical Education and Research (PGIMER), Chandigarh,
dency from PGIMER, Chandigarh, India.
India. He has completed his Clinical Research Fellowship
(subspecialty of vitreoretina and uveitis) in the Ocular
Imaging Research and Reading Center, Stanley
M. Truhlsen Eye Institute, Omaha, Nebraska, USA
between 2014 and 2016. He did his ophthalmology resi-
dency and Surgical Vitreoretina and Uveitis Fellowship at
the PGIMER, Chandigarh, India. He is the recipient of
prestigious awards such as the Bayer Global
Ophthalmology Association Project (GOAP) Fellowship,
Carl Camras Best Researcher Award, J. M. Pahwa Award
by Vitreoretina Society of India (VRSI), Narsing Rao
Award by Uveitis Society of India (USI), and the Carl
Herbort Award by the USI. In 2015, he was felicitated by
the Hon. Prime Minister of India for his excellent contri-
bution. He has authored more than 150 publications and
36 book chapters. His areas of interest include uveitis, as
well as medical and surgical diseases of the retina. He is
an expert in ocular imaging and has numerous interna-
tional presentations and collaborations for the same.
Part I
Basic Science
Pathogenesis of Choroidal
Neovascularization 2
Mayss Al-Sheikh and Daniel Barthelmes
Choroidal neovascularization (CNV) is a major blood flow decreases in patients with AMD; how-
cause of severe vision loss in patients with age- ever, the exact nature of impairment remains
related macular degeneration (AMD) [1]. unclear. Interestingly, eyes with neovascular
Neovascular AMD is characterized by the devel- AMD showed CNV development in areas of
opment of a neovascular membrane, emerging hypofluorescence in the macula and areas of
from the choroid, which may remain underneath watershed zone [5]. For decades, different modal-
the retinal pigment epithelium (RPE) or extend ities were used to investigate blood flow in
through the Bruch’s membrane and the RPE to patients with neovascular AMD [6–8]. Using
the subretinal space [2]. Early signs of CNV are fundus fluorescence angiography, a prolonged
hemorrhage, macular edema, lipid deposition, or choroidal filling phase was identified in many
detachment of the retinal pigment epithelium. patients [6]. Using indocyaningreen dye angiog-
End stages are characterized by a scar formation raphy, an attenuation of choriocapillaris blood
[3]. For the localization of CNV in the central flow was revealed [7]. Recently, using optical
macula, the thinning of the Bruch’s membrane coherence tomography angiography, an impair-
with its elastic and collagenous laminae in the ment of the perfusion of the choriocapillaris
foveal region has been proposed to play a role around the CNV lesion was described [9]. Those
[4]. The lamina elastica of Bruch’s membrane is observations led to the hypothesis that abnormal-
described to be 3–6 times thinner and 2–5 times ities of the choriocapillaris may reduce diffusion
more porous in the macular region than it is in the of debris material derived from the RPE into the
peripheral region at all ages, especially in elderly. intravascular space leading to accumulating into
This large discontinuities within the macular the Bruch’s membrane and consecutively to its
lamina elastica may explain the predilection thickening.
toward CNV formation in the macular region Friedman reported that the development of
since for a neovessel to grow from the choroid CNV is associated with hemodynamic changes in
into the sub-RPE space, its cells must be able to the ocular vessels [10, 11]. Accumulation of lip-
pass physically through Bruch’s membrane. ids in the sclera and the Bruch’s membrane,
The exact pathophysiology of CNV is not yet which leads to a stiffening of the tissue, increases
understood. There is an evidence that choroidal the resistance of choroidal blood flow. This, in
turn, results in decreased choroidal perfusion and
impaired RPE transport function, which leads to
M. Al-Sheikh (*) · D. Barthelmes the formation of drusen, RPE atrophy, and lipid
University Hospital Zurich, University of Zurich, infiltration of the Bruch’s membrane, while
Zurich, Switzerland

8 M. Al-Sheikh and D. Barthelmes
increased intravascular pressure is suggested to turn decreases diffusion of oxygen from the cho-
lead to RPE detachment and CNV formation [5, roid to the RPE and outer retina in aging patients
8, 12, 13]. as well as reduce the clearance of debris from
Regardless of what is the initial cause of CNV, RPE and Bruch’s membrane. Those two factors,
angiogenic factors are critically involved in the hypoxia and Bruch’s membrane degradation,
development of CNV. Angiogenesis is defined as induce VEGF secretion and promote CNV devel-
the development of new capillaries from preex- opment. On the other hand, this hypothesis is less
isting networks. Angiogenesis is a critical pro- likely to occur in other types of CNV that are not
cess in the embryologic phase, somatic growth as age-related, such as those associated with young
well as in tissue and wound repair. An important myopic patients and patients with ocular
factor in the angiogenesis process is the balance histoplasmosis.
between pro-angiogenic and antiangiogenic fac- While VEGF is crucial, especially in the early
tors; however, if there is an excessive stimulus stages of development of blood vessels, angio-
and/or reduced inhibitory effects, then patholo- poietins are involved in stabilization and matura-
gies such as CNV may result. Another important tion of vessels in later stages. Angiopoietin-1,
factor in angiogenesis is the extracellular matrix produced by pericytes, induces endothelial cells
(ECM) molecules that are involved in several to recruit periendothelial support cells and asso-
ways in the regulation of the growth of new blood ciate with the extracellular matrix and mesen-
vessels. chyme, promoting vascular integrity and
Vascular endothelial growth factor (VEGF) maintenance of adult vasculature and affects vas-
was initially discovered as a peptide secreted by cular tight junctions [28–30]. In contrast to
tumor cells that leads to increased vascular per- VEGF, which its overexpression leads to the
meability [14]. Later it was explained to be an development of leaky vessels due to a distur-
angiogenic growth factor with high specificity for bance of the blood–retina barrier, overexpression
vascular endothelial cells [15, 16]. The synthesis of angiopoietin-1 leads to an increased density
of VEGF is driven by hypoxia [17] where the and caliber of non-leaky vessels, and modulate
concentration is increased in the retina and vitre- VEGF-induced growing and existing vessels
ous [18, 19]. [31–33]. In other words, angiopoietin-1 promotes
In contrast to CNV, we better understand neo- the maturation of the immature vessels induced
vascularization development in the retina and the by VEGF. Angiopoietin-2 is an antagonist to
iris [20–23]. Many studies have reported a central angiopoietin-1 that is found where vascular
role of hypoxia or ischemia in the development of remodeling takes place.
retinal neovascularization in ischemic diseases However, VEGF and angiopoietins are not the
like diabetic retinopathy, retinal vein occlusion, only factors that potentially play a role in CNV
and neovascular glaucoma [24, 25]. In contrast, formation, basic fibroblast growth factor
the role of hypoxia and ischemia in CNV is not (bFGF2) is detectable in the RPE cells in surgi-
completely clear [8]. The VEGF secretion of cally excised CNV membranes [34, 35]. It is also
RPE is described to be polarized with a higher overexpressed in RPE cells, choroidal vascular
secretion toward the Bruch’s membrane and a endothelial cells, and fibroblasts in laser-induced
lower secretion rate toward the photoreceptors CNV [36]. It has been postulated that bFGF2 has
[26]. Under abnormal conditions, specifically an angiogenic action only in the setting of cellu-
hypoxia, this disparity may exaggerate. Since lar injury [37].
VEGF has a trophic function for the choriocapil- Another important factor in angiogenesis is
laris by its ability to induce endothelial fenestra- the pigment epithelium-derived factor (PEDF).
tion [27], it is conceivable that thickening of the It is a neurotrophic growth factor for photore-
Bruch’s membrane with lipophilic material depo- ceptors that has antiangiogenic action. Its reac-
sition may prevent VEGF from reaching the cho- tion to oxygen is reciprocal to that of VEGF. In
riocapillaris causing its atrophy—which in its many animal models, it was shown that PEDF
2 Pathogenesis of Choroidal Neovascularization 9
inhibits ischemia-induced retinopathy, VEGF- brane, and by releasing cytokines, inflammatory

induced leakage as well as laser-induced CNV cells might foster CNV growth.
formation [38, 39]. Considering the antiangio- Another important point is the leucocytes-
genetic activity, it seems that endogenous mediated angiogenesis and its role in the interac-
PEDF does not prevent the development of tion of cellular adhesion molecules and
CNV. One reason might be the concentration of VEGF. VEGF induces the expression of intracel-
this factor. It has been shown that PEDF pro- lular adhesion molecule-1 (ICAM-1) on tumor
duction decreases with age and that the vitre- cells as well as vascular endothelial cells and
ous concentration is decreased in patients with regulates leukocyte adhesion to endothelial cells
AMD [40] which is then overwhelmed by [53]. ICAM-1 blockade decreases VEGF-induced
angiogenic factors. leukostasis in the retina [54]. These systems are
Along with the pro-angiogenic and antiangio- intertwined: leukocytes, which possess receptors
genic factors, extracellular matrix (ECM) mole- for and migrate in response to VEGF can also
cules also participate in several ways in the produce and release VEGF [55].
regulation of angiogenesis. Degradation of ECM
releases and/or activates pro-angiogenic factors.
Pro-angiogenic factors stimulate proteolytic 2.1 Choroidal
activity, migration, proliferation, and tube forma- Neovascularization
tion in endothelial cells [41, 42]. in Other Diseases
Extracellular matrix molecules are able to
directly inhibit or stimulate endothelial cell pro- The development of choroidal neovascularization
cesses involved in angiogenesis by binding to is found in a heterogeneous group of diseases. In
integrins, which on its turn can upregulate and addition to choroidal neovascularization second-
downregulate various intracellular signaling ary to age-related macular degeneration, neovas-
pathways. On the other hand, pro-angiogenic fac- cular membranes are also found in patients with
tors may act in part by altering the expression of high myopia, pseudoxanthoma elasticum, after
integrins on endothelial cells [43–47]. This pro- trauma, or after inflammatory diseases that affect
cess is potentiated by the secretion of proteolytic the choroid.
enzymes. Two proteolytic systems have been
implicated in the breakdown of ECM during
angiogenesis, one involving urokinase type of 2.2 hat Do These Patients Have
W
plasminogen activator and one involving matrix in Common?
metalloproteinases (MMPs) which are present in
excised CNV specimens and increased in laser- Various diseases that are associated with an
induced CNV [48, 49]. increased risk for the development of CNV have
Inflammation has also been proposed to play a a disorder of the Bruch’s membrane in common.
role in the formation of CNV. Several histopatho- In patients with high myopia or patients with
logical studies have identified inflammatory reac- Pseudoxanthoma elasticum who are known to
tions in autopsy eyes with CNV [50, 51]. It has have thinning of the Bruch’s membrane, there is
been shown that the Bruch’s membrane has thin an increased risk of developing CNV [56].
areas around the breaks where CNV goes through Likewise, mechanical (i.e., trauma) or thermal
to the subretinal space and that the choroid under- (i.e., laser photocoagulation) damage to the
neath those thinned areas is infiltrated with Bruch’s membrane leads to an increased risk for
inflammation cells such as lymphocytes, macro- CNV. Another category of increased risk for
phages, fibroblasts, and myofibroblasts [52]. CNV development is inflammatory diseases of
Activated macrophages and other inflammatory the choroid such as multifocal choroiditis or ocu-
cells secrete proteolytic enzymes such as collage- lar histoplasmosis. In Sorsby fundus dystrophy,
nase and elastase that can degrade Bruch’s mem- an autosomal dominant disease, deposits on the
Bruch’s membrane lead to CNV formation angiogenic growth factor does not seem suffi-
[57, 58]. Sorsby is caused by mutations in the tis- cient to induce CNV formation. The intact
sue inhibitor of metalloproteinases 3 (TIMP-3) Bruch’s membrane apparently forms a mechani-
gene. The gene product is an inhibitor of metal- cal or biochemical barrier to VEGF from the ret-
loproteinases, which is involved in the regulation ina into the choroid. Another transgenic mouse
of ECM turnover [59]. Malattia Leventinese and line with overexpression of VEGF in the RPE
Doyne honeycomb retinal dystrophy are two showed the development of intrachoroidal CNV
other autosomal dominant disorders with drusen that did not penetrate through the intact Bruch’s
formation and CNV, both caused by a mutation in membrane and RPE [65]. Viral models with a
the EFEMP1 gene that encodes a protein of recombinant adenovirus vector encoding VEGF
extracellular matrix (“EGF-containing fibrillin- that was injected into the subretinal space showed
like extracellular matrix protein 1”). the development of CNV that breached the
Therefore, these diseases and CNV secondary Bruch’s membrane and reached the subretinal
to AMD that exhibits abnormalities of the Bruch’s space [66]. However, the localization into the
membrane suggest that alteration of ECM of the subretinal space may be due to iatrogenic breaks
RPE predisposes the development of CNV. in Bruch’s membrane during the subretinal injec-
tion of the virus. Taking the abovementioned data
together, it seems that the development of CNV
2.3 What Do We Know requires different factors including the imbalance
from Animal Models? of the angiogenesis process as well as a defect in
the Bruch’s membrane.
A central limitation in studies of CNV is the lack
of adequate animal models that accurately reflect Key Learning Points
changes in AMD. In an established model, a rup- • Choroidal neovascularization is one of the
ture of the Bruch’s membrane induced by laser leading causes of vision impairment in devel-
burns results in CNV formation [60]. Although oped countries.
laser-induced CNV in animal models may not • The pathogenesis of choroidal neovascular-
represent CNV secondary to AMD in humans, it ization is not completely clear.
gives us insights into important features of the • Using different image modalities impairment
human condition. Laser photocoagulation that of choroidal blood flow has been shown to play
disrupts the Bruch’s membrane, especially in the a role in the development of choroidal
macula, can induce CNV in humans [60, 61]. neovascularization.
Other similar features are migration of choroidal • Hemodynamic changes in the ocular vessels
vascular endothelial cells and newly formed ves- induced by accumulation of lipids leads to
sels into the subretinal space through the dis- increased intravascular pressure and RPE
rupted Bruch’s membrane, accumulation of detachment and CNV formation.
subretinal fluid, presence of leucocytes adjacent • Imbalance of the angiogenic process including
to the neovascular membrane as well as fibrovas- different factors such as vascular endothelial
cular scar formation [51, 62, 63]. growth factor, angiopoietins, basal fibroblast
Another animal model is the transgenic model. growth factor and pigment epithelium-derived
A transgenic mouse line with overexpression of factor as well as degradation of extracellular
VEGF in the photoreceptors showed new vessels matrix are crucial for the development of CNV.
originating from the deep retina capillary extend- • Inflammation including infiltration of the cho-
ing through the photoreceptors to the subretinal roid with inflammatory cells as well as
space. This model did not show choroidal neo- leucocytes-mediated angiogenesis are part of
vascularization [64]. A sole overexpression of an CNV formation.
References sion characteristics in fellow eyes of patients with uni-

lateral neovascular age-related macular degeneration.
Br J Ophthalmol. 1995;79(6):558–61.
1. Bourne RR, Stevens GA, White RA, Smith JL,
14. Senger DR, Galli SJ, Dvorak AM, Perruzzi CA,
Flaxman SR, Price H, et al. Causes of vision loss
Harvey VS, Dvorak HF. Tumor cells secrete a vascu-
worldwide, 1990–2010: a systematic analysis. Lancet
lar permeability factor that promotes accumulation of
Glob Health. 2013;1(6):e339–49.
ascites fluid. Science. 1983;219(4587):983–5.
2. Chang TS, Freund KB, de la Cruz Z, Yannuzzi
15. Keck PJ, Hauser SD, Krivi G, Sanzo K, Warren
LA, Green WR. Clinicopathologic correlation of
T, Feder J, et al. Vascular permeability factor, an
choroidal neovascularization demonstrated by
endothelial cell mitogen related to PDGF. Science.
indocyanine green angiography in a patient with
1989;246(4935):1309–12.
retention of good vision for almost four years. Retina.
16. Leung DW, Cachianes G, Kuang WJ, Goeddel
1994;14(2):114–24.
DV, Ferrara N. Vascular endothelial growth fac-
3. Gass JD, Agarwal A, Lavina AM, Tawansy KA. Focal
tor is a secreted angiogenic mitogen. Science.
inner retinal hemorrhages in patients with dru-
1989;246(4935):1306–9.
sen: an early sign of occult choroidal neovascu-
17. Shweiki D, Itin A, Soffer D, Keshet E. Vascular
larization and chorioretinal anastomosis. Retina.
endothelial growth factor induced by hypoxia may
2003;23(6):741–51.
mediate hypoxia-initiated angiogenesis. Nature.
4. Chong NH, Keonin J, Luthert PJ, Frennesson CI,
1992;359(6398):843–5.
Weingeist DM, Wolf RL, et al. Decreased thickness
18. Adamis AP, Miller JW, Bernal MT, D’Amico DJ,
and integrity of the macular elastic layer of Bruch’s
Folkman J, Yeo TK, et al. Increased vascular endothe-
membrane correspond to the distribution of lesions
lial growth factor levels in the vitreous of eyes with
associated with age-related macular degeneration. Am
proliferative diabetic retinopathy. Am J Ophthalmol.
J Pathol. 2005;166(1):241–51.
1994;118(4):445–50.
5. Ross RD, Barofsky JM, Cohen G, Baber WB, Palao
19. Pe’er J, Shweiki D, Itin A, Hemo I, Gnessin H, Keshet
SW, Gitter KA. Presumed macular choroidal water-
E. Hypoxia-induced expression of vascular endothe-
shed vascular filling, choroidal neovascularization,
lial growth factor by retinal cells is a common fac-
and systemic vascular disease in patients with age-
tor in neovascularizing ocular diseases. Lab Investig.
related macular degeneration. Am J Ophthalmol.
1995;72(6):638–45.
1998;125(1):71–80.
20. Adamis AP, Shima DT, Tolentino MJ, Gragoudas ES,
6. Pauleikhoff D, Spital G, Radermacher M, Brumm
Ferrara N, Folkman J, et al. Inhibition of vascular
GA, Lommatzsch A, Bird AC. A fluorescein and indo-
endothelial growth factor prevents retinal ischemia-
cyanine green angiographic study of choriocapillaris
associated iris neovascularization in a nonhuman pri-
in age-related macular disease. Arch Ophthalmol.
mate. Arch Ophthalmol. 1996;114(1):66–71.
1999;117(10):1353–8.
21. Amano S, Rohan R, Kuroki M, Tolentino M, Adamis
7. Grunwald JE, Hariprasad SM, DuPont J. Effect
AP. Requirement for vascular endothelial growth
of aging on foveolar choroidal circulation. Arch
factor in wound- and inflammation-related cor-
Ophthalmol. 1998;116(2):150–4.
neal neovascularization. Invest Ophthalmol Vis Sci.
8. Grunwald JE, Hariprasad SM, DuPont J, Maguire
1998;39(1):18–22.
MG, Fine SL, Brucker AJ, et al. Foveolar choroi-
22. Ferrara N, Henzel WJ. Pituitary follicular cells secrete
dal blood flow in age-related macular degeneration.
a novel heparin-binding growth factor specific for
Invest Ophthalmol Vis Sci. 1998;39(2):385–90.
vascular endothelial cells. Biochem Biophys Res
9. Moult EM, Alibhai AY, Rebhun C, Lee B, Ploner S,
Commun. 1989;161(2):851–8.
Schottenhamml J, et al. Spatial distribution of cho-
23. Miller JW, Adamis AP, Shima DT, D’Amore PA,
riocapillaris impairment in eyes with choroidal neo-
Moulton RS, O’Reilly MS, et al. Vascular endothelial
vascularization secondary to age-related macular
growth factor/vascular permeability factor is tempo-
degeneration: a quantitative OCT angiography study.
rally and spatially correlated with ocular angiogenesis
Retina. 2020;40(3):428–45.
in a primate model. Am J Pathol. 1994;145(3):574–84.
10. Friedman E. A hemodynamic model of the patho-
24. Ashton N. Retinal vascularization in health and dis-
genesis of age-related macular degeneration. Am J
ease: proctor award lecture of the Association for
Ophthalmol. 1997;124(5):677–82.
Research in Ophthalmology. Am J Ophthalmol.
11. Friedman E. The role of the atherosclerotic process in
1957;44(4 Pt 2):7–17.
the pathogenesis of age-related macular degeneration.
25. Shimizu K, Kobayashi Y, Muraoka K. Midperipheral
Am J Ophthalmol. 2000;130(5):658–63.
fundus involvement in diabetic retinopathy.
12. Mori F, Konno S, Hikichi T, Yamaguchi Y, Ishiko
Ophthalmology. 1981;88(7):601–12.
S, Yoshida A. Pulsatile ocular blood flow study:
26. Blaauwgeers HG, Holtkamp GM, Rutten H, Witmer
decreases in exudative age related macular degenera-
AN, Koolwijk P, Partanen TA, et al. Polarized
tion. Br J Ophthalmol. 2001;85(5):531–3.
vascular endothelial growth factor secretion by
13. Remulla JF, Gaudio AR, Miller S, Sandberg
human retinal pigment epithelium and localiza-
MA. Foveal electroretinograms and choroidal perfu-
tion of vascular endothelial growth factor receptors
on the inner choriocapillaris. Evidence for a tro- age-related macular degeneration. Am J Ophthalmol.
phic paracrine relation. Am J Pathol. 1999;155(2): 2002;134(2):220–7.
421–8. 41. Gross JL, Moscatelli D, Rifkin DB. Increased capil-
27. Roberts WG, Palade GE. Increased microvascular lary endothelial cell protease activity in response to
permeability and endothelial fenestration induced angiogenic stimuli in vitro. Proc Natl Acad Sci USA.
by vascular endothelial growth factor. J Cell Sci. 1983;80(9):2623–7.
1995;108(Pt 6):2369–79. 42. Sato Y, Rifkin DB. Autocrine activities of basic
28. Suri C, Jones PF, Patan S, Bartunkova S, Maisonpierre fibroblast growth factor: regulation of endothelial
PC, Davis S, et al. Requisite role of angiopoietin-1, a cell movement, plasminogen activator synthesis, and
ligand for the TIE2 receptor, during embryonic angio- DNA synthesis. J Cell Biol. 1988;107(3):1199–205.
genesis. Cell. 1996;87(7):1171–80. 43. Asahara T, Chen D, Takahashi T, Fujikawa K,
29. Carmeliet P. Mechanisms of angiogenesis and arterio- Kearney M, Magner M, et al. Tie2 receptor ligands,
genesis. Nat Med. 2000;6(4):389–95. angiopoietin-1 and angiopoietin-2, modulate VEGF-
30. Potente M, Gerhardt H, Carmeliet P. Basic induced postnatal neovascularization. Circ Res.
and therapeutic aspects of angiogenesis. Cell. 1998;83(3):233–40.
2011;146(6):873–87. 44. Campochiaro PA, Soloway P, Ryan SJ, Miller
31. Suri C, McClain J, Thurston G, McDonald DM, JW. The pathogenesis of choroidal neovascularization
Zhou H, Oldmixon EH, et al. Increased vasculariza- in patients with age-related macular degeneration.
tion in mice overexpressing angiopoietin-1. Science. Mol Vis. 1999;5:34.
1998;282(5388):468–71. 45. Murata T, He S, Hangai M, Ishibashi T, Xi XP, Kim
32. Thurston G, Suri C, Smith K, McClain J, Sato TN, S, et al. Peroxisome proliferator-activated receptor-
Yancopoulos GD, et al. Leakage-resistant blood ves- gamma ligands inhibit choroidal neovascularization.
sels in mice transgenically overexpressing angiopoi- Invest Ophthalmol Vis Sci. 2000;41(8):2309–17.
etin-1. Science. 1999;286(5449):2511–4. 46. Seo MS, Kwak N, Ozaki H, Yamada H, Okamoto N,
33. Witzenbichler B, Maisonpierre PC, Jones P, Yamada E, et al. Dramatic inhibition of retinal and
Yancopoulos GD, Isner JM. Chemotactic properties choroidal neovascularization by oral administration of
of angiopoietin-1 and -2, ligands for the endothelial- a kinase inhibitor. Am J Pathol. 1999;154(6):1743–53.
specific receptor tyrosine kinase Tie2. J Biol Chem. 47. Miller JW, Stinson WG, Folkman J. Regression of
1998;273(29):18514–21. experimental iris neovascularization with systemic
34. Amin R, Puklin JE, Frank RN. Growth factor local- alpha-interferon. Ophthalmology. 1993;100(1):9–14.
ization in choroidal neovascular membranes of age- 48. Cornelius LA, Nehring LC, Roby JD, Parks WC,
related macular degeneration. Invest Ophthalmol Vis Welgus HG. Human dermal microvascular endo-
Sci. 1994;35(8):3178–88. thelial cells produce matrix metalloproteinases in
35. Frank RN, Amin RH, Eliott D, Puklin JE, Abrams response to angiogenic factors and migration. J Invest
GW. Basic fibroblast growth factor and vascular Dermatol. 1995;105(2):170–6.
endothelial growth factor are present in epiretinal and 49. Pepper MS, Vassalli JD, Montesano R, Orci
choroidal neovascular membranes. Am J Ophthalmol. L. Urokinase-type plasminogen activator is induced
1996;122(3):393–403. in migrating capillary endothelial cells. J Cell Biol.
36. Ogata N, Matsushima M, Takada Y, Tobe T, Takahashi 1987;105(6 Pt 1):2535–41.
K, Yi X, et al. Expression of basic fibroblast growth 50. Green WR, Key SN 3rd. Senile macular degeneration:
factor mRNA in developing choroidal neovasculariza- a histopathologic study. Trans Am Ophthalmol Soc.
tion. Curr Eye Res. 1996;15(10):1008–18. 1977;75:180–254.
37. Yamada H, Yamada E, Kwak N, Ando A, Suzuki 51. Penfold PL, Provis JM, Billson FA. Age-related mac-
A, Esumi N, et al. Cell injury unmasks a latent ular degeneration: ultrastructural studies of the rela-
proangiogenic phenotype in mice with increased tionship of leucocytes to angiogenesis. Graefes Arch
expression of FGF2 in the retina. J Cell Physiol. Clin Exp Ophthalmol. 1987;225(1):70–6.
2000;185(1):135–42. 52. van der Schaft TL, Mooy CM, de Bruijn WC, de Jong
38. Mori K, Gehlbach P, Yamamoto S, Duh E, Zack PT. Early stages of age-related macular degeneration:
DJ, Li Q, et al. AAV-mediated gene transfer of pig- an immunofluorescence and electron microscopy
ment epithelium-derived factor inhibits choroidal study. Br J Ophthalmol. 1993;77(10):657–61.
neovascularization. Invest Ophthalmol Vis Sci. 53. Becker MD, Kruse FE, Azzam L, Nobiling R,
2002;43(6):1994–2000. Reichling J, Volcker HE. In vivo significance of
39. Stellmach V, Crawford SE, Zhou W, Bouck ICAM-1—dependent leukocyte adhesion in early
N. Prevention of ischemia-induced retinopathy by corneal angiogenesis. Invest Ophthalmol Vis Sci.
the natural ocular antiangiogenic agent pigment 1999;40(3):612–8.
epithelium-derived factor. Proc Natl Acad Sci USA. 54. Miyamoto K, Khosrof S, Bursell SE, Moromizato Y,
2001;98(5):2593–7. Aiello LP, Ogura Y, et al. Vascular endothelial growth
40. Holekamp NM, Bouck N, Volpert O. Pigment factor (VEGF)-induced retinal vascular permeabil-
epithelium-derived factor is deficient in the vitreous ity is mediated by intercellular adhesion molecule-1
of patients with choroidal neovascularization due to (ICAM-1). Am J Pathol. 2000;156(5):1733–9.
55. Harmey JH, Dimitriadis E, Kay E, Redmond HP, 2-dimensional reconstruction and confocal scanning
Bouchier-Hayes D. Regulation of macrophage pro- laser microscopic imaging of choroidal neovascular-
duction of vascular endothelial growth factor (VEGF) ization in eyes with age-related maculopathy. Arch
by hypoxia and transforming growth factor beta-1. Ophthalmol. 2000;118(5):625–9.
Ann Surg Oncol. 1998;5(3):271–8. 63. Nishimura T, Goodnight R, Prendergast RA, Ryan
56. Green WR, Wilson DJ. Choroidal neovascularization. SJ. Activated macrophages in experimental sub-
Ophthalmology. 1986;93(9):1169–76. retinal neovascularization. Ophthalmologica.
57. Hoskin A, Sehmi K, Bird AC. Sorsby’s pseudoin- 1990;200(1):39–44.
flammatory macular dystrophy. Br J Ophthalmol. 64. Okamoto N, Tobe T, Hackett SF, Ozaki H, Vinores
1981;65(12):859–65. MA, LaRochelle W, et al. Transgenic mice with
58. Sorsby A, Mason ME. A fundus dystrophy with increased expression of vascular endothelial growth
unusual features. Br J Ophthalmol. 1949;33(2):67–97. factor in the retina: a new model of intraretinal
59. Weber BH, Vogt G, Pruett RC, Stohr H, Felbor and subretinal neovascularization. Am J Pathol.
U. Mutations in the tissue inhibitor of metallopro- 1997;151(1):281–91.
teinases-3 (TIMP3) in patients with Sorsby’s fundus 65. Schwesinger C, Yee C, Rohan RM, Joussen AM,
dystrophy. Nat Genet. 1994;8(4):352–6. Fernandez A, Meyer TN, et al. Intrachoroidal neo-
60. Ryan SJ. Subretinal neovascularization. Natural his- vascularization in transgenic mice overexpress-
tory of an experimental model. Arch Ophthalmol. ing vascular endothelial growth factor in the retinal
1982;100(11):1804–9. pigment epithelium. Am J Pathol. 2001;158(3):
61. Francois J, De Laey JJ, Cambie E, Hanssens M, 1161–72.
Victoria-Troncoso V. Neovascularization after argon 66. Baffi J, Byrnes G, Chan CC, Csaky KG. Choroidal
laser photocoagulation of macular lesions. Am J neovascularization in the rat induced by adenovirus
Ophthalmol. 1975;79(2):206–10. mediated expression of vascular endothelial growth
62. Grossniklaus HE, Cingle KA, Yoon YD, Ketkar N, factor. Invest Ophthalmol Vis Sci. 2000;41(11):
L’Hernault N, Brown S. Correlation of histologic 3582–9.
Dr. Mayss Al-Sheikh is a medical retina consultant at Daniel Barthelmes Director, Department of
the University Hospital of Zurich, Switzerland. She Ophthalmology, University Hospital Zurich and Chair of
achieved her medical degree at the University of Ophthalmology, University of Zurich
Dusseldorf in Germany. Head, Clinical Neurosciences/Head and Neck Disease,
Dr. Mayss Al-Sheikh completed her residency at the University Hospital Zurich
University Hospital of Zurich, Switzerland, and an inter- Academic qualifications: MD, PhD, executive MBA,
national fellowship in medical retina at the Doheny Eye FEBO, FMH Ophthalmology and Ophthalmic Surgery.
Institute and Stein Eye Institute, University of California, After his specialty training in Switzerland, DB moved
Los Angeles, USA. to Sydney (Australia). Besides his PhD in basic research
Dr. Mayss Al-Sheikh’s main professional interest is on vascular stem cells at The University of Sydney, DB
multimodal and advanced imaging. She published more also did a clinical fellowship in vitreoretinal surgery and
than 40 peer-reviewed articles and book chapters. She was closely involved in the Fight Retinal Blindness
received her “venia legendi” in ophthalmology, in particu- Project (FRB!). He was the first Swiss participant in the
lar in retinal imaging from the University of Zurich in European Leadership Development Program, finished the
2019. She is involved in training of medical students and executive MBA program at the University of Zurich in
residents in ophthalmology. March 2018 and was appointed director of the Department
of Ophthalmology at the University Hospital Zurich and rent research involves both clinical and basic research top-
chair of ophthalmology at the University of Zurich in ics. He did his MD thesis on gene identification using a
August 2018. gene-trap approach and his PhD thesis on vascular stem
DB’s research topics for more than 12 years are age- cells. DB has coauthored more than 110 original
related macular degeneration (AMD), hereditary retinal manuscripts.
degeneration, and retinal vascular disease. Past and cur-
Histopathology of Choroidal
Evangelina Esposito, Julio A. Urrets-Zavalia,
and Pablo Zoroquiain
3.1 General Histopathology body and the iris. Its thickness varies in humans
of CNV from 0.1 mm anteriorly and 0.22 mm posteriorly
(Fig. 3.1); however, it decreases by age 90 to about
The globe is composed of three layers: fibrous 80 μm [1]. The choroid is composed of vessels that
sheet, vascular sheet, and nervous sheet. The first are derived from the anastomosis of branches of
is composed of the cornea and the sclera, and pro- the ophthalmic artery. These are the posterior cili-
vides the structure of the eye. The second, called ary arteries, and penetrate the sclera posteriorly,
the uveal tract, is composed of the iris, ciliary approximately 6 mm far from the optic nerve. The
body, and the choroid. The third is composed of arteries then branch into terminal arterioles that
the retina. The choroid is a pigmented and highly feed the choriocapillaris. These subsequently drain
vascularized component of the uveal tract in the into venules that merge to form the 4–5 vortex
eye, allowing for light absorption and providing veins at the equator of the sclera [2].
oxygen and nutrients to the outer retina. The choroid and the retina are anatomically
Anatomically, the choroid extends from the ora and functionally related, the retinal pigmentary
serrata to the optic nerve head and is located at the epithelium (RPE), photoreceptors and the chorio-
posterior two-thirds of the eye between the sclera capillaris are described as a functional unit [3].
and retina. Anteriorly, it is followed by the ciliary Choroidal neovascularization is controlled by a
dynamic balance between membrane-bound and
diffusible substances with properties that either
E. Esposito (*) promote or inhibit blood vessel development [4].
Department of Ophthalmology, University Clinic Choroidal neovascularization is a major cause
Reina Fabiola, Universidad Catolica de Cordoba, of blindness, and is characterized by the three
Cordoba, Argentina
patterns of growth of newly formed vessels from
Department of Pathology, University Clinic Reina the choriocapillaris through Bruch’s membrane,
Fabiola, Universidad Catolica de Cordoba,
Cordoba, Argentina infiltrating sub-RPE space (type 1) (Fig. 3.2),
between retina and RPE (type 2) or combined
J. A. Urrets-Zavalia
Department of Ophthalmology, University Clinic (type 3) [5]. The mechanism of this neovascular-
Reina Fabiola, Universidad Catolica de Cordoba, ization is not well elucidated.
Cordoba, Argentina Any damage in the Bruch’s membrane or RPE
P. Zoroquiain may lead to CNV, which represents an altered
Department of Pathology, School of Medicine, healing process secondary to a chorioretinal
Pontificia Universidad Católica de Chile,
injury. In this scenario, no single cause for CNV
Santiago, Chile

16 E. Esposito et al.
a b
Fig. 3.1 Histology of the choroid. (a) The choroid is the posterior aspect of the uveal tract, located between the retina
and the sclera (7×). (b) The sclera, choroid, and retina (400×)
a b
Fig. 3.2 Choroidal neovascularization. (a) SD-OCT between Bruch’s membrane and RPE. Bruch’s membrane
image showing sub-RPE proliferation. (b) The histopa- in this image is thickened and showed some external
thology of CNV, note the fibrovascular membrane located excrescences (400×)
is identifiable. Rather, it represents a broad spec- In each disease, CNV may be accompanied by
trum of conditions arising from different etiolo- findings associated with the primary injury. In
gies. On light microscopy, different amounts and children and young adults, the development of
types of blood vessels, inflammatory exudate or CNV usually is secondary to choroidal osteoma,
infiltrate, fibrosis and scaring process are seen. A pathologic myopia, punctate inner choroidopa-
spectrum of findings can be found, with more thy, hereditary macular dystrophy, and angioid
“active” lesions, similar to any other granulation streaks but may also be idiopathic [7].
tissue or more “scarring” lesions on the other
side. On electron microscopy, the most common
cellular components are RPE, macrophages, 3.2 Inflammatory Associated
erythrocytes, fibrocytes, and vascular endothe- CNV
lium. The most common extracellular compo-
nents are 24-nm collagen and fibrin [6]. Clinically, Both infectious and noninfectious uveitic entities
this altered healing process is called CNV mem- can lead to CNV [8]. Of the clinically evident
branes. If this repair process is composed only by inflammatory CNV, the vast majority are classic
fibrous tissue without the proliferation of vessels CNV on fluorescein angiography and type 2
above Bruch’s membrane (or the blood vessels CNV on optic coherence tomography imaging
are fully regressed) the term scar is used. (OCT) [9, 10]. Inflammatory associations of
3 Histopathology of Choroidal Neovascularization 17
CNV are usually related to breaks in RPE or 3.2.1.1 Presumed Ocular

Bruch’s membrane. Moreover, they are usually Histoplasmosis Syndrome
associated with granulomas, scars, or choroidal (POHS)
granulomas [11]. In focal disease processes, such as POHS, anti-
gen deposition in the area of the Bruch’s mem-
brane leads to a focal inflammatory response, a
3.2.1 Non-granulomatous break in the Bruch’s membrane, and granulation
Inflammation tissue proliferation (CNV) into the subretinal
space [5]. CNV may appear peripapillary
On histopathology, non-granulomatous processes (Fig. 3.3) or juxtafoveal, and is known to be a
can be defined as a predominantly exudative and prominent feature in POHS [12, 13].
distortive process. Contrary to this, proliferative
changes are subtle (i.e., granulomas, lymphoid 3.2.1.2 P unctate Inner Choroidopathy
aggregates, exuberant granulation tissue). As a (PIC)
sequela of uveitis, the choroid may show focal or PIC is a multifocal choroiditis that affects young
diffuse areas of atrophy or scarring. myopic women. It presents with blurred vision,
Retinochoroiditis or chorioretinitis may destroy photopsias, or paracentral scotomas. Multiple
Bruch’s membrane and the retinal pigment epi- small, round, subretinal yellow-white lesions are
thelium. Due to the fact that the regenerative observed in the posterior pole that heal and sub-
capabilities of these tissues are poor, most cases sequently form small atrophic scars. Sometimes,
will develop CNV with the possibility of fibrosis a shallow neurosensory detachment may overlay
and chorioretinal fusion [9]. the lesions. CNV may complicate PIC in more
a b
Fig. 3.3 Peripapillary CNV (a) Funduscopic image (b) Fluorescein angiography (c) OCT image
than 50% of cases, and they develop within 1 year with hypopyon, aphthous ulcers of the mouth and
of initial disease [9]. genitalia, dermatitis, arthralgia, thrombophlebi-
tis, and neurologic disturbances. The disease is
3.2.1.3 Serpiginous Choroiditis (SC) most common in men, especially between the
Serpiginous choroiditis is a rare chronic, pro- ages of 20 and 30 years.
gressive, recurrent, bilateral asymmetric, poste- Pathological examination of the eyes diag-
rior uveitis of unknown etiology. It is very nosed as Behçet’s disease show a serohemor-
important to differentiate between classic SC rhagic exudate containing polymorphonuclear
and serpiginous-like choroiditis before initiat- leukocytes in the vitreous and in the anterior
ing aggressive immunomodulatory therapy, and posterior chambers. There are extensive
knowing the relationship of the latter with tuber- areas of retinal necrosis. Depending on the stage
culosis [14]. of the disease, mononuclear and polymorpho-
The disease extends centrifugally from the nuclear leukocytes can be found in the choroid.
peripapillary region toward the posterior pole. The choroidal infiltrate is predominantly com-
Visual acuity may be severely compromised posed of CD4 T lymphocytes, with some B lym-
when the disease progresses through the macula, phocytes and plasma cells (Fig. 3.5). If retinal
or when a submacular CNV membrane develops. necrosis affects Bruch’s membrane it may
CNV complicates serpiginous choroiditis in up to develop CNV.
35% of cases. As it occurs within an area of cho-
rioretinal disturbance, it is sometimes difficult to 3.2.1.6 Pars Planitis
detect clinically, and is more readily detected by This disease usually affects children or young
fluorescein angiography and OCT. adults. The histopathologic features include
detachment and collapse of the vitreous body
3.2.1.4 A cute Posterior Multifocal with fibrous organization of the vitreous base,
Placoid Pigment Epitheliopathy chronic inflammatory cells in the vitreous,
(APMPPE) edema of the optic nerve head and macula, reti-
APMPPE is a rare inflammatory bilateral intra- nal phlebitis and periphlebitis, preretinal mem-
ocular disease that affects generally healthy branes associated with breaks in the internal
young adults, characterized by sudden onset of limiting membrane, anterior traction of the
paracentral scotomas, photopsia, and blurred peripheral retina, and no significant choroiditis,
vision, and the appearance of multifocal cyclitis, or peripheral chorioretinal atrophy
yellowish-white placoid lesions of different (Fig. 3.6) [17]. Choroidal neovascularization is a
sizes in the posterior pole and mid-periphery. rare complication of intermediate uveitis, and
Visual symptoms recover after a course of a few pathophysiologic consideration suggests that
weeks, and healing of fundus lesions leaves a chronic disc edema may be a risk factor for this
mottled RPE or an irregularly pigmented and condition [18].
atrophic chorioretinal scar [15]. In the acute
phase, on fluorescein angiography lesions show
early hypofluorescence followed by late hyper- 3.2.2 Granulomatous Inflammation
fluorescence (Fig. 3.4). Very rarely, a CNV
membrane may develop within an area of a Granulomatous inflammation is a type of chronic
healed lesion [16]. inflammation characterized by a cellular infiltrate
of histiocytes. In addition, lymphocytes, plasma
3.2.1.5 Behçet’s Disease cells, and polymorphonuclear cells, such as
Behçet’s disease (syndrome) is characterized by eosinophils and neutrophils may be also observed
retinal vasculitis, recurrent bilateral iridocyclitis [19, 20].
a b
Fig. 3.4 Acute posterior multifocal placoid pigment epitheliopathy (a) Funduscopic image (b) Fluorescein angiogra-
phy (c) SD-OCT image
a b
Fig. 3.5 Behçet’s disease (a) Funduscopic image (b) Full 1st edition, Esposito, et al, Choroidal Histopathology,
thickness choroidal inflammatory infiltrate composed of Pages No. 21-48, Copyright 2017, with permission from
lymphocytes, neutrophils, and plasma cells are seen. No Elsevier
vasculitis is present. Reprinted from Choroidal Disorders,
a b
Fig. 3.6 Pars planitis (a) Funduscopic image showing peripheral vasculitis (arrowheads) (b) Fluorescein angiography
showing macular edema
ity of the cases are diagnosed clinically and CNV

3.2.2.1 Toxoplasmosis complication is uncommon [25]. Although the
Ocular toxoplasmosis is a parasitic infection of parasite is confined to the retina, breaks in
the eye caused by the protozoan Toxoplasma Bruch’s membrane will permit the contact of the
gondii. Infections may be congenital or acquired choroid with the infectious antigen, thereby caus-
through the ingestion of uncooked and infected ing an inflammatory response. This may lead to
meat, contaminated vegetables or water [21]. CNV that appears at the border of the scar and the
This disease typically affects the posterior healthy retina [25].
pole of the eye and the lesions can be solitary or Ocular toxoplasmosis often presents as extensive
multiple and can further be subclassified as active granulomatous inflammatory infiltration of the
or scaring. Active lesions are gray-white and choroid and areas of necrosis in Bruch’s mem-
accompanied by exudation, vasculitis, and cho- brane. In immunocompromised patients, the
roiditis (Fig. 3.7). inflammatory infiltrate may be minimal or absent.
The normally clear vitreous is compromised Therefore, the focal areas of necrosis are impor-
and becomes hazy due to the infiltration of tant clues to make the correct diagnosis of toxo-
inflammatory cells. The patient will generally plasmosis [24].
not complain of pain, but rather of an increase
in floaters and a possible decrease in vision in
the affected eye [22]. The scarring begins from 3.2.3 Tuberculosis
the periphery of the lesion, and progresses
toward the center with variable pigmentation Tuberculosis is an infectious disease caused by
changes [23]. the acid-fast bacilli Mycobacterium tuberculosis
T. gondii primarily affects the retina and sec- and is characterized pathologically by the for-
ondarily the choroid, although choroidal lesions mation of granulomas with a central area of
do not occur in the absence of retinal infection. caseous necrosis. The most frequent route that
Histopathological confirmation may be the bacilli reaches the eye is through the blood-
obtained by chorioretinal biopsies, and more stream [19].
rarely, enucleation. The toxoplasma cysts, brady- In posterior uveitis caused by tuberculosis,
zoites, and tachyzoites can be identified with the ocular changes can be divided into four
hematoxylin and eosin (H&E), immunohisto- groups: choroidal tubercles, choroidal tubercu-
chemistry, or by PCR [24]. However, the major- loma, subretinal abscess, and serpiginous-like
a b
c d
Fig. 3.7 Toxoplasmosis (a) Funduscopic image (b) The (arrow) H&E 200×. (c) Bradyzoites cyst (arrow) H&E
inflammation extends to the inner part of the choriocapil- 200×. (d) The microorganisms are highlighted with anti-
laris. Note the dense lymphocytic and plasmocytic reac- toxoplasmosis immunohistochemistry (DAB 200×)
tion of the choroid and the multinucleated giant cell
choroiditis (SLC) [26]. The exact mechanism of Langhans type, and a rim of small lymphocytes
SLC in tuberculosis remains unknown. It may (Fig. 3.8).
represent an immune-mediated hypersensitivity These inflammatory phagocytes in turn are
reaction (type IV) without the presence of acid- surrounded by lymphocytes. The necrotic area
fast bacteria in the choroid or retinal pigment usually contains few bacteria, which can be visu-
epithelium [26]. The lesions are usually multi- alized on Ziehl–Neelsen acid-fast stain as red
focal, bilateral, noncontiguous to optic disc, and rod-shaped organisms. However, several organ-
are commonly associated with mild vitreous isms can also be seen in the necrotic macrophages
inflammation. There are two distinct clinical that line caseous necrosis. In some granulomas,
patterns: discrete, multifocal choroiditis lesions the organisms can be seen in multinucleated giant
that are initially noncontiguous but later prog- cells (Fig. 3.3) or more frequently, they may not
ress to form diffuse lesions with an active edge, be detected by the histologic staining techniques
resembling serpiginous choroiditis; and less [26]. Currently PCR-based diagnostic tools are
commonly, a solitary, plaque-like lesion. highly sensitive and specific [29].
Neovascularization, when present, has been In the choroid, these tubercles/tuberculomas
reported as Type 1 (sub-RPE) [27]. may involve all layers of the choroid. In the early
The disease is characterized pathologically by stages, the overlying RPE remains normal but is
the formation of one or multiple granulomas disrupted during later stages as the tubercles
[28]. The histology of the granuloma reveals cen- increase in size. The surrounding choroid is
tral necrosis surrounded by histiocytes/epitheli- essentially normal except for some lymphocytic
oid cells mixed with multinucleated giant cells, infiltration [26, 30–33].
a b
Fig. 3.8 (a) Fundus image of the left eye of a patient with ated giant cells (arrows). A rim of lymphocytes
tuberculosis (b) Chronic choroidal granulomatous inflam- surrounding the granuloma is seen
mation with caseous necrosis and Langhans multinucle-
3.2.4 Vogt–Koyanagi–Harada (VKH) 9% of cases [9, 40]. Type 2 accounts for all cases
Syndrome of CNV complicating these cases [9] and the
macular and peripapillary areas seem to be the
Vogt–Koyanagi–Harada (VKH) syndrome is a most frequently affected areas by this complica-
bilateral granulomatous uveitis that is associated tion [41].
with integumentary, auditory, and central nervous The histopathology of the choroid depends on
manifestations [34]. the stage of the disease. In acute VKH, there is gen-
The pathogenesis underlying VKH is thought eralized granulomatous inflammation of the cho-
to be autoimmune, with T cells mounting a roid with uveal thickening [42]. This is due to the
response against melanocytes. There are acute and infiltration with lymphocytes, macrophages, epi-
chronic stages of the disease, with the former thelioid cells [36], and plasma cells. The sensory
responding well to corticosteroid treatment; retina may be detached from the pigment epithe-
chronic disease may have recurrent bouts of acute lium by a protein exudate with eosinophils [43].
activity [35]. In the eye, VKH presents with poste- Dalen-Fuchs nodules, which are clusters of macro-
rior uveitis or diffuse granulomatous panuveitis, in phages and RPE cells located over Bruch’s mem-
association with serous exudative detachment and brane, may also be observed (Fig. 3.9) [35]. In the
disc hyperemia, secondary to increased permeabil- early stages, the choriocapillaris is usually spared.
ity and leakage of the choroidal vessels [34, 36]. In contrast, the choroidal inflammation in
Anterior segment inflammation can also occur chronic VKH is non-granulomatous [38]. Infiltrates
concomitantly with subclinical posterior uveitis are still primarily lymphocytic, and there is marked
[37]. Chronic VKH can lead to the pathognomonic thinning of the uvea. There may be obliteration of
sunset glow fundus, corresponding to the degen- the choriocapillaris. Dalen- Fuchs nodules are
eration of the retinal pigment epithelium [38]. absent; instead, there is loss of the melanin gran-
Chronic VKH can also result in peripapillary ules in the retinal pigment epithelium. Nearby
atrophy [36] and subretinal fibrosis leading to these are focal areas of hyperpigmentation, which
neovascularization, which are poor prognostic are compensatory hyperproliferations of RPE;
factors [39]. Submacular choroidal neovascular- these tend to be arranged in papillary or tubular pat-
ization may be another cause of significant vision terns [35]. Chronic recurrent VKH resembles acute
loss in VKH syndrome and may occur in up to VKH on histopathology, characterized by granulo-
a b
Fig. 3.9 Vogt–Koyanagi–Harada syndrome. (a) RPE cells overlying Bruch’s membrane. These are charac-
Funduscopic image (b) Fluorescein angiography (c) teristic of acute VKH and chronic recurrent VKH. Note
Dalen-Fuchs nodules are clusters of macrophages and the dense choroidal lymphocytic infiltrate
matous inflammation and Dalen-Fuchs nodules but complex is not common, though when present,
with less uveal thickening and loss of choroidal has been reported as type 2 CNV [10].
melanocytes [43]. Sarcoid nodules or tubercles are pathogno-
monic of this disease. On histology, these are
circumscribed noncaseating granulomas com-
3.2.5 Sarcoidosis posed of primarily lymphocytes in association
with Langerhans giant cells and macrophages
Sarcoidosis is a noninfectious inflammatory [48]. Tubercles of the same size are usually seen
granulomatous disease that may involve a single in isolation, although these may sometimes
or multiple systems. The lungs, lymph nodes, coalesce. Clusters of the epithelioid and
skin, the central nervous system, and the eye can Langerhans cells are usually surrounded by
be involved. lymphocytes or plasma cells and may either be
Posterior segment involvement is reported in separated by connective tissue or form con-
14–28% of patients [44], and these can present as glomerates (Fig. 3.10) [49].
posterior uveitis and sarcoid nodules of the optic Also observed are perivascular exudates that
nerve, retina, and choroid [45]; vitritis with or correlate with “candle-wax drippings” on indi-
without inflammatory snowballs [46], retinal vas- rect ophthalmoscopy, perivascular lymphocytic
culitis, chorioretinitis, vascular occlusions, mac- and neutrophilic infiltration, and vascular sheath-
ular edema, papilledema [44] and retinal ing [50]. Acid fast and Gomori methenamine sil-
detachment [47]. The development of a CNV ver stains for fungi and bacteria, respectively,
Fig. 3.10 Sarcoid

nodules. Note the
discrete, compact
granulomas consisting
of histiocytes,
lymphocytes, and giant
cells. The granulomas
are separated by
connective tissue. No
necrosis is seen
should be negative, and there should be no signs Almost always, choroidal submacular neovas-
of a foreign body inciting the reaction [51]. cularization occurs in the context of preexisting
Sarcoidosis resembles other granulomatous clinical manifestations of dry or non-exudative
diseases on histology, such as histoplasmosis, AMD, such as macular retinal pigment epithe-
leprosy, and tuberculosis based on the discrete lium (RPE) irregularities, drusen, and/or patchy
pattern in which cellular infiltration is arranged or geographic atrophy [55].
[49]; the infrequency of necrosis and the occa- Biomicroscopy of the fundus shows a local-
sional distinctive asteroid and Schaumann bodies ized area of a shallow neurosensory detachment
that can be seen on sarcoid nodules [52]. that may be exudative, hemorrhagic, or mixed.
Also, an RPE detachment may be the initial clini-
cal sign or accompany the neurosensory detach-
3.3 Degeneration Associated ment, and is clinically observed as grayish or
CNV dark, well-delineated dome-shaped subretinal
elevation.
3.3.1 Neovascular Age-Related CNV development among patients with AMD
Macular Degeneration (AMD) can be characterized as type 1 (subretinal), type 2
(outer retinal), or mixed based on clinical and
AMD classically presents in a person over examination (including imaging) findings
50 years old with sudden blurred central vision, Fig. 3.11 [56]. A majority of CNV are type 2
metamorphopsia, and/or a central or paracentral lesions with abnormal growth of vasculature into
relative scotoma. the outer retinal space. CNV seen in AMD are
Neovascular AMD, also known as wet or exu- usually subfoveal and are associated with the
dative AMD, affects 10–20% of all AMD patients, presence of drusen and retinal pigment epithelial
and is a leading cause of severe visual impair- abnormalities due to the accumulation of lipofus-
ment among the elderly population living in cin material. On the other hand, the retinal pig-
high- or middle-income countries [53]. ment epithelium is often intact in individuals
Inflammation can play a key role in the treatment with CNV [57]. The proposed mechanism of
and pathophysiology of this condition. It is con- development of CNV is the focal breach of the
sidered that the primary event is the deposition of retinal pigment epithelium due to infection/
extracellular material (drusen). This material inflammation leading to growth and entry.
seems to be highly pro-inflammatory leading to Most of AMD CNV are type 2 (external sen-
the inflammatory state [54]. sory retinal) and are accompanied by drusen and
a b
c d
Fig. 3.11 Neovascular age-related macular degeneration brane with diffuse drusen formation is seen. Note on the
type 1 (a) Funduscopic image (b) Fluorescein angiogra- right two ghost vessels with red blood cells. (e) Dry AMD
phy (c) SD-OCT image (d) A thickened Bruch’s mem- with drusen (arrows) and no CNV H&E (200×)
RPE abnormalities similar to those of Dry AMD disruption of the submacular RPE, accompanied
such lipofuscin deposition [56] and is nicely by subretinal hemorrhage or fluid, disruption of
delineated by fluorescein angiography [58]. the outer neurosensory retina, and neurosensory
Type 1 CNV can also be seen under the RPE, cystoid edema. Type 2 CNV is easily identified
and presents clinically as an RPE detachment. between RPE and neurosensory retina, and also
Later, CNV membranes disrupt the RPE and invade neurosensory detachment and edema are gener-
the subretinal space (type 3). Type 2 gets through ally observed.
RPE directly into the sub-neurosensory space. Progressively, in untreated or unresponsive-
With optical coherence tomography (OCT), to-treatment cases, a disciform fibrotic scarring
type 1 is observed as a well-delineated RPE of the macula forms, surrounded by an area of
detachment, or as an irregular elevation and later chorioretinal atrophy [58].
3.3.2 Myopic Neovascular Choroidal neovascularization in high myopia

Maculopathy is less aggressive and expansive than AMD and
tends to regress and cicatrize spontaneously,
Progressive elongation of the globe observed in leaving a subretinal scar area of irregular cicatri-
high myopia produces a biomechanical stretch- cial hyperplasia of the RPE named Fuchs spot,
ing of the retina, RPE, and choroid [59] accom- surrounded by progressive chorioretinal atrophy
panied by a straightening and thinning of retinal [71].
vessels with reduction of retinal vascular flow, Fluorescein angiography delineates quite well
and a diminished density of the retinal capillary the generally small neovascular membrane.
network and choriocapillaris [60–62]. These However, in cases with advanced macular patchy
events appear to induce several degenerative chorioretinal atrophy the lesion may become
processes of the macular region [63, 64]. In one hardly identifiable.
study evaluating patients with unilateral myopia In OCT, the acute neovascular membrane
complicated with neovascular maculopathy by appears as a well-circumscribed non-
means of color Doppler imaging, higher resis- homogenously hyper-reflective lesion in the sub-
tivity index in posterior ciliary arteries were retinal space, with a subtle amount of subretinal
found when compared with the non-myopic fel- exudation and intraretinal cystoid edema, not
low eye [65]. always present in the initial phase.
Myopic neovascular maculopathy is one of Long-term visual prognosis is poor for
the most frequent and severe vision-threatening untreated or unresponsive-to-treatment lesions.
complications in highly myopic patients and is At 10 years after onset 96% of cases will have a
the most frequent cause of submacular choroidal visual acuity of less than 20/200 [72].
neovascularization in persons under the age of 50
years [66]. It may be observed in up to 10% of
patients with high myopia, being more frequent 3.3.3 Pachychoroid Related CNV
in women [67].
Although its pathogenesis is not well estab- In normal subjects, submacular choroidal thick-
lished, some predisposing risk factors have been ness measured by means of OCT ranges between
found, such as degenerative changes in Bruch’s 250 and 330 μm, and is thinner temporally and
membrane, thinning of choriocapillaris, and nasally. It may decrease with aging, in high axial
slowing of the choroidal circulation [68, 69]. length, and in certain diseases such as glaucoma,
Biomicroscopically, the subretinal choroidal pseudoxanthoma elasticum, and birdshot chorio-
neovascularization is generally observed as a retinopathy [73]. A diffuse choroidal thickening
round, flat, or minimally elevated brown or gray- may also be observed in normal eyes without any
ish spot, sometimes accompanied by a small pathologic consequence.
hemorrhage in the surroundings. Although it may Pachychoroid is a term to describe focal or
be observed in a highly myopic eye without evi- diffuse choroidal thickening, dilated vessels in
dent myopic fundus changes at the posterior pole Haller’s layer (pachyvessels) that may repre-
[66], it presents generally in an eye with diffuse sent the full extent of choroidal thickness, and
macular atrophy, patchy atrophy in the macular thinning of Sattler’s layer and choriocapillaris
area, and lacquer cracks [70]. Frequently, the [74, 75].
subretinal neovascular membrane develops at the Another important feature of pachychoroid is
edges of a lacquer crack, atrophy plaque, or steep choroidal vascular hyperpermeability, evidenced
staphylomatous area [70]. by indocyanine green (ICG) angiography [76].
About 35% of bilateral highly myopic patients Pachychoroid constitutes a clinical entity
with neovascular maculopathy in one eye may comprising a spectrum of diseases that includes
develop neovascular maculopathy in the contra- central serous chorioretinopathy, pachychoroid
lateral eye within 8 years [70]. pigment epitheliopathy, polypoidal choroidal
vasculopathy, and pachychoroid neovasculopa- long as the anomalous network is not covered by
thy [74]. a dense hemorrhage plaque [81].
The latter consists in a choroidal neovascular
complex that develops overlying the RPE (type 1
neovascularization) in a localized area of dilated 3.3.5 Macular Telangiectasia
choroidal vessels and choroidal thickening, diag-
nosed by means of OCT, and may progress to Macular telangiectasia comprises a variety of
polypoidal choroidal vasculopathy [77]. clinical pictures characterized by altered juxta- or
Clinically, besides some focal RPE abnor- parafoveolar retinal capillaries. It may be classi-
malities, choroidal neovascularization occurs in fied into two major forms. Type 1 is the congeni-
the absence of age-related degenerative changes tal form, also known as mac tel type 1, and is
such as drusen and atrophy, or other types of usually unilateral. Type 2, also known as mac tel
degenerative disease [77]. Irregular and shallow type 2, is presumably acquired, affecting both
RPE detachments are very characteristic of this eyes of middle-aged and older individuals [82].
entity. Under this area, a complex choroidal vein Intraretinal neovascularization is a frequent
network may be visualized by OCT angiogra- complication of mac tel type 2, and is more
phy [77, 78]. often observed temporal to the fovea, and may
A variant of this syndrome is peripapillary extend to the subretinal space and choriocapil-
pachychoroid, described recently by laris. Intraretinal and subretinal lipid deposits,
Phasukkijwatana et al., and is characterized by subretinal or retinal hemorrhage, and macular
pachychoroid around the optic disc, choroidal edema, and a right-angle retinal venule are
folds, hyperopia, short axial length, nasal macu- reaching the center of the lesion are key clinical
lar edema and/or detachment, and occasional findings [83].
optic disc edema [79].
The pathogenesis of pachychoroid and its
associations remains unknown. 3.3.6 NV Secondary to Angioid
C
Streaks
3.3.4 Polypoidal Choroidal Histopathology has shown that angioid streaks

Vasculopathy (PCV) are caused by breaks within an abnormally thick-
ened and often calcified Bruch’s membrane [84].
Polypoidal choroidal vasculopathy (PCV) is a The overlying RPE is atrophic but not necessarily
variant of type 1 choroidal neovascularization, discontinuous and the blood–retinal barrier is
characterized by abnormal focal vascular branch- intact. CNV is represented by blood vessels that
ing network and terminal saccular dilatation invade the break, breach the epithelium, and pro-
resembling polyps that may bleed and exudate liferate in the subretinal space [85].
profusely under the neurosensory retina of the
posterior pole in longstanding lesions. Polypoidal
lesions may sometimes be identified as a subreti- 3.3.7 Best Associated CNV
nal orange bulging lesion principally located
around the optic disc and the macular area, are Usually the ganglion cell layer and inner and
frequently bilateral, and are less frequently outer plexiform layers of the central retina are
observed with drusen than in AMD [80]. Risk edematous, and the outer segments showed focal
factors include male gender, smoking, hyperlip- atrophy, lipofuscin-like material under the macu-
idemia, obesity, and hypertension. ICGA is the lar RPE. Bruch’s membrane changes such as dis-
gold standard for their diagnosis, but they can ruption or thickening were also observed along
also be observed with fluorescein angiography as with CNV [86].
3.4 Traumatic Associated CNV well-defined flat or slightly elevated mass, found
around the optic disc or in the macula. The high
3.4.1 Post-trauma CNV tissue density lesion is also evidenced by ultraso-
nography and CT of the orbit [91]. Vision may
Ocular blunt trauma is a major cause of prevent- decrease slowly as the result of degenerative
able visual loss secondary to posterior segment changes at the choriocapillaris and RPE over the
complications, such as vitreous hemorrhage, reti- years. However, sudden visual loss may occur as
nal detachment, macular hole, traumatic optic the consequence of serous or hemorrhagic macu-
neuropathy, choroidal rupture, and CNV [87]. lar detachment from choroidal neovasculariza-
Choroidal rupture is a break in the choroid, tion (CNV) [92].
Bruch’s membrane, and RPE, and 5–10% of The histopathological description of choroidal
those eyes may develop a late CNV [88]. osteoma in the literature is limited. On histopa-
Choroidal ruptures located closer to the fovea, as thology, choroidal osteomas appear as mature
well as longer ruptures, were at higher risk for bone with marrow spaces connected to each other
developing CNV, 82% occurring within the first- and filled with connective tissue [93, 94]. The
year post trauma [88]. Clinically they present as a bony trabeculae contain osteoblasts, osteocytes,
relatively small, flat, or slightly elevated hemor- and occasionally osteoclasts [94]. Superficially,
rhagic or grayish plaques that typically develop some of the marrow spaces will have feeder ves-
within or at the border of a chorioretinal trau- sels that connect to choriocapillaris beneath
matic scar. Bruch’s membrane. The lumen of the choriocap-
illaris can be obliterated by the tumor in most
areas [94]. An atrophied and depigmented RPE
3.4.2 LASER Induced CNV layer is usually seen overlying the tumor [93, 94].
In addition, subretinal exudates and a degener-
Classically, immediately after laser treatment ated photoreceptor layer above the tumor, with
histopathology shows coagulative necrosis loss of nuclei, can be observed [93, 94].
involving all retinal layers that in 3 weeks evolve
to a marked retinal attenuation and chorioretinal
scarring [89]. The irradiated cells show vacuola- 3.5.2 Choroidal Nevus
tion damage to the inner retinal layers and some
of them vacuolation in their inner segments, pyk- Choroidal nevi are benign tumors, generally oph-
notic nuclei, or degeneration in the fiber layer of thalmoscopically observed as a dark-gray flat or
Henle. Increase in output energy results in minimally elevated lesion, and most frequently
increasing trauma at damaged sites. According to observed posterior to the equator and extrafoveo-
the LASER power, a breakdown of the pigment lar. Patients with subfoveolar nevi are at certain
epithelium and Bruch’s membrane may be seen risk of visual loss because of choriocapillaris dis-
[90]. Neovascularization shares the common his- ruption and subsequent disruption of photorecep-
topathological characteristics at the margin of the tor outer segments. Also, a CNV complex may
LASER scar. develop at the margins or over the choroidal
nevus generating a serous neurosensory detach-
ment [95]. In longstanding cases lipid deposition
3.5 Tumoral Associated CNV may precipitate in the subretinal space contour-
ing the CNV lesion (Fig. 3.12).
3.5.1 Choroidal Osteoma Choroidal nevi arise from melanocytic cells
derived from the neural crest [96, 97]. In general,
Choroidal osteoma is a benign ossifying tumor the cells of a nevus cell are “plumper” than the
affecting more frequently young women that normal melanocytes of the choroid and ciliary
presents ophthalmoscopically as a white or pale body. Morphologically, nevi are classified into
Fig. 3.12 Choroidal

nevus. Note the CNV
characterized by a
vascularized fibrous
membrane over Bruch’s
membrane. RPE and
outer retinal segments
are disrupted, H&E
(100×)
four types: polyhedral, fusiform, spindle, and occasionally even a small nucleolus.
balloon cells [97]. Furthermore, the cytoplasm is more abundant
than in the spindle melanocytes.
1. Polyhedral nevus cells: This is the most com- 4. Balloon cells: This subtype of cells is similar
mon cell type found in nevi of the ciliary to cells found in cutaneous nevi. These are
body and choroid. Its voluminous cytoplasm large cells with abundant foamy cytoplasm.
is densely packed with melanin, obscuring
nuclear details. When sections are thor- Histopathological alterations in the structures
oughly bleached, a rather small, round, uni- surrounding choroidal nevi include narrowing or
formly basophilic nucleus without a dilatation of the choriocapillaris, atrophy and
prominent nucleolus is seen. This cell clumping of the retinal pigment epithelial (RPE),
accounts for over two-thirds of the mass in drusen formation, serous detachment of the neu-
the majority of these nevi. It is indistinguish- rosensory retina, subretinal neovascularization,
able from those cells that compose the mela- and lipofuscin deposition [97–99].
nocytomas of the optic disc and from those
that diffusely thicken the choroid in congeni-
tal ocular melanocytosis. 3.6 evelopment Alterations or
D
2. Spindle nevus cells: This is considered the Malformations Associated
second most frequent cell type in benign pig- CNV
mented tumors of the choroid. It is a small,
spindle-shaped cell with a slender, intensely 3.6.1 Coloboma
basophilic nucleus. Unlike polyhedral cells,
these cells consistently contain little or no pig- Choroidal coloboma is a congenital malforma-
ment and are often distributed in a striking tion characterized by a failure during embryo-
manner in the outer portions of a nevus. Only genesis of the inner and outer layers to fuse along
rarely does this cell type make up the bulk of the optic fissure [100].
a nevus, which is then characteristically only The simultaneous closure of the fissure and
lightly pigmented. retinal differentiation, together with subsequent
3. Fusiform and dendritic nevus cells: These retinal growth, are vital with respect to the patho-
cells are less intensely pigmented than the physiology of the fissure in that only the inner
polyhedral cells and display a larger nucleus neuroblastic layer of the retina, with its Müller
with a slightly loose chromatin pattern and cells, fuses across the fissure. The outer neuro-
a b
c d
Fig. 3.13 Coloboma. (a) Histology of a normal develop- (d) SD-OCT image Histology of coloboma. The choroid
ing eye in the 8th gestational week. Note the primitive two is replaced only by a loose connective tissue. The retina in
layers of the retina, immature choroid and sclera H&E this image is dysplastic with rosette formation
200× (b) Funduscopic image (c) Fluorescein angiography
blastic layer reverts and connects to the RPE the RPE. Also variable degrees of vascularization
(Fig. 3.13) [101]. at the junction between the external limiting
As described by Schubert (2005), common membrane and the RPE, the choroid ends together
histologic findings are [101]: with the RPE and may be thickened and the sclera
Pediatric coloboma is characterized by an is thinned. Intrascleral rosettes may be present.
intercalary membrane in direct contact with the In adult coloboma marginal retinal vessels,
sclera, central glial triangle, point of reversal and marginal choroidal thickening and hyperplasia of
duplication of the photoreceptor layer, locus the RPE are encountered together with vessels
minoris resistentiae, and lateral displacement of with intimal hyperplasia and absence of the glial
triangle, point of reversal, duplication, and locus lar surgery trials report no. 7. Arch Ophthalmol.
2005;123(7):914–21.
minoris resistentiae. The choroid is absent at the 7. Barth T, Zeman F, Helbig H, Oberacher-Velten
center and the sclera is thinned. I. Etiology and treatment of choroidal neovascu-
Choroidal neovascularization favors the mar- larization in pediatric patients. Eur J Ophthalmol.
gin of colobomas, with vessels tending to grow 2016;26(5):388–93.
8. Kuo IC, Cunningham ET Jr. Ocular neovasculariza-
toward the intercalary membrane [102–104]. tion in patients with uveitis. Int Ophthalmol Clin.
2000;40(2):111–26.
9. Wu K, Zhang X, Su Y, Ji Y, Zuo C, Li M, et al.
3.7 Unknown: Idiopathic CNV Clinical characteristics of inflammatory choroidal
neovascularization in a Chinese population. Ocul
Immunol Inflamm. 2016;24(3):261–7.
The pathophysiological processes in idiopathic 10. Querques L, Querques G, Miserocchi E, Modorati
CNV remain unclear. Histopathology of this G, Bandello F. Intravitreal aflibercept for choroi-
lesion may correspond to the general description dal neovascularization in ocular sarcoidosis. Eur J
Ophthalmol. 2016;26(5):e124–7.
of CNV [6]. 11. Roy R, Saurabh K, Bansal A, Kumar A, Majumdar
AK, Paul SS. Inflammatory choroidal neovascular-
Key Learning Points ization in Indian eyes: etiology, clinical features, and
CNV is a repair response during the healing pro- outcomes to anti-vascular endothelial growth factor.
Indian J Ophthalmol. 2017;65(4):295–300.
cess after injury due to the poor regenerative 12. Krill AE, Archer D. Choroidal neovascularization
properties of the retina, Bruch’s membrane, and in multifocal (presumed histoplasmin) choroiditis.
choriocapillaris. Arch Ophthalmol. 1970;84(5):595–604.
Choroidal neovascularization represents a 13. Macular Photocoagulation Study Group. Laser pho-
tocoagulation for neovascular lesions nasal to the
nonspecific response to different etiologies. fovea. Results from clinical trials for lesions second-
In each disease, CNV may be accompanied by ary to ocular histoplasmosis or idiopathic causes.
findings associated with the primary etiologic Arch Ophthalmol. 1995;113(1):56–61.
condition. 14. Dutta Majumder P, Biswas J, Gupta A. Enigma
of serpiginous choroiditis. Indian J Ophthalmol.
CNV may be localized in any part of the ret- 2019;67(3):325–33.
ina, but macular involvement is common. 15. Xerri O, Salah S, Monnet D, Brezin AP. Untreated
acute posterior multifocal placoid pigment epitheli-
opathy (APMPPE): a case series. BMC Ophthalmol.
2018;18(1):76.
References 16. Bowie EM, Sletten KR, Kayser DL, Folk JC. Acute
posterior multifocal placoid pigment epitheli-
1. Ramrattan RS, van der Schaft TL, Mooy CM, de opathy and choroidal neovascularization. Retina.
Bruijn WC, Mulder PG, de Jong PT. Morphometric 2005;25(3):362–4.
analysis of Bruch’s membrane, the choriocapillaris, 17. Pederson JE, Kenyon KR, Green WR, Maumenee
and the choroid in aging. Invest Ophthalmol Vis Sci. AE. Pathology of pars planitis. Am J Ophthalmol.
1994;35(6):2857–64. 1978;86(6):762–74.
2. Kiel JW. The ocular circulation. In: Granger DN, 18. Mehta S, Hariharan L, Ho AC, Kempen
Granger JP, editors. Integrated systems physiology: JH. Peripapillary choroidal neovascularization
from molecule to function to disease. San Rafael, in pars planitis. J Ophthalmic Inflamm Infect.
CA: Morgan & Claypool Life Sciences; 2010. 2013;3(1):13.
3. Strauss O. The retinal pigment epithelium in visual 19. Yanoff M, Sassani JW. Ocular pathology. 6th ed.
function. Physiol Rev. 2005;85(3):845–81. Edinburgh: Mosby/Elsevier; 2009. p. x, 789.
4. Hanahan D, Folkman J. Patterns and emerging 20. Williams GT, Williams WJ. Granulomatous inflam-
mechanisms of the angiogenic switch during tumori- mation—a review. J Clin Pathol. 1983;36(7):723–33.
genesis. Cell. 1996;86(3):353–64. 21. Silveira C, Belfort R Jr, Burnier M Jr, Nussenblatt
5. Grossniklaus HE, Green WR. Choroidal neo- R. Acquired toxoplasmic infection as the cause of
vascularization. Am J Ophthalmol. 2004;137(3): toxoplasmic retinochoroiditis in families. Am J
496–503. Ophthalmol. 1988;106(3):362–4.
6. Grossniklaus HE, Miskala PH, Green WR, Bressler 22. Nussenblatt RB, Belfort R Jr. Ocular toxoplasmosis.
SB, Hawkins BS, Toth C, et al. Histopathologic An old disease revisited. JAMA. 1994;271(4):304–7.
and ultrastructural features of surgically excised 23. Commodaro AG, Belfort RN, Rizzo LV, Muccioli C,
subfoveal choroidal neovascular lesions: submacu- Silveira C, Burnier MN Jr, et al. Ocular toxoplasmo-
sis: an update and review of the literature. Mem Inst 38. Takahashi H, Takase H, Ishizuka A, Miyanaga M,
Oswaldo Cruz. 2009;104(2):345–50. Kawaguchi T, Ohno-Matsui K, et al. Choroidal
24. Belfort RN, Rasmussen S, Kherani A, Lodha N, thickness in convalescent Vogt-Koyanagi-Harada
Williams G, Fernandes BF, et al. Bilateral progres- disease. Retina. 2014;34(4):775–80.
sive necrotizing retinochoroiditis in an immunocom- 39. Lertsumitkul S, Whitcup SM, Nussenblatt RB, Chan
promised patient: histopathological diagnosis. Acta CC. Subretinal fibrosis and choroidal neovascular-
Ophthalmol. 2010;88(5):614–5. ization in Vogt-Koyanagi-Harada syndrome. Graefes
25. Benevento JD, Jager RD, Noble AG, Latkany Arch Clin Exp Ophthalmol. 1999;237(12):1039–45.
P, Mieler WF, Sautter M, et al. Toxoplasmosis- 40. Rubsamen PE, Gass JD. Vogt-Koyanagi-Harada syn-
associated neovascular lesions treated successfully drome. Clinical course, therapy, and long-term visual
with ranibizumab and antiparasitic therapy. Arch outcome. Arch Ophthalmol. 1991;109(5):682–7.
Ophthalmol. 2008;126(8):1152–6. 41. Inomata H, Minei M, Taniguchi Y, Nishimura
26. Gupta V, Gupta A, Rao NA. Intraocular F. Choroidal neovascularization in long-standing
tuberculosis—an update. Surv Ophthalmol. case of Vogt-Koyanagi-Harada disease. Jpn J
2007;52(6):561–87. Ophthalmol. 1983;27(1):9–26.
27. Aggarwal K, Agarwal A, Sharma A, Sharma K, 42. Lubin JR, Ni C, Albert DM. Clinicopathological
Gupta V, Group OS. Detection of type 1 choroidal study of the Vogt-Koyanagi-Harada syndrome. Int
neovascular membranes using optical coherence Ophthalmol Clin. 1982;22(3):141–56.
tomography angiography in tubercular posterior 43. Rao NA. Pathology of Vogt-Koyanagi-Harada dis-
uveitis. Retina. 2019;39(8):1595–1606. ease. Int Ophthalmol. 2007;27(2–3):81–5.
28. Centers for Disease Control and Prevention. Case 44. Verma A, Biswas J. Choroidal granuloma as an
definitions for infectious conditions under pub- initial manifestation of systemic sarcoidosis. Int
lic health surveillance. MMWR Recomm Rep. Ophthalmol. 2010;30(5):603–6.
1997;46(RR-10):1–55. 45. Cook BE Jr, Robertson DM. Confluent choroidal
29. Biswas J, Kazi MS, Agarwal VA, Alam MS, Therese infiltrates with sarcoidosis. Retina. 2000;20(1):1–7.
KL. Polymerase chain reaction for Mycobacterium 46. Kawaguchi T, Hanada A, Horie S, Sugamoto Y,
tuberculosis DNA detection from ocular fluids Sugita S, Mochizuki M. Evaluation of charac-
in patients with various types of choroiditis in a teristic ocular signs and systemic investigations
referral eye center in India. Indian J Ophthalmol. in ocular sarcoidosis patients. Jpn J Ophthalmol.
2016;64(12):904–7. 2007;51(2):121–6.
30. Duke-Elder S, Perkin ES. System of ophthalmology: 47. Olk RJ, Lipmann MJ, Cundiff HC, Daniels
diseases of the uveal tract. St Louis: CV Mosby; 1966. J. Solitary choroidal mass as the presenting sign in
31. Lyon CE, Grimson BS, Peiffer RL Jr, Merritt systemic sarcoidosis. Br J Ophthalmol. 1983;67(12):
JC. Clinicopathological correlation of a soli- 826–9.
tary choroidal tuberculoma. Ophthalmology. 48. Whitcup SM, Chan CC, Geiger GL. Diagnosis of
1985;92(6):845–50. corticosteroid resistant ocular sarcoidosis by chorio-
32. Vrioni G, Levidiotou S, Matsiota-Bernard P, Marinis retinal biopsy. Br J Ophthalmol. 1999;83(4):504–5.
E. Molecular characterization of Mycobacterium 49. Hogan MJ, Zimmerman LE. Ophthalmic pathol-
tuberculosis isolates presenting various drug suscep- ogy, an atlas and textbook. 2nd ed. Philadelphia:
tibility from Greece using three DNA typing meth- W.B. Saunders; 1962. p. xiv, 797.
ods. J Infect. 2004;48(3):253–62. 50. Gass JD, Olson CL. Sarcoidosis with optic nerve
33. Spencer WH. Ophthalmic pathology. 4rd ed. and retinal involvement. Arch Ophthalmol.
Philadelphia: W.B. Saunders; 1990. 1976;94(6):945–50.
34. Read RW, Rao NA. Utility of existing Vogt- 51. Frank KW, Weiss H. Unusual clinical and histo-
Koyanagi- Harada syndrome diagnostic criteria pathological findings in ocular sarcoidosis. Br J
at initial evaluation of the individual patient: a Ophthalmol. 1983;67(1):8–16.
retrospective analysis. Ocul Immunol Inflamm. 52. Spencer WH. Ophthalmic pathology: an atlas and
2000;8(4):227–34. textbook. 4th ed. Philadelphia: W.B. Saunders; 1996.
35. Inomata H, Rao NA. Depigmented atrophic lesions 53. Mitchell P, Liew G, Gopinath B, Wong
in sunset glow fundi of Vogt-Koyanagi-Harada dis- TY. Age-related macular degeneration. Lancet.
ease. Am J Ophthalmol. 2001;131(5):607–14. 2018;392(10153):1147–59.
36. Nakayama M, Keino H, Okada AA, Watanabe 54. Anderson DH, Mullins RF, Hageman GS, et al. A
T, Taki W, Inoue M, et al. Enhanced depth imag- role of for local inflammation in the formation of dru-
ing optical coherence tomography of the chosen in the aging eye. Am J Ophthalmol. 2002;134(3):
roid in Vogt-Koyanagi-Harada disease. Retina. 411–31.
2012;32(10):2061–9. 55. Kim JH, Chang YS, Kim JW, Kim CG, Lee
37. Bacsal K, Wen DS, Chee SP. Concomitant choroidal DW. Age-related differences in the prevalence of
inflammation during anterior segment recurrence in subtypes of Neovascular age-related macular degen-
Vogt-Koyanagi-Harada disease. Am J Ophthalmol. eration in the first diagnosed eye. Graefes Arch Clin
2008;145(3):480–6. Exp Ophthalmol. 2019;257(5):891–8.
56. Told R, Sacu S, Hecht A, Baratsits M, Eibenberger choroidal neovascularisation in pathological myo-
K, Kroh ME, et al. Comparison of SD-optical coher- pia. Br J Ophthalmol. 2003;87(5):570–3.
ence tomography angiography and indocyanine 71. Yokoi T, Ohno-Matsui K. Diagnosis and Treatment
green angiography in type 1 and 2 neovascular age- of Myopic Maculopathy. Asia Pac J Ophthalmol.
related macular degeneration. Invest Ophthalmol Vis 2018;7(6):415–21.
Sci. 2018;59(6):2393–400. 72. Yoshida T, Ohno-Matsui K, Yasuzumi K, Kojima
57. D’Ambrosio E, Tortorella P, Iannetti L. Management A, Shimada N, Futagami S, et al. Myopic cho-
of uveitis-related choroidal neovascularization: roidal neovascularization: a 10-year follow-up.
from the pathogenesis to the therapy. J Ophthalmol. Ophthalmology. 2003;110(7):1297–305.
2014;2014:450428. 73. Nag TC, Kumari C. Electron microscopy of the
58. Yonekawa Y, Miller JW, Kim IK. Age-related macu- human choroid. In: Chhablani J, Ruiz-Medrano J,
lar degeneration: advances in management and diag- editors. Choroidal disorders. London: Academic
nosis. J Clin Med. 2015;4(2):343–59. Press; 2017. p. 7–20.
59. Wu PC, Chen YJ, Chen CH, Chen YH, Shin SJ, Yang 74. Akkaya S. Spectrum of pachychoroid diseases. Int
HJ, et al. Assessment of macular retinal thickness Ophthalmol. 2018;38(5):2239–46.
and volume in normal eyes and highly myopic eyes 75. Cheung CMG, Lee WK, Koizumi H, Dansingani K,
with third-generation optical coherence tomography. Lai TYY, Freund KB. Pachychoroid disease. Eye.
Eye. 2008;22(4):551–5. 2019;33(1):14–33.
60. Shimada N, Ohno-Matsui K, Harino S, Yoshida T, 76. Ersoz MG, Arf S, Hocaoglu M, Sayman Muslubas
Yasuzumi K, Kojima A, et al. Reduction of retinal I, Karacorlu M. Indocyanine green angiography
blood flow in high myopia. Graefes Arch Clin Exp of pachychoroid pigment epitheliopathy. Retina.
Ophthalmol. 2004;242(4):284–8. 2018;38(9):1668–74.
61. Shih YF, Fitzgerald ME, Norton TT, Gamlin 77. Pang CE, Freund KB. Pachychoroid neovasculopa-
PD, Hodos W, Reiner A. Reduction in choroidal thy. Retina. 2015;35(1):1–9.
blood flow occurs in chicks wearing goggles that 78. Dansingani KK, Balaratnasingam C, Klufas MA,
induce eye growth toward myopia. Curr Eye Res. Sarraf D, Freund KB. Optical coherence tomography
1993;12(3):219–27. angiography of shallow irregular pigment epithelial
62. Sayanagi K, Ikuno Y, Uematsu S, Nishida K. Features detachments in pachychoroid spectrum disease. Am
of the choriocapillaris in myopic maculopathy iden- J Ophthalmol. 2015;160(6):1243–54.e2.
tified by optical coherence tomography angiography. 79. Phasukkijwatana N, Freund KB, Dolz-Marco
Br J Ophthalmol. 2017;101(11):1524–9. R, Al-Sheikh M, Keane PA, Egan CA, et al.
63. Steidl SM, Pruett RC. Macular complications asso- Peripapillary pachychoroid syndrome. Retina.
ciated with posterior staphyloma. Am J Ophthalmol. 2018;38(9):1652–67.
1997;123(2):181–7. 80. Kumar A, Kumawat D, Sundar MD, Gagrani M,
64. Ohsugi H, Ikuno Y, Shoujou T, Oshima K, Ohsugi E, Gupta B, Roop P, et al. Polypoidal choroidal vascu-
Tabuchi H. Axial length changes in highly myopic lopathy: a comprehensive clinical update. Ther Adv
eyes and influence of myopic macular complications Ophthalmol. 2019;11:2515841419831152.
in Japanese adults. PLoS One. 2017;12(7):e0180851. 81. Goldhardt R, Rosen BS. Polypoidal choroidal
65. Dimitrova G, Tamaki Y, Kato S, Nagahara vasculopathy. Curr Ophthalmol Rep. 2019;7(1):
M. Retrobulbar circulation in myopic patients with 66–72.
or without myopic choroidal neovascularisation. Br 82. Engelbert M, Chew EY, Yannuzzi LA. Macular tel-
J Ophthalmol. 2002;86(7):771–3. angiectasia. In: Ryan SJ, editor. Retina. 2. London:
66. Cohen SY, Laroche A, Leguen Y, Soubrane G, Elsevier; 2013. p. 1050–7.
Coscas GJ. Etiology of choroidal neovascu- 83. Yannuzzi LA, Bardal AM, Freund KB, Chen KJ,
larization in young patients. Ophthalmology. Eandi CM, Blodi B. Idiopathic macular telangiecta-
1996;103(8):1241–4. sia. Arch Ophthalmol. 2006;124(4):450–60.
67. Hotchkiss ML, Fine SL. Pathologic myopia and 84. Dreyer R, Green WR. The pathology of angioid
choroidal neovascularization. Am J Ophthalmol. streaks: a study of twenty-one cases. Trans Pa Acad
1981;91(2):177–83. Ophthalmol Otolaryngol. 1978;31(2):158–67.
68. Fujiwara T, Imamura Y, Margolis R, Slakter JS, 85. Clarkson JG, Altman RD. Angioid streaks. Surv
Spaide RF. Enhanced depth imaging optical Ophthalmol. 1982;26(5):235–46.
coherence tomography of the choroid in highly myo- 86. Atmaca-Sonmez P, Heckenlively JR. Genetic disor-
pic eyes. Am J Ophthalmol. 2009;148(3):445–50. ders of the retina and optic nerve. In: Klintworth GK,
69. Ikuno Y, Jo Y, Hamasaki T, Tano Y. Ocular risk fac- Garner A, editors. Garner and Klintworth’s pathobi-
tors for choroidal neovascularization in pathologic ology of ocular disease. 3rd ed. New York: Informa
myopia. Invest Ophthalmol Vis Sci. 2010;51(7): Healthcare; 2008. p. 753–94.
3721–5. 87. Venkatesh R, Bavaharan B, Yadav NK. Predictors
70. Ohno-Matsui K, Yoshida T, Futagami S, Yasuzumi for choroidal neovascular membrane formation and
K, Shimada N, Kojima A, et al. Patchy atrophy and visual outcome following blunt ocular trauma. Ther
lacquer cracks predispose to the development of Adv Ophthalmol. 2019;11:2515841419852011.
88. Ament CS, Zacks DN, Lane AM, Krzystolik M, 97. Naumann G, Yanoff M, Zimmerman
D’Amico DJ, Mukai S, et al. Predictors of visual LE. Histogenesis of malignant melanomas of the
outcome and choroidal neovascular membrane uvea. I. Histopathologic characteristics of nevi of
formation after traumatic choroidal rupture. Arch the choroid and ciliary body. Arch Ophthalmol.
Ophthalmol. 2006;124(7):957–66. 1966;76(6):784–96.
89. Smiddy WE, Hernandez E. Histopathologic charac- 98. Naumann G, Zimmerman LE, Yanoff M. Visual
teristics of diode laser-induced chorioretinal adhe- field defect associated with choroidal nevus. Am J
sions for experimental retinal detachment in rabbit Ophthalmol. 1966;62(5):914–7.
eyes. Arch Ophthalmol. 1992;110(11):1630–3. 99. Muscat S, Parks S, Kemp E, Keating D. Secondary
90. Marshall J, Hamilton AM, Bird AC. Histopathology retinal changes associated with choroidal naevi
of ruby and argon laser lesions in monkey and human and melanomas documented by optical coherence
retina. A comparative study. Br J Ophthalmol. tomography. Br J Ophthalmol. 2004;88(1):120–4.
1975;59(11):610–30. 100. Ammon V. Anatomische Untersuchung von
91. Lehto KS, Tommila PV, Karma A. Choroidal oste- Coloboma bulbi. Z Ophthalmol. 1830;1830:I.
oma: clues to diagnosis. Acta Ophthalmol Scand. 101. Schubert HD. Structural organization of choroidal
2007;85(2):218–20. colobomas of young and adult patients and mecha-
92. Shields CL, Sun H, Demirci H, Shields JA. Factors nism of retinal detachment. Trans Am Ophthalmol
predictive of tumor growth, tumor decalcification, Soc. 2005;103:457–72.
choroidal neovascularization, and visual outcome in 102. Leff SR, Britton WA Jr, Brown GC, Lucier AC,
74 eyes with choroidal osteoma. Arch Ophthalmol. Brown JF. Retinochoroidal coloboma associ-
2005;123(12):1658–66. ated with subretinal neovascularization. Retina.
93. Gass JD, Guerry RK, Jack RL, Harris G. Choroidal 1985;5(3):154–6.
osteoma. Arch Ophthalmol. 1978;96(3):428–35. 103. Takenaka J, Yamane K, Minamoto A, Mishima
94. Williams AT, Font RL, Van Dyk HJ, Riekhof HK, Hayashida H. Subretinal neovascularization
FT. Osseous choristoma of the choroid simulating associated with retinochoroidal coloboma. Eur J
a choroidal melanoma. Association with a positive Ophthalmol. 2005;15(6):815–7.
32P test. Arch Ophthalmol. 1978;96(10):1874–7. 104. Spitzer M, Grisanti S, Bartz-Schmidt KU, Gelisken
95. Chien JL, Sioufi K, Surakiatchanukul T, Shields JA, F. Choroidal neovascularization in retinochoroi-
Shields CL. Choroidal nevus: a review of preva- dal coloboma: thermal laser treatment achieves
lence, features, genetics, risks, and outcomes. Curr long-term stabilization of visual acuity. Eye.
Opin Ophthalmol. 2017;28(3):228–37. 2006;20(8):969–72.
96. Jakobiec FA. Ocular anatomy, embryology, and tera-
tology. Philadelphia: Harper & Row; 1982.
cian in the National University of Córdoba, Argentina

(2004–2009), and completed residency training in
Ophthalmology at the Catholic University of Córdoba,
Argentina (2011–2014). She finished an Ocular Pathology
and Ocular Oncology fellowship at McGill University,
Montreal, Canada (2015–2017). Dr. Esposito was awarded
by the Argentinian Council of Ophthalmology (CAO) as a
Distinguished Young Ophthalmologist in 2014 and by the
McGill University Health Centre Foundation with the
Leonard Ellen Ocular Pathology Fellowship in 2015. She
also was awarded by the International Council of
Ophthalmology (ICO) with The David E I Pyott Master of
Surgery in Clinical Ophthalmology Scholarship (2017),
and finished the second year of the ChM in Clinical
Ophthalmology at the University of Edinburgh (2017–
2019) with a distiction. She participates actively in teach-
ing and research.
Evangelina Esposito , M.D., ChM., is an ophthalmolo-

gist from Argentina. Her fields are Ocular Pathology,
Ocular Oncology, and Uveitis. She graduated as a physi-
Julio A. Urrets-Zavalia, M.D., Ph.D., is a vitreo-retinal

subspecialist and Chairman of the Department of
Ophthalmology at the University Clinic Reina Fabiola,
Catholic University of Cordoba, Argentina, since 1991.
He obtained his medical degree from the Faculty of Pablo Zoroquiain is an Assistant Professor in the
Medical Sciences, National University of Cordoba, and Department of Pathology and Ophthalmology, at
his Ph.D. in Medicine from the Faculty of Health Pontificia Universidad Católica de Chile, Santiago, Chile.
Sciences of the Catholic University of Cordoba, Argentina, He graduated from Universidad de los Andes, School of
and performed a fellowship in Retina at Department of Medicine, and did a residency in Anatomical Pathology at
Ophthalmology, Hôpital de la Croix-Rousse, Université Pontificia Universidad Católica de Chile, Santiago, Chile.
Claude Bernard, Lyon, France. He completed an Ocular Pathology Clinical and Research
Dr. Urrets-Zavalia is profesor and Chair of Fellowship at McGill University and a Doctorate in
Ophthalmology Clinic, and director of the Postgraduate Ophthalmology and Visual Science at Universidad Federal
Degree in Ophthalmology at the Faculty of Health de Sao Paulo (UNIFESP). Currently, he is an Ocular
Sciences, Catholic University of Cordoba, Argentina. Pathologist, Cytopathologist, and the Director of the
Cytopathology Laboratory at UC-Christus Health Center.
He is author and co-author of over 80 publications includ-
ing, 50 peer-reviewed papers, 32 peer-reviewed abstracts,
and 2 book chapters. Dr. Zoroquiain is distinguished with
several national and international awards. He has served
as Guest Speaker at many conferences and symposia in
the United States, England, Thailand, Brazil, Peru, and
Chile.
Animal Models 4
Takayuki Baba
4.1 Introduction The ideal animal model needs to have the fol-
lowing characteristics; (1) similarity to human
The wet type of age-related macular degenera- CNV; (2) to clear ethical issues and to be justified
tion (AMD) is typically caused by choroidal neo- for experimental use; and (3) the affordability for
vascularization (CNV). The growing population multiple purposes. Unfortunately, the perfect ani-
of patients with wet AMD needs an effective mal model which can duplicate human CNV is
modality of treatment. The treatment burden for still not available. However, it is most important
patients and drug cost is still problematic that the CNV should be very close to human
although the extensive use of intravitreal anti- CNV when the disease pathogenesisi is investi-
vascular endothelial growth factor (VEGF) injec- gated. The ethical issue should be considered
tion enabled to treat wet AMD patients. To before starting experiments. For example, CNV
develop new drugs or treatment modality, an in primate models is expected to have a similar
experimental model using animals is quite useful. response to the treatments as human CNV does.
Since the CNV grows only in vivo, there is no However, the advantage of using a primate model
experimental model of CNV using cell or tissue is limited to the ability to test the CNV located at
culture. Therefore, the animal model which mim- macular. If the location of CNV does not matter
ics human CNV is critical to test various mole- for the experiment, the small animals such as
cules which have the potential to cure the CNV mice or rats are good to be used. The use of small
[1, 2]. The other pivotal role of an animal model animals is also beneficial to reduce the cost of
is to investigate the mechanism of disease pro- experiments. The price for animals themselves
gression [3]. The exact cause of the development but also the expenses for maintaining animals are
of CNV has not been determined yet although much less than keeping large animals and
there were numerous studies conducted to eluci- primates.
date the mechanism of disease. Since the tissues The history of an animal model of CNV
of active CNV are quite challenging to obtain, the started with the primate model of CNV created
investigation of an animal model is essential to induced by laser [4]. They used Argon laser and
understand the disease process. found the subretinal neovascularization in 4 in 5
eyes after disruption of Bruch’s membrane.
T. Baba (*) Interestingly, the incidence of CNV formation is
Department of Ophthalmology and Visual Science, higher in eyes with laser-induced branch retinal
Chiba University Graduate School of Medicine, vein occlusion in addition to the disruption of
Chiba, Japan Bruch’s membrane (9/11 eyes). It suggests the
e-mail: babatakayuki@nifty.com

38 T. Baba
substantial effect of ischemia and inflammation have macular and do have a different system of
which helps to grow CNV. The ischemic retina retinal vascular such as medullary ray from other
drives intraocular VEGF and seems to accelerate animals including mice, rats, and primates. The
the development of CNV. At that point, this size of the pig eye is similar to that of the human
model is very reasonable. eye and surgical manipulation including vitrec-
Based on Ryan’s CNV model, it has already tomy is available. The disadvantages of using pig
suggested that the experimental CNV requires eyes are that the pig eye does not have macular
several factors to develop. First, the compensa- and the expenses for obtaining and maintaining
tion of Bruch’s membrane is always necessary. animals is high. The primate models have advan-
This can be explained by the fact that the tages such as the large size of the eyeball, the
increased level of angiogenic factors is not existence of macular, and close approximately to
enough to cause the CNV [5, 6]. On the other the human retina. The primate eyes also used for
hand, the CNV formation is confirmed simply testing the drug delivery system and even used
after the disruption of Bruch’s membrane using for clinical trials proceed to the human trial. The
laser breaks or mechanical or chemical damage primate eye model is ideal for studying CNV but
of Bruch’s membrane [7]. For this instance, the there is an ethical issue of using primate eyes
size of CNV is relatively small suggesting the when eyes of small rodents are already available
damage of Bruch’s membrane is enough to grow as reliable models. The cost for animals itself and
CNV but not enough to extend to the large size. maintaining are also high.
To create a large volume of CNV, the growth fac-
tors such as VEGF play an important role. The
additional use of growth factor facilitates to 4.2 Animal Models of CNV
obtain large and long-acting CNV [8]. The local
inflammation and wound-healing process are 4.2.1 Laser-Induced Model
also necessary to grow the large CNV. The
inflammatory cytokines and growth factors can The first laser-induced CNV model was reported
be increased in such occasions and result in help- by Ryan et al. in 1979 [4]. They used rhesus mon-
ing the growth of the CNV. key and applied relatively strong Argon laser
The species mainly used for animal experi- such as 200–900 mW with a duration of 0.1 s.
ments are as follows: mice, rats, rabbits, pigs, and The spot size was small about 50–100 microme-
primates. The most commonly used are mice and ters and resulted in laser break of Bruch’s mem-
rats. The size of the eyeball is small in mice but it brane. The formation of CNV was confirmed in 4
is possible to create laser-induced CNV and sub- of 5 eyes with laser break only and 9 of 11 eyes
retinal injection with some difficulty. The avail- with a combination of laser break and retinal vein
ability of genetically engineered animals is an occlusion. Criswell et al. reported the character-
advantage of using mice models. On the other istics of laser CNV in squirrel monkeys [9]. They
hand, the size of the rat’s eyeball is larger than used 532 nm and 664 nm diode laser mounted
that of mice and enables subretinal injections slit-lump and fundus was visualized using a
more comfortable. The surgically induced CNV Goldmann-type contact lens. Nine to twelve pho-
models with subretinal injection of various mate- tocoagulations were placed as a grid pattern at
rials are available using rat. The low cost for the posterior fundus. Their laser setting was with
maintaining those small rodents is another advan- laser power 104–650 mW, size with 70–75
tage when the cost is compared with those for micrometers, duration 0.01–1.0 s. They found
large animals. The rabbit eye is much larger than 65% of lesions had CNV in squirrel monkey and
the eyes of mice and rats and is more easily 37% of lesions developed CNV in the macaque
manipulated. The longitude of CNV is greater monkey. Histologically, the localized CNV tissue
than the eyes of small rodents in the rabbit eyes. had the moderate to extensive thickening of cho-
The disadvantage of rabbit eye is that they do not riocapillaris layer. Diffuse fibrovascular tissues
4 Choroidal Neovascularization Animal Models 39
expanded beyond laser spot and invaded into the only faint tissue staining at 1–4 weeks postopera-
retina in 76% of CNV in squirrel monkeys and tively suggesting the proliferative retinal pigment
27% of macaque monkeys. They suggested that epithelial cells ensheathed the CNV.
the CNVs have characteristics which are unique Tobe et al. used the mouse model of laser-
to spieces and squirrel monkey has more diffuse induced CNV in 1998 [12]. The C57BL/6 J mice
CNV allows studying neovascularization away were used for the experiments. The laser was
from the trauma site. mounted on a slit-lump and a slide glass was used
The rat model of laser-induced CNV was first as a contact lens. They used Krypton laser with a
reported by Dobi et al. in 1989 [10]. They used spot size of 50 micrometers, 350–400 mW, and a
647 nm Krypton laser with a spot size of 100 duration of 0.05 s. They found CNV in 57% of
micrometers, 50–160 mW, and duration of 0.1 s. treated eyes, and the central bubble formation
The laser was mounted on the Zeiss-slit lump and which is the same as rat CNV is a good sign of
the cover-slip was served as a contact lens. Long– initiating the CNV formation. Four weeks after
Evans rats were used and the CNV was observed the laser, the dye leakage was confirmed in 39%
in 24–60% of eyes. One hundred and twenty of the area by fluorescein angiography.
megawatts of laser intensity was most reliable for Histological findings such as cellular tissue
inducing CNV in their settings. They also found extending from choroid to subretinal space
the adequate power delivered shows the central through the rupture of Bruch’s membrane and
bubble at the irradiated area and develops good pigment-laden macrophage-like cells and sur-
CNV. If the laser is too weak or too strong, the rounding retinal pigment epithelial cells were
CNV did not grow well. The excessively power- observed at 3 days after the laser. One week after
ful laser made avascular scars without any CNV the laser, newly formed vessels with wide lumen
formation. The fluorescein angiography showed from the choroid to subretinal space were found
staining at 1 day corresponding to the acute with pigment epithelial cells partly covering the
injury and this hyperfluorescence disappeared at new vessel.
1 week. The leakage of dye associated CNV was A pig model was used by Saishin et al. for
observed 2–6 weeks after the laser treatment. testing the effect of protein kinase inhibitors [13].
Sixty percent of the lesions presented the capil- Four-week-old young Yorkshire pigs were treated
laries originating within the choroid extended by diode laser after the intubation and general
through the disruption of Bruch’s membrane into anesthesia maintained with inhalation of 1–3%
the subretinal space histologically. The overlying halothane, 2% O2, and 1% NO2. The Bruch’s
retina degenerated as well as pigment epithelium, membrane rupture was made by laser with a
outer nuclear layer, and outer plexiform layer. power of 400 mW, size of 75 micrometers, and
The disrupted area of Bruch’s membrane had duration of 0.1 s. The minimal laser power was
pigment-laden cells, fibroblasts, and collagen determined when the vaporization bubble was
fibers. observed. They found CNV formation in 100%
The rabbit CNV model was reported in 1991 of eyes treated and it was confirmed by histologi-
by El Dirini et al. [11] They used subretinal irra- cally. To test the ability to cease the formation of
diation of laser using trans-vitreal glass pipette CNV via blocking VEGF, they measured the size
delivered into subretinal space. The argon laser of CNV by calculating the area of the GSA-lectin
power was 200–1100 mW and spot size was 100 labeled vascular tissue in each slide through the
micrometer with a duration of 0.05 s. They found laser spot.
no CNV formation in the eyes with low powers Overall, the laser-induced model was strongly
of 0.2–0.5 W but found CNV in 100% of lesions associated with the rupture of Bruch’s membrane
by histopathologically if the laser was applied and if there is no damage to the Bruch’s mem-
with a high intensity such as 1000 mW. The fluo- brane, there was no CNV formation. At that
rescein angiogram showed the hyperfluorescence point, this model is a wound-healing model and
at 1 and 3 days after the treatment but showed has a self-limited course. Although the mechanism
40 T. Baba
of developing CNV is different from the human Julien and Kreppel et al. injected high-capacity
CNV, the advantage of the laser-induce CNV adenoviral vector expressing VEGF-A165 into the
model should be evaluated. The simple technique subretinal space of Chinchilla bastard rabbits [8].
of using the laser, the reproducibility of CNV if The leakage from CNV was observed by scan-
the proper setting of laser is used, the short time- ning laser ophthalmoscopy with fluorescein and
course of CNV to develop are the merits of this indocyanine green at 2 and 4 weeks in 15 of 18
model. Multiple lesions created in one eye eyes (83%). Pathological subretinal vessel for-
enables the multiple evaluations of treatment mation was confirmed with activated endothelial
modalities and contributes to increasing the effi- cells and macrophages. They demonstrated the
cacy of trials. elevation of basic fibrous growth factor as well as
VEGF. Cui et al. used VEGF-impregnated gelatin
microspheres injected intravitreally into the sub-
4.3 Surgically Induced Model retinal space of Macaca mulatta monkeys [16].
The CNV developed in 12 of 13 eyes (92%) and
The formation is CNV can be obtained by inject- the leakage from CNV was confirmed by fluores-
ing various materials into the subretinal space. cein angiography from 2 to 12 weeks after the
For the technical reason, larger animals are pre- injection. The immunohistochemical findings
ferred but there were numerous reports using showed the CD31-positive new vessels at
mouse and rat for the background of this model. 1–8 weeks and glial fibrillary acidic protein at 6
Subretinal injection is necessary to the process of weeks suggesting glial proliferation.
this model; the mechanically induced CNV mod- Macrophages engaged early point at day 3 but
els are surgically induced models of CNV. The not found on day 7 and after.
nature of these models varies based on the mate- Schmack and Berglin et al. injected polysty-
rials or drug injected subretinally. rene microbeads and retinal epithelial cells into
VEGF is a major player to develop neovascu- the subretinal space of C57BL6 mice [17]. The
larization. The induction of rat VEGF164 cDNA overall incidence of CNV was 94%. The CNV
using adenovirus vector which is injected in sub- thickness became maximal at 7 days post injec-
retinal space of rat caused CNV in 86% of eyes tion of C57BL6-derived RPE cells and polysty-
[14]. The overexpression of mRNA of VEGF was rene microbeads simultaneously. The activated
observed in retinal pigment epithelial cells by retinal pigment cells expressed not only VEGF
10–31 days postoperatively. The formation of but other chemotactic and inflammatory cyto-
CNV was confirmed by histologically at 80 days kines such as MCP-1, MMP-2, MMP-9 which
after the injection concomitant with the degener- lead to the development of CNV. The cell-2 defi-
ation of overlying retina. The severity of CNV cient mouse showed smaller CNV after subreti-
depended on the amount of vector. Baffi et al. nal injection suggested the recruitment of
used adenovirus vector coding VEGF165 to pro- macrophages in the process of CNV formation.
mote CNV by subretinal injection [15]. They Matrigel, a basement membrane extract, is
used Long–Evans rat and found the elevated used for solidifying tissue and it can stimulate
expression of VEGF at retinal pigment epithe- angiogenesis. Shen et al. used Matrigel to induce
lium 1 week after the injection. The formation of CNV by injecting subretinally [18]. They used
CNV was confirmed by fluorescein angiography, wildtype and Ccl2-deficient mouse and found the
histology using flatmount choroid, and transmis- CNV formation in 31% and 53% of eyes, respec-
sion electron microscope from 2 to 4 weeks post tively. Ccl2, a member of MCP of CC-chemokine
injection. Although their concentration of adeno- is known to have a protective role against
virus coding VEGF165 was much smaller than that AMD. In their study, the CNV developed in 31%
used in Spilsbury’s study, the mitogenic effect on of eyes in wildtype and 53% of eyes in the Ccl2-
vascular endothelial tissue was still significant. deficient group suggesting the involvement of
macrophages and chemotaxis in the formation of at 1 week but leakage was evident at 2 and 4
CNV. Qiu et al. used pigmented rabbit and weeks postoperatively. After the injection of
injected Matrigel with or without a various con- LHP, Armstrong et al. found that the level of
centration of VEGF [19]. They found CNV in VEGF, transforming growth factor-b, and
100% of eyes injected Matrigel and none of the platelet-derived growth factor peaked at 24 hours
eyes injected PBS. The CNV developed and sustained at the high level throughout the fol-
2–4 weeks after injection and lasted at 9 weeks lowing 3 weeks in the postoperative period [22].
after that. The activity of CNV was confirmed by Those angiogenic cytokines play roles to develop
fluorescein angiography and the CNV was histo- CNV after the injection of LHP.
logically confirmed. They also performed optical We used Sprague–Dawley rats to develop a
coherence tomography to reveal the subretinal lipid-induced CNV model [23]. 13(S)-Hydro
fluid suggesting the leaky vessel within the peroxy-9Z,11E-octadecadienoic acid (HpODE)
lesion. is one of several oxidized lipids, which exist
It resembles the laser rupture of Bruch’s mem- under normal conditions in cell membranes of
brane that the perforation of Bruch’s membrane the human body [24]. We injected 13(S)-HPODE
by a needle developed CNV in rabbit eyes [7]. into the subretinal space using a glass needle and
The subretinal neovascularization was observed a pico-injector. The formation of CNV was con-
arising from the site of the perforation of Bruch’s firmed by histology including JB-4 embedded
membrane in the eyes of pigmented rabbits. The thin sections, immunohistochemistry with a con-
injury was made by a scalpel blade trans-sclerally. focal microscope, and transmission electron
The degeneration of photoreceptors and prolifera- microscope. The course of CNV development
tion of retinal pigment epithelium were also was observed by fluorescein angiography.
observed. The neovascularization extended Typically, some small hyperfluorescent spots
beyond the scar of Bruch’s membrane. Lassota were found at 1 week and more intense leakage
et al. used 12-week-old domestic pigs and com- was observed at 2 weeks and the leakage was
pared the techniques of perforating Bruch’s mem- most prominent at 3–4 weeks and decreased at 5
brane transvitreally with and without removal of weeks suggesting the reduced or matured CNV
the retinal pigment epithelium [20]. They found (Fig. 4.1). We found the CNV in 6 of 7 eyes
CNV formation with both techniques at 14 days (85.7%) which were injected 30 microgram of
after the injection but the CNV was more promi- HPODE in 2 microliters of borate buffer
nent in eyes with Bruch’s rupture without retinal (Fig. 4.2). Lipids were confirmed at inside of
pigment epithelial removal. They suggested that CNV and surrounding retina, and ED1 (CD68)
the VEGF secreted by retinal pigment epithelium positive circulating monocyte and ED2 (CD163)
was reduced after the removal of cells and resulted positive resident macrophage were recruited in
in reduced stimuli to grow CNV. This model does the CNV (Fig. 4.3). We suggested the 13(S)
not require any special agents or drugs to promote HPODE-stimulated inflammatory cells or retinal
CNV growth. At that point, this is a good simple pigment epithelial cells degraded the Bruch’s
model but is eventually a wound-healing process membrane by tissue protease and stimulated
and the lifetime of CNV is unknown. endothelial cells by secreting growth factors such
Tamai et al. injected 12.5–25 micrograms of as VEGF and eventually developed CNV. Since
linoleic acid hydroperoxide (18:2/LHP) into the HPODE exists in Bruch’s membrane naturally
subretinal space of rabbits and found the CNV in and was found more in the eyes of aged human
11 of 14 eyes (46%) histologically [21]. When and eyes with age-related macular degeneration
they injected 150–200 micrograms of LHP, [24], this model using subretinal HPODE is a
severe retinal and choroidal atrophy was good model that mimics human CNV with high
observed, but no CNV developed. The fluores- reproducibility and ideal time course to evaluate
cein angiography showed no hyper fluorescence therapeutic modalities.
42 T. Baba
a b
c d
Fig. 4.1 Fluorescein angiographic image of the eye after observed at 1 week (arrow). Greater leakage was observed
subretinal injection of 30 micrograms of HpODE. The at 2 weeks (b, arrow) and was greatest at 3 weeks (c,
timing in each angiogram was about one and a half min- arrow). (d) The leakage area decreased at 5 weeks (arrow).
utes after the injection of dye. (a) A small dye leakage was (From Baba T et al., 2010 [23])
4.4 Genetically Engineered and the accumulation of C5a and IgG suggested
Model the source of VEGF.
The Ccl2/Cx3cr1 deficient double knockout
The Ccr2/Ccl2 deficient mouse model was devel- mouse was developed by Chan et al. [27] Their
oped by Ambati et al. [25, 26]. Ccr2/Ccl2 defi- mouse had increased N-retinylidene-N-
cient transgenic mouse fails to recruit retinylethanolamine (A2E) and decreased DRp20
macrophages to the area of the RPE and Bruch’s levels in the RPE. These mice had the changes of
membrane because those chemotactic cytokines retinal pigment epithelium and drusen-like
and receptors control inflammatory cells. They lesions if the mouse fed a diet low in omega-3
found 25% of the eyes of their transgenic animals polyunsaturated fatty acids (PUFAs). Ccl2 is
a b
c d
Fig. 4.2 Images of HpODE-induced CNV. GSA-lectin (b) and 15 micrograms (c) of HpODE caused CNV (arrows);
visualized the neovascularization in flatmount choroid at Relatively small CNV was observed. (d) Thirty micrograms
3 weeks after injection. (a) No CNV was observed after one of HpODE caused significantly large CNV (arrow). Scale
microgram of HpODE injection. (b and c) Five micrograms bar: 500 micrometers. (From Baba T et al., [23])
necessary to recruit monocyte and CX3CR1 major scavenger of oxidative stress. In 3 of 30

moves circulating monocytes to tissues and animals (10%) which is older than 10 months,
makes them into resident macrophages and den- they found the CNV histologically. In the ani-
dritic cells [28]. In these mice, CNV was found mals that had a CNV formation, the fibrovascular
histologically in 15% of animals in 6 weeks. tissue extended through the rupture of Bruch’s
The Cu, Zn-superoxide dismutase (SOD1) membrane. The drusen formation and thickened
deficient mice had funduscopically and histologi- Bruch’s membrane were commonly observed in
cally evident CNV in 8.3% and 10% of animals the older mice. Therefore, the senecient
[29]. One of the antioxidative systems in the eye SOD1−/− mice had characteristics of dry and
is Cu, Zn-SOD1 in the cytosol and the most wet human AMD.
44 T. Baba
PAS Oil Red O vWf ED1

a b c d
Control
3 Days
e f g h
3 Days
i j k l
1 Wk
HpODE
m n o p
2 Wk
q r s t
3 Wk
Fig. 4.3 Histological images after buffer (A–D) or paired arrows). (k) Some vWF structures suggesting reti-
HpODE (E–T) injection. PAS-positive RPE cells were nal capillaries are observed in the inner nuclear layer. (m)
observed at 3 days after buffer injection (a, paired arrows). At 2 weeks after the injection of HpODE, the degenerated
The RPE cells contained no lipids (b, paired arrows). photoreceptors were observed. (m and n) In the subretinal
There was no labelling by vWF (c, paired arrows) or ED1 space, PAS-positive, lipid-laden cells were observed
of the cells (d, paired arrows) in the subretinal space. (e) (paired arrows). (o) There is a vWF structure suggesting a
At 3 days after the HpODE injection, PAS-positive cells deep retinal capillary in the inner nuclear layer. (p) There
were abundant (paired arrows) in the space between retina were some subretinal ED1-positive cells (paired arrows).
and RPE. (f) These cells contained lipids (paired arrows) (s) By 3 weeks, a subretinal CNV formation was clearly
and soluble vWF (g, paired arrows) was observed around labeled with vWF (open arrows). PAS-positive material
the ED1-positive cells (h, paired arrows). One week after and cells are observed in subretinal space (q, paired
the injection of HpODE there were PAS-positive cells on arrows), lipids (r, paired arrows), and some ED1-labeled
Bruch’s membrane (i, paired arrows) with lipids (j, paired cells (t, paired arrows) above the CNV. Scale bar 30
arrows). There were ED1-positive cells among them (L, micrometers. (From Baba T et al., [23])
The nuclear factor erythroid-2 related factor 2 sition and accumulation of lipofuscin, deposition
(Nrf2) is important for cell response to oxidative of sub pigment epithelial inflammatory protein
stress and toxification. The Nrf2 deficient mouse [30]. The CNV was confirmed using histology in
had an age-dependent degenerative change of 3 of 17 eyes (18%) of animals with age
retinal pigment epithelium and drusen-like depo- 11–17 months. The eyes with CNV also presented
subretinal hemorrhage and exudate which are References

typical findings of human CNV.
1. Grossniklaus HE, Kang SJ, Berglin L. Animal mod-
els of choroidal and retinal neovascularization. Prog
Retin Eye Res. 2010;29:500–19.
4.5 Conclusion 2. Pennesi ME, Neuringer M, Courtney RJ. Animal
models of age related macular degeneration. Mol Asp
Excellent quality of experimental models needs Med. 2012;33:487–509.
3. Campochiaro PA. Molecular pathogenesis of retinal
the comparability to human disease and repro- and choroidal vascular diseases. Prog Retin Eye Res.
ducibility for repeating experiments. Also, the 2015;49:67–81.
affordability of animal is vital for continuing the 4. Ryan SJ. The development of an experimental
project and clearing ethical problems is also nec- model of subretinal neovascularization in disciform
macular degeneration. Trans Am Ophthalmol Soc.
essary to justify the use of animals for research 1979;77:707–45.
for a human being. Therefore, the perfect animal 5. Ohno-Matsui K, Hirose A, Yamamoto S, et al.
model is desirable, but it is quite challenging to Inducible expression of vascular endothelial growth
obtain. In terms of the CNV model for wet AMD, factor in adult mice causes severe proliferative
retinopathy and retinal detachment. Am J Pathol.
there is a variety of experimental models. Three 2002;160:711–9.
major approaches to develop CNV are laser, sur- 6. Oshima Y, Oshima S, Nambu H, et al. Increased
gery, and genetic manipulation. Each option can expression of VEGF in retinal pigmented epithelial
be used to small to large animals except for cells is not sufficient to cause choroidal neovascular-
ization. J Cell Physiol. 2004;201:393–400.
genetic engineering is mainly for mice. It is 7. Hsu HT, Goodnight R, Ryan SJ. Subretinal choroidal
important to use the proper model suitable for neovascularization as a response to penetrating reti-
what to see through the experiments. Because the nal injury in the pigmented rabbit. Jpn J Ophthalmol.
response to drug is different from human CNV, 1989;33:358–66.
8. Julien S, Kreppel F, Beck S, et al. A reproducible and
e.g., less response of rodent CNV to bevacizumab quantifiable model of choroidal neovascularization
and ranibizumab, great caution should be paid to induced by VEGF A165 after subretinal adenoviral
test the treatment modalities [31]. gene transfer in the rabbit. Mol Vis. 2008;14:1358–72.
9. Criswell MH, Ciulla TA, Hill TE, et al. The squir-
rel monkey: characterization of a new-world primate
Key Learning Messages model of experimental Choroidal neovascularization
• There are three major types of experimental and comparison with the macaque. Invest Ophthalmol
CNV models. Vis Sci. 2004;45:625–34.
• A laser-induced CNV model is reproducible 10. Dobi ET, Puliafito CA, Destro M. A new model of
experimental choroidal neovascularization in the rat.
but the mechanism is a wound-healing process Arch Ophthalmol. 1989;107:264–9.
which has a relatively short time course. 11. elDirini AA, Ogden TE, Ryan SJ. Subretinal endo-
• A surgically induced CNV model is more photocoagulation. A new model of subretinal neovas-
physiological but technically challenging in cularization in the rabbit. Retina. 1991;11:244–9.
12. Tobe T, Ortega S, Luna JD, et al. Targeted disrup-
some models using small rodents. tion of the FGF2 gene does not prevent choroidal
• Some of the genetically engineered models neovascularization in a murine model. Am J Pathol.
mimic human disease very well but it takes a 1998;153:1641–6.
relatively long time to develop CNV. 13. Saishin Y, Silva RL, Saishin Y, et al. Periocular injec-
tion of microspheres containing PKC412 inhibits
• Models using small rodents are affordable and Choroidal neovascularization in a porcine model.
reproducible but do not have macular. Invest Ophthalmol Vis Sci. 2003;44:4989–93.
• Models using large animals including pri- 14. Spilsbury K, Garrett KL, Shen WY, et al.
mates have high comparability to human CNV Overexpression of vascular endothelial growth factor
(VEGF) in the retinal pigment epithelium leads to the
but have ethical and financial issues. development of choroidal neovascularization. Am J
• The selection of models varies based on the Pathol. 2000;157:135–44.
purpose of investigation or trial for treatment 15. Baffi J, Byrnes G, Chan CC, Csaky KG. Choroidal
modalities. neovascularization in the rat induced by adenovirus
46 T. Baba
mediated expression of vascular endothelial growth 23. Baba T, Bhutto IA, Merges C, et al. A rat model
factor. Invest Ophthalmol Vis Sci. 2000;41:3582–9. for choroidal neovascularization using subreti-
16. Cui JZ, Kimura H, Spee C, et al. Natural history of nal lipid hydroperoxide injection. Am J Pathol.
choroidal neovascularization induced by vascular 2010;176:3085–97.
endothelial growth factor in the primate. Graefes Arch 24. Spaide RF, Ho-Spaide WC, Browne RW, Armstrong
Clin Exp Ophthalmol. 2000;238:326–33. D. Characterization of peroxidized lipids in Bruch’s
17. Schmack I, Berglin L, Nie X, et al. Modulation of membrane. Retina. 1999;19:141–7.
choroidal neovascularization by subretinal injection 25. Ambati J, Anand A, Fernandez S, et al. An ani-
of retinal pigment epithelium and polystyrene micro- mal model of age-related macular degeneration in
beads. Mol Vis. 2009;15:146–61. senescent Ccl-2- or Ccr-2-deficient mice. Nat Med.
18. Shen D, Wen R, Tuo J, et al. Exacerbation of reti- 2003;9:1390–7.
nal degeneration and choroidal neovasculariza- 26. Takeda A, Baffi JZ, Kleinman ME, et al. CCR3 is a
tion induced by subretinal injection of Matrigel target for age-related macular degeneration diagnosis
in CCL2/MCP-1-deficient mice. Ophthalmic Res. and therapy. Nature. 2009;460:225–30.
2006;38:71–3. 27. Chan CC, Ross RJ, Shen D, et al. Ccl2/Cx3cr1-
19. Qiu G, Stewart JM, Sadda S, et al. A new model of deficient mice: an animal model for age-related macu-
experimental subretinal neovascularization in the rab- lar degeneration. Ophthalmic Res. 2008;40:124–8.
bit. Exp Eye Res. 2006;83:141–52. 28. Geissmann F, Jung S, Littman DR. Blood monocytes
20. Lassota N, Kiilgaard JF, Prause JU, et al. Surgical consist of two principal subsets with distinct migra-
induction of choroidal neovascularization in a por- tory properties. Immunity. 2003;19:71–82.
cine model. Graefes Arch Clin Exp Ophthalmol. 29. Imamura Y, Noda S, Hashizume K, et al. Drusen, cho-
2007;245:1189–98. roidal neovascularization, and retinal pigment epithe-
21. Tamai K, Spaide RF, Ellis EA, et al. Lipid hydroper- lium dysfunction in SOD1-deficient mice: a model of
oxide stimulates subretinal choroidal neovasculariza- age-related macular degeneration. Proc Natl Acad Sci
tion in the rabbit. Exp Eye Res. 2002;74:301–8. U S A. 2006;103:11282–7.
22. Armstrong D, Ueda T, Ueda T, et al. Lipid hydro- 30. Zhao Z, Chen Y, Wang J, et al. Age-related retinopathy
peroxide stimulates retinal neovascularization in in NRF2-deficient mice. PLoS One. 2011;6:e19456.
rabbit retina through expression of tumor necrosis 31. Lu F, Adelman RA. Are intravitreal bevacizumab and
factor-alpha, vascular endothelial growth factor and ranibizumab effective in a rat model of choroidal neo-
platelet-derived growth factor. Angiogenesis. 1998;2: vascularization? Graefes Arch Clin Exp Ophthalmol.
93–104. 2009;247:171–7.
Electron Microscopy
of the Choroidal 5
Neovascularization
Tapas C. Nag and Sneha Gupta
5.1 Introduction The initial course of the disease, called dry

age-related macular degeneration (accounting for
Age-related macular degeneration is a serious 90% of all cases), begins with the formation and
ocular problem of the elderly (>50 years), espe- deposition of some yellowish materials, called
cially those living in the advanced countries. In drusen, in the macula with aging. Electron
this disease, the capillary layer of the choroid, microscopy has confirmed these materials as
called the choriocapillaris [1–3] is critically degenerative tissue remnants that most likely
affected, along with Bruch’s membrane, a collag- arise from degenerated retinal pigment epithe-
enous layer that separates the retina from it [4], lium, besides other undefined sources. There is a
and the outer retina comprising the retinal pig- limitation in visual capabilities (e.g., difficulty in
ment epithelium and photoreceptor cells [5–9]. reading) in the affected individuals, though it is
Aging is considered a risk factor that initiates the not severe at this early stage with dry age-related
development of age-related macular degenera- macular degeneration. The situation progresses
tion. There are gradual alterations in the contents slowly in untreated cases (and if the diet lacks
of retinal pigment epithelium, Bruch’s membrane carotenoids and vitamins) and over a period of
(BM), and the choriocapillaris with progressive years, there is substantial death of macular photo-
aging, resulting in the death of capillaries. Due to receptors and retinal pigment epithelium. The
some unknown situations, these changes culmi- loss of the retinal pigment epithelium may be
nate in the disease manifestations, wherein exces- extensive in another clinical condition, called
sive tissue changes and capillary loss result in the geographic atrophy [7, 9, 10]. In rather unfavor-
diminished supply of oxygen and vital nutrients able situations in about 10% of the total victims,
to the retinal pigment epithelium. As a result, the the dry age-related macular degeneration may
retinal pigment epithelium becomes hypofunc- take a sight threatening form, called wet or exu-
tional, and gradually photoreceptor cells die from dative age-related macular degeneration.
the macula. The latter is ultimately responsible Choroidal neovascularization (CNV) is clinically
for vision loss, affecting the regular activities and a devastating event in wet age-related macular
lifestyle patterns in the elderly. degeneration. This chapter deals with the elec-
tron microscope data of CNV reported in affected
individuals secondary to age-related macular
T. C. Nag (*) · S. Gupta degeneration.
Neurobiology Laboratory, Department of Anatomy,
All India Institute of Medical Sciences,
New Delhi, India

48 T. C. Nag and S. Gupta
5.1.1 Choroidal Neovascularization retinal pigment epithelium (subretinal pigment

epithelium pattern; Figs. 5.2 and 5.3); (2) type 2,
Figure 5.1 demonstrates the capillary layer of the when vessel growth is reached between the reti-
choriocapillaris in a normal, aged human donor. nal pigment epithelium and retina (subretinal pat-
In wet age-related macular degeneration, CNV tern); and (3) the combined pattern, when
results in abnormal development of vessels from subretinal pigment epithelium new vessels extend
the choriocapillaris. It has two well-recognized into the subretinal space [11–13].
categories: (1) occult CNV, when the growth of
vessels is not marked, and there is minimal leak-
age from those vessels, and (2) classical CNV, 5.2 Risk Factors for CNV
when the growth of vessels is prominent beneath
the retina, which leads to more leakage of blood In wet age-related macular degeneration, the
and severe loss of central vision. Depending on abnormal development of new vessels is linked
the site and spreading of new vessels, CNV can with certain altered histophysiological altera-
be categorized into three types: (1) type 1, when tions of the BM and retinal pigment epithelium.
CNV is limited predominantly between BM and These changes are likely to be the risk factors
Fig. 5.1 Light

micrograph of toluidine
blue-stained choroid RPE
from a normal 64-year-
old donor, showing
capillaries (arrows) of
the choriocapillaris
layer, from which
pathological
neovascularization
begins in wet AMD
50 µm
Fig. 5.2 Light

micrograph showing a
capillary lying in the
RPE layer (arrowhead)
of 89-year-old donor
retina. Bruch’s
membrane (BM) and
RPE
choriocapillaris (CC) are
indicated
BM
CC
50 µm
5 Electron Microscopy of the Choroidal Neovascularization 49
Fig. 5.3 Diagram showing the formation of a new vessel (arrow) between Bruch’s membrane (BM) and RPE (sub-RPE
pattern). The vessel originates from the choriocapillaris (CC)
for CNV. Currently, the identified major risk fusion barrier for the retinal pigment epithelium
factors are: [16, 34], thereby depriving it from receiving oxy-
gen and water-soluble substances from the cho-
1. Drusen characteristics roidal circulation.
2. Calcification and disintegration of BM
3. Focal hyperpigmentation of the macular RPE
4. Accumulation of polar phospholipids in BM 5.2.2 Focal Macular
5. Choriocapillaris loss Hyperpigmentation
of the Retinal Pigment
The significance of these individual factors Epithelium
predisposing CNV is analyzed below from the
published work in this field. The reports of Bressler et al. [14], Abdelsalam
et al. [19], Wang et al. [18], and CAPT Research
Group [35] asserted that focal macular hyperpig-
5.2.1 ole of Drusen and Basal
R mentation is linked with an increased risk of
Deposits in CNV emergent CNV.
The unusual build-up of drusen in the macula is a

characteristic feature for age-related macular 5.3 Calcification of Bruch’s
degeneration [14–18]. Drusen occupy two Membrane
favored sites: (1) in spaces created between the
retinal pigment epithelium basal and plasma Histological examination of eyes with different
membrane, and (2) in BM, especially between stages of age-related macular degeneration [36]
the retinal pigment epithelium basal membrane revealed high grade of calcification and disinte-
and inner collagenous layer. The histological and gration of BM in exudative age-related macular
histochemical characteristics of drusen are to degeneration, unlike in nonexudative cases.
specific forms of age-related macular degenera- These changes imply that calcification as well as
tion and CNV pathogenesis. Soft, diffuse, large, fragmentation of BM may be involved in CNV,
and hyperfluorescent drusen, which originate especially by helping ingrowth of choroidal neo-
from the degeneration of the retinal pigment epi- vascular membranes [33, 36–38].
thelium, are considered a major risk factor for the
progress of CNV [7, 17, 19–27]. Besides drusen,
other amorphous and fibrillar deposits are associ- 5.3.1 ole of Phospholipids in CNV
R
ated with CNV (frank as well as occult) second- Growth
ary to the pathogenesis of age-related macular
degeneration [8, 24, 25, 28–33]. The slow, diffuse The soft, large, and diffuse drusen of the macula
thickening of BM and progressive accumulation contain predominantly polar phospholipids and
of basal linear deposits in this layer create a dif- are found to be associated with the development
of CNV [23, 24]. Lommatzsch et al. [31] showed vascular endothelial growth factor by retinal pig-
by electron microscopy the presence of vesicles ment epithelium, stimulating neovascularization
and droplets in basal deposits that stained from preexisting choriocapillaris.
intensely for phospholipids. The phospholipid
content most likely undergoes oxidative damage
and the lipid deposition in association with 5.4.1 The Process of CNV in General
changes in BM are believed to induce inflamma-
tory reactions, resulting in the ingrowth of cho- The early features of CNV in wet age-related
roidal new vessels beneath the retinal pigment macular degeneration were described by examin-
epithelium [23]. Experimentally, in rodent mod- ing tissues from patients with an advanced form
els, subretinal injection of oxidized lipids can of the disease. Neovascularization appears to
induce CNV [39, 40] by inducing increased proceed from preexisting choroidal capillaries
expressions of genes involved in inflammatory and venules [24, 44–46]. This is also evident in
cascade [26]. The latter can be inhibited by experimental studies using laser photocoagula-
blocking monocyte chemoattractant protein-1, a tion in monkeys [47] and in newly formed sub-
potent molecule involved in inflammation [40]. retinal vessels after photocoagulation,
Finally, the chemical modification of BM likely demonstrating features that are essentially pres-
generates a furrow between the inner collagenous ent in choroidal capillaries [48]. In CNV, initially
layer and the basal lamina of the retinal pigment there is budding and sprouting of endothelial
epithelium through which new vessels can grow. cells (Fig. 5.4) and enlargement of pericytes, the
latter appear to surround endothelial cell sprouts
in neovascular development [45]. These vessels
5.4 Choriocapillaris Loss enter the inner collagenous layer of BM and
extend into BM and basal laminar deposits,
There is clear evidence for CNV development in spreading over the subretinal pigment epithelial
response to choriocapillaris loss. Histological zone [8]. Killingsworth [45] described two
studies revealed degenerative capillaries to be phases of neovascular growth: (1) the initial
frequent in cases with exudative age-related mac- phase, being restricted within the choroid, is low
ular degeneration and loss of capillaries was in turnover of vessels and these vessels penetrate
attended by neovascularization [41–43]. The reti- BM. This is followed by a “high-turnover” phase,
nal pigment epithelium then suffers from showing extensive proliferation of endothelial
hypoxia, a condition that favors the release of cells and migration and invasion of new vessels
Fig. 5.4 A capillary of

the choriocapillaris
(CC), showing slender, BM
endothelial sprouts
(arrows), passing
through the basal lamina
into the perivascular
matrix. BM, Bruch’s
EN
membrane; EN,
endothelial cell.
CC
Reprinted from
Killingsworth (Graefe’s
Arch Clin Exp EN
Ophthalmol 1995; 233:
313–323), with
permission from
1 µm
Springer-Verlag
in the subretinal pigment epithelial zone. neovascularization [46, 51]. The new vessels of
Hemorrhage and lipid exudates in the subretinal the choriocapillaris can penetrate BM in the
or subretinal pigment epithelial zones are consid- absence of a prior break [52], or the activated leu-
ered to be the earliest sign of CNV symptoms. kocytes and macrophages may cause focal thin-
The new vessels are leaky, and clinically this can ning and break the integrity of BM [50, 53],
be identified from hyperfluorescence in fluores- favoring neovascular growth in age-related mac-
cein angiography, indicating leakage of fluores- ular degeneration [54]. This is evident from the
cein from the new vessels [24, 48]. Excessive appearance of macrophages and lymphocytes
leakage of serum fluid can lead to the formation near the region of breaks of BM (Fig. 5.5), sug-
of disciform scars that trigger photoreceptor cell gesting their involvement in the growth of new
death. vessels. Matrix metalloproteinases, collagenase,
and elastase released by them disrupts the integ-
rity of BM, allowing the incursion of new vessels
5.4.2 Role of Macrophages into the subretinal pigment epithelial to subreti-
and Leukocytes in CNV nal space [55–57]. Nishimura et al. [58] reported
the presence of activated macrophages in experi-
The spatiotemporal distribution of macrophages mental subretinal neovascularization, implying
and leukocytes in areas of neovascularization them to play a potential angiogenic role [59].
(Figs. 5.5 and 5.6) led several authors to imply Other cellular phenotypes, namely fibrocytes,
the involvement of these cells in the development mast cells, monocytes, lymphocytes, and colla-
of neovascular age-related macular degeneration gen and fibrin as extracellular materials have
[46, 49–51]. These cells, upon activation, can lib- been found in and around the new vessels [13, 29,
erate vasoproliferative agents as part of inflam- 32, 46, 49, 53, 60, 61]. The cellular elements
matory response, thereby inducing noted are possibly involved in the formation of
PC
EL
V C
L
Fig. 5.5 Electron micrograph showing a capillary (C) diately above Bruch’s membrane. EL, elastic layer of
and a venule (V) ingress the sub-RPE area via a break in Bruch’s membrane. Final magnification x1,226.4.
Bruch’s membrane (B). Leukocytes (L) lie closely Reprinted from Penfold et al. (Graefe’s Arch Clin Exp
adhered to Bruch’s membrane and on the retinal side of Ophthalmol 1987; 225:70–76), with permission from
Bruch’s membrane. A pigment clump (PC) is seen imme- Springer-Verlag
Fig. 5.6 Electron

micrograph depicting L
the accumulation of L
lymphocytes (L)
adherent to the
abluminal surface of a
subretinal vessel.
Arrows denote tight
junctions between
apposing endothelial L
cells that lack fenestrae.
Reprinted from Penfold
et al. (Prog Ret Eye Res
2001; 20:385–414, with
permission from
Elsevier)
L
L
L 10 µm
breaks in BM in age-related macular degenera- also antiangiogenic molecules like pigment

tion [46, 53], favoring CNV growth and epithelial-
derived factor from retinal pigment
progression [62, 63], while fibrin is suggested to epithelium [71] that can arrest CNV progression
act as a scaffold for the growth of new vessels [72]. A balance between the secretory levels of
[49]. There is evidence for the dissolution of elas- vascular endothelial growth factor and pigment
tic lamina of BM in CNV, allowing new vessels epithelial-derived factor therefore determines the
to grow toward the retinal pigment epithelium. course of CNV pathogenesis [73].
Also, experimental studies hint to the possible
involvement of abnormal elastin metabolism in
the formation of neovascular membrane in age- 5.5 Conclusions
related macular degeneration [64]. In addition,
leukocytes are implicated in the formation as Although extensive research has been carried out
well as exudation from new vessels [46]. Thus, it in the last two decades, the pathological basis of
is likely that inflammation plays a critical role in CNV development in wet age-related macular
CNV pathogenesis in age-related macular degen- degeneration remains poorly understood.
eration [31, 54, 65]. Therapeutic measures to combat CNV can be
useful only when we have an unambiguous
understanding of the process. Earlier, the Eye
5.4.3 Role of RPE in CNV Disease Case–Control Study Group [74] in its
preliminary report pointed out that carotenoid-
The retinal pigment epithelium is vital for the based antioxidant supplementation could retard
regular maintenance of the choriocapillaris [66] neovascular growth in age-related macular
via the release of trophic substances. However, it degeneration, an approach that needs additional
plays a role in CNV by releasing an excess of studies for evaluation. Second, because numer-
certain angiogenic cytokines [56], e.g., VEGF ous reports have linked the presence of inflam-
[42, 43, 49, 67]. The presence of VEGF immuno- matory cells with CNV pathogenesis,
reactivity in the neovascular membrane has been experimental studies to decipher their precise
shown [32, 68, 69]. Other factors, e.g., fibroblast contributory role (friends or foe? [75]) in this dis-
growth factors, released by injured RPE could ease and therapeutic interventions to suppress/
also be involved in this process [70]. There are modulate the activities of local inflammatory
cells can be considered. Further, inhibition of 9. Zarbin MA. Current concepts in the pathogenesis of
age related macular degeneration. Arch Ophthalmol.
angiogenesis proper, in a strict sense, seems a 2004;122:598–614.
promising plan for the management of CNV. For 10. McLeod DS, Taomoto M, Otsuji T, Green WR,
example, inhibitors of metalloproteinases, e.g., Sunness JS, Lutty GA. Quantifying changes in RPE
TIMP-3 [76], have been shown to offer good pro- and choroidal vasculature in eyes with age-related
macular degeneration. Invest Ophthalmol Vis Sci.
tection. The precise regulation of secretion of 2002;43:1986–93.
pigment epithelial-derived factor in CNV is not 11. Gass JD. Biomicroscopic and histopathologic con-
clear; it is known that its loss could be a problem siderations regarding the feasibility of surgical exci-
for ischemia-induced neovascularization [77]. sion of subfoveal neovascular membranes. Am J
Ophthalmol. 1994;118:285–98.
Paradoxically, its high level is reported to increase 12. Grossniklaus HE, Gass JDM. Clinicopathologic
CNV development [78], a fact that just repeats correlations of surgically excised type 1 and type 2
that much remains to be known about this factor. submacular choroidal neovascular membranes. Am J
Considering that it is a potent antiangiogenic fac- Ophthalmol. 1998;126:59–69.
13. Grossniklaus HE, Green WR. Choroidal neovascular-
tor [77], its use can be manipulated to have ben- ization. Am J Ophthalmol. 2004;137:496–503.
efit over CNV growth. Intraocular or subretinal 14. Bressler NM, Bressler SB, Seddon JM, Gragoudas
injection of adeno-associated viral vector with ES, Jacobson LP. Drusen characteristics in patients
PEDF gene is reported to partially inhibit laser- with exudative versus non-exudative age-related mac-
ular degeneration. Retina. 1998;8:109–14.
induced neovascularization in murine models 15. Mullins RF, Johnson MN, Faidley EA, Skeie JM,
[79], raising the hope that sustained release of Huang J. Choriocapillaris vascular dropout related
pigment epithelial-derived factor may cause to density of drusen in human eyes with early age-
regression even in established related macular degeneration. Invest Ophthalmol Vis
Sci. 2011;52:1606–12.
neovascularization. 16. Pauleikhoff D, Chen J, Chisholm I, Bird A. Choroidal
perfusion abnormality with age-related Bruch’s
membrane change. Am J Ophthalmol. 1990;109:
211–7.
References 17. Sarks SH. Council lecture: drusen and their rela-
tionship to senile macular degeneration. Aust J
1. Krebs W, Krebs I. Fine structure of the choroid. In: Ophthalmol. 1980;8:117–30.
Primate retina and choroid. New York, NY: Springer; 18. Wang JJ, Foran S, Smith W, Mitchell P. Risk of age-
1991. related macular degeneration in eyes with macular
2. Nag TC, Kumari C. Electron microscopy of the drusen or hyperpigmentation: the Blue Mountains eye
human choroid. In: Chhablani J, Ruiz-Medrano J, edi- study cohort. Arch Ophthalmol. 2003;121:658–63.
tors. Choroidal disorders. London: Academic Press; 19. Abdelsalam A, Del Priore L, Zarbin MA. Drusen
2017. p. 7–20. in age-related macular degeneration: pathogenesis,
3. Nickla DL, Wallman J. The multifunctional choroid. natural course, and laser photocoagulation-induced
Prog Ret Eye Res. 2010;29:144–68. regression. Surv Ophthalmol. 1999;44:1–29.
4. Nag TC, Wadhwa S. Ultrastructure of human retina 20. Bressler SB, Maguire MG, Bressler NM, Fine
in aging and various pathological states. Micron. SL. Relationship of drusen and abnormalities of
2012;43:859–71. the retinal pigment epithelium to the prognosis of
5. Ambati J, Ambati BK, Yoo SH, Ianchulev S, Adamis neovascular macular degeneration. The macular
AP. Age-related macular degeneration: etiol- photocoagulation study group. Arch Ophthalmol.
ogy, pathogenesis, and therapeutic strategies. Surv 1990;108:1442–7.
Ophthalmol. 2003;48:257–93. 21. Bressler NM, Silva JC, Bressler SB, Fine SL,
6. Bhutto I, Lutty G. Understanding age-related macu- Green WR. Clinicopathologic correlation of dru-
lar degeneration (AMD): relationships between the sen and retinal pigment epithelial abnormalities in
photoreceptor/retinal pigment epithelium/Bruch’s age-related macular degeneration. Retina. 1994;14:
membrane/choriocapillaris complex. Mol Asp Med. 130–42.
2012;33:295–317. 22. Fine SL, Berger JW, Maguire MG, Ho AC. Age-
7. Green WR. Histopathology of age-related macular related macular degeneration. New Eng J Med.
degeneration. Mol Vis. 1999;5:27. 2000;342:483–92.
8. Green WR, Enger C. Age-related macular 23. Pauleikhoff D. Drusen in Bruch’s membrane. Their
degeneration histopathologic studies: the 1992 significance for the pathogenesis and therapy of age-
Lorenz E. Zimmerman Lecture. Ophthalmology. associated macular degeneration. Ophthalmologe.
1993;100:1519–35. 1992;89:363–86. [Article in German]
24. Sarks SH, Van Driel D, Maxwell L, Killingsworth atrophy in the complications of age-related macu-
M. Softening of drusen and subretinal neovasculariza- lar degeneration prevention trial. Ophthalmology.
tion. Trans Ophthalmol Soc UK. 1980;100:414–22. 2008;115:1474–9.
25. Sarks JP, Sarks SH, Killingsworth MC. Evolution of 36. Spraul CW, Grossniklaus HE. Characteristics of
soft drusen in age-related macular degeneration. Eye. Drusen and Bruch’s membrane in postmortem
1994;8:269–83. eyes with age-related macular degeneration. Arch
26. Shaw PX, Zhang L, Zhang M, Du H, Zhao L, Lee C, Ophthalmol. 1997;115:267–73.
Grob S, Lim SL, Hughes G, Lee J, Bedell M, Nelson 37. Friedman E. The pathogenesis of age-related macular
MH, Lu F, Krupa M, Luo J, Ouyang H, Tu Z, Su Z, degeneration. Am J Ophthalmol. 2008;146:348–9.
Zhu J, Wei X, Feng Z, Duan Y, Yang Z, Ferreyra H, 38. Spraul CW, Lang GE, Grossniklaus HE, Lang
Bartsch DU, Kozak I, Zhang L, Lin F, Sun H, Feng GK. Characteristics of drusen and changes in Bruch’s
H, Zhang K. Complement factor H genotypes impact membrane in eyes with age-related macular degenera-
risk of age-related macular degeneration by interaction. Histological study. Ophthalmologe. 1998;95:73–
tion with oxidized phospholipids. Proc Natl Acad Sci 9. [Article in German]
U S A. 2012;109:13757–62. 39. Lu Z, Lin V, May A, Che B, Xiao X, Shaw DH, Su
27. Sarks S, Cherepanoff S, Killingsworth M, Sarks F, Wang Z, Du H, Shaw PX. HTRA1 synergizes
J. Relationship of basal laminar deposit and membra- with oxidized phospholipids in promoting inflamma-
nous debris to the clinical presentation of early age- tion and macrophage infiltration essential for ocular
related macular degeneration. Invest Ophthalmol Vis VEGF expression. PLoS One. 2019;14:e0216808.
Sci. 2007;48:968–77. 40. Suzuki M, Tsujikawa M, Itabe H, Du ZJ, Xie P,
28. Chang TS, Freund KB, de la Cruz Z, Yannuzzi LA, Matsumura N, Fu X, Zhang R, Sonoda KH, Egashira
Green WR. Clinicopathologic correlation of choroi- K, Hazen SL, Kamei M. Chronic photo-oxidative
dal neovascularization demonstrated by indocyanine stress and subsequent MCP-1 activation as causative
green angiography in a patient with retention of good factors for age-related macular degeneration. J Cell
vision for almost four years. Retina. 1994;14:114–24. Sci. 2012;125:2407–15.
29. Grossniklaus HE, Green WR. Histopathologic and 41. Biesemeier A, Taubitz T, Julien S, Yoeruek E,
ultrastructural findings of surgically excised choroi- Schraermeyer U. Choriocapillaris breakdown pre-
dal neovascularization. Submacular surgery trials cedes retinal degeneration in age-related macular
research group. Arch Opthalmol. 1998;116:745–9. degeneration. Neurobiol Aging. 2014;35:2562–73.
30. Hermans P, Lommatzsch A, Bornfeld N, Pauleikhoff 42. McLeod DS, Grebe R, Bhutto I, Merges C, Baba T,
D. Angiographic-histological correlation of late Lutty GA. Relationship between RPE and chorio-
exudative age-related macular degeneration. capillaris in age-related macular degeneration. Invest
Ophthalmologe. 2003;100:378–83. Ophthalmol Vis Sci. 2009;50:4982–91.
31. Lommatzsch A, Hermans P, Müller KD, Bornfeld 43. Seddon JM, McLeod DS, Bhutto IA, Villalonga
N, Bird AC, Pauleikhoff D. Are low inflamma- MB, Silver RE, Wenick AS, Edwards MM, Lutty
tory reactions involved in exudative age-related GA. Histopathological insights into choroidal vas-
macular degeneration? Morphological and immun- cular loss in clinically documented cases of age-
histochemical analysis of AMD associated with related macular degeneration. JAMA Ophthalmol.
basal deposits. Graefes Arch Clin Exp Ophthalmol. 2016;134:1272–80.
2008;246:803–10. 44. Green WR, Key SN 3rd. Senile macular degeneration:
32. Lopez PF, Grossniklaus HE, Lambert HM, Aaberg a histopathologic study. Trans Am Ophthalmol Soc.
TM, Capone A Jr, Sternberg P Jr, L’Hernault 1977;75:180–254.
N. Pathologic features of surgically excised subreti- 45. Killingsworth MC. Angiogenesis in early choroidal
nal neovascular membranes in age-related macular neovascularization secondary to age-related macular
degeneration. Am J Ophthalmol. 1991;112:647–56. degeneration. Graefes Arch Clin Exp Ophthalmol.
33. Spraul CW, Lang GE, Grossniklaus HE, Lang 1995;233:313–23.
GK. Histologic and morphometric analysis of the 46. Penfold PL, Provis JM, Billson FA. Age-related mac-
choroid, Bruch’s membrane, and retinal pigment epi- ular degeneration: ultrastructural studies of the rela-
thelium in postmortem eyes with age-related macular tionship of leucocytes to angiogenesis. Graefes Arch
degeneration and histologic examination of surgi- Clin Exp Ophthalmol. 1987;225:70–6.
cally excised choroidal neovascular membranes. Surv 47. Archer DB, Gardiner TA. Electron microscopic fea-
Ophthalmol. 1999;44(Suppl 1):S10–32. tures of experimental choroidal neovascularization.
34. Moore DJ, Hussain AA, Marshall J. Age-related vari- Am J Ophthalmol. 1981;91:433–57.
ation in the hydraulic conductivity of Bruch’s mem- 48. Miller H, Miller B, Ryan SJ. Newly-formed subretinal
brane. Invest Ophthalmol Vis Sci. 1995;36:1290–7. vessels. Fine structure and fluorescein leakage. Invest
35. Complications of Age-related Macular Degeneration Ophthalmol Vis Sci. 1986;27:204–13.
Prevention Trial (CAPT) Research Group. Risk fac- 49. Grossniklaus HE, Ling JX, Wallace TM, Dithmar
tors for choroidal neovascularization and geographic S, Lawson DH, Cohen C, Elner VM, Elner SG,
Sternberg P Jr. Macrophage and retinal pigment 63. Grossniklaus HE, Cingle KA, Yoon YD, Ketkar N,
epithelium expression of angiogenic cytokines in L’Hernault N, Brown S. Correlation of histologic
choroidal neovascularization. Mol Vis. 2002;8: 2-dimensional reconstruction and confocal scanning
119–26. laser microscopic imaging of choroidal neovascular-
50. Killingsworth MC, Sarks JP, Sarks SH. Macrophages ization in eyes with age-related maculopathy. Arch
related to Bruch’s membrane in age-related macular Ophthalmol. 2000;118:625–9.
degeneration. Eye. 1990;4:613–21. 64. Skeie JM, Mullins RF. Elastin-mediated choroidal
51. Ryan SJ. Subretinal neovascularization after laser endothelial cell migration: possible role in age-related
coagulation. Graefes Arch Clin Exp Ophthalmol. macular degeneration. Invest Ophthalmol Vis Sci.
1980;215:29–42. 2008;49:5574–80.
52. Heriot WJ, Henkind P, Bellhorn DVN, Burns 65. Penfold PL, Madigan MC, Gillies MC, Provis
MS. Choroidal neovascularisation can digest JM. Immunological and aetiological aspects of macu-
Bruch’s membrane: a prior break is not essential. lar degeneration. Prog Ret Eye Res. 2001;20:385–414.
Ophthalmology. 1984;91:1603–8. 66. Korte GE, Repucci V, Henkind P. RPE destruction
53. Penfold P, Killingsworth M, Sarks S. An ultrastruc- causes choriocapillary atrophy. Invest Ophthalmol Vis
tural study of the role of leucocytes and fibroblasts Sci. 1984;25:1135–45.
in the breakdown of Bruch’s membrane. Aust NZ J 67. Spilsbury K, Garrett KL, Shen WY, Constable IJ,
Ophthalmol. 1984;12:23–31. Rakoczy PE. Overexpression of vascular endothelial
54. Penfold PL, Killingsworth MC, Sarks SH. Senile growth factor (VEGF) in the retinal pigment epithe-
macular degeneration: the involvement of immuno- lium leads to the development of choroidal neovascu-
competent cells. Graefes Arch Clin Exp Ophthalmol. larization. Am J Pathol. 2000;157:135–44.
1985;223:69–76. 68. Frank RN, Amin RH, Eliott D, Puklin JE, Abrams
55. Kadonosono K, Yazama F, Itoh N, et al. Expression of GW. Basic fibroblast growth factor and vascular
matrix metalloproteinase-7 in choroidal neovascular endothelial growth factor are present in epiretinal and
membranes in age-related macular degeneration. Am choroidal neovascular membranes. Am J Ophthalmol.
J Ophthalmol. 1999;128:382–4. 1996;122:393–403.
56. Lambert V, Munaut C, Jost M, Noël A, Werb Z, 69. Kvanta A, Algvere PV, Berglin L, Seregard
Foidart JM, Rakic JM. Matrix metalloproteinase-9 S. Subfoveal fibrovascular membranes in age-
contributes to choroidal neovascularization. Am J related macular degeneration express vascular endo-
Pathol. 2002;161:1247–53. thelial growth factor. Invest Ophthalmol Vis Sci.
57. van der Schaft TL, Mooy CM, de Bruijn WC, de Jong 1996;37:1929–34.
PT. Early stages of age-related macular degeneration: 70. Amin R, Puklin JE, Frank RN. Growth factor local-
an immunofluorescence and electron microscopy ization in choroidal neovascular membranes of age-
study. Br J Ophthalmol. 1993;77:657–61. related macular degeneration. Invest Ophthalmol Vis
58. Nishimura T, Goodnight R, Prendergast RA, Ryan Sci. 1994;35:3178–88.
SJ. Activated macrophages in experimental sub- 71. Campochiaro PA, Sugg R, Grotendorst G, Hjelmeland
retinal neovascularization. Ophthalmologica. LM. Retinal pigment epithelial cells produce PDGF-
1990;200:39–44. like proteins and secrete them into their media. Exp
59. Oh H, Takagi H, Takagi C, et al. The potential angio- Eye Res. 1989;49:217–27.
genic role of macrophages in the formation of cho- 72. Mori K, Duh E, Gehlbach P, Ando A, Takahashi K,
roidal neovascular membranes. Invest Ophthalmol Vis Pearlman J, Mori K, Yang HS, Zack DJ, Ettyreddy
Sci. 1999;40:1891–8. D, Brough DE, Wei LL, Campochiaro PA. Pigment
60. Grossniklaus HE, Miskala PH, Green WR, Bressler epithelium derived factor inhibits retinal and cho-
SB, Hawkins BS, Toth C, Wilson DJ, Bressler roidal neovascularization. J Cell Physiol. 2001;188:
NM. Histopathologic and ultrastructural features of 253–63.
surgically excised subfoveal choroidal neovascular 73. Bhutto IA, McLeod DS, Hasegawa T, Kim SY,
lesions: submacular surgery trials report no. 7. Arch Merges C, Tong P, Lutty GA. Pigment epithelium-
Ophthalmol. 2005;123:914–21. derived factor (PEDF) and vascular endothelial
61. Lafaut BA, Aisenbrey S, Vanden Broecke C, Di Tizio growth factor (VEGF) in aged human choroid and
F, Bartz-Schmidt KU. Clinicopathological correlation eyes with age-related macular degeneration. Exp Eye
in exudative age-related macular degeneration: recur- Res. 2006;82:99–110.
rent choroidal neovascularization. Graefes Arch Clin 74. Eye Disease Case Control Study Group. Antioxidant
Exp Ophthalmol. 2001;239:5–11. status and neovascular age-related macular degenera-
62. Gehrs KM, Heriot WJ, de Juan E Jr. Transmission tion. Arch Ophthalmol. 1993;111:104–9.
electron microscopic study of a subretinal choroidal 75. Skeie JM, Mullins RF. Macrophages in neovascular
neovascular membrane due to age-related macular age-related macular degeneration: friends or foes?
degeneration. Arch Ophthalmol. 1992;110:833–7. Eye (Lond). 2009;23:747–55.
76. Anand-Apte B, Pepper MS, Voest E, Montesano R, 78. Ferguson TA. Kaplan H, Tezel T, et al: augmentation
Olsen B, Murphy G, Apte SS, Zetter B. Inhibition of of choroidal neovascularization in a laser induced
angiogenesis by tissue inhibitor of metalloproteinase- mouse model by PEDF. ARVO Annual Meeting,
3. Invest Ophthalmol Vis Sci. 1997;38:817–23. Abstract. 2002;2756
77. Dawson DW, Volpert OV, Gillis P, Crawford SE, 79. Mori K, Gehlbach P, Yamamoto S, et al. AAV-
Xu H, Benedict W, Bouck NP. Pigment epithelium mediated gene transfer of pigment epithelium-derived
derived factor: a potent inhibitor of angiogenesis. factor inhibits choroidal neovascularization. Invest
Science. 1999;285:245–8. Ophthalmol Vis Sci. 2002;43:1994–2000.
Dr. TC Nag is a professor in the Department of Anatomy,

All India Institute of Medical Sciences, New Delhi, India.
His main research interest is to see age-related changes in
Dr. Sneha Gupta is currently in her final year of resi-
the human retina and choroid, to find a basis/link that trig-
dency (MD) in the Department of Anatomy, All India
gers for the pathogenesis of age-related macular
Institute of Medical Sciences, New Delhi, India. Her
degeneration.
research interest includes light-induced retinal changes in
He has published about 142 papers in peer-reviewed
rodent models, especially to understand the role of con-
journals and 7 invited chapters in books published by
tinuous exposure to light on the pathophysiology of age-
Elsevier and Springer-Verlag.
related macular degeneration.
Genomics in Choroidal
Kenji Yamashiro
6.1 Genes Associated samples [3–6]. In 2010, two GWASs using 2688
with Age-Related Macular and 3307 samples found further AMD suscepti-
Degeneration bility genes of LIPC and TIMP3, respectively [7,
8]. The first GWAS from Japan discovered a sig-
Age-related macular degeneration (AMD) is a nificant association between TNFRSF10A and
multifactorial disease, and its development is AMD using 827 cases and 3323 controls [9], and
affected by both genetic and environmental fac- the first GWAS combining East Asian samples of
tors. Aging and smoking are the major environ- 2119 cases and 5691 controls discovered genome-
mental factors for AMD, and genes affecting AMD wide significance of CETP with AMD [10].
development are called susceptibility genes. In 2013, a large-scale GWAS comparing
Since the discovery of associations between 2,442,884 genetic variations between 7650 cases
the CFH gene and AMD development in 2005 and 51,844 controls discovered seven new loci
[1], many genome-wide association studies associated with AMD [11]. This study group fur-
(GWASs) have investigated AMD pathogenesis. ther performed a large-scale GWAS comparing
The first GWAS successfully discovered an asso- 12,023,830 genetic variations between 16,144
ciation between CFH and AMD by comparing cases and 17,832 controls and discovered 34
116,204 genetic variations between 96 cases and AMD-susceptibility loci in 2016 (Table 6.1) [12].
50 controls [1]. In 2006, GWASs comparing Since GWAS can sometimes provide false posi-
97,824 genetic variations between 130 cases and tive results, the results should be carefully under-
96 controls discovered the second AMD suscep- stood by referring to reports from other groups.
tibility locus, ARMS2/HTRA1 [2]. After these
discoveries, several candidate-gene approach
studies confirmed significant associations 6.2 Genes Associated
between AMD and complement pathway genes, with Neovascular AMD
such as C2/CFB, C3, and CFI, using 1287–2172
To date, the mechanisms by which susceptibility
K. Yamashiro (*) genes affect AMD development have not been
Department of Ophthalmology, Otsu Red Cross thoroughly elucidated. Some AMD susceptibility
Hospital, Otsu, Japan genes might promote drusen accumulation, vul-
Department of Ophthalmology and Visual Sciences, nerability of the Bruch’s membrane or retinal pig-
Kyoto University Graduate School of Medicine, ment epithelium, CNV development, or g eographic
Kyoto, Japan atrophy (GA) development. Since most previous
e-mail: yamashro@kuhp.kyoto-u.ac.jp

58 K. Yamashiro
GWASs on AMD have included both neovascular nificantly associated only with neovascular AMD,
AMD and atrophic AMD in case samples, it is and it was not significantly associated with GA
unclear whether susceptibility genes for CNV [12]. To confirm the reproducibility of these find-
development and GA development are similar. ings, associations between MMP9 and CNV or
However, one study reported that MMP9 was sig- GA should be further evaluated in the future.
Table 6.1 Susceptibility genes for age-related macular

degeneration 6.3 enes Associated with CNV
G
Chromosome Locus Subtypes
1 CFH
2 COL4A3 Neovascular AMD is usually divided into three sub-
3 ADAMTS9-AS types: typical AMD, polypoidal choroidal vascu-
3 COL8A1
lopathy (PCV), and retinal angiomatous proliferation
4 CFI
(RAP). Typical AMD develops as type 1 or type 2
5 SPEF2
5 C9
CNV, PCV develops as type 1 CNV with polypoidal
6 C2/CFB-SKIV2L lesions at its edge, and RAP develops as type 3
6 VEGFA CNV. Among previously reported susceptibility
7 KMT2E-SRPK2 genes for AMD, ARMS2/HTRA1 has demonstrated
7 PILRB-PILRA different associations to these three subtypes [13–
8 TNFRSF10A 16]. Its association is weaker for PCV and stronger
9 TRPM3 for RAP (Table 6.2). ARMS2/HTRA1 might deter-
9 MIR6130-RORB mine the type of CNV in eyes with AMD.
9 TGFBR1
Although most susceptibility genes are signifi-
9 ABCA1
cantly associated with all AMD subtypes, FGD6
10 ARHGAP21
10 ARMS2-HTRA1 might be associated only with PCV. One study
12 RDH5-CD63 from East Asia reported that a rare variant of FGD6
12 ACAD10 was significantly associated with PCV, but not with
13 B2GALTL typical AMD [17]. Therefore, associations between
14 RAD51B FGD6 and PCV or typical AMD should be evalu-
15 LIPC ated in Caucasian populations. Further investiga-
16 CETP tion of FGD6 might determine whether it is a true
16 CTRB2-CTRB1 susceptibility gene of PCV only.
17 TMEM97-VTN
17 NPLOC4-TSPAN10
19 CNN2
19 C3 6.4 Genes Associated
19 APOE with Natural Course of PCV
20 MMP9
20 C20orf85 In addition to its potential roles in the divergence
22 SYN3-TIMP3 of AMD into typical AMD, PCV, or RAP,
22 SLC16A8 ARMS2/HTRA1 might be useful to predict PCV
Table 6.2 Genetic associations to polypoidal choroidal vasculopathy

Association to
Genes PCV tAMD RAP Ref
ARMS2/HTRA1 Association (+) (weak) Association (+) Association (+) (strong) 13–16
FGD6 Association (+) Association (−) 17
PCV polypoidal choroidal vasculopathy, tAMD typical age-related macular degeneration, RAP retinal angiomatous
proliferation, Ref references
6 Genomics in Choroidal Neovascularization 59
prognosis. One Japanese group reported that the associations between ARMS2/HTRA1 and AMD
T allele in ARMS2/HTRA1 rs10490924 was a risk bilaterality [29], many previous studies reported
allele for vitreous hemorrhage, retinal hemor- positive associations between ARMS2/HTRA1
rhage, hemorrhagic pigment epithelial detach- and AMD bilaterality. Additionally, one study
ment (PED), and serous PED in eyes with PCV reported further genome-wide level associations.
[18, 19]. Since hemorrhagic complications sig- Genotype data of ARMS2/HTRA1 could be useful
nificantly worsen visual prognosis of eyes with to predict future development of AMD in the
PCV, ARMS2/HTRA1 would determine the natu- unaffected eye of patients with unilateral AMD.
ral prognosis of PCV. In contrast to ARMS2/ In a study on 207 patients with unilateral
HTRA1, CFH has not shown clear associations to AMD, the rate of AMD development in the unaf-
the natural course of PCV in previous studies. fected eye at 10 years was reportedly about 10%,
50%, and 70% in patients with GG, GT, and TT
genotype in ARMS2/HTRA1 rs10490924, respec-
6.5 enes Associated with CNV
G tively [30].
Size in AMD
Since CNV size determines scotoma size in 6.7 Genes Associated

patients with AMD, genes associated with CNV with Treatment Outcomes
size are clinically important. In a GWAS from for CNV in AMD
Japan, ARMS2/HTRA1 and MMP20 were discov-
ered as significantly associated with CNV size in After approval of anti-VEGF drugs for AMD,
AMD [20]. The association between ARMS2/ anti-VEGF treatment has become the first-line
HTRA1 and CNV size in AMD has been reported treatment for AMD. As such, photodynamic ther-
in previous studies on ARMS2/HTRA1 associa- apy (PDT) is rarely used to treat AMD. However,
tion to AMD [21–23]. Therefore, ARMS2/HTRA1 recent studies suggest that PDT would be a better
would be a key gene to determine CNV size in treatment for some eyes with PCV [31, 32].
AMD. As for MMP20, its association with CNV Although outcomes of PDT treatment for AMD
size should be confirmed using other cohorts, due are not always favorable, the genetic information
to the potential of false-positive results from of ARMS2/HTRA1 might be useful to predict the
GWAS. Similar to genetic associations in PCV response to PDT in AMD. Three studies from
prognosis, CFH does not show a clear associa- Japan reported that patients with risk alleles for
tion to CNV size in AMD. AMD in ARMS2/HTRA1 showed worse visual
outcomes after PDT for typical AMD and
PCV. Genotype information of ARMS2/HTRA1
6.6 enes Associated with AMD
G might enable prediction of PDT treatment out-
Bilaterality comes for typical AMD and PCV. In Caucasian
populations, however, two reports did not find
Early genetic studies for AMD reported that asso- associations between ARMS2/HTRA1 and PDT
ciations between ARMS2/HTRA1 and AMD were outcomes for AMD [33, 34]. Since PDT is rarely
stronger for patients with bilateral AMD than used for typical AMD, yet PDT can be a suitable
patients with unilateral AMD [24, 25]. Later stud- treatment for some PCV, associations between
ies confirmed significant genotype distribution ARMS2/HTRA1 and PDT outcomes should be
differences in ARMS2/HTRA1 between patients further confirmed in eyes with PCV to enable
with bilateral AMD and patients with unilateral precision/personalized medicine with PDT for
AMD [19, 26–28]. In 2017, a GWAS from Japan PCV.
confirmed a genome-wide level significant asso- In addition to ARMS2/HTRA1, several genes
ciation between ARMS2/HTRA1 and AMD bilat- have been reported to be associated with treat-
erality. Although the CATT study did not show ment outcomes after PDT for AMD, including
60 K. Yamashiro
VEGFA [35, 36], PEDF [37], CRP [38], and Although GWASs on associations for AMD treat-
coagulation balance genes [39, 40]. Therefore, ment outcomes are rare, one GWAS from Japan
use of genetic information might be able to reported that ARMS2/HTRA1 might be useful to
improve the prediction accuracy of treatment out- predict the number of anti-VEGF drug injections
comes for AMD. However, limited evidence is
available for the associations between these Table 6.4 Associations of ARMS2/HTRA1 gene with
genes and treatment outcomes. Further studies treatment outcome after anti-VEGF treatment for AMD
are warranted to use these genes for precision/ Association to treatment
personalized medicine in AMD. N outcome
In contrast to genes associated with PDT out- Kang, et al. [57] 75 Significant association
McKibbin, et al. [41] 104 Significant association
comes, genes associated with treatment outcomes
Abedi, et al. [55] 224 Significant association
after anti-VEGF treatment have not been clearly
Tian, et al. [43] 144 Significant association
elucidated. In candidate gene approach studies, Teper, et al. [56] 90 Significant association
various genes have been evaluated for their asso- Yuan, et al. [59] 168 Significant association
ciations to anti-VEGF treatment outcomes, such Kitchens, et al. [58] 101 Significant association
as ARMS2/HTRA1, CFH, VEGFA, C2/CFB, C3, Smailhodzic, et al. 420 No association
CFI, APOE, IL8, and PEDF. Although some [47]
studies reported significant associations of these Van Asten, et al. [63] 391 No association
genes with treatment outcomes after anti-VEGF Orlin, et al. [52] 150 No association
Hagstrom, et al. [54] 834 No association
treatment for AMD, many other studies did not
Hata, et al. [53] 105 No association
find associations (Tables 6.3, 6.4, 6.5, and 6.6). Kloeckener-Gruissem, 122 No association
et al. [44]
Table 6.3 Associations of CFH gene with treatment out- Yamashiro, et al. [51] 75 No association
come after anti-VEGF treatment for AMD Brantley, et al. [45] 86 No association
Chang, et al. [60] 102 No association
Association to treatment
N outcome Park et al. [62] 273 No association
Imai, et al. [50] 83 No association
McKibbin, et al. [41] 104 Significant association
Menghini, et al. [42] 204 Significant association
Tian, et al. [43] 144 Significant association Table 6.5 Associations of VEGFA gene with treatment
Kloeckener-Gruissem, 122 Significant association outcome after anti-VEGF treatment for AMD
et al. [44]
Association to treatment
Brantley, et al. [45] 86 Significant association N outcome
Dikmetas, et al. [46] 193 Significant association Abedi, et al. [64] 201 Significant association
Smailhodzic, et al. 420 Significant association Nakata, et al. [36] 94 Significant association
[47]
Kitchens, et al. [58] 101 Significant association
Lee, et al. [48] 156 Significant association
Park et al. [62] 273 Significant association
Nischler, et al. [49] 197 Significant association
Imai, et al. [50] 83 Significant association
Imai, et al. [50] 83 Significant association
Lazzeri, et al. [65] 64 Significant association
Yamashiro, et al. [51] 75 No association
McKibbin, et al. 104 No association
Orlin, et al. [52] 150 No association [41]
Hata, et al. [53] 105 No association Smailhodzic, et al. 420 No association
Hagstrom, et al. [54] 834 No association [47]
Abedi, et al. [55] 224 No association van Asten, et al. 391 No association
Teper, et al. [56] 90 No association [63]
Kang, et al. [57] 75 No association Hagstrom, et al. 834 No association
Kitchens, et al. [58] 101 No association [54]
Yuan, et al. [59] 168 No association Wang, et al. [66] 106 No association
Chang, et al. [60] 102 No association Boltz, et al. [67] 141 No association
Habibi, et al. [61] 70 No association Yuan, et al. [59] 168 No association
Park et al. [62] 273 No association Chang, et al. [60] 102 No association
Table 6.6 Associations of other genes with treatment adjust to each patient’s needs. TAE regimen can
outcome after anti-VEGF treatment for AMD
theoretically prevent undertreatment for AMD,
Association to and many studies have reported favorable treat-
treatment
Loci N outcome
ment results [72, 73]. Therefore, genetic associa-
Kloeckener- C2/ 122 No association tion studies analyzing samples treated with the
Gruissem, et al. [44] CFB TAE regimen could reveal usefulness of genetic
Abedi, et al. [55] C2/ 224 No association information for precision/personalized medicine
CFB in AMD.
Park, et al. [62] C2/ 273 No association
CFB
Abedi, et al. [55] C3 224 No association
Hagstrom, et al. [54] C3 834 No association 6.8 Genes Associated
Park, et al. [62] CFI 273 No association with Pachychoroid Diseases
Park, et al. [62] APOE 273 No association
Wickremasinghe, APOE 192 Significant Although late AMD has been thought to develop
et al. [68] association from early AMD with drusen and retinal pigmen-
Hautamaki, et al. IL8 96 Significant tary abnormalities, some eyes with late AMD do
[69] association
not have drusen or pigmentary abnormalities. In
Hautamaki, et al. IL8 50 Significant
[70] association 2012, it was proposed that some eyes with previ-
Park, et al. [62] PEDF 273 No association ously diagnosed neovascular AMD should be
Imai, et al. [50] PEDF 83 Significant diagnosed as having CNV secondary to central
association serous chorioretinopathy (CSC) rather than AMD
[74]. Before this proposal, it had been demon-
strated that eyes with CSC had thicker choroid
required to maintain dry macula after initial treat- [75]. To mention the thick choroid status, a new
ment [71]. In AMD, ARMS2/HTRA1 would be the term “pachychoroid” was coined in 2013, and the
most attractive gene for precision/personalized CNV secondary to CSC was termed as “pachy-
medicine with anti-VEGF treatment. choroid neovasculopathy” that develops in asso-
Most previous genetic association studies on ciation with choroidal thickening [76]. Detailed
treatment outcomes after anti-VEGF treatment clinical phenotypes of pachychoroid neovascu-
for AMD analyzed samples treated with a pro re lopathy were described in 2015 [77], and the
nata (PRN) regimen. Eyes with AMD treated genetic characteristics of pachychoroid neovas-
with the PRN regimen receive an initial treatment culopathy were reported in 2016 [78].
of anti-VEGF injections to inactivate CNV and In Japanese populations, about 20% of previ-
then were followed up monthly to determine ously diagnosed AMD cases were reportedly
recurrence of retinal exudative changes. Although pachychoroid neovasculopathy [79]. To date,
additional treatment is required immediately fol- clinically significant differences have not been
lowing the detection of retinal exudative change reported on treatment outcomes between neovas-
recurrence with the PRN regimen, many patients cular AMD and pachychoroid neovasculopathy.
do not follow-up monthly and receive immediate However, a recent genetic study confirmed the
treatment partly due to patient decisions in the significance of differentiating pachychoroid neo-
real world. As a result, insufficient treatment can vasculopathy from AMD. Risk alleles for AMD
lead to false results in genetic association studies in CFH were determined to function as protective
on treatment outcomes. alleles for pachychoroid, while risk alleles for
Recently, the treat and extend (TAE) regimen pachychoroid were protective alleles for AMD,
has become popular to treat AMD with anti- suggesting that AMD and pachychoroid are
VEGF drugs. In this regimen, patients continue genetically distinct relative to CFH [80].
to receive treatment at every visit, and the inter- Thus far, genetic studies on pachychoroid dis-
val between visits is extended or shortened to eases have been limited. Previously reported fre-
62 K. Yamashiro
Table 6.7 Frequencies of risk alleles for age-related macular degeneration in Caucasians
Risk Risk allele frequencies
Loci SNP allele Pachychoroid Control Pachychoroid neovasculopathy AMD Ref
CFH Y402H C 0.24 0.31 0.46 0.63 48
ARMS2/HTRA1 A69S T 0.15 0.22 0.41 0.44
Table 6.8 Frequencies of risk alleles for age-related macular degeneration in Japanese
Risk allele frequencies
Loci SNP Risk allele Control Pachychoroid neovasculopathy AMD Ref
CFH I62V G 0.59 0.59 0.75 49
ARMS2/HTRA1 A69S T 0.39 0.51 0.65
quencies of risk alleles for AMD in CFH and studies are warranted to confirm the association
ARMS2/HTRA1 are summarized in Tables 6.7 between ARMS2/HTRA1 and CNV development
and 6.8. The C allele frequencies of CFH Y402H in pachychoroid neovasculopathy.
were found to be 0.24 in pachychoroid subjects Elucidation of genetic background for
and 0.31 in controls among Caucasians, suggest- pachychoroid diseases might enable highly
ing that the AMD risk allele in CFH, C, has pro- accurate precision/personalized medicine for
tective roles against pachychoroid development AMD. Previous genetic studies on AMD treat-
in Caucasians. The higher frequency of C allele ment outcomes have included pachychoroid
(0.46) in pachychoroid neovasculopathy patients neovasculopathy as AMD. However, AMD and
suggests that the C allele is a risk allele for pro- pachychoroid disease are genetically opposite
gression of pachychoroid to pachychoroid neo- diseases from the viewpoint of CFH. Genetic
vasculopathy after CNV development. The risk studies on treatment outcomes for pachycho-
allele of CFH for AMD would be a risk allele for roid neovascularization and AMD will lead to
CNV development in pachychoroid neovascu- highly accurate precision/personalized medi-
lopathy in Caucasian subjects. cine for pachychoroid neovasculopathy and
Although the risk allele frequency of CFH in AMD.
pachychoroid subjects has not been reported in
Japanese populations, recent GWAS has discov-
ered that the risk allele of CFH for AMD has a 6.9 Genes Associated
protective effect against pachychoroid [80]. The with Myopia
risk allele frequency in pachychoroid must be
lower than the control. Taken together with the Myopia is also a multifactorial disease, and its
risk allele frequency of 0.59 in controls, the risk heritability has long been widely known. The first
allele frequency of 0.59 in pachychoroid neovas- established susceptibility gene for myopia,
cularization suggests that the CFH risk allele in GJD2, was discovered in 2010 through GWAS
AMD contributes to development of pachycho- [81]. Its association with high myopia was also
roid neovasculopathy. CFH should have similar confirmed [82]. Since the effect of each suscepti-
effects on CNV development in pachychoroid bility gene is weak for myopia, large-scale
neovasculopathy, as it affects CNV development GWAS is required to discover susceptibility
in AMD both in Caucasian and Japanese popula- genes for myopia. In 2018, the CREAM
tions. As for ARMS2/HTRA1, it might also con- Consortium conducted a meta-analysis of
tribute to CNV development both in pachychoroid GWASs using 160,420 samples and discovered
neovasculopathy and AMD. However, further 161 genetic regions associated with myopia [83].
Table 6.9 Susceptibility genes for myopia 1

Vascular endothelium
Cornea Lens Retina (cell type not specified) (Choroid) Extracellular matrix (Sclera)
CHD7 BMP4 ACP2 MYO1D BMP2 ANTXR2
CLU CLU AKAP6 MYO5B BMP4 COL10A1
COL6A1 KCNA4 API5 NCOA2 CD34 EFEMP1
CYP26A1 MAF ARID2 NDUFBI CD55 LAMA2
EFEMP1 SIX3 C14orf39 PCAT4 FLT1 SNTB1
FBN1 ST8SIA1 C2CD5 PDE11A MED1 TCF7L2
FLT1 CD55 PDE3A TGFBR1 TGFBR1
PBX1 CHD7 PNPT1 TMEM98 TMEM98
RASGRF1 CLU PTPRR TNFSF12 VIPR2
SETMAR CYP26A1 RALY ZIC2
TCF7L2 DNAJB12 RASGRFI
TGFBR1 DRD1 RPP14
VIPR2 GRIK1 SH3GL2
KCNA4 SIX3
KCNMA1 SNTB1
KIRREL SYN3
LINC00461 TFAP2D
LYPLAL1 THEM184A
MYCN
Many myopia susceptibility genes were expressed 6.10 Genes Associated

in the retina (Tables 6.9 and 6.10), and it is now with Myopic CNV
thought that the retina plays important roles to
promote myopia development. Established susceptibility genes for CNV devel-
When myopia progression is severe, it development secondary to high myopia have not been
ops into high myopia. High myopia is usually discovered. Some susceptibility genes for AMD,
defined as eyes with refractive error greater myopia, and high myopia have been evaluated
than −6.0D or axial length more than 26.0 mm. for their associations to myopic CNV develop-
Although similarities between backgrounds of ment [20, 82, 91–95]. However, associations
myopia and high myopia development have not were not found for most genes and myopic CNV.
been cleared, most myopia susceptibility genes A recent GWAS from Japan discovered that
are suspected to susceptibility genes for high CCDC102B was associated with fundus patho-
myopia. In addition to myopia susceptibility logic change secondary to high myopia, but not
genes, GWASs on high myopia have reported with myopia development [96]. Therefore,
possible susceptibility genes for high myopia CCDC102B might contribute to the development
(Table 6.11). Interestingly, most susceptibility of pathologic changes in high myopia.
genes for high myopia have not been included CCDC102B might be associated with myopic
as myopia susceptibility genes discovered CNV development, and elucidation of
through a recent large-scale myopia CCDC102B mechanisms contributing to patho-
GWAS. There might be background that is logic change development could lead to preven-
associated only with high myopia development, tive treatments for myopic CNV in highly myopic
but not myopia. eyes.
64
Table 6.10 Susceptibility genes for myopia 2

Retina
Cone OFF Retinal pigment
Ganglion cell Amacrine cell Rod bipolar cell Cone ON bipolar cell bipolar cell Mueller cell Horizontal cell Cone Rod epithelium
BMP4 FRMPD2 GRM3 GJD2 LRlT2 CA8 DLG2 CABP4 CABP4 ANO2
GRIA4 GJD2 KCNA4 GNB3 EDN2 TFAP2B CACNA1D CACNA1D BMP2
GRM3 GRM3 KCNMA1 GRIA4 KCNJ2 ZEB2 EDN2 DYNLRB2 BMP4
KCNJ2 KCNA4 RORB TJP2 PRSS56 GJD2 EDN2 C8orf84
MED1 KCNMA1 RGR GNB3 FRMPD2 CLU
NMT POU6F2 KCNQ5 GJD2 EFEMP1
POU6F2 SEMA3D LRITI1 KCNQ5 IL4
ZIC2 TJP2 MED1 KCNA4 KCNJ5
TFAP2B RP1L1 LRIT1 KCNQ5
ZEB2 THRB LRlT2 MED1
TSPAN10 MAF RDH5
TTC8 MED1 RGR
RORB TGFBR1
RP1L1 TMEM98
TSPAN10 TTC8
TTC8 TRAF1
K. Yamashiro
Table 6.11 Previously reported possible susceptibility associated loci influence susceptibility to age-related
genes for high myopia macular degeneration. Proc Natl Acad Sci U S A.
2010;107(16):7401–6.
Nakanishi et al. [84] 11q24.1
8. Neale BM, Fagerness J, Reynolds R, et al. Genome-
Li et al. [85] CTNND2 wide association study of advanced age-related
Li et al. [86] 4q25 macular degeneration identifies a role of the hepatic
Shi et al. [87] 13q12.12 lipase gene (LIPC). Proc Natl Acad Sci U S A.
Fan et al. [88] 1q41 2010;107(16):7395–400.
Shi et al. [89] VIPR2 9. Arakawa S, Takahashi A, Ashikawa K, et al.
Shi et al. [89] SNTB1 Genome-wide association study identifies two
susceptibility loci for exudative age-related macular
Khor et al. [90] ZFHX1B
degeneration in the Japanese population. Nat Genet.
2011;43(10):1001–4.
Key Learning Points: 5–7 Key Messages 10. Cheng CY, Yamashiro K, Chen LJ, et al. New loci
and coding variants confer risk for age-related
1. Major susceptibility genes for AMD are CFH macular degeneration in east Asians. Nat Commun.
and ARMS2/HTRA1. 2015;6:6063.
2. MMP9 might be a susceptibility gene for neo- 11. Fritsche LG, Chen W, Schu M, et al. Seven new loci
vascular AMD but not for dry AMD. associated with age-related macular degeneration. Nat
Genet. 2013;45(4):433–9. 439e431-432
3. FGD6 might be a susceptibility gene for PCV 12. Fritsche LG, Igl W, Bailey JN, et al. A large genome-
but not for typical AMD. wide association study of age-related macular degen-
4. ARMS2/HTRA1 might be useful for precision/ eration highlights contributions of rare and common
personalized medicine for AMD. variants. Nat Genet. 2016;48(2):134–43.
13. Hayashi H, Yamashiro K, Gotoh N, et al. CFH and
5. Pachychoroid disease and AMD are geneti- ARMS2 variations in age-related macular degenera-
cally distinct diseases, as related to CFH. tion, polypoidal choroidal vasculopathy, and retinal
6. GJD2 is a major susceptibility gene for angiomatous proliferation. Invest Ophthalmol Vis Sci.
myopia. 2010;51(11):5914–9.
14. Tanaka K, Nakayama T, Yuzawa M, et al. Analysis
7. CCDC102B is a major susceptibility gene for of candidate genes for age-related macular degenera-
pathologic myopia. tion subtypes in the Japanese population. Mol Vis.
2011;17:2751–8.
15. Liang XY, Lai TY, Liu DT, et al. Differentiation
of exudative age-related macular degeneration
References and polypoidal choroidal vasculopathy in the
ARMS2/HTRA1 locus. Invest Ophthalmol Vis Sci.
1. Klein RJ, Zeiss C, Chew EY, et al. Complement factor 2012;53(6):3175–82.
H polymorphism in age-related macular degeneration. 16. Yuan D, Yuan D, Yuan S, Liu Q. The age-related macu-
Science. 2005;308(5720):385–9. lopathy susceptibility 2 polymorphism and polypoidal
2. Dewan A, Liu M, Hartman S, et al. HTRA1 promoter choroidal vasculopathy in Asian populations: a meta-
polymorphism in wet age-related macular degenera- analysis. Ophthalmology. 2013;120(10):2051–7.
tion. Science. 2006;314(5801):989–92. 17. Huang L, Zhang H, Cheng CY, et al. A missense vari-
3. Gold B, Merriam JE, Zernant J, et al. Variation in fac- ant in FGD6 confers increased risk of polypoidal cho-
tor B (BF) and complement component 2 (C2) genes roidal vasculopathy. Nat Genet. 2016;48(6):640–7.
is associated with age-related macular degeneration. 18. Sakurada Y, Kubota T, Mabuchi F, Imasawa M,
Nat Genet. 2006;38(4):458–62. Tanabe N, Iijima H. Association of LOC387715
4. Maller JB, Fagerness JA, Reynolds RC, Neale BM, A69S with vitreous hemorrhage in polypoi-
Daly MJ, Seddon JM. Variation in complement factor dal choroidal vasculopathy. Am J Ophthalmol.
3 is associated with risk of age-related macular degen- 2008;145(6):1058–62.
eration. Nat Genet. 2007;39(10):1200–1. 19. Sakurada Y, Kubota T, Imasawa M, et al. Role of
5. Yates JR, Sepp T, Matharu BK, et al. Complement C3 complement factor H I62V and age-related macu-
variant and the risk of age-related macular degenera- lopathy susceptibility 2 A69S variants in the clini-
tion. N Engl J Med. 2007;357(6):553–61. cal expression of polypoidal choroidal vasculopathy.
6. Fagerness JA, Maller JB, Neale BM, Reynolds RC, Ophthalmology. 2011;118(7):1402–7.
Daly MJ, Seddon JM. Variation near complement fac- 20. Akagi-Kurashige Y, Yamashiro K, Gotoh N, et al.
tor I is associated with risk of advanced AMD. Eur J MMP20 and ARMS2/HTRA1 are associated with
Hum Genet. 2009;17(1):100–4. Neovascular lesion size in age-related macular degen-
7. Chen W, Stambolian D, Edwards AO, et al. Genetic eration. Ophthalmology. 2015;122(11):2295–302.
variants near TIMP3 and high-density lipoprotein- e2292
66 K. Yamashiro
21. Gotoh N, Yamada R, Nakanishi H, et al. Correlation neovascular age-related macular degeneration in pop-
between CFH Y402H and HTRA1 rs11200638 ulations from Israel. Mol Vis. 2008;14:2263–71.
genotype to typical exudative age-related macular 34. Brantley MA Jr, Edelstein SL, King JM, et al.
degeneration and polypoidal choroidal vasculopa- Association of complement factor H and LOC387715
thy phenotype in the Japanese population. Clin Exp genotypes with response of exudative age-related
Ophthalmol. 2008;36(5):437–42. macular degeneration to photodynamic therapy. Eye
22. Andreoli MT, Morrison MA, Kim BJ, et al. (Lond). 2009;23(3):626–31.
Comprehensive analysis of complement factor H and 35. Immonen I, Seitsonen S, Tommila P, et al. Vascular
LOC387715/ARMS2/HTRA1 variants with respect endothelial growth factor gene variation and
to phenotype in advanced age-related macular degen- the response to photodynamic therapy in age-
eration. Am J Ophthalmol. 2009;148(6):869–74. related macular degeneration. Ophthalmology.
23. Bessho H, Honda S, Kondo N, Negi A. The asso- 2010;117(1):103–8.
ciation of age-related maculopathy susceptibility 2 36. Nakata I, Yamashiro K, Nakanishi H, Tsujikawa A,
polymorphisms with phenotype in typical neovascu- Otani A, Yoshimura N. VEGF gene polymorphism
lar age-related macular degeneration and polypoidal and response to intravitreal bevacizumab and triple
choroidal vasculopathy. Mol Vis. 2011;17:977–82. therapy in age-related macular degeneration. Jpn J
24. Seddon JM, Francis PJ, George S, Schultz Ophthalmol. 2011;55(5):435–43.
DW, Rosner B, Klein ML. Association of CFH 37. Nakata I, Yamashiro K, Yamada R, et al. Genetic
Y402H and LOC387715 A69S with progres- variants in pigment epithelium-derived factor influ-
sion of age-related macular degeneration. JAMA. ence response of polypoidal choroidal vasculopa-
2007;297(16):1793–800. thy to photodynamic therapy. Ophthalmology.
25. Seddon JM, Reynolds R, Maller J, Fagerness JA, 2011;118(7):1408–15.
Daly MJ, Rosner B. Prediction model for preva- 38. Feng X, Xiao J, Longville B, et al. Complement
lence and incidence of advanced age-related macu- factor H Y402H and C-reactive protein polymor-
lar degeneration based on genetic, demographic, and phism and photodynamic therapy response in age-
environmental variables. Invest Ophthalmol Vis Sci. related macular degeneration. Ophthalmology.
2009;50(5):2044–53. 2009;116(10):1908–12. e1901
26. Chen H, Yang Z, Gibbs D, et al. Association of 39. Parmeggiani F, Costagliola C, Gemmati D, et al.
HTRA1 polymorphism and bilaterality in advanced Predictive role of coagulation-balance gene poly-
age-related macular degeneration. Vis Res. morphisms in the efficacy of photodynamic therapy
2008;48(5):690–4. with verteporfin for classic choroidal neovasculariza-
27. Chen Y, Zeng J, Zhao C, et al. Assessing susceptibil- tion secondary to age-related macular degeneration.
ity to age-related macular degeneration with genetic Pharmacogenet Genomics. 2007;17(12):1039–46.
markers and environmental factors. Arch Ophthalmol. 40. Parmeggiani F, Costagliola C, Gemmati D, et al.
2011;129(3):344–51. Coagulation gene predictors of photodynamic therapy
28. Schwartz SG, Agarwal A, Kovach JL, et al. The ARMS2 for occult choroidal neovascularization in age-related
A69S variant and bilateral advanced age-related mac- macular degeneration. Invest Ophthalmol Vis Sci.
ular degeneration. Retina. 2012;32(8):1486–91. 2008;49(7):3100–6.
29. Maguire MG, Daniel E, Shah AR, et al. Incidence of 41. McKibbin M, Ali M, Bansal S, et al. CFH, VEGF
choroidal neovascularization in the fellow eye in the and HTRA1 promoter genotype may influence the
comparison of age-related macular degeneration treat- response to intravitreal ranibizumab therapy for
ments trials. Ophthalmology. 2013;120(10):2035–41. neovascular age-related macular degeneration. Br J
30. Tamura H, Tsujikawa A, Yamashiro K, et al. Ophthalmol. 2012;96(2):208–12.
Association of ARMS2 genotype with bilateral 42. Menghini M, Kloeckener-Gruissem B, Fleischhauer
involvement of exudative age-related macular degen- J, et al. Impact of loading phase, initial response and
eration. Am J Ophthalmol. 2012;154(3):542–8. e541 CFH genotype on the long-term outcome of treatment
31. Koh A, Lee WK, Chen LJ, et al. EVEREST study: for neovascular age-related macular degeneration.
efficacy and safety of verteporfin photodynamic ther- PLoS One. 2012;7(7):e42014.
apy in combination with ranibizumab or alone versus 43. Tian J, Qin X, Fang K, et al. Association of genetic
ranibizumab monotherapy in patients with symp- polymorphisms with response to bevacizumab
tomatic macular polypoidal choroidal vasculopathy. for neovascular age-related macular degeneration
Retina. 2012;32(8):1453–64. in the Chinese population. Pharmacogenomics.
32. Koh A, Lai TYY, Takahashi K, et al. Efficacy and 2012;13(7):779–87.
safety of Ranibizumab with or without Verteporfin 44. Kloeckener-Gruissem B, Barthelmes D, Labs S, et al.
photodynamic therapy for Polypoidal Choroidal Genetic association with response to intravitreal
vasculopathy: a randomized clinical trial. JAMA ranibizumab in patients with neovascular AMD. Invest
Ophthalmol. 2017;135(11):1206–13. Ophthalmol Vis Sci. 2011;52(7):4694–702.
33. Chowers I, Meir T, Lederman M, et al. Sequence 45. Brantley MA Jr, Fang AM, King JM, Tewari A, Kymes
variants in HTRA1 and LOC387715/ARMS2 and SM, Shiels A. Association of complement factor H and
phenotype and response to photodynamic therapy in LOC387715 genotypes with response of exudative
age-related macular degeneration to intravitreal beva- 58. Kitchens JW, Kassem N, Wood W, et al. A pharmaco-
cizumab. Ophthalmology. 2007;114(12):2168–73. genetics study to predict outcome in patients receiving
46. Dikmetas O, Kadayifcilar S, Eldem B. The effect of anti-VEGF therapy in age related macular degenera-
CFH polymorphisms on the response to the treatment tion. Clin Ophthalmol. 2013;7:1987–93.
of age-related macular degeneration (AMD) with 59. Yuan D, Yuan D, Liu X, Yuan S, Xie P, Liu Q. Genetic
intravitreal ranibizumab. Mol Vis. 2013;19:2571–8. association with response to intravitreal ranibizumab
47. Smailhodzic D, Muether PS, Chen J, et al. Cumulative for neovascular age-related macular degeneration
effect of risk alleles in CFH, ARMS2, and VEGFA in the Han Chinese population. Ophthalmologica.
on the response to ranibizumab treatment in age- 2013;230(4):227–32.
related macular degeneration. Ophthalmology. 60. Chang W, Noh DH, Sagong M, Kim
2012;119(11):2304–11. IT. Pharmacogenetic association with early response
48. Lee AY, Raya AK, Kymes SM, Shiels A, Brantley to intravitreal ranibizumab for age-related macu-
MA Jr. Pharmacogenetics of complement factor H lar degeneration in a Korean population. Mol Vis.
(Y402H) and treatment of exudative age-related mac- 2013;19:702–9.
ular degeneration with ranibizumab. Br J Ophthalmol. 61. Habibi I, Sfar I, Kort F, et al. Y402H polymorphism in
2009;93(5):610–3. complement factor H and age-related macular degen-
49. Nischler C, Oberkofler H, Ortner C, et al. eration in the Tunisian population. Ophthalmic Res.
Complement factor H Y402H gene polymorphism 2013;49(4):177–84.
and response to intravitreal bevacizumab in exudative 62. Park UC, Shin JY, Kim SJ, et al. Genetic factors asso-
age-related macular degeneration. Acta Ophthalmol. ciated with response to intravitreal ranibizumab in
2011;89(4):e344–9. Korean patients with neovascular age-related macular
50. Imai D, Mori K, Horie-Inoue K, et al. CFH, VEGF, degeneration. Retina. 2014;34(2):288–97.
and PEDF genotypes and the response to intravitre- 63. van Asten F, Rovers MM, Lechanteur YT, et al.
ous injection of bevacizumab for the treatment of age- Predicting non-response to ranibizumab in patients
related macular degeneration. J Ocul Biol Dis Infor. with neovascular age-related macular degeneration.
2010;3(2):53–9. Ophthalmic Epidemiol. 2014;21(6):347–55.
51. Yamashiro K, Tomita K, Tsujikawa A, et al. Factors 64. Abedi F, Wickremasinghe S, Richardson AJ, et al.
associated with the response of age-related macular Variants in the VEGFA gene and treatment out-
degeneration to intravitreal ranibizumab treatment. come after anti-VEGF treatment for neovascular
Am J Ophthalmol. 2012;154(1):125–36. age-related macular degeneration. Ophthalmology.
52. Orlin A, Hadley D, Chang W, et al. Association 2013;120(1):115–21.
between high-risk disease loci and response to 65. Lazzeri S, Figus M, Orlandi P, et al. VEGF-A poly-
anti-vascular endothelial growth factor treatment morphisms predict short-term functional response
for wet age-related macular degeneration. Retina. to intravitreal ranibizumab in exudative age-
2012;32(1):4–9. related macular degeneration. Pharmacogenomics.
53. Hata M, Tsujikawa A, Miyake M, et al. Two-year 2013;14(6):623–30.
visual outcome of ranibizumab in typical neovascu- 66. Wang JJ, Rochtchina E, Smith W, et al. Combined
lar age-related macular degeneration and polypoi- effects of complement factor H genotypes, fish con-
dal choroidal vasculopathy. Graefes Arch Clin Exp sumption, and inflammatory markers on long-term
Ophthalmol. 2015;253(2):221–7. risk for age-related macular degeneration in a cohort.
54. Hagstrom SA, Ying GS, Pauer GJT, et al. Am J Epidemiol. 2009;169(5):633–41.
Pharmacogenetics for genes associated with age- 67. Boltz A, Ruiss M, Jonas JB, et al. Role of vascu-
related macular degeneration in the comparison of lar endothelial growth factor polymorphisms in
AMD treatments trials (CATT). Ophthalmology. the treatment success in patients with wet age-
2013;120(3):593–9. related macular degeneration. Ophthalmology.
55. Abedi F, Wickremasinghe S, Richardson AJ, Islam 2012;119(8):1615–20.
AF, Guymer RH, Baird PN. Genetic influences on the 68. Wickremasinghe SS, Xie J, Lim J, et al. Variants in
outcome of anti-vascular endothelial growth factor the APOE gene are associated with improved out-
treatment in neovascular age-related macular degen- come after anti-VEGF treatment for neovascular
eration. Ophthalmology. 2013;120(8):1641–8. AMD. Invest Ophthalmol Vis Sci. 2011;52(7):4072–9.
56. Teper SJ, Nowinska A, Pilat J, Palucha A, Wylegala 69. Hautamaki A, Kivioja J, Vavuli S, et al. Interleukin 8
E. Involvement of genetic factors in the response promoter polymorphism predicts the initial response
to a variable-dosing ranibizumab treatment regi- to bevacizumab treatment for exudative age-related
men for age-related macular degeneration. Mol Vis. macular degeneration. Retina. 2013;33(9):1815–27.
2010;16:2598–604. 70. Hautamaki A, Kivioja J, Seitsonen S, et al. The IL-8,
57. Kang HK, Yoon MH, Lee DH, Chin VEGF, and CFH polymorphisms and bevacizumab in
HS. Pharmacogenetic influence of L OC387715/ age-related macular degeneration. Ophthalmology.
HTRA1 on the efficacy of bevacizumab treatment for 2014;121(4):973–973.e971.
age-related macular degeneration in a Korean popula- 71. Yamashiro K, Mori K, Honda S, et al. A prospective
tion. Korean J Ophthalmol. 2012;26(6):414–22. multicenter study on genome wide associations to
68 K. Yamashiro
ranibizumab treatment outcome for age-related macu- 85. Li YJ, Goh L, Khor CC, et al. Genome-wide associa-
lar degeneration. Sci Rep. 2017;7(1):9196. tion studies reveal genetic variants in CTNND2 for
72. Okada M, Kandasamy R, Chong EW, McGuiness high myopia in Singapore Chinese. Ophthalmology.
M, Guymer RH. The treat-and-extend injection regi- 2011;118(2):368–75.
men versus alternate dosing strategies in age-related 86. Li Z, Qu J, Xu X, et al. A genome-wide association
macular degeneration: a systematic review and meta- study reveals association between common variants in
analysis. Am J Ophthalmol. 2018;192:184–97. an intergenic region of 4q25 and high-grade m yopia
73. Silva R, Berta A, Larsen M, et al. Treat-and-extend ver- in the Chinese Han population. Hum Mol Genet.
sus monthly regimen in Neovascular age-related mac- 2011;20(14):2861–8.
ular degeneration: results with Ranibizumab from the 87. Shi Y, Qu J, Zhang D, et al. Genetic variants at 13q12.12
TREND study. Ophthalmology. 2018;125(1):57–65. are associated with high myopia in the Han Chinese
74. Fung AT, Yannuzzi LA, Freund KB. Type 1 (sub- population. Am J Hum Genet. 2011;88(6):805–13.
retinal pigment epithelial) neovascularization in 88. Fan Q, Barathi VA, Cheng CY, et al. Genetic variants
central serous chorioretinopathy masquerading on chromosome 1q41 influence ocular axial length
as neovascular age-related macular degeneration. and high myopia. PLoS Genet. 2012;8(6):e1002753.
Retina. 2012;32(9):1829–37. 89. Shi Y, Gong B, Chen L, et al. A genome-wide meta-
75. Imamura Y, Iida T, Maruko I, Zweifel SA, Spaide analysis identifies two novel loci associated with high
RF. Enhanced depth imaging optical coherence myopia in the Han Chinese population. Hum Mol
tomography of the sclera in dome-shaped macula. Am Genet. 2013;22(11):2325–33.
J Ophthalmol. 2011;151(2):297–302. 90. Khor CC, Miyake M, Chen LJ, et al. Genome-wide
76. Warrow DJ, Hoang QV, Freund KB. Pachychoroid association study identifies ZFHX1B as a suscep-
pigment epitheliopathy. Retina. 2013;33(8):1659–72. tibility locus for severe myopia. Hum Mol Genet.
77. Pang CE, Freund KB. Pachychoroid neovasculopathy. 2013;22(25):5288–94.
Retina. 2015;35(1):1–9. 91. Fernandez-Robredo P, Maestre SR, Zarranz-Ventura
78. Dansingani KK, Perlee LT, Hamon S, et al. Risk alleles J, Mulero HH, Salinas-Alaman A, Garcia-Layana
associated with neovascularization in a Pachychoroid A. Myopic choroidal neovascularization genetics.
phenotype. Ophthalmology. 2016;123(12):2628–30. Ophthalmology. 2008;115(9):1632. 1632 e1631
79. Miyake M, Ooto S, Yamashiro K, et al. Pachychoroid 92. Nakanishi H, Gotoh N, Yamada R, et al. ARMS2/
neovasculopathy and age-related macular degenera- HTRA1 and CFH polymorphisms are not associated
tion. Sci Rep. 2015;5:16204. with choroidal neovascularization in highly myopic
80. Hosoda Y, Yoshikawa M, Miyake M, et al. CFH and eyes of the elderly Japanese population. Eye (Lond).
VIPR2 as susceptibility loci in choroidal thickness and 2010;24(6):1078–84.
pachychoroid disease central serous chorioretinopa- 93. Leveziel N, Yu Y, Reynolds R, et al. Genetic fac-
thy. Proc Natl Acad Sci U S A. 2018;115(24):6261–6. tors for choroidal neovascularization associated
81. Solouki AM, Verhoeven VJ, van Duijn CM, et al. A with high myopia. Invest Ophthalmol Vis Sci.
genome-wide association study identifies a suscepti- 2012;53(8):5004–9.
bility locus for refractive errors and myopia at 15q14. 94. Miyake M, Yamashiro K, Nakanishi H, et al.
Nat Genet. 2010;42(10):897–901. Evaluation of pigment epithelium-derived factor and
82. Hayashi H, Yamashiro K, Nakanishi H, et al. complement factor I polymorphisms as a cause of
Association of 15q14 and 15q25 with high myo- choroidal neovascularization in highly myopic eyes.
pia in Japanese. Invest Ophthalmol Vis Sci. Invest Ophthalmol Vis Sci. 2013;54(6):4208–12.
2011;52(7):4853–8. 95. Oishi M, Yamashiro K, Miyake M, et al. Association
83. Tedja MS, Wojciechowski R, Hysi PG, et al. Genome- between ZIC2, RASGRF1, and SHISA6 genes and
wide association meta-analysis highlights light- high myopia in Japanese subjects. Invest Ophthalmol
induced signaling as a driver for refractive error. Nat Vis Sci. 2013;54(12):7492–7.
Genet. 2018;50(6):834–48. 96. Hosoda Y, Yoshikawa M, Miyake M, et al.
84. Nakanishi H, Yamada R, Gotoh N, et al. A genome- CCDC102B confers risk of low vision and blindness
wide association analysis identified a novel suscep- in high myopia. Nat Commun. 2018;9(1):1782.
tible locus for pathological myopia at 11q24.1. PLoS
Genet. 2009;5(9):e1000660.
Kenji Yamashiro received MD from Kyoto University

Medical School in1995 and PhD from Kyoto University
Graduate School of Medicine in 2002 after joining
Angiogenesis lab in Massachusetts Eye and Ear Infirmary
at Harvard Medical School. Dr. Yamashiro is an author on
over 180 scientific publications (impact factor: over 850)
and currently a Director of Ophthalmology at Otsu Red
Cross Hospital and a Clinical Professor of Ophthalmology
at the Kyoto University Graduate School of Medicine,
Japan. His recent interest is in improving medical and sur-
gical treatment for retinal diseases.
Part II
Clinical Entities
Neovascular AMD: Clinical
Features and Imaging 7
Cláudia Farinha and Rufino Silva
7.1 Introduction patients manifest the neovascular form of the

disease, which if left untreated leads to severe
Age-related macular degeneration (AMD) is irreversible visual loss. Neovascular AMD
the leading cause of irreversible visual impair- (nAMD) is, in fact, responsible for approxi-
ment and blindness in the elderly in developed mately 80% of cases of severe vision loss due
countries [1, 2]. It is well recognized that to AMD [2]. Therefore, a worldwide increase
global population aging is expected in the next in the burden of nAMD and associated risk of
decades, and therefore, the number of patients blindness seems to be certain in the near future
affected by AMD is expected to increase sig- [5, 6].
nificantly. Recent estimates point to a projected Neovascular AMD reports to the development
number of people with AMD in 2020 of 196 of choroidal neovascularization (CNV) and the
million, and 288 million in 2040 [3, 4]. presence of its associated features, such as retinal
Importantly, approximately 10% of AMD pigment epithelial detachment (PED), retinal and
subretinal hemorrhage and fluid, fibrovascular
disciform scarring and retinal pigment epithelial
tear [6]. This designation also includes specific
C. Farinha neovascular phenotypes such as retinal angioma-
Ophthalmology Department, Coimbra Hospital and tous proliferation (RAP), or type 3 neovascular-
Universitary Centre (CHUC), Coimbra, Portugal ization, and polypoidal choroidal vasculopathy
Association for Innovation and Biomedical Research (PCV) [7–9].
on Light and Image (AIBILI), Coimbra, Portugal Histologically, a CNV is a neovascular pro-
Faculty of Medicine, University of Coimbra (FMUC), liferation that grows through breaks in the
Coimbra, Portugal Bruch’s membrane and progresses laterally
R. Silva (*) between the RPE and Bruch’s. Grossniklaus
Ophthalmology Department, Coimbra Hospital and and Gass described two different histologic
Universitary Centre (CHUC), Coimbra, Portugal types of neovascular growth: type 1 (confined
Association for Innovation and Biomedical Research beneath the retinal pigmented epithelium
on Light and Image (AIBILI), Coimbra, Portugal (RPE)) and type 2 (when the neovascularization
Faculty of Medicine, Coimbra Institute for Clinical grows between the neurosensory retina and the
and Biomedical Research. University of Coimbra RPE) [10, 11]. However, the two types can
(iCBR- FMUC), Coimbra, Portugal
often coexist in AMD (mixed CNV). In 1996,
Coimbra Medical Space, Coimbra, Portugal Hartnett et al. [12] described a new distinct
e-mail: rufino.silva@oftalmologia.co.pt

74 C. Farinha and R. Silva
phenotype of neovascularization, which origi-

nates from the deeper retina and eventually
extends posteriorly toward the choroid, this was
later denominated by Yannuzzi and colleagues
as RAP, and more recently as type 3 neovascu-
larization [7, 13]. The diagnosis of different
neovascular histologic subtypes in AMD is
nowadays regarded as non-crucial in deciding
treatment, as all neovascular AMD can be
treated with anti-vascular endothelial growth
factors (VEGF). However, it may be important
for predicting therapeutic response and progno-
sis and for consideration of adjunctive photody-
namic therapy in PCV cases [14].
Fig. 7.1 Clinical presentation of nAMD. Retinal edema
and elevation associated with retinal hemorrhages, hard
exudates, and subretinal fibrosis are typical signs of late
7.2 Clinical Presentation exudative AMD
of Neovascular AMD
Neovascular AMD presents clinically by typical 7.2.1 Pigment Epithelial

signs and symptoms, usually in the elder patient. Detachment
Blurred or decreased vision and distortion in the
image (metamorphopsia), especially when read- A PED refers to the anatomical separation of the
ing, are the main complaints. Micropsia and cen- RPE from the underlying Bruch’s membrane. In
tral scotoma can be referred, but sometimes there neovascular AMD they can be serous, fibrovascu-
may be no complaints at all. The patient can be lar (type 1 CNV), mixed (with both serous and
truly asymptomatic, or the visual changes are too fibrovascular components), and might also be
subtle in the early stages for the patient to notice hemorrhagic as a complication of the associated
them. CNV [15, 16]. In eyes with nAMD, it is not
In fundoscopy, there are several features that uncommon to see more than one type of PED.
can be appreciated. The CNV may appear as a On funduscopic examination, a serous PED is
gray lesion located deep to the retina, elevating seen as a sharply demarcated circular or ovoid
it, and often with associated presence of blood dome-shaped elevation of the RPE with a clear or
(intra and/or subretinal), hard exudates, and yellowish–orange color. When chronic there may
intraretinal edema with or without a neurosen- be pigment migration and deposition in a reticu-
sorial detachment (Fig. 7.1). This gray color of lar or radial pattern over the PED [6]. The serous
the CNV is not always seen, and the presence of PED is usually related to CNV in nAMD and an
the remaining features in an older patient with associated notched appearance is considered typ-
AMD and vision loss indicates the presence of ical. However, avascular PEDs may also be seen
the membrane. Associated serous or hemor- without evidence of CNV. Of patients with neo-
rhagic PED is also typical, with the latter being vascular AMD only 1% present with a pure
more frequent in the context of PCV. In other serous PED, whereas 31% present with a vascu-
cases, the only presenting sign is a small neuro- larized serous PED [15]. The natural history of a
sensory detachment, and a careful analysis of serous PED is relatively more favorable, com-
the fundus together with a high clinical suspi- pared to vascularized PEDs associated with type
cion and imagiological ancillary tests is funda- 1 (sub-RPE) neovascularization, which have a
mental to the correct diagnosis and timely high risk for vision loss. Subfoveal avascular
treatment [6]. serous PEDs can, in fact, remain unchanged over
7 Neovascular AMD: Clinical Features and Imaging 75
significant periods of time and visual acuity may CNV, leaving bare Bruch’s membrane and cho-
stabilize without treatment [15]. roid in the area exposed, which gives rise to a
A hemorrhagic PED, this is blood within a sharply depigmented area, usually with a
PED, implies a diagnosis of CNV. It appears as crescent-
shaped configuration. The curled and
an elevated, dark-red-greenish mound. The hem- rolled RPE is seen adjacent to this hypopig-
orrhage can dissect through the RPE into the sub- mented area as an area of increased brownish
retinal space or into the retina. Rarely, it may pigmentation [15]. Sometimes concentric tears
pass through the retina into the vitreous cavity, may develop and only an island of scrolled RPE
causing hemovitreous [6]. remains at the center of the tear. Diagnosing
In some cases, a pure fibrovascular PED small tears is easier with fundus autofluorescence
(fvPED) can present with a homogeneous dark imaging, as the tear crescent area shows profound
red color in fundoscopy. This feature can be mis- hypoautofluorescence, with a slight increase in
interpreted as blood in the sub-RPE space; how- autofluorescence in the area of the torn and
ever, it corresponds only to an engorged retracted RPE [17].
neovascular complex with large-caliber vessels.
These vessels may be more mature and possibly
contain stagnated blood that appears dark red. 7.2.3 Disciform Scar
One should be aware of this variant of fibrovas-
cular PEDs because these large neovascular ves- A CNV is a fibrovascular tissue since the neoves-
sels may bleed and necessitate aggressive sels are accompanied by fibrous tissue. With
treatment [15]. time, and despite treatment, the fibrotic turnover
might progress, and the plane of the RPE is
destroyed, making the CNV classification by
7.2.2 Retinal Pigment type no longer discernible. When the fibrous tis-
Epithelial Tear sue becomes clinically visible, the CNV and
fibrous tissue complex are called a disciform
Tears of the RPE were first described in 1981 by scar.
Hoskin and colleagues [17]. A retinal pigment Clinically, disciform lesions vary in color
epithelial tear usually occurs in an eye with a from white to yellow. Hyperpigmented areas may
serous or fibrovascular PED, and can occur in up be present depending on the degree of RPE
to 27% of eyes, depending on the study popula- hyperplasia within the scar tissue. Disciform
tion. The spontaneous tear rate in the natural his- scars may continue to grow, with new areas of
tory of fvPEDs in nAMD has been reported to be neovascularization proliferating along their edge.
between 10% and 12.5%. There are conflicting Reading vision is severely compromised at this
reports on the incidence of RPE tears after anti- point and rarely is better than 20/200 [6].
VEGF therapies, but it is estimated to affect 15%
to 20%, especially when the PED is 600 μm or
greater in height. Visual acuity usually drops sud- 7.3 Imaging in Neovascular AMD
denly with the event if the tear involves the foveal
center; however, it is often stabilized if anti- Imaging is essential in the diagnosis and treat-
VEGF therapy is continued [16, 17]. ment of nAMD. Fluorescein angiography (FA) is
A RPE tear probably results from growing traditionally considered the gold standard for the
tension in the junction of the attached and diagnosis, because it can provide useful informa-
detached areas of the RPE, from increasing fluid tion concerning lesion subtype and activity
beneath it and/or from contraction of an associ- through the dynamic visualization of dye leakage
ated CNV when an RPE-adherent fibrovascular [18]. In the last years, however, Spectral Domain
component is present in the PED [6]. The edge of Optical Coherence Tomography (SD-OCT) has
the elevated RPE curls in the direction of the become an indispensable tool in the diagnosis
and follow-up of nAMD, as it allows for noninva- of dye leakage. Other features of an active CNV
sive high-resolution visualization of the CNV such as neurosensorial detachment and serous
and associated features related to angiographic PED are easily assessed by FA, as dye pools in
leakage, such as intraretinal and subretinal fluid these potential spaces, with increased fluores-
[19]. In the clinical setting, other ancillary exams cence through the angiogram [18].
used in nAMD management include indocyanine The nomenclature of classic vs occult CNV
green angiography (ICGA) and more recently provided by FA is, however, becoming less used
OCT Angiography (OCT-A). in the clinical setting because of the relatively
OCT Angiography is a new technology that is uniform treatment given to these patients with
rapidly growing and gaining popularity in the anti-VEGFs, and because of the major role attrib-
daily practice of AMD clinics because it allows uted to SD-OCT in the diagnosis and follow-up
to noninvasively detect the neovascular network, of nAMD. Despite this, the knowledge is still
giving information on flow and vessel density, essential because response to treatment may be
while simultaneously providing all the structural different on different lesion subtypes, and
B-scan information of a conventional OCT because most clinical trials still rely on these
device. angiographic features for the inclusion of
patients.
7.3.1 Fluorescein Angiography 7.3.1.1 Classic CNV

A classic CNV is seen in the early phases of
As stated above, FA is still considered today the the angiogram as a hyperfluorescent area, with
gold standard for the diagnosis and management well-
defined borders, sometimes surrounded
of nAMD in clinical practice. In the era of OCT, by a hypofluorescent halo. The individual ves-
FA remains important in CNV diagnosis because sels in the CNV may be seen, often with a
it reduces the possibility of error or misdiagnosis. “wagon wheel” appearance composed by the
In fact, multimodal imaging including FA and feeding central vessels and the centripetally
OCT is currently the state of the art in the assess- oriented neovessels that radiate from them.
ment of nAMD. Fluorescein angiography is also The hyperfluorescence associated with the
used to this day in major nAMD clinical trials classic CNV typically increases as the angio-
and the imaging features analyzed are recognized gram progresses, with marked leakage extend-
as valid and reliable endpoints [20]. ing beyond the borders of the CNV in the late
The study of a CNV lesion with FA allows the frames (Fig. 7.2) [6, 21–23].
clinic to determine patterns of fluorescence such
as classic and occult, and the relative composi- 7.3.1.2 Occult CNV
tion (predominantly classic, minimally classic, The term occult CNV reports to two patterns in
and occult CNV). There is an assumed relation- angiography: fibrovascular PED and late leakage
ship between the histopathologic types 1 and 2 of undetermined source (LLUS).
and the occult and classic CNV patterns, respec- A fvPED is best appreciated at 1–2 min in the
tively, however, there is little evidence to support angiogram, as an irregular elevated area of stip-
the universality of this correlation. FA also allows pled or pinpoint-like hyperfluorescence, with or
to analyze on the boundaries of the lesion as without leakage or with staining in the late
“well defined” or “poorly defined” and the loca- frames. The hyperfluorescence, in this case, is
tion of the CNV in respect to the fovea (subfo- much weaker compared to classic CNVs, and the
veal, juxtafoveal if the border is within 199 μm of borders of the detachment are often indistinct
the center, and extrafoveal if the margin is at least (Fig. 7.3). Fibrovascular PEDs must be distin-
200 μm from the foveal center) [6]. Fluorescein guished from purely serous PEDs, but it is not
angiography is particularly useful in assessing uncommon to visualize a serous PED contiguous
CNV activity through the dynamic visualization with a fvPED.
a b c
Fig. 7.2 Classic CNV in FA. (a) The classic lesion pres- further enhancing the visualization of the neovessels;
ents with well-defined limits in early phase and the neo- (c) in late phase there is intense leakage obscuring the
vascular network is seen with great detail; (b) in CNV borders and neovascular network details
intermediate phase it increases significantly in brightness
a b c
Fig. 7.3 Multimodal imaging of occult CNV/type 1 CNV. (d) SD-OCT scans show the presence of the fvPED cor-
(a) CFP image shows the presence of a PED in the central responding to type 1 CNV, seen as irregularly elevated
macula with soft drusen and subretinal fluid, and an infe- RPE with underlying heterogenous laminar tissue, sepa-
rior subretinal hemorrhage with dehemoglobinized blood; rating it from the Bruch’s membrane; there is a fluid cleft
(b, c) FA shows occult CNV with fibrovascular PED pat- beneath the fibrovascular tissue (triple layer sign) (yellow
tern as an irregular area of stippled hyperfluorescence in arrows). Subretinal fluid involving the foveal center and
intermediate transit phase, with persistent staining and subretinal hemorrhage, hyperreflective, and blocking the
discrete leakage in late phase; there is also blocked fluo- RPE and choroidal details, are also seen (green arrow)
rescence of the choroid from the subretinal blood;
The LLUS term is used when there is late 7.3.3 Optical Coherence
choroidal-based leakage with no clearly corre- Tomography
sponding identifiable classic CNV or fvPED in
the early or mid-phase of the angiogram. It Optical coherence tomography has become an
appears as speckled hyperfluorescence, often indispensable tool in the assessment of retinal
with overlying subretinal pooling of dye, and the pathology, and by extension in the diagnosis
CNV boundaries cannot be determined precisely and follow-up of patients with
[6, 15, 21–23]. nAMD. Compared to dye-based FA and ICGA,
OCT is a noninvasive and less-time consuming
exam that gives precious information on the
7.3.2 Indocyanine Green morphologic microstructural characteristics of
Angiography a CNV and associated features, and their
response to treatment. Currently, two OCT
Indocyanine green is a dye that is more protein- technologies are in wide-spread use: the spec-
bound than sodium fluorescein and fluoresces in tral-domain OCT (SD-OCT), that operates at
the near-infrared wavelength, which enhances its scanning rates of approximately 27,000–
transmission through the RPE, but also through 70,000 A-scans/second; and the swept source
blood and exudative material, thus providing a OCT (SS-OCT). The latter has faster scanning
more detailed visualization of the choroidal vas- speeds (100,000–400,000 A-scans/second) and
culature and its pathology [23, 24]. operates at a longer wavelength, increasing the
ICGA is very useful in the context of nAMD visualization of structures beneath the RPE
for accurate detection of occult CNV, especially [26, 27].
in cases that otherwise would be difficult to dis- In nAMD, the use of OCT combined with
cern using only FA [25], for example, when FA and ICGA provided a better understanding
occult CNV does not fluoresce enough to be on how to correlate the abundant information
clearly seen in FA or when a serous or hemor- from multimodal imaging with the previous
rhagic PED mask the underlying membrane. histologic findings described by Grossniklaus
ICGA is also of great interest in diagnosing spe- and Gass on types 1 and 2 neovascular mem-
cific neovascular phenotypes such as RAP branes [10, 11]. This multimodal approach is
lesions, especially in the early stages, and PCV superior to FA alone to determine the presence,
[20]. ICGA is the gold standard for the diagnosis location and size of CNV, and these are factors
of the latter because it allows the clear detection that may be predictive of prognosis and treat-
of polyps or aneurysmatic dilatations, as more ment response.
recently described. Thus, a new imagiological CNV classification
emerged based on this OCT-driven multimodal
ICGA Patterns in nAMD: approach: type 1, type 2 and type 3 neovascular-
• Plaque: The most common pattern, is a hyper- ization [23]. Type 1 and type 3 neovascular mem-
fluorescent area larger than one disk area, usu- branes are the most prevalent subtypes of
ally well defined, that shows relatively late in neovascular AMD, when using FA and OCT
the angiogram. The plaques may exhibit pro- combined, and are estimated to represent 40%
gressive growth. and 34% of cases, respectively. Pure type 2
• Hot spot or focal CNV: When lesions appear lesions account for only 9% and mixed
as occult CNV on FA but demonstrate hyper- membranes are seen in 17% of cases, as reported
fluorescence in ICGA of less than one disk by Jung et al. [28].
area. They usually show by the mid-phase of Type 1 and 2 lesions characteristics in OCT
the angiogram and often correspond to ele- are presented below. Type 3 neovascular mem-
ments of RAP or PCV [23]. brane or RAP and PCV, as a variant of type 1,
Area of ill-defined fluorescence. will be discussed separately.
7.3.3.1 Type 1 Neovascular Membrane by OCT-A. Grossniklaus and Green hypothesized

Corresponds to a CNV that occurs and grows that the type 1 CNV may be a compensatory
beneath the RPE, creating a separation between response supporting an ischemic outer retina, and
the RPE and the Bruch’s membrane. A type 1 theoretically that could protect against the advent
CNV corresponds to a vascularized PED and is or progression of geographic atrophy [33].
the most common in nAMD. Eyes become symptomatic when subretinal or
On OCT it appears as moderate to high reflec- intraretinal exudation develops and this is seen in
tive material located between the RPE and the OCT as subretinal (SRF) or intraretinal fluid
Bruch, clearly separating them. Some authors (IRF). IRF occurs either diffusely, with increased
have called this the “double layer sign” when it is retinal thickness and reduced retinal reflectivity, or
relatively thin [29]. This fibrovascular tissue is localized in nonreflective cysts (cystic macular
seen with various degrees of thickness and vol- edema). Subretinal fluid is the predominant form
ume, usually with a fusiform or nodular shape. Its of exudation in type 1 CNV. OCT is thus of great
inner reflectivity varies according to the composi- value in assessing the CNV activity during follow-
tion, typically having a heterogeneous multi- up [15]. Although the clinical course is variable, a
laminar structure with moderate to high reflectivity, type 1 CNV tends to behave less aggressively than
that adheres to the elevated RPE (Fig. 7.3). Spaide types 2 and 3, as these cause more active exuda-
et al. [30] proposed that the CNV vessels adhere to tion and more rapid vision loss [15].
the basal surface of the RPE when the RPE A type 1 CNV corresponds usually to an
becomes elevated by sub- RPE exudation. occult CNV in FA (both fvPED and LLUS), and
Hyporeflective oval spaces and channels may be to a plaque pattern in ICGA.
seen within the fibrovascular tissue and are thought
to correspond to the neovessels’ vascular lumen. 7.3.3.2 Type 2 Neovascular Membrane
Fluid clefts are also observed as optically Corresponds to a CNV that has penetrated the
empty spaces within or beneath the fibrovascular RPE/Bruch’s membrane complex and prolifer-
tissue. Khan et al. [31] described the “triple layer ates in the subretinal space. Pure type 2 CNV is
sign” when a hyporeflective space or cleft rare in nAMD and Naysan and colleagues [34]
appears immediately underneath the fibrovascu- reported that type 2 CNV occurs almost exclu-
lar multilayered PED. The “onion sign” was sively in AMD patients with pure subretinal
described by Mukkamala et al. [32], and corre- drusenoid deposits (reticular pseudodrusen) phe-
sponds to a pattern of layered hyperreflective notype and thin choroids. In fact, in AMD a type
bands beneath the RPE line and within the fvPED 2 CNV is usually seen in association with a type
that correlates clinically with areas of yellow 1 that has grown vertically to penetrate the RPE
exudates. They hypothesized that these hyperre- and is now proliferating in the subretinal space,
flective bands correspond to exudation of lipid in this is, a mixed CNV (Fig. 7.4) [23, 28].
the sub-RPE space from the CNV. In OCT type 2 neovessels are seen above the
A type 1 CNV may be associated with larger RPE band and beneath the photoreceptor outer seg-
serous components, and the CNV may be seen ments. The CNV is seen as moderately to hyperre-
adjacent to the borders of the serous detachment flective material that may appear more laminar in
and/or as hyperreflective material that lines along structure or with a more amorphous appearance
the back surface of the elevated PED. The appear- with fuzzy borders. This depends on the size,
ance of vascularized and serous PEDs may be degree of exudation, and associated features such
similar in OCT and FA/ICGA is essential to iden- as fibrin deposition or presence of subretinal hem-
tify the associated CNV. orrhage. Usually, there is disorganization of the
Type 1 CNVs may be non-exudative, slowly photoreceptors’ layers with disruption of the ellip-
enlarging over years, with the patient remaining soid zone and external limiting membrane. In this
asymptomatic. This silent growth can be nonin- context, intraretinal cystic edema is common and
vasively monitored by OCT, and more recently predominates over subretinal fluid [23].
a b
c d e
Fig. 7.4 Multimodal imaging of mixed type 1 and 2 shows a classic CNV in the early phase corresponding to
CNV. (a) CFP image shows the presence hard exudates in the type 2 lesion, with well-defined limits and a surround-
the macula and some degree of subretinal fibrosis tempo- ing hypofluorescent halo; (d, e) in intermediate and late
ral to the fovea; (b) SD-OCT scan shows the presence of a FA phases occult CNV (type 1) is seen as irregular, stip-
fvPED corresponding to type 1 CNV, in a double layer pled hyperfluorescence adjacent to the classic CNV and in
configuration (yellow arrow), with overlying subretinal the nasal fovea, with leakage in late phases less intense
type 2 CNV (green arrow) and subretinal fluid; (c) FA compared to the classic component
As the CNV matures, progression to subreti- PED types that can be present in AMD. These
nal fibrosis may ensue and appear on OCT as a include drusenoid PED not associated with CNV,
hyperreflective band overlying the RPE with sub- fibrovascular PED or type 1 CNV, serous PED
stantial disorganization and thinning of the outer (vascularized or avascular), and hemorrhagic
retina. PED [6, 15]:
A type 2 CNV corresponds usually to a classic
CNV in FA. In ICGA, the vascular networks are –– Drusenoid PEDs are not a feature of nAMD
seen but can be difficult to detect overlying the but they will be referred as it is important to
intense hyperfluorescence of the background accurately differentiate them from fvPEDs.
choroidal circulation. A hypofluorescent outer They are seen as an extensive area of large
border can surround the CNV [21]. confluent soft drusen, often with overlying
reticulated pigment clumping and a scal-
loped border. They fluoresce faintly during
7.3.4 ther Features of nAMD
O the angiogram and become less fluorescent
in Multimodal Imaging in the late phase. They do not progress to
bright hyperfluorescence like a serous PED,
7.3.4.1 Pigment Epithelial neither present with stippled hyperfluores-
Detachments cence, differentiating them from
The FA pattern through the angiogram together fvPED. Focal hypofluorescent areas corre-
with OCT and, whenever necessary, ICGA, spond to the blocking effect of overlying
allows the differentiation between the different pigment hyperplasia [35].
–– In ICGA using a confocal scanning laser oph- –– On OCT a serous PED appears as a smooth,
thalmoscope (SLO) system, the PED will sharply demarcated dome-shaped elevation of
appear as a homogeneous hypofluorescent the RPE, overlying a homogeneously hypore-
area during the transit, because the drusen-like flective space, with a clearly visible Bruch’s
content will block the fluorescence from the membrane beneath it, seen as a thin hyperre-
choroidal vasculature. With ICGA using a tra- flective line [37].
ditional fundus camera-based system the –– A hemorrhagic PED presents in FA as a
drusenoid PED can appear isofluorescent or dark area through the angiogram, as the
slightly hypofluorescent. blood beneath the RPE will block choroidal
–– On OCT the PED will appear as a smooth fluorescence. In ICGA, a focal or plaque-like
elevation of the RPE band with or without area of hyperfluorescence corresponding to
undulation, with material underneath that the CNV may be seen through the blood or at
typically has a dense homogeneous appear- its border. The OCT shows a dome-shaped
ance with moderate or high reflectivity. Sub- elevation of the RPE with the blood beneath
RPE fluid pockets above the deposited appearing hyperreflective. Because of this,
material may be present. Hyperreflective foci there is substantial attenuation of the signal
with posterior shadowing are commonly seen from deeper structures and the sub-RPE
immediately above the PED and correspond CNV, as well as the Bruch’s membrane and
to pigment clumping from RPE in histologic the choroid, are not easily discerned or not
examination [36]. Subretinal or intraretinal seen at all.
fluid is not typical and should raise suspicion
of CNV. However, there may be a small 7.3.4.2 RPE Tear
pocket of benign subretinal fluid or an In FA, the areas where the RPE is missing cause
acquired vitelliform lesion. This must be dif- a window defect showing the early choroidal
ferentiated from an active CNV as treatment fluorescence, with staining till the late frames,
is not indicated. but generally there is no leakage from the stripped
–– A serous PED appears in FA as a uniform area. The area with the retracted and curled RPE
bright hyperfluorescent area in the early phase blocks the background fluorescence, thus being
with smooth borders, increasing in intensity hypofluorescent through all angiographic exami-
toward the late phase, with little or no leakage nations. There may be leakage from this area
along its borders. A serous PED might obscure later from an underlying CNV.
any CNV contained in its area. However, a Sarraf et al. [38] published a grading system
typical fluorescein pattern of the vascularized for RPE tears that is based on the greatest linear
serous PED is when a serous PED has a notch diameter of the tear measured with FA: grade 1
or hot spot corresponding to CNV. The FA are those <200 μm, grade 2 when between
shows early intense hyperfluorescence and 200 μm and 1 disk diameter, grade 3 if >1 disk
late homogeneous pooling of the serous com- diameter, and grade 4 if they are >1 disk diameter
ponent while the notch has stippled hyperfluo- and involve the center of the fovea. This grading
rescence indicating the associated occult system was shown to have prognostic signifi-
membrane. cance in the response to treatment with anti-
–– In ICGA, the serous PED is hypofluorescent VEGF and functional outcome.
in both the early and the late phases when In ICGA, the bare choroid is hypofluorescent
using a confocal SLO system, because fluid or isofluorescent, and the rolled RPE is moder-
blocks the normal fluorescence of the cho- ately hypofluorescent [17].
roid. If there is early hyperfluorescence and The OCT shows disruption of the RPE
late staining adjacent to the serous PED or monolayer in the PED, varying from a micro-
within, this corresponds to associated occult scopic defect to a large discontinuity. The
CNV. retracted RPE is seen with an irregular shape,
often with the free edge curled under the PED, On OCT the scar corresponds to a highly
showing dense hyperreflectivity from duplica- reflective outer retinal or subretinal lesion. There
tion of the RPE, and a shadowing effect beneath is extensive damage to the photoreceptors inner
it. The bare choroid area shows hyperreflectiv- and outer segments with disruption of the ellip-
ity as a result of deeper penetration due to the soid zone and external limiting membrane and
absence of the RPE. The retina overlying this thinning of the outer nuclear layer [37]. At this
zone remains intact with or without associated point, outer-retinal tubulations (ORTs) may
subretinal fluid. OCT is the preferred method appear over the scar and in associated areas of
for definite diagnosis of RPE tear (Fig. 7.5) outer retina and RPE atrophy (Fig. 7.6). ORTs
[17, 37]. should be differentiated from intraretinal fluid
from exudation. They were first described histo-
7.3.4.3 Disciform Scarring logically as interconnecting tubes containing
Usually the disciform scar is hyperfluorescent degenerate photoreceptors and enveloping Müller
from both fluorescein dye leakage and staining. cells in AMD patients. In OCT ORTs appear as
The fluorescence may fade away in the late round or oval lesions with hyporeflective lumen
frames or increase, if there is still exudation with and hyperreflective borders located in the outer
fluid. Hypofluorescent areas may be seen when nuclear layer, overlying an RPE that is either dis-
RPE pigmentary reaction blocks fluorescence. In rupted or absent. They may form branching net-
ICGA anastomoses between the retina and the works of channels best appreciated in en face
choroid are often observed [21]. OCT [39, 40]. Degenerative cysts, usually small
a b
Fig. 7.5 RPE tear. (a) CFP shows the ruptured and SD-OCT B-scan shows the elevated fibrovascular PED
retracted RPE as an area of increased pigmentation (white with RPE retraction and shadowing effect over the under-
arrow) and an associated geographic area of depigmenta- lying choroid from the folded RPE (white arrow) and the
tion from denuded choroid (yellow arrow); fibrosis and area of bare choroid, where the RPE is absent, with
retinal hemorrhages are also seen; (b) the corresponding increased reflectivity (yellow arrow)
Fig. 7.6 Disciform scar in OCT. The scar is seen as a and thinning of the outer nuclear layer. Outer-retinal tubu-
highly reflective subretinal lesion with disruption of the lations with hyperreflective borders are seen over the
overlying ellipsoid zone and external limiting membrane fibrotic lesion (yellow arrow)
and more quadrangular in shape, may also appear presence of reticular pseudodrusen and thinner
overlying the scar and/or atrophy and are thought choroid also seems to be associated with RAP
to be related to a degenerative process, perhaps development [8].
Müller cell degeneration, rather than to exuda- Clinically, RAP lesions are usually seen in the
tion. In accordance with this, leakage is usually parafoveal region presenting with retinal edema,
absent in FA [41]. Both ORTs and degenerative hard exudates, and preretinal, intraretinal, and
cysts detected in OCT are important for the clini- subretinal small spot hemorrhages. Sometimes it
cian to recognize because they can interfere with is possible to see the deepening of the dilated
treatment decisions. anastomotic vessel at 90° orientation, typically
over a PED. Retinal–retinal anastomosis is also
seen in RAP. When a larger fibrovascular PED
7.4 Specific Neovascular develops in later stages other features of nAMD
Phenotypes in AMD may appear making the diagnosis of this distinct
phenotype more difficult based only in fundos-
7.4.1 Retinal Angiomatous copy [13, 15, 46].
Proliferation (Type 3 On fluorescein angiography, a RAP lesion in
Neovascularization) the initial stage typically presents as a focal area
of hyperfuorescence in the early phase located
First described by Hartnett et al. [12] as intrareti- outside the foveal avascular zone. This focal area
nal vascular proliferation without any choroidal progressively increases in brightness during the
vascular involvement or “deep retinal anomalous angiogram with leakage in the late frames. A
complexes,” this type of neovascularization was retinal-retinal or retinal–choroidal anastomosis
later denominated by Yannuzzi and colleagues as may be seen. If a fvPED develops from progres-
retinal angiomatous proliferation or RAP [13]. sion of the type 3 membrane, signs of occult
There was, however, doubt on the primary origin CNV will appear and diagnosis of RAP becomes
of the lesion: the retina or the choroid, the latter difficult with FA only. Both intraretinal leakage
being proposed by Gass [42]. Studies with with pooling in cystoid spaces and subretinal
SD-OCT imaging concluded that the neovascular leakage may be seen.
process could originate from either the retina or ICG angiography is very useful in identifying
choroid, but histopathologic studies suggest that the intraretinal neovascular complex as a
neovascularization is in the retina [7, 43, 44]. characteristic “hot spot.” Another typical feature
Following the intraretinal origin, a RAP staging is the intraretinal leakage in the late phase of the
system was proposed by Yannuzi et al. [13]: ICGA into the cystoid macular edema (CME)
Stage 1—intraretinal neovascularization; Stage spaces, presumably because of the presence of
2—subretinal neovascularization with retinal– fibrin in these and the affinity of the ICG mole-
retinal anastomosis; and Stage 3—neovascular- cule to fibrin [7, 8, 13].
ization with a vascularized pigment epithelial Findings in OCT also depend on the stage of
detachment and a retinal–choroidal anastomosis. the RAP lesion. An incipient RAP lesion can be
Despite this, RAP was more recently proposed seen as a punctate or focal intraretinal hyperre-
to be designated as type 3 neovascularization, flective lesion located in the outer retina above
following the logic of type 1 and 2 neovascular- the external limiting membrane (ELM), overly-
ization classification [7, 13, 23, 43]. ing an area of outer nuclear layer thinning and
RAP is estimated to represent about 10% to photoreceptors loss, or overlying a PED. There
15% of new nAMD cases. Another important can be associated CME at this point (Fig. 7.7)
notion is that a patient with RAP in one eye has a [37]. In more advanced stages, the most com-
high chance of developing another RAP lesion in mon appearance is that of a serous PED (or
the fellow eye. This is estimated to be 80% after sometimes a fvPED) with overlying CME. It is
1 year, and increases to 100% at 3 years [45]. The possible to observe a focal break in the RPE
band of the PED through which a continuous Recently OCT-A proved to be useful not only
hyperreflective band connects the intraretinal in the diagnosis but also in the study of RAP
component to the sub-RPE component. A dis- pathophysiology. Characteristic findings of RAP
tinctive feature is the disruption of the external in OCT-A will be discussed further below.
limiting membrane that accompanies the CME
above the PED as opposed to a Type 2 CNV that
is usually associated with a neurosensory detach- 7.4.2 Polypoidal Choroidal
ment and an intact ELM. Even in highly active Vasculopathy/Aneurysmal
lesions with extensive sub-RPE fluid and CME, Type 1 Neovascularization
the latter can extend greatly along the outer
plexiform layer without significant subretinal Polypoidal choroidal vasculopathy is currently
fluid [8, 15, 47]. considered a subtype of type 1 neovasculariza-
Fig. 7.7 Multimodal imaging of stage 1 RAP lesion pro- increased reflectivity extending to the RPE, accompanied
gression over 2 months. (a) At baseline, SD-OCT scan by the development of cystoid spaces corresponding to
shows drusenoid-like PED with relatively homogenous cystoid macular edema (orange arrow); CFP shows only
content and hyperreflective foci overlying the RPE (yel- increased pigmentation in the area above the soft drusen
low arrow). (b) after 2 months, one can see in the same and ICGA (top, right) shows a small hot spot representing
orientation SD-OCT scan the disruption of the RPE below the incipient RAP lesion (yellow arrow)
the hyperreflective foci, with an ill-defined band of
tion in which polypoidal or aneurysmal vascular clearly image the polypoidal lesion. FA usually
dilations develop, usually in the fringe of the neo- shows only occult CNV patterns with intermedi-
vascular network or “branch vascular network” ate or late leakage from undetermined source.
(BVN). But it can be useful to analyze the lesion’s activ-
This entity was first coined by Yannuzi and ity, as leakage is better imaged in FA compared to
colleagues as Idiopathic Polypoidal Choroidal ICGA [9, 52].
Vasculopathy in 1982, as it presented in a group ICGA is still considered the gold standard in
of patients with distinct demographic characteris- PCV diagnosis as it allows to image both the
tics—younger middle-aged women of black BVN and the aneurysmal dilatations or polyps.
American background—and had better prognosis Single or multiple polyps organized in grape-like
compared to “conventional” nAMD [48, 49]. clusters are better seen in the early phase of
Subsequent reports extended the superior preva- ICGA and make the diagnosis [37]. The area sur-
lence to the Asian population and recent studies rounding the vascular network is hypofluorescent
have shown that a greater proportion of these in early phases but in the late phases there is often
patients may be white and older. a reversal of the pattern with hyperfluorescence
Controversy persists if PCV is a subset of around the polypoidal lesion and hypofluores-
nAMD or a separate entity as polypoidal lesions cence in the center. In very late phases ICGA
can be seen in the context of other clinical enti- shows washout in non-active lesions.
ties, such as in the pachychoroid spectrum [50]. The EVEREST study group defined in the
Besides, PCV was shown in multimodal imaging grading protocol PCV as the presence of early
studies to represent a neovasculopathy, rather subretinal ICGA hyperfluorescence (appearing
than a distinct abnormality of the inner choroidal within the first 5–6 minutes of dye injection) and
vasculature as initially proposed. This led to the at least one of the following criteria: (1) nodular
notion that PCV should not be seen as a single appearance of the polyp on stereoscopic viewing;
disease, but as a common manifestation of many. (2) hypofluorescent halo around the nodule; (3)
In this regard, PCV is seen today as a variant of abnormal vascular channel(s) supplying the pol-
type 1 CNV predominantly affecting eyes with yps, this is, the BVN; (4) pulsatile filling of pol-
longstanding neovascularization, and the desig- yps in dynamic ICGA; (5) orange subretinal
nation of Aneurysmal Type 1 Neovascularization nodules in the fundus corresponding to the hyper-
was introduced [31, 51]. Being a type 1 CNV, fluorescent nodular areas in ICGA; and (6) mas-
both the BVN and the polyps or aneurysms are sive submacular hemorrhage. Nodular
located in the sub-RPE space, but rarely the pol- hyperfluorescence on stereoscopic ICGA was
yps can erode through the RPE and are seen as a seen in the EVEREST study in 91.8% of cases, a
type 2 lesion [15, 31]. hypofluorescent halo in 68.9%, and pulsatile pol-
Clinically the PCV patient presents with mul- yps in only 6.6%. Byeon et al. described pulsatile
tiple, recurrent, sero-hemorrhagic pigment epi- polyps as more prone to rupture and causing
thelial and neurosensorial detachments, severe hemorrhage [54, 55].
sometimes with intraretinal lipid deposition, OCT findings are highly reliable in suggesting
around the optic nerve or in the central macula. a PCV phenotype: an isolated highly protruded
Orange nodules are sometimes seen in the bor- RPE with underlying moderate reflectivity, some-
ders of the lesion, or within the top of the PED, times round shaped with hyperreflective borders
and correspond to polyps in ICGA. Vitreous and hyporeflective lumen; a notched PED; nodular
hemorrhage is possible when subretinal hemor- structures adhering to the elevated back surface of
rhage is significant. PCV often follows a the RPE like “pearls on a string”—all these fea-
remitting-relapsing course and bilateral involve- tures corresponding to the polyps; and the double-
ment is common (up to 86%) [52, 53]. layer sign, composed by the hyperreflective lines
The use of fluorescein angiography in PCV formed by the separation of the RPE and Bruch’s
diagnosis is limited because it does not allow to membrane, with an intervening layer of moderate
to high reflectivity that corresponds to the BVN/ 7.5 Optical Coherence

type 1 CNV [31]. De Salvo and colleagues reported Tomography Angiography
that with a combination of OCT-based signs: mul- in nAMD
tiple PEDs, sharp PED peak, PED notch, and
rounded sub-RPE hyporeflective lesions, the sen- OCT angiography is a new imaging modality that
sitivity and specificity of PCV diagnosis with uses the OCT principles to detect blood flow
SD-OCT alone was of 94.6% and 92.9%, respec- within the vascular structures of the eye without
tively [56]. Therefore, the OCT can be very help- the need to intravenously administer fluorescent
ful as a screening tool for PCV suspected cases dyes. Currently, available OCT-A devices pro-
that should undergo ICGA procedure for confir- vide detailed 3-dimensional volumes of the reti-
mation, as ICGA is not routinely performed in the nal and choroidal vasculatures that can be
majority of clinics worldwide (Fig. 7.8). analyzed layer-by-layer in both en-face or cross-
OCT is also very useful to detect associated sectional display, simultaneously co-localizing
intraretinal, subretinal, and sub-RPE fluid, thus blood flow with microstructural changes in con-
inferring on lesion activity. In PCV different ventional OCT B-scans.
components of the lesion may be active or inac- Technically, OCT-A refers to a group of
tive. The predominant active component can be OCT imaging methods that allow the visualiza-
determined by co-localization of fluid in OCT tion of moving particles, such as erythrocytes in
with the polyps, the BVN, or both in ICGA. This biological tissues. These methods ignore the
multimodal analysis of the lesion activity is quantitative measurement of blood flow and use
important as it may impact on the choice of treat- motion as a contrast mechanism to visualize the
ment (anti-VEGF, photodynamic therapy, or location of the moving cells [57]. These meth-
both) [9, 23]. ods detect differences in amplitude, intensity or
a b c
Fig. 7.8 Multimodal imaging of PCV lesion. (a) CFP BVN. (d) The SD-OCT scan shows round structures
shows typical orange round lesions (yellow arrow) with adhering to the back surface of the elevated RPE, with a
hard exudates in the macula. (b) FA only shows occult hypo-reflective lumen and hyperreflective borders, corre-
CNV pattern and it is with ICGA (c) that the clusters of sponding to the polyps, along with the double-layer sign
polyps are identified (yellow arrow), surrounded by a corresponding to the BVN located between the RPE and
hypofluorescent halo and located in the fringes of the the Bruch’s membrane
phase variance in a series of sequential B-scans the normally avascular outer retina or under the
taken at the same location, with high-frequency RPE, with several reports published detailing the
and dense volumetric scanning, from which OCT-A characteristics of type 1 and 2 neovascu-
proprietary algorithms such as the split-spec- lar membranes [60]. Appropriate segmentation of
trum amplitude-decorrelation angiography the layer containing the CNV is necessary to
(SSADA), the optical microangiography good visualization, with some platforms allowing
(OMAG), and the OCT angiography ratio anal- for customized layer segmentation.
ysis (OCTARA), are used to reconstruct the Neovascular membranes in AMD may present
blood flow network in the acquired volume. In as well-circumscribed high-flow networks with
general, OCT-A systems can be divided into larger feeder vessel trunks, fine arborizing vessels
SD-OCT or SS-OCT-based systems. Some plat- in the periphery and anastomotic loops or arcades
forms already give automated, objective, quan- at vessel termini, producing patterns described as
titative data of flow (angio-analytics), but these “lacy wheel,” “sea-fan,” or “medusa- head”
features are not yet fully developed and must be (Fig. 7.9); or poorly circumscribed, irregular vas-
validated prior to widespread use in clinics and cular networks, said as “filamentous,” “tangled,”
clinical trials [58, 59]. Compared to conven- or “dead tree” CNVs (Fig. 7.10) [14]. Coscas
tional angiography the primary limitation is the et al. suggested different patterns and features for
absence of dynamic information such as filling differentiating active from inactive CNVs in OCT-
speed and leakage, however, OCT-A can show A: (1) shape, a well-defined (lacy-wheel or sea-
collections of fluid in structural analysis, sug- fan shaped) CNV lesion in contrast to one with
gesting that leakage is present. Motion and reti- long filamentous linear vessels; (2) branching,
nal vascular projections are additional numerous tiny capillaries, in contrast to rare large
limitations, but these tend to be corrected as mature vessels; (3) the presence of anastomoses
technology improves [20]. and loops; (4) morphology of the vessel termini,
In nAMD, CNVs can be visualized as a net- assessing the presence of a peripheral arcade in
work of abnormal, dilated, tortuous neovessels in contrast to a “dead tree” appearance; and (5) pres-
Fig. 7.9 OCT-A 3x3 mm en face and B-scans with super- double-layer sign, corresponding to the CNV and of sub-
imposed flow (in red) of two cases presenting with type 1 retinal fluid, a marker of lesion activity. En face images
neovascularization (a, b). In both cases the en face recon- show the CNV typical lacy pattern with peripheral arcades
struction shows more clearly the neovascular network in and loops, and surrounding hyposignal halo, correlating to
the choriocapillaris slab segmented bellow the RPE; the activity status
simultaneous B-scan allows for identification of the
Fig. 7.9 (continued)
a b
Fig. 7.10 “Dead tree” CNV pattern in OCT-A. (a) A “tangled,” or “dead tree” patterns are associated with
poorly circumscribed, irregular vascular network is inactive or chronic CNVs; (b) the CNV corresponds to a
observed after treatment with anti-VEGF. Larger mature shallow PED in SD-OCT (double-layer sign), and
vessels predominate, and no loops, peripheral arcade, or subretinal fluid is seen indicating lesion activity
perilesional hyposignal halo is seen. The “filamentous,”
ence of a perilesional hypointense halo, consid- are reported to vary from 6% to 27% [37, 62].
ered as regions of choriocapillaris alteration, The importance of the recognition of these
either corresponding to flow impairment, steal or lesions is the associated risk of exudation. De
localized atrophy [61]. However, despite the Oliveira Dias and colleagues reported an inci-
attempts in establishing criteria for activity status dence at 1-year of 21% in exudation when qCNV
in OCT-A, these are not validated, and the clini- was present at baseline compared to 4% when it
cian must complement information with struc- was not, which translated to superior a risk of
tural OCT showing indirect signs of leakage, such exudation of 15.2 times. Although no treatment
as subretinal or intraretinal fluid. for qCNVs is recommended at the present time,
OCT-A also improves the detection of CNV in closer monitoring is warranted in order to detect
nAMD before it develops exudation, this is, qui- exudation as early as possible and start appropri-
escent CNV (qCNV) (Fig. 7.11). Rates of qCNV ate treatment [62, 63].
a b
Fig. 7.11 Two exemplifying cases with quiescent type 1 preserved retinal architecture. En face OCT-A clearly
CNV detected in OCT-A (a, b). Structural OCT scans shows the CNV presence in the corresponding B-scan
show RPE irregularity and elevation (a) and a small PED areas, with typical lacy pattern in the choriocapillaris slab
(b), without associated intraretinal or subretinal fluid, and
OCT-A combined with structural OCT thus and less active phenotype over time, with larger
emerges as a noninvasive tool with high potential vessel diameters, less network branching, and
for both the diagnosis and follow-up of the capillary rarefaction [62].
nAMD patient, and in the evaluation of the Altogether, the advantages of OCT-A in
response to treatment over time. Treatment-naive nAMD could lead to decreasing necessity of con-
CNV shows numerous fine, arborizing vessels, ventional invasive angiography in the clinical set-
but during treatment with anti-VEGF there is a ting. Several studies compared the sensitivity and
regression of the most fine vessels, while central, specificity of OCT-A alone or in combination
perhaps more mature trunks containing pericyte with structural OCT to the gold standard FA
protected endothelial cells, seem to prevail [60, (and/or ICGA) [67]. These reports vary between
64]. The area of CNV can also be measured by 50% and 87% in sensitivity and 91% to 98% in
OCT-A, and its change with treatment is reported specificity for nAMD diagnosis with OCT-
(Fig. 7.12) [37, 65, 66]. Other parameters, such A. Lower sensitivities for CNV are usually asso-
as fractal dimension analysis, are being used to ciated with the presence of hemorrhage and large
measure the branching pattern complexity of the PEDs [62]. Perrott-Reynolds and colleagues con-
CNV. Studies reported on reduced fractal dimen- ducted a review on the diagnostic accuracy of
sion after treatment, suggesting less complexity OCT-A for nAMD compared to FA/ICGA, and
due to the attenuation and pruning of small-cali- found only moderate consistency of sensitivity
ber vessels [60]. OCT-A-based studies further and specificity among the different studies ana-
revealed that CNV re-proliferation with the lyzed. The majority had small sample sizes and
reopening of neovessels is observed from week 2 none reported confidence intervals, limiting their
to week 4 post-treatment [64]. The revascularized results [68]. Compared to ICGA, measurements
CNV progressively changes to a more mature of CNV area in nAMD were significantly smaller
a b c
Fig. 7.12 CNV change with anti-VEGF therapy in OCT- more peripheral capillaries; (c) after the initial response
A. (a) CNV lesion area is marked in yellow in en face there is repermeabilization of the peripheral capillaries
image, B-scan shows type 1 CNV with subretinal fluid; with visualization of loops, peripheral arcades, and sur-
(b) after treatment the CNV area decreases, with persis- rounding hyposignal halo, indicating that CNV activity is
tence of the central trunk and disappearance of flow in the likely increasing
in OCT-A in both type 1 and type 2 membranes, the deep capillary plexus dragging toward the
despite the good correlation between measure- outer retinal layers can be observed. In some
ments [14, 69]. cases, a clew-like flow signal in the choriocapil-
Because of this inconsistent data, and despite laris is also seen underneath the tuft-shaped
the clear utility of OCT-A in nAMD diagnosis lesion and corresponds to proliferation in the
and follow-up, FA and ICGA are still considered sub-RPE space. Careful assessment combined
the gold standard for a comprehensive evaluation with structural OCT with superimposed flow sig-
of lesion location and morphology, and OCT-A nal is needed, as the RAP connections may be
stands for now as an auxiliary imaging method. missed if only the en face image is examined [14,
Others propose, however, a new algorithm in that 71, 72]. OCT-A was shown to detect growth and
OCT and OCT-A are to be performed first and if morphologic changes in the incipient RAP lesion
the diagnosis is certain one should proceed to before the exudation develops and also enables
treatment, however, if the result is negative or the distinction from avascular hyperreflective
there is doubt, dye-based examinations should foci from pigment migration seen in structural
always be performed [70]. OCT [73].
OCT-A is also useful in documenting the
response of type 3 lesions to treatment, as studies
7.5.1 AP and PCV Phenotypes
R showed that the microvascular structure of RAP
in OCT-A is highly responsive to anti-VEGF therapy, with a
clear decrease in size [74].
OCT-A features of RAP lesions are now well At present, the role of OCT-A in PCV is less
documented. RAP lesions present in the earliest clear compared to other types of CNV. The BVN
stages as small intraretinal neovascular com- is visualized in OCT-A as any type 1 CNV, but
plexes seen as tufts-shaped lesions in the outer polyps are not very well detected, at least using
retinal layers, descending from the deep capillary only automated segmentation. The detection
plexus to the RPE (Fig. 7.13). Feeder vessels in rates of polyps using OCT-A are in fact much
a b c
d e
Fig. 7.13 RAP lesion in OCT-A. (a, b) A high-flow tuft- posed blood flow (in red) where the flow is seen deepen-
shaped abnormal neovascular network is seen mainly in ing and merging with the PED, with focal disruption of
deep retinal plexus and outer retina slabs; (c, d) there is a the RPE layer; (e) correspondent FA showing a hot spot
connection of the neovascular process to the subretinal from RAP lesion
space; this is best appreciated in the B-scan with superim-
lower compared with ICGA, which is the gold is very well seen between the RPE and the
standard in PCV diagnosis. The hypothesis is Bruch’s membrane and in patterns similar to
that turbulent or very slow flow inside the polyp other type 1 CNVs, including seafan, medusa,
is below the threshold of motion detection by the and tangle [76, 77].
device [14, 75]. One group using a novel OCT-A algorithm
Analyzing the B-scans obtained with OCT- termed variable interscan time analysis (VISTA),
A, the blood flow in polyps was shown to be that allows to visualize variations in relative
just below the top of the PED, or focally located blood flow, in conjunction with a prototype high-
in the upper part of the peak caused by the polyp speed swept-source OCT, showed heterogeneous
[76]. Because of this, it may be necessary to blood flow speeds within the polyps. Some had
analyze the automated outer retina slab or to faster blood flow in the periphery and slower
change manually the segmentation to above the blood flow in the center, perhaps from turbulence
choriocapillaris slab (where the BVN is seen), within. The BVNs showed relatively faster flow
to detect the flow in the polyps in en face in the larger trunk vessels and slower flow speeds
images. in the smaller vessels [78, 79].
According to level of segmentation and flow When there is a severe distortion of the retinal
characteristics within the polyps, they are seen architecture from large serosanguineous PEDs,
as a hyposignal round structure, a hypersignal visualization of the BVN and polyps may not be
ring of flow devoid of flow inside, or they can possible or is very poor. This is another limitation
present in hypersignal patterns such as nodular, of OCT-A in PCV where this type of complica-
dot, or cluster (Figs. 7.14 and 7.15). The BVN tion is common. Therefore, OCT-A is currently
a b
Fig. 7.14 Macular PCV lesion in OCT-A (a) and ICGA superimposed flow (a, below, in the left) one can appreci-
(b). OCT-A en face image in the left (a) shows the BVN ate the presence of flow (in red) in the top of the peaked
with more detail compared to ICGA (b), however, the pol- RPE elevations corresponding to polyps in ICGA (b). En
yps are devoid of flow in the choriocapillaris slab where face structural OCT (a, above, in the right) shows very
the BVN is best seen, and appear as round hyposignal nicely the BVN and polyps of the PCV lesion in close cor-
structures (yellow arrows); in the B-scan image with respondence with the PCV lesion observed in ICGA (b)
a b
Fig. 7.15 PCV lesion in OCT-A (a), ICGA (b), and (yellow arrows). (c) SD-OCT shows the peaked RPE ele-
SD-OCT (c). In this case, the polyp component is seen in vation (orange arrow) typical for the underlying presence
both ICGA and OCT-A as a round hypersignal structure of polyps and the double layer sign in both sides repre-
with flow, surrounded by a hyposignal halo (orange senting the BVN, along with subretinal fluid
arrow). The BVN is best appreciated in en face OCT-A
unable to replace ICGA in the assessment of 2. Clinical and imagiological characteristics of

PCV but can provide important supplementary nAMD are subject to constant study and
information [9, 77]. In the future, further development, as new technologies provide
improvements in OCT-A will allow this technol- new means to accurately diagnose and to eval-
ogy to play a greater role in PCV diagnosis and uate the response to treatment.
management. 3. Currently, there is a transition in the way
nAMD is classified. The classification in
occult/classic CNV patterns diagnosed with
7.6 Concluding Remarks the gold standard fluorescein angiography
changed to a multimodal-based classification
Neovascular AMD is an entity extensively char- mainly guided by SD-OCT and closer to the
acterized by the clinical and imagiological per- histology of different nAMD phenotypes.
spectives. First with dye-based examinations, 4. Neovascularization type 1 (bellow the RPE),
which for years served as the cornerstone for sys- type 2 (above the RPE), and type 3 (RAP
tematic clinical classification and treatment guid- lesion) classifications are now preferred. PCV
ance, and afterward with the historical is considered more as a variant of type 1 neo-
introduction of OCT. The latter, in constant vascularization in which polypoidal or aneu-
improvement, lead to refinement in CNV classifi- rysmal vascular dilations develop.
cation in the last decade. A conciliation was 5. Neovascular AMD phenotype distinction
finally achieved between the earliest histopatho- might not be as important as in the past in
logic studies conducted by Gass and colleagues treatment decision, but might be important for
and the new imaging findings at the microstruc- prognosis, possible involvement of fellow eye
tural level that OCT provided complemented and in considering adjuvant therapies such as
with FA and ICGA. PDT.
Today with the introduction of OCT-A we are 6. Multimodal-imaging with FA and SD-OCT
again on the verge of exciting changes in both (and ICGA when necessary) is the gold stan-
how nAMD is diagnosed and how patients will dard for nAMD diagnosis and follow-up.
be managed during treatment. New algorithms However, OCT-A is a breakthrough technol-
and ultra-high-speed OCT systems are being ogy recently introduced in clinical practice
developed and improved. Noninvasive, fast, and that might change this paradigm. Neovascular
reliable new technologies will not only allow to AMD diagnosis and decision to treat might
detect CNVs’ flow and vessel density, correlating soon change to a purely non-invasive imaging
them to microstructural analysis at the histologi- approach.
cal level, but they will enable flow velocity mea-
surements, with distinction of regions of high and Acknowledgments We are grateful to Ana Rita Santos
for her help in assembling the figures.
low flow that were not seen before. Progression
to quantitative absolute measurements of flow in
the future might be of great value, especially in
analyzing for CNV activity and response to References
treatment.
1. Tomany SC, Wang JJ, Van Leeuwen R, et al. Risk fac-
Multimodal imaging is the future in nAMD tors for incident age-related macular degeneration:
assessment, and the future is promising. pooled findings from 3 continents. Ophthalmology.
2004;111(7):1280–7.
2. Shao J, Choudhary MM, Schachat AP. Neovascular
Key Learning Points
age-related macular degeneration. Dev Ophthalmol.
1. Neovascular AMD must be promptly recog- 2016;55:125–36.
nized, as treatments must be implemented 3. Wong WL, Su X, Li X, et al. Global prevalence of
early in the natural history of the disease to age-related macular degeneration and disease bur-
den projection for 2020 and 2040: a systematic
prevent severe visual loss.
review and meta-analysis. Lancet Glob Health. branes in age-related macular degeneration. Retina
2014;2(2):e106–16. (Philadelphia, Pa). 2019;39(2):281–7.
4. Buch H, Nielsen NV, Vinding T, Jensen GB, Prause 19. Giani A, Luiselli C, Esmaili DD, et al. Spectral-
JU, La cour M. 14-year incidence, progression, domain optical coherence tomography as an indi-
and visual morbidity of age-related maculopathy: cator of fluorescein angiography leakage from
the Copenhagen City eye study. Ophthalmology. choroidal neovascularization. Invest Ophthalmol Vis
2005;112(5):787–98. Sci. 2011;52(8):5579–86.
5. Ferris FL, Fine SL, Hyman L. Age-related macu- 20. Holz FG, Sadda SR, Staurenghi G, et al. Imaging
lar degeneration and blindness due to neovascu- protocols in clinical studies in advanced age-
lar maculopathy. Arch Ophthalmol. 1984;102(11): related macular degeneration: recommendations
1640–2. from classification of atrophy consensus meetings.
6. Bressler NM, Bressler SB. Neovascular (exuda- Ophthalmology. 2017;124(4):464–78.
tive or “wet”) age-related macular degeneration. 21. Ditthmar F, Holz FG. Fluorescein angiography in oph-
6th ed. Retina Fifth Edition. Elsevier Inc; 2012. thalmology. Heidelberg: Springer; 2007. p. 56–88.
p. 1345–1372. https://www.springer.com/gp/book/9783540783596.
7. Yannuzzi LA, Freund KB, Takahashi BS. Review of 22. Subfoveal neovascular lesions in age-related macular
retinal angiomatous proliferation or type 3 neovascu- degeneration. Guidelines for evaluation and treat-
larization. Retina. 2008;28(3):375–84. ment in the macular photocoagulation study. Macular
8. Tsai ASH, Cheung N, Gan ATL, et al. Retinal photocoagulation study group. Arch Ophthalmol.
angiomatous proliferation. Surv Ophthalmol. 1991;109(9):1242–57.
2017;62(4):462–92. 23. Freund KB, Zweifel SA, Engelbert M. Do we need
9. Cheung CMG, Lai TYY, Ruamviboonsuk P, et al. a new classification for choroidal neovasculariza-
Polypoidal Choroidal vasculopathy: definition, patho- tion in age-related macular degeneration? Retina
genesis, diagnosis, and management. Ophthalmology. (Philadelphia, Pa). 2010;30(9):1333–49.
2018;125(5):708–24. 24. Yannuzzi LA, Slakter JS, Sorenson JA, Guyer
10. Gass JD. Biomicroscopic and histopathologic con- DR, Orlock DA. Digital Indocyanine Green
siderations regarding the feasibility of surgical exci- Videoangiography and Choroidal neovascularization.
sion of subfoveal neovascular membranes. Am J Retina. 1992;12:191–223.
Ophthalmol. 1994;118(3):285–98. 25. Regillo CD, Benson WE, Maguire JI, Annesley
11. Grossniklaus HE, Gass JD. Clinicopathologic corre- WH Jr. Indocyanine Green angiography and occult
lations of surgically excised type 1 and type 2 sub- Choroidal neovascularization. Ophthalmology.
macular choroidal neovascular membranes. Am J 1994;101(2):280–8.
Ophthalmol. 1998;126(1):59–69. 26. Gabriele ML, Wollstein G, Ishikawa H, et al. Optical
12. Hartnett ME, Weiter JJ, Staurenghi G, Elsner coherence tomography: history, current status,
AE. Deep retinal vascular anomalous complexes and laboratory work. Invest Ophthalmol Vis Sci.
in advanced age-related macular degeneration. 2011;52(5):2425–36.
Ophthalmology. 1996;103(12):2042–53. 27. Potsaid B, Baumann B, Huang D, et al. Ultrahigh
13. Yannuzzi LA, Negrão S, Iida T, et al. Retinal speed 1050nm swept source/Fourier domain OCT
Angiomatous proliferation in age-related macular retinal and anterior segment imaging at 100,000
degeneration. Retina. 2001;21:416–34. to 400,000 axial scans per second. Opt Express.
14. Talks SJ, Aftab AM, Ashfaq I, Soomro T. The role 2010;18(19):20029–48.
of new imaging methods in managing age-related 28. Jung JJ, Chen CY, Mrejen S, et al. The incidence of
macular degeneration. Asia Pac J Ophthalmol (Phila). neovascular subtypes in newly diagnosed neovascular
2017;6(6):498–507. age-related macular degeneration. Am J Ophthalmol.
15. Mrejen S, Sarraf D, Mukkamala SK, Freund 2014;158(4):769–79. e2
KB. Multimodal imaging of pigment epithelial detach- 29. Sato T, Kishi S, Watanabe G, Matsumoto H, Mukai
ment: a guide to evaluation. Retina (Philadelphia, Pa). R. Tomographic features of branching vascular net-
2013;33(9):1735–62. works in polypoidal choroidal vasculopathy. Retina
16. Khanani AM, Eichenbaum D, Schlottmann PG, (Philadelphia, Pa). 2007;27(5):589–94.
Tuomi L, Sarraf D. Optimal Management of Pigment 30. Spaide RF. Enhanced depth imaging optical coher-
Epithelial Detachments in eyes with Neovascular age- ence tomography of retinal pigment epithelial detach-
related macular degeneration. Retina (Philadelphia, ment in age-related macular degeneration. Am J
Pa). 2018;38(11):2103–17. Ophthalmol. 2009;147(4):644–52.
17. Ersoz MG, Karacorlu M, Arf S. Sayman muslubas I, 31. Khan S, Engelbert M, Imamura Y, Freund
Hocaoglu M. retinal pigment epithelium tears: clas- KB. Polypoidal choroidal vasculopathy: simultaneous
sification, pathogenesis, predictors, and management. indocyanine green angiography and eye-tracked spec-
Surv Ophthalmol. 2017;62(4):493–505. tral domain optical coherence tomography findings.
18. Ravera V, Giani A, Pellegrini M, et al. Comparison Retina (Philadelphia, Pa). 2012;32(6):1057–68.
among different diagnostic methods in the study of 32. Mukkamala SK, Costa RA, Fung A, Sarraf D,
type and activity of Choroidal Neovascular mem- Gallego-pinazo R, Freund KB. Optical coherence
tomographic imaging of sub-retinal pigment epithe- 47. Nagiel A, Sarraf D, Sadda SR, et al. Type 3 neovascu-
lium lipid. Arch Ophthalmol. 2012;130(12):1547–53. larization: evolution, association with pigment epithe-
33. Grossniklaus HE, Green WR. Choroidal neovascular- lial detachment, and treatment response as revealed
ization. Am J Ophthalmol. 2004;137(3):496–503. by spectral domain optical coherence tomography.
34. Naysan J, Jung JJ, Dansingani KK, Balaratnasingam Retina (Philadelphia, Pa). 2015;35(4):638–47.
C, Freund KB. Type 2 (subretinal) neovasculariza- 48. Yannuzzi LA, Sorenson J, Spaide RF, Lipson
tion in age-related macular degeneration associated B. Idiopathic polypoidal choroidal vasculopathy
with pure reticular pseudodrusen phenotype. Retina (IPCV). Retina. 1990;10:1–8.
(Philadelphia, Pa). 2016;36(3):449–57. 49. Yannuzzi LA, Wong DW, Sforzolini BS, et al.
35. Roquet W, Roudot-Thoraval F, Coscas G, Soubrane Polypoidal choroidal vasculopathy and neovas-
G. Clinical features of drusenoid pigment epithelial cularized age-related macular degeneration. Arch
detachment in age related macular degeneration. Br J Ophthalmol. 1999;117(11):1503–10.
Ophthalmol. 2004;88(5):638–42. 50. Dansingani KK, Balaratnasingam C, Naysan J, Freund
36. Balaratnasingam C, Messinger JD, Sloan KR, KB. En face imaging of Pachychoroid Spectrum dis-
Yannuzzi LA, Freund KB, Curcio CA. Histologic orders with swept-source optical coherence tomogra-
and optical coherence tomographic correlates in phy. Retina (Philadelphia, Pa). 2016;36(3):499–516.
Drusenoid pigment epithelium detachment in age- 51. Dansingani KK, Gal-or O, Sadda SR, Yannuzzi LA,
related macular degeneration. Ophthalmology. Freund KB. Understanding aneurysmal type 1 neo-
2017;124(5):644–56. vascularization (polypoidal choroidal vasculopathy):
37. Waheed NK, Kashani AH, Alexandre C, Garcia DA, a lesson in the taxonomy of ‘expanded spectra’–a
Duker JS, Rosenfeld PJ. Optical coherence tomog- review. Clin Exp Ophthalmol. 2018;46(2):189–200.
raphy. 6th ed. Ryan’s Retina. Elsevier Inc; 2019. 52. Silva R. Neovascular phenotypes: Polypoidal
p. 77–119. Choroidal vasculopathy. In: Age-related macular
38. Sarraf D, Reddy S, Chiang A, Yu F, Jain A. A new degeneration. 1st ed; 2010. p. 129–39.
grading system for retinal pigment epithelial tears. 53. Li X. 74–Polypoidal Choroidal vasculopathy. 6th ed.
Retina (Philadelphia, Pa). 2010;30(7):1039–45. Ryan’s Retina. Elsevier Inc; 2019. p. 1458–1464.
39. Schaal KB, Freund KB, Litts KM, Zhang Y, Messinger 54. Tan CS, Ngo WK, Chen JP, Tan NW, Lim
JD, Curcio CA. Outer retinal Tubulation in advanced TH. EVEREST study report 2: imaging and grading
age-related macular degeneration: optical coherence protocol, and baseline characteristics of a randomised
tomographic findings correspond to histology. Retina controlled trial of polypoidal choroidal vasculopathy.
(Philadelphia, Pa). 2015;35(7):1339–50. Br J Ophthalmol. 2015;99(5):624–8.
40. Preti RC, Govetto A, Filho RGA, et al. Optical coher- 55. Byeon SH, Lew YJ, Lee SC, Kwon OW. Clinical fea-
ence tomography analysis of outer retinal Tubulations, tures and follow-up results of pulsating polypoidal
sequential evolution and pathophysiological insights. choroidal vasculopathy treated with photodynamic
Retina (Philadelphia, Pa). 2018;38(8):1518–25. therapy. Acta Ophthalmol. 2010;88(6):660–8.
41. Cohen SY, Dubois L, Nghiem-buffet S, et al. Retinal 56. De Salvo G, Vaz-Pereira S, Keane PA, Tufail A, Liew
pseudocysts in age-related geographic atrophy. Am J G. Sensitivity and specificity of spectral-domain opti-
Ophthalmol. 2010;150(2):211–7. e1 cal coherence tomography in detecting idiopathic pol-
42. Gass JD, Agarwal A, Lavina AM, Tawansy KA. Focal ypoidal choroidal vasculopathy. Am J Ophthalmol.
inner retinal hemorrhages in patients with drusen: an 2014;158(6):1228–38.
early sign of occult choroidal neovascularization and 57. Gorczynska I, Migacz JV, Zawadzki RJ, Capps
chorioretinal anastomosis. Retina (Philadelphia, Pa). AG, Werner JS. Comparison of amplitude-decorre-
2003;23(6):741–51. lation, speckle-variance and phase-variance OCT
43. Freund KB, Ho IV, Barbazetto IA, et al. Type 3 neo- angiography methods for imaging the human ret-
vascularization: the expanded spectrum of retinal ina and choroid. Biomed Opt Express. 2016;7(3):
angiomatous proliferation. Retina (Philadelphia, Pa). 911–42.
2008;28(2):201–11. 58. Tan ACS, Tan GS, Denniston AK, et al. An overview of
44. Monson DM, Smith JR, Klein ML, Wilson the clinical applications of optical coherence tomog-
DJ. Clinicopathologic correlation of retinal raphy angiography. Eye (Lond). 2018;32(2):262–86.
Angiomatous proliferation. Arch Ophthalmol. 59. De carlo TE, Romano A, Waheed NK, Duker JS. A
2008;37(8):1664–8. review of optical coherence tomography angiography
45. Gross NE, Aizman A, Brucker A, Klancnik JM, (OCTA). Int J Retina Vitreous. 2015;1:5.
Yannuzzi LA. Nature and risk of neovascularization 60. Al-Sheikh M, Iafe NA, Phasukkijwatana N, Sadda
in the fellow eye of patients with unilateral retinal SR, Sarraf D. Biomarkers of Neovascular activity in
angiomatous proliferation. Retina (Philadelphia, Pa). age-related macular degeneration using optical coher-
2005;25(6):713–8. ence tomography angiography. Retina (Philadelphia,
46. Silva R. Neovascular phenotypes: retinal Angiomatous Pa). 2018;38(2):220–30.
proliferation (RAP) or type 3 neovascularization. 61. Coscas GJ, Lupidi M, Coscas F, Cagini C, Souied
In: Age-related macular degeneration. 1st ed; 2010. EH. Optical coherence tomography angiography ver-
p. 117–27. sus traditional multimodal imaging in assessing the
activity of exudative age-related macular degenera- 71. Miere A, Querques G, Semoun O, et al. Optical coher-
tion: a new diagnostic challenge. Retina (Philadelphia, ence tomography angiography changes in early type 3
Pa). 2015;35(11):2219–28. neovascularization after anti-VEGF treatment. Retina
62. Schneider EW, Fowler SC. Optical coherence tomog- (Philadelphia, Pa). 2017;37(10):1873–9.
raphy angiography in the management of age-related 72. Querques G, Miere A, Souied EH. Optical coherence
macular degeneration. Curr Opin Ophthalmol. tomography angiography features of type 3 neovas-
2018;29(3):217–25. cularization in age-related macular degeneration. Dev
63. De Oliveira dias JR, Zhang Q, JMB G, et al. Natural Ophthalmol. 2016;56:57–61.
history of subclinical neovascularization in non- 73. Tan ACS, Dansingani KK, Yannuzzi LA, Sarraf
exudative age-related macular degeneration using D, Freund KB. Type 3 neovascularization imaged
swept-source OCT angiography. Ophthalmology. with cross-sectional and en face optical coherence
2018;125(2):255–66. tomography angiography. Retina (Philadelphia, Pa).
64. Mar ques JP, Costa JF, Marques M, Cachulo ML, 2017;37(2):234–46.
Figueira J, Silva R. Sequential morphological 74. Phasukkijwatana N, Tan ACS, Chen X, Freund KB,
changes in the CNV net after Intravitreal anti-VEGF Sarraf D. Optical coherence tomography angiography
evaluated with OCT angiography. Ophthalmic Res. of type 3 neovascularisation in age-related macu-
2016;55(3):145–51. lar degeneration after antiangiogenic therapy. Br J
65. Coscas G, Lupidi M, Coscas F, Français C, Cagini C, Ophthalmol. 2017;101(5):597–602.
Souied EH. Optical coherence tomography angiog- 75. Tomiyasu T, Nozaki M, Yoshida M, Ogura
raphy during follow-up: qualitative and quantitative Y. Characteristics of Polypoidal Choroidal vas-
analysis of mixed type I and II choroidal neovascu- culopathy evaluated by optical coherence tomog-
larization after vascular endothelial growth factor trap raphy angiography. Invest Ophthalmol Vis Sci.
therapy. Ophthalmic Res. 2015;54(2):57–63. 2016;57(9):OCT324–30.
66. Pilotto E, Frizziero L, Daniele AR, et al. Early OCT 76. Wang M, Zhou Y, Gao SS, et al. Evaluating Polypoidal
angiography changes of type 1 CNV in exudative Choroidal vasculopathy with optical coherence
AMD treated with anti-VEGF. Br J Ophthalmol. tomography angiography. Invest Ophthalmol Vis Sci.
2019;103(1):67–71. 2016;57(9):OCT526–32.
67. De Carlo TE, Bonini filho MA, Chin AT, et al. Spectral- 77. Inoue M, Balaratnasingam C, Freund KB. Optical
domain optical coherence tomography angiography coherence tomography angiography of polypoi-
of choroidal neovascularization. Ophthalmology. dal choroidal vasculopathy and polypoidal choroi-
2015;122(6):1228–38. dal neovascularization. Retina (Philadelphia, Pa).
68. Perrott-reynolds R, Cann R, Cronbach N, et al. The 2015;35(11):2265–74.
diagnostic accuracy of OCT angiography in naive and 78. Arya M, Rashad R, Sorour O, Moult EM, Fujimoto
treated neovascular age-related macular degeneration: JG, Waheed NK. Optical coherence tomography
a review. Eye (Lond). 2019;33(2):274–82. angiography (OCTA) flow speed mapping technol-
69. Told R, Sacu S, Hecht A, Baratsits M, Eibenberger ogy for retinal diseases. Expert Rev Med Devices.
K, et al. Comparison of SD-optical coherence tomog- 2018;21:1–8.
raphy angiography and Indocyanine Green angi- 79. Rebhun CB, Moult EM, Novais EA, et al. Polypoidal
ography in type. Investig Opthalmology Vis Sci. Choroidal vasculopathy on swept-source optical
2018;59:2393–400. coherence tomography angiography with variable
70. Imaging of exudative age-related macular degenera- Interscan time analysis. Transl Vis Sci Technol.
tion: toward a shift in the diagnostic paradigm? Retina 2017;6(6):4.
(Philadelphia, Pa). 2017;38(Suppl 1):1–12.
Medical Retina Unit at the Ophthalmology Department,

Centro Hospitalar e Universitário de Coimbra, Portugal
(CHUC) and an investigator at the Association for
Innovation and Biomedical Research on Light and Image
(AIBILI), Coimbra, Portugal.
Prof. Rufino Silva is an ophthalmologist with more
than 25 years of experience, primarily committed to the
study and treatment of retinal diseases, mainly focused on
AMD, diabetic retinopathy, retinal vein occlusion (RVO),
and pathologic myopia. His research is mainly focused on
AMD: epidemiology, diet and lifestyle, imaging biomark-
Rufino Silva MD, PhD, FEBO is an Assistant Professor ers, metabolomics, and genetics. He is author and co-
of Ophthalmology at the Faculty of Medicine, University author of more than 140 articles, chapters, and books.
of Coimbra, Portugal (FMUC). He is also the Head of Orcid: https://orcid.org/0000-0001-8676-0833
Management Strategies
for Neovascular AMD 8
Irmela Mantel
8.1 he Story of Anti-VEGF

T pegaptanib (brand name Macugen), licensed in
Treatment: Molecules 2000. It showed in a phase 3 clinical study an
and Regimens efficacy for nAMD which was yet limited to a
reduction in visual loss compared to sham [6],
The current treatment of neovascular age-related similar to the efficacy of PDT with verteporfin
macular degeneration (nAMD) is mainly based on [7]. Pegaptanib is a molecule that competitively
the inhibition of the vascular endothelial growth binds to the VEGF isoform 165, which was con-
factor (VEGF) in the retinal tissue by intravitreal sidered the main pathogenic isoform. However,
injection of anti-VEGF agents. This approach later studies showed that this approach neglected
allowed for the first time in history, a mean visual other VEGF isoforms with significant pathogenic
improvement from treatment start (baseline) [1– potential that need to be targeted as well.
3], whereas preceding treatment options such as Ranibizumab (brand name Lucentis), a specific
laser photocoagulation or photodynamic therapy affinity-mature fragment of a recombinant
(PDT) with verteporfin were only able to reduce humanized IgG1 monoclonal antibody that neu-
the degree of visual loss. Although both laser and tralizes all active VEGF-A isoforms of the human
PDT are still occasionally used, these treatment VEGF protein, was licensed in 2006. With its
options are reserved for exceptional cases. arrival, a true efficacy revolution was started:
VEGF was identified in the 1990s as a key fac- Using monthly intravitreal injections, a mean
tor in the development of neovascular membranes visual acuity improvement was obtained after 1
[4, 5]. VEGF was not only sufficient to promote a and after 2 years of treatment, ranging between a
neovascular response [5] but also required; when mean gain of 5.4 ETDRS letters and 11.3 letters
VEGF was blocked in a nonhuman primate, no according to the study [1, 2, 8].
vasoproliferation was detected [4]. Subsequent These pivotal trials did set the reference for
studies have well established the central role of the best visual outcomes under currently avail-
VEGF in neovascular disorders. The first com- able treatment options. This astounding improve-
mercially available anti-VEGF molecule was ment in the prognosis of nAMD was achieved on
the basis of fixed monthly injections. The high
treatment frequency placed a heavy burden on
I. Mantel (*) the management of patients with chronic nAMD,
Department of Ophthalmology, University thereby requiring many clinical and therapeutic
of Lausanne, Jules Gonin Eye Hospital, Fondation interventions over the course of a patient’s lifes-
Asile des Aveugles, CP, Lausanne, Switzerland pan due to the repetitive treatment scheme.
e-mail: irmela.mantel@fa2.ch

100 I. Mantel
Therefore, many attempts were made to reduce 2 months for aflibercept (as compared to monthly
the burden, both on the level of the total number injections with ranibizumab) and a PRN regimen
of injections (cost factor) as well as on the level in the second year. In both study periods, the effi-
of the number of monitoring visits (human and cacy of aflibercept was equivalent to that of
technical resource factor). Quarterly fixed ranibi- ranibizumab. Thus, aflibercept was labeled for
zumab injections resulted in a loss of the initial intravitreal injection in nAMD, and the recom-
VA improvement and were shown to be signifi- mended regimen was a fixed injection every
cantly inferior to monthly injections, although a 2 months. Although this is a valid approach, con-
subset of patients did well on this regimen [9– cerns were raised due to a subset of patients with
11]. A small study with the first pro re nata (PRN) significant monthly recurrences, leading to a zig-
dosing showed the feasibility and efficiency of a zag curve of structural OCT outcomes when
re-treatment strategy based on monthly assess- treating every other month [15]. It is unclear
ments of visual acuity, fundus examination, and whether these patients would have had a better
most importantly optical coherence tomography functional outcome with a monthly instead of the
(OCT), allowing for a reduction in the mean bimonthly aflibercept treatment.
number of injections [12, 13]. However, the inter- Very recently, the new anti-VEGF agent bro-
individual range of treatment needs varied lucizumab, a very small single-chain antibody
widely, between monthly injections and no re- fragment allowing a high drug concentration, has
treatment after the loading dose. A later study shown its efficacy in a phase 3 study. The non-
showed that the mean treatment interval in a PRN inferiority in comparison to aflibercept was
regimen is approximately 2 months for ranibi- shown for visual acuity and structural results,
zumab [14]. The PRN regimen was the first although the regimen was based on fixed injec-
approach to be widely adopted in clinical prac- tions every 3 months. Apparently, this new mol-
tice. It was validated as a valuable treatment ecule might have a long-lasting activity, possibly
option with a non-inferior outcome as compared related to the higher molecular dose. However,
to a fixed monthly re-treatment [15]. Both ranibi- roughly half of the patients showed some disease
zumab and the entire antibody bevacizumab were activity when treated at 3-month intervals, and
shown to be adequate treatment options [15]. their regimen was changed to an injection every
However, some differences were found favoring 2 months. Brolucizumab has recently become
ranibizumab: Pathological fluid was less fre- available on the market.
quently present when using ranibizumab, and the While fixed regimens have the major advan-
functional outcome of ranibizumab in the fixed tage of the ability to plan ahead with little neces-
monthly treatment was significantly better than sity for monitoring visits, they result inevitably in
bevacizumab in a PRN regimen [15]. In addition, overtreatment of some patients and/or undertreat-
bevacizumab is not licensed for intravitreal use ment of others due to the large variability in treat-
but for intravenous use in oncology. In conclu- ment needs. In addition, economic concerns,
sion, bevacizumab is a less expensive second- limited resources, and considerations regarding
choice off-label anti-VEGF drug, which has possible side effects of the intravitreal injection
shown close to optimal efficacy for intravitreal procedure (endophthalmitis) lead to an inclina-
use in nAMD. tion to reduce the number of injections.
Aflibercept is a recombinant fusion protein Individually adjusted variable dosing regimens
consisting of VEGF-binding portions from the have the potential to respond adequately to these
extracellular domains of the human VEGF recep- requirements. Variable dosing regimens aiming
tors 1 and 2 that are fused to the Fc portion of the for an individualized treatment, use the minimum
human IgG1 immunoglobulin. Aflibercept was number of injections needed for the best possible
first labeled in 2011 for an intravitreal use in outcome. Independent of the choice of the anti-
nAMD. The pivotal phase 3 clinical trial used in VEGF agent, variable dosing regimens rely on
the first year a fixed injection regimen every best practice in terms of visit intervals and
8 Management Strategies for Neovascular AMD 101
re-treatment criteria. Major efficacy losses arise strategy are comparable to those with a PRN reg-
if the follow-up is neglected or unsuitable re- imen and statistically non-inferior to monthly re-
treatment criteria are used. Real-life results of treatment [16]. The treat-and-extend regimen
PRN regimens have shown that visit interval lon- needs a significantly lower number of monitoring
ger than a month are unsuitable to achieve the visits (the mean value is approximately eight per
best outcome. This is not surprising, as not only year), and planning is clearly facilitated as nearly
the treatment needs are extremely variable among each monitoring visit goes along with an injec-
patients, but longer monitoring intervals in a tion. Thus, this strategy has been widely adopted
PRN regimen lead in many patients also to pro- as the first-choice regimen for anti-VEGF treat-
tracted recurrence periods. Thus, re-treatment ment in nAMD.
may come too late in order to prevent progressive The observe-and-plan regimen also applies the
retinal damage due to recurrence. In other words, idea to schedule treatment intervals, however, to a
PRN with anti-VEGF antibodies is only a valu- series of injections instead of single injection inter-
able strategy if monthly monitoring visits and vals. Based on the regularity of individual treatment
prompt re-treatment in case of a recurrence can needs [17], the observe-and-plan regimen evaluates
be guaranteed. The re-treatment criteria are dis- after the loading doses the time to the first recur-
cussed in a separate paragraph below. rence signs using monthly monitoring visits
The need for strict monthly monitoring visits (“observe”), thereafter applying this interval slightly
in a PRN regimen with anti-VEGF drugs has rap- shortened by 2 weeks in a series of planned injec-
idly revealed the capacity limits of healthcare tions with a fixed interval (up to three injections,
providers. The high incidence of new cases and interval between 1 and 3 months, treatment plan up
the required treatment chronicity have led to an to 6 months) [18, 19]. The monitoring visits after a
overwhelming number of patients to be cared for. series of fixed injections are therefore less frequent
Moreover, PRN has the major disadvantage that (the mean being approximately four per year) than
planning is impossible, which is a logistic chal- in the treat-and-extend regimen, and the ability to
lenge to the institution and a psychological bur- plan ahead is excellent. However, the longest treat-
den for the patient. Thus, alternative regimens ment interval without a monitoring visit should not
have been developed, based on the idea that the exceed 3 months. This regimen has the potential for
past experience with a given eye may allow antic- excellent patient care with limited ressources.
ipating the future need for injections.
The treat-and-extend regimen applies this
idea. After a loading dose of one to three monthly 8.2 Re-Treatment Criteria
injections, the treatment interval is progressively for Variable Dosing Regimen
extended by 2 weeks as long as no signs of activ-
ity (the criteria are discussed below) are detected. Sensitive re-treatment criteria are the cornerstone
Monitoring is performed immediately before an of any variable dosing regimen with anti-VEGF
injection, and its result determines the length of drugs. Most studies consider primarily OCT cri-
the next interval, without modifying the planned teria, combined with fundus examination and
imminent injection. As soon as a monitoring visit visual acuity loss.
reveals any signs of activity, the interval is short-
ened by 2 weeks. The interval between visits • Visual acuity: Visual loss relative to any pre-
combining monitoring with treatment should ceding monitoring visit (typically 5 ETDRS
usually not exceed approximately 3 months (12– letters or 1 line) is a clinically meaningful but
16 weeks). If the macula of a treated eye remains unreliable criterion for re-treatment with anti-
dry at this three-month interval, a choice between VEGF drugs. This re-treatment criterion was
the default treatment every 3 months or a change used due to its ease of examination. However,
to monitoring visits without injections every pathological changes are usually first detect-
2 months is suggested. The outcomes with this able by OCT and manifest as a decline in
102 I. Mantel
visual acuity later. In addition, vision loss may sive peaks or sustained Valsalva pressure might
be caused by many other conditions than exu- be the cause. Thus, hemorrhage might occur
dative recurrence, and they may be completely even under monthly anti-VEGF treatments and
unresponsive to anti-VEGF treatment, such as a complete VEGF suppression.
progressive macular atrophy or cataract.
Therefore, it cannot be a stand alone criterion. In summary, SD-OCT is the most important
In addition, retreatment should not wait for investigation for variable dosing regimens with
visual loss to happen, if other signs of activity anti-VEGF drugs. The general strategy is to apply
are present. For best functional results it is the minimal number of injections in order to
important to treat before irreversible damage nearly completely suppress VEGF activity.
occurs to the photoreceptors. Re-treatment is indicated at the earliest signs of
• OCT criteria: With the arrival of high-precision exudative recurrence such as the presence of
spectral-domain OCT (SD-OCT) allowing a intra- or subretinal fluids, in particular foveal flu-
precise examination of the entire 6 × 6 mm ids. However, even extrafoveal fluids represent a
macular cube, OCT has become the corner- threat to visual acuity due to underlying reactiva-
stone of re-treatment strategies. Any pathologi- tion of neovascular processes. Any such reactiva-
cal retinal fluid is easily detected by scanning tion might ultimately lead to further growth of the
through all acquired B-scans. However, a single neovascular complex. As this could result in fur-
line would be insufficient for decision making, ther loss of vision, any identification of relevant
as not only the fovea but also extrafoveal areas fluids is considered a criterion for re-treatment in
are relevant, needing treatment to prevent fur- order to prevent a disease progression.
ther neovascuar growth. Intra- or subretinal Recent studies revealed a difference in the
fluid is commonly considered a solid re-treat- functional and prognostic relevance of intra- ver-
ment criterion because it is usually a sign of sus subretinal fluids. It has been shown that intra-
active exudation from the neovascular mem- retinal fluids are associated with a worse visual
brane, requiring anti-VEGF re-treatment. outcome, whereas subretinal fluids appear to be
However, fluid under the retinal pigment epi- well tolerated [20, 21]. Another recent study
thelium is only by some studies considered a revealed that subretinal fluids up to 200 μm can be
re-treatment criterion [15]. This type of fluid tolerated even under the fovea without a visual dis-
is - after an initial improvement - insufficiently advantage, and the authors found non-inferiority
responsive to anti-VEGF treatment and appears versus no fluid tolerance [22]. Thus, the re-treat-
to be not relevant for re-treatment. Considering ment criteria in variable dosing regimens with
this type of fluid a re-treatment criterion does anti-VEGF drugs for nAMD might undergo some
not change the visual prognosis. However, changes in the near future, differentiating the roles
cases of high retinal pigment epithelium of subretinal from those of intraretinal fluids.
detachment generally require anyway a high Some open questions still need to be addressed:
number of injections as intra- or subretinal fluid
is frequently present and poorly responsive. • Degenerative fluid accumulation due to a loss
• Retinal hemorrhage: The fundus examination of retinal substance might simulate exudative
is the most useful investigation technique to fluids. However, reliable criteria to differenti-
identify retinal hemorrhage. New retinal hem- ate between them are missing. The best indi-
orrhage may be a sign of nAMD activity requir- cators of degenerative fluids might be so far:
ing anti-VEGF re-treatment. Most variable non-responsiveness to anti-VEGF drugs,
dosing regimens consider a new hemorrhage to overlying atrophy or fibrosis, no retinal thick-
be a re-treatment criterion. However, hemor- ness increase, and small intraretinal spaces.
rhage might occur independently of VEGF lev- • A subset of eyes presents refractory fluids
els and may not necessarily represent nAMD despite monthly treatments with anti-VEGF
activity. For instance, large neovascular feeder medication. The causes of such a refractory
vessels in combination with systemic hyperten- fluid are variable and often difficult to identify.
Some patients might just have high intraretinal anti-VEGF drugs. While there is an agreement
VEGF levels and an early recurrence, requiring that exudative fluid should be treated indepen-
ongoing monthly re-treatment with anti-VEGF dently of its location with respect to the fovea, it
drugs, while others might have a pathological might well be that a location within a nonfunc-
exudation of an origin other than VEGF, being tional atrophic or fibrotic area is not a good re-
completely non-responsive to anti-VEGF treat- treatment criterion: As long as functional
ment and, thus, not requiring monthly anti- regions of the retina are not threatened by exu-
VEGF injections. A non-responsive fluid might dation or neovascular membrane growth, it
be linked to inflammation, degenerative might not be needed to insist on VEGF suppres-
changes, central serous chorioretinopathy, pol- sion. However, the available evidence is insuf-
ypoidal choroidal vasculopathy, or other disor- ficient to give recommendations on this point.
ders. Obviously, adjuvant or alternative
treatment strategies could be considered
according to the cause of the refractory fluid. 8.3 Screening and Early
Steroids will be most useful in cases involving Discovery
inflammatory components, while PDT with
verteporfin would be an interesting treatment Anti-VEGF treatment has introduced a new era in
adjuvant for polypoidal choroidal vasculopathy the treatment of nAMD improving its prognosis.
(recently termed aneurysmal type 1 neovascu- Its efficacy with strong control over exudation
larization). A triple therapy combining anti- and vasoproliferation allows for good visual acu-
VEGF medication with both intravitreal ity improvements as compared to the treatment
steroids and PDT has also been suggested. baseline. The relative improvement is particularly
• The role of switching from one anti-VEGF good in eyes with poorer baseline vision [25].
molecule to another has been largely dis- However, the resulting vision level is the most
cussed in the literature, however, without a relevant outcome for patients. Even if a lower
convincing conclusion. It seems that refrac- baseline vision gains statistically more with anti-
tory cases might sometimes benefit from VEGF drugs, the resulting visual acuity remains
changing the anti-VEGF agent, and drug tol- lower than that of an early treated nAMD. A
erance may play a role [23]. However, clear long-standing untreated nAMD will show more
clinical indicators are missing. fibrosis and irreversible retinal damage, limiting
• The degree of structural and functional damage the potential functional gain of any treatment.
to the retina is currently poorly taken into Thus, it is well recognized that the early discov-
account in the treatment regimen. Progressive ery of a neovascular complication in AMD is
fibrosis and atrophy are the main reasons for extremely important for the final visual outcome.
progressive visual loss despite careful re- A variety of screening approaches are available
treatments with anti-VEGF drugs [24]. Anti- for the clinician and the patient. Clinical visits
VEGF treatment only addresses the exudative with visual acuity and SD-OCT are sensitive, but
part of the disorder. However, if visual loss they cannot be performed frequently enough for
becomes severe (<0.1) due to irreversible fibro- efficient screening. The cornerstone of screening
sis or atrophy, the usefulness of continued intra- is the home monitoring of well-educated patients.
vitreal injections is very limited. At some point, The oldest method goes back to Prof. Amsler, who
the anti-VEGF treatment does not make sense developed a simple grid with a central fixation
anymore. In the absence of a clearly defined point, printed on paper, to identify metamorphop-
limit, most clinicians will abandon the treatment sia (monocular examination reading correction
when the vision is reduced to counting fingers. and reading illumination). The appearance of new
• The topographic correlation between exudation metamorphopsia is an early sign of retinal defor-
and structural–functional retinal damage is mation, usually in the context of macular edema.
another point that has so far been poorly In a patient with known early signs of AMD such
addressed by variable dosing regimens with as drusen and pigmentary changes, there is a high
104 I. Mantel
probability that new metamorphopsia indicates slightly more cardiovascular events in the anti-
early exudative changes of nAMD. Such a patient VEGF arms have initiated meta-analysis studies.
needs to be seen by an ophthalmologist as soon as In a meta-analysis including 21 studies with 9557
possible within 2 weeks. patients, anti-VEGF treatment did not signifi-
Recently, a specially designed home monitor- cantly increase overall mortality or cardiovascu-
ing device (ForseeHome™) has been tested for lar mortality [27]. The occurrence of serious
its sensitivity and specificity in the screening of systemic adverse events was comparable across
neovascular complications in AMD [26]. It uses anti-VEGF-treated groups and control groups.
preferential hyperacuity perimetry. The use of Ocular inflammation and increased intraocular
this device led to the earlier recognition of neo- pressure after intravitreal injection were the most
vascular complications and a better visual out- frequently reported serious ocular adverse events
come than regular office visits (in combination [28]. Endophthalmitis was reported in fewer than
with the Amsler grid at the discretion of the 1% of anti-VEGF-treated participants; no cases
investigator). It is the first commercially avail- were reported in control groups. The transient
able device for this use and approved by the US increase in intraocular pressure after intravitreal
Food and Drug Administration (FDA). injection of 0.05 ml of the anti-VEGF drug might
In recent years, many electronic applications need special consideration in advanced glaucoma
have emerged, proposing nAMD screening based patients with low target pressure.
on both Amsler grid analysis and hyperacuity
perimetry. There are many minor and major vari-
ations available, and it is impossible to give clear 8.5 Combination Treatment
recommendations for one or the other applica- and Alternative Treatment
tion. However, a minimum of scientific evalua- Options
tion should be required in order to guarantee the
quality of size, contrast, color setting, etc. Anti-VEGF monotherapy is currently the best
Approval by the FDA is available for some of the available treatment strategy for nAMD. Its strong
applications. control over exudation and vasoproliferation
results in a high level of disease control. No alter-
native or adjuvant treatment has been able to
8.4 Safety of Anti-VEGF improve this excellent outcome further. However,
Treatment nAMD is a complex, multifactorial disorder.
Therefore, it appears attractive to add a second
Anti-VEGF treatment for nAMD is applied using line of action by an adjuvant treatment, poten-
intravitreal injections. Both ocular and systemic tially complementary to the anti-VEGF action.
safety concerns of the drug and the injection pro- So far, no combination treatment was able to
cedure need to be considered. improve the visual results. Recently, the promis-
Safety issues of systemic treatments, including approach of combining anti-VEGF antibod-
ing arterial hypertension and thromboembolic ies with pegpleranib E10030 (Fovista), an
events, are not applicable in the same way as the anti-platelet-derived growth factor antibody,
eye globe is a relatively closed system. However, failed to reach superiority to anti-VEGF mono-
small quantities of anti-VEGF antibodies are therapy (ClinicalTrials.gov Identifier:
absorbed into the systemic circulation, and in NCT01940900), although its pathophysiological
some patients, the systemic VEGF levels were basis was highly attractive (acting on pericytes in
critically reduced by 50%. However, the reported semi-mature neovessels in order to increase the
rates of systemic adverse events are low in all anti-VEGF sensitivity).
studies. No study showed a significant difference However, several other combination therapies
between anti-VEGF and comparison arms. are able to reduce the number of anti-VEGF
However, minor nonsignificant differences with injections. This appears to be particularly
attractive for the so-called “anti-VEGF-refrac- [33], whereas others found a very high rate of
tory nAMD.” However, little is known about pos- disease control in anti-VEGF monotherapy with
sible unwanted effects on visual acuity in these no added benefit from additional PDT treatment
cases. This would be important to avoid unneces- [34]. Thus, both options are currently considered
sary additional risks of unwanted effects due to valuable treatment approaches. In a setting with
adjuvant treatments. readily available anti-VEGF medication, PCV
PDT with verteporfin was the standard treat- patients will undergo anti-VEGF monotherapy as
ment for subfoveal nAMD before the arrival of a first-line treatment and change to a combination
anti-VEGF drugs. Although it improves the final with PDT in case of persisting exudative signs.
outcome compared to sham to some extent, a loss However, in cases with a priority for reduced
in the mean visual acuity was still a reality and numbers of injections and appointments, the
the efficacy, therefore, not satisfactory [29]. first-line combination of anti-VEGF drugs and
However, in cases of a contraindication of the PDT might be an interesting option. However,
anti-VEGF treatment for whatever reason, PDT this approach requires indocyanine green angiog-
might be an alternative to reduce visual loss, but raphy for diagnosis and treatment guidance.
patients need to be informed about its inferiority A broad combination treatment with anti-
compared to anti-VEGF therapy. VEGF antibodies, PDT, and intraocular steroid
The combination of anti-VEGF drugs with injections (triamcinolone) has also been pro-
PDT has been investigated in numerous studies. posed as the so-called triple therapy [35]. The
A recent meta-analysis included 16 randomized rationale combines the anti-VEGF and anti-
controlled trials comparing anti-VEGF mono- inflammatory effects to act on both the neovascu-
therapy and the combination treatment of anti- lar processes and the PDT-related unwanted
VEGF antibodies with PDT (standard or reduced effects, increasing the benefits of adjuvant PDT
fluence) [30]. This meta-analysis confirmed that on vascular occlusion. Although this treatment
they only differed in the number of anti-VEGF has been reported to be beneficial, there is no
injections needed, whereas visual acuity and cen- controlled clinical trial available. Therefore, no
tral retinal thickness changes did not differ [30]. clear recommendation can be made. This treat-
Interestingly, a subgroup analysis of adjuvant ment option might be of interest in some highly
full-fluence PDT did reveal a lower central retinal exudative, anti-VEGF-refractory cases.
thickness. This could be a warning sign, as full- The combination of anti-VEGF medication
fluence PDT has the potential to damage the cho- with stereotactic radiotherapy has been investi-
riocapillaris and the pigment epithelium. gated in nAMD eyes with high anti-VEGF
Particular interest has been given to the role of demand. The INTREPID trial found a reduced
PDT in polypoidal choroidal vasculopathy number of anti-VEGF injections over 2 years fol-
(PCV), a special subtype of nAMD. Before the lowing a single stereotactic radiotherapy session
arrival of anti-VEGF therapy, PDT showed some [36]. However, some radiation-related unwanted
promising results in PCV [31], while being rela- effects were also described but considered non-
tively poorly efficient in other types of significant for visual outcomes. Currently, a
nAMD. Since the arrival of the anti-VEGF treat- larger clinical trial is underway (ClinicalTrials.
ment, an improved visual benefit was found not gov Identifier: NCT02243878), and first results
only for nAMD in general but also for might be expected for 2024.
PCV. However, several studies have shown in Large submacular hemorrhage is a particular
PCV an interesting potential of anti-VEGF drugs challenge in the management of nAMD. Dramatic
with adjuvant PDT [32]. Their results are moder- visual loss occurs if the hemorrhage is central,
ately variable; some of them suggesting a supe- and photoreceptors rapidly suffer from irrevers-
rior visual outcome and a higher rate of polyp ible damage when in direct contact with a thick
closure along with a decreased number of anti- layer of blood (subretinal hemorrhage). In such
VEGF injections in the combination therapy cases, a combination of vitrectomy, subretinal
106 I. Mantel
tissue plasminogen activator, and intravitreal gas plays a key role in follow-up and
might be considered, if the bleeding is very decision-making.
recent. Outcomes vary widely, and systematic tri- • A variety of treatment regimens are available
als are lacking [37]. to the clinician. The key to functional success
Very recently, the mineralocorticoid receptor is not the choice of the regimen but the careful
pathway has been suggested to be implicated in application of its rules, including monitoring
the pathogenesis of choroidal neovascularization, intervals and re-treatment criteria.
based on animal models [38]. The clinical appli- • Undertreatment threatens the vision more than
cation has only been tested in a clinical pilot overtreatment.
study [38] but may potentially become an inter- • Most importantly, early discovery is crucial
esting future option. for good final visual results. Screening meth-
ods and patient education are most helpful.
• Some cases do not entirely respond to anti-
8.6 Future Challenges VEGF treatment. Their best management is
not well established. Combination treatment
Despite the breakthrough in the nAMD prognosis with other modalities may be considered.
due to anti-VEGF treatment options, there are
many unmet needs in nAMD. The underlying
degenerative process frequently leads to macular
atrophy and profound visual loss despite “suc- References
cessful” control of the neovascular process.
1. Brown DM, Kaiser PK, Michels M, et al. Ranibizumab
Fibrosis is not inhibited by anti-VEGF treatment,
versus verteporfin for neovascular age-related
a major problem particularly in type 2 neovascu- macular degeneration. N Engl J Med. 2006;355:
larization (classic neovascular membrane), and 1432–44.
responsible for irreversible visual losses. Massive 2. Rosenfeld PJ, Brown DM, Heier JS, et al. Ranibizumab
for neovascular age-related macular degeneration. N
macular hemorrhage may occur in nAMD despite
Engl J Med. 2006;355:1419–31.
maximal monthly re-treatment with anti-VEGF 3. Heier JS, Brown DM, Chong V, et al. Intravitreal
medication, causing a profound acute loss in aflibercept (VEGF trap-eye) in wet age-related macu-
visual acuity. The chronicity of the exudative dis- lar degeneration. Ophthalmology. 2012;119:2537–48.
4. Adamis AP, Shima DT, Tolentino MJ, et al. Inhibition
order requiring repetitive intravitreal injections is
of vascular endothelial growth factor prevents retinal
a major problem for patients and healthcare pro- ischemia-associated iris neovascularization in a non-
viders. More durable and less expensive solutions human primate. Arch Ophthalmol. 1996;114:66–71.
would be beneficial. Long-term delivery systems 5. Tolentino MJ, Miller JW, Gragoudas ES,
Chatzistefanou K, Ferrara N, Adamis AP. Vascular
or small interfering RNAs are promising future
endothelial growth factor is sufficient to produce
possibilities. Artificial intelligence might become iris neovascularization and neovascular glau-
a useful tool to monitor and manage large num- coma in a nonhuman primate. Arch Ophthalmol.
bers of patients. Finally, it could become a reality 1996;114:964–70.
6. Gragoudas ES, Adamis AP, Cunningham ET Jr,
to recover lost vision with stem cell techniques or
Feinsod M, Guyer DR. Group VISiONCT. Pegaptanib
artificial retinal implants. Obviously, the best for neovascular age-related macular degeneration. N
solution would be to find an efficient prophylac- Engl J Med. 2004;351:2805–16.
tic treatment to prevent that AMD compromises 7. Bressler NM, Bressler SB. Photodynamic therapy
with verteporfin (Visudyne): impact on ophthalmol-
the vision.
ogy and visual sciences. Invest Ophthalmol Vis Sci.
2000;41:624–8.
Key Learning Points 8. Brown DM, Michels M, Kaiser PK, Heier JS, Sy JP,
Ianchulev T. Ranibizumab versus verteporfin photo-
• The clinical management of nAMD requires
dynamic therapy for neovascular age-related macular
careful monitoring and sensitive re-treatment degeneration: two-year results of the ANCHOR study.
criteria for anti-VEGF treatment. SD-OCT Ophthalmology. 2009;116:57–65.
9. Abraham P, Yue H, Wilson L. Randomized, double- 22. Guymer RH, Markey CM, McAllister IL, Gillies MC,
masked, sham-controlled trial of ranibizumab for Hunyor AP, Arnold JJ. Tolerating subretinal fluid in
neovascular age-related macular degeneration: PIER neovascular age-related macular degeneration treated
study year 2. Am J Ophthalmol. 2010;150:315–24. with ranibizumab using a treat-and-extend regimen:
10. Schmidt-Erfurth U, Eldem B, Guymer R, et al. FLUID study 24-month results. Ophthalmology.
Efficacy and safety of monthly versus quarterly ranibi- 2019;126:723–34.
zumab treatment in neovascular age-related macular 23. Mantel I, Gillies MC, Souied EH. Switching between
degeneration: the EXCITE study. Ophthalmology. ranibizumab and aflibercept for the treatment of
2011;118:831–9. neovascular age-related macular degeneration. Surv
11. Regillo CD, Brown DM, Abraham P, et al. Ophthalmol. 2018;63:638–45.
Randomized, double-masked, sham-controlled trial 24. Rosenfeld PJ, Shapiro H, Tuomi L, et al.
of ranibizumab for neovascular age-related macular Characteristics of patients losing vision after 2 years
degeneration: PIER study year 1. Am J Ophthalmol. of monthly dosing in the phase III ranibizumab clini-
2008;145:239–48. cal trials. Ophthalmology. 2011;118:523–30.
12. Fung AE, Lalwani GA, Rosenfeld PJ, et al. An opti- 25. Kaiser PK, Brown DM, Zhang K, et al. Ranibizumab
cal coherence tomography-guided, variable dosing for predominantly classic neovascular age-related
regimen with intravitreal ranibizumab (Lucentis) for macular degeneration: subgroup analysis of first-year
neovascular age-related macular degeneration. Am J ANCHOR results. AmJ Ophthalmol. 2007;144:850–7.
Ophthalmol. 2007;143:566–83. 26. Group AHSR, Chew EY, Clemons TE, et al.
13. Lalwani GA, Rosenfeld PJ, Fung AE, et al. A Randomized trial of a home monitoring system for
variable-dosing regimen with intravitreal ranibi- early detection of choroidal neovascularization home
zumab for neovascular age-related macular degenera- monitoring of the eye (HOME) study. Ophthalmology.
tion: year 2 of the PrONTO study. Am J Ophthalmol. 2014;121:535–44.
2009;148:43–58. 27. Thulliez M, Angoulvant D, Le Lez ML, et al.
14. Busbee BG, Ho AC, Brown DM, et al. Twelve-month Cardiovascular events and bleeding risk associated
efficacy and safety of 0.5 mg or 2.0 mg ranibizumab in with intravitreal antivascular endothelial growth factor
patients with subfoveal neovascular age-related macu- monoclonal antibodies: systematic review and meta-
lar degeneration. Ophthalmology. 2013;120:1046–56. analysis. JAMA Ophthalmol. 2014;132:1317–26.
15. Martin DF, Maguire MG, Ying GS, Grunwald JE, 28. Solomon SD, Lindsley K, Vedula SS, Krzystolik MG,
Fine SL, Jaffe GJ. Ranibizumab and bevacizumab for Hawkins BS. Anti-vascular endothelial growth fac-
neovascular age-related macular degeneration. N Engl tor for neovascular age-related macular degeneration.
J Med. 2011;364:1897–908. Cochrane Database Syst Rev. 2014;8(8):CD005139.
16. Silva R, Berta A, Larsen M, et al. Treat-and-extend 29. Bressler NM, Treatment of Age-Related Macular
versus monthly regimen in Neovascular age-related Degeneration with Photodynamic Therapy Study
macular degeneration: results with Ranibizumab from G. Photodynamic therapy of subfoveal choroidal
the TREND study. Ophthalmology. 2018;125:57–65. neovascularization in age-related macular degenera-
17. Mantel I, Deli A, Iglesias K, Ambresin A. Prospective tion with verteporfin: two-year results of 2 random-
study evaluating the predictability of need for retreat- ized clinical trials-tap report 2. Arch Ophthalmol.
ment with intravitreal ranibizumab for age-related 2001;119:198–207.
macular degeneration. Graefe’s Arch Clin Exp 30. Gao Y, Yu T, Zhang Y, Dang G. Anti-VEGF
Ophthalmol. 2013;251:697–704. Monotherapy versus photodynamic therapy and anti-
18. Gianniou C, Dirani A, Ferrini W, et al. Two-year out- VEGF combination treatment for Neovascular age-
come of an observe-and-plan regimen for neovascular related macular degeneration: a meta-analysis. Invest
age-related macular degeneration: how to alleviate Ophthalmol Vis Sci. 2018;59:4307–17.
the clinical burden with maintained functional results. 31. Nowak-Sliwinska P, van den Bergh H, Sickenberg
Eye (London). 2015;29:342–9. M, Koh AH. Photodynamic therapy for polypoi-
19. Mantel I, Niderprim SA, Gianniou C, Deli A, dal choroidal vasculopathy. Prog Retin Eye Res.
Ambresin A. Reducing the clinical burden of ranibi- 2013;37:182–99.
zumab treatment for neovascular age-related macular 32. Teo KYC, Gillies M, Fraser-Bell S. The use of vascu-
degeneration using an individually planned regimen. lar endothelial growth factor inhibitors and comple-
Br J Ophthalmol. 2014;98:1192–6. mentary treatment options in Polypoidal Choroidal
20. Jaffe GJ, Martin DF, Toth CA, et al. Macular mor- vasculopathy: a subtype of Neovascular age-related
phology and visual acuity in the comparison of macular degeneration. Int J Mol Sci. 2018;19(9):2611.
age-related macular degeneration treatments trials. 33. Koh A, Lai TYY, Takahashi K, et al. Efficacy and
Ophthalmology. 2013;120:1860–70. safety of Ranibizumab with or without Verteporfin
21. Sharma S, Toth CA, Daniel E, et al. Macular mor- photodynamic therapy for Polypoidal Choroidal
phology and visual acuity in the second year of vasculopathy: a randomized clinical trial. JAMA
the comparison of age-related macular degenera- Ophthalmol. 2017;135:1206–13.
tion treatments trials. Ophthalmology. 2016;123: 34. Lee WK, Iida T, Ogura Y, et al. Efficacy and safety
865–75. of Intravitreal Aflibercept for Polypoidal Choroidal
108 I. Mantel
vasculopathy in the PLANET study: a randomized 37. Stanescu-Segall D, Balta F, Jackson TL. Submacular
clinical trial. JAMA Ophthalmol. 2018;136:786–93. hemorrhage in neovascular age-related macular
35. Augustin AJ, Puls S, Offermann I. Triple therapy for degeneration: a synthesis of the literature. Surv
choroidal neovascularization due to age-related macu- Ophthalmol. 2016;61:18–32.
lar degeneration: verteporfin PDT, bevacizumab, and 38. Zhao M, Mantel I, Gelize E, et al. Mineralocorticoid
dexamethasone. Retina. 2007;27:133–40. receptor antagonism limits experimental choroidal
36. Jackson TL, Chakravarthy U, Slakter JS, et al. neovascularization and structural changes associated
Stereotactic radiotherapy for neovascular age-related with neovascular age-related macular degeneration.
macular degeneration: year 2 results of the INTREPID Nat Commun. 2019;10:369.
study. Ophthalmology. 2015;122:138–45.
Irmela Mantel, MD is currently a medical retina lead

consultant at the Jules Gonin Eye Hospital in Lausanne,
Switzerland. She is actively involved in a number of
research projects investigating retinal degenerative disor-
ders and in particular age-related macular degeneration.
Dr. Mantel received her medical degree from the
University of Zurich, Switzerland. She has been a medical
retina fellow to Professor Alan Bird at Moorfields Eye
Hospital in London, England.
Myopic Choroidal
Seung-Young Yu and Kiyoung Kim
9.1 Incidence and Related the macula. Through slit examination, gray circu-
Lesions lar or oval-shaped pigmentation is observed in
the border of macular lesions, and lesions can be
Pathologic myopia, termed “myopic macular confined in the macular or elsewhere. Fuch’s spot
degeneration,” “myopic maculopathy,” or is regarded as evidence of myopic CNV, which as
“degenerative myopia,” is a consequence of myo- Curtin and Karlin detected Fuch’s spot in 5.2% of
pia progression, particularly in eyes with high myopic eyes longer than 26.5 mm, and Hotchkiss
myopia (typically defined as spherical equivalent and Fine found neovascularization in 40.7% of
less than −6.0 diopters). One of the most serious high myopia. Incidence of bilateral Fuch’s spot
complications of pathologic myopia is myopic has been reported between 12 and 41% by differ-
choroidal neovascularization (CNV), which fre- ent authors. Particularly, 62% of total myopic
quently leads to a sudden but progressive loss of CNV has developed in younger aged than 50, and
central vision. when CNV occurs in one eye, the possibility of
Myopic CNV occurs in 5–11% of pathologic development in the fellow eye is estimated to be
myopia, and the second common lesion in macu- higher than 30% within 8 years.
lar CNV followed by age-related macular degen-
eration (AMD). The prevalence and incidence of
myopic CNV may differ between each geo- 9.2 Clinical Manifestation
graphic area and maybe underestimated or inac-
curate because of insufficient screening. Since Although, vision progressively decreases from
1991, Fuchs named fibrovascular tissue elevated the 40s in general pathologic myopia, a rapid
with dome-shaped retinal pigment epithelium vision decreases and metamorphopsia cab be
(RPE) as Foester–Fuch’s spot in high myopia, caused by marginal exudation and hemorrhage in
Gass defined that acute hemorrhage detachment myopic CNV. Focal retinal detachment and hem-
primary occurs after the formation of neovascular orrhage can develop around CNV, but hard exu-
membrane from Fuch’s spot, secondly, organized dates are rarely detected [1].
sub-RPE hemorrhage accompanied with RPE Neovascular tissue is usually located under-
proliferation remains as a blackish brown spot in neath fovea. 58–74% of subretinal CNV involve
fovea and size is less than half of optic disc. The
S.-Y. Yu (*) · K. Kim other CNVs also exist within 100–300 μm of
Department of Ophthalmology, Kyung Hee fovea. These foveal CNV is one of the specific
University Hospital, Seoul, South Korea clinical characteristics of myopic CNV [2].
e-mail: syyu@khu.ac.kr

110 S.-Y. Yu and K. Kim
9.3 Fluorescein Angiography 9.4 Optical Coherence

and Indocyanine Green Tomography (OCT) Findings
Angiography
High-resolution OCT is a rapid, noninvasive
Early hyperfluorescence and late leakage of technique that has been used efficiently to deter-
foveal CNV were observed in Fluorescein angi- mine disease activity of myopic CNV. Several
ography (FA). However, leakage is not severely studies have successfully demonstrated the effec-
increased as CNV of AMD, and confined to the tiveness of SD-OCT in assessing the activity and
border of neovascular membrane. If a relatively stage of the disease [1]. Compared with FA, myo-
large amount of leakage is present in old myopic pic CNV presents as a hyper-reflective material
patients, AMD should be considered as a main above the RPE band (type 2 CNV) on SD-OCT,
cause. with a variable amount of subretinal fluid.
Avila et al. [3] reported that 91% of myopic Especially, it is helpful to evaluate the effective-
eyes with severe choroidal degeneration did ness of treatment by measuring subretinal fluid
not present fluorescein leakage in CNV and from CNV. If CNV locates above RPE layer and
80% of myopic eyes with mild choroidal enough space is available around the lesion, it
degeneration presented fluorescein leakage. can be an indication for surgical removal of CNV.
This result supports that choroidal change of
pathologic myopia is related to neovasculariza-
tion which comes from the choroid. Choroidal 9.5 Optical Coherence
blood flow decreases and choriocapillaris cir- Tomography Angiography
culation are severely impaired in pathologic (OCTA)
myopia. Furthermore, extensive choroidal
atrophy causes loss of choroidal stroma and The recent development of OCTA has allowed
partial occlusion of the vessel, which conse- accurate delineation of structural and blood flow
quently induces the formation of a relatively information from different retina and choroid
small size of CNV. In the early phase of indo- layers. OCTA has been investigated to the study
cyanine green angiography (ICGA), CNV of several retinal disorders, and it also has proven
appears as a hyperfluorescent lesion of similar particularly useful for the diagnosis of CNV in
intensity with choroidal background. ICGA pathologic myopia. In OCTA, CNV appears as a
can be used in addition to FA, because it can large hyperintense vascular anastomotic network
provide more informative data on the choroidal (Fig. 9.1). The use of OCTA may further clarify
circulation, along with the presence and extent the role of choroidal blood flow and retinal
of lacquer cracks. CVN is delineated with function.
hypofluorescent border and hyperfluorescent A recently developed swept-source OCTA
CNV continuously increases with choroidal (SS-OCTA) uses longer wavelength of approxi-
arterial filling and maintains during choroidal mately 1050 nm compared with spectral-domain
venous phase. Neovascular membrane can OCTA (SD-OCTA: 840–880 nm). The main
remain as relatively hypofluorescent spot in advantages of SS-OCTA over SD-OCTA is the
late phase, but leakage does not occur. higher scanning speed, which allows for denser
Although, it can be difficult to identify the A-scan and B-scans and larger scan areas for a
exact margin and range of neovascular mem- same acquisition time. Additional advantages of
brane, ICGA is a useful method to investigate SS-OCTA technology are enhanced light pene-
the relationship between CNV and lacquer tration through the RPE, as well as better detec-
crack and detect hidden CNV which cannot be tion of signals from the deeper retinal layers. This
discovered with FA in myopic CNV patients potentially allows for more detailed visualiza-
with subretinal hemorrhage. tion, enhanced penetration through exudative
9 Myopic Choroidal Neovascularization 111
a b
c d
e f
Fig. 9.1 Multimodal imaging of a 56-year-old female spectral-domain and swept-source OCTA shows a clear
patient presenting myopic choroidal neovascularization vascular network in myopic CNV (yellow circle). (g)
(CNV). (a) Fundus color photo, (b) Optical coherence AngioPlex (spectral-domain OCTA): Carl Zeiss Meditec,
tomography (OCT), (c–d) Early and late fluorescein angi- Inc., Dublin, CA, USA; (h) Spectralis (Spectral-domain
ography (FA), (e–f) early and late indocyanine green angi- OCTA): Heidelberg Engineering, Heidelberg, Germany;
ography (g–i) Comparisons of three different OCT (i) Swept-source OCTA: PLEX Elite 9000; Carl Zeiss
angiography (OCTA) devices in myopic CNV. Both Meditec, Inc., Dublin, CA, USA
g h i
material or subretinal fluid, and better detection 29.6%. However, at 10 years of final visit, only
of neovascular complex compared with one eye remains with visual acuity over 20/40,
SD-OCTA. SS-OCTA can also provide better and 96.3% of patients presented decreased
visualization of the choroid and choriocapillaris visual acuity lower than 20/200. Hemorrhage
layer, which can detect neovascularization or associated with CNV is absorbed after an aver-
abnormal choriocapillaris circulation. Novais age 7.6 months, and rebleeding is rare. All neo-
et al. [4] showed that SS-OCTA is more accu- vascularization generally totally regresses and
rately able to demarcate the full lesions of CNV flattens after 5–10 years of occurrence, and is
vasculature compared with SD-OCTA. even difficult to identify in some cases. CNV
Although OCTA may delineate the total area activity shows generally weaker than AMD and
of neovascularization, there are still clinical limi- spontaneous remission. However, regressed
tations owing to its diagnostic sensitivity and neovascular membrane itself and surrounding
specificity. OCTA is not able to reflect the level atrophy progress and expands. This enlarge-
of activity and reveals new vessels with blood ment of atrophy can cause progressive central
flow even in the scar or atrophic phase of myopic visual loss.
CNV. Secondly, the presence of macular hemor- According to the long-term retrospective
rhage may result in the masking of CNV signals. observational study, the visual prognosis of myo-
In these cases, OCTA may not detect any blood pic CNV is determined by age at onset, with
flow signal, while FA presents leakage in corre- patients aged >40 years having a worse prognosis
sponding lesions. compared to early-onset disease. Therefore,
patient age at which an individual develops dis-
ease should be considered when evaluating the
9.6 Clinical Course therapeutic management and course. Kojima
et al. [6] reported that areas of chorioretinal atro-
Yoshida et al. [5]. reported natural course of phy are the most significant factors of poor long-
myopic CNV for 10 years. At baseline, the per- term visual prognosis, and older age and larger
centage of patients with visual acuity better CNV area were also associated factors for the
than 20/40 was 22.2%, lower than 20/200 was onset of atrophy.
9.7 Treatment removal of CNV: visual increase in 39%, stable

visual acuity in 26%, and visual decrease in
9.7.1 Laser 35%. However, Ruiz-Moreno et al. reported no
significant visual increase in 22 eyes.
Krypton red laser photocoagulation has been Recurrence rate was reported between 18
developed to prevent visual loss in extra-foveal and 57%.
myopic CNV. Krypton red laser was used to treat Macular translocation is a surgical method
myopic eyes with less macular pigment because that relocates the fovea away from CNV. There
red laser light of a krypton laser is preferentially are two surgical methods which are vitrectomy,
absorbed in melanin of choroid and occlude posterior vitreous detachment, relocating
choroid derived neovascularization. At 2 years foveal with air–fluid exchange, and moving
follow-up, there was a significant difference in macula enough by 360 degree peripheral reti-
treatment and control group in regard to far and notomy. Tano et al. reported outcomes of mac-
near vision, but no significant difference at 5-year ular translocation in 28 eyes, in which 75%
follow-up. Central visual acuity decreased as showed BCVA improvement over 2 lines, 11%
recurrence of CNV and enlargement of laser scar showed stable BCVA, and 14% exhibited
after photocoagulation. Recurrence rate is BCVA decreases. However, this method is
31–71% and usually occurs with laser scars. With recently rarely performed due to postoperative
other wavelength lasers, treatment scar progres- complications such as retinal detachment, rota-
sively expands in 92–100%. As photocoagulation tional diplopia, and difficulty of surgical
scar and atrophic lesion are progressively procedure.
enlarged and cause final visual loss, photocoagu-
lation treatment is not recommended; otherwise,
lesion locates far away from the fovea. 9.7.4 Anti-vascular Endothelial
Growth Factor (Anti-VEGF)
Treatment
9.7.2 Photodynamic Therapy (PDT)
Since 2005, Nguyen et al. reported visual
VIP (verteporfin in photodynamic therapy) study increase after bevacizumab injection in two
group attempted PDT in myopic CNV occurring myopic CNV, anti-VEGF treatment including
in the fovea. Percentage of subjects with best- intravitreal ranibizumab injection has been
corrected visual acuity (BCVA) increasing over widely used as myopic CNV. A total of four
eight letters was 72%, higher than the control intravitreal anti-VEGF drugs have been clini-
group (44%) at 1 year, but there was no signifi- cally studied for the treatment of myopic
cant difference between the two groups at 2-year CNV. Of these, only Ranibizumab and
follow-up. VIP group suggested that even results Aflibercept are currently approved for the use of
based on increase of BCVA over eight letters did myopic CNV [7–9]. Twenty-four months’ com-
not show a significant difference; overall BCVA parison study between PDT and intravitreal bev-
distribution is better in the treatment group than acizumab injection showed that PDT can regress
the control group, which can be an evidence of CNV and decrease central retinal thickness
superiority for PDT in myopic CNV. while being less effective in BCVA improvement
[10]. Intravitreal bevacizumab treatment was
reported to maintain stable visual acuity until
9.7.3 Surgery 24 months [11]. A systematic review and meta-
analysis of anti-VEGF therapy for myopic CNV
Surgical method was tried to remove CNV or has concluded that intravitreal anti-VEGF injec-
macular translocation. Uemura and Thomas tion should be considered as a first-line treat-
reported results of 23 eyes with surgical ment [12, 13].
9.7.5 rognostic Factors Associated

P CNV activity is confirmed using FA with OCT,
with Anti-VEGF Treatment prompt treatment with intravitreal anti-VEGF ther-
Outcomes apy should be administered. Beneficial evidences
of pro re nata (PRN) dosing regimen is supported
Yang et al. [14] reported that baseline CNV lesion by the results of the REPAIR (Ranibizumabfor
size was significantly associated with recurrence, trEatment of CNV secondary to Pathological myo-
and recurrence rate during follow-up was 23%. pia: An Individualized Regimen) and RADIANCE
Ahn et al. [15] found that a thinner baseline cho- (Ranibizumab And PDT evAluation iN myopic
roid thickness was associated with incomplete Choroidal nEovascularization) studies, in which
regression of myopic CNV after a single anti- patients were treated with a single injection of
VEGF injection and with higher 1-year recur- ranibizumab followed by PRN dosing based on the
rence of myopic CNV. Baseline BCVA and disease activity present at each follow-up visit
location of neovascularization have been signifi- (Fig. 9.2). The results 1 or 3 number of initial dos-
cantly correlated with visual outcome at the end ing schedules of intravitreal injection have been
of follow-up and associated with the need for re- compared, resulting that similar visual outcome
treatment. A systematic review of intravitreal was reported in both injection strategies [20].
bevacizumab studies identified a lower rate of
development of chorioretinal atrophy, younger
age, and smaller myopic CNV size as the factors 9.8 Follow-Up
most consistently associated with BCVA
improvement [16, 17]. Other factors identified as After anti-VEGF treatment, SD-OCT should be
less frequently correlated with visual outcome performed monthly to identify the presence or
included spherical equivalent, chorioretinal atro- absence of CNV activity and the need for addi-
phy enlargement, choroidal thickness, presence tional treatment. Disease activity is determined
of lacquer crack extending to the fovea, and size by decreased visual acuity, new or persistent
of peripapillary chorioretinal atrophy [18]. visual symptoms or signs of myopic CNV dis-
ease activity on FA and on SD-OCT. OCTA
might be also helpful to confirm the presence of
9.7.6 First-Line Therapy newly developed CNV. FA is recommended to
determine CNV activity when considering re-
The first-line treatment for myopic CNV recently treatment. For the recurrent CNV, intravitreal
has become anti-VEGF therapy [19]. If myopic anti-VEGF should be promptly re-administered.
a b
Fig. 9.2 Changes of SD-OCT after intravitreal anti- ment junction (b) Six months after 3 times anti-VEGF
VEGF injection in myopic CNV. (a) The baseline spectral- injection, SD-OCT scan shows the disappearance of the
domain-optical coherence tomography (SD-OCT) scan subretinal fluid with a reduced size of the subretinal
shows subretinal hyper-reflective material with fuzzy hyper-reflective material
margin and the absence of the inner segment/outer seg-
Annual retinal screening and examination the safety of ranibizumab in subjects with neovascu-
lar age-related macular degeneration. Ophthalmology.
should be conducted with an increased risk of 2009;116:1731–9.
developing myopic CNV without any visual 8. Bressler NM, Boyer DS, Williams DF, Butler S,
symptoms. Bilateral myopic CNV can occur in Francom SF, Brown B, Di Nucci F, Cramm T, Tuomi
up to one-third of patients. So, the fellow eye is LL, Ianchulev T, Rubio RG. Cerebrovascular acci-
dents in patients treated for choroidal neovasculariza-
also required to be examined for the development tion with ranibizumab in randomized controlled trials.
of CNV. In a view of a stable period over 1 year Retina. 2012;32:1821–8.
without recurrence, the major concern is the 9. Rosenfeld PJ, Brown DM, Heier JS, Boyer DS, Kaiser
development of chorioretinal atrophy which pro- PK, Chung CY, Kim RY. Ranibizumab for neovascu-
lar age-related macular degeneration. N Engl J Med.
gressively involves central fovea. The most chal- 2006;355:1419–31.
lenging problem of long-term therapy for myopic 10. Verteporfin in Photodynamic Therapy Study Group.
CNV is decreased visual acuity due to CNV- Photodynamic therapy of subfoveal choroidal neo-
related macular atrophy even after successful vascularization in pathologic myopia with vertepor-
fin. 1-year results of a randomized clinical trial–VIP
CNV remission. Therefore, regular check-up report no. 1. Ophthalmology. 2001;108:841–52.
should include fundus autofluorescence imaging 11. Sarao V, Veritti D, Macor S, Lanzetta P. Intravitreal
to detect early changes of macular atrophy. bevacizumab for choroidal neovascularization due
to pathologic myopia: long-term outcomes. Graefes
Arch Clin Exp Ophthalmol. 2016;254:445–54.
12. Ruiz-Moreno JM, Arias L, Montero JA, Carneiro A,
References Silva R. Intravitreal anti-VEGF therapy for choroidal
neovascularisation secondary to pathological myopia:
1. Liang IC, Shimada N, Tanaka Y, Nagaoka N, 4-year outcome. Br J Ophthalmol. 2013;97:1447–50.
Moriyama M, Yoshida T, Ohno-Matsui K. Comparison 13. Ruiz-Moreno JM, Montero JA, Araiz J, Arias L,
of clinical features in highly myopic eyes with and Garcia-Layana A, Carneiro A, Figueroa MS, Silva
without a dome-shaped macula. Ophthalmology. R. Intravitreal anti-vascular endothelial growth fac-
2015;122:1591–600. tor therapy for choroidal neovascularization second-
2. Ohno-Matsui K, Yoshida T. Myopic choroidal neovas- ary to pathologic myopia: six years outcome. Retina.
cularization: natural course and treatment. Curr Opin 2015;35:2450–6.
Ophthalmol. 2004;15:197–202. 14. Yang HS, Kim JG, Kim JT, Joe SG. Prognostic fac-
3. Avila MP, Weiter JJ, Jalkh AE, Trempe CL, Pruett RC, tors of eyes with naïve subfoveal myopic choroidal
Schepens CL. Natural history of choroidal neovascu- neovascularization after intravitreal bevacizumab. Am
larization in degenerative myopia. Ophthalmology. J Ophthalmol. 2013;156(6):1201–10.
1984;91:1573–81. 15. Ahn SJ, Park KH, Woo SJ. Subfoveal choroidal
4. Novais EA, Adhi M, Moult EM, Louzada RN, Cole thickness changes following anti-vascular endo-
ED, Husvogt L, Lee B, Dang S, Regatieri CV, Witkin thelial growth factor therapy in myopic choroidal
AJ, Baumal CR, Hornegger J, Jayaraman V, Fujimoto neovascularization. Invest Ophthalmol Vis Sci.
JG, Duker JS, Waheed NK. Choroidal neovasculariza- 2015;56:5794–800.
tion analyzed on ultrahigh-speed swept-source opti- 16. Shimada N, Ohno-Matsui K, Hayashi K, Yoshida
cal coherence tomography angiography compared to T, Tokoro T, Mochizuki M. Macular detachment
spectral-domain optical coherence tomography angi- after successful intravitreal bevacizumab for myo-
ography. Am J Ophthalmol. 2016;164:80–8. pic choroidal neovascularization. Jpn J Ophthalmol.
5. Yoshida T, Ohno-Matsui K, Yasuzumi K, Kojima 2011;55:378–82.
A, Shimada N, Futagami S, Tokoro T, Mochizuki 17. Vance SK, Khan S, Klancnik JM, Freund
M. Myopic choroidal neovascularization: a 10-year KB. Characteristic spectral-domain optical coherence
follow-up. Ophthalmology. 2003;110:1297–305. tomography findings of multifocal choroiditis. Retina.
6. Kojima A, Ohno-Matsui K, Teramukai S, Yoshida T, 2011;31:717–23.
Ishihara Y, Kobayashi K, Shimada N, Yasuzumi K, 18. Kim YM, Yoon JU, Koh HJ. The analysis of lacquer
Futagami S, Tokoro T, Mochizuki M. Factors asso- crack in the assessment of myopic choroidal neovas-
ciated with the development of chorioretinal atrophy cularization. Eye (Lond). 2011;25:937–46.
around choroidal neovascularization in pathologic 19. Wakabayashi T, Ikuno Y, Oshima Y, Hamasaki T,
myopia. Graefes Arch Clin Exp Ophthalmol. Nishida K. Aqueous concentrations of vascular
2004;242:114–9. endothelial growth factor in eyes with high myopia
7. Boyer DS, Heier JS, Brown DM, Francom SF, with and without choroidal neovascularization. J
Ianchulev T, Rubio RG. A phase IIIb study to evaluate Ophthalmol. 2013;2013:257381.
20. Wolf S, Balciuniene VJ, Laganovska G, Menchini

U, Ohno-Matsui K, Sharma T, Wong TY, Silva R,
Pilz S, Gekkieva M, RADIANCE Study Group.
RADIANCE: a randomized controlled study of
ranibizumab in patients with choroidal neovas-
cularization secondary to pathologic myopia.
Ophthalmology. 2014;121(3):682–92.
Kiyoung Kim, MD, PhD, is a current clinical associate

professor at the Department of Ophthalmology, Kyung
Hee University Hospital. His current research interests
include updated retinal imaging in retinal vascular dis-
eases and AMD. His recent published articles have
focused on OCTA imaging in diabetic retinopathy.
Seung-Young Yu, MD, PhD, is a current professor at

the Department of Ophthalmology, Kyung Hee University
Hospital. She has an in-depth understanding of the latest
systems in retinal imaging, particularly in AMD and dia-
betic retinopathy, including OCT technologies and fundus
autofluorescence systems.
Pachychoroid-Related Choroidal
Apoorva Ayachit and Jay Chhablani
10.1 Introduction logical changes, focally. The choroidal thickness

is normally thickest subfoveally, followed by
The word pachychoroid is derived from the Greek temporal fovea and the nasal fovea. In peripapil-
word παχύ (pachy—thick). The pachychoroid lary pachychoroid, however, the nasal choroid is
spectrum refers to a group of conditions that are thicker than the subfoveal and temporal choroid.
characterized by a thickened choroid. The first dis- Choroidal thickness can be influenced by age,
ease described in this spectrum was pachychoroid axial length, refractive error, blood pressure, and
pigment epitheliopathy (PPE) (Warrow and col- diurnal variation. It varies across ethnicities and
leagues). PPE was described as a condition causing the imaging platforms used. A wide range of val-
RPE changes similar to central serous chorioreti- ues have hence been reported in normal subjects
nopathy (CSC) but without the attendant subretinal and there is no set threshold for thickness defin-
fluid. Subsequently CSC, neovasculopathy, and ing pachychoroid. Eyes with CSC and PNV have
polypoidal choroidal vasculopathy characterized been noted to have choroidal thickness of 345–
by the commonality of a thickened choroid were 505 μm [1, 2] and 223–590 μm [3, 4], respec-
considered a part of the pachychoroid spectrum. tively. In these studies, the contralateral normal
Recently, Cheung et al. included focal choroidal eyes also had increased choroidal thickness [2].
excavation and peripapillary pachychoroid syn- In view of the wide range of choroidal thickness,
drome in the expanded spectrum of pachychoroid. researchers often use the cutoff of 300 μm for
subfoveal choroidal thickness for defining
pachychoroid.
10.2 Pathophysiology The key feature of pachychoroid, however, is
of Pachychoroid not only the quantitative measurement of thick-
ness. Focal/diffuse choriocapillaris effacement/
Increased choroidal thickness is considered a attenuation is considered characteristic of pachy-
marker of pachychoroid. The thickening could be choroid. The attenuation is accompanied by dila-
either diffuse or correspond spatially to patho- tation of the outer choroid, i.e., the Haller’s layer.
The thinning of choriocapillaris may cause the
A. Ayachit outer choroid to occupy the full extent of the cho-
Department of Vitreo-retina, roid [5]. Considering the various factors affecting
M M Joshi Eye Institute, Hubballi, Karnataka, India choroidal thickness, pachychoroid phenotype
J. Chhablani (*) may be identified in eyes even with normal or
Department of Ophthalmology, University thinner choroid [5]. The increase in the volume of
of Pittsburgh Medical Center, Pittsburgh, PA, USA

118 A. Ayachit and J. Chhablani
the outer choroid is also offset by a reduction of PNV, nAMD, and controls. Responses to intravit-
the choriocapillaris and inner choroid volume real injection of aflibercept were also analyzed in
causing a given eye to have less than usual cho- the PNV and nAMD groups. In the PNV group,
roid thickness in pachychoroid phenotype [6]. VEGF-A was significantly lower than in the
Morphology of the choroid is thus more valuable nAMD group (p = 0.03) but was similar to that in
than quantitative measurements alone. Whether the control group (p = 0.86). The nAMD group
the dilation of the Haller vessels (pachyvessels) showed positive correlations between interleukin
occurs first or is secondary to the inner choroidal (IL-6 and IL-8 (p < 0.001), IL-6 and monocyte
attenuation is still unknown. It has been hypoth- chemoattractant protein (MCP)-1 (p = 0.002) and
esized that the inner choroidal circulation suffers IL-8 and MCP-1 (p = 0.002). There was a propor-
ischemic and inflammatory insults due to loss of tionate decrease in VEGF and PIGF after 1 month
volume and may result in hypoxic venous dilata- of aflibercept injection in dry maculae compared
tion seen as pachyvessels. to wet maculae in the nAMD group. In PNV eyes
Understanding of pachychoroid has spawned however, there were no differences in the
great interest in its clinical implications. VEGF-A/PIGF levels between dry and wet mac-
Choroidal pathology causes RPE/Bruch’s mem- ulae suggesting that other cytokines may be driv-
brane irregularities, elevation, and thinning. The ing the neovascular activity in PNV [10].
effacement and effective loss of the choriocapil- Major AMD susceptibility genes ARMS2 and
laris due to expansion of Haller vessel volume CFH4–7 has led to the understanding that AMD
induces damage to the RPE/Bruch’s membrane is a genetically homogenous disease. However,
complex. If the RPE can overcome the chronic the differences in the phenotype of nAMD
fluid overload it is known as uncomplicated between Asians and Caucasians have led to spec-
pachychoroid. ulation about the possible differences in geno-
If the RPE however fails to overcome the cho- type. Considering that PCV/polypoidal variant of
roidal congestion, features of pachychoroid pig- nAMD is commoner in Asians, and soft drusen
ment epitheliopathy (PPE) like RPE changes and are less prevalent in Asians, susceptibilities for
outer retinal thinning occur, that signifies pro- PCV and in turn AMD maybe differently driven
gressive dysfunction of RPE. With continuing [11]. Miyake et al. studied genetic susceptibili-
RPE damage, breakdown of the RPE barrier and ties in Japanese patients diagnosed with pachy-
the accompanying choriocapillaris loss may lead choroid neovasculopathy and neovascular
to the development of sub-RPE and subretinal AMD. Pachychoroid neovasculopathy was
fluid, i.e., CSC. New vessels may develop in observed in 39 individuals (19.5%). Their genetic
response to the ischemia causing pachychoroid susceptibility to AMD genes was significantly
neovasculopathy (PNV). Eyes with PPE lesions lower than that of neovascular AMD; ARMS2
have been noted to have a higher risk of develop- (p = 0.029), CFH (p = 0.013), and genetic risk
ing PNV [7, 8]. score calculated from 11 AMD susceptibility
Finally, the ends of the neovessels develop genes (p = 0.0038). These results showed that
aneurysms and convert to polypoidal choroidal patients with pachychoroid neovasculopathy
vasculopathy (PCV). Thus, in eyes with PCV were genetically less susceptible to AMD [12].
precursor lesions like PPE, CSC and even PNV Lehmann and associates have hypothesized
may be present. Angiogenic factors that are over- that pachychoroid characteristics may have an
expressed in these eyes due to loss of choriocap- autosomal dominant mode of inheritance [13].
illaris may cause the development of PNV and
subsequent PCV. Chronic inflammation involv-
ing choriocapillaris may also produce angiogen- 10.2.1 Clinical Features
esis [9].
Terao et al. analyzed and compared the cyto- Patients with PNV are younger than those with
kine profiles in the aqueous humor of eyes with CNVM due to wet AMD. Miyake and colleagues
10 Pachychoroid-Related Choroidal Neovascularization 119
in their study of 200 patients with CNVM showed

that patients with PNV were significantly younger
than those with wet AMD (68.7 years vs.
75.6 years) [12]. Patients with PNV may present
for the first time with symptoms of decreased
vision, metamorphopsia, and/or central scotoma.
However, more often history of CSC, i.e.,
repeated episodes of central scotomas, may be
elicited. Patients on routine follow-up for chronic
CSC may also have rebound fluid due to a sub-
clinical neovascular membrane. Rarely, photop-
sias have been described in patients with
extramacular PNV [14]. The hitherto described
entity “CSC-associated CNVM” is increasingly
being recognized as in fact PNV because of the
shared features of choroidal hyperpermeability
and congestion. Further, CSC-associated CNVM
maybe declared an incorrect label because the
exact mechanism of how CSC by itself causes Fig. 10.1 AF image of a case of PNV showing hyperAF
CNVM has not been proven. descending tracts owing to chronic CSC (green arrow).
Acute onset vision loss may be noted if there PNV was seen at the area showing hypoAF superior to the
is a hemorrhage or new-onset fluid in a patient fovea (yellow arrow)
without history of CSC who has now developed
PNV. A quiescent PNV may also be detected 10.2.4 OCT
incidentally in the fellow eyes of patients with
established PNV/PCV. Quiescent PNVs are 10.2.4.1 Cross-sectional OCT
detected incidentally on OCT angiography or by As mentioned earlier, increased choroidal thick-
FFA/ICGA. ness with inner choroidal attenuation is the hall-
mark of pachychoroid on OCT. Querques et al. in
their study on quiescent PNV found subfoveal
10.2.2 Clinical Examination choroidal thickness was 399 ± 72 microns in their
series [15]. PNVs may also correspond spatially
Eyes with PNV exhibit features of pachychoroid to areas of focal pachychoroid. Extramacular
such as reduced tessellation of fundus giving an PNV may be due to extramacular foci of pachy-
orange appearance. RPE abnormalities are com- choroid. Dansingani et al. in their study evaluat-
monly seen. In patients with chronic CSC, ing enface features of pachychoroid disorders
descending tracts of hyperpigmentation may be found choroidal thickness of >300 microns sub-
seen. PNV usually presents as a small greenish foveally. In the eyes with normal subfoveal thick-
membrane with subretinal hemorrhage and sub- ness, there was an extrafoveal focus of maximal
retinal fluid. choroidal thickness exceeding 300 μm [16]. In
these cases, choroidal thickness mapping
revealed that the extrafoveal focus of maximal
10.2.3 Autofluorescence choroidal thickness exceeded the choroidal thick-
ness subfoveally by at least 50 μm. The impor-
The descending tracts of CSC may be more tance of noting the area of maximal choroidal
apparent. Tracts are generally hyperautofluores- thickness was to correlate the disease foci with
cent with or without hypoautofluorescent borders the distribution of the thickened choroidal vessels
denoting chronicity (Fig. 10.1). [16] (Fig. 10.2).
Fig. 10.2 AF + OCT

image of a case of
extrafoveal PNV. At the
fovea (top), although the
choroidal thickness is
increased; the
extrafoveal inferior
focus with the flat
irregular PED (DLS) has
pachyvessels with
effacement of
choriocapillaris (bottom)
OCT may reveal areas of PPE, which is con- whereas DLS was seen in only 11 of the 35 eyes
sidered a forme-fruste of CSC. Pachydrusen may of CSC (31.43%). DLS was significantly associ-
be apparent. PNV, being a type 1 CNVM, is usu- ated with PCV and PNV as compared to
ally noted as a flat irregular PED. This is referred CSC. Further in eyes with PNV and PCV, the
to as a double layer sign (DLS) wherein there is a DLS had a characteristic moderate hyperreflec-
membrane insinuated between RPE and Bruch’s tivity in contrast to CSC eyes which had a hypo-
membrane [17]. There may also be fluid accumu- reflective DLS (P < 0.001). DLS height and width
lation between RPE and Bruch’s membrane were significantly more in the PCV and PNV
resulting from leakage from the network of groups as compared with the CSC group. They
abnormal vessels. Sheth et al. comprehensively concluded that the changes in the dimensions of
evaluated EDI OCT images of patients with DLS from one stage of pachychoroid spectrum
pachychoroid variant to discern various tomo- (i.e., CSC) to the terminal stages (i.e., PNV and
graphic features of DLS to better characterize the finally PCV) represents a progressive worsening
pachychoroid disease spectrum. Of the 102 eyes of RPE–Bruch’s membrane complex impair-
of 79 patients analyzed, they found that none of ment. The DLS site also corresponded to the mid-
the PPE eyes had DLS. Fifteen eyes with PNV phase choroidal hyperfluorescence, dilated
had DLS (93.75%). In total, 32 of the 35 PCV choroidal veins, and abnormal network of vessels
eyes (91.43%) showed the presence of DLS, on ICGA [18] (Fig. 10.3).
Fig. 10.3 ICGA + OCT (top)—hyperreflective DLS cor- Simultaneous FFA and ICGA shows late leakage and an
responding to a network seen on ICGA. The OCT also abnormal network, respectively
shows intraretinal edema with subretinal fluid.
SRF in PNV maybe clear. However, if associ- rial may be present. In all, the presence of a
ated with CSC, the SRF may be associated with hyperreflective DLS with SRF, SHRM is indica-
fibrin and a vacuole sign indicating the active tive of an active PNV (Fig. 10.4).
leakage point of the CSC (Fig. 10.4). In this situ- Frequently in patients on follow-up with
ation it is indeed difficult to attribute the fluid to chronic CSC with dry macula, there may be a
either CSC or PNV alone. IRF and spongy edema hyperreflective DLS with no exudation (SRF/
are also frequently seen in the setting of IRF). These are termed quiescent CNVs and
PNV. Subretinal hemorrhage is seen as hyperre- do not meet the criteria for treatment. These
flectivity with back-shadowing. Subretinal hyper- may require frequent follow-up and an OCTA
reflective material (SHRM) may be indicative of may be warranted (refer to the section on
a PNV. Rarely intraretinal hyperreflective mate- OCTA).
ICGA as well. The property of inward displace-

ment of pachyvessels can be further confirmed by
the corresponding cross-sectional OCT in which
the choroidal vessels occupy the entire choroidal
thickness.
The PNV network can be identified noninva-
sively using OCTA. PNV networks have been
found to be corresponding to the hyperreflective
DLS on OCT. Dansingiani et al. reported that in
eyes with pachychoroid features, a hyperreflec-
tive DLS on OCT has more diagnostic value
than dye-based angiography. Given the diffi-
culty in detecting latent type 1 neovasculariza-
tion in pachychoroid eyes using conventional
retinal imaging and the fact that untreated PNVs
can lead to hemorrhage, exudation, atrophy, and
fibrosis; OCTA can be valuable in detecting a
subclinical/ quiescent PNV. The frequency with
which quiescent neovascularization evolves to
clinical significance is unknown, but its pres-
Fig. 10.4 Case of CSC with PNV showing subretinal
membrane with subretinal hyperreflective material (top). ence may indicate a closer follow-up of the
In lower sections, line scans showing SRF with fibrin (yel- patient [19]. In conclusion, OCT angiography
low arrow) (bottom) detection rate for type 1 neovascularization
under a hyperreflective DLS or a flat irregular
10.2.4.2 OCT Angiography PED (FIPED) was found to be 95% compared to
Freund et al. studied the enface features of eyes the dye angiography detection rate of 29% [19]
with pachychoroid. In 66 eyes of 33 patients (Fig. 10.5).
including uncomplicated pachychoroid, PPE, Querques et al. studied six eyes with quiescent
CSC, PNV, and PCV, there was focal choriocapil- PNV on OCTA. The diagnosis of treatment-naïve
laris atrophy with deep choroidal vessels occupy- “quiescent” CNV was defined as a flat irregular
ing the entire choroidal thickness. Enface PED (FIPED) with moderate reflectivity with no
SS-OCT in all eyes demonstrated dilated outer intraretinal or subretinal fluid on OCT, late-phase
choroidal vessels in the temporal quadrants, ill-defined hyperfluorescent lesion, leakage on
extending into the macular area. The dilation fluorescein angiography, and hyperfluorescent
seen on the enface correlated spatially with the network in the early and mid-phases of ICGA
areas of maximal choroidal thickness. These along with the presence of pachychoroid charac-
were referred to as “pachyvessels.” Diffuse teristics. The most frequently observed features
pachyvessels were seen in only 5 eyes, whereas on OCTA were irregular shape and well-defined
28 eyes had focally distributed pachyvessels. The margin. In addition, 100% of the vascular net-
appearance of the choroidal vessels on the enface works were seen on OCTA [15] (Fig. 10.6).
represented their inward displacement as a con- Querques et al. have noted in a study that qui-
sequence of choriocapillaris thinning [16]. escent CNVs enlarge over time with no
Freund et al. further stated that these dilated consequent activity seen on OCT [20]. It remains
deep choroidal vessels retain their caliber as they to be seen if the same is true of PNV networks. It
traverse the macula. Unlike normal choroidal is easy to monitor PNV networks on OCTA non-
vessels, pachyvessels do not taper, and terminate invasively, including monitoring the area of CNV
at their posterior extent, and appear to terminate over time. In quiescent CNV secondary to AMD,
abruptly. This same property is appreciated on Querques and group suggested that the absence
Fig. 10.5 OCTA in a case of PNV showing a well-defined network in the slab corresponding to the DLS (orange
arrow)
Fig. 10.6 A large quiescent CNV corresponding to the DLS in a case of PNV. There is no associated fluid
of a detectable core vessel may represent a pro- 10.2.4.3 FFA

tective factor against increased activity from qui- Differentiating chronic CSC from AMD can be
escent CNV [21]. Thus, as an extrapolation, the challenging as the two conditions may have very
higher prevalence of a detectable core vessel in similar features on conventional angiography,
quiescent pachychoroid neovasculopathy may be characterized by RPE atrophy and diffuse leak-
associated with a higher tendency to activation in age owing to diffuse retinal pigment epitheliopa-
pachychoroid neovasculopathy than AMD. thy. Importantly, the areas of type 1
New fluid in a case of chronic CSC can be neovascularization are correlated spatially to
either a recurrence of CSC or due to the develop- areas displaying pachychoroid features. This
ment of a PNV network. In cases of refractory OCT correlation is possible during simultaneous
CSCs not responding to conventional treatments, FFA + OCT [5]. Early stippled fluorescence with
a subclinical PNV found on OCTA may be the late leakage is the feature of PNV, being a sub-
cause of recurrent fluid. Identification of vascular RPE membrane (Fig. 10.7). It may sometimes be
networks assumes paramount importance in such difficult to differentiate between the diffuse stip-
cases. pling of hyperfluorescence of chronic CSC from
a b
c d
Fig. 10.7 The same eye as Fig. 10.1. (a) AF + OCT showing SRF at macula. (b) DLS corresponding to hypoAF. (c)
late leakage on FFA and abnormal vascular network on ICGA (d)
the stippling of a PNV. The PNV would demon- have late washout. This makes diagnosis chal-
strate late leakage in addition to the background lenging. Therefore, it is emphasized that OCTA
of mottled hyperfluorescence due to diffuse reti- is a better tool for diagnosing quiescent
nal pigment epitheliopathy of chronic CSC. In networks.
CSC co-existing with PNV, there may be dot-
block leaks/smokestack leaks suggesting acute
on chronic CSC. 10.2.5 Treatment
10.2.4.4 ICGA 10.2.5.1 Anti-VEGF therapy

Flash-based ICGA shows an early hypofluores- Jung et al., in a retrospective study, compared
cence followed by a late hot spot/plaque. With the efficacy of intravitreal injection of ranibi-
the advent of scanning laser ophthalmoscopy zumab and aflibercept for patients with
(SLO) imaging, networks of PNV are more obvi- PNV. Eyes were initially treated with 3 monthly
ous. A clear well-defined abnormal vascular net- loading injections of ranibizumab or afliber-
work is apparent in the late phases of the cept. They found that at 3 months, the rate of
angiogram. In addition, large caliber choroidal complete fluid absorption was higher in the
vessels are seen traversing the fundus. These do aflibercept-treated group than in the ranibi-
not taper toward the macula. Areas of choroidal zumab-treated group (p = 0.018). The mean
hyperpermeability classical of pachychoroid are reduction of subfoveal choroidal thickness was
also seen on ICGA. These findings may be pres- also greater in the aflibercept group than in the
ent in the apparently normal fellow eye as well ranibizumab group (p = 0.013). However, there
(Fig. 10.8). Areas of hyperpermeability are char- was no difference in visual acuity or central
acteristically seen in areas where there is intact macular thickness. Also, complete fluid absorp-
choriocapillaris. This is also a feature seen on tion was noted after switching from ranibi-
simultaneous ICGA + OCT [5]. This holds zumab to aflibercept in 13 of 15 eyes (86.7%).
importance when planning photodynamic ther- Adjunctive photodynamic therapy was required
apy for PNV/chronic CSC associated with in six eyes. They concluded that anti-VEGF
PNV. Furthermore, unlike quiescent CNV in treatments helped improve visual acuity at
AMD characterized by late hyperfluorescence, in 12 months. Aflibercept was better than ranibi-
PNV eyes only the early-mid phases of ICGA zumab in causing fluid resorption and reducing
show the quiescent neovascular network as these choroidal thickness at 3 months [22].
Fig. 10.8 Other eye of the patient depicted in Fig. 10.3. choroidal hyperpermeability in late phases. OCT line
Early frames of ICGA show large caliber vessels in the scans through these areas show enlarged choroidal
midperiphery (yellow arrows) corresponding to areas of vessels
Hata et al. reported lower intraocular VEGF lev- PNV are distinct and help in the distinction from
els in pachychoroid neovasculopathy than in AMD-related and other secondary CNVMs. PNV
nAMD [23]. Despite this, the authors noted favor- is characterized by CNVM in younger patients
able response to anti-VEGF therapy in eyes with along with features of thick choroid and chorio-
PNV Almost 90% of patients responded with ini- capillaris attenuation on imaging. Quiescent
tial anti-VEGF therapy and only 11.1% needed PNV is to be monitored closely for the develop-
adjunctive PDT therapy [24]. To minimize the ment of activity. There is a lower dependence of
deteriorating effects of repeated PDT, applying PNV on VEGF levels, thus requiring a smaller
PDT as a deferred or rescue option has been sug- number of injections and having longer
gested in PCV [24]. They suggested PNV may also retreatment-free intervals. PDT is a viable option
need PDT only as a rescue option like PCV [25]. as a rescue treatment for PNV unresponsive to
Azuma et al. investigated the effect of anti- anti-VEGF therapy.
VEGF therapy on choroidal structure of PNV and
other types of neovascular age-related macular Key Points
degeneration (non-PNV). Twenty-one eyes with • PNV is commoner in Asian populations than
PNV and 34 eyes with non-PNV who underwent Caucasians. Patients tend to be younger than
anti-VEGF treatment were reviewed. CNV area at AMD age group and have a distinct genetic
baseline was measured with FFA. The luminal and susceptibility.
stromal areas in the choroid were measured at base- • PNV has lower dependence on VEGF levels
line and 1 month. The association between dry compared to nAMD.
macula or visual acuity at 1 month and baseline val- • Clinically, pigmentary alterations, reduced tes-
ues or changes in the luminal or stromal area at sellation of the fundus, and features of CNVM
1 month, baseline CNV area, were analyzed in like fluid and hemorrhage may be seen.
patients with or without PNV. In non-PNV, change • EDI-OCT shows increased choroidal thick-
of luminal area (p = 0.0001), baseline CNV area ness (predominantly Haller layer expansion)
(p = 0.033), and aflibercept versus ranibizumab with concomitant choriocapillaris attenuation.
(p = 0.0048) were shown to be predictors for dry PNV corresponds spatially to the thickened
macula [26]. choroid and has a double layer sign.
A retrospective study of 46 eyes with PNV • FFA shows stippled fluorescence and may
treated with intravitreal anti-VEGF therapy demon- show diffuse changes of chronic CSC. AF
strated improvement in visual acuity after a mean may also show descending tracts.
follow-up of 38.3 months [27]. In the MINERVA • ICGA shows pachyvessels, areas of choroidal
study, eyes with PNV treated with ranibizumab permeability along with hot spot/plaque of the
experienced a higher letter gain compared to the PNV. PNV shows late washout phenomenon
sham group at 2 months [28]. It has also been in contrast to nAMD.
observed that PNV eyes have a longer retreatment- • OCTA is a noninvasive modality that can
free interval compared to AMD eyes following load- detect active and quiescent PNV with high
ing doses. As mentioned earlier, this is due to lower sensitivity and specificity.
dependence on VEGF levels in PNV eyes [12]. • PNV has shown favorable response to anti-
VEGF treatments and has a longer treatment-
free interval. PDT may be used as a rescue
10.3 Conclusion option in cases refractory to anti-VEGF
therapy.
PNV is a distinct entity that is to be differentiated
from other causes of CNVM. SD-OCT (EDI), Conflict of Interest None of the authors have any con-
SS-OCT, FFA, ICGA, and OCTA findings of flict of interest.
References 13. Lehmann M, Bousquet E, Beydoun T, Behar-Cohen

F. PACHYCHOROID: an inherited condition? Retina.
2015;35:10–6.
1. Imamura Y, Fujiwara T, Margolis R, Spaide
14. Kiss S, Rusu I, Seidman C, Orlin A, D’amico D, Gupta
RF. Enhanced depth imaging optical coherence
M. Pachychoroid neovasculopathy in extramacu-
tomography of the choroid in central serous chorio-
lar choroidal neovascularization. Clin Ophthalmol.
retinopathy. Retina. 2009;29:1469–73.
2016;10:1275–82.
2. Fujiwara T, Imamura Y, Margolis R, Slakter JS,
15. Carnevali A, Capuano V, Sacconi R, Querques L,
Spaide RF. Enhanced depth imaging optical coher-
Marchese A, Rabiolo A, et al. OCT angiography
ence tomography of the choroid in highly myopic
of treatment-Naïve quiescent Choroidal neovas-
eyes. Am J Ophthalmol. 2009;148:445–50.
cularization in Pachychoroid Neovasculopathy.
3. Koizumi H, Yamagishi T, Yamazaki T, Kinoshita
Ophthalmology Retina. 2017;1(4):328–32.
S. Relationship between clinical characteristics
16. Dansingani K, Balaratnasingam C, Naysan J, Freund
of polypoidal choroidal vasculopathy and choroi-
K. En face imaging of pachychoroid spectrum disor-
dal vascular hyperpermeability. Am J Ophthalmol.
ders with swept-source optical coherence tomogra-
2013;155:305–13. e301
phy. Retina. 2016;36(3):499–516.
4. Gupta P, Ting DSW, Thakku SG, et al. Detailed
17. Sato T, Kishi S, Watanabe G, et al. Tomographic fea-
characterization of choroidal morphologic and vas-
tures of branching vascular networks in polypoidal
cular features in age-related macular degeneration
choroidal vasculopathy. Retina. 2007;27:589–94.
and polypoidal choroidal vasculopathy. Retina.
18. Sheth J, Anantharaman G, Chandra S, Sivaprasad
2017;37:2269–80.
S. “Double-layer sign” on spectral domain optical
5. Dansingani KK, Balaratnasingam C, Naysan J,
coherence tomography in pachychoroid spectrum dis-
Freund KB. Enface imaging of pachychoroid spec-
ease. Indian J Ophthalmol. 2018;66(12):1796.
trum disorders with swept source optical coherence
19. Dansingani K, Balaratnasingam C, Klufas M, Sarraf
tomography. Retina. 2016;36:499–516.
D, Freund K. Optical coherence tomography angi-
6. Balaratnasingam C, Lee WK, Koizumi H, Dansingani
ography of shallow irregular pigment epithelial
K, Inoue M, Freund KB. Polypoidal choroidal vascu-
detachments in pachychoroid spectrum disease. Am J
lopathy: a distinct disease or manifestation of many?
Ophthalmol. 2015;160(6):1243–1254.e2.
Retina. 2016;36:1–8.
20. Querques G, Srour M, Massamba N, et al. Functional
7. Roisman L, Zhang Q, Wang RK, et al. Optical
characterization and multimodal imaging of treatment-
coherence tomography angiography of asymptom-
naïve “quiescent” choroidal neovascularization. Invest
atic neovascularization in intermediate age-related
Ophthalmol Vis Sci. 2013;54:6886e6892.
macular degeneration. Ophthalmology. 2016;123:
21. Carnevali A, Cicinelli MV, Capuano V, et al.
1309–19.
Optical coherence tomography angiography: a use-
8. Yanagi Y, Mohla A, Lee WK, et al. Prevalence
ful tool for diagnosis of treatment-naïve quiescent
and risk factors for nonexudative neovascular-
choroidal neovascularization. Am J Ophthalmol.
ization in fellow eyes of patients with unilateral
2016;169:189e198.
age-related macular degeneration and polypoidal
22. Jung B, Kim J, Lee J, Baek J, Lee K, Lee W. Intravitreal
choroidal vasculopathy. Invest Ophthalmol Vis Sci.
aflibercept and ranibizumab for pachychoroid neovas-
2017;58:3488–95.
culopathy. Sci Rep. 2019;9:1–7.
9. Peiretti E, Ferrara DC, Caminiti G, Mura M, Hughes
23. Hata M, et al. Intraocular vascular endothelial growth
J. Choroidal neovascularization in caucasian patients
factor levels in pachychoroid neovasculopathy and
with longstanding central serous chorioretinopathy.
neovascular age-related macular degeneration. Invest
Retina. 2015;35:1360–7.
Ophthalmol Vis Sci. 2017;58:292–8.
10. Terao N, Koizumi H, Kojima K, Yamagishi T,
24. Lee JH, Lee WK. One-year results of adjunctive
Yamamoto Y, Yoshii K, et al. Distinct aqueous
photodynamic therapy for type 1 neovasculariza-
humour cytokine profiles of patients with pachycho-
tion associated with thickened choroid. Retina.
roid neovasculopathy and neovascular age-related
2016;36:889–95.
macular degeneration. Sci Rep. 2018;8(1):1–10.
25. Gomi F, et al. Initial versus delayed photodynamic
11. Coscas G, et al. Comparison of exudative age-
therapy in combination with ranibizumab for treat-
related macular degeneration subtypes in Japanese
ment of polypoidal choroidal vasculopathy: the
and French Patients: multicenter diagnosis
fujisan study. Retina. 2015;35:1569–76.
with multimodal imaging. Am J Ophthalmol.
26. Azuma K, Tan X, Asano S, Shimizu K, Ogawa A,
2014;158:309–318e2.
Inoue T, et al. The association of choroidal structure
12. Miyake M, Ooto S, Yamashiro K, Takahashi A,
and its response to anti-VEGF treatment with the
Yoshikawa M, Akagi-Kurashige Y, et al. Pachychoroid
short-time outcome in pachychoroid neovasculopa-
neovasculopathy and age-related macular degenera-
thy. PLoS One. 2019;14(2):e0212055.
tion. Sci Rep. 2015;5:1–11.
27. Chhablani J, Kozak I, Pichi F, et al. Outcomes of 28. Lai TYY, Staurenghi G, Lanzetta P, et al. Efficacy and
treatment of choroidal neovascularization asso- safety of ranibizumab for the treatment of choroidal
ciated with central serous chorioretinopathy neovascularization due to uncommon cause: twelve-
with intravitreal antiangiogenic agents. Retina. month results of the MINERVA study. Retina. 2017
2015;35:2489–97. Jul 12;38(8):1464–77.
Dr. Apoorva Ayachit

Qualifications
Dr. Jay Chhablani is a Vitreo-Retina Specialist at the
1. MS Ophthalmology (Delhi University—2015) University of Pittsburgh Eye Center. He completed a clini-
2. Fellow of International Council of Ophthalmology cal vitreo-retina fellowship at Sankara Nethralaya,
(FICO), ICO—London 2016 Chennai, India, and was an International Council of
3. Diplomate of National Board—2016 Ophthalmology (ICO) fellow at Jules Gonin Eye Hospital,
4. Fellowship in Vitreoretinal Surgery (FVRS—RGUHS Switzerland, in 2009. He was a Clinical Instructor at the
affiliated)—M M Joshi Eye Institute, Hubballi Jacobs Retina Center at Shiley Eye Center, University of
5. FAICO—vitreoretina 2018—Awarded at AIOC California, San Diego, USA (2010–2012) before joining
Coimbatore the faculty at L V Prasad Eye Institute, Hyderabad, India
(2012–2019). His areas of interest are macular disorders
and recent imaging techniques. He has published more
Current post Consultant, Department of Vitreoretina, than 300 articles in peer-reviewed journals with a focus on
M M Joshi Eye Institute the field of the choroid. He is the editor of the books
She has multiple publications, book chapters to her Choroidal Disorders and Central Serous
credit. Special interest in macular imaging. Chorioretinopathy, and he is on the review boards of all
the high-impact ophthalmology journals. He is also on the
editorial board of several specialty journals, including the
American Journal of Ophthalmology. He is a member of
the American Academy of Ophthalmology’s Global ONE
network committee. He has won several national and
international awards and delivered the inaugural Ian
Constable lecture at the Asia-Pacific Vitreo-Retina Society
in 2016. He received the Inaugural Namperumalsamy
Young Researcher Award in 2018, presented by the
Vitreo-Retina Society of India.
Inflammatory Choroidal
Alvaro Olate-Perez, Carolina Bernal-Morales,
Aina Moll-Udina, Alfredo Adan,
and Javier Zarranz-Ventura
11.1 Introduction is yet to be elucidated, and there are mainly two

mechanisms that may concur, an inflammatory
Inflammatory choroidal neovascularization mediated angiogenic stimulus and the abovemen-
(CNV) membranes are generated in the cho- tioned frequent disruption in the RPE–Bruch’s
riocapillaris in the context of local inflamma- membrane complex seen in these clinical condi-
tion and frequently sprout to the retina through tions, among other complex processes [1, 2].
defects in the retinal pigment epithelium (RPE)– The diagnostic tools and treatment options for
Bruch’s membrane complex, that are commonly inflammatory CNV have experienced significant
seen in infectious and noninfectious uveitis [1]. advances in the last decade. As in other fields, the
After age-related macular degeneration (AMD) development of new noninvasive imaging tech-
and myopia, uveitis syndromes appear as the niques such as optical coherence tomography
third leading cause for CNVs and are frequently angiography (OCTA) beyond the structural opti-
present in specific clinical conditions such as cal coherence tomography (OCT) scans has com-
punctate inner choroidopathy (PIC), multifocal plemented the classic fluorescein angiography
choroiditis (MFC), multiple evanescent white (FA) and autofluorescence examination tech-
dot syndromes (MEWDS), serpiginous cho- niques and has provided new insights to diagnose
roiditis (SC), presumed ocular histoplasmosis and monitor CNV progression. Currently, the
(POHS), toxoplasma retinochoroiditis, and Vogt– first-line therapy for inflammatory CNVs is intra-
Koyanagi–Harada’s disease (VKH) [2]. The vitreal anti-VEGF drugs, although beneficial out-
exact pathophysiology of inflammatory CNVs comes have also been reported for other treatment
options such as corticosteroids and immunosup-
pressive agents [3].
A. Olate-Perez · C. Bernal-Morales
Institut Clínic de Oftalmología, Hospital Clínic,
Barcelona, Spain 11.2 Epidemiology
A. Moll-Udina · A. Adan · J. Zarranz-Ventura (*)
Institut Clínic de Oftalmología, Hospital Clínic, Inflammatory CNVs represent a severe compli-
Barcelona, Spain cation which may cause irreversible central
Institut de Investigacions Biomediques August Pi i visual loss. Uveitis syndromes are more frequent
Sunyer (IDIBAPS), Barcelona, Spain in young people, often in active working-age, and

130 A. Olate-Perez et al.
Table 11.1 Prevalence and incidence of main infectious and noninfectious uveitis
Infectious uveitis Noninfectious uveitis
Prevalence Prevalence Incidence
POHS Frequent Vogt–Koyanagi–Harada 0.3% [7]; 6.4% [7]
9–15% cases [2]
Toxoplasmosis 0.3–19% [2] Multifocal choroiditis 4.6% [7]; 3.6% [7]
33–50% cases [2]
West Nile virus chorioretinitis Few cases Punctate inner choroidopathy 11.8% [7] 13.2% [7]
reported 76–100% cases [2]
Intraocular Tuberculosis Uncommon Serpiginous choroiditis 4.7% [7] 2.7% [7]
10–25% cases [2]
for this reason early detection and timely treat- diagnosis, being rarely found in anterior and
ment of CNV is essential to reduce the risk of intermediate uveitis patients. Moreover, in this
permanent irreversible visual loss which could be same series posterior uveitis presented a higher
invalidating for their daily activities [3, 4]. prevalence of CNV (2.7%) compared to panuve-
With regards to the etiology, two groups of itis (0.8%) [6]. Another study reported a cumula-
inflammatory CNV are described, those second- tive incidence of inflammatory CNV in
ary to noninfectious or infectious uveitis. The noninfectious uveitis of 2.7% at 2 years, present-
main causes of infectious uveitis-associated ing the majority of the cases within the first 6
CNVs are toxoplasma and presumed ocular his- months after diagnosis [7]. In this series, no sig-
toplasmosis syndrome (POHS). Toxoplasma is a nificant difference was found in terms of inci-
relatively common type of uveitis and the only dence between panuveitis and posterior uveitis,
protozoan affecting the eye which has been asso- but VKH and PIC were associated with increased
ciated with inflammatory CNV. POHS is pro- risk of developing CNV than other noninfectious
duced by histoplasma capsulatum, a species of uveitis syndromes [7]. Prevalence and incidence
dimorphic fungi, which occurs mainly in central of CNV at presentation are summarized in
and eastern United States CNV [4]. The preva- Table 11.1. Finally, the main risk factors for the
lence of these two features varies according to development of inflammatory CNV in uveitis
geographical area. According to previously pub- syndromes include active inflammation status,
lished data, toxoplasmosis presents higher rates concomitant retinal neovascularization, and pre-
in South Asia and POHS in North America [2]. vious CNV diagnosis in the fellow eye [3, 4].
Other infectious uveitis types are rarely associ-
ated with CNV development and only single case
reports and small case series have been reported. 11.3 Diagnosis
This is the case of inflammatory CNV secondary
to Tuberculosis, Toxocara, West Nile virus, 11.3.1 Signs and Symptoms
Congenital Rubella Retinopathy, Candida albi-
cans, Cryptococcus neoformans, and Aspergillus Symptoms may be different in each specific case
fumigatus [5]. Interestingly, some cases have also and may ultimately depend on the location of the
been reported secondary to endophthalmitis [2]. inflammatory CNV complex. According to this,
The prevalence and incidence of inflammatory subfoveal CNVs are usually more symptomatic,
CNV secondary to noninfectious uveitis is higher, causing visual loss, metamorphopsia or central
according to more robust epidemiological data. scotoma, and extrafoveal CNV are less symp-
The main conditions associated to inflammatory tomatic or even asymptomatic, and may often be
CNVs are PIC, MFC, MEWDS, SC, and VKH detected incidentally by direct examination or
disease. Baxter et al. reported in a multicenter ancillary tests such as OCT or OCTA in routine
study of 15,137 noninfectious uveitis eyes that clinical exams.
2% of posterior uveitis or panuveitis cases had a On clinical examination concomitant signs of
history of CNV (active or past) at the moment of posterior uveitis or panuveitis are often seen, and
11 Inflammatory Choroidal Neovascularization 131
a b c e
d f
g h i k
j l
Fig. 11.1 A 29-year-old male with LE choroidal neovas- shows a choroidal neovascularization of glomerular pat-
cularization adjacent to an old toxoplasma macular scar. tern in the avascular plexus. One month after the second
Fundus picture (a) shows a juxtafoveal inferotemporal injection of Aflibercept (Eylea®) the macular hemorrhage
subretinal hemorrhage, represented by a hypofluorescent disappeared (g), FAF showed a mild decrease of the hypo-
area in the fundus autofluorescence (FAF) image (b) sur- autofluorescent halo (h), structural OCT revealed com-
rounded by a hyperfluorescent halo. A focal area of retinal plete resolution of subretinal fluid and a residual solid
thickening is seen in the thickness map of the structural pigment epithelium detachment lesion compatible with
OCT scan (c), corresponding to a dense hyperreflective the old scar (i, j), and OCTA (k, l) showed complete
material lesion with associated subretinal fluid detected in regression of the glomerular-shape CNV in the avascular
the cross-sectional image of the b-scan (d). OCTA (e, f) plexus
more rarely intermediate uveitis. In funduscopy, • Multifocal choroiditis (MC): Multiple

findings suggestive of CNV include green-gray yellow-white punched-out lesions are located
subretinal discoloration areas, intra- and subreti- peripapillary and in the mid-periphery with
nal fluid, hemorrhages, and scars. minimal vitritis. CNV is usually associated
The main clinical characteristics are described with active inflammatory retinal lesions,
below disclosed by etiologies [1–3]: which may be located in the foveal or extrafo-
veal macular regions.
• Presumed ocular histoplasmosis syndrome • Punctate inner choroidopathy (PIC):
(POHS): Three classic findings are described, Multiple, small, round, yellow-white lesions
which include peripapillary atrophy, focal are seen in the posterior pole with no signs of
oval-shaped chorioretinal lesions (histo spots), vitreous inflammation. CNV frequently is
and no vitreous inflammation. CNVs appear associated with the inflammatory lesions
usually in the peripapillary or subfoveal located in the subfoveal region (Fig. 11.2).
regions, frequently in the edges of preexisting • Serpiginous choroiditis (SC): Gray-white
chorioretinal scars. lesions with a geographical distribution affect-
• Ocular toxoplasmosis: It commonly appears ing the posterior pole. CNV lesions may
as an oval whitish focal retinochoroiditis area appear in the edges of chorioretinal atrophy
adjacent to atrophic scar lesions, with dense plaques distributed in the peripapillary, subfo-
overlying vitritis. CNV is located at the edge of veal, and/or extrafoveal regions (Fig. 11.3).
old lesions and may be associated to hemor- • Vogt–Koyanagi–Harada (VKH) disease:
rhages and intra- or subretinal fluid (Fig. 11.1). Multiple exudative retinal detachments are
a b c e
f g h j
Fig. 11.2 A 34-year-old female diagnosed with punctate sponding to two independent active CNVs. OCTA con-
inner choroidopathy and previous history of CNV treated firmed both CNV lesions (superotemporal and inferonasal
with two intravitreal injections of anti-VEFG 2 years ago. to fovea) in the avascular plexus level (d, e). Five months
She presented with a new relapse of her disease with new after an intravitreal implant of dexamethasone (Ozurdex®)
white dots in the posterior pole adjacent to old atrophic (f), autofluorescence showed hypoautofluorescence of the
scars (a). Fundus autofluorescence (b) showed a hyperau- posterior pole and isolated lesions (g), and regression of
tofluorescent diffuse pattern in the posterior pole with the CNV lesions with well-defined edges in the structural
hyperautofluorescent lesions corresponding to the white OCT (h) and OCTA (i, j) images. Of note, OCTA images
dots in mid periphery seen in funduscopy. Structural OCT showed a reduction in the lesion size and vessel density of
(c) showed two subretinal hyperreflective material lesions both neovascular complexes compared to baseline images
with poorly defined edges in two different areas, corre-
a b e g
f h
c d i k
j l
Fig. 11.3 A 76-year-old female with serpiginous choroi- a lacy wheel-shaped CNV was seen in both the structural
ditis associated to tuberculosis. Fundus retinographies (a, OCT (e, f) and OCTA (g, h). Three months after a single
b) showed chorioretinal atrophy lesions in both eyes. aflibercept (Eylea®) injection a reduction in the lesion
Fundus autofluorescence showed hypoautofluorescent thickness and size is seen in the structural OCT (i, j) and
lesion corresponding to the atrophy areas (c, d). In the LE, regression of the CNV is observed in the OCTA (k, l)
seen, often associated with granulomatous 11.3.5 Optical Coherence

anterior uveitis. CNV is usually extrafoveal and Tomography
associated with hemorrhages and exudation.
Optical coherence tomography (OCT) is essential
for the diagnosis and management of inflammatory
11.3.2 Fluorescein Angiography CNV lesions due to its noninvasive nature and high
reproducibility in sequential follow-up scans [10].
Fluorescein angiography (FA) has classically Inflammatory CNVs frequently appear in OCT as
been the principal technique employed for CNV hyperreflective lesions in the subretinal space,
diagnosis. Inflammatory CNV lesions typically between the neurosensory retina and the RPE fol-
show early hyperfluorescence and late leakage, lowing a type 2 CNV pattern [11]. In some occa-
as other types of CNV. In selected cases, it is sions, multiple hyperreflective projections
often difficult to differentiate between CNV and extending from the CNV may appear (“pitchfork
uveitic lesions and additional tests are required. sign”). This sign has been suggested as a distinctive
Active chorioretinal lesions commonly show an characteristic of inflammatory CNV by some
early hypofluorescent pattern with late leakage, authors [12]. OCT activity signs for inflammatory
whereas inactive lesions show early hypofluores- CNV include subretinal and/or intraretinal fluid, as
cence without leakage [8]. well as increased retinal thickness or blurry mar-
gins of the lesion edges. These parameters can also
be helpful in the follow-up of these lesions [10].
11.3.3 Fundus Autofluorescence
Fundus autofluorescence (FAF) may also be 11.3.6 Optical Coherence

helpful in differentiating active CNV from Tomography Angiography
inflammatory lesions. Active inflammatory
lesions often show a hyperautofluorescent pat- Optical coherence tomography angiography
tern, whereas inflammatory CNVs frequently (OCTA) is an evolution of structural OCT that
appear with a more patchy hypoautofluorescent allows the reconstruction of the retinal vascular
signal, especially if these are adjacent to chorio- plexuses by sequential image captures and image
retinal atrophy patches [2]. In these situations, processing algorithms, based on the detection of
FAF may also be useful to differentiate areas of blood movement without the need of dye injection
intact RPE from subretinal hemorrhages (hyper- [13]. Thanks to this ability to identify perfused and
and hypoautofluorescent, respectively). non-perfused structures, inflammatory CNV can
be diagnosed and differentiated from inflammatory
lesions as these do not present vascular flow [14].
11.3.4 Indocyanine green This is especially important and helpful to inform
angiography treatment decisions in such circumstances, in other
to decide whether intravitreal corticosteroids or
Indocyanine green (ICG) angiography is fre- anti-VEGF drugs should be administered in each
quently used to evaluate the status of the choroid, individual case. Moreover, in comparison with
due to the higher molecular weight and greater other multimodal image techniques, OCTA is able
binding ability to proteins than fluorescein. For to detect structural and functional changes [15].
these reasons, it is an exceptional tool to identify Recent publications have highlighted the role of
CNV but also to evaluate other choroidal findings OCTA in the diagnosis and follow-up of CNV sec-
as granulomas, choriocapillaris hypoperfusion, ondary to PIC [16], multifocal choroiditis [17], and
and choroiditis, being really helpful in the differ- serpiginous choroiditis [18], among other inflam-
ential diagnosis of such conditions in case of matory causes. Zahid et al. published the first case
diagnostic dilemmas [9]. of CNV diagnosed by OCTA in a multifocal choroi-
ditis patient, demonstrating that OCTA may be use- associated CNVs normally require repeated and
ful also during the follow-up of these lesions prolonged treatments with anti-VEGF therapies,
revealing decreases in flux and size after treatment whereas inflammatory CNV typically requires a
[19]. Nevertheless, the most frequent OCTA finding lower number of injections and shorter duration
related with inflammatory CNV is hypoperfusion of treatments, as described in several different case
the choriocapillaris adjacent to the neovascular series.
membrane (perilesional hypointense halo), which
diminishes with anti-VEGF treatment. However,
OCTA has some significant limitations. For exam- 11.5 Management
ple, the edges of an active CNV are often not well
defined in OCTA images, which may well be a con- 11.5.1 Anti-VEGF
sequence of real changes in the CNV vascular struc-
ture or an artifact produced by the hyperreflective It is known that a pro-inflammatory environment
material that often surrounds the neovascular lesion,in the retina (as occurs in uveitis) could generate
a frequent finding in mature CNVs. These features the release of VEGF, a mediator of angiogenesis
may be a significant limitation to obtain helpful [1, 21]. Despite this, clinical studies show contro-
information about CNV activity with OCTA. Finally, versial results. Paroli et al. reported higher VEGF
manual segmentation of OCTA images is very often levels in the aqueous humor of VKH patients
required to avoid artifact and segmentation errors, compared to healthy controls [22]; however, this
especially in cases with abnormal retinal anatomy could not be confirmed by Banerjee et al. in CNV
as in inflammatory CNVs associated to posterior secondary to PIC [23].
uveitis syndromes [15]. Several recent case series have reported suc-
cessful outcomes with anti-VEGF therapy in
inflammatory CNV lesions secondary to infec-
11.4 Comparing AMD-Associated tious and noninfectious uveitis. In the first group,
CNV and Inflammatory CNV Schadlu et al. demonstrated visual improvement
or stabilization in 85% of eyes treated with intra-
Most knowledge of inflammatory CNV is extrap- vitreal bevacizumab in cases with POHS [24].
olated from AMD studies. According to the OCT- Korol et al. demonstrated clinical improvement
based classification, AMD-associated CNV can in CNV secondary to toxoplasmosis with intra-
be classified as type 1 (subretinal pigment epithe- vitreal aflibercept [25]. Other cases successfully
lium), type 2 (subretinal), or type 3 (retinal angi- treated with anti-VEGF include ocular tuberculo-
omatous proliferation). Inflammatory CNV sis and toxocariasis [26]. With regards to nonin-
presents a higher percentage of type 2 mem- fectious CNV, cases successfully treated with
branes [2]. In both type 2 AMD-associated and intravitreal anti-VEGF in multifocal choroiditis
inflammatory CNVs the alteration occurs at the [27–33], PIC [26, 31, 34, 35], serpiginous choroi-
RPE level, leading to abnormal new vessels ditis [18, 33, 36, 37], and VKH [26, 28, 31] have
growth from the choriocapillaris toward the sub- been published. Table 11.2 summarizes the main
retinal space. However, RPE defects are usually studies of inflammatory CNV treated with anti-
focal in inflammatory CNV and diffuse in AMD- angiogenic therapy.
associated CNV [20]. In the pathogenesis of
AMD, several factors contribute to CNV devel-
opment such as inflammation, complement cas- 11.5.2 Corticosteroids
cade activation, macrophage dysregulation, and
abnormal growth factors levels [20]. These fac- Systemic and local corticosteroids have been
tors and a healthier RPE related to the younger widely used in the treatment of CNV secondary
age of uveitis patients could explain why AMD- to uveitis due to its anti-inflammatory effects.
11
Table 11.2 Summary of studies showing antiangiogenic therapy in inflammatory CNV. MC Multifocal choroiditis, PIC Punctate inner choroidopathy, SC Serpiginous choroi-
ditis, VKH Vogt–Koyanagi–Harada syndrome, Toxo Ocular toxoplasmosis, POHS Presumed ocular histoplasmosis syndrome, SO Sympathetic ophthalmia, Sarco Ocular sarcoid-
osis, VA Stab Number of patients with visual acuity stabilization at the end of follow-up, Mean IVT Mean intravitreal therapy injections in the study
Inflammatory Choroidal Neovascularization
135
There is evidence to suggest that inflammatory 11.5.5 Immunosuppressive

cytokines produced in uveitis could promote the Treatment
development of CNV [38]. Flaxel et al. described
in a series of subfoveal CNV secondary to MFC Immunosuppressive agents have been introduced
and PIC vision stabilization in 83% of patients as an alternative treatment for noninfectious
that were treated with 1 mg/kg/day of oral corti- CNV, mainly for patients where corticosteroid
costeroid for 3–5 days followed by a tapering treatment is contraindicated.
dose period [39]. Martidis et al. compared the Systemic immunosuppressive therapies can
effect of oral prednisone versus a single sub- be effective in the resolution of inflammatory
Tenon triamcinolone acetonide injection for sub- CNV as demonstrated by Dees et al. in a series of
foveal CNV secondary to POHS. Oral prednisone CNV in posterior uveitis cases treated with cyclo-
group resulted in a greater initial VA improve- sporin [42]. Immunosuppressive therapies may
ment at 2 weeks that remain stable at 3 months, be also used in combination with other treat-
but no differences were observed in the final VA ments. Hogan et al. showed that a combination
between both treatment options [40]. In another treatment with oral mycophenolate and photody-
study of CNV secondary to POHS, Rechtman namic therapy achieved good disease control in
et al. reported VA improvement in 30% and stabi- patients with inflammatory CNV associated to
lization in 50% of eyes treated with intravitreal posterior uveitis for a relatively long follow-up
triamcinolone alone [41]. period [43]. Combination therapies with systemic
methotrexate and anti-VEGF drugs have also
been reported with positive outcomes in inflam-
11.5.3 Laser Photocoagulation matory CNV [44]. The use of intravitreal metho-
trexate has also been reported in inflammatory
Laser photocoagulation causes the destruction CNV related to multifocal choroiditis [45].
of the CNV, and for this reason it is not utilized
for subfoveal CNV due to the development of Key Points
permanent scotomas. The macular photocoagu- • Choroidal neovascularization (CNV) is a
lation study (MPS) showed post laser therapy severe complication of uveitis.
benefits for juxtafoveal and extrafoveal CNV in • CNV may be present at diagnosis or appear
POHS, with reduced relative risk of visual acu- with the evolution of the disease, so home-
ity loss of six lines at 5 years [1, 2]. However, monitoring with Amsler grid charts, high
since the advent of anti-VEGF therapies, laser awareness, and close follow-up are required
photocoagulation treatment has been relegated despite the stability of the associated uveitis
even in these cases and it is no longer used for syndrome.
CNV. • OCT and OCTA appear as the most employed
retinal imaging techniques for the diagnosis
and follow-up of CNV lesions.
11.5.4 Photodynamic Therapy • Anti-VEGF therapy is the first-line treatment
for both infectious and noninfectious CNV
Photodynamic therapy (PDT) with verteporfin membranes.
causes vascular damage and occlusion after laser
photoactivation. PDT has been used as mono-
therapy or in combination with other treatments References
(anti-VEGF, corticosteroids, and immunosup-
pression) in inflammatory CNV associated to 1. Dhingra N, Kelly S, Majid MA, Bailey CB,
Dick AD. Inflammatory choroidal neovascu-
MC, PIC, SC, VKH syndrome, toxoplasmic reti- lar membrane in posterior uveitis-pathogenesis
nochoroiditis, and POHS. Stabilization of lesions and treatment. Indian J Ophthalmol. 2010;58(1):
was common in most of the studies [3]. 3–10.
2. Agarwal A, Invernizzi A, Singh RB, Foulsham W, 16. Levison AL, Baynes KM, Lowder CY, Kaiser PK,
Aggarwal K, Handa S, Agrawal R, Pavesio C, Gupta Srivastava SK. Choroidal neovascularization on opti-
V. An update on inflammatory choroidal neovascular- cal coherence tomography angiography in punctate
ization: epidemiology, multimodal imaging, and man- inner choroidopathy and multifocal choroiditis. Br J
agement. J Ophthalmic Inflamm Infect. 2018;8(1):13. Ophthalmol. 2017;101(5):616–22.
3. Miller DG, Singerman LJ. Vision loss in younger 17. Amer R, Priel E, Kramer M. Spectral-domain opti-
patients: a review of choroidal neovascularization. cal coherence tomographic features of choroidal
Optom Vis Sci. 2006;83(5):316–25. neovascular membranes in multifocal choroiditis and
4. Walia HS, Shah GK, Blinder KJ. Treatment of punctate inner choroidopathy. Graefes Arch Clin Exp
CNV secondary to presumed ocular histoplasmosis Ophthalmol. 2015;253(6):949–57.
with intravitreal aflibercept 2.0 mg injection. Can J 18. Parodi MB, Iacono P, La Spina C, Knutsson KA,
Ophtalmol. 2016;51(2):91–6. Mansour A, Arevalo JF, Bandello F. Intravitreal bevaci-
5. Neri P, Lettieri M, Fortuna C, et al. Inflammatory zumab for choroidal neovascularisation in serpiginous
choroidal neovascularization. Middle East Afr J choroiditis. Br J Ophthalmol. 2014;98(4):519–22.
Ophthalmol. 2009;16:245–51. 19. Zahid S, Chen KC, Jung JJ, Balaratnasingam C,
6. Symons RCA, Shah SM, Do DV, Hanout M, Sepah Ghadiali Q, Sorenson J, Rofagha S, Freund KB,
YJ, Nguyen QD. Neovascularization. In: Zierhut Yannuzzi LA. Optical coherence tomography angiog-
M, editor. Intraocular inflammation. 1st ed. Berlin: raphy of chorioretinal lesions due to idiopathic multi-
Springer; 2016. p. 471–83. focal choroiditis. Retina. 2017;37(8):1451–63.
7. Baxter SL, Pistilli M, Pujari SS, Liesegang TL, Suhler 20. D'Ambrosio E, Tortorella P, Iannetti L. Management
EB, Thorne JE, Foster CS, Jabs DA, Levy-Clarke GA, of uveitis-related choroidal neovascularization:
Nussenblatt RB, Rosenbaum JT, Kempen JH. Risk from the pathogenesis to the therapy. J Ophthalmol.
of choroidal neovascularization among the uveitides. 2014;2014:450428.
Am J Ophthalmol. 2013;156(3):468–77.e2. 21. Xiong M, Elson G, Legarda D, Leibovich
8. Kotsolis AI, Killian FA, Ladas ID, Yannuzzi SJ. Production of vascular endothelial growth factor
LA. Fluorescein angiography and optical coherence by murine macrophages: regulation by hypoxia, lac-
tomography concordance for choroidal neovascu- tate, and the inducible nitric oxide synthase pathway.
larisation in multifocal choroidtis. Br J Ophthalmol. Am J Pathol. 1998;153(2):587–98.
2010;94(11):1506–8. 22. Paroli MP, Teodori C, D'Alessandro M, Mariani P,
9. Atmaca LS, Batioğlu F, Atmaca P. Evaluation of Iannucci G, Paroli M. Increased vascular endothelial
choroidal neovascularization in age-related macu- growth factor levels in aqueous humor and serum of
lar degeneration with fluorescein and indocya- patients with quiescent uveitis. Eur J Ophthalmol.
nine green videoangiography. Ophthalmologica. 2007;17(6):938–42.
1996;210(3):148–51. 23. Banerjee S, Savant V, Scott RA, Curnow SJ, Wallace
10. Drexler W, Fujimoto JG. State-of-the-art retinal GR, Murray PI. Multiplex bead analysis of vitreous
optical coherence tomography. Prog Retin Eye Res. humor of patients with vitreoretinal disorders. Invest
2008;27(1):45–88. Ophthalmol Vis Sci. 2007 May;48(5):2203–7.
11. Wu K, Zhang X, Su Y, Ji Y, Zuo C, Li M, Wen 24. Schadlu R, Blinder KJ, Shah GK, Holekamp NM,
F. Clinical Characteristics of Inflammatory Choroidal Thomas MA, Grand MG, Engelbrecht NE, Apte
Neovascularization in a Chinese Population. Ocul RS, Joseph DP, Prasad AG, Smith BT, Sheybani
Immunol Inflamm. 2016;24(3):261–7. A. Intravitreal bevacizumab for choroidal neovascu-
12. Hoang QV, Cunningham ET Jr, Sorenson JA, Freund larization in ocular histoplasmosis. Am J Ophthalmol.
KB. The “pitchfork sign” a distinctive optical coher- 2008;145(5):875–8.
ence tomography finding in inflammatory choroidal 25. Korol AR, Zborovska O, Kustryn T, Dorokhova O,
neovascularization. Retina. 2013;33(5):1049–55. Pasyechnikova N. Intravitreal aflibercept for cho-
13. Yu S, Lu J, Cao D, Liu R, Liu B, Li T, Luo Y, Lu roidal neovascularization associated with chorio-
L. The role of optical coherence tomography angi- retinitis: a pilot study. Clin Ophthalmol. 2017;11:
ography in fundus vascular abnormalities. BMC 1315–20.
Ophthalmol. 2016;16:107. 26. Lott MN, Schiffman JC, Davis JL. Bevacizumab
14. Cheng L, Chen X, Weng S, Mao L, Gong Y, Yu S, Xu in inflammatory eye disease. Am J Ophthalmol.
X. Spectral-domain optical coherence tomography 2009;148(5):711–7.e2.
angiography findings in multifocal choroiditis with 27. Adan A, Mateo C, Navarro R, Bitrian E, Casaroli-
active lesions. Am J Ophthalmol. 2016;169:145–61. Marano RP. Intravitreal bevacizumab (avastin) injec-
15. Adan A, Llorens V, Mesquida M, Carreño E. OCTA tion as primary treatment of inflammatory choroidal
en neovascularización coroidea inflamatoria. In: neovascularization. Retina. 2007;27(9):1180–6.
Adan A, Zarranz-Ventura J, editors. Angiografía 28. Tran TH, Fardeau C, Terrada C, Ducos De Lahitte G,
por Tomografía de Coherencia Óptica (OCTA). Bodaghi B, Lehoang P. Intravitreal bevacizumab for
Mesa Redonda 93 Congreso Sociedad Española de refractory choroidal neovascularization (CNV) sec-
Oftalmología. 1st ed. Madrid: Sociedad Española de ondary to uveitis. Graefes Arch Clin Exp Ophthalmol.
Oftalmología; 2017. p. 133–40. 2008 Dec;246(12):1685–92.
29. Fine HF, Zhitomirsky I, Freund KB, Barile GR, SK. Three-year visual and anatomic results of admin-
Shirkey BL, Samson CM, Yannuzzi LA. Bevacizumab istrating intravitreal bevacizumab in inflamma-
(avastin) and ranibizumab (lucentis) for choroidal tory ocular neovascularization. Can J Ophthalmol.
neovascularization in multifocal choroiditis. Retina. 2012;47(3):269–74.
2009 Jan;29(1):8–12. 38. Reddy VM, Zamora RL, Kaplan HJ. Distribution of
30. Doctor PP, Bhat P, Sayed R, Foster CS. Intravitreal growth factors in subfoveal neovascular membranes
bevacizumab for uveitic choroidal neovascularization. in age-related macular degeneration and presumed
Ocul Immunol Inflamm. 2009 Mar–Apr;17(2):118–26. ocular histoplasmosis syndrome. Am J Ophthalmol.
31. Kramer M, Axer-Siegel R, Jaouni T, Reich E, 1995;120(3):291–301.
Hemo I, Priel E, Averbukh E, Ehrlich R, Chowers I, 39. Flaxel CJ, Owens SL, Mulholland B, Schwartz SD,
Weinberger D, Amer R. Bevacizumab for choroidal Gregor ZJ. The use of corticosteroids for choroidal
neovascularization related to inflammatory diseases. neovascularisation in young patients. Eye (Lond).
Retina. 2010 Jun;30(6):938–44. 1998;12(Pt 2):266–72.
32. Julián K, Terrada C, Fardeau C, Cassoux N, Français 40. Martidis A, Miller DG, Ciulla TA, Danis RP, Moorthy
C, LeHoang P, Bodaghi B. Intravitreal bevacizumab as RS. Corticosteroids as an antiangiogenic agent for
first local treatment for uveitis-related choroidal neo- histoplasmosis-related subfoveal choroidal neovascu-
vascularization: long-term results. Acta Ophthalmol. larization. J Ocul Pharmacol Ther. 1999;15(5):425–8.
2011;89(2):179–84. 41. Rechtman E, Allen VD, Danis RP, Pratt LM, Harris
33. Rouvas A, Petrou P, Douvali M, Ntouraki A, Vergados A, Speicher MA. Intravitreal triamcinolone for
I, Georgalas I, Markomichelakis N. Intravitreal choroidal neovascularization in ocular histoplas-
ranibizumab for the treatment of inflammatory cho- mosis syndrome. Am J Ophthalmol. 2003;136(4):
roidal neovascularization. Retina. 2011;31(5):871–9. 739–41.
34. Menezo V, Cuthbertson F, Downes SM. Positive 42. Dees C, Arnold JJ, Forrester JV, Dick
response to intravitreal ranibizumab in the treatment AD. Immunosuppressive treatment of choroidal neo-
of choroidal neovascularization secondary to punctate vascularization associated with endogenous posterior
inner choroidopathy. Retina. 2010;30(9):1400–4. uveitis. Arch Ophthalmol. 1998;116(11):1456–61.
35. Cornish KS, Williams GJ, Gavin MP, Imrie FR. Visual 43. Hogan A, Behan U, Kilmartin DJ. Outcomes after
and optical coherence tomography outcomes of combination photodynamic therapy and immunosup-
intravitreal bevacizumab and ranibizumab in inflam- pression for inflammatory subfoveal choroidal neovas-
matory choroidal neovascularization secondary to cularisation. Br J Ophthalmol. 2005;89(9):1109–11.
punctate inner choroidopathy. Eur J Ophthalmol. 44. Kolomeyer AM, Roy MS, Chu DS. The use of intra-
2011;21(4):440–5. vitreal ranibizumab for choroidal neovascularization
36. Iannetti L, Paroli MP, Fabiani C, Nardella C, associated with Vogt-Koyanagi-Harada syndrome.
Campanella M, Pivetti-Pezzi P. Effects of intravit- Case Rep Med. 2011;2011:747648.
real bevacizumab on inflammatory choroidal neo- 45. Mateo-Montoya A, Baglivo E, de Smet
vascular membrane. Eur J Ophthalmol. 2013;23(1): MD. Intravitreal methotrexate for the treatment of
114–8. choroidal neovascularization in multifocal choroidi-
37. Mansour AM, Arevalo JF, Fardeau C, Hrisomalos EN, tis. Eye (Lond). 2013;27(2):277–8.
Chan WM, Lai TY, Ziemssen F, Ness T, Sibai AM, 46. Roy R, Saurabh K, Bansal A, Kumar A, Majumdar
Mackensen F, Wolf A, Hrisomalos N, Heiligenhaus AK, Paul SS. Inflammatory choroidal neovasculariza-
A, Spital G, Jo Y, Gomi F, Ikuno Y, Akesbi J, tion in Indian eyes: Etiology, clinical features, and
LeHoang P, Adan A, Mahendradas P, Khairallah M, outcomes to anti-vascular endothelial growth factor.
Guthoff R, Ghandour B, Küçükerdönmez C, Kurup Indian J Ophthalmol. 2017;65(4):295–300.
Dystrophy-Related Choroidal
Pierluigi Iacono, Stefano Da Pozzo,
Alessandro Papayannis, Francesco Romano,
Alessandro Arrigo, and Maurizio Battaglia Parodi
12.1 Introduction related to this topic derives from case series or

single case reports. As observed in the more fre-
Choroidal neovascularization (CNV) can compli- quent forms of CNV associated with age-related
cate the clinical course of chorioretinal dystro- macular degeneration or pathological myopia,
phies. In the forms with primary involvement of the advent of anti-VEGF drug therapy has pro-
the macular area, the occurrence of the CNV vided a more reassuring perspective on the evolu-
adds a high and significant risk of rapid deterioration of the disease and the efficacy of the therapy.
tion of the function of central vision and overall Below we will examine the most recent studies
the age of onset is very early compared to the on CNV associated with retinal dystrophies and
more common form associated with age-related the available treatment options.
macular degeneration (AMD) [1, 2]. To date,
there is no precise data on the prevalence of CNV
in retinal dystrophies and most of our knowledge 12.2 Retinitis Pigmentosa
P. Iacono (*) Retinitis Pigmentosa (RP) is an inherited retinal

IRCCS-Fondazione Bietti, Rome, Italy dystrophy caused by the loss of photoreceptors
TS-RETINA, Trieste, Italy and characterized by retinal pigment deposits
S. Da Pozzo visible on fundus examination. The prevalence of
TS-RETINA, Trieste, Italy RP is approximately 1/5000 worldwide, so that
A. Papayannis RP is the most common inherited disease of the
Azienda Sanitaria Universitaria Giuliano Isontina, retina. The most common form of RP is a rod-
S.O.C. di Oculistica di Monfalcone e Gorizia, cone dystrophy and it is initially characterized by
Monfalcone, Italy
night blindness, followed by the progressive loss
F. Romano in the peripheral visual field in daylight, and
Eye Clinic, Department of Biomedical and Clinical
Science, Luigi Sacco Hospital, University of Milan,
eventually leading to blindness after several
Milan, Italy decades. The course of the disease may be fre-
A. Arrigo
quently complicated by cataract formation in
Department of Ophthalmology of San Raffaele Scientific young age and cystoid macular edema. The exact
Institute, University Vita-Salute, Milan, Italy prevalence of CNV in RP is unknown. Two inter-
M. B. Parodi esting and recent studies explored in retrospec-
TS-RETINA, Trieste, Italy tive analysis the morphological alterations of the
Department of Ophthalmology of San Raffaele Scientific macular area by mean of optical coherence
Institute, University Vita-Salute, Milan, Italy tomography in a large sample of patients affected
140 P. Iacono et al.
by RP at different stages of evolution [3, 4]. ing finger at 2 m, received an initial single intra-
Triolo et al. identified three cases of CNV in a vitreal bevacizumab 1.25 mg injection [7]. One
group of 176 eyes with a calculated prevalence of month later, the CNV regressed with a complete
1.7%; no evidence of CNV was demonstrated in resolution of the serous detachment, a meaning-
the cohort of 183 eyes examined by Braima et al. ful decrease in central retinal thickness, passing
The low prevalence of this uncommon com- from 415 to 290 μm, and no leakage on fluores-
plication is clearly reflected in the scarce litera- cein angiography. Although a single injection
ture available with reference to this topic. To date achieved a complete stabilization of the neovas-
and according to our research, only a few case cular lesion with no recurrence in the following
reports have been published which, however, 12 months of follow-up, no benefit on visual acu-
offer some interesting considerations. ity could be recorded.
Ivakiri et al. report data on the natural history Bevacizumab was also employed as a first-line
in two patients with CNV associated with RP and treatment for juxtafoveal CNV associated with RP
followed for a period of 14 and 15 years, respec- by Battaglia Parodi et al. [8]. A 66-year-old woman
tively [5]. The lesions were juxtafoveal and sub- with baseline visual acuity of 20/200 in the left
foveal and received no type of treatment. In the eye, classic juxtafoveal CNV evidenced on fluo-
first case, the visual acuity was 0.3 at the first rescein angiography, neurosensory detachment
visit and was reduced to 0.04 at 14 years of overlying the hyperreflective lesion corresponding
follow-ups. In the second case, visual acuity to the CNV as demonstrated on OCT examination,
passed from 0.03 at the first visit to hand motion received intravitreal injections of bevacizumab
at 15 years of follow-ups. As expected, the visual (1.25 mg) during a monthly monitoring planned of
function declined progressively. follow-up. At 1 month of examination, the visual
Marano et al. describes a single case with acuity improved to 20/100 and a complete resolu-
CNV subfoveal, followed for 15 months and tion of neurosensory detachment was demon-
receiving a partial laser treatment of the strated on OCT. During the 1-year follow-up, the
CNV. The patient showed a stabilization of patient experienced reactivation of the CNV with
visual acuity that was equal to 0.1 with some recurrent neurosensory detachment and visual
phases of CNV reactivation and bleeding in the acuity deterioration and received four additional
follow-up period [1]. injections. The patient was monitored for visual
Cheng et al. reported the results of PDT with field and electroretinogram responses which
verteporfin in two patients [6]. In the first case, a turned out to be stable during the follow-up.
63-year-old female, with extrafoveal predomi- Finally, at 12 months of examination, visual acuity
nantly, classic CNV, subretinal hemorrhage and was stable at 20/100 and no sign of CNV activity
visual acuity 6/15 received a single session of was demonstrated, especially with no subfoveal
PDT. Her visual acuity improved to 6/9 at 3-months extension. Compared with the case presented by
examination with slow resolution of the subretinal Malik, Battaglia Parodi et al. demonstrated the
hemorrhage and after 2 years of follow-up showed importance to customize the monitoring of the
no recurrence. A second case, a 72-year-old male, neovascular lesion as it is possible to observe a fre-
presented a subfoveal predominantly classic CNV quent reactivation of the CNV. Another important
and a baseline visual acuity of 6/120. The patient observation provided by Battaglia Parodi et al.
received a single session of PDT with a subsequent concerns the safety of the treatment. In particular,
improvement of visual acuity, a progressive reso- the neuroprotective effect of VEGF on retinal gan-
lution of the CNV, and no recurrence. glion cells and the role for endogenous VEGF in
The last treatment option considered for CNV the maintenance and function of Muller cells and
related to RP is anti-VEGF. The first experience photoreceptors is well known; these findings
on the use of bevacizumab was reported by Malik should be particularly taken into consideration in
et al. A 40-year-old male with subfoveal classic RP, characterized by progressive anatomical dam-
CNV and a serious visual acuity decline, count- age involving the same cell populations. In this
12 Dystrophy-Related Choroidal Neovascularization 141
single case, the visual field and the electrooculo- nine green angiography. In particular, no descrip-
gram responses did not get worse during the 1-year tion was provided with regard to the evaluation of
follow-up; this suggests that a limited anti-VEGF type 3 neovascularization in association with cys-
therapy may not lead to further damage. toid macular edema and which methodology was
A similar conclusion was achieved in a report applied during the anti-VEGF therapy monitor-
presented by Miyata et al. A 56-year-old woman ing to differentiate the retinal fluid originating
received an anti-VEGF treatment in a pro-re-nata from the neovascularization and the fluid collec-
regimen for multiple recurrences of CNV activity tion secondary to the cystoid macular edema.
over a 2-year follow-up. Despite a tailored moni-
toring, the patient experienced a visual acuity
deterioration to a value of 20/100 starting from 12.3 Best Vitelliform Macular
the baseline value of 20/50. Subsequently the Dystrophy
case was treated in a fixed bimonthly regimen
that continued for 6 years with 34 anti-VEGF CNV is a possible complication of Best vitelli-
injections administered in the whole period. form macular dystrophy (BVMD). CNV may
During the follow-up a slight decline in the cen- show exudative manifestations, with subretinal
tral visual field was registered without a differ- fluid and subretinal hemorrhages, leading to a
ence between the treated and untreated eye, with visual acuity worsening [10]. Typically exudative
a mean deviation reduction of 0.32 and 0.68 dB/ CNV develops in the initial stages of the disease,
year, in the treated and untreated eye. At the final before the vitelliruptive evolution. On the other
visit the CNV did not show activity and the visual hand, CNV without exudation is detectable in
acuity was stable at a solid 20/100. almost 90% of eyes affected by BVMD in the
Finally, an additional interesting case report more advanced stages of the disease [11]. Focal
was described by Sayadi et al. [9]. In this case choroidal excavation can be associated with the
OCT angiography was applied to describe a type CNV growth, probably due to the mechanical
3 neovascularization, formerly defined retinal contraction secondary to the CNV [12].
angiomatous proliferation, in a patient with RP Identification of the CNV can be achieved by
and cystoid macular edema and receiving 8 intra- means of conventional techniques including fluo-
vitreal ranibizumab injections over a 1-year fol- rescein and indocyanine green angiography but
low-up. Unfortunately, the authors did not report also using optical coherence tomography [13].
the effects of anti-VEGF therapy for the type 3 Optical coherence tomography angiography can
neovascularization, the change in the visual acu- effectively visualize CNV associated with
tiy, the comparison of the OCT angiography BVMD, depicting the whole neovascular exten-
image with conventional fluorescein, or indocya- sion (Figs. 12.1, 12.2, 12.3, and 12.4) [11].
Fig. 12.1 Choroidal neovascularization in Best vitelliform macular dystrophy. Nine-year-old male in Best pseudohy-
popyon stage on blue-light fundus autofluorescence and on optical coherence tomography
Fig. 12.2 Choroidal neovascularization in Best vitelli- cence due to subretinal hemorrhage. OCT disclosed a
form macular dystrophy. Same case as Fig. 12.1, 1 year wide subretinal detachment. Fluorescein angiography
later, when choroidal neovascularization developed. Blue- clearly showed a type II choroidal neovascularization in
light fundus autofluorescence revealed a blocked fluores- the early phase, with dye leakage in more advanced phases
Fig. 12.3 Choroidal neovascularization in Best vitelli- hyperreflective signal. OCT scan reveals the subretinal
form macular dystrophy. Same case as Fig. 12.1, same fluid resolution, whereas fluorescein angiography dis-
patient 6 months after a single ranibizumab injection. closes no dye leakage from the scar
Blue-light fundus autofluorescence shows an irregular
Fig. 12.4 Choroidal neovascularization in Stargardt dis- rescence reveals the characteristic hyperautofluorescence
ease/fundus flavimaculatus. Fluorescein angiography in areas of lipofuscin accumulation at the posterior pole.
clearly shows leakage from choroidal neovascularization Optical coherence tomography angiography well identi-
on a background of “silent” choroid and multiple hyper- fies the neovascular network. B-scan OCT demonstrates
fluorescent pisciform flecks. Blue-light fundus autofluo- the CNV with small amount of intraretinal fluid
Compared with fluorescein angiography, OCT-A of the CNV and a more precise quantification of
allows a better characterization of the CNV, the CNV area also in the presence of vitelliform
clearly delineating the CNV network in at least deposits able to mask the CNV borders; OCT-A
four patterns, namely, dense net, loose net, depicts more clearly the vasculature across the
unidentifiable, and in a pattern more recently different retinal layers. However, OCT-A pres-
defined as “ring-shape.” Additional advantages of ents some limitations, the main being the inabil-
OCT-A are represented by a better visualization ity to detect the leakage and as a consequence it
fails to provide suitable information on the activ- the pathologic lesions involving the photorecep-
ity of the CNV [14, 15]. tors, the retinal pigment epithelium (RPE) and
The management of CNV secondary to the retinal–choroidal vascularization [20, 21]. In
BVMD is still unclear. Overall, treatment, espe- particular, it has been demonstrated that the loss
cially with intravitreal anti-VEGF, is associated of the inner and outer segments of the photore-
with better functional outcomes with respect to ceptors precedes the degenerative damage involv-
observation alone [10]. At present, possible treating the RPE [22]. The presence of hyperreflective
ment options include photodynamic therapy and foci (HF) is an additional morphological element
intravitreal anti-VEGF. Photodynamic therapy is to be taken into account in the evolutionary pro-
effective in stopping the neovascular growth, but cesses of Stargardt’s disease. In detail, the HF
it may bring about atrophic changes and subreti- turns out to be more frequent in the pathological
nal fibrosis causing functional damage over the edge compared with the healthy edge of the atro-
follow-up. Anti-VEGF approach is currently the phy. In addition, HF number in the outer retina of
most used treatment, requiring a number of intra- the pathological edge significantly decreases dur-
vitreal injections from 1 to 2, and leading to the ing the different stages of evolution with an
stabilization of the CNV activity [16]. enlargement of the area as the HF number pro-
gressively reduces [23]. OCT-A has further con-
tributed to the understanding of degenerative
12.4 Stargardt Disease/Fundus mechanisms affecting retinal and choroidal vas-
Flavimaculatus cular laminae. Guduru et al. have shown that
alterations in the RPE appear to be significantly
Since its first report in medical literature, the larger than the choriocapillaris layer vessel loss,
clinical spectrum of Stargardt disease (SD) has which suggests that RPE damage might precede
enriched increasingly. Usually, in early stages, that of choriocapillaris [24]. Battaglia Parodi
patients sustain a progressive visual loss, present- et al. demonstrated a progressive vascular impair-
ing in childhood or early adulthood with normal ment with an early involvement of the superficial
fundus appearance [17]. Later in life, the disease and the deep retinal plexuses. In eyes with
progresses toward macular atrophy and the advanced atrophic changes a greater reduction in
appearance of yellowish retinal flecks. The final choriocapillaris density was demonstrated in
result is a visual acuity deterioration up to 20/200 comparison with eyes without atrophy.
or worse. SD transmission is autosomal recessive Mastropasqua et al. convey to similar conclu-
and it is due to mutations in ABCA4 gene (OMIM sions. Eyes with advanced Stargardt’s disease
601691), which encodes an adenosine showed a meaningful reduction of the vascular
triphosphate- binding cassette transporter [18, density of the superficial capillary plexus, deep
19]. In the absence of normal ABCA4 function, capillary plexus, and choriocapillaris related to a
photoreceptor metabolism is compromised and progressive reduction of the macular thickness
A2E, a component of lipofuscin, accumulates in [25]. The morphological characterization of the
photoreceptors and the retinal pigment epithe- retinal alterations in the Stargardt’s disease needs
lium (RPE), causing injury to both. further study. However, numerous potential bio-
Laboratory and hystopathologic findings markers have been made available as sensitive
described the essence of SD to be a diffuse lipo- anatomic outcome measures for defining the
fuscin deposits at the RPE level. On fluorescein more appropriate endpoints for clinical trials.
angiography, choroidal details are frequently Another clinical entity, defined as fundus fla-
masked by these deposits, a sign also known as vimaculatus (FFM) is an SD-like phenotype, first
“dark” choroid. described by Franceschetti and François. It is
In more recent years, optical coherence characterized by late-onset of symptoms with
tomography (OCT) provided meaningful infor- slow deterioration, but still presenting a progres-
mation regarding the location and progression of sive central atrophy and late loss of visual acuity
[26]. SD and FFM are terms frequently used growth of atrophy in SD/FFM. Lipofuscin-
interchangeably in the literature, but probably it filled RPE cells may be more vulnerable to fur-
is more proper to consider them as different phe- ther degenerative processes secondary to the
notypes of a single, heterogeneous disease. effects of anti-VEGF molecules. on RPE and
Degenerative changes involving the RPE may choriocapillaris. Also the high retreatment rate
extend to the Bruch’s membrane, paving the way may imply that ranibizumab may control the
to subsequent choroidal neovascularization CNV activity only in part. OCT detection of
(CNV). On average, CNV is an infrequent com- subretinal hyperreflective material associated
plication in late stages of retinal dystrophies, In with hyperreflective dots may be interpreted as
the available literature there are only a few case the indirect manifestation of an inflammatory
series and case reports about CNV associated reaction related to these biochemical alterations
with SD/FFM. [41, 42].
CNV is a rare complication in SD/FFM (up to In conclusion, even if CNV complicating SD/
2% according to Armstrong et al. [27]), but FF is quite rare, available protocols of treatment
associated with a rapid progression and poor provide suboptimal results in terms of visual acu-
prognosis (Fig. 12.4) [27–29]. More recently ity preservation.
even retinal angiomatous proliferation or type 3
CNV was described in a case of FF [30].
Natural history in terms of final visual acuity 12.5 Pattern Dystrophy
has been described [27, 31] and resulted to be
unfavorable. Data about treatment options are The definition of pattern dystrophy (PD) of the
strictly dependent on the time period of the retinal pigment epithelium (RPE) includes a het-
reports. Laser photocoagulation was proposed erogeneous group of inherited retinal diseases
but is now abandoned. with a natural history that is not entirely known.
Photodynamic therapy (PDT) with vertepor- These disorders are characterized by abnormal
fin was then proposed and administered pigment distribution in the RPE due to the accu-
[32–36]. Only case series with a maximum of mulation of lipofuscin [15, 43, 44].
three eyes or single case reports are available. Overall, five main PD subforms can be iden-
On average a stabilization in visual acuity was tified on the basis of the RPE alterations
achieved with one or more PDT sessions. [15, 43, 45]:
However, the rarity of SD/FFM and CNV asso-
ciation did not allow to plan a controlled study. –– Adult-Onset Foveomacular Vitelliform
Encouraging results can be obtained also Dystrophy (AOFVD)
administering intravitreal injection with anti- –– Butterfly-shaped pigment dystrophy
VEGF drugs. As well as with PDT only small –– Reticular dystrophy (RD)
case series or single case reports were reported, –– Multifocal PD simulating fundus
both with ranibizumab [30, 37–40] and bevaci- flavimaculatus
zumab [41]. Battaglia Parodi et al. reported that –– Fundus pulverulentus
the frequency of injections did not differ sig-
nificantly from corresponding values described Mutations in the peripherin/RDS gene and a
for CNV in age-related macular degeneration high phenotypic variability have been observed
(7 per year). Anti-VEGF usually can halt CNV between individuals. Although the natural his-
activity but visual acuity does not improve [40]. tory of PD is not clearly known the prognosis is
In the same paper, the authors pointed out some generally favorable. The visual function is gen-
potential drawbacks of anti-VEGF treatment. erally preserved for a long time and the patients
Enlargement of the atrophic changes and devel- usually have their visual acuity maintained
opment of outer retinal tubulation may indicate until the sixth decade. The onset of atrophy or
that anti-VEGF treatment may aggravate the CNV may complicate the clinical course,
leading to a progressive visual acuity deteriora- the presence of yellow-white crystals deposits in
tion. Nevertheless, some cases with subfoveal the retina and in about one-third of cases also into
CNV may show a relatively stable visual func- the cornea [57]. Progressive atrophy and degen-
tion in the short-term follow-up, with a visual eration of the retinal pigment epithelium and cho-
loss in the long run [43, 46, 47]. roid are observed as BCD evolves; symptoms are
Variable therapeutic options have been proposed similar to those of RP with reduced visual acuity,
for the management of subfoveal CNV related to hemeralopia, visual field loss, and impaired color
PD including laser photocoagulation, photody- vision. It is caused by mutations in the CYP4V2
namic therapy with verteporfin (vPDT), and anti- gene, which encodes for a protein whose structure
VEGF therapy. vPDT has been used in the treatment suggests that it may play an active role in fatty
of subfoveal CNV secondary to PD in some studies acid metabolism [58]. CNV may occur during the
obtaining a temporary short-term functional stabili- first stages of course disease and in young age
zation, followed subsequently by a progressive [59–63]. Two hypotheses are currently formu-
long-term visual acuity deterioration [48, 49]. More lated to explain the pathogenetic mechanisms
specifically, Battaglia Parodi et al. demonstrated a leading to the CNV formation. One theory pro-
visual acuity decline of 3 lines in 70% of treated poses that the CNV might be caused by the
cases at the 3-year examination. Laser photocoagu- chronic irritation of Bruch’s membrane by the
lation may bring about a functional deterioration, crystals deposits [64, 65]. The second theory sug-
due to the following scar enlargement, as described gests a main role of the drusen-like deposits char-
in a single case of extrafoveal CNV secondary to acterizing the BCD [59]. Le Tien and Nachiappan
fundus pulverulentus treated with conventional described two young patients, 29- and 33-years
laser photocoagulation [50]. old, receiving intravitreal ranibizumab [60, 63]. In
A larger number of studies are available with both cases ranibizumab stabilized the CNV after a
regard to anti-VEGF therapy for CNV associated course of 3 monthly intravitreal injections with a
with different PD subtypes. In spite of the limita- visual acuity improvement from 20/50 to 20/32 in
tions related to the administration of different a case and a visual acuity stabilization at 6/9 in the
drugs, study design, and treatment protocols, the second case.
clinical reports describe overall positive anatomi-
cal and functional effects, with resolution of the
CNV activity, visual acuity stabilization or 12.7 Miscellaneous
improvement in 87–92% of cases, over a 12–24-
month follow-up [51–56]. Owing to the rarity of CNV may be an uncommon complication in
CNV complicating PD, the studies share similar very rare retinal disorders including choroider-
and obvious limitations including the small num- emia, gyrate atrophy, cone dystrophy, enhanced
ber of patients, short follow-up, uneven genetic S-cone syndrome, congenital retinoschisis, and
characterization, and absence of a control group. only a handful of case reports are presented in
Nevertheless, the anti-VEGF approach represents the current literature [66–76]. Although the
the most promising treatment for the manage- occurrence of CNV can be considered infre-
ment of CNV secondary to PD. quent, clinicians should always bear in mind
this possible complication that generally appears
at a younger age and in subjects with a more
12.6 Bietti Crystalline unfavorable functional prognosis. Intravitreal
Corneoretinal Dystrophy anti-VEGF therapy seems to be a valid treat-
ment option; however, there are no guidelines
Bietti crystalline dystrophy (BCD) is a rare auto- for therapeutic drug monitoring related to a spe-
somal recessive chorioretinal degeneration first cific retinal disorder and each subject should
described by Bietti in 1939 and characterized by receive a tailored approach.
References 15. Shahzad R, Siddiqui MA. Choroidal neovasculariza-

tion secondary to best vitelliform macular dystrophy
detected by optical coherence tomography angiogra-
1. Marano F, Deutman AF, Leys A, Aandekerk
phy. J AAPOS. 2017;21(1):68–70.
AL. Hereditary retinal dystrophies and choroidal neo-
16. Padhi TR, Anderson BJ, Abbey AM, et al. Choroidal
vascularization. Graefes Arch Clin Exp Ophthalmol.
neovascular membrane in paediatric patients: clinical
2000;238:760–4.
characteristics and outcomes. BJO. 2018;102:1232–7.
2. Battaglia Parodi M, Iacono P, Bandello F. Antivascular
17. Battaglia Parodi M. Eredodistrofie Corioretiniche.
endothelial growth factor in hereditary dystrophies.
Rome: Verduci Editore; 2002. p. 21–60.
Dev Ophthalmol. 2010;46:107–10.
18. Allikmets R, Singh N, Sun H, et al. A photoreceptor
3. Triolo G, Pierro L, Parodi MB, De Benedetto U,
cell specific ATP-binding transporter gene (ABCR) is
Gagliardi M, Manitto MP, Bandello F. Spectral
mutated in recessive Stargardt macular dystrophy. Nat
domain optical coherence tomography findings in
Genet. 1997;15:236–46.
patients with retinitis pigmentosa. Ophthalmic Res.
19. Koenekoop RK. The gene for Stargardt disease,
2013;50:160–4.
ABCA4, is a major retinal gene: a mini review.
4. Braimah IZ, Dumpala S, Chhablani J. Outer reti-
Ophthalmic Genet. 2003;24:75–80.
nal tubulation in retinal dystrophies. Retina.
20. Jampol LM, Shankle R, Tornambe P, et al. Diagnostic
2017;37:578–84.
and therapeutic challenges. Retina. 2006;26:1072–6.
5. Iwakiri R, Okinami S, Hirata A. [Two cases of reti-
21. Margolis R, Mukkamala SK, Jampol LM, et al. The
nitis pigmentosa associated with choroidal neo-
expanded spectrum of focal choroidal excavation.
vascularization]. Nippon Ganka Gakkai Zasshi.
Arch Ophthalmol. 2011;129:1320Y5.
2007;111:606–611.
22. Alabduljalil T, Patel RC, Alqahtani AA, Gao SS, Gale
6. Cheng JY, Adrian KH. Photodynamic therapy for
MJ, Zhang M, Jia Y, Huang D, Chiang PW, Chen R,
choroidal neovascularization in stargardt disease
Wang J, Weleber RG, Pennesi ME, Yang P. Correlation
and retinitis pigmentosa. Retin Cases Brief Rep.
of outer retinal degeneration and choriocapillaris loss
2009;3:388–90.
in Stargardt disease using en face optical coherence
7. Malik A, Sood S, Narang S. Successful treatment of
tomography and optical coherence tomography angi-
choroidal neovascular membrane in retinitis pigmen-
ography. Am J Ophthalmol. 2019;202:79–90.
tosa with intravitreal bevacizumab. Int Ophthalmol.
23. Guduru A, Lupidi M, Gupta A, Jalali S, Chhablani
2010;30:425–8.
J. Comparative analysis of autofluorescence and OCT
8. Battaglia Parodi M, De Benedetto U, Knutsson KA,
angiography in Stargardt disease. Br J Ophthalmol.
Scotti F, Librando A, Bandello F, Iacono P. Juxtafoveal
2018;102(9):1204–7.
choroidal neovascularization associated with retinitis
24. Battaglia Parodi M, Sacconi R, Romano F, Bandello
pigmentosa treated with intravitreal bevacizumab. J
F. Hyperreflective foci in Stargardt disease: 1-year
Ocul Pharmacol Ther. 2012;28:202–4.
follow-up. Graefes Arch Clin Exp Ophthalmol.
9. Sayadi J, Miere A, Souied EH, Cohen SY. Type 3 neo-
2019;257(1):41–8.
vascularization associated with retinitis pigmentosa.
25. Mastropasqua R, Toto L, Borrelli E, Di Antonio
Case Rep Ophthalmol. 2017;8:245–9.
L, Mattei PA, Senatore A, Di Nicola M, Mariotti
10. Khan KN, Mahroo OA, Islam F, et al. Functional and
C. Optical coherence tomography angiogra-
anatomical outcomes of choroidal neovasculariza-
phy findings in Stargardt disease. PLoS One.
tion complicating BEST1-related retinopathy. Retina.
2017;12(2):e0170343.
2017;37:1360–70.
26. Franceschetti A, François J. Fundus flavimaculatus.
11. Battaglia Parodi M, Romano F, Cicinelli MV, et al.
Arch Ophthalmol. 1965;25:505–30.
Retinal vascular impairment in best vitelliform macu-
27. Armstrong J, Meyer D, Xu S, Elfervig J. Long-term
lar dystrophy assessed by means of optical coher-
follow-up of Stargardt’s disease and fundus flavimac-
ence tomography angiography. Am J Ophthalmol.
ulatus. Ophthalmology. 1998;105:448–58.
2018;187:61–70.
28. Klein R, Lewis RA, Meyers SM, Myers FL. Subretinal
12. Battaglia Parodi M, Casalino G, Iacono P, et al. The
neovascularization associated with fundus flavimacu-
expanding clinical spectrum of choroidal excavation
latus. Arch Ophthalmol. 1978;96:2054–7.
in macular dystrophies. Retina. 2018;38:2030–4.
29. Pawlak D, Souied E, Mimoun G, et al. Fundus fla-
13. Battaglia Parodi M, Iacono P, Romano F, Bandello
vimaculatus and choroidal neovascularisation. J Fr
F. Spectral domain optical coherence tomography
Ophtalmol. 2006;29:188–94.
features in different stages of best vitelliform macular
30. Quijano C, Querques C, Massamba N, Soubrane G,
dystrophy. Retina. 2018;38:1041–6.
Souied EH. Type 3 choroidal neovascularization asso-
14. Guduru A, Gupta A, Tyagi M, Jalali S, Chhablani
ciated with fundus flavimaculatus. Ophthalmic Res.
J. Optical coherence tomography angiography charac-
2009;42:152–4.
terisation of best disease and associated choroidal neo-
31. Rotenstreich Y, Fishman GA, Anderson RJ. Visual
vascularisation. Br J Ophthalmol. 2018;102(4):444–7.
acuity loss and clinical observations in a large series
of patients with Stargardt disease. Ophthalmology. 46. Taillanter-Francoz N, Remy C, Bonnet M, Baserer
2003;110:1151–8. T. Choroidal neovessels associated with reticular
32. Valmaggia C, Niederberger H, Helbig dystrophy of the pigment epithelium: case report (in
H. Photodynamic therapy for choroidal neovas- French). Bull Soc Ophtalmol Fr. 1981;81:539–41.
cularisation in fundus flavimaculatus. Retina. 47. Marano F, Deutman AF, Pinckers AJ, et al. Reticular
2002;22:111–3. dystrophy of the retinal pigment epithelium and cho-
33. Souied EH, Pawlak D, Algan M, Sayag D, Coscas roidal neovascularization: a fluorescein and ICGV
G, Soubrane G. Photodynamic therapy for choroidal study. Acta Ophthalmol Scand. 1997;75:22–7.
neovascularization on late-onset fundus flavimacula- 48. Parodi MB, Da Pozzo S, Ravalico G. Photodynamic
tus. Am J Ophthalmol. 2005;140:312–4. therapy for choroidal neovascularization associated
34. Japiassù RM, Ferreira Moura Brasil O, Soares Maia with pattern dystrophy. Retina. 2003;23:171–6.
H, Soares Maia A, Moura Brasil O. Photodynamic 49. Parodi MB, Liberali T, Pedio M, et al. Photodynamic
therapy for bilateral and simultaneous choroidal neo- therapy of subfoveal choroidal neovascularization
vascularization in Stargardt disease. Retin Cases Brief secondary to reticular pattern dystrophy: three-year
Rep. 2008;2:9–11. results of an uncontrolled, prospective case series. Am
35. Kim HH, Carvounis PE, Albini TA, Eric R, Holz J Ophthalmol. 2006;141:1152–4.
ER, Mieler WF. Photodynamic therapy for bilateral 50. Parodi MB. Choroidal neovascularization in fun-
subfoveal choroidal neovascularization complicating dus pulverulentus. Acta Ophthalmol Scand.
Stargardt macular dystrophy. Retin Cases Brief Rep. 2002;80:559–60.
2008;2:6–8. 51. Parodi MB, Iacono P, Cascavilla M, Zucchiatti I,
36. Cheng JY, Koh HCA. Photodynamic therapy for Kontadakis DS, Bandello F. Intravitreal bevacizumab
choroidal neovascularization in Stargardt disease for subfoveal choroidal neovascularization associated
and retinitis pigmentosa. Retin Cases Brief Rep. with pattern dystrophy. Invest Ophthalmol Vis Sci.
2009;3:388–90. 2010;51:4358–61.
37. Querques G, Bocco MC, Soubrane G, Souied 52. Schutt F, Davies S, Kopitz J, et al. Photodamage
EH. Intravitreal ranibizumab (Lucentis) fr choroi- to human RPE cells by A2-E, a retinoid compo-
dal neovascularization associated with Stargardt’s nent of lipofuscin. Invest Ophthalmol Vis Sci.
disease. Graefes Arch Clin Exp Ophthalmol. 2000;41:2303–8.
2008;246:319–21. 53. Mimoun G, Caillaux V, Querques G, et al.
38. Tejerina A, García-Pous M, Palacios Pozos E, Desco Ranibizumab for choroidal neovascularization associ-
Esteban MC, Mataix Boronat J. Ranibizumab for cho- ated with adult-onset foveomacular vitelliform dys-
roidal neovascularization in fundus flavimaculatus. trophy: one-year results. Retina. 2013;33:513–21.
Retin Cases Brief Rep. 2008;2:250–2. 54. Empeslidis T, Vardarinos A, Deane J, Banerjee
39. Koh V, Naing T, Chee C. Fundus flavimaculatus S. Intravitreal ranibizumab in the treatment of
and choroidal neovascularization in a young patient butterfly-shaped pattern dystrophy associated with
with normal electroretinography: case report. Can J choroidal neovascularization: a case report. Case Rep
Ophthalmol. 2012;47:e3–5. Ophthalmol. 2012;3:77–82.
40. Battaglia Parodi M, Munk MR, Iacono P, Bandello 55. Vaclavik V, Tran HV, Gaillard MC, et al. Pattern dys-
F. Ranibizumab for subfoveal choroidal neovascu- trophy with high intrafamilial variability associated
larisation associated with Stargardt disease. Br J with Y141C mutation in the peripherin/RDS gene
Ophthalmol. 2015;99:1268–70. and successful treatment of subfoveal CNV related to
41. Roy R, Kumar A, Ghosh S, Lobo A, Agarwal P. All multifocal pattern type with anti-VEGF (ranibizumab)
that glitters are not flecks: inflammatory choroidal intravitreal injections. Retina. 2012;32:1942–9.
neovascularization in fundus flavimaculatus. Ocul 56. Montero JA, Ruiz-Moreno JM, De La Vega
Immunol Inflamm. 2015;23(2):188–90. C. Intravitreal bevacizumab for adult-onset vitel-
42. Radu RA, Hu J, Yuan Q, et al. Complement system liform dystrophy: a case report. Eur J Ophthalmol.
dysregulation and inflammation in the retinal pigment 2007;17:983–6.
epithelium of a mouse model for Stargardt macular 57. Bietti G. Ueber familiaeres vorkommen von ‘reti-
degeneration. J Biol Chem. 2011;286:18593–601. nitis punctata albescens’ (verbunden mit ‘dystro-
43. Gass JDM. Stereoscopic atlas of macular diseases: phia marginalis cristallinea corneae’), glitzern des
diagnosis and treatment. 4th ed. St. Louis: CV Mosby; glaskoerpers und anderen degenerativen augenver-
1997. p. 314–25. aenderungen. Klin Monatsbl Augenheilkd. 1939;99:
44. Francis PJ, Schultz DW, Gregory AM, et al. Genetic 737–57.
and phenotypic heterogeneity in pattern dystrophy. Br 58. Li A, Jiao X, Munier FL, Schorderet DF, Yao W, Iwata
J Ophthalmol. 2005;89:1115–9. F, Hayakawa M, Kanai A, Shy Chen M, Alan Lewis R,
45. Marmor MF, Byers B. Pattern dystrophy of the Heckenlively J, Weleber RG, Traboulsi EI, Zhang Q,
retinal pigment epithelium. Am J Ophthalmol. Xiao X, Kaiser-Kupfer M, Sergeev YV, Hejtmancik
1976;13:112–6. JF. Bietti crystalline corneoretinal dystrophy is caused
by mutations in the novel gene CYP4V2. Am J Hum 67. Gulkilik G, Erdur SK, Eliacik M, Odabasi M, Ozsutcu
Genet. 2004;74(5):817–26. M, Demirci G, Kocabora MS. A case of cone dys-
59. Li Q, Li Y, Zhang X, Xu Z, Zhu X, Ma K, She H, trophy associated with choroidal neovascularization.
Peng X. Utilization of fundus autofluorescence, Retin Cases Brief Rep. 2018;12:111–4.
spectral domain optical coherence tomography, and 68. Chatziralli I, Theodossiadis G, Emfietzoglou I,
enhanced depth imaging in the characterization of Theodossiadis P. Intravitreal ranibizumab for choroi-
Bietti crystalline dystrophy in different stages. Retina. dal neovascularization secondary to gyrate atrophy in
2015;35:2074–84. a young patient: a multimodal imaging analysis. Eur J
60. Le Tien V, Atmani K, Querques G, Massamba N, Ophthalmol. 2015;25:e119–22.
Souied EH. Ranibizumab for subfoveal choroidal 69. Marano F, Deutman AF, Pinckers AJ, Aandekerk
neovascularization in Bietti crystalline retinopathy. AL. Gyrate atrophy and choroidal neovascularization.
Eye (Lond). 2010;24:1728–9. Arch Ophthalmol. 1996;114:1295.
61. Gupta B, Parvizi S, Mohamed MD. Bietti crystal- 70. Campos-Pavón J, Torres-Peña JL. Choroidal neovas-
line dystrophy and choroidal neovascularisation. Int cularization secondary to choroideremia. Arch Soc
Ophthalmol. 2011;31:59–61. Esp Oftalmol. 2015;90:289–91.
62. Atmaca LS, Muftuoglu O, Atmaca-Sonmez 71. Palejwala NV, Lauer AK, Weleber RG. Choroideremia
P. Peripapillary choroidal neovascularization in Bietti associated with choroidal neovascularization treated
crystalline retinopathy. Eye (Lond). 2007;21:839–42. with intravitreal bevacizumab. Clin Ophthalmol.
63. Nachiappan K, Krishnan T, Madhavan J. Ranibizumab 2014;8:1675–9.
for choroidal neovascular membrane in a rare case of 72. Endo K, Yuzawa M, Ohba N. Choroideremia asso-
Bietti’s crystalline dystrophy: a case report. Indian J ciated with subretinal neovascular membrane. Acta
Ophthalmol. 2012;60:207–9. Ophthalmol Scand. 2000;78:483–6.
64. Yin X, Yang L, Chen N, Cui H, Zhao L, Feng L, Li A, 73. Robinson D, Tiedeman J. Choroideremia associated
Zhang H, Ma Z, Li G. Identification of CYP4V2 muta- with a subretinal neovascular membrane. Case Rep
tion in 36 Chinese families with Bietti crystalline cor- Retina. 1987;7:70–4.
neoretinal dystrophy. Exp Eye Res. 2016;146:154–62. 74. Bertoli F, Pignatto S, Rizzetto F, Lanzetta P. A 5-year-
65. Mamatha G, Umashankar V, Kasinathan N, Krishnan old case of choroidal neovascularization in enhanced
T, Sathyabaarathi R, Karthiyayini T, Amali J, Rao C, S-cone syndrome treated with ranibizumab. Case Rep
Madhavan J. Molecular screening of the CYP4V2 Ophthalmol. 2018;9:510–5.
gene in Bietti crystalline dystrophy that is associ- 75. Broadhead GK, Grigg JR, McCluskey P, Korsakova
ated with choroidal neovascularization. Mol Vis. M, Chang AA. Bevacizumab for choroidal neovas-
2011;17:1970–7. cularisation in enhanced S-cone syndrome. Doc
66. Inanc M, Tekin K, Teke MY. Bilateral choroidal neo- Ophthalmol. 2016;133:139–43.
vascularization associated with gyrate atrophy man- 76. Ranchod TM, Faia LJ, Drenser KA. Peripapillary
aged with intravitreal bevacizumab. Int Ophthalmol. choroidal neovascularization in congenital reti-
2018;38:1351–5. noschisis. Retin Cases Brief Rep. 2011;5:
336–8.
Associated with Angioid Streaks 13
Christof Hänsli and Sandrine A. Zweifel
13.1 Angioid Streaks Secondary choroidal neovascularization

and Associated Diseases (CNV) occurs in up to 65% of patients with AS
[5]. Prognosis of untreated secondary CNV due
13.1.1 Introduction to AS is poor. Development of a disciform scar,
typically at the posterior pole, leads to a relevant
Angioid streaks (AS) are a clinical finding with vision loss with a best corrected visual acuity
linear or branching patterns in the posterior pole. (BCVA) of 20/200 or less at the fourth to fifth
They were first described in 1889 [1], named decade of life [6].
“angioid” in 1892 [2] according to their vessel-
like appearance, and found to be a pathology of
Bruch’s membrane (BM) in 1917 [3, 4]. 13.1.2 Pseudoxanthoma Elasticum:
Angioid streaks are a frequent finding in pseu- Incidence, Genetics,
doxanthoma elasticum (PXE), with an incidence and Histopathology
of 59–87% of patients. After 20 years of disease,
almost all patients with PXE have AS [3]. They Pseudoxanthoma elasticum is an autosomal
also occur in 8–15% of patients with Paget’s dis- recessive disorder, with a prevalence of 1:25,000
ease, and can occur in Ehlers–Danlos syndrome, to 1:100,000 [7]. It is caused by a mutation of the
Marfan syndrome, sickle cell hemoglobinopa- ABCC6 gene, a part of the subclass C of the ATP-
thies, and hypercalcemia. They have also been binding cassette transporter (ABC). The gene
described in an enlarging list of other diseases ABCC6 is highly expressed in kidney and liver,
including chronic congenital hyperphosphatemia with lower expression rates in skin, retina, and
and Sturge–Weber syndrome. A systemic associ- vessel walls, which are the clinically affected tis-
ation can be identified in at least 50% of patients, sues in PXE [7, 8].
of which PXE is the most frequent, but AS are Histologic findings of the skin in PXE classi-
also observed as an isolated ocular finding (idio- cally include elastin abnormalities in the mid-
pathic) [3]. epidermis with swelling, granular degeneration,
and fragmentation. The first detectable sign is
calcification of elastin fibers in electron micros-
C. Hänsli · S. A. Zweifel (*) copy [8]. Similar alterations are found in the BM
University Hospital Zurich, Zurich, Switzerland of the eye. Calcification and thickening make BM
University of Zurich, Zurich, Switzerland brittle, leading to breaks in BM, clinically appear-
e-mail: christof.haensli@usz.ch; ing as AS [8]. Later, ingrowth of fibrovascular
sandrine.zweifel@usz.ch

152 C. Hänsli and S. A. Zweifel
tissue from the choriocapillaris may occur, lead- elastic fibers. Thus, its deficiency may lead to
ing to CNV and development of a disciform scar BM calcification and BM cracks, apparent as AS
with subretinal fibrosis and atrophy, especially [10]. Secondary CNV is a possible complication
when CNV remains untreated [7, 8]. In small and causing visual loss in young patients [10].
medium-sized vasculature of the body, calcifica-
tion of elastic fibers seems to resemble normal
atherosclerosis [8]. 13.1.4 Hemoglobinopathies
Signaling cascades leading to these changes
are complex, and not yet fully understood. The Elastic tissue disorders, similar to PXE, with
substrate of ABCC6 and its role in the pathogen- ocular involvement are a frequent finding in
esis of PXE are largely unknown [9]. The patho- hemoglobinopathies. Angioid streaks have been
logic mechanism is possibly not due to a single reported in 1–22% of patients with sickle cell dis-
gene, but modulated by many genetic modifiers ease, depending on patients’ age, in 20% with
[9]. Main important pathways seem to involve β-thalassemia, and in 10% with sickle thalas-
transforming growth factor beta (TGF-β), bone semia [11]. The changes seem to be acquired as a
morphogenetic proteins (BMPs), and inorganic consequence of the underlying disease, with a
phosphates (PPi). All are known factors in the generally milder clinical course and later onset,
homeostasis of extracellular matrix, its mineral- compared to PXE [11].
ization, and vascular calcification [9].
First clinical signs of PXE are asymptomatic
small yellow papules on the skin, beginning at 13.1.5 Other Causes
the neck and shoulder in the adolescence [7].
Cardiovascular involvement of the disease can Other mechanisms have been proposed in the for-
lead to early-onset atherosclerosis, coronary or mation of AS or provoking secondary CNV in
valvular heart disease, and restrictive cardiomy- preexisting AS. For example, acceleration/decel-
opathy, with unknown prevalence but possibly eration trauma may cause hyperextension of the
resulting in death [7]. The systemic manifesta- eyeball in its equator line, leading to ruptures of
tions with possibly life-threatening vascular BM and/or the secondary formation of CNV in
complications and the vision-threatening ocular preexisting AS [12].
complications make the knowledge of the disease
important. Early recognition enables an appropri-
ate screening for systemic or ocular complica- 13.2 Clinical Features
tions and thus treatment in early stages leads to
better treatment outcomes. 13.2.1 Incidence and Visual Loss
Angioid streaks, independent of underlying dis-

13.1.3 Paget’s Disease ease, usually occur bilaterally and are equally
distributed in men and women [5]. They affect
Juvenile Paget’s disease is a rare recessively younger patients of average 43–57 years [5, 13].
inherited disorder, associated with accelerated Secondary CNV occurs in 67.5% of patients with
bone turnover, long bone deformity, growth AS, and is bilaterally in 31.2% of patients with
impairment, and fractures in early life. It is asso- AS. Prior to the anti-VEGF era, a decreased
ciated with progressive retinopathy characterized BCVA less than 20/200 was shown in 40.4% of
by the development of AS. Retinal abnormalities eyes with AS. In younger patients with AS with-
increase with age. The underlying mutation leads out CNV, BCVA can be normal [14]. Most (86%)
to a loss of function of osteoprotegerin. This gly- of the CNV lesions are macular or both macular
coprotein has an important function in regulating and extramacular, only 14% of lesions were
bone turnover and may inhibit calcification of extramacular [5].
13 Choroidal Neovascularization Associated with Angioid Streaks 153
In PXE, 94% of patients have associated AS, visual acuity <20/50 in at least one eye occurs in
with macular involvement in more than half of 20% of younger patients, but increases to 75% of
the eyes (Fig. 13.1). Sixty percent of patients patients older than 52 years. Visual impairment
have bilateral BCVA > 20/50 on presentation. with bilateral BCVA <20/50 occurs in 17% of
Visual loss increases with age. Best corrected patients with PXE.
a b c
e f
Fig. 13.1 Multimodal imaging findings of angioid laser ophthalmoscopy (cSLO) (b) with indication of cor-
streaks (AS). Multimodal imaging of the left eye of a responding OCT scan (green line) and hypoautofluores-
50-year-old patient with histologically proven pseudoxan- cent (c). The corresponding OCT image with its enlarged
thoma elasticum with AS without secondary choroidal section (d) demonstrates Bruch’s membrane break (white
neovascularization. BCVA was 20/30 in the left eye. arrow). Fluorescein angiography shows window defect
Angioid streaks emanating from the optic disc in the pos- with hyperfluorescence in the arteriovenous phase (e) and
terior pole appear reddish-brown in fundus photography late staining (f)
(a, white arrows), hyporeflective on confocal scanning
13.2.2 Ocular Imaging Findings show a higher prevalence of BM undulations,

suggesting a continuous disease progress from
Diagnosis and treatment monitoring of complica- BM undulations to BM breaks and later second-
tions associated with AS is based on multimodal ary CNV [19, 20]. Subretinal hyperreflective
imaging including color fundus photography, material above and beneath RPE/BM complex
near-infrared imaging (NIR), fundus autofluores- can indicate CNV. This sign of CNV often
cence (FAF), fluorescein-angiography (FA), occurs at locations of larger BM breaks, sug-
indocyanine green angiography (ICG), optical gesting the development of CNV at locations of
coherence tomography (OCT), and OCT angiog- BM breaks [19].
raphy (OCTA). Smaller defects in BM may be difficult to
detect in OCT B-scans or appear obscured by
13.2.2.1 Angioid Streaks complications such as hemorrhages. In these
Angioid streaks appear as linear crack-like lines cases, en-face reconstruction of BM layer from
in the posterior pole, emanating radially from the OCT may visualize breaks in BM which are not
optic disc with a diminishing width toward the visible in other imaging modalities [21].
periphery [15]. The color of AS is brown in 50%, Angioid streaks appear as hyperfluorescent
but can also appear gray or reddish-brown window defects in early frames with sharp bor-
(Fig. 13.1a) [5]. ders and without leakage in later frames in FA
Angioid streaks are best visualized using NIR (Figs. 13.1e and f) and ICG [15]. Typically they
or confocal scanning laser ophthalmoscopy seem to appear more prominent in ICG, even
(cSLO) imaging (Fig. 13.1b) [16]. They appear though few cases are better highlighted by other
as dark hyporeflective linear bands in the poste- modalities [15].
rior pole with sharp borders. The lesions corre- Optical coherence tomography angiography
spond to BM breaks in OCT (Fig. 13.1d). (OCTA) offers the possibility to visualize cho-
Fundus autofluorescence can reveal larger riocapillaris vessels underlying RPE in a very
areas of retinal pigment epithelium (RPE) dam- high resolution, not achievable with other
age than visible on clinical examination or color modalities. In OCTA, AS without CNV typi-
fundus photography [17, 18]. Angioid streaks cally shows underlying choriocapillaris rarefi-
appear as hypoautofluorescent lines, even though cation, corresponding to the dark areas in NIR
not as distinct as in NIR imaging (Fig. 13.1c) imaging and RPE/BM hyporeflectivity in
[16]. A hyperautofluorescent margin of the streak SD-OCT [22]. These findings are irrelevant of
can occur, called the parastreak phenomenon. It the underlying disease (PXE, sickle cell disease,
is thought to result from RPE cell proliferation or idiopathic). A minor part of AS, with a bright
and secondary lipofuscin accumulation. This and adjacent dark appearance in NIR imaging
parastreak phenomenon is more often observed and subretinal hyperreflective material in OCT,
in patients without CNV. The mechanism for this shows irregular choriocapillaris vascular net-
distribution remains unknown [17]. work, for which fibrovascular tissue remodeling
Spectral domain (SD-) OCT offers the pos- is hypothesized [22].
sibility to show retinal architecture in a near- Changes in the choriocapillaris vascular net-
histological resolution. Angioid streaks are work can be analyzed using flow voids in
visualized as breaks in BM (Fig. 13.1d). The OCTA. Choriocapillaris changes can be
overlying RPE can remain intact, be altered, detected in AS, even in the absence of RPE
focally detached, or completely absent [16]. atrophy. Besides remarkable structural differ-
Additional OCT findings are BM undulations ences in the appearance of the choriocapillaris
(Fig. 13.1d) and large dehiscences of BM [19]. layer, higher proportions of larger flow voids
Over time, BM undulations can progress to BM (areas greater than 10,000 or 40,000 μm2 with
breaks, especially at vortexes and margins of lack of flow signal) can be measured, compared
BM undulations [19]. Eyes with CNV due to AS to normal eyes [23].
13.2.2.2 eau d’Orange and Reticular

P bodies show round irregularly shaped hyperre-
Pseudodrusen flective borders with a hyporeflective space
Peau d’ orange is usually the first ocular sign, within interrupted retinal layers in SD-OCT [14,
preceding AS for 1–8 years [7]. It is a non- 30]. Underlying RPE is altered or atrophic, creat-
confluent yellowish granular finding located pre- ing a hypertransmission artifact in SD-OCT [14].
dominantly in the temporal macula, resembling
the skin of an orange: in French “peau d’orange” 13.2.2.4 Optic Disc Drusen
[24] (Fig. 13.3a). It is best visualized as small Optic disc drusen can be diagnosed in 8–25% of
dark flecks in NIR imaging, appearing as isore- patients with PXE; they are often only visualized
flective non-confluent spots in NIR imaging in an using FAF [7].
area of hyperreflectivity (Fig. 13.3b) [25], with
corresponding round hypocyanescent spots in 13.2.2.5 Pattern Dystrophy-Like
ICG [15]. Areas of peau d’orange show hyperre- Changes
flectivity of the outer band of the RPE/BM com- Multiple irregular hyperautofluorescent lines and
plex in SD-OCT (Fig. 13.3f) [14, 16, 26]. spots can cause various presentations of pattern
Posteriorly adjacent to peau d’orange, a yel- dystrophy-like appearance, visible in FAF [16,
low confluent area can be seen, with a reduced 28, 31, 32]. These changes seem to occur more
visualization of underlying choroidal vessels often in patients with CNV in at least one eye,
[26]. Angioid streaks are located within this area and may thus be a risk factor for the development
and stretching to its borders. It is thought to be of CNV [17]. Fundus pulverulentus is a rare sub-
the representation of calcium deposits in the BM form of pattern dystrophy. In patients with PXE
layer [16, 26]. Cracks of BM may propagate who show fundus pulverulentus, subretinal fibro-
within this hypercalcified and thus brittle layer, sis is more often present than without fundus pul-
just as in a shell of an egg. Hence, the French verulentus (34.6 vs. 4.2%, respectively). This
term “coquille d’œuf” has been proposed for this subretinal fibrosis often leads to vision loss due
phenomenon [26]. This area can be highlighted to macular atrophy and/or CNV [33].
by separating color channels of a fundus photo-
graph and multiplying red and green channel 13.2.2.6 Atrophy
with an image processing software. In atrophic areas, underlying choroidal vessels
Reticular pseudodrusen (subretinal drusenoid are visible in color fundus photography, but AS
deposits) [27] may look similar to peau d’orange under atrophic areas are less visible. Atrophy as a
in fundoscopy and color fundus photography. sign of additional RPE damage appears hypoau-
They are located more posteriorly than peau tofluorescent in FAF. In eyes with AS, atrophy
d’orange and appear as hyporeflective non- can appear as RPE rips, sharply demarcated mul-
confluent spots in NIR [25]. Optical coherence tilobular RPE atrophy, or poorly defined RPE
tomography shows multiple small subretinal atrophy [17, 18]. Most eyes with AS and atrophy
hyperreflective accumulations leading to undula- show small dot-like hyperautofluorescent pat-
tions of the outer retinal layers, appearing in 50% terns in the perilesional area (Fig. 13.5b). Atrophy
of PXE patients with AS [25, 28]. shows the loss of the ellipsoid and RPE bands
with choroidal hypertransmission and hyperre-
13.2.2.3 Crystalline Bodies flective spots within the outer retinal layers in
and Comet Lesions SD-OCT [34, 35]. Angioid streaks appear as
Crystalline bodies are small chorioretinal atro- breaks in BM in SD-OCT, which also shows
phies, appearing as nodular crystalline or white underlying choroidal thinning [34].
bodies, and can be associated with a white tail of Atrophy of the RPE is detectable in 32% of
RPE atrophy extending posteriorly, resembling a patients with PXE [34]. In 7% it occurs without
comet [29]. FA reveals early transmission and concomitant CNV, most commonly between 60
late staining hyperfluorescence [14]. Crystalline and 70 years of age. Being relatively rare below
40 years, the incidence of atrophy increases with 1 CNV can develop into type 2 CNV [38]. The
age to virtually all patients over 70 years. In most clinical course of eyes with AS and secondary
of the patients with AS and RPE atrophy, second- CNV shows progressive visual loss, compared to
ary CNV is adjacent or within the atrophic area. eyes without CNV. Vision loss is more pro-
Before the development of atrophy, eyes typi- nounced in patients with type 2 NV, and even
cally show pattern dystrophy-like changes, retic- more with fibrotic scars secondary to previous
ular pseudodrusen, and reduced subfoveal CNV [38].
choroidal thickness [34]. Typically, separate atro- Eyes with type 2 CNV also have more second-
phic areas develop individually and later coalesce, ary complications compared to type 1 CNV, such
sparing the fovea until advanced stages [34]. The as subretinal hemorrhage (75 vs. 25%, respec-
progression rate of RPE atrophy and its size tively) or cystoid retinal fluid (78 vs. 13%,
increasingly vary with age. In eyes with PXE and respectively) [38]. But severe visual loss due to
pattern dystrophy, progression is faster [34]. hemorrhage may also occur in type 1 CNV [40].
Visual deterioration depends on the precise loca- Thus, it may be a continuous disease progress of
tion [36]. CNV developing under AS before crossing the
Quantitative analysis of FAF is reduced in RPE, which increases morbidity.
patients with PXE, compared to normal eyes, Subretinal detachment or subretinal fluid may
especially in the nasal macula, with a more probe signs of CNV activity. However, subretinal
nounced effect in advanced disease [31]. fluid can occur in patients without evidence of
CNV in FA [28, 41]. In these cases, subretinal
13.2.2.7 Choroidal fluid is not responsive to treatment with anti-
Neovascularization VEGF. Additionally, subretinal fluid can appear
Secondary to Angioid at remote locations without connection to active
Streaks CNV, which also may not be responding to anti-
Choroidal neovascularizations in the era of VEGF treatment or cause visual deterioration
SD-OCT can be graded as type 1–3 neovascular- [28, 41].
izations (NV) based on their anatomic location, Hyperreflective foci are discrete, dot-shaped,
first described in age-related macular degenera- hyperreflective lesions solely identifiable on
tion (AMD). Type 1 NV was defined as the SD-OCT imaging (Fig. 13.4) [42]. Hyperreflective
growth of new vessels below the RPE, type 2 NV foci are more present in patients with active
as the growth of vessels in the subretinal space, CNV. They may be an early sign before the recur-
and type 3 NV within the retina. Polypoidal cho- rence of CNV activity under anti-VEGF treat-
roidal vasculopathy (PCV) is considered as a spe- ment in a pro re nata (PRN) regimen.
cific subform of type 1 NV in AMD [37]. Hyperreflective foci thus may possibly be used as
Choroidal neovascularization in AS can be of a marker for monitoring CNV activity [42].
types 1 or 2, or PCV [38]. In classic CNV (type 2 Subretinal detachments are less frequent in
NV), FA typically shows early well-defined hyper- CNV due to AS than in CNV due to AMD (17 vs.
fluorescence with dye leakage in later frames 59%) [20]. Intraretinal or subretinal fluid thus is
(Figs. 13.2a–e). Occult CNV (type 1 NV) may be not a necessary finding in secondary CNV [43].
hidden in FA but can be detected with early hyper- Outer retinal tubulations are tubular structures,
cyanescence and late leakage using ICG [15]. visible in SD-OCT, and first described in AMD
OCT can show intra- and subretinal fluid as a sign [44]. The incidence of outer retinal tubulations is
of CNV activity and/or subretinal hyperreflective higher in secondary CNV due to AS (71%), com-
material (Figs. 13.2e, 13.3, and 13.4). pared to CNV due to AMD (34%) [20]. Most
Polypoidal choroidal vasculopathy may (74%) of them appear at the border of a CNV
appear as a primary lesion [38] or may develop lesion. Submacular choroidal thickness and vol-
during the follow-up in association with other ume remain normal in patients with AS without
types of neovascularizations [39]. Over time type CNV. But they are increased in patients with
a c d
b e
Fig. 13.2 Choroidal neovascularization associated with pigment epithelial detachment on OCT (e, f arrowheads)
angioid streaks. Fundus photography (a), near-infrared associated with subretinal fluid (e, white arrow) corre-
imaging (b), middle (c) and late (d) frames of fluorescein sponding with the hyperfluorescent lesion on FA. Ten
angiography (FA) of the same patient as shown in years later under close monitoring and continuous anti-
Fig. 13.1 three years later after initial presentation, now VEGF therapy vision remained 20/20. In the OCT (g)
demonstrating a secondary choroidal neovascularization larger dehiscences in the Bruch’s membrane (between
(CNV). BCVA was 20/20, but the patient was symptom- white arrows), advanced CNV with fibrotic tissue, but no
atic and reported metamorphopsia. Note the fibrovascular subretinal fluid could be detected
s econdary CNV due to AS, independently of pre- tion, and without the need for dye injection. Due
vious treatment or treatment modality [45]. to its noninvasive properties and relatively small
recording time, OCTA can be used more fre-
Optical Coherence Tomography quently compared to FA. This provides the basis
Angiography for a potential use in regular follow-up and treat-
Optical coherence tomography angiography ment monitoring CNV.
offers the possibility to visualize retinal vessels Optical coherence tomography angiography is
including CNV membranes at very high resolu- able to detect the majority of, but not all, CNV
a b
f d
Fig. 13.3 Peau d’orange, angioid streaks, and CNV in a visit) showing macular thickening with intra- and subreti-
patient with pseudoxanthoma elasticum. Left eye of a nal fluid. Spectral-domain (SD)-OCT imaging 1 year later
52-year-old female with secondary choroidal neovascu- under treatment with four anti-VEGF injections with bev-
larization (CNV) due to angioid streaks (AS) with pseu- acizumab: note the fibrovascular pigment epithelial
doxanthoma elasticum (PXE). BCVA at baseline was detachment (e, white arrow), subretinal fluid and subreti-
20/200, she has been noticing a central scotoma since a nal hyperreflective material (e, white arrowhead).
few days. Fundus photography (a) showing macular hem- Hyperreflectivity of the outer retinal pigment epithelium/
orrhage associated with a neovascular membrane, AS and Bruch’s membrane complex appears in SD-OCT in the
peau d’orange (white arrow), fluorescein angiography areas of peau d’orange (e, small arrows). Near-infrared
showing (b) early well-defined hyperfluorescence and (c) image (f) indicating the location of B-scan and showing
late leakage, and OCT (d, time domain at the baseline peau d’orange (white arrow)
a c
Fig. 13.4 OCT angiography of secondary CNV due to choriocapillaris (a) there was evidence of a neovascular
idiopathic angioid streaks. Right eye of a 58-year-old network (white arrows), corresponding to flow on the
female patient with secondary choroidal neovasculariza- B-Scan (b, white arrow). In the structural OCT subretinal
tion (CNV) due to idiopathic angioid streaks (AS). hyperreflective material (SHRM), suggestive for active
Recurrency occurred 6 months after anti-VEGF therapy CNV could be observed (d, white arrow). Note that 2
was paused with the development of CNV in the macular months prior SHRM was not present (c). Outer retinal
area. At this time point the patient had still good visual tubulation overlying a mature neovascular complex can be
acuity (20/20, 91 ETDRS letters) but noticed metamor- observed suprapapillary (e, white arrow)
phopsia. In the en face OCTA slab of the outer retina and
proven by FA or ICG (Figs. 13.4, 13.5) [43]. On SD-OCT and FA. The interlacing pattern could
the other hand, OCTA may be able to detect and therefore itself be a possible sign of activity
monitor the growth of CNV without signs of [43]. In the tangled appearing CNV, no activity
leakage in FLA or fluid in OCT [40]. was described [22].
In patients after initiation of anti-VEGF treat- Of note, the mentioned classification is based
ment, the CNV membrane may appear as a “tan- on qualitative criteria and its reproducibility and
gled vascular network” also named “pruned applicability in a clinical setting and treatment
vascular tree” with a loose lace appearance with decisions remain to be proven.
filamentous vessels and few large branches or vas-
cular trunks. Alternatively, CNV can appear “inter-
lacing” with a densely ramified vascular 13.3 Treatment and Monitoring
hyperintensity with a convoluted or cobweb shape, Regimen
multiple and tortuous thin vessels, and sometimes
a perilesional halo. A mixed form is also described Treatment initiation is mainly based on FA and
[22, 43]. The majority of CNV show foveal or jux- ICG, but OCTA can add additional guidance in
tafoveal involvement in OCTA [22, 43]. treatment initiation and control, as mentioned
The interlacing or mixed network is more above. It may also be a suitable tool to exclude
often associated with CNV activity, and with active CNV in the case of possibly misleading
previous treatment in the last 6 months and findings in other imaging modalities, such as sub-
activity signs in multimodal imaging using retinal fluid. Due to the lack of larger studies, in
a b
d f
e g
h
Fig. 13.5 Long-term follow-up of CNV associated with OCTA slab of the outer retina and choriocapillaris (d),
angioid streaks. Same patient as shown in Fig. 13.3, 10 B-scan with flow overlay in red (e), confirms the neovas-
years later undergoing anti-VEGF therapy in both eyes cular network (white arrow). The OCT of the left eye (h)
with an average of six injections per year due to persistent 6 weeks after the last injection, shows subretinal hyperre-
activity. Six weeks after the last injection in the right eye, flective material and adjacent subretinal fluid (white
BCVA was 20/32. Fundus photography (a) and fundus arrow), but no definite sign of a neovascular network in
autofluorescence (b) of the right eye show extrafoveal the en face OCTA slab of the outer retina and choriocapil-
atrophy and pigment clumping and hyperautofluorescent laris (f), (g) B-scan with flow overlay (g), BCVA was
spots surrounding the atrophy and fibrovascular lesion 20/50. In the right eye, treatment was continued in an
(b). The OCT of the right eye (c) shows an advanced extended interval, in the left eye, findings remained stable
(mature) CNV without intra- or subretinal fluid, en face without treatment
borderline cases, an individual treatment decision men are similar to reported monthly treatment for
will be based on multimodal imaging interpreted 1 year [47, 48].
by experienced ophthalmologists. A cost-effectiveness analysis of the study pop-
ulation in the MINERVA trial proved PRN intra-
vitreal ranibizumab to be a highly cost-effective
13.3.1 Ranibizumab intervention in patients with secondary CNV due
to other causes than AMD and myopia [49] with
13.3.1.1 MINERVA Trial costs of 1363£ per quality-adjusted life year in a
The efficacy of intravitreal ranibizumab in the 2018 United Kingdom setting.
treatment of secondary CNV due to AS has been
shown in a phase 3 randomized controlled trial. A 13.3.1.2 Long-Term Treatment
significant and clinically relevant treatment effect Outcomes
after 2 months using a pro re nata (PRN) regimen, Long-term treatment outcomes exist with level 2
compared to sham in CNV because of an uncom- evidence from previous non-randomized real-life
mon cause could be shown [46]. In the subgroup trials.
of secondary CNV due to AS, after 2 months a The PIXEL study was a retro- and prospective
BCVA gain (in early treatment of diabetic retinop- non-randomized trial to evaluate the long-term
athy study letters) of +11.0 letters in the ranibi- efficacy of ranibizumab in a real-life setting with a
zumab arm versus –3.5 letters in the sham arm was follow-up time of up to 4 years in 29% of patients
found. After 2 months, all patients received open- (mean 39.2 months, SD ±27.3) [50]. In 98 eyes of
label ranibizumab according to PRN. 72 patients, mean BCVA remained stable from
In the overall study population, after 12 baseline to 4 years of follow-up [50]. The propor-
months with ranibizumab according to PRN, the tion of patients with BCVA stabilization within 15
change of BCVA was +11.0 letters compared to ETDRS letters from baseline was 77.3% after 2
baseline, with a mean of 5.8 injections. Gain of years and 52% after 4 years. Best corrected visual
15 ETDRS letters occurred in 48.7%, and loss of acuity loss of more than 15 ETDRS letters
more than 15 ETDRS in 2.7%. Central subfield occurred in 11.4% after 2 years, and 26.3% after 4
thickness as a marker of disease activity decreased years. The proportions of eyes with CNV dropped
by −77.0 μm after month 2, compared to +9.8 μm from 68.4% at baseline to 32.1% after four years,
in the sham group. After 12 months the effect and the proportion of CNV leakage from 39.8% to
remained stable in the original treatment group 14.3%, respectively [50]. The mean annual num-
with −102.7 μm [46]. Thus, ranibizumab is the ber of injections was 4.1 in the first and 2.7 in the
only medication with level 1 evidence for treat- second year [50]. Other levels 2 and 3 studies and
ment of secondary CNV due to AS, yet. case reports show similar long-term outcomes for
Functional and anatomic effects with a PRN regi- up to 6 years [51–53].
13.3.2 Bevacizumab as a second-line treatment option. To date, this

evidence remains level 3 [65, 66].
A series of retrospective or non-randomized
smaller studies or case series using bevaci-
zumab showed similar results. Best corrected 13.3.4 Historic Treatment Regimens
visual acuity stabilization within 2 ETDRS
lines or BCVA increase compared to baseline, 13.3.4.1 Laser Photocoagulation
were shown in 65.2–100% of eyes with a fol- and Transpupillary
low-up time of 12–100 months [20, 54–58]. Thermotherapy
Signs of disease activity or recurrence were fre- Thermal photocoagulation was aimed to coagu-
quent even after treatment. Eyes with early dis- late CNV to stop vessel hyperpermeability and
ease showed significantly better visual gain disease progression. It was historically the first
than with advanced disease [56]. An initial gain treatment for AMD-related CNV [6]. Results of
of mean BCVA could not always be maintained various reports in secondary CNV due to AS
over 24 or 36 months, but mean BCVA remained overall were not convincing, at best leading to
stable compared to the initial baseline [59]. In stabilization or slowing down of visual loss in
younger patients, CNV seems to behave more most cases [6]. Frequent retreatments were nec-
aggressively [60]. essary with a delay of the natural history of the
A retrospective analysis of 52 eyes of 39 disease being the best achievable outcome.
patients treated with both ranibizumab and beva- Additionally, laser photocoagulation leads to an
cizumab showed an overall deterioration in mean immediate drop in retinal function at the treated
BCVA of an average of 6.8 ETDRS letters with a area. Thus, it is frequently resulting in vision loss
mean follow-up of 33.8 months (SD 19.6) [61]. and central scotoma, due to the central localiza-
Within this group, the patients with subfoveal tion of CNV.
CNV had higher BCVA loss, compared to non- Transpupillary thermotherapy was tried
subfoveal, with a loss of 11 versus 3 ETDRS let- intending to obliterate CNV vessels, without a
ters per year [61]. Summing up, there is a level 2 positive or lasting effect [6].
evidence for bevacizumab in CNV due to AS.
13.3.4.2 Surgical Treatments
Subretinal CNV extraction and macular translo-
13.3.3 Aflibercept cation surgery have been reported to treat CNV
secondary to AS [6]. Short-term stabilization of
Intravitreal injections with aflibercept using a visual acuity was achieved, but recurrences and
PRN regimen has been reported as an effective complications were frequent.
therapy in treatment-naïve CNV due to AS in
case reports [42, 62]. In a case report treat and 13.3.4.3 Photodynamic Therapy
extend (T&E) regimen using aflibercept has been Photodynamic therapy (PDT) with verteporfin is
reported as being effective after 1 year with a more selectively destroying CNV with less col-
total of six injections [63]. lateral tissue damage and proved favorable over
Whether a PRN regimen is superior to T&E the natural history of the disease in AMD-related
remains an object of clinical reasoning based on CNV [6]. Most conducted studies showed slow-
the understanding of the disease. Most studies, ing down of disease progression and visual dete-
among the MINERVA trial, used PRN [64]. rioration. The effect was independent of standard
Aflibercept has shown functional and ana- 3-monthly or more frequent treatment protocols.
tomic improvement in treating refractory visual A meta-analysis from 2013 showed an overall
loss and CNV activity after previous bevaci- visual loss after PDT, compared to baseline [6].
zumab and ranibizumab treatment in CNV due to Lesion size also increased under PDT treatment
AS in case reports, possibly offering the potential in all studies which reported this outcome param-
eter. The same meta-analysis showed the benefit also be of help in cases of isolated subretinal fluid
of anti-VEGF agents compared to PDT prior to without concomitant CNV, and thus refractory to
the randomized controlled trial [6]. Combination treatment (Figs. 13.4, 13.5).
therapy of PDT and anti-VEGF did not show a
benefit over anti-VEGF alone.
13.5 Key Learning Points
13.4 Summary • Angioid streaks have a complex pathogenesis,

can be related to many clinical conditions, of
Angioid streaks are a rare ophthalmologic find- which PXE is the most frequent, but are also
ing. They are frequently associated with systemic observed as an isolated ocular finding
disease. Pseudoxanthoma elasticum is the most (idiopathic).
frequent association and, besides ocular morbid- • Deformation and ruptures of BM are associ-
ity, carries a cardiovascular risk. Secondary CNV ated with angioid streaks as shown by histol-
occur in a large proportion of patients with ogy and OCT.
AS. Untreated, they have an unfavorable course • The development of CNV is suggested to
with progredient vision loss in young patients. occur at locations of BM breaks. Secondary
Thus, regular examinations of patients with AS is CNV can lead to severe visual loss in rela-
recommended, yearly or biyearly under 40 years tively young patients.
of age, and twice a year thereafter [13]. • Type 1 and 2 CNV seem to be a continuous
The current mainstay of treatment of second- disease progress emanating at locations of
ary CNV due to AS are intravitreal anti-VEGF BM breaks.
agents using a PRN regimen with level 1 evi- • In active CNV treatment is recommended
dence for ranibizumab, level 2 evidence for beva- using intravitreal anti-VEGF agents in a pro re
cizumab, and level 3 evidence for aflibercept [6, nata regimen. A treat and extend regimen
46]. Treatment outcome is better at early stages might be a good alternative; however, there is
of disease. Together with the progressive natural only limited evidence supporting this.
course of the disease this prompts for regular • OCT angiography can provide additional
screening and early treatment of affected patients information in less active CNV or subretinal
with AS and secondary CNV, respectively. Using fluid without CNV. Classification of CNV
Anti-VEGF, an early BCVA gain of mean 2 activity using OCTA has been based on quali-
ETDRS lines after 2 months has been reported, tative criteria; however, its reproducibility
remaining stable for 12 months with a mean of remains to be proven.
5.8 injections [46]. Stabilization within 15 • Multimodal imaging is important in the diag-
ETDRS letters was retrospectively shown in nosis and treatment of CNV associated with
77.3% after 2 years and 52% after 4 years [50]. AS.
Multimodal imaging is necessary for diagno-
sis, treatment decision, and monitoring. AS are
best visualized with NIR. FAF additionally shows
atrophy and pattern dystrophy-like changes, References
which can be precursors of vision loss and
1. Doyne R. Choroidal and retinal changes. The results
CNV. Active CNVs demonstrate leakage in angi- of blows on the eyes. Trans Ophthalmol Soc UK.
ography and are often associated with hyperre- 1889;9:128.
flective foci, subretinal hyperreflective material, 2. Knapp H. On formation of dark angiod streaks as an
and intra- or subretinal fluid in OCT. Optical unusual metamorphosis of retinal hemorrhage. Arch
Ophthalmol (Chicago, Ill: 1960). 1892;21:289.
coherence tomography angiography offers addi- 3. Chatziralli I, Saitakis G, Dimitriou E, Chatzirallis
tional guidance in therapy, specifically in detect- A, Stoungioti S, Theodossiadis G, et al. ANGIOID
ing the growth of otherwise inactive CNV. It may streaks: a comprehensive review from pathophysi-
ology to treatment. Retina (Philadelphia, PA). 18. Sawa M, Ober MD, Freund KB, Spaide RF. Fundus
2019;39(1):1–11. autofluorescence in patients with pseudoxanthoma
4. Kumudhan D, Wallace EJ, Roxburgh ST. Angioid elasticum. Ophthalmology. 2006;113(5):814–20.e2.
streaks in identical twins. Br J Ophthalmol. 19. Marchese A, Parravano M, Rabiolo A, Carnevali
2004;88(6):837–8. A, Corbelli E, Cicinelli MV, et al. Optical coher-
5. Abujamra S, Negretto AD, Saraceno JJ, Oliveira TL, ence tomography analysis of evolution of Bruch’s
Gomes AM. Angioid streaks: fundoscopic analysis of membrane features in angioid streaks. Eye.
317 cases. Arq Bras Oftalmol. 2008;71(6):819–21. 2017;31(11):1600–5.
6. Gliem M, Finger RP, Fimmers R, Brinkmann 20. Ellabban AA, Hangai M, Yamashiro K, Nakagawa
CK, Holz FG, Charbel IP. Treatment of choroi- S, Tsujikawa A, Yoshimura N. Tomographic fundus
dal neovascularization due to angioid streaks: a features in pseudoxanthoma elasticum: comparison
comprehensive review. Retina (Philadelphia, PA). with neovascular age-related macular degeneration in
2013;33(7):1300–14. Japanese patients. Eye (Lond). 2012;26(8):1086–94.
7. Finger RP, Charbel Issa P, Ladewig MS, Gotting C, 21. Hanhart J, Greifner H, Rozenman Y. Locating and
Szliska C, Scholl HP, et al. Pseudoxanthoma elasti- characterizing angioid streaks with en face opti-
cum: genetics, clinical manifestations and therapeutic cal coherence tomography. Retin Cases Brief Rep.
approaches. Surv Ophthalmol. 2009;54(2):272–85. 2017;11(3):203–6.
8. Hu X, Plomp AS, van Soest S, Wijnholds J, de Jong 22. Corbelli E, Carnevali A, Marchese A, Cicinelli MV,
PT, Bergen AA. Pseudoxanthoma elasticum: a clini- Querques L, Sacconi R, et al. Optical coherence
cal, histopathological, and molecular update. Surv tomography angiography features of angioid streaks.
Ophthalmol. 2003;48(4):424–38. Retina (Philadelphia, PA). 2018;38(11):2128–36.
9. Van Gils M, Nollet L, Verly E, Deianova N, Vanakker 23. Spaide RF. Choriocapillaris signal voids in mater-
OM. Cellular signaling in pseudoxanthoma elasti- nally inherited diabetes and deafness and in pseu-
cum: an update. Cell Signal. 2019;55:119–29. doxanthoma elasticum. Retina (Philadelphia, PA).
10. Kerr NM, Cassinelli HR, DiMeglio LA, Tau C, 2017;37(11):2008–14.
Tuysuz B, Cundy T, et al. Ocular manifestations of 24. Smith JL, Gass JD, Justice J Jr. Fluorescein fundus
juvenile Paget disease. Arch Ophthalmol (Chicago, photography of angioid streaks. Br J Ophthalmol.
Ill: 1960). 2010;128(6):698–703. 1964;48:517–21.
11. Aessopos A, Farmakis D, Loukopoulos D. Elastic tis- 25. Gliem M, Hendig D, Finger RP, Holz FG, Charbel
sue abnormalities resembling pseudoxanthoma elasti- IP. Reticular pseudodrusen associated with a diseased
cum in beta thalassemia and the sickling syndromes. bruch membrane in pseudoxanthoma elasticum.
Blood. 2002;99(1):30–5. JAMA Ophthalmology. 2015;133(5):581–8.
12. Fajardo Sanchez J, Chau Ramos CE, Mazagatos 26. Spaide RF. Peau d’orange and angioid streaks: mani-
Used PJ, Aparicio Hernandez-Lastras MJ. The effect festations of Bruch membrane pathology. Retina
of the acceleration/deceleration trauma in angioid (Philadelphia, PA). 2015;35(3):392–7.
streaks: a pathogenic hypothesis. Archivos de la 27. Zweifel SA, Spaide RF, Curcio CA, Malek G, Imamura
Sociedad Espanola de Oftalmologia. 2016;91(9): Y. Reticular pseudodrusen are subretinal drusenoid
446–9. deposits. Ophthalmology. 2010;117(2):303–12.e1.
13. Orssaud C, Roche O, Dufier JL, Germain DP. Visual 28. Zweifel SA, Imamura Y, Freund KB, Spaide
impairment in pseudoxanthoma elasticum: a sur- RF. Multimodal fundus imaging of pseudox-
vey of 40 patients. Ophthalmic Genet. 2015;36(4): anthoma elasticum. Retina (Philadelphia, PA).
327–32. 2011;31(3):482–91.
14. Ari Yaylali S, Akcakaya AA, Erbil HH, Salar S, 29. Gass JD. “Comet” lesion: an ocular sign of pseu-
Karakurt Y. Optical coherence tomography findings doxanthoma elasticum. Retina (Philadelphia, PA).
in pseudoxanthoma elasticum. Eur J Ophthalmol. 2003;23(5):729–30.
2010;20(2):397–401. 30. Barteselli G, Viola F. Comet lesions in pseudoxan-
15. Arvas S, Akar S, Yolar M, Yetik H, Kizilkaya M, thoma elasticum: a spectral domain optical coher-
Ozkan S. Optical coherence tomography (OCT) and ence tomography analysis. Retina (Philadelphia, PA).
angiography in patients with angioid streaks. Eur J 2015;35(5):1051–3.
Ophthalmol. 2002;12(6):473–81. 31. Gliem M, Muller PL, Birtel J, McGuinness MB,
16. Charbel Issa P, Finger RP, Holz FG, Scholl Finger RP, Herrmann P, et al. Quantitative fundus
HP. Multimodal imaging including spectral domain autofluorescence in pseudoxanthoma elasticum.
OCT and confocal near infrared reflectance for Invest Ophthalmol Vis Sci. 2017;58(14):6159–65.
characterization of outer retinal pathology in pseu- 32. Agarwal A, Patel P, Adkins T, Gass JD. Spectrum
doxanthoma elasticum. Invest Ophthalmol Vis Sci. of pattern dystrophy in pseudoxanthoma elasticum.
2009;50(12):5913–8. Arch Ophthalmol (Chicago, Ill: 1960). 2005;123(7):
17. Finger RP, Charbel Issa P, Ladewig M, Gotting C, 923–8.
Holz FG, Scholl HP. Fundus autofluorescence in 33. Ebran JM, Martin L, Leftheriotis, Navasiolava N,
pseudoxanthoma elasticum. Retina (Philadelphia, Ferre M, Milea D, et al. Subretinal fibrosis is asso-
PA). 2009;29(10):1496–505. ciated with fundus pulverulentus in pseudoxanthoma
elasticum. Graefes Arch Clin Exp Ophthalmol. results of the MINERVA study. Retina (Philadelphia,
2018;256(4):699–707. PA). 2018;38(8):1464–77.
34. Gliem M, Muller PL, Birtel J, Hendig D, Holz FG, 47. Finger RP, Charbel Issa P, Hendig D, Scholl HP, Holz
Charbel IP. Frequency, phenotypic characteristics and FG. Monthly ranibizumab for choroidal neovascular-
progression of atrophy associated with a diseased izations secondary to angioid streaks in pseudoxan-
Bruch’s membrane in pseudoxanthoma elasticum. thoma elasticum: a one-year prospective study. Am J
Invest Ophthalmol Vis Sci. 2016;57(7):3323–30. Ophthalmol. 2011;152(4):695–703.
35. Sadda SR, Guymer R, Holz FG, Schmitz-Valckenberg 48. Heier JS, Brown D, Ciulla T, Abraham P, Bankert
S, Curcio CA, Bird AC, et al. Consensus definition for JM, Chong S, et al. Ranibizumab for choroidal neo-
atrophy associated with age-related macular degen- vascularization secondary to causes other than age-
eration on OCT: classification of atrophy report 3. related macular degeneration: a phase I clinical trial.
Ophthalmology. 2018;125(4):537–48. Ophthalmology. 2011;118(1):111–8.
36. Schoenberger SD, Agarwal A. Geographic chorio- 49. McCarthy G, Fenu E, Bennett N, Almond
retinal atrophy in pseudoxanthoma elasticum. Am J C. Intravitreal Ranibizumab for the treatment of
Ophthalmol. 2013;156(4):715–23. visual impairment due to choroidal neovasculariza-
37. Freund KB, Zweifel SA, Engelbert M. Do we need tion associated with rare diseases: cost-effectiveness
a new classification for choroidal neovasculariza- in the UK. Adv Ther. 2019;36(3):632–44.
tion in age-related macular degeneration? Retina 50. Mimoun G, Ebran JM, Grenet T, Donati A, Cohen
(Philadelphia, PA). 2010;30(9):1333–49. SY, Ponthieux A. Ranibizumab for choroidal neo-
38. Nakagawa S, Yamashiro K, Tsujikawa A, Otani A, vascularization secondary to pseudoxanthoma elasti-
Tamura H, Ooto S, et al. The time course changes cum: 4-year results from the PIXEL study in France.
of choroidal neovascularization in angioid streaks. Graefes Arch Clin Exp Ophthalmol. 2017;255(8):
Retina (Philadelphia, PA). 2013;33(4):825–33. 1651–60.
39. Baillif-Gostoli S, Quaranta-El Maftouhi M, Mauget- 51. Tilleul J, Mimoun G, Querques G, Puche N, Zerbib
Faysse M. Polypoidal choroidal vasculopathy in a J, Lalloum F, et al. Intravitreal ranibizumab for
patient with angioid streaks secondary to pseudoxan- choroidal neovascularization in angioid streaks:
thoma elasticum. Graefes Arch Clin Exp Ophthalmol. four-year follow-up. Retina (Philadelphia, PA).
2010;248(12):1845–8. 2016;36(3):483–91.
40. Andreanos KD, Rotsos T, Koutsandrea C, Kymionis 52. Zebardast N, Adelman RA. Intravitreal ranibizumab
GD, Georgalas I, Ladas ID. Detection of nonexu- for treatment of choroidal neovascularization sec-
dative choroidal neovascularization secondary to ondary to angioid streaks in pseudoxanthoma elas-
angioid streaks using optical coherence tomography ticum: five-year follow-up. Semin Ophthalmol.
angiography. Eur J Ophthalmol. 2017;27(5):e140–e3. 2012;27(3-4):61–4.
41. Hansen MS, Klefter ON, Larsen M. Retinal degen- 53. Savastano MC, Minnella AM, Zinzanella G, Falsini
eration and persistent serous detachment in the B, Caporossi A. Successful long-term management
absence of active choroidal neovascularization of choroidal neovascularization secondary to angioid
in pseudoxanthoma elasticum. Acta Ophthalmol. streaks in a patient with pseudoxanthoma elasticum: a
2014;92(2):e156–7. case report. J Med Case Rep. 2014;8:458.
42. Parodi MB, Arrigo A, Romano F, Aragona E, 54. Lekha T, Prasad HN, Sarwate RN, Patel M,
Marchese A, Cicinelli MV, et al. Hyperreflective foci Karthikeyan S. Intravitreal bevacizumab for choroidal
number correlates with choroidal neovascularization neovascularization associated with Angioid streaks:
activity in angioid streaks. Invest Ophthalmol Vis Sci. long-term results. Middle East Afr J Ophthalmol.
2018;59(8):3314–9. 2017;24(3):136–42.
43. Chapron T, Mimoun G, Miere A, Srour M, El Ameen 55. El Matri L, Kort F, Bouraoui R, Karim B, Chebil A,
A, Semoun O, et al. Optical coherence tomography Chaker N. Intravitreal bevacizumab for the treatment
angiography features of choroidal neovascularization of choroidal neovascularization secondary to angi-
secondary to angioid streaks. Eye. 2018;33:385–91. oid streaks: one year of follow-up. Acta Ophthalmol.
44. Zweifel SA, Engelbert M, Laud K, Margolis R, Spaide 2011;89(7):641–6.
RF, Freund KB. Outer retinal tubulation: a novel opti- 56. Finger RP, Charbel Issa P, Schmitz-Valckenberg
cal coherence tomography finding. Arch Ophthalmol S, Holz FG, Scholl HN. Long-term effectiveness
(Chicago, Ill: 1960). 2009;127(12):1596–602. of intravitreal bevacizumab for choroidal neovas-
45. Ellabban AA, Tsujikawa A, Matsumoto A, Ogino K, cularization secondary to angioid streaks in pseu-
Hangai M, Ooto S, et al. Macular choroidal thickness doxanthoma elasticum. Retina (Philadelphia, PA).
and volume in eyes with angioid streaks measured by 2011;31(7):1268–78.
swept source optical coherence tomography. Am J 57. Mimoun G, Tilleul J, Leys A, Coscas G, Soubrane
Ophthalmol. 2012;153(6):1133–43.e1. G, Souied EH. Intravitreal ranibizumab for cho-
46. Lai TYY, Staurenghi G, Lanzetta P, Holz FG, Melissa roidal neovascularization in angioid streaks. Am J
Liew SH, Desset-Brethes S, et al. Efficacy and safety Ophthalmol. 2010;150(5):692–700.e1.
of ranibizumab for the treatment of choroidal neovas- 58. Sawa M, Gomi F, Tsujikawa M, Sakaguchi H, Tano
cularization due to uncommon cause: twelve-month Y. Long-term results of intravitreal bevacizumab injec-
tion for choroidal neovascularization secondary to 62. Vaz-Pereira S, Collaco L, De Salvo G, van Zeller
angioid streaks. Am J Ophthalmol. 2009;148(4):584– P. Intravitreal aflibercept for choroidal neovas-
90.e2. cularisation in angioid streaks. Eye (Lond).
59. Iacono P, Battaglia Parodi M, La Spina C, Bandello 2015;29(9):1236–8.
F. Intravitreal bevacizumab for nonsubfoveal cho- 63. Diago T. Treat and extend regimen with aflibercept
roidal neovascularization associated with angioid for choroidal neovascularization in angioid streaks.
streaks: 3-year follow-up study. Am J Ophthalmol. Eye (Lond). 2016;30(4):637–9.
2016;165:174–8. 64. Vaz-Pereira S, Collaco L, De Salvo G, van Zeller
60. Alagoz C, Alagoz N, Ozkaya A, Celik U, Turan P. Intravitreal aflibercept for choroidal neovasculari-
MF, Yazici AT, et al. Intravitreal bevacizumab in sation in angioid streaks. Eye (Lond). 2016;30(4):639.
the treatment of choroidal neovascular membrane 65. Tetikoglu M, Sagdik HM, Aktas S, Ozcura
due to angioid streaks. Retina (Philadelphia, PA). F. Intravitreal aflibercept for refractory choroidal
2015;35(10):2001–10. neovascularization secondary to angioid streaks. Eye
61. Giacomelli G, Finocchio L, Biagini I, Sodi A, Murro (Lond). 2016;30(6):894–5.
V, Introini U, et al. Long-term follow-up of cho- 66. Esen E, Sizmaz S, Demircan N. Intravitreal afliber-
roidal neovascularization due to angioid streaks cept for management of subfoveal choroidal neovas-
with pro re nata intravitreal anti-VEGF treatment. cularization secondary to angioid streaks. Indian J
Ophthalmologica. 2017;238(1-2):44–51. Ophthalmol. 2015;63(7):616–8.
Idiopathic Choroidal
Faisal A. Almarek and Sulaiman M. Alsulaiman
14.1 Introduction tion [5]. Because it affects younger patients in

their working age, a significant impact on visual
Choroidal neovascularization (CNV) per se is a function over a longer life span may ensue.
description of a pathological entity rather than a
diagnostic term, when it is attributed to an ocular
or systemic cause such as age-related macular 14.2 Pathogenesis
degeneration (AMD), degenerative myopia,
inflammatory/infectious causes, pseudoxan- The pathogenesis of CNV is speculated to be a
thoma elasticum, etc., a diagnosis is reached. nonspecific healing mechanism to a specific
More than 30 ocular conditions have been associ- stimulus that is the underlying ocular or systemic
ated with choroidal neovascularization [1]. disease. It is characterized by new vessel forma-
Formerly named hemorrhagic macular cho- tion that passes through stages of activity and
roidopathy, idiopathic focal subretinal neovascu- involution [6]. Gass has classified the neovascu-
larization, or focal macular choroidopathy, lar formation based on biomicroscopic and histo-
idiopathic CNV is considered a diagnosis of pathologic findings from surgical specimens to
exclusion in which no primary ocular or systemic type 1 lesion, that is under the retinal pigment
disease is identified. It has been described to epithelium (RPE), which is usually found in
occur in healthy patients under the age of 50 AMD, or type 2 lesion located under the retina,
years [2–4]. The majority of the ophthalmic that is usually associated with presumed ocular
research on choroidal neovascularization is histoplasmosis (POHS) [7]. Histopathological
focused on age-related macular degeneration, studies were conducted on surgically excised
which is considered the most common cause of subfoveal idiopathic CNV and were compared to
CNV. In comparison, studies on idiopathic CNV other specimens in patients with CNV secondary
are scarce. This could be attributed to the rela- to AMD and POHS. The microscopic examina-
tively limited number of patients with this condition identified the presence of RPE cells, vascular
endothelial cells, photoreceptors, macrophages,
F. A. Almarek myofibroblasts, glial cells, and erythrocytes. The
College of medicine, Imam Mohammad bin Saud histopathological findings in idiopathic CNV
Islamic University, Riyadh, Saudi Arabia were similar to those of AMD and POHS with the
S. M. Alsulaiman (*) exception of basal laminar linear deposits found
Vitreoretinal Division, King Khaled Eye Specialist in membranes from patients with AMD [8].
Hospital, Riyadh, Saudi Arabia
e-mail: ssulaiman@kkesh.med.sa

168 F. A. Almarek and S. M. Alsulaiman
Although histopathological studies did not associated with increased SCT. Thicker subfo-
show a significant difference among CNV sam- veal choroid may be related to choroidal hyper-
ples from different etiologies, Ho et al. [2] pro- permeability, which is associated with the
posed that subfoveal idiopathic CNV could be development of idiopathic CNV, and SCT
focal in nature and these membranes are envel- changes may reflect disease activity [11].
oped with a bilayer of RPE similar to those seen Although no systemic diseases are usually
in POHS type 2 CNV membranes that result in identified in idiopathic CNV, one study by
smaller lesions and greater preservation of the Sasahara and colleagues [12] looked into the
photoreceptors, and hence a more favorable out- circulating hematopoietic stem cells (HSCs) in
come compared to CNV secondary to AMD. patients with idiopathic CNV, which was com-
Idiopathic CNV has the tendency toward pared to a control group with no CNV or sys-
affecting the posterior pole rather than the periph- temic disease. HSCs play a role in normal and
eral retina. The cause of this remains speculative. pathological postnatal angiogenesis. These
Bottoni and associates [4] proposed that the cells are derived from bone marrow and can dif-
nature of the choroidal vasculature in the macular ferentiate into various cells, which are thought
area plays a role. As a watershed area of end arte- to play an important role in tissue injury repair.
rial supply (branches of short posterior arteries) In their study, HSCs were collected from
to this area has a temporally diminished blood peripheral blood, followed by an evaluation of
supply which secondarily affects the Bruch’s the functional activity of these cells between
membrane, RPE and causes neovascularization. the groups. They found that some of the func-
Although no abnormality in the lobular choroidal tional activities were reduced in patients with
filling of the macular area was established in their idiopathic CNV especially in patients with pro-
study. gressing lesions. Based on these findings, they
Lee and colleagues [9] further studied the top- concluded that the functional activity of circu-
ographical relation between choroidal watershed lating HSCs may be a novel biomarker for idio-
zones (CWZs) and idiopathic CNV. Based on pathic CNV.
indocyanine green angiography (ICGA) findings Choroidal inflammation is a known cause of
of previous studies by Hayashi and de Laey [10], CNV. Many uveitic entities such as Vogt–
CWZs were classified into four categories: I, Koyanagi–Harada disease, serpiginous choroidi-
nasal filling; II-A, temporal filling; II-B, tempo- tis, white dot syndromes, and others could be
ral filling with delayed fluorescence in the foveal complicated by choroidal neovascular membrane
area; III, upper or lower temporal quadrant fill- formation. In idiopathic CNV, no apparent signs
ing; and IV, scattered dye filling. They found that of inflammation are detectable by clinical exami-
CWZs involving the fovea were seen in more nation but some investigators proposed that sub-
patients with idiopathic CNV than in the control clinical inflammation might play a role, whether
group, suggesting that CWZs were related to it is localized to the choroid [13] or systemic.
idiopathic CNV topographically, and CWZ clas- Yang et al. [14] reported in their study of serum
sification was a predictor of the functional out- inflammatory markers in patients with idiopathic
comes. This indicates that choroidal perfusion CNV that serum VEGF was significantly higher
could play a role in the development of CNV in in patients with idiopathic CNV compared to
young to middle-age patients. healthy subjects. This might lead to an increased
Subfoveal choroidal thickness (SCT) in VEGF in the macula through circulation given
patients with unilateral idiopathic CNV was the nature of rich blood supply by the fenestrated
compared to the fellow normal eyes and was choriocapillaris. This finding was consistent with
found to be significantly thicker. Furthermore, studies on patients with CNV secondary to AMD
treatment with intravitreal anti-vascular endothe- [15, 16]. On the other hand, local (intraocular)
lial growth factor (VEGF) led to a decrease in the inflammatory markers are believed to be contrib-
SCT, and reactivation of idiopathic CNV was uting to the pathogenesis of CNV in AMD [17–
14 Idiopathic Choroidal Neovascularization 169
20]. In eyes with idiopathic CNV, multiple studies 14.3 linical Features, Multimodal
C
looked at the inflammatory markers of samples Imaging, and Differential
obtained from the aqueous humor. Yin et al. [21] Diagnosis
studied the level of multiple inflammatory mark-
ers in patients with idiopathic CNV including 27 As mentioned earlier, idiopathic CNV usually
cytokines. Compared to a control group, there affects the posterior pole. The proposed theory
was a significantly higher level of IL-2, IL-10, could be related to the nature of the choroidal cir-
IL-15, IL-17, basic FGF, and GM-CSF. After culation in the macular area [4, 10]. Fundus
adjusting for the axial length between the groups, examination often reveals a grayish membrane in
lL-17 was the only significantly higher cytokine the macula with or without sub-macular hemor-
in the idiopathic CNV group. IL-17 is an inflam- rhage. It is usually unilateral but bilateral cases
matory cytokine that plays a role in multiple have been reported (Fig. 14.1) [3, 22, 23].
autoimmune and inflammatory diseases. It is also Fluorescein angiography (FA) and ICGA are
known to promote the angiogenesis effect of commonly used for diagnosing CNV of various
VEGF. etiologies. FA usually shows an early phase
Red-free
Fig. 14.1 A 36-year old female presented with diminution showed a type 2 CNV with subretinal fluid. The optical
of vision since 2 months in her left eye. Clinical examina- coherence tomography angiography shows the presence of
tion revealed the presence of choroidal neovascular mem- neovascular network in the outer retina and choriocapillaris
brane (CNV) with subretinal heamorrhage and subretinal segments (bottom). A diagnosis of idiopathic CNV was
fluid. Late venous phase of the fluorescein angiogram (top made based on exclusion and the patient received three
left) and early phase of the indocyanine green angiogram(top monthly anti-VEGF injections following which her vision
right) showed a well defined classic CNV. Optical coher- improved significantly. Courtesy Dr. Vishal Gowindhari,
ence tomography (middle) passing through the lesion LV Prasad Eye Institute, Bhubhaneswar, India
Angiography (Superficial) Angiography (Deep) Angiography (Outer retina) Angiography (Choriocapillaris)
ILM + 2.6 µm ~ IPL/INL + 15.6 µm IPL/INL + 15.6 µm ~ IPL/INL + 70.2 µm IPL/INL + 70.2 µm ~ BM + 0.0 µm BM + 0.0 µm ~ BM + 10.4 µm
hyperfluorescence that increases with time and modality. The use of OCTA in clinical settings
progresses to leakage in the late phase. Given that has been increasing in the last few years. Few
idiopathic CNV is usually a type 2 CNV and is studies evaluated OCTA in idiopathic CNV [24,
expected to behave as a classic CNV on FA. ICGA 26–28]. An irregular close-knit formation of
is used if the diagnosis is not certain after FA and intertwined and convoluted vessels was
optical coherence tomography (OCT). It is of observed at the level of the outer retina that is
higher value in cases of occult CNV, polypoidal normally supposed to be avascular so any flow
choroidal vasculopathy, retinal angiomatous pro- at that layer could be interpreted as a neovascu-
liferation, or in cases associated with RPE larization. Factors that might affect the quality
detachment. Kumar et al. [24] reported ICGA of the OCTA images such as pigment epithelial
findings of seven eyes with type 2 idiopathic detachment, high refractive error, posterior
CNV. ICGA showed a patch of diffuse hyperfluo- staphyloma, or poor fixation are usually not
rescence in three of seven eyes, a network of encountered in patients with idiopathic CNV,
abnormal vessels in four of seven eyes, and all which makes it a useful tool in evaluating these
eyes showed a hypofluorescent halo around the lesions [24].
CNV. The ring of hypofluorescence was also Chen et al. [27] studied the OCTA features of
observed in all of the 12 eyes with idiopathic 17 eyes with idiopathic CNV and monitored the
CNV reported by Toju et al. [25]. response after treatment with anti-VEGF. They
OCT is now the gold standard in diagnosing found that the most common morphological
and monitoring treatment response in structure in idiopathic CNV was “tree-in-bud”
CNV. Subretinal hyper-reflectivity is usually morphology that differs from AMD-related CNV,
observed that might be associated with other in which a large central trunk of “sea fan” vessels
findings such as subretinal or intraretinal hypore- is commonly seen. The authors speculated that
flective signal corresponding to intraretinal or the difference in the morphological structures
subretinal fluid. might play a role in the prognosis and response to
Optical coherence tomography angiography treatment. Interestingly, these membranes were
(OCTA) is a noninvasive dye-less imaging monitored at day 1 post treatment with intravit-
real anti-VEGF (ranibizumab) with OCT and include punctate inner choroidopathy, central
OCTA images. Selected areas of flow in the CNV serous chorioretinopathy, multifocal choroidi-
were significantly smaller on OCTA but no tis, subtle angioid streaks, and buried optic
change in fluid with B-scan OCT images was nerve head drusens. Laser-induced CNV has
observed. Based on this finding, OCTA could be been reported with the use of handheld laser
a more sensitive tool in predicting outcome pointers and cosmetic laser devices [32, 33].
instead of relying on OCT and visual acuity The history of exposure to handheld laser
solely. devices in young children is usually difficult to
Near-infrared autofluorescence (IR-AF) elicit and the diagnosis may be significantly
has been investigated as an imaging tool in delayed [34].
CNV. Keilhauer et al. [29] proposed that
images of IR-AF are derived from the melanin
in RPE and choroid. IR-AF features in 12 eyes 14.4 Natural History
with idiopathic CNV were studied by Toju and and Prognosis
colleagues [25]. In six of the studied eyes, a
well-defined ring of hyperautofluorescence Data on the natural history of CNV is derived
surrounding the hypoautofluorescent CNV from the era prior to anti-VEGF for lesions that
lesion at baseline was noted before treatment are mainly located sub-foveally in which laser
with bevacizumab. After treatment, all eyes photocoagulations was not applied based on the
showed a ring of hyperautofluorescence that criteria of the macular photocoagulation study
became more prominent as the lesions [30]. Eyes with idiopathic CNV that were
regressed which was confirmed by the absence assigned to observation had three times the risk
of leakage on FA. Three of the treated eyes of severe vision loss compared to the treated eyes
showed signs of recurrence of CNV activity after one year of follow-up. Lesions that spared
that were observed by leakage on FA and the center of foveal avascular zone had pro-
accumulation of subretinal fluid on OCT. The gressed during the one-year follow-up to involve
hyperautofluorescent ring became partially the center in 33% of untreated eyes. Ho et al. [2]
obscured and became clearly visible after retrospectively studied the natural history of 19
retreatment. The hyperautofluorescent ring patients with unilateral idiopathic CNV located
corresponded with a hypofluorescent halo or in the center of the foveal avascular zone, with a
dark rim on ICGA during the regression of the follow-up period ranging from 5 to 230 months.
lesions. It has been proposed that this well- They found that patients with idiopathic CNV
defined hyperautofluorescent ring represents had a significantly better long-term visual prog-
the melanin in the proliferating RPE cells at nosis compared to patients with POHS or AMD
the outer margin of the CNV lesion as the with subfoveal CNV. Up to 90% of patients with
lesions involute which also appear as a hypo- idiopathic subfoveal CNV had improved or main-
fluorescent rim on ICGA. tained vision in the follow-up period. This was
Certain entities with subtle or atypical clini- not the same for patients with POHS or AMD
cal findings or prior to the appearance of the full subfoveal CNV, where vision was maintained or
clinical picture may be mistakenly labeled as improved in 29 and 15%, respectively. The size
idiopathic CNV. Some eyes enrolled in the mac- of CNV was the only parameter associated with
ular photocoagulation study for the idiopathic the final visual acuity, lesions less than or equal
CNV arm [30, 31] manifested the clinical pic- to 1 disc diameter in size were associated with
ture of senile macular degeneration with more better final visual acuity of 20/60 or better. It is
than minimal drusen or atrophic histoplasmosis speculated that favorable prognosis of idiopathic
scar during the first year of follow-up. Other CNV compared to AMD might be related to the
conditions that may be complicated by CNV age of the affected patients; younger patients
and could be overlooked in clinical examination have a more active RPE that is more successful in
enveloping the CNV and promoting auto- based on the recommendation of the data review-
involution of the lesion. Another reason could be ing committee, as it was unlikely that a sufficient
related to the nature of the subretinal neovascular number of subjects would be recruited during the
membrane which is focal in nature in idiopathic expected study lifetime. However, accumulated
CNV rather than the diffuse form, which is seen data showed a pattern of treatment benefit in idio-
with AMD, thus affecting a smaller area and pathic CNV similar to that seen in AMD and
allowing preservation of the photoreceptors. POHS. Treatment complications reported by
In another study, Lindblom et al. [3] con- MPS study included hemorrhage, perforation of
cluded that idiopathic CNV carries a more favor- Bruch’s membrane, retinal hole, too extensive
able prognosis compared to AMD and treatment treatment, extension of laser scar (runoff) to the
recommendations, at that time, with laser photo- center of foveal avascular zone, and wrinkling of
coagulation (prior to the availability of the internal limiting membrane.
anti-
VEGF) should be reconsidered given the
possible complications of laser treatment includ-
ing inadvertent foveal burn, RPE tear, or creeping 14.5.2 Photodynamic Therapy
laser scars.
Although the natural history of idiopathic Photodynamic therapy (PDT) has the ability to
CNV is more favorable compared to AMD and selectively occlude the abnormal CNV with min-
POHS, individual variations still exist and imal effect on the overlying retina. A diode laser
cases with irreversible visual loss have been with low energy is used to activate the verteporfin
observed [2]. dye and initiate a photochemical reaction that
leads to occlusion of the abnormal CNV. PDT for
CNV has been proven effective in patients with
14.5 Treatment AMD, pathological myopia, and POHS [35–37].
Its efficacy in preventing moderate to severe
14.5.1 Laser Photocoagulation vision loss in predominantly classic CNV sec-
ondary to AMD has been established [35].
Laser photocoagulation to macular CNV is con- Idiopathic CNV usually is a classic CNV angio-
sidered a historical treatment after the introduc- graphically and PDT may be effective in occlud-
tion of less destructive treatment modalities such ing CNV as in predominantly classic CNV seen
as photodynamic therapy and anti-VEGF therapy. in AMD. Chan and associates [23] prospectively
Macular photocoagulation study (MPS) was a studied 17 eyes with idiopathic CNV. Eyes were
multicenter randomized clinical trial that evalu- treated with PDT for subfoveal and juxtafoveal
ated the effect of argon laser [30] and krypton lesions. There was a mean of 2.3 lines improve-
laser [31] photocoagulation in the treatment of ment in best-corrected visual acuity (BCVA) at
CNV associated with AMD, POHS, or idiopathic the end of 12 months follow-up with most gain
CNV versus observation in preventing or delay- noticed in the first 3 months. At 12-month visit,
ing severe vision loss defined by the study as six 41% had a BCVA of 20/40 or better, and 82% had
or more lines decrease in visual acuity. Lesions a BCVA of 20/100 or better. These results were
eligible for treatment based on study criteria better compared to the natural history in the
included lesions located 200–2500 μm from the observational study reported by Ho et al. [2]. The
center of the foveal avascular zone for argon laser need for retreatment was assessed every 3
trial; lesions within 1–199 μm for the krypton months, 47% of eyes needed retreatment with a
laser trial were treated with laser photocoagula- mean number of 1.8 treatments per eye in 1 year,
tion. After 1 year, up to three times of untreated which is less than 3.4 treatments for AMD and
eyes compared to treated eyes had lost six or pathological myopia, and 2.9 treatments for
more lines of visual acuity. Further randomiza- POHS. Factors associated with better final visual
tion of patients with idiopathic CNV was halted outcome were: active lesions on presentation,
smaller lesion size, and lesions that underwent either PDT alone or PDT after a course of sys-
only one PDT treatment. Reported complications temic steroids. Steroid regimen was given as
included blurry of vision in one case and submac- 1-gram methylprednisolone intravenously for 3
ular hemorrhage in three cases. No serious sys- days followed by oral prednisolone of 1 mg/kg
temic complications were encountered. per day and tapered according to individual basis.
Kang and colleagues [22] retrospectively The group with combination therapy had a better
studied the long-term effect of PDT on 30 eyes reduction of CNV size, better final BCVA, and
with idiopathic CNV. At 5-year follow-up period, less number of PDT treatments. Hyperplastic
46.7% gained vision, whereas 20% lost vision fibrotic scars were observed in the PDT only
when compared to baseline. Comparing this to treatment group. This suggests that steroids have
the natural history reported by Ho et al. [2] where a suppressing effect on fibroblast activity.
only 5% of eyes showed significant vision loss,
long-term effect of PDT for idiopathic CNV
appears to be limited. Eyes with juxtafoveal 14.5.3 Anti-vascular Endothelial
lesions lost more gain compared to subfoveal Growth Factors
ones despite better visual acuity at baseline. The
author proposed that the harmful effect of PDT With the introduction of anti-VEGF for the treat-
on foveal center and RPE cells could explain the ment of CNV, clinical practice has changed dra-
limited long-term outcome especially in juxtafo- matically. It is considered the mainstay treatment
veal CNV lesions. of CNV with a proven benefit observed in many
Inflammation of the choroid has been pro- studies, the majority of these looked into AMD-
posed as an underlying cause of idiopathic CNV related CNV. Idiopathic CNV has been studied to
[13]. Steroids are frequently used to control non- a lesser extent. The main outcome of treatment at
infectious uveitis and choroidal inflammation. the time of laser photocoagulation was to prevent
Combining PDT with systemic steroids in the or delay severe visual loss as defined by the MPS
treatment of idiopathic CNV was studied by [30] while with anti-VEGF, a significant improve-
Giovannini and associates [38]. Two groups of 10 ment in visual acuity is observed compared to
eyes each with idiopathic CNV were assigned to previous treatment modalities (Fig. 14.2).
Fig. 14.2 Fundus photograph showing a choroidal neo- Following treatment (lower images), resolution of intra-
vascular membrane (CNV) involving the fovea. Optical retinal and subretinal fluid was observed with restoration
Coherence Tomography (OCT) through the lesion of the foveal contour, although a pigment epithelial
revealed a type 2 CNV with subretinal and intraretinal detachment persisted. OCTA showed a less prominent
fluid. Optical Coherence Tomography Angiography vascular network. Courtesy Dr. Sumit Randhir Singh, LV
(OCTA) confirmed the presence of a vascular network. Prasad Eye Institute, Visakhapatnam, India
Angiography (Superfical) Angiography (Deep) Angiography (Outer retina) Angiography (Choriocapillaris)
Angiography (Superfical) Angiography (Deep) Angiography (Outer retina) Angiography (Choriocapillaris)
14.5.4 Bevacizumab resolution of pigment epithelial detachment

(PED) on OCT images. Maximum reduction of
A number of studies investigated the efficacy of CMT and maximum improvement of visual acu-
different anti-VEGF agents in the treatment of ity was observed in the first 4 weeks. Subretinal
idiopathic CNV. Mandal and colleagues [39] fluid resolved faster than intraretinal fluid whereas
reported their findings in treating 32 eyes with PED resolved last. At the end of 12 weeks follow-
idiopathic CNV using bevacizumab up, 59% eyes had improved BCVA, 34% remained
1.25 mg/0.05 ml. The need for reinjection was stable, and 6% had dropped BCVA. The mean
assessed monthly based on the reduction of sub- number of injections was 1.7 per eye with 12
retinal fluid, central macular thickness (CMT), or patients requiring 2 injections, 4 patients requir-
ing 3 injections, and 1 patient requiring 4 injec- humor samples were obtained and inflammatory
tions over a mean follow-up period of 4.2 months. cytokines were assessed prior and after treat-
Zhang et al. [40] prospectively evaluated 40 ment. Twenty-four eyes (61.5%) required retreat-
eyes with subfoveal idiopathic CNV treated using ment during the 1-year follow-up period with a
intravitreal bevacizumab 1.25 mg/0.05 ml once mean of 2.1 injections per eye. All eyes gained
then as needed (PRN) regimen. At 1-year follow- vision at the end of follow-up. Although the
up, 70% showed improvement in vision while aqueous level of VEGF was not higher compared
30% remained stable and none lost vision. Of to the control group prior to treatment, its level
note, 60% of eyes needed reinjection with a mean was significantly decreased compared to pretreat-
of two injections per eye, all were performed ment. Subgroup analysis of inflammatory cyto-
within 3 months of the initial treatment with no kines comparing the single injection group with
recurrence on follow-up. At 12-month follow-up, the multiple injections group showed a signifi-
all lesions were cicatrized. Although there is no cantly higher level of IL-10 and macrophage
consensus on the frequency or the number of inflammatory protein 1b (MIP-1b) in the multiple
injections in idiopathic CNV, the authors pro- injection group. The authors suggested that other
posed that given the nature of idiopathic CNV molecules besides VEGF may play a crucial role
being smaller in size and affecting younger in idiopathic CNV formation.
patients with healthier RPE a single injection fol-
lowed by PRN approach may be more effective
in comparison to monthly injection. Inou et al. 14.5.6 Comparative Studies
[41] reported similar findings for seven eyes with
subfoveal and juxtafoveal idiopathic CNV. All Multiple studies compared between bevaci-
eyes either improved or remained stable at 1-year zumab and ranibizumab for the treatment of
follow-up. The mean number of bevacizumab wet AMD [43–46]. Both medications were
injections was 2.7 per eye. equally effective; however, there are concerns
A larger retrospective study by Qi H-J et al. that bevacizumab may have a worse side effect
[42] in 77 Chinese patients with idiopathic CNV profile due to its greater systemic
(subfoveal and juxtafoveal) who were treated bioavailability.
with bevacizumab found that 79% gained vision, In a retrospective study by Zhou et al. [47],
20% remained stable, and 1% lost > 2 lines of 60 eyes with idiopathic CNV were treated with
vision. Reinjection was needed in 81% of eyes either ranibizumab or bevacizumab with 30
with a mean of 2.47 injections per eye over a eyes in each group. At 2-year follow-up, both
mean follow-up period of 14.3 months. drugs were effective in the treatment of CNV
with no significant difference in the clinical
outcome between the groups. Most patients
14.5.5 Ranibizumab required three or more injections in both groups
(90% of bevacizumab group and 97% in the
Unlike bevacizumab that is used as off-label for ranibizumab group) with a mean number of
the treatment of CNV, ranibizumab is approved by 3.20 and 3.13 injections in bevacizumab and
the United States Food and Drug Administration ranibizumab, respectively. Central retinal thick-
for the treatment of wet AMD. The evidence on ness was significantly less in the ranibizumab
the safety of one medication over the other remains group but no difference in the final visual acuity
contradictory. compared to the bevacizumab group was
In a prospective study by Yin and associates observed. Both medications were equally safe
[21], 39 eyes with idiopathic CNV received intra- with no reported systemic or ocular
vitreal ranibizumab 0.5 mg/0.05 ml. Aqueous complications.
Similar findings in a retrospective compara- 14.6 Key Learning Points

tive study of 47 eyes with idiopathic CNV by
Sudhalkar et al. [48] were reported. Both medica- 1. Idiopathic CNV usually affects the posterior
tions were equally effective in improving visual pole in patients younger than 50 years of age.
acuity with no difference in final central macular 2. It is usually unilateral but could be bilateral.
thickness between the treated groups. The mean 3. It is a diagnosis of exclusion. Findings related
number of injections was 2.4 injections over a to other etiologies may appear over time.
1-year period. No ocular or systemic complica- 4. It carries a better prognosis when compared to
tions were observed. age-related macular degeneration.
From the previous studies, it is noted that the 5. It requires significantly less number of intra-
mean number of injections is significantly lower vitreal anti-VEGF injections to achieve quies-
compared to that reported for the treatment of cence when compared to wet AMD.
wet AMD [44]. There is no consensus on the best
treatment regimen for idiopathic CNV. Several
studies with relatively small numbers reported References
using monthly injections while others used treat-
ment as needed. Further prospective controlled 1. Green WR, Wilson DJ. Choroidal neovascularization.
Ophthalmology. 1986;93:1169–76.
studies on idiopathic CNV to assess the best regi- 2. Ho AC, Yannuzzi LA, Pisicano K, Derosa J. The natu-
men are needed. ral history of idiopathic subfoveal choroidal neovas-
cularization. Ophthalmology. 1995;102:782–9.
3. Lindblom B, Andersson T. The prognosis of idiopathic
choroidal neovascularization in persons younger than
14.5.7 Surgical Treatment 50 years of age. Ophthalmology. 1998;105:1816–20.
4. Bottoni FG, Deutman AF. Idiopathic sub-retinal
Submacular surgery and macular translocation neovascular membranes in the macula (hemor-
are possible options for the treatment of subfoveal rhagic macular choroidopathy of young adults). Doc
Ophthalmol. 1987;64:311–43.
idiopathic CNV. Few studies looked into surgical 5. Cohen SY, Laroche A, Leguen Y, Soubrane G, Coscas
options in idiopathic CNV as well as other non- GJ. Etiology of choroidal neovascularization in young
AMD related CNV such as POHS, pathological patients. Ophthalmology. 1996;103:1241–4.
myopia, or uveitis. [49, 50] As part of the sub- 6. Grossniklaus HE, Green W. Choroidal neovascular-
ization. Am J Ophthalmol. 2004;137:496–503.
macular surgery trials (SST), surgical removal 7. Gass JDM. Biomicroscopic and histopathologic con-
versus observation of subfoveal CNV associated siderations regarding the feasibility of surgical exci-
with ocular histoplasmosis or idiopathic form trial sion of subfoveal Neovascular membranes. Am J
found no benefit or a smaller benefit to surgery Ophthalmol. 1994;118:285–98.
8. Thomas JW, Grossniklaus HE, Lambert HM, Aaberg
than the trial was designed to detect. The risk of TM, Lʼhernault N. Ultrastructural features of surgi-
retinal detachment was 5% and the risk of recur- cally excised idiopathic subfoveal neovascular mem-
rence of CNV was ≥50% [50]. branes. Retina. 1993;13:93–8.
Surgical treatment for CNV in the era of anti- 9. Lee JE, Shin MK, Chung IY, Lee JE, Kim HW, Lee
SJ, Park SW, Byon IS. Topographical relationship
VEGF is limited. The disadvantages of surgery between the choroidal watershed zone and submac-
include being invasive and requiring more com- ular idiopathic choroidal neovascularisation. Br J
plex surgical skills. Complications such as retinal Ophthalmol. 2015;100:652–9.
detachment, cataract progression, and high inci- 10. Hayashi K, Laey JD. Indocyanine green angiography
of submacular choroidal vessels in the human eye.
dence of recurrence have been reported. The ben- Ophthalmologica. 1985;190:20–9.
efit over risk is difficult to be established 11. Ahn SJ, Kim TW, Ahn J, Woo SJ, Park KH, Lee
especially in idiopathic CNV which usually has a BR. Subfoveal choroidal thickness in idiopathic
favorable prognosis and good response to less choroidal neovascularization. Ophthalmology.
2014;121(7):1486–7.
invasive treatment options such as anti-VEGF 12. Sasahara M, Otani A, Yodoi Y, Yoshimura
therapy. N. Circulating hematopoietic stem cells in patients
with idiopathic choroidal neovascularization. Invest 26. Wang Q, Chan SY, Jonas JB, Wei WB. Optical coher-
Ophthalmol Vis Sci. 2009;50:1575. ence tomography angiography in idiopathic choroidal
13. Giovannini A, Scassellati-Sforzolini B, Mariotti C, neovascularization. Acta Ophthalmol. 2015;94:415–7.
D’Altobrando E. Indocyanine green angiographic 27. Chen Q, Yu X, Sun Z, Dai H. The application of
findings in idiopathic choroidal neovascularization. OCTA in assessment of anti-VEGF therapy for idio-
Int Ophthalmol. 1996;20(4):171–9. pathic choroidal neovascularization. J Ophthalmol.
14. Yang F, Dou H-L, Ma Z, Li Y-L, Lu X-R, Wang X, He 2016;2016:1–8.
P-P. Serum inflammatory factors in patients with idio- 28. Liu B, Bao L, Zhang J. Optical coherence tomogra-
pathic choroidal neovascularization. Ocul Immunol phy angiography of pathological myopia sourced and
Inflamm. 2010;18:390–4. idiopathic choroidal neovascularization with follow-
15. Tsai D-C, Charng M-J, Lee F-L, Hsu W-M, up. Medicine. 2016;95(14):e3264.
Chen S-J. Different plasma levels of vascu- 29. Keilhauer CN, Delori FC. Near-infrared autofluores-
lar endothelial growth factor and nitric oxide cence imaging of the fundus: visualization of ocular
between patients with choroidal and retinal neo- melanin. Invest Ophthalmol Vis Sci. 2006;47:3556.
vascularization. Ophthalmologica. 2006;220: 30. Macular Photocoagulation Study Group. Argon laser
246–51. photocoagulation for idiopathic neovascularization.
16. Lip P-L, Blann AD, Hope-Ross M, Gibson JM, Lip Arch Ophthalmol. 1983;101:1358.
GY. Age-related macular degeneration is associ- 31. Krypton laser photocoagulation for idiopathic neovas-
ated with increased vascular endothelial growth cular lesions. Arch Ophthalmol. 1990;108:832.
factor, hemorheology and endothelial dysfunction. 32. Veronese C, Maiolo C, Huang D, Jia Y, Armstrong
Ophthalmology. 2001;108:705–10. GW, Morara M, Ciardella AP. Optical coherence
17. Stefansson E, Geirsdottir A, Sigurdsson H. Metabolic tomography angiography in pediatric choroidal
physiology in age related macular degeneration. Prog neovascularization. Am J Ophthalmol Case Rep.
Retinal Eye Res. 2011;30(1):72–80. 2016;2:37–40.
18. Wankun X, Wenzhen Y, Min Z, Weiyan Z, Huan C, 33. Wang R, Wykoff CC, Christie L, Croft DE, Major
Wei D, et al. Protective effect of paeoniflorin against JC, Fish RH, Brown DM. Choroidal neovasculariza-
oxidative stress in human retinal pigment epithelium tion secondary to alexandrite laser exposure. Retinal
in vitro. Mol Vis. 2011;17:3512–22. Cases Brief Rep. 2016;10:244–8.
19. Zhao M, Bai Y, Xie W, Shi X, Li F, Yang F, et al. Inter- 34. Raoof N, Bradley P, Theodorou M, Moore AT,
leukin-1beta level is increased in vitreous of patients Michaelides M. The new pretender: a large UK case
with neovascular age-related macular degenera- series of retinal injuries in children secondary to
tion (nAMD) and polypoidal choroidal vasculopathy handheld lasers. Am J Ophthalmol. 2016;171:88–94.
(PCV). PLoS One. 2015;10(5):e0125150. 35. Group1A VI. Photodynamic therapy of subfoveal
20. Miao H, Tao Y, Li XX. Inflammatory cytokines in choroidal neovascularization in age-related macu-
aqueous humor of patients with choroidal neovascular degeneration with verteporfin. Arch Ophthalmol.
larization. Mol Vis. 2012;18:574–80. 1999;117:1329.
21. Yin H, Fang X, Ma J, et al. Idiopathic choroidal neo- 36. Bressler NM. Photodynamic therapy of subfoveal
vascularization: intraocular inflammatory cytokines choroidal neovascularization in pathologic myopia
and the effect of intravitreal Ranibizumab treatment. with verteporfin. Ophthalmology. 2001;108:841–52.
Scientific Reports. 2016;25(6):31880. 37. Saperstein DA, Rosenfeld PJ, Bressler NM, Rosa
22. Kang HM, Koh HJ. Five-year visual outcome and RH, Sickenberg M, Sternberg P, Aaberg TM, Aaberg
visual prognostic factors after photodynamic therapy TM, Reaves TA. Photodynamic therapy of subfoveal
with verteporfin for idiopathic choroidal neovascular- choroidal neovascularization with verteporfin in the
ization. ARC J Ophthalmol. 2016;1:12–8. ocular histoplasmosis syndrome. Ophthalmology.
23. Chan W-M, Lam DS, Wong T-H, Lai TY, Kwok AK, 2002;109:1499–505.
Tam BS, Li KK. Photodynamic therapy with vertepor- 38. Giovannini A, Neri P, Mercanti L, Brue
fin for subfoveal idiopathic choroidal neovasculariza- C. Photodynamic treatment versus photodynamic
tion. Ophthalmology. 2003;110:2395–402. treatment associated with systemic steroids for idio-
24. Kumar A, Vohra R, Agrawal S, Chawla R, Azad SV, pathic choroidal neovascularisation. Br J Ophthalmol.
Venkatesh P, Sharma A. Characterization of idio- 2007;91:620–3.
pathic choroidal neovascularization using fluorescein 39. Mandal S. Intravitreal bevacizumab for subfo-
angiography, indocyanine green angiography, and veal idiopathic choroidal neovascularization. Arch
optical coherence tomography angiography. Ophthal Ophthalmol. 2007;125:1487.
Surg Lasers Imaging Retina. 2018;49:516–22. 40. Zhang H, Liu Z-L, Sun P, Gu F. Intravitreal beva-
25. Toju R, Iida T, Sekiryu T, Saito M, Maruko I, Kano cizumab for treatment of subfoveal idiopathic
M. Near-infrared autofluorescence in patients with choroidal neovascularization: results of a 1-year
idiopathic submacular choroidal neovascularization. prospective trial. Am J Ophthalmol. 2012;153(2):
Am J Ophthalmol. 2012;153:314–9. 300–6.
41. Inoue M, Kadonosono K, Watanabe Y, Sato S, 46. Kodjikian L, Souied EH, Mimoun G, Mauget-Faÿsse
Kobayashi S, Yamane S, Ito R, Arakawa A. Results M, Behar-Cohen F, Decullier E, Huot L, Aulagner
of 1-year follow-up examinations after intravit- G. Ranibizumab versus bevacizumab for neovas-
real bevacizumab administration for idiopathic cular age-related macular degeneration: results
choroidal neovascularization. Retina. 2010;30: from the GEFAL noninferiority randomized trial.
733–8. Ophthalmology. 2013;120:2300–9.
42. Qi H-J, Li X-X, Tao Y. Outcome of intravitreal 47. Zhou P, Yang L, Jin X. Ranibizumab versus bevaci-
bevacizumab for idiopathic choroidal neovascular- zumab for the treatment of idiopathic choroidal neo-
ization in a Chinese population. Can J Ophthalmol. vascularization: 2-year results. Eur J Ophthalmol.
2010;45:381–5. 2015;26:262–7.
43. CATT Research Group. Ranibizumab and bevaci- 48. Sudhalkar A, Yogi R, Chhablani J. Anti-vascular endo-
zumab for neovascular age-related macular degenera- thelial growth factor therapy for naive idiopathic cho-
tion. N Engl J Med. 2011;364:1897–908. roidal neovascularization. Retina. 2015;35:1368–74.
44. Chakravarthy U, Harding SP, Rogers CA, Downes SM, 49. Essex RW, Tufail A, Bunce C, Aylward GW. Two-
Lotery AJ, Wordsworth S, Reeves BC. Ranibizumab year results of surgical removal of choroidal neovas-
versus bevacizumab to treat neovascular age- cular membranes related to non-age-related macular
related macular degeneration. Ophthalmology. degeneration. Br J Ophthalmol. 2007;91:649–54.
2012;119:1399–411. 50. Submacular Surgery Trials Research Group. Surgical
45. Krebs I, Schmetterer L, Boltz A, Told R, Vécsei- removal vs observation for subfoveal choroidal neo-
Marlovits V, Egger S, Schönherr U, Haas A, Ansari- vascularization, either associated with the ocular his-
Shahrezaei S, Binder S. A randomised double-masked toplasmosis syndrome or idiopathic: I. ophthalmic
trial comparing the visual outcome after treatment findings from a randomized clinical trial: submacular
with ranibizumab or bevacizumab in patients with surgery trials (SST) group H trial: SST report no. 9—
neovascular age-related macular degeneration. Br J correction. Arch Ophthalmol. 2005;123:28.
Ophthalmol. 2013;97:266–71.
Faisal A. Almarek, MD received his medical degree in

2010. Completed residency training in ophthalmology at
King Khaled Eye Specialist Hospital and King Saud
University, Riyadh, Saudi Arabia. He then joined the vit- Sulaiman M. Alsulaiman, MD received his medical
reoretinal surgery fellowship at King Khaled Eye degree in 2006 and completed residency in ophthalmol-
Specialist Hospital. Dr. Almarek is currently an assistant ogy at King Khaled Eye Specialist Hospital and King
professor at the department of ophthalmology, Imam Saud University, Riyadh, Saudi Arabia. He then com-
Mohammed Bin Saud Islamic University, Riyadh, Saudi pleted vitreoretinal surgery and uveitis fellowships at
Arabia. King Khaled Eye Specialist Hospital followed by pediat-
ric retina fellowship at the Eye and Ear infirmary,
University of Illinois, Chicago. Dr Alsulaiman is currently
an attending vitreoretinal surgeon at the King Khaled Eye
Specialist Hospital.
Subretinal Neovascularization
Associated with Idiopathic 15
Juxtafoveal Telangiectasia
Matthew R. Starr and Sophie J. Bakri
15.1 Introduction vasculature in origin, but as newer imaging

modalities and postmortem studies have become
Macular telangiectasia (MacTel) or idiopathic available, it is now believed that MacTel is pos-
juxtafoveal telangiectasia includes several phe- sibly a neurodegenerative disorder affecting the
notypic subtypes of retinal vascular disease that perifoveal Müller cells and photoreceptors. The
primarily affect the posterior pole. Gass pub- atrophy of these cells, in turn, leads to the vascu-
lished an extensive classification system on the lar abnormalities and retinal changes seen in
various subtypes of macular telangiectasia in MacTel type 2 [3, 5–8].
1993 [1], but despite the complex classification There are two population-based studies
systems, there are in essence two forms. One has examining the incidence and prevalence of
a congenital abnormality that is typically unilat- MacTel type 2 in primarily Caucasian popula-
eral and may be a part of the Coats disease spec- tions [9, 10]. The Beaver Dam Eye Study, found
trum, referred to as MacTel type 1, and an a prevalence of MacTel type 2 of 0.1% [10],
acquired form that is typically bilateral found in while the Melbourne Collaborative Cohort
middle-aged adults that is referred to as MacTel study found a prevalence of 0.0045–0.022% [9].
type 2. The classification system was further The prevalence of MacTel in a primarily African
refined by Yanuzzi in 2006 using multimodal population of nearly 8600 patients was found to
imaging analyses [2]. Yanuzzi and colleagues be 0.06%, similar to the other two studies [11].
classified the original MacTel type 1 as an aneu- Although no specific gene has been identified, it
rysmal macular telangiectasia, while MacTel is widely accepted that the condition is autoso-
type 2 was labeled as an idiopathic macular telan- mal dominant with an incomplete penetrance
giectasia. This chapter will focus on the subreti- [12, 13].
nal neovascularization associated with idiopathic
juxtafoveal telangiectasia or MacTel type 2.
The vascular changes in MacTel are thought to 15.2 Clinical Exam and History
affect the capillaries of the temporal fovea first,
hence the term juxtavofeal telangiectasia [3, 4]. Patients with MacTel tend to present with mild
These abnormalities were initially thought to be visual complaints or even a central scotoma in
their fifth or sixth decades of life [3]. The charac-
M. R. Starr · S. J. Bakri (*) teristic fundus findings include the loss of retinal
Department of Ophthalmology, Mayo Clinic, transparency, and the presence of capillary dilata-
Rochester, MN, USA tion and telangiectasias, right-angle vessels, and
e-mail: bakri.sophie@mayo.edu

180 M. R. Starr and S. J. Bakri
retinal pigment epithelial (RPE) hyperplasia. The ally extends to involve the entire perifoveal area
earliest manifestation of MacTel type 2 tends to [10]. Involvement of both the inner and the outer
be loss of the retinal transparency (Fig. 15.1a), capillary plexuses has been described [2]. The
which begins temporal to the fovea, but eventu- hard exudates seen on funduscopic examination
a b
d e
Fig. 15.1 Left eye imaging of a 41 year-old female with poral to fovea (b). Optical coherence tomography revealed
macular telangiectasia type 2 with associated vitelliform- a flat-top atrophic cyst, bullseye pattern ellipsoid zone
like lesion in the left eye. On fundus photography (a) there layer loss, and scattered intraretinal cysts (c). On early
is evidence of decreased retinal transparency with blunted phase fluorescein angiography at 34 s there is evidence of
retinal vessels and yellow foveal depositis. Fundus auto- telangiectatic vasculature just temporal to the fovea (d)
fluoresence revealed loss of the central hypo- with diffuse late leakage of these vessels on late phase
autofluorescent fovea with a hyper-autofluorescent lesion fluorescein angiography at 9 minutes (e)
at the fovea and a stippled hyperautofluorescent area tem-
15 Subretinal Neovascularization Associated with Idiopathic Juxtafoveal Telangiectasia 181
of those patients with MacTel type 1 are typically mild hyperfluorescence in the absence of abnor-
not seen in patients with MacTel type 2, however, mal capillaries thought to be staining of the
crystalline deposits can be seen in the vitreoreti- outer retina [1], but this might actually be a win-
nal interface in those patients with MacTel type 2 dow defect due to loss of the macular pigment in
[14]. As venules course toward the fovea in the perifoveal region [21]. Over the course of
patients with MacTel type 2, they dilate and make the disease, the telangiectatic vessels will even-
right-angle turns into the deeper retinal layers. tually involve the entire parafoveal region.
RPE hyperplasia is also noted along the courses During the later stages of the FA, there is diffuse
of these right-angle venules and lastly, atrophic hyperfluorescence presumed due to the abnor-
retinal changes occur [2]. This atrophy can poten- mal vasculature in the absence of choroidal neo-
tially lead to the formation of full-thickness mac- vascularization (CNV, Fig. 15.1e). OCT
ular holes [15–17]. angiography (OCTA) reveals the abnormally
Subretinal neovascularization may occur in dilated vessels tend to be drawn to a temporal
patients with MacTel, and is typically preceded focal point in the fovea and can highlight the
by the right-angle venules with subsequent RPE point where venules dive down to the deeper
hyperplasia [1]. These patients will typically vasculature [22]. Newer studies have identified
manifest with acute vision changes, loss of retinal–choroidal anastomoses that develop
vision, metamorphopsia, or with a central sco- before subretinal neovascularization and in con-
toma. The neovascularization is retinal in origin cordance with the development of right-angle
as evidenced by the feeder vessels from the reti- venules. Spaide et al. postulated that the abnor-
nal arteries draining into the venules [18]. Lastly, mal vessels moved posterior toward the choroid,
a disciform scar can manifest as the advanced rather than into the vitreous, due to the meta-
stage of the neovascularization process, leading bolic derangements these patients have; one of
to a significant decline in visual acuity. which is potentially high cholesterol [22].
15.3 Diagnostic Imaging 15.4 ssociated Systemic Medical

A
Conditions
On fundus autofluorescence, there is a loss of
the foveal hypoautofluorescent center, thought Recent studies have found that patients with
to be due to the depletion of the retinal pigment MacTel type 2 have a higher prevalence of sys-
at the fovea (Fig. 15.1b) [19]. On optical coher- temic vascular comorbidities such as diabetes
ence tomography (OCT), patients may fre- mellitus, hyperlipidemia, hypertension, obesity,
quently present with temporal enlargement of and cancer [3, 23, 24]. Additionally, it appears
the foveal pit, hyporeflective cavities of the that patients with MacTel may be more likely to
inner and outer retina, normal retinal thickness have a history of smoking [3]. It is very possible
despite angiographic evidence of leakage, dis- that diabetes and hypertension lead to increased
ruption of the inner and outer photoreceptor susceptibility of the macular vasculature to
segments, increased reflectivity of the inner reti- develop the abnormalities seen in MacTel
nal layers, hyperreflective intraretinal lesions, patients. Due to the associations with these dis-
and outer retinal atrophy late in the disease. eases, it is plausible other vascular diseases
Infrequently, patients may present with lamellar could be associated with MacTel, but no other
or full-thickness macular holes, foveal detach- systemic medical risk factors have been identi-
ments, or subfoveal debris (Fig. 15.1c) [2, 3, fied. Recently, there was no association found
20]. Fluorescein angiography (FA) reveals early between MacTel and obstructive sleep apnea
telangiectatic capillaries in the temporal fovea [25], while there remain no definitive ocular
in the early phase of the FA (Fig. 15.1d). During associations that have been found with macular
the early stages of the disease, there may be a telangiectasia.
15.5 Subretinal to that of retinal angiomatous proliferation lesions

Neovascularization with proliferation of the deep capillary circulation
Associated with Idiopathic served by these abnormal anastomoses [2]. Gass
Macular Telangiectasia also noted that certain end-stage eyes were noted to
have anastomoses between these subretinal neovas-
Subretinal neovascular complexes are a late mani- cular complexes and the choroid [1]. These choroi-
festation of MacTel type 2 and are thought to be dal anastomoses were also found in a separate study
distinct from the CNV lesions seen in patients with by Engelbrecht and colleagues [18]. These anasto-
age-related macular degeneration (AMD) and neo- moses and neovascular complexes were difficult to
vascularization elsewhere of diabetes [23]. These image, though, due to the leaky nature of both the
lesions may present at any time during the natural neovascular complexes and abnormally dilated
course of the disease [1], but are typically preceded venules that typically were overlying them. OCTA
by the right-angle venules and RPE hyperplasia imaging, however, nicely is able to delineate these
[22]. Gass classified the neovascularization seen subretinal neovascular complexes [26]. These find-
these patients as type 2 subretinal neovasculariza- ings were further explained by Spaide et al. who
tion that developed from the proliferation and noted the presence of retinal choroidal anastomoses
remodeling of the outer capillary plexus and in 65% of eyes with MacTel type 2 and that these
extended underneath the retina and RPE but above anastomoses occur before and independent to the
Bruch’s membrane [1]. Yanuzzi in 2006 proposed development of the subretinal neovascular com-
that the subretinal neovascularization originated plexes associated with late-stage MacTel type 2
from the deep retinal circulation and in his study [22]. This led the group to conclude that MacTel
observed retinal to retinal and retinal to subretinal type 2 may not be a pure neurodegenerative process
anastomoses [2]. Yanuzzi and colleagues felt the with secondary retinal and vascular changes, as the
subretinal neovascularization in MacTel was similar choroid appears to be involved as well (Fig. 15.2).
a b c d
e f
Fig. 15.2 A 68-year-old male with subretinal neovascu- on the late phase angiogram at 7 min (d). Optical coher-
larization presumed due to idiopathic macular telangiecta- ence tomography (e) of the left eye of the same patient
sia type 2. Color fungus photograph (a) of the left eye with subretinal hyper-reflective material consistent with a
with loss of retinal transparency, retinal pigment epithelial potential neovascular network. Optical coherence tomog-
changes, and blunted retinal vasculature. Fluorescein raphy angiography (f) of the left eye reveals a fine net-
angiography of the left eye (B-D) with the early phase (b) work or retinal neovascularization temporal to the fovea.
at 28 s revealing a retinal neovascular network temporal to Photographs courtesy of Dr. Avantika Dogra, LV Prasad
the fovea. This lesion progressive leaked through the Eye Institute, Hyderabad, India
course of the angiography, evidenced at both 51 s (c) and
15.6 Treatment of Subretinal [33]. There were two series of 6 patients each,
Neovascularization both with follow up less than 2 years, that found
in Macular Telangiectasia the majority of eyes maintained visual acuity
Type 2 through the course of their respective studies [32,
34]. Snyers et al. reported that one eye in their
As previously mentioned, the subretinal neovas- series of 5 had vision loss persist after PDT with
cular complexes in eyes with macular telangiec- persistent fluorescein leakage without enlarge-
tasia type 2 are very different from the CNV ment of the neovascular lesion [35]. Hershberger
lesions seen in patients with AMD. Park and col- and colleagues also noted that in their series of
leagues in 1996 described the natural history of eyes, there was rapid refilling of the neovascular
these fibrovascular lesions in 11 eyes. The cohort lesions thought to be due to the choroidal–retinal
had an average follow up of 44 months and only anastomoses that develop during the natural
2 eyes lost 2 or more lines of vision after the course of the disease [31]. These anastomoses
detection of the subretinal neovascularization perhaps can lead to treatment failure if eyes are
[27]. Additionally, Park et al. studied the effects followed long enough. In the series by Hussain,
of grid laser photocoagulation in patients with all 6 eyes had evidence of associated retinal pig-
macular edema and subretinal neovascularization ment epithelial damage [32] and in another
and found no benefits in improving or stabilizing report, there was atrophy of the RPE following
long-term visual acuity in a cohort of 10 eyes PDT [36]. There is reasonable concern that the
compared to a control population [28]. Park con- verteporfin may leak from the telangiectatic ves-
cluded that perhaps focal laser photocoagulation sels and cause direct toxic damage to the sur-
may not be beneficial as the vision does not tend rounding neurosensory retina and RPE [3]. Given
to worsen over time. Given that the average fol- the small reported numbers with minimal follow-
low-up in both studies was around 5 years, they up, potentially toxic side effects, and in the set-
may not adequately represent the eyes that will ting of newer antiangiogenic medications, PDT
deteriorate much later in the course of the dis- may play a limited role in the management of
ease. Regardless, both studies reflect how differ- subretinal neovascularization associated with
ent these fibrovascular complexes are compared MacTel type 2, but perhaps can be reserved for
to typical CNV lesions. certain, refractory neovascular lesions.
Other studies have examined the use of photo- With the success intravitreal anti-VEGF injec-
dynamic therapy (PDT) in MacTel patients with tions have had on visual recovery and stabiliza-
macular edema, but without subretinal neovascu- tion with other causes of CNV [37, 38], several
larization lesions which found mixed results [29, studies evaluated the visual and anatomical out-
30]. All are small case reports or case series comes in MacTel patients with subretinal neovas-
which make it difficult to glean the clinical utility cularization (Table 15.1) [39–47]. Jorge and
of these studies for nonproliferative MacTel. colleagues were the first to show an improvement
MacTel associated with subretinal neovascular- in visual acuity in a patient with subretinal neo-
ization complexes, though, seems to respond well vascularization in 2006 [41]. This led to larger
to PDT. Several small case series have shown case series and retrospective studies that all
improvement or stability in visual acuity over showed an improvement in visual acuity and a
time in eyes with subretinal neovascularization decrease in retinal thickness over the treatment
treated with PDT [31–35]. The first report of course. In the series of 9 eyes by Roller and col-
visual improvement was a case report by Potter leagues, eyes needed an average of 4.9 injections
and colleagues in 2002 that noted visual improve- over a mean period of approximately 1.5 years
ment that was stable 7 months after treatment. [46]. In a similar study by Narayanan et al. who
Interestingly, the fluorescein leakage from the evaluated 16 eyes over a 1 year time period, eyes
telangiectatic vessels persisted, but the leakage only required on average 1.9 injections. These
from the subretinal neovascularization stopped reports conclude that these subretinal neovascu-
Table 15.1 Summary of studies examining the use of intravitreal anti-vascular endothelial growth factor injections for
the treatment of idiopathic macular telangiectasia type 2
Mean number of Baseline visual Final visual
Number of Intravitreal intravitreal acuity (Snellen acuity (Snellen Mean follow-up
Study eyes agent injections acuity) acuity) (years)
Charbel 6 Bevacizumab 2 20/80 20/50 1.5
et al. [39]
Do et al. 10 Ranibizumab 6 20/64 20/50 0.5
[40]
Jorge et al. 1 Bevacizumab 1 20/40 20/20 2
[41]
Kovach and 5 Bevacizumab 2.8 20/86 20/48 1.3
Rosenfeld
[43]
Mandal 6 Bevacizumab 1 20/200 20/100 0.5
et al. [44]
Narayanan 16 Bevacizumab 1.9 20/120 20/70 1
et al. [45] and
Ranibizumab
Roller et al. 9 Bevacizumab 4.9 20/80 20/63 1.5
[46]
Toygar 25 Bevacizumab 8.4 20/91 20/62 3.3
et al. [47]
lar complexes appear to respond well to intravit- period of 8.6 months with 11 eyes that had stable
real anti-VEGF therapy with fewer injections visual acuity, 1 eye with improved vision, and 1
than their AMD counterparts. The study with the eye with a decrease in visual acuity [49]. Both
longest follow-up period was reported by Toygar combination PDT with triamcinolone and PDT
and colleagues on 25 eyes with an average fol- with anti-VEGF have been reported successfully
low-up of 3.5 years [47]. These eyes required a [29, 50, 51], but the same concerns with PDT in
mean of 8.4 injections over that time frame with these eyes persists and thus the data is limited
an improvement in visual acuity and reduction in regarding combination PDT therapy. Lastly, sur-
retinal thickness that was maintained over the gical removal of the subretinal neovascular mem-
study period. Conversely, intravitreal anti-VEGF branes has been tried in two separate reports,
for the management of macular edema without both of which report difficulty removing the sub-
subretinal neovascularization in MacTel patients retinal complexes without damaging or removing
has been largely unsuccessful with no functional the adjacent neurosensory retina making surgery
or visual benefit [48]. Thus, the preferred treat- obsolete in the management of subretinal neovas-
ment for patients with MacTel who experience a cularization associated with macular telangiecta-
sudden decline in visual acuity due to subretinal sia [52, 53].
neovascularization is with intravitreal anti-VEGF
therapy. Compared to other diseases causing Key Learning Points
CNV, patients with MacTel may require fewer 1. Patients with macular telangiectasia type 2
injections in order to maintain visual acuity gains may develop subretinal neovascular com-
and retinal thickness reduction. plexes that originate from the retinal
Other modalities that have been tried for these vasculature.
subretinal neovascular complexes include trans- 2. These complexes almost always form after the
pupillary thermotherapy (TTT), combination development of right-angle venules and RPE
PDT with both triamcinolone and anti-VEGF hyperplasia.
agents, and even surgical removal of the lesions. 3. The neovascularization may develop indepen-
The report on TTT followed 13 eyes over a mean dently from the formation of retinal–choroidal
anastomoses seen in patients with macular two African populations. Ophthalmic Epidemiol.
2012;19(4):185–9.
telangiectasia. 12. Parmalee NL, Schubert C, Figueroa M, et al.
4. Nonproliferative macular telangiectasia does Identification of a potential susceptibility locus
not respond to anti-VEGF therapy. for macular telangiectasia type 2. PLoS One.
5. The subretinal neovascular complexes in mac- 2012;7(8):e24268.
13. Ronquillo CC, Wegner K, Calvo CM, Bernstein
ular telangiectasia do not respond to well to PS. Genetic penetrance of macular telangiectasia type
focal laser or photodynamic therapy. 2. JAMA Ophthalmol. 2018;136(10):1158–63.
6. Anti-VEGF for the management of neovascu- 14. Moisseiev J, Lewis H, Bartov E, Fine SL, Murphy
lar complexes in macular telangiectasia RP. Superficial retinal refractile deposits in juxtafoveal
telangiectasis. Am J Ophthalmol. 1990;109(5):604–5.
patients maintains visual acuity gains and 15. Charbel Issa P, Scholl HP, Gaudric A, et al. Macular
reductions in retinal thickness. full-thickness and lamellar holes in association with
7. Patients with macular telangiectasia may type 2 idiopathic macular telangiectasia. Eye (Lond).
require fewer anti-VEGF injections than other 2009;23(2):435–41.
16. Patel B, Duvall J, Tullo AB. Lamellar macular hole
patients with typical choroidal neovascular associated with idiopathic juxtafoveolar telangiecta-
lesions. sia. Br J Ophthalmol. 1988;72(7):550–1.
17. Karth PA, Raja SC, Brown DM, Kim JE. Outcomes of
macular hole surgeries for macular telangiectasia type
2. Retina. 2014;34(5):907–15.
References 18. Engelbrecht NE, Aaberg TM Jr, Sung J, Lewis
ML. Neovascular membranes associated with idio-
1. Gass JD, Blodi BA. Idiopathic juxtafoveolar retinal pathic juxtafoveolar telangiectasis. Arch Ophthalmol.
telangiectasis. Update of classification and follow-up 2002;120(3):320–4.
study. Ophthalmology. 1993;100(10):1536–46. 19. Solberg Y, Dysli C, Wolf S, Zinkernagel
2. Yannuzzi LA, Bardal AM, Freund KB, Chen KJ, MS. Fluorescence lifetime patterns in macular telangi-
Eandi CM, Blodi B. Idiopathic macular telangiecta- ectasia type 2. Retina. 2020; https://doi.org/10.1097/
sia. Arch Ophthalmol. 2006;124(4):450–60. IAE.0000000000002411.
3. Charbel Issa P, Gillies MC, Chew EY, et al. 20. Sallo FB, Leung I, Clemons TE, Peto T, Bird
Macular telangiectasia type 2. Prog Retin Eye Res. AC, Pauleikhoff D. Multimodal imaging in type
2013;34:49–77. 2 idiopathic macular telangiectasia. Retina.
4. Tikellis G, Gillies MC, Guymer RH, McAllister 2015;35(4):742–9.
IL, Shaw JE, Wong TY. Retinal vascular caliber 21. Gillies MC, Zhu M, Chew E, et al. Familial asymptom-
and macular telangiectasia type 2. Ophthalmology. atic macular telangiectasia type 2. Ophthalmology.
2009;116(2):319–23. 2009;116(12):2422–9.
5. Eliassi-Rad B, Green WR. Histopathologic study 22. Spaide RF, Yannuzzi LA, Maloca PM. Retinal-
of presumed parafoveal telangiectasis. Retina. choroidal anastomosis in macular telangiectasia type
1999;19(4):332–5. 2. Retina. 2018;38(10):1920–9.
6. Gass JD. Histopathologic study of presumed parafo- 23. Chew EY, Murphy RP, Newsome DA, Fine
veal telangiectasis. Retina. 2000;20(2):226–7. SL. Parafoveal telangiectasis and diabetic retinopathy.
7. Powner MB, Gillies MC, Zhu M, Vevis K, Hunyor AP, Arch Ophthalmol. 1986;104(1):71–5.
Fruttiger M. Loss of Muller’s cells and p hotoreceptors 24. Clemons TE, Gillies MC, Chew EY, et al. Medical
in macular telangiectasia type 2. Ophthalmology. characteristics of patients with macular telangiectasia
2013;120(11):2344–52. type 2 (MacTel Type 2) MacTel project report no. 3.
8. Wubben TJ, Zacks DN, Besirli CG. Retinal neuropro- Ophthalmic Epidemiol. 2013;20(2):109–13.
tection: current strategies and future directions. Curr 25. Lee MG, Marshall NS, Clemons TE, et al. No associa-
Opin Ophthalmol. 2019;30(3):199–205. tion between sleep apnoea and macular telangiectasia
9. Aung KZ, Wickremasinghe SS, Makeyeva G, Robman type 2 and its markers of severity and progression:
L, Guymer RH. The prevalence estimates of macular a case-control study and retrospective cohort study.
telangiectasia type 2: the Melbourne collaborative Clin Exp Ophthalmol. 2019; https://doi.org/10.1111/
cohort study. Retina. 2010;30(3):473–8. ceo.13363.
10. Klein R, Blodi BA, Meuer SM, Myers CE, Chew EY, 26. Balaratnasingam C, Yannuzzi LA, Spaide RF. Possible
Klein BE. The prevalence of macular telangiectasia choroidal neovascularization in macular telangiecta-
type 2 in the Beaver Dam eye study. Am J Ophthalmol. sia type 2. Retina. 2015;35(11):2317–22.
2010;150(1):55–62. e52 27. Park D, Schatz H, McDonald HR, Johnson
11. Sallo FB, Leung I, Mathenge W, et al. The preva- RN. Fibrovascular tissue in bilateral juxtafoveal tel-
lence of type 2 idiopathic macular telangiectasia in angiectasis. Arch Ophthalmol. 1996;114(9):1092–6.
28. Park DW, Schatz H, McDonald HR, Johnson 41. Jorge R, Costa RA, Calucci D, Scott IU. Intravitreal
RN. Grid laser photocoagulation for macular edema bevacizumab (Avastin) associated with the regression
in bilateral juxtafoveal telangiectasis. Ophthalmology. of subretinal neovascularization in idiopathic juxtafo-
1997;104(11):1838–46. veolar retinal telangiectasis. Graefes Arch Clin Exp
29. De Lahitte GD, Cohen SY, Gaudric A. Lack of appar- Ophthalmol. 2007;245(7):1045–8.
ent short-term benefit of photodynamic therapy in 42. Karagiannis D, Georgalas I, Ladas I, Eustratios P,
bilateral, acquired, parafoveal telangiectasis without Mitropoulos P. A case of subretinal neovasculariza-
subretinal neovascularization. Am J Ophthalmol. tion treated with intravitreal ranibizumab in a patient
2004;138(5):892–4. with idiopathic juxtafoveal retinal telangiectasis. Clin
30. Kotoula MG, Chatzoulis DZ, Karabatsas CH, Interv Aging. 2009;4:63–5.
Tsiloulis A, Tsironi EE. Resolution of macu- 43. Kovach JL, Rosenfeld PJ. Bevacizumab (avastin)
lar edema in idiopathic juxtafoveal telangiecta- therapy for idiopathic macular telangiectasia type
sis using PDT. Ophthalmic Surg Lasers Imaging. II. Retina. 2009;29(1):27–32.
2009;40(1):65–7. 44. Mandal S, Venkatesh P, Abbas Z, Vohra R, Garg
31. Hershberger VS, Hutchins RK, Laber S. Intravitreal bevacizumab (Avastin) for subretinal
PW. Photodynamic therapy with verteporfin for neovascularization secondary to type 2A idiopathic
subretinal neovascularization secondary to bilat- juxtafoveal telangiectasia. Graefes Arch Clin Exp
eral idiopathic acquired juxtafoveolar telangiectasis. Ophthalmol. 2007;245(12):1825–9.
Ophthalmic Surg Lasers Imaging. 2003;34(4):318–20. 45. Narayanan R, Chhablani J, Sinha M, et al. Efficacy
32. Hussain N, Das T, Sumasri K, Ram LS. Bilateral of anti-vascular endothelial growth factor therapy in
sequential photodynamic therapy for sub-retinal neo- subretinal neovascularization secondary to macular
vascularization with type 2A parafoveal telangiecta- telangiectasia type 2. Retina. 2012;32(10):2001–5.
sis. Am J Ophthalmol. 2005;140(2):333–5. 46. Roller AB, Folk JC, Patel NM, et al. Intravitreal
33. Potter MJ, Szabo SM, Chan EY, Morris bevacizumab for treatment of proliferative and non-
AH. Photodynamic therapy of a subretinal neo- proliferative type 2 idiopathic macular telangiectasia.
vascular membrane in type 2A idiopathic juxta- Retina. 2011;31(9):1848–55.
foveolar retinal telangiectasis. Am J Ophthalmol. 47. Toygar O, Guess MG, Youssef DS, Miller DM. Long-
2002;133(1):149–51. term outcomes of intravitreal bevacizumab therapy
34. Potter MJ, Szabo SM, Sarraf D, Michels R, Schmidt- for subretinal neovascularization secondary to
Erfurth U. Photodynamic therapy for subretinal idiopathic macular telangiectasia type 2. Retina.
neovascularization in type 2A idiopathic juxtafo- 2016;36(11):2150–7.
veolar telangiectasis. Can J Ophthalmol. 2006;41(1): 48. Charbel Issa P, Finger RP, Kruse K, Baumuller
34–7. S, Scholl HP, Holz FG. Monthly ranibizumab for
35. Snyers B, Verougstraete C, Postelmans L, Leys A, nonproliferative macular telangiectasia type 2: a
Hykin P. Photodynamic therapy of subfoveal neo- 12-month prospective study. Am J Ophthalmol.
vascular membrane in type 2A idiopathic juxta- 2011;151(5):876–86. e871
foveolar retinal telangiectasis. Am J Ophthalmol. 49. Shukla D, Singh J, Kolluru CM, Kim R,
2004;137(5):812–9. Namperumalsamy P. Transpupillary thermotherapy
36. Shanmugam MP, Agarwal M. RPE atrophy follow- for subfoveal neovascularization secondary to group
ing photodynamic therapy in type 2A idiopathic 2A idiopathic juxtafoveolar telangiectasis. Am J
parafoveal telangiectasis. Indian J Ophthalmol. Ophthalmol. 2004;138(1):147–9.
2005;53(1):61–3. 50. Rishi P, Shroff D, Rishi E. Combined photody-
37. Brown DM, Michels M, Kaiser PK, et al. Ranibizumab namic therapy and intravitreal ranibizumab as pri-
versus verteporfin photodynamic therapy for neovas- mary treatment for subretinal neovascular membrane
cular age-related macular degeneration: two-year (SRNVM) associated with type 2 idiopathic macular
results of the ANCHOR study. Ophthalmology. telangiectasia. Graefes Arch Clin Exp Ophthalmol.
2009;116(1):57–65. e55 2008;246(4):619–21.
38. Rosenfeld PJ, Brown DM, Heier JS, et al. Ranibizumab 51. Smithen LM, Spaide RF. Photodynamic therapy and
for neovascular age-related macular degeneration. N intravitreal triamcinolone for a subretinal neovascu-
Engl J Med. 2006;355(14):1419–31. larization in bilateral idiopathic juxtafoveal telangiec-
39. Charbel Issa P, Finger RP, Holz FG, Scholl tasis. Am J Ophthalmol. 2004;138(5):884–5.
HP. Eighteen-month follow-up of intravitreal bevaci- 52. Berger AS, McCuen BW 2nd, Brown GC, Brownlow
zumab in type 2 idiopathic macular telangiectasia. Br RL Jr. Surgical removal of subfoveal neovasculariza-
J Ophthalmol. 2008;92(7):941–5. tion in idiopathic juxtafoveolar retinal telangiectasis.
40. Do DV, Bressler SB, Cassard SD, et al. Retina. 1997;17(2):94–8.
Ranibizumab for macular telangiectasia type 2 in 53. Davidorf FH, Pressman MD, Chambers
the absence of subretinal neovascularization. Retina. RB. Juxtafoveal telangiectasis-a name change?
2014;34(10):2063–71. Retina. 2004;24(3):474–8.
Peripapillary Choroidal
Sumit Randhir Singh and Jay Chhablani
16.1 Introduction event can happen either concomitantly or after a

gap of several years [5].
Peripapillary choroidal neovascular membrane Most common location of CMVM is nasal to
(CNVM) is a distinct terminology used to describe the optic disc in up to 40% of the cases [7], how-
CVNM that is located near the optic nerve head. ever, these patients may be asymptomatic
Broadly speaking, any CNVM a portion of which whereas the most common location in symptom-
is located within one disc diameter is considered atic patients is temporal due to involvement of
as peripapillary CVNM [1]. Approximately 10% papillomacular bundle [8]. Histopathological
of all CNVM are peripapillary in a location with a studies show the presence of these CNVM lesions
female predilection [1–3]. in apparently normal eyes suggestive of a higher
Description of peripapillary CNVM includes prevalence of peripapillary CNVM as previously
the number of clock hours, quadrants, size of thought [8, 9].
CNVM complex, and distance from fovea. Ruben Peripapillary CNVM has certain differences
et al. devised a grading system to classify the with respect to subfoveal CNVM. There is a vari-
CNVM in different zones. Each zone centered ation in terms of etiology, location of the lesions.
around the disc covers 60 degrees. Zone 1 extends A majority (up to 2/3rd) of peripapillary CNVM
30 degrees above and below a straight line con- may be asymptomatic for long duration [7], and
necting optic disc and fovea [2]. Although no therefore treatment may not be warranted in con-
consensus exists, few authors have defined large trast to subfoveal CNVM, which invariably needs
peripapillary CNVM as lesions with size more treatment. The site of CNVM formation in
than 3.5 disc areas or involving >50% of disc cir- PCNVM is through the breaks in Bruch’s mem-
cumference [4]. These large lesions may show brane (57%) or through the end of Bruch’s mem-
unpredictable growth and may harbor a predomi- brane at the edge of the optic disc (43%) more
nantly occult component [4]. Fellow eye involve- commonly on nasal side [8].
ment is common in up to 20–62% of cases in
patients more than 70 years of age [5, 6]. This
16.2 Etiologies
S. R. Singh
Smt. Kanuri Santhamma Centre for Vitreo-Retinal
A myriad of ocular diseases can lead to the forma-
Diseases, L V Prasad Eye Institute, Hyderabad, India
tion of peripapillary CNVM, the most common
J. Chhablani (*)
being neovascular age-related macular degenera-
Department of Ophthalmology, University of
Pittsburgh Medical Center, Pittsburgh, PA, USA tion (AMD), inflammatory chorioretinal disorders,

188 S. R. Singh and J. Chhablani
angioid streaks, and optic disc anomalies [10]. No Pathological Myopia It is defined as degenera-
identifiable cause can be ascertained in a signifi- tive changes in eyes with high myopia (≥6 D,
cant proportion (up to 40%) of these eyes, hence diopters) [20]. Often termed as myopic maculopa-
termed idiopathic [6]. While AMD and idiopathic thy or myopic macular degeneration, a consensus
causes predominate in elderly, inflammatory, optic definition according to META analysis for
disc anomalies, and angioid streaks are more com- Pathologic Myopia (META-PM) study includes
monly seen in younger age groups [11]. the presence of diffuse/patchy chorioretinal atro-
phy or macular atrophy or presence of staphyloma
[21]. There is sparse literature on peripapillary
16.3 Chorioretinal Degenerations CNVM in myopic eyes with one case series report-
ing the incidence as 4.2% (11 of 260 myopic
Neovascular AMD AMD is characterized by CNVM) [22]. This is despite a high overall inci-
drusen of size more than 63 μ, retinal pigment epi- dence (5–11%) of myopic CNVM in eyes with
thelial (RPE) atrophy, fibrovascular pigment epi- pathological myopia [23]. These CNVM are
thelial detachments (FVPED), and/or disciform known to develop at the edge of conus around the
scars in the macular area with reports suggesting optic disc and more commonly seen with large
an association of peripapillary CNVM with conus [22]. Their source of origin is possible
advanced stages of AMD [7, 12]. However, these through a defect in Bruch’s membrane at the site
macular changes in the form of RPE atrophy and of lacquer cracks near the myopic conus. Moreover,
drusen (including subretinal drusenoid deposits, they tend to regress even without treatment [22].
SDD) can be present in the peripapillary area also
which may be a harbinger of peripapillary CNVM Post Laser Photocoagulation Though laser
similar to macular CNVM. Overall, 45.2–84.6% photocoagulation has been used to treat CNVM
of all cases of peripapillary CNVM are caused due in multiple clinical scenarios, the risk of recur-
to AMD [6, 13]. Wilde et al. reported the preva- rence and CNVM formation at the edge of the
lence of peripapillary CNVM secondary to AMD lesion is well documented. Govindhari et al. have
as 3.9% (9 eyes among 231 neovascular AMD shown exacerbation of CNVM after 3 months of
cases) [3]. Fellow eye involvement was seen in laser in a case of angioid streak [24].
19.8% and 54% patients with AMD-related peri-
papillary CNVM over 2 and 7 years, respectively
[5, 6]. Bilaterality up to 62% was more common in 16.4 Inflammatory
patients with age more than 70 years [5].
A host of inflammatory chorioretinal disorders
Angioid Streaks CNVM is a rare complication can lead to CNVM formation. Common causes
in eyes with angioid streaks that are characterized include infectious (toxoplasma retinochoroiditis,
by a break in brittle and calcified Bruch’s mem- presumed ocular tuberculosis) and inflammatory
brane [14]. There can be systemic associations pathologies (sarcoidosis, serpiginous-like cho-
such as pseudoxanthoma elasticum, Ehlers– roiditis (SLC), Punctate inner choroidopathy
Danlos syndrome, Paget’s disease of bone, sickle (PIC), multifocal choroiditis and panuveitis
cell disease, or other hemoglobinopathies in half (MCP), presumed ocular histoplasmosis syn-
of the cases whereas in remaining cases, the cause drome (POHS), Vogt–Koyanagi–Harada (VKH)
is not identifiable [15]. The risk of CNVM forma- syndrome and other inflammatory chorioretinop-
tion increases with age and also appears related to athies) [10, 25–28]. The background inflamma-
the width, length, and location of angioid streaks, tory pathology may or may not be active once
especially in close proximity to disc [16, 17]. The clinical manifestations of CNVM ensue. Animal
reported incidence of CNVM in these eyes has models suggest that neovascularization can occur
been as high as 72–86% [18, 19]. CNVM in these even in the absence of disease activity [29].
eyes, however, are small and have a higher prob- The etiologies vary widely according to geo-
ability of being asymptomatic. graphic location. For instance, POHS is com-
16 Peripapillary Choroidal Neovascularization 189
monly reported from Ohio and Mississippi river Morning glory syndrome is a congenital con-
valleys and is very rare outside this endemic zone dition associated with retinal dysplasia, and exca-
[30]. POHS is the most common cause of inflam- vation of optic disc filled with glial tissue and
matory CNVM in children in Western literature pigment ring. Few case reports have shown an
[31]. Ocular tuberculosis is more commonly seen association of CNVM with morning glory syn-
in the Indian subcontinent [25, 27]. drome and tilted disc syndrome in peripapillary
Perentes et al. have reported a prevalence of location [40–43]. CNVM have also been reported
1.9% (12/648 patients) of subretinal membranes in eyes with retinochoroidal coloboma in peri-
[10]. In contrast to other causes, the primary papillary location, optic nerve coloboma, con-
treatment for ocular inflammation in the form of genital optic disc pit, or optic neuritis [44–48].
local or systemic steroid or immunomodulatory
therapy needs to be continued as they may com-
plement and help reduce inflammation. 16.7 Dystrophies
Pozzoni et al. reported a case in a 12-year male

16.5 Idiopathic with peripapillary CNVM in a case of Best dis-
ease. Vitelliform lesion was present at macula
This category where no underlying choroidal or with subretinal heme and CNVM adjacent to
retinal pathology can be established is a diagno- optic disc margin [49]. Browning et al. reported 3
sis of exclusion and includes 5–40% of patients eyes of peripapillary CNVM due to pattern dys-
with peripapillary CNVM [6, 32]. Idiopathic trophy among which 2 and 1 were occult and
CNVM similar to AMD-related CNVM are pre- classic CNVM, respectively [6].
dominantly occult in majority of eyes (64.1%)
[6]. Although the recurrence rates appear compa-
rable between idiopathic and AMD CNVM, the 16.8 Congenital/Benign/
visual outcomes in idiopathic cases are relatively Malignant Lesions of Retina
better [6]. and Choroid
Very rarely, choroidal neoplastic lesions can be

16.6 Optic Disc Anomalies associated with CNVM. Less than 1% of giant
choroidal nevi (≥10 mm basal dimension)
Optic nerve head drusen (ONHD) are an inciden- develop CNVM over a mean period of 61 months
tal finding in majority of the patients with an esti- [50]. Peripapillary CNVM associated with cho-
mated prevalence of 3.4–24 individuals per 1000 roidal nevi was seen in a total of 4 eyes (17%)
population [33]. These originate due to disturbed among 24 eyes with choroidal nevi associated
axonal metabolism in association with a small CNVM [51]. Tuncer et al. reported a case of peri-
scleral canal [33, 34]. Duncan et al. have reported papillary CNVM in a 16-year female with cho-
a 25% (24 of 98 eyes) incidence rate of CNVM in roidal nevus [52]. Optic disc melanocytoma and
pediatric population with ONHD. The majority choroidal osteoma also are associated with
(21 out of 24) were located nasal to disc while the CNVM [6, 53–56]. Kase et al. reported a case of
remaining 3 were located temporally [34]. retinoblastoma with peripapillary CNVM in a
Idiopathic intracranial hypertension (IIH) has 1-year old who underwent enucleation [57].
also been rarely associated with CNVM (0.53%) Congenital, benign lesions like combined
[35–38]. The proposed mechanism is a combina- hamartoma of the retina and retinal pigment epi-
tion of hypoxia due to axonal swelling combined thelium (CHRRPE) are also associated with
with pressure deformity of Bruch’s membrane at CNVM in 6% of eyes which characteristically
the edge of optic nerve head leading to localized exude and leak less compared to CNVM due to
discontinuity. These events lead to the process of other causes [58–60]. The descriptions in previ-
angiogenesis and formation of CNVM [36, 39]. ous reports, however, had no mention of exact
location of CNVM lesions. They are usually occult or type 1 CNVM, and assess choroidal per-
present at the junction of abnormal hamartoma- fusion status in inflammatory choroidal disorders.
tous and normal retina. OCT angiography, there-
fore, plays an important role in identifying OCT Angiography OCT Angiography (OCTA)
CNVM in these eyes [60]. in recent years has evolved as an alternative to
dye-based imaging techniques. Based on identifi-
cation of motion contrast and sequential OCT b
16.9 Multimodal Imaging scans to estimate the decorrelation signals, OCTA
can identify blood flow in the retinal and choroi-
Optical Coherence Tomography Optical dal vessels up to choriocapillaris [63]. This can
Coherence Tomography (OCT) plays an impor- be especially useful in identifying CNVM in con-
tant role in establishing the diagnosis and deter- ditions such as choroidal osteoma, CHRRPE,
mining the disease activity during the follow-up. inflammatory CNVM where FFA may not iden-
AMD-related and idiopathic CNVM are usually tify CNVM due to diffuse RPE or choroidal
located below RPE (type 1) whereas inflamma- changes [55, 60].
tory CNVM are above RPE level (type 2). Sub-
retinal hyper-reflectivity in the form of
hemorrhage, exudation or neovascular membrane, 16.10 Natural History
or sub-RPE hyper-reflectivity such as FVPED can
be identified using structural OCT b scans. Visual prognosis and natural history are influ-
enced by underlying etiology, proximity to pap-
Fundus Fluorescein Angiography Fundus illomacular bundle, and involvement of a
Fluorescein Angiography (FFA) is an invasive number of clock hours of the optic disc. Reports
modality and is able to determine the location of show contrasting results with few showing that
CNVM either at subretinal (classic) or sub RPE visual prognosis is poor if left untreated [5, 11].
(occult) level. FFA in classic CNVM shows well- In contrast, Lin et al. have shown that in eyes
defined margins of CNVM with hyperfluores- with peripapillary CNVM post AMD, observa-
cence in the early phase followed by a late tion with prompt treatment using anti-VEGF
leakage. Occult CNVM is characterized by irreg- agents is a good strategy. Up to 42% of patients
ular hyperfluorescence, ill-defined margins in with fovea sparing SRF or exudation were
early phase and late leakage. AMD-related and observed for periods ranging from months to
idiopathic variety CNVM are predominantly years (median 16 months) [13]. Spontaneous
occult whereas inflammatory CNVM are com- regression of p eripapillary CNVM due to patho-
monly classic [61]. Choroidal nevi associated logical myopia was seen in five eyes with a
CNVM are predominantly classic (20 eyes, 83%) reduction in visual acuity in only one eye due to
[51]. Kies and Bird reported a higher predomi- chorioretinal atrophy around the regressed
nance of occult lesions (93%) in their cohort CNVM [22].
[62]. This is explained by higher incidence of Though previous authors have suggested a
AMD and idiopathic causes of peripapillary conservative approach in cases with extrafoveal
CNVM in their sample. lesions, these studies were done prior to the
widespread availability of anti-VEGF agents [5].
Indocyanine Angiography Indocyanine Angi- Silvestri et al. reported the natural history of peri-
ography (ICGA) like FFA is an invasive diagnos- papillary CNVM in a cohort of 20 eyes. Causes
tic modality and is a gold standard test in the of vision loss were extension of membrane at
diagnosis of polypoidal choroidal vasculopathy subfoveal level, exudation, subretinal fluid, and
(PCV). PCV is characterized by the presence of hemorrhage involving the fovea [5]. They also
branching vascular networks with terminal, aneu- proposed that younger patients (<40 years) of age
rysmal polypoidal dilatations called polyps. had unilateral disease and showed better visual
ICGA may be particularly helpful in cases with prognosis compared to older patients.
As a general rule, inflammatory CNVM tend wide margin of normal-looking retina around
to fare better compared to CNVM related to CNVM [62].
AMD. The possible reasons could be small sized, Laser photocoagulation using argon laser in
classic nature of CNVM in otherwise young, angioid streaks has been associated with either
healthy individuals with overall good health of persistence or recurrence [19]. Shah et al. showed
RPE–Bruch’s complex [27, 64, 65]. efficacy of laser in a case of Inflammatory
CNVM in sarcoidosis [69]. Smaller CNVM
located outside the vascular arcades if treated
16.11 Treatment promptly carry a good prognosis. There is no sig-
nificant utility in treating eyes with large lesion
Multiple treatment options have been described or associated large subretinal hemorrhage. Laser
in the literature including close observation, laser is a treatment of choice in pregnancy when PDT
photocoagulation, photodynamic therapy (PDT), and anti-VEGF agents are not considered safe
anti-vascular endothelial growth factors (VEGF), [70]. However, this is associated with multiple
and rarely, surgical extraction. adverse effects and carries a risk of permanent
scotoma, increase risk of scar expansion, damage
Observation Condition such as IIH where the to papillomacular bundle, vitreous hemorrhage,
papilledema is controlled adequately using medi- and branch retinal artery occlusion [11, 71].
cal or surgical techniques, spontaneous regres- There is also a high risk of recurrence (19–50%)
sion of CNVM has been reported. This approach in literature [6, 62, 67, 72]. In view of no signifi-
of close follow-up is especially useful in cases cant reduction of recurrence, side effect profile,
with localization of CNVM to peripapillary area and availability of better alternatives, laser pho-
alone with no subfoveal extension [35]. tocoagulation is no longer the main treatment
Peripapillary CNVM extending more than 750 μ modality.
from fovea, not involving 1.5 clock hours of tem-
poral peripapillary region may be carefully Photodynamic Therapy Various reports have
observed [66]. However, one should bear in mind shown the utility of PDT with verteporfin
that these CNVM are known to show unpredict- (Visudyne; Novartis Pharmaceuticals
able growth patterns [62]. Corporation, New Jersey) in eyes with peripap-
illary CNVM secondary to AMD, POHS, and
Laser Photocoagulation This was the first idiopathic CNVM [73, 74]. PDT appears to
modality used in clinical practice. A pre-anti- cause a lesser tissue damage compared to laser
VEGF era report suggests no significant differ- [75]. Some authors suggest avoiding PDT in
ence of visual acuity with respect to position or cases with CNVM in close proximity to the
size of CNVM complex [2]. On the other hand, optic disc (within 200 μ) whereas others have
Cialdini et al. reported maintained or improved refuted this proposition and reported no obvi-
visual acuity in 20 eyes (80%) post laser photo- ous damage to optic nerve even after including
coagulation [67]. According to macular photo- the optic nerve within the field of PDT applica-
coagulation study (MPS) guidelines, temporal tion [73, 74, 76]. Pozzoni et al. successfully
1.5 clock hours of temporal retina need to be treated a case of peripapillary CNVM with best
spared to avoid iatrogenic damage to papillo- disease with repeated sessions of reduced flu-
macular bundle. A total of 14% of patients ence PDT [49].
treated with laser photocoagulation lost 6 or Rosenblatt et al. reported at least 2 line visual
more lines of visual acuity compared to 26% of acuity gain in 6 out of 7 eyes with CNVM due to
untreated eyes, the difference was however not AMD, POHS, and a follow-up ranging from 6 to
significant (p = 0.29) [68]. In view of the high 13.5 months [74]. Long-term visual acuity, how-
recurrence rates with laser and possibility of ever, is not known and it is likely that short-term
missing the extent of CNVM, especially in gain may not be maintained over long term.
occult lesions, Kies and Bird proposed to treat a Moreover, there is an additional need for multiple
sessions of PDT along with a risk of iatrogenic Pederson et al. showed anatomical resolution
damage to the optic disc. It is therefore advisable with maintained visual acuity in one eye with
to use PDT in cases nonresponsive to anti-VEGF peripapillary CNVM [88]. Lin et al. treated eyes
agents. with AMD related peripapillary CNVM with a
median of 3 anti-VEGF injections to obtain com-
Submacular Surgery and Macular plete resolution of SRF [13]. This is comparable
Translocation Introduced in 1991 by Thomas to previous reports of 3.96 injections in macular
and Kaplan, submacular surgery was attempted CNVM due to AMD [89].
initially in cases with foveal CNVM [77]. Mansour et al. reported results of bevacizumab
Submacular surgery in peripapillary CNVM in eyes with inflammatory CNVM through
remained an option in large CNVM with exten- 3 months. Though this study had only 8 eyes with
sive exudation where PDT and laser were rela- peripapillary CNVM, the visual outcomes were
tively contraindicated. encouraging in the short term [26]. Figueroa et al.
Atebara et al. reported successful visual out- studied 6 eyes of peripapillary CNVM treated
comes in peripapillary CNVM secondary to with bevacizumab. Five eyes showed complete
POHS confined to extrafoveal location (3 eyes, anatomical resolution with a mean improvement
20/20) and 14 eyes with extension involving sub- of 4 lines [90]. Hoeh et al. reported that among 4
foveal location as well. A total of 7 eyes reached eyes treated with intravitreal bevacizumab
20/40 visual acuity. Recurrence was noted in (mean ± SD:3.5 ± 3.1) through 34 weeks, two lost
4/17 eyes at 32 months [78]. Other authors have visual acuity while one maintained vision [91].
also reported visual and anatomical outcomes Anti VEGF agents can help in resolution of
with CNVM related to histoplasmosis, AMD, CNVM with no damage to papillomacular bun-
and idiopathic nature [79, 80]. Bains et al. in dle. However, need for frequent injections,
their study of 17 eyes with age > 55 years showed absence of level I evidence in eyes with
that only one-third patients had improvement or peripapillary CNVM can pose questions in using
stabilization of visual acuity [80]. Tran et al. anti-VEGF agents as first-line agents.
reported a successful excision of CNVM in case Very few reports exist in literature, which
of melanocytoma with visual acuity of 20/25 at questions the efficacy of anti-VEGF agents and
14 months [54]. prove better outcomes with laser photocoagula-
However, with a host of complications, vari- tion [69]. Munie et al. reported the successful
able outcomes (35% to 88% patients maintain or outcomes of bevacizumab in eyes with CNVM
gain visual acuity) and easy availability of anti- due to choroidal nevi. Though their cohort had 4
VEGF agents, these surgical techniques have eyes with peripapillary CNVM, these eyes were
fallen out of favor [76, 80–84]. not analyzed separately [51]. Other authors have
reported similar successful outcomes in eyes
Anti-VEGF Agents Multiple anti-VEGF agents with CNVM due to melanocytoma with bevaci-
including intravitreal bevacizumab, ranibizumab, zumab [53]. Mansour et al. reported successful
aflibercept, and ziv-aflibercept have been used to outcomes using bevacizumab (2 eyes) and ranibi-
treat CNVM with successful outcomes [69, 85, zumab (one eye) in cases with peripapillary
86]. Representative cases are shown as (Figs. 16.1 CNVM due to choroidal osteoma. Two patients
and 16.2). One major drawback, however, is the had an improvement in vision while one main-
need for multiple injections due to short-term tained visual acuity [56].
action of these agents. Though various landmark
studies showed the importance of anti-VEGF Combination Therapy Studies have also
agents in subfoveal CNVM, Nguyen et al. shown a beneficial effect of combination treat-
reported the anatomical resolution and outcomes ment comprising laser photocoagulation and
of intravenous bevaziumab in patients with peri- anti-VEGF agents [92]. Adrean et al. have
papillary CNVM due to AMD and myopia [87]. reported significant visual and anatomical
a b c
e f
Fig. 16.1 Multimodal imaging of a 58-year-old male the lesion and fovea revealed the presence of intraretinal
with acute onset diminution of vision and a best-corrected and subretinal fluid and subretinal hyperreflectivity sug-
visual acuity (BCVA) of 20/25. Fundus examination gestive of type 2 CNVM (d). A diagnosis of idiopathic
revealed peripapillary subretinal heme and subretinal yel- CNVM was made and patient was treated with intravitreal
lowish lesion superior to disc (a). Fundus fluorescein bevacizumab (IVB). Post 2 IVB injections, BCVA
angiography early and late phase showed early hyperfluo- improved to 20/20. OCT showed resolution of SRF/IRF
rescence and late leak suggestive of classic CNVM (b, c). with fundus photography showing resolution of subretinal
Optical coherence tomography (OCT, line scan) through heme (e, f)
improvement in 5 eyes with peripapillary CNVM 16.12 Conclusions

treated with combination treatment of intravitreal
bevacizumab and laser therapy [92]. Anti-VEGF agents form the first line of therapy
Inflammatory CNVM deserve special mention for the treatment of peripapillary CNVM. Laser
because underlying inflammation may remain photocoagulation and PDT can be used in isola-
active in up to 40% of individuals at the time of tion or as a combination therapy in select condi-
CNVM diagnosis. Therefore, a combination of tions only. Recurrences are common, depend on
steroids (either oral or periocular) may be helpful the underlying pathology and may need pro-
to counteract the inflammation along with anti- longed treatment leading to an overall fair visual
VEGF agents [25]. Tables 16.1 and 16.2 list the prognosis. Young age and inflammatory CNVM
etiologies leading to peripapillary CNVM forma- fare better compared to CNVM in elderly sec-
tion and a brief review of literature respectively. ondary to AMD.
a b c
d e f
g h i
Fig. 16.2 Multimodal imaging of 39-year male with was made. Post five intravitreal bevacizumab injections
diminution of vision in left eye (20/50) and fundus show- over 24 months, visual acuity improved to 20/20. Fundus
ing subretinal heme and lesion temporal to optic disc (a). photograph and AF showed peripapillary atrophy, multi-
Autofluorescence (AF) showed blocked fluorescence due ple hypoAF lesion over posterior pole which was better
to heme, hyperautofluorescence corresponding to lesion delineated on AF (d, e) along with subretinal scar tempo-
and multiple small, dot-like hypoAF suggestive of focal ral to disc which was confirmed on OCT (f). Right eye
RPE loss (b). OCT confirmed the presence of type 2 showed peripapillary scarring and multiple small,
CNVM (c). A presumptive diagnosis of punctate inner hypoAF lesions (g, h) with OCT showing a normal foveal
choroidopathy (PIC) with secondary CNVM in left eye contour (i)
Table 16.1 List of conditions leading to formation of Table 16.1 (continued)

peripapillary choroidal neovascular membrane (CNVM) Sarcoidosis
Degenerative Multifocal Choroiditis with Panuveitis (MFC)
Age-related macular degeneration (AMD) Punctate inner choroidopathy (PIC)
Pathological myopia Eales/Vasculitis
Angioid streaks Serpiginous choroiditis/Serpiginous-like choroiditis
Post-laser photocoagulation Birdshot retinochoroidopathy
Idiopathic Pars planitis
Inflammatory Optic disc anomalies
Presumed ocular histoplasmosis syndrome (POHS) Optic disc drusen
Presumed ocular tuberculosis (POTB) Idiopathic intracranial hypertension (IIH)
Vogt–Koyanagi–Harada syndrome (VKH) Optic nerve coloboma
Table 16.1 (continued) Table 16.1 (continued)

Morning glory syndrome Congenital/benign/malignant lesions of retina and
Optic disc pit choroid
Optic neuritis Choroidal nevus
Tilted disc syndrome Choroidal osteoma
Traumatic choroidal ruptures Malignant melanoma
Dystrophies Retinoblastoma
Best disease Combined Hamartoma of the Retina and Retina
Pattern dystrophy Pigment Epithelium (CHRRPE)
Table 16.2 Compilation of previous studies on peripapillary choroidal neovascular membrane (CNVM) treated using
various modalities
Change in
BCVA
Author Number (logMAR/ Follow-up Treatment
(year) of eyes Pathology Snellen) (mean, months) (mean ± SD) Outcomes
Kies and 55 AMD and 23 ± 6 Laser Recurrence seen in 10
Bird Idiopathic eyes (76.9%) post laser
(1988) and in some eyes outside
[62] the treatment area.
Cialdini 27 AMD (17) and Maintained/ 37.7 Laser Out of 25 treated eyes,
et al. Idiopathic (10) improved anatomical resolution in
(1989) (80%) 23 eyes (92%) and
[67] recurrence in 7 eyes
(28%).
Ruben 20 Idiopathic NA 20 Laser (2 Final BCVA not
et al. sessions) dependent on size/
(1994) [2] location of CNV or
treatment status (lasered
or not)
Blinder 11 AMD Maintained/ 23 Surgical Surgical removal may be
et al. improved removal of helpful in cases
(2005) (64%) CNVM ineligible for PDT and
[83] laser treatment.
Rosenblatt 7 AMD and 20/200– 6–13.5 PDT At least 2 line BCVA
et al. Inflammatory 20/150 to gain was noted in 6 eyes
(2005) (POHS) 20/20–20/80 while one eye
[74] maintained 20/20 vision.
Wendel 6 IIH NA 11a Obs (3), Prevalence was 6 (out of
et al. laser (2), 1140; 0.53%), no
(2005) PDT(1) recurrence was seen
[35] after resolution of
papilledema.
Aisenbrey 8 AMD 20/63–20/40 26 Surgical Surgical extraction in
et al. removal of eyes with extensive
(2007) CNVM CNVM can be helpful to
[76] prevent subfoveal
extension.
Bernstein 7 AMD 0.63 ± 0.31– 3–33 PDT Inclusion of optic nerve
et al. 0.41 ± 0.43 in treatment zone of
(2008) standard-dose
[73] verteporfin PDT may not
lead to optic nerve
damage.
(continued)
Table 16.2 (continued)

Change in
BCVA
Mansour 8b Inflammatory 0.82 ± 0.36– 3 IVB IVB showed anatomical
et al. 0.47 ± 0.38b (1.1 ± 0.4) resolution of CNVM
(2008) over short term (3
[26] months) in varying
etiology of inflammatory
CNVM.
Almony 40 Inflammatory 20/200–20/50 68 Surgical Surgical removal of
et al. (POHS) (median) removal of CNVM may be an
(2008) CNVM alternative in eyes with
[84] extensive CNVM when
laser may be
contraindicated.
Hoeh et al. 4 0.65 ± 0.52 8.5 ± 5 IVB Two patients lost BCVA,
(2008) to 0.5 ± 0.44 (3.5 ± 3.1) while one each improved
[91] and maintained BCVA.
Hamoudi 12 AMD (3) and 59–66 letters 15.9 IVR (6.2) Seven eyes (58%)
et al. Idiopathic (9) achieved complete
(2011) inactivation of CNVM
[93] using IVR.
Nagaoka 11 Pathological 0.47 ± 0.32– 28.3 ± 35.8 Obs (5), TA Incidence of 4.2%
et al. myopia 0.49 ± 0.45 (1), PDT(1), (11/260 eyes), 5 out of
(2011) IVB (4) 11 regressed with no
[22] treatment without any
future recurrence.
Davis et al. 20 AMD (12), 20/40–20/30 13.5 IVB (5.6) Anatomical resolution
(2012) Inflammatory seen in 17 (85%) with a
[32] (6), Trauma (1), mean of 1 line gain of
and Idiopathic visual acuity.
(1)
Mansour 3 Choroidal 0.82 ± 0.42– 15.3 IVB, IVR Significant visual
et al. osteoma 0.53 ± 0.68 (2.67) recovery using
(2014) monotherapy of
[56] anti-VEGF agents.
Duncan 24 ONH drusen 20/15–20/25c Cross- Not Among 24 eyes, 21 eyes
et al. sectional mentioned (87.5%) had nasal
(2016) CNVM whereas
[34] remaining 3 eyes
(12.5%) showed
temporal CNVM.
Grewal 8 Coloboma 21.4 ± 12.1 IVB (2), CNVM was seen as a
et al. laser (1) temporal grayish
(2017) membrane at a temporal
[94] margin of coloboma.
Roy et al. 4 Inflammatory 0.60 ± 0.49– 17.93 ± 14.28 IVB, IVR Monotherapy of
(2017) 0.40 ± 0.49c (2.76) anti-VEGF agents
[27] appears to have an
acceptable response in
inflammatory CNVM.
Adrean 5 AMD (4) 20/50–20/30 31 IVB (3.8) + Combination of IVB and
et al. (median) laser laser led to no recurrence
(2017) in 5 eyes through mean
[92] follow-up of 31 months.
Table 16.2 (continued)

Change in
BCVA
D’souza 8 Inflammatory 0.82 ± 0.96– 19 ± 15.6 IVB (5 Anti-VEGF need to be
et al. 0.51 ± 0.44 eyes), IVR supplemented with
(2018) (3 eyes), immunosuppression in
[25] 2.8 ± 1.3 cases of active
inflammation.
Lin et al. 33 AMD 0.34 ± 0.32– 6–165 IVB/IVA (6, Ten eyes (out of 24;
(2018) 0.60 ± 0.51 3)d 42%) had dry fovea
[13] throughout follow-up.
Ozgonul 13 IIH 8 IVB, IVR, Seven eyes were
et al. IVA observed, and 6 eyes
(2019) were given anti-VEGF
[95] agents. All patients
received oral
acetazolamide.
a
Only for the three eyes treated with laser and PDT
b
Analysis of 8 eyes of peripapillary CNVM among a cohort of 84 eyes with inflammatory CNVM
c
BCVA of the entire cohort
d
Median number of injection at year 1 and 2 respectively; 10 eyes (out of 23) were switched to IVA in view of nonre-
sponse to IVB
BCVA best-corrected visual acuity; logMAR logarithm of minimum angle of resolution; SD standard deviation; AMD
age-related macular degeneration; NA not available; CF counting fingers; PDT photodynamic therapy; IVB intravitreal
bevacizumab; POHS presumed ocular histoplasmosis syndrome; IVR intravitreal ranibizumab; IIH idiopathic intracra-
nial hypertension; IVA intravitreal aflibercept; ONH optic nerve head; Obs observation; TA triamcinolone acetonide
Key Learning Points

• Peripapillary CNVM are a rare cause of References
CNVM with an incidence of 10% and a female
1. Lopez PF, Green WR. Peripapillary subretinal
preponderance. neovascularization. A review. Retina. 1992;12:
• Age-related macular degeneration (AMD), 147–71.
pathological myopia, inflammatory ocular 2. Ruben S, Palmer H, Marsh RJ. The visual outcome of
pathologies, and angioid streaks are the most peripapillary choroidal neovascular membranes. Acta
Ophthalmol (Copenh). 1994;72:118–21.
common pathologies leading to the formation 3. Wilde C, Patel M, Lakshmanan A, et al. The diag-
of peripapillary CNVM. nostic accuracy of spectral-domain optical coher-
• Etiology is not established in up to 40% of ence tomography for neovascular age-related macular
eyes, therefore labeled as idiopathic CNVM. degeneration: a comparison with fundus fluorescein
angiography. Eye (Lond). 2015;29:602–9. quiz 610
• Anti-VEGF agents are the first-line of drug in 4. Flaxel CJ, Bird AC, Hamilton AM, et al. Partial laser
these cases. Laser photocoagulation and PDT ablation of massive peripapillary subretinal neovascu-
can be used as an adjunct in select cases either larization. Ophthalmology. 1996;103:1250–9.
as monotherapy or combination. 5. Silvestri G, Archer DB, Johnston PB. Peripapillary
subretinal neovascular membranes: the natural history.
• Inflammatory CNVM and IIH need treatment Eye (Lond). 1993;7(Pt 3):398–402.
of underlying disease with corticosteroids 6. Browning DJ, Fraser CM. Ocular conditions associ-
and/ or immunosuppressive agents or acet- ated with peripapillary subretinal neovascularization,
azolamide if disease active is still present. their relative frequencies, and associated outcomes.
Ophthalmology. 2005;112:1054–61.
7. Wilde C, Poostchi A, Mehta RL, et al. Prevalence 25. D’Souza P, Ranjan R, Babu U, et al. Inflammatory
of peripapillary choroidal neovascular membranes choroidal neovascular membrane: long-term visual
(PPCNV) in an elderly UK population-the Bridlington and anatomical outcomes after intravitreal anti-
eye assessment project (BEAP): a cross-sectional vascular endothelial growth factor therapy. Retina.
study (2002-2006). Eye (Lond). 2019;33:451–8. 2018;38:1307–15.
8. Sarks SH. New vessel formation beneath the retinal 26. Mansour AM, Mackensen F, Arevalo JF, et al.
pigment epithelium in senile eyes. Br J Ophthalmol. Intravitreal bevacizumab in inflammatory ocular neo-
1973;57:951–65. vascularization. Am J Ophthalmol. 2008;146:410–6.
9. Curcio CA, Saunders PL, Younger PW, et al. 27. Roy R, Saurabh K, Bansal A, et al. Inflammatory
Peripapillary chorioretinal atrophy: Bruch’s mem- choroidal neovascularization in Indian eyes: etiol-
brane changes and photoreceptor loss. Ophthalmology. ogy, clinical features, and outcomes to anti-vascular
2000;107:334–43. endothelial growth factor. Indian J Ophthalmol.
10. Perentes Y, Van Tran T, Sickenberg M, et al. 2017;65:295–300.
Subretinal neovascular membranes complicating uve- 28. Fine SL, Owens SL, Haller JA, et al. Choroidal neo-
itis: frequency, treatments, and visual outcome. Ocul vascularization as a late complication of ocular toxo-
Immunol Inflamm. 2005;13:219–24. plasmosis. Am J Ophthalmol. 1981;91:318–22.
11. Jutley G, Jutley G, Tah V, et al. Treating peripapillary 29. Forrester JV, Liversidge J, Dua HS, et al. Comparison
choroidal neovascular membranes: a review of the of clinical and experimental uveitis. Curr Eye Res.
evidence. Eye (Lond). 2011;25:675–81. 1990;9(Suppl):75–84.
12. Ambati J, Ambati BK, Yoo SH, et al. Age-related mac- 30. Parnell JR, Jampol LM, Yannuzzi LA, et al.
ular degeneration: etiology, pathogenesis, and thera- Differentiation between presumed ocular histoplas-
peutic strategies. Surv Ophthalmol. 2003;48:257–93. mosis syndrome and multifocal choroiditis with
13. Lin T, Dans K, Meshi A, et al. Age-related macular panuveitis based on morphology of photographed
degeneration-associated peripapillary choroidal neo- fundus lesions and fluorescein angiography. JAMA
vascularization in the era of anti-vascular endothe- Ophthalmol. 2001;119:208–12.
lial growth factor therapy. Retina. 2018; https://doi. 31. Shakarchi FI. Ocular tuberculosis: current perspec-
org/10.1097/IAE.0000000000002272. tives. Clin Ophthalmol. 2015;9:2223–7.
14. Shields JA, Federman JL, Tomer TL, et al. Angioid 32. Davis AS, Folk JC, Russell SR, et al. Intravitreal
streaks. I. Ophthalmoscopic variations and diagnostic bevacizumab for peripapillary choroidal neovascular
problems. Br J Ophthalmol. 1975;59:257–66. membranes. Arch Ophthalmol. 2012;130:1073–5.
15. Clarkson JG, Altman RD. Angioid streaks. Surv 33. Auw-Haedrich C, Staubach F, Witschel H. Optic disk
Ophthalmol. 1982;26:235–46. drusen. Surv Ophthalmol. 2002;47:515–32.
16. Mansour AM, Ansari NH, Shields JA, et al. Evolution 34. Duncan JE, Freedman SF, El-Dairi MA. The inci-
of angioid streaks. Ophthalmologica. 1993;207:57–61. dence of neovascular membranes and visual field
17. Shilling JS, Blach RK. Prognosis and therapy defects from optic nerve head drusen in children. J
of angioid streaks. Trans Ophthalmol Soc U K. AAPOS. 2016;20:44–8.
1975;95:301–6. 35. Wendel L, Lee AG, Boldt HC, et al. Subretinal neo-
18. Georgalas I, Papaconstantinou D, Koutsandrea C, vascular membrane in idiopathic intracranial hyper-
et al. Angioid streaks, clinical course, complications, tension. Am J Ophthalmol. 2006;141:573–4.
and current therapeutic management. Ther Clin Risk 36. Sathornsumetee B, Webb A, Hill DL, et al. Subretinal
Manag. 2009;5:81–9. hemorrhage from a peripapillary choroidal neo-
19. Lim JI, Bressler NM, Marsh MJ, et al. Laser treatment vascular membrane in papilledema caused by idio-
of choroidal neovascularization in patients with angi- pathic intracranial hypertension. J Neuroophthalmol.
oid streaks. Am J Ophthalmol. 1993;116:414–23. 2006;26:197–9.
20. Cheung CMG, Arnold JJ, Holz FG, et al. Myopic cho- 37. Belliveau MJ, Xing L, Almeida DR, et al. Peripapillary
roidal neovascularization: review, guidance, and con- choroidal neovascular membrane in a teenage boy:
sensus statement on management. Ophthalmology. presenting feature of idiopathic intracranial hyperten-
2017;124:1690–711. sion and resolution with intravitreal bevacizumab. J
21. Ohno-Matsui K, Kawasaki R, Jonas JB, et al. Neuroophthalmol. 2013;33:48–50.
International photographic classification and grading 38. Lee IJ, Maccheron LJ, Kwan AS. Intravitreal beva-
system for myopic maculopathy. Am J Ophthalmol. cizumab in the treatment of peripapillary choroi-
2015;159:877–883.e877. dal neovascular membrane secondary to idiopathic
22. Nagaoka N, Shimada N, Hayashi W, et al. intracranial hypertension. J Neuroophthalmol.
Characteristics of periconus choroidal neovascu- 2013;33:155–7.
larization in pathologic myopia. Am J Ophthalmol. 39. Morse PH, Leveille AS, Antel JP, et al. Bilateral
2011;152:420–427.e421. juxtapapillary subretinal neovascularization associ-
23. Wong TY, Ferreira A, Hughes R, et al. Epidemiology ated with pseudotumor cerebri. Am J Ophthalmol.
and disease burden of pathologic myopia and myopic 1981;91:312–7.
choroidal neovascularization: an evidence-based sys- 40. Sobol WM, Bratton AR, Rivers MB, et al. Morning
tematic review. Am J Ophthalmol. 2014;157:9–25.e12. glory disk syndrome associated with subretinal neo-
24. Govindhari V, Chhablani J. To laser or not? Indian J vascular membrane formation. Am J Ophthalmol.
Ophthalmol. 2018;66:463–6. 1990;110:93–4.
41. Chuman H, Nao-i N, Sawada A. A case of morning 57. Kase S, Parikh JG, Rao NA. Peripapillary subretinal
glory syndrome associated with contractile movement neovascularization in retinoblastoma. Graefes Arch
of the optic disc and subretinal neovascularization. Clin Exp Ophthalmol. 2008;246:931–4.
Nippon Ganka Gakkai Zasshi. 1996;100:705–9. 58. Shields CL, Thangappan A, Hartzell K, et al.
42. Cennamo G, Rossi C, Velotti N, et al. Ranibizumab in Combined hamartoma of the retina and retinal pigment
the treatment of choroidal neovascularization associ- epithelium in 77 consecutive patients visual outcome
ated with morning glory syndrome. Acta Ophthalmol based on macular versus extramacular tumor location.
(Copenh). 2015;93:e516–7. Ophthalmology. 2008;115:2246–2252.e2243.
43. Özkaya A, Yilmaz I, Alkin Z, et al. Intravitreal ranibi- 59. Echevarria L, Villena O, Nievas T, et al. Combined
zumab in the treatment of choroidal neovasculariza- hamartoma of the retina and retinal pigment epithe-
tion secondary to morning glory syndrome in a child. lium. Anti-VEGF treatment of the associated choroi-
Saudi J Ophthalmol. 2016;30:140–3. dal neovascular membranes. Arch Soc Esp Oftalmol.
44. Dailey JR, Cantore WA, Gardner TW. Peripapillary cho- 2015;90:87–93.
roidal neovascular membrane associated with an optic 60. Gupta R, Pappuru R, Dave V, et al. Choroidal neo-
nerve coloboma. Arch Ophthalmol. 1993;111:441–2. vascularization associated with combined hamartoma
45. Guirgis MF, Lueder GT. Choroidal neovascular mem- of retina and retinal pigment epithelium: multimodal
brane associated with optic nerve coloboma in a imaging. Indian J Ophthalmol. 2018;66:1866–8.
patient with CHARGE association. Am J Ophthalmol. 61. Wolf S, Wald KJ, Remky A, et al. Evolving peripap-
2003;135:919–20. illary choroidal neovascular membrane demonstrated
46. Shaikh S, Trese M. Infantile choroidal neovascular- by indocyanine green choroidal angiography. Retina.
ization associated with choroidal coloboma. Retina. 1994;14:465–7.
2003;23:585–6. 62. Kies J, Bird A. Juxtapapillary choroidal neovas-
47. Duchnowski E, Rodman JA. Juxtapapillary choroidal cularization in older patients. Am J Ophthalmol.
neovascular membrane following optic neuritis. Clin 1988;105:11–9.
Exp Optom. 2010;93:91–4. 63. de Carlo TE, Romano A, Waheed NK, et al. A review
48. Borodic GE, Gragoudas ES, Edward WO, et al. of optical coherence tomography angiography
Peripapillary subretinal neovascularization and (OCTA). Int J Retina Vitreous. 2015;1:5.
serous macular detachment. Association with con- 64. Chan WM, Lai TY, Lau TT, et al. Combined photo-
genital optic nerve pits. Arch Ophthalmol. 1984;102: dynamic therapy and intravitreal triamcinolone for
229–31. choroidal neovascularization secondary to punctate
49. Pozzoni MC, Fine HF, Ferrara DC, et al. Peripapillary inner choroidopathy or of idiopathic origin: one-year
choroidal neovascularization in best disease. Retin results of a prospective series. Retina. 2008;28:71–80.
Cases Brief Rep. 2012;6:176–8. 65. Cohen SY, Laroche A, Leguen Y, et al. Etiology of
50. Li HK, Shields CL, Mashayekhi A, et al. Giant cho- choroidal neovascularization in young patients.
roidal nevus clinical features and natural course in Ophthalmology. 1996;103:1241–4.
322 cases. Ophthalmology. 2010;117:324–33. 66. Reddy H, Loewenstein J, Lane A, et al. Peripapillary
51. Munie MT, Demirci H. Management of choroidal choroidal neovascularization. Invest Ophthalmol Vis
neovascular membranes associated with choroidal Sci (Suppl). 2004;45:3077.
nevi. Ophthalmol Retina. 2018;2:53–8. 67. Cialdini AP, Jalkh AE, Trempe CL, et al. Argon green
52. Tuncer S, Tugal-Tutkun I. Choroidal neovascu- laser treatment of peripapillary choroidal neovascular
larization secondary to choroidal nevus simulat- membranes. Ophthalmic Surg. 1989;20:93–9.
ing an inflammatory lesion. Indian J Ophthalmol. 68. Macular Photocoagulation Study Group. Laser photo-
2013;61:305–6. coagulation for neovascular lesions nasal to the fovea.
53. Al-Halafi AM. Successful treatment of melanocy- Results from clinical trials for lesions secondary
toma associated choroidal neovascular membrane to ocular histoplasmosis or idiopathic causes. Arch
with intravitreal bevacizumab. Saudi J Ophthalmol. Ophthalmol. 1995;113:56–61.
2013;27:117–9. 69. Shah SP, Hubschman JP, Bourges JL, et al. Limited
54. Tran HV, Bovey EH, Uffer S, et al. Peripapillary long-term efficacy of intravitreous anti-VEGF phar-
choroidal neovascularization associated with mela- macotherapy in sarcoidosis complicated by peri-
nocytoma of the optic disc: a clinicopathologic papillary choroidal neovascular membrane. Acta
case report. Graefes Arch Clin Exp Ophthalmol. Ophthalmol. 2010;88:e243–4.
2006;244:1367–9. 70. Schmidt-Erfurth U, Chong V, Loewenstein A, et al.
55. Szelog JT, Bonini Filho MA, Lally DR, et al. Optical Guidelines for the management of neovascular age-
coherence tomography angiography for detecting related macular degeneration by the European Society
choroidal neovascularization secondary to choroidal of Retina Specialists (EURETINA). Br J Ophthalmol.
osteoma. Ophthalmic Surg Lasers Imaging Retina. 2014;98:1144–67.
2016;47:69–72. 71. Lee J, Ferrucci S. Peripapillary subretinal neovascular
56. Mansour AM, Arevalo JF, Al Kahtani E, et al. Role membranes: a review. Optometry. 2011;82:681–8.
of intravitreal antivascular endothelial growth fac- 72. Annesley WH Jr, Shah HG, Mansour AM, et al.
tor injections for choroidal neovascularization due Krypton red laser photocoagulation of peripapillary
to choroidal Osteoma. J Ophthalmol. 2014;2014: subretinal neovascular membranes. Graefes Arch Clin
210458. Exp Ophthalmol. 1986;224:101–5.
73. Bernstein PS, Horn RS. Verteporfin photodynamic ocular histoplasmosis syndrome. Ophthalmology.
therapy involving the optic nerve for peripapillary 2008;115:540–545.e545.
choroidal neovascularization. Retina. 2008;28:81–4. 85. Nochez Y, Le Lez ML, Pisella PJ. Intravitreal injec-
74. Rosenblatt BJ, Shah GK, Blinder K. Photodynamic tions of ranibizumab in treatment of large peripapil-
therapy with verteporfin for peripapillary choroidal lary choroidal neovascularization. J Fr Ophtalmol.
neovascularization. Retina. 2005;25:33–7. 2009;32:25–31.
75. Blumenkranz MS, Bressler NM, Bressler SB, et al. 86. Soliman W, Lund-Andersen H, Larsen M. Resolution
Verteporfin therapy for subfoveal choroidal neovas- of subretinal haemorrhage and fluid after intravit-
cularization in age-related macular degeneration: real bevacizumab in aggressive peripapillary sub-
three-year results of an open-label extension of 2 retinal neovascularization. Acta Ophthalmol Scand.
randomized clinical trials – TAP Report no. 5. Arch 2006;84:707–8.
Ophthalmol. 2002;120:1307–14. 87. Nguyen QD, Shah S, Tatlipinar S, et al. Bevacizumab
76. Aisenbrey S, Gelisken F, Szurman P, et al. Surgical suppresses choroidal neovascularisation caused
treatment of peripapillary choroidal neovascularisa- by pathological myopia. Br J Ophthalmol.
tion. Br J Ophthalmol. 2007;91:1027–30. 2005;89:1368–70.
77. Thomas MA, Kaplan HJ. Surgical removal of sub- 88. Pedersen R, Soliman W, Lund-Andersen H, et al.
foveal neovascularization in the presumed ocu- Treatment of choroidal neovascularization using
lar histoplasmosis syndrome. Am J Ophthalmol. intravitreal bevacizumab. Acta Ophthalmol Scand.
1991;111:1–7. 2007;85:526–33.
78. Atebara NH, Thomas MA, Holekamp NM, et al. 89. Carneiro AM, Mendonça LS, Falcão MS, et al.
Surgical removal of extensive peripapillary choroi- Comparative study of 1+PRN ranibizumab versus
dal neovascularization associated with presumed bevacizumab in the clinical setting. Clin Ophthalmol.
ocular histoplasmosis syndrome. Ophthalmology. 2012;6:1149–57.
1998;105:1598–605. 90. Figueroa MS, Noval S, Contreras I. Treatment of
79. Kertes PJ. massive peripapillary subretinal neovas- peripapillary choroidal neovascular membranes
cularization: an indication for submacular surgery. with intravitreal bevacizumab. Br J Ophthalmol.
Retina. 2004;24:219–25. 2008;92:1244–7.
80. Bains HS, Patel MR, Singh H, et al. Surgical treat- 91. Hoeh AE, Schaal KB, Ach T, et al. Treatment of peri-
ment of extensive peripapillary choroidal neovascu- papillary choroidal neovascularization with intravit-
larization in elderly patients. Retina. 2003;23:469–74. real bevacizumab. Eur J Ophthalmol. 2009;19:163–5.
81. Connor TB Jr, Wolf MD, Arrindell EL, et al. Surgical 92. Adrean SD, Grant S, Chaili S. Bevacizumab (avastin)
removal of an extrafoveal fibrotic choroidal neovascu- and thermal laser combination therapy for peripapil-
lar membrane with foveal serous detachment in age- lary choroidal neovascular membranes. J Ophthalmol.
related macular degeneration. Retina. 1994;14:125–9. 2017;2017:5.
82. Kokame GT, Yamaoka S. Subretinal surgery for peri- 93. Hamoudi H, Sørensen TL. Effect of intravitreal ranibi-
papillary subretinal neovascular membranes. Retina. zumab in the treatment of peripapillary choroidal neo-
2005;25:564–9. vascularisation. J Ophthalmol. 2011;2011:602729.
83. Blinder KJ, Shah GK, Thomas MA, et al. Surgical 94. Grewal DS, Tran-Viet D, Vajzovic L, et al. Association
removal of peripapillary choroidal neovasculariza- of Pediatric Choroidal Neovascular Membranes at the
tion associated with age-related macular degen- Temporal Edge of Optic Nerve and Retinochoroidal
eration. Ophthalmic Surg Lasers Imaging. 2005;36: Coloboma. Am J Ophthalmol. 2017;174:104–12.
358–64. 95. Ozgonul C, Moinuddin O, Munie M, et al.
84. Almony A, Thomas MA, Atebara NH, et al. Long- Management of Peripapillary Choroidal Neovascular
term follow-up of surgical removal of extensive peri- Membrane in Patients With Idiopathic Intracranial
papillary choroidal neovascularization in presumed Hypertension. J Neuroophthalmol. 2019;39:451–7.
(JIPMER), Puducherry and was awarded a gold medal as

the best outgoing student in his postgraduation examina-
tion. He also won the best paper award in clinical sciences
in JIPMER scientific society (JSS awards 2015-16). He
subsequently completed a two-year fellowship in vitreo-
retina and uveitis at LVPEI, Hyderabad.
His main areas of clinical interest include diseases of the
macula and newer imaging modalities such as OCT
angiography, wide-field OCT, FFA, and imaging bio-
markers for various retinal and choroidal diseases. He
has to his credit more than 30 indexed publications
in various national and international journals. He has
co-authored 3 book chapters and 7 non-peer-reviewed
publications with presentations at national and inter-
Sumit completed his basic medical degree and Master of national meetings.
Surgery (MS) in Ophthalmology from Jawaharlal Institute
of Postgraduate Medical Education & Research
in Pediatric Population 17
Şengül Özdek and Hatice Tuba Atalay
Choroidal neovascularization (CNV) is quite rare time of systemic infection or autoimmune trig-
in children and adolescents. In this young popu- ger, may lead to the formation of an atrophic scar
lation, CNV has been reported in association with disruption of Bruch’s membrane. A break in
with infection, inflammation, optic disc anoma- Bruch’s membrane may then lead to the forma-
lies, retinal dystrophies, disruption of Bruch’s tion of a CNV [5].
membrane, choroidal tumors, and trauma. Often,
no cause is found and the CNV is said to be idio-
Table 17.1 Etiology
pathic (Table 17.1) [1–3].
Inflammatory/infectious
Presumed ocular histoplasmosis syndrome
Toxoplasmosis
17.1 Etiology Toxocara canis
Rubella retinopathy
17.1.1 Inflammatory/Infectious Vogt–Koyanagi–Harada syndrome
Chronic uveitis
Inflammatory CNV is the most common type of Sarcoidosis
CNV among children and adolescents. The most Inherited retinal disorders
common cause of CNV secondary to inflamma- Best’s vitelliform macular dystrophy
tion is presumed ocular histoplasmosis syndrome Stargardt disease
Choroideremia
(POHS) [4]. The clinical syndrome is most com-
Disruption of Bruch’s membrane
mon in the United States, which is endemic for
Choroidal rupture
Histoplasma capsulatum. The classical triad of Angioid streaks
POHS is atrophic chorioretinal spots, peripapil- Pathologic myopia-lacquer cracks
lary atrophy, and CNV. The CNV may manifest Optic disc anomalies
at multiple sites, mostly at peripapillary and sub- Congenital optic pits
foveal locations. The pathogenesis of CNV in Optic nerve head drusen
POHS is thought to be nonspecific. Focal infec- Optic nerve coloboma
tion or inflammation of the choroid, either at the Coats’ Disease
Tumors of choroid and RPE
Choroidal osteoma
Combined retinal pigment epithelium: retinal
Ş. Özdek (*) · H. T. Atalay
hamartomas
Ophthalmology Department, Gazi University Medical
School, Ankara, Turkey Idiopathic

204 Ş. Özdek and H. T. Atalay
Another important cause is ocular toxoplasmo- 17.1.3 Disruption of Bruch’s

sis. Although the frequency of CNV in ocular toxo- Membrane
plasmosis in children is uncertain, most case series
of pediatric CNV include this diagnosis [6]. Active Choroidal ruptures due to blunt trauma may
inflammation may lead directly to CNV, particu- occur directly at the site of injury or, more com-
larly in cases of sarcoidosis and syphilis [3]. monly, at a location distant from the site of injury
Childhood CNV has also been reported in associa- due to contrecoup injury. The retina is often
tion with Vogt–Koyanagi–Harada syndrome which spared due to its inherent elasticity, whereas the
is known to have a poor visual prognosis [7]. sclera is spared due to its rigidity. CNV has been
well recognized as an infrequent but very visu-
ally significant complication of choroidal rup-
17.1.2 Inherited Retinal Disorders ture. Longer ruptures and ruptures within the
arcades increase the risk of CNV development. It
CNV has been reported in a number of retinal dys- was noted that 81.2% of patients with CNV after
trophies which may be symptomatic in childhood. choroidal rupture developed it within the first
These are Best disease, Stargardt disease, and year after the injury, suggesting that follow-up
Choroideremia [1, 8]. Visual prognosis of Best should occur more frequently in the first year
disease is usually good despite large vitelliform after trauma [9, 10].
lesions; however, during the disruptive phase, it Angioid streaks result from breaks in a weak-
may be complicated by subretinal hemorrhage and ened Bruch’s membrane, which allows ingrowth
CNV, a major vision-threatening complication [8]. of CNV into the subretinal space, can lead to
Figure 17.1 shows a case with CNV associ- visual impairment. 72–86% of angioid streaks
ated with presumed Stargardt’s disease. cases present with CNV, the spontaneous evolu-
Figure 17.2 shows a case as a representative tion is rare and frequently results in legal blind-
for CNV related to Best disease. ness [11].
a b
Fig. 17.1 A 2-year-old girl who presented with low blocked fluorescence caused by hemorrhage, and staining
vision and esotropia OU (a) Fundus picture (OD), subreti- of subretinal fibrosis in both eyes (c, d). (e, f) Color fun-
nal hemorrhage associated with presumed Stargardt’s dis- dus picture of both eyes after monthly injections of ranibi-
ease with subretinal fibrosis in the macular area of the zumab (3X1) OD. Note the complete resolution of
right eye. (b) Fundus picture (OS), perifoveal subretinal subretinal hemorrhage OD. (g, h) FA 6 months of follow-
fibrosis in the left eye. (c) Late-phase fluorescein angiog- up. Note that there is no late leakage anymore in the RE
raphy of the right eye shows late leakage of CNV in the after anti-VEGF injections
superotemporal and inferonasal areas associated with
17 Choroidal Neovascularization in Pediatric Population 205
c d
e f
g h

a b e
c d f
g h i
Fig. 17.2 An 8-year-old boy presented with decreased no hemorrhage associated with the lesion in the color pic-
visual acuity OU that was recognized by his teacher in ture (i). OCT revealed an increase in subfoveal fluid and
school. Family history was unremarkable. At presenta- new intraretinal cyst formation, which is also in favor of
tion, BCVA was 0.7 in the right eye and 0.6 in the left eye. CNV (j). OCTA, on the other hand, shows the CNV very
(a, b) Fundus examination of the right and left eyes clearly. OCTA is a very valuable tool for the diagnosis of
showed yellow vitelliform lesions in the submacular area CNV associated with Best disease. It is possible to see the
that is typical for Best Disease. (c, d) Fundus autofluores- CNV vessels directly which may not be so clear in fluo-
cence images showing macular hyperautofluorescence in rescein angiography (k). OS underwent intravitreal
both eyes. (e, f) OCT of the right and left eyes revealed the ranibizumab injections 2 times 1 month apart and vision
typical subfoveal hyperreflective material representing the improved to 0.5 and was stabile afterward for a year (l).
vitelliform material. VA of the LE dropped to 0.1 within Color fundus photo of the left eye 1 year after treatment
6 months. It was difficult to be sure that there is an associ- showing a further decrease in the amount of vitelliform
ated CNV by looking at the FA since the staining in late material. (m) OCT shows subfoveal fluid and persistent
stages may be secondary to vitelliruptive stage (g, h). The intraretinal cysts. (n) Vessels of CNV in OCTA were
amount of the vitelliform material was less, but there was much less than before and represented the inactive lesion
j k
l m n
Pathologic myopia can induce CNV, com- Optic nerve coloboma, optic nerve pit, and
monly in older patients, being rare in children. morning glory syndrome are optic nerve anoma-
This may be because of aging which might be lies that are associated with CNV in children [1].
important for the development of predisposing Figure 17.3 shows a case of CNV associated
factors such as retinal pigment epithelium (RPE) with optic disc coloboma.
atrophy and lacquer cracks [12]. However, myo-
pic CNV has been reported in children [13].
Secretan et al. performed a retrospective study of 17.1.5 Coats’ Disease
50 treated and 50 untreated eyes with pathologic
myopia and CNV. The investigators found that Daruich et al. reported their series of 40 Coats’
the eyes had a poor prognosis, but treated eyes patients and found that a subfoveal nodule was
seemed to do better than untreated eyes at 2 years detected in 21 patients (52.5%) [17]. Of 18
of follow-up [14]. Patients with pathologic myo- patients with subfoveal nodules who underwent
pia appear to have a very high recurrence rate. FA, neovascularization was detected in 2 eyes
(11.1%).
Figure 17.4 shows a case of subfoveal nodule
17.1.4 Optic Disc Anomalies with CNV associated with Coats’ Disease.
Optic nerve head drusen (ONHD) are a common,

benign congenital anomaly of the optic nerve. 17.1.6 T
umors of Choroid and Retinal
They have a reported prevalence of 0.4% in chil- Pigment Epithelium
dren [15]. ONHD has been reported to cause
peripapillary CNV which can cause central vision Choroidal osteoma can lead to important vision-
loss by subfoveal progression of the CNV, serous threatening problems. Tumor growth, tumor
macular detachment, or hemorrhage [16]. decalcification, and related CNV can contribute
a b
Fig. 17.3 (a) Color fundus photo of a 2.5-year-old of papillomacular bundle area. (b) FA shows frank late
patient with optic disc coloboma associated with a small leakage indicative of active CNV: arrows indicate CNV
subretinal hemorrhage caused by CNV in the inferior part
a b
c d
Fig. 17.4 A 10-year-old boy with Coats disease. (a) a CNV lesion. (c, d) Early and late frames of FA showing
Color picture shows the submacular exudation and central late leakage associated with subfoveal CNV and typical
fibrosis together with temporal peripheral telangiectatic telangiectatic vessels associated with capillary dropout
vessels. (b) OCT shows the subfoveal fibrotic nodule with areas in the temporal periphery
to substantial visual loss and poor visual acuity. 17.2 Symptoms

Tumors with overlying hemorrhage and irregular
surfaces were at the greatest risk for develop- Children may complain of metamorphopsia and
ment of CNV. Disruption of the RPE and thin- blurred vision, but often they do not inform their
ning or loss of the Bruch membrane and caretakers of any trouble. Fundus findings may
choriocapillaris might contribute to the develop- include a dark gray membrane, subretinal or
ment of CNV [18]. intraretinal hemorrhage, hard exudates, and pig-
mentary changes, most commonly in the macula 17.4 Treatment

or peripapillary region [19].
CNV in the pediatric population is uncommon.
Thus, little has been written in the ophthalmic
17.3 Diagnostic Tests literature concerning the management of CNV in
pediatric patients.
Fluorescein angiography (FA) has always been
considered the gold standard for the diagnosis,
classification, localization, and activity monitor- 17.4.1 Observation
ing of CNVs. Indocyanine green angiography
(ICGA) is crucial for the diagnosis of neovascu- Pediatric CNV is a rare but sight-threatening reti-
lar AMD, due to its ability to image sub-RPE nal disease. CNV in children has a more favor-
structures and to identify the presence and exten- able prognosis than in adults with age-related
sion of the neovascular lesions. The spectral macular degeneration. Goshorn et al. reported
domain OCT (SD-OCT), has radically changed spontaneous involution of CNV in 11 of 19 (58%)
the diagnostic approach to macular diseases by untreated eyes in children; nine patients achieved
offering a noninvasive imaging method, highly visual acuity better than or equal to 20/50 [2].
sensitive in identifying retinal and subretinal Rishi et al. reported spontaneous regression of
abnormalities associated with both neovascular CNV in 15 of 17 untreated eyes in their pediatric
or dry AMD. patients [12]. Observation of CNVs in children
Optical coherence tomography angiography might seem a reasonable approach; but, it is dif-
(OCTA) enables a depth-resolved assessment ficult to predict which CNV will regress or prog-
of the retinal and choroidal blood flow, far ress. Rishi et al. noted that visual outcome in eyes
exceeding the levels of detail commonly with treated CNV was better than in eyes with
obtained with dye angiographies. One of the spontaneously regressed CNV [12].
first applications of OCTA was in detecting the
presence of CNV and establishing its position
in relation to the retinal pigmented epithelium 17.4.2 Laser Photocoagulation
and Bruch’s membrane, and thereby classify-
ing the CNV as type 1, type 2, type 3, or mixed Laser photocoagulation of extrafoveal and juxtafo-
lesions. OCTAs, due to the longer wavelength veal CNV can be utilized in cooperative children in
used by OCT, showed a more distinct CNV a similar fashion used in adults, but it is more chal-
vascular pattern than FA, since there is less lenging in younger children [1]. Wilson and Mazur
suffering from light scattering or is less reported 5 cases of CNV of various causes in chil-
obscured by overlying subretinal hemorrhages dren aged 18 years or younger. Of these patients, 3
or exudation [20]. with extrafoveal CNV had laser photocoagulation,
OCTA is a very invaluable tool for the diagno- and visual acuity improved in 2 patients [4].
sis of CNV associated with Best disease. Since
there is already a hyperfluorescence in FA and
subfoveal fluid in OCT associated with vitellirup- 17.4.3 Photodynamic Therapy
tive stage of Best disease, it may be difficult to
diagnose a CNV lesion without OCTA. However, Photodynamic therapy (PDT) with verteporfin
CNV lesions are very apparent and easily detected can be considered in the pediatric population.
with OCTA (Figs. 17.2 and 17.6). On the other Sodi et al. reported long-term results of PDT in
hand, there may be no intra/subretinal fluid in young patients affected by Best vitelliform mac-
OCT and there may be very subtle changes in FA ular dystrophy (BVMD) complicated by
in eyes with degenerative myopia associated CNV CNV. PDT was a safe procedure in their series,
where OCTA is again very informative. and it was followed by a CNV regression and a
a b c
d e f
Fig. 17.5 An 8-year-old girl presented with decreased space. The apex of this nodular lesion looked in apposi-
visual acuity. Family history was remarkable with similar tion with the neurosensory retina without any intraretinal
complaints in her brother, mother, maternal uncle, and fluid in the right eye (c) and splitting of the RPE with an
grandfather. At presentation, BCVA was 20/400 in the optically empty space was shown in the OCT of both
right eye and 20/32 in the left eye. Fundus examination of eyes. The Arden ratio was abnormal (1.12) in both eyes.
the right eye revealed a submacular grayish vitelliform One session of PDT with verteporfin was performed in
lesion with subretinal hemorrhage at its inferior margin the right eye according to the standard treatment proto-
(a). The left eye showed a vitelliform lesion in the atro- col. BCVA increased to 20/100 in the first month with
phic phase at the macular region. There were some sub- clearing of the subretinal hemorrhage and only minimal
retinal fleck-like lesions in the macular areas in both staining of the lesion in FA, which was maintained
eyes. FA demonstrated subfoveal classic CNV pattern throughout the 26-month follow-up period (d, e). OCT
and blocked hypofluorescence due to the subretinal hem- revealed that the lesion got smaller with only splitting of
orrhage in her right eye (b), and hyperfluorescence asso- the RPE and some optically empty space (f).
ciated with pigment epithelial window defect in her left (Photodynamic therapy for best disease complicated by
eye. OCT of the right eye revealed a subfoveal elevated choroidal neovascularization in children. J Pediatr
hyperreflective nodular lesion at the level of the retinal Ophthalmol Strabismus. 2012;49 (4):216–21. Permission
pigment epithelium (RPE) bulging through subretinal is obtained to use the figures)
consequent improvement in visual acuity that complex. 10 of 12 eyes improved from immedi-
continued to progress even several years after the ate preoperative visual acuity, and four eyes
treatment [21]. developed recurrence of CNV over a mean fol-
We have reported five eyes of four children low-up of 18 months. They concluded that
diagnosed as having BVMD. All eyes responded selected eyes of children with subfoveal CNV
well to one session of PDT with significant and no evidence of membrane regression may
increases in visual acuity in four of the five eyes.benefit from submacular surgery [22].
Increase or stabilization of visual acuity was In a case series of 17 eyes undergoing surgical
reported during a mean follow-up period of removal of CNV of various causes in patients
25 months (Fig. 17.5) [8]. aged 18 years and younger, Uemura et al. reported
that 10 (72%) had improvement of 2 or more
Snellen lines after surgery, and 6 eyes (43%) had
17.4.4 Submacular Surgery final visual acuity of 20/40 or better. They have
noted that removal of these membranes may be a
Submacular surgery for the treatment of CNV viable alternative to laser photocoagulation in
had been popular for a short time in the pre anti- pediatric patients [23]. Nevertheless, in these
vascular endothelial growth factor (Anti-VEGF) series in 25–35% of children had a recurrence of
era. Sears et al. reported 12 eyes of 12 consecu- CNV. As a conclusion, the role of submacular
tive children with subfoveal CNV treated by vit- surgery for the treatment of CNV is very limited
rectomy and excision of the choroidal neovascular in the anti-VEGF era.
17.4.5 Anti-Vascular Endothelial with IVB or IVR for CNV over a mean follow-up
Growth Factor Therapy period of 12.8 months. A mean of 2.2 injections
per eye was required for treatment. An improve-
Case studies have shown that intravitreal bevaci- ment in visual acuity of 3 lines was seen in 22
zumab (IVB) and intravitreal ranibizumab (IVR) eyes (49%), and only 1 eye had worsened vision
are effective in the treatment of CNV associated after treatment [31].
with Best disease, toxoplasmosis, noninfectious Figures 17.1, 17.2, and 17.6 show a case with
uveitis, optic nerve head drusen, optic disc colo- CNV related to Best Disease and treated success-
boma, choroidal rupture, and other etiologies in fully with anti-VEGF injection.
children [24–29]. In a study by Henry et al., IVB Laser photocoagulation, photodynamic ther-
was found to be efficient in the management of apy with verteporfin, and submacular surgery
children with CNV due to different underlying have been employed successfully for the treat-
conditions. The majority of these patients were ment of pediatric CNVs. However, anti-VEGF
successfully treated with a single dose of bevaci- therapy and PDT remains the preferred treatment
zumab. Generally, few injections (mean, 3.6 per options. The need for retreatments is definitely
eye; median, 2) were needed in these patients [30]. much less than adults usually between 1 and 3.
Kozak and colleagues retrospectively ana- But still, the long-term effects of suppressing
lyzed the results of 45 eyes of 39 children treated VEGF in children is uncertain.
a b
Fig. 17.6 A 5-year-old girl presented with decreased the LE. (g, h) OCT revealed subfoveal fluid and some
visual acuity during a routine eye examination. At presen- hyperreflective material OD and a subfoveal elevated
tation, BCVA was 20/20 OD, 20/200 OS. (a) Fundus highly reflective nodular lesion at the level of the RPE
examination of the right eye revealed a vitelliform lesion protruding through the subretinal space associated with
at the macula indicating Best Disease. (b) The left eye subretinal fluid OS. This young girl was diagnosed as Best
showed a submacular yellow vitelliform lesion associated Disease in vitelliruptive stage bilaterally, which is compli-
with subretinal hemorrhage around the vitelliform mate- cated with CNV OS. (i) Fundus picture of the LE 7 months
rial. (c, d) FAF images show splitted hyperautofluores- after intravitreal ranibizumab injections (3×). Her BCVA
cence of macula more in the RE. (e, f) FA demonstrated a in the LE improved to 20/60, the hemorrhage resolved
mild hyperfluorescence in the macular area secondary to totally. (j) OCTA revealed inactive CNV and (k) OCT
vitelliruptive stage of Best disease in the RE and a frank showed contraction of the lesion and a decrease in the
hyperfluorescent leakage consistent with subfoveal CNV subretinal fluid
and blocked fluorescence around it due to hemorrhage in
c d
e f

i j

Key Points 12. Rishi P, Gupta A, Rishi E, Shah BJ. Choroidal neo-
vascularization in 36 eyes of children and adolescents.
1. Choroidal neovascularization (CNV) is quite Eye (Lond). 2013;27(10):1158–68.
rare in children and adolescents. 13. Potter MJ, Szabo SM, Ho T. Combined photody-
2. CNV has been reported in association with namic therapy and intravitreal triamcinolone for the
infection, inflammation, optic disc anomalies, treatment of myopic choroidal neovascularization in a
13-year-old girl. Graefes Arch Clin Exp Ophthalmol.
and retinal dystrophies. 2006;244(5):639–41.
3. Often, no cause is found and the CNV is said 14. Secretan M, Kuhn D, Soubrane G, Coscas G. Long-
to be idiopathic. term visual outcome of choroidal neovascularization
4. FA, OCT, recently OCTA are ancillary diag- in pathologic myopia: natural history and laser treat-
ment. Eur J Ophthalmol. 1997;7:307–16.
nostic tests. 15. Erkkilä H. Clinical appearance of optic disc drusen
5. CNV in children has a more favorable progno- in childhood. Albrecht Von Graefes Arch Klin Exp
sis than in adults with age-related macular Ophthalmol. 1975;193(1):1–18.
degeneration. 16. Basdekidou C, Caputo G. Optic nerve abnormali-
ties in pediatric ocular diseases. In: Hartnett ME,
6. Anti-VEGF therapy and PDT remains the pre- Trese M, Capone A, BJB K, Steidl SM, editors.
ferred treatment options. Pediatric retina. 2nd ed. Philadelphia: Wolters
Kluver, Lippincott Williams & Wilkins; 2014.
p. 148–72.
References 17. Daruich AL, Moulin AP, Tran HV, Matet A, Munier
FL. Subfoveal nodule in Coats’ disease: toward an
updated classification predicting visual prognosis.
1. Sivaprasad S, Moore AT. Choroidal neovascularisa- Retina. 2017;37(8):1591–8.
tion in children. Br J Ophthalmol. 2008;92(4):451–4. 18. Shields CL, Sun H, Demirci H, Shields JA. Factors
2. Goshorn EB, Hoover DL, Eller AW, Friberg TR, predictive of tumor growth, tumor decalcification,
Jarrett WH, Sorr EM. Subretinal neovascularization choroidal neovascularization, and visual outcome in
in children and adolescents. J Pediatr Ophthalmol 74 eyes with choroidal osteoma. Arch Ophthalmol.
Strabismus. 1995;32(3):178–82. 2005;123(12):1658–66.
3. Spaide RF. Choroidal neovascularization in younger 19. Hollander DA, Stewart JM. Acquired and other retinal
patients. Curr Opin Ophthalmol. 1999;10:177–81. diseases. In: Lambert SR, Lyons CJ, editors. Taylor &
4. Wilson ME, Mazur DO. Choroidal neovascularization Hoyt’s pediatric ophthalmology and strabismus. 4th
in children: report of five cases and literature review. J ed. Philadelphia: Elsevier; 2013. p. 522–9.
Pediatr Ophthalmol Strabismus. 1988;25(1):23–9. 20. Lupidi M, Cerquaglia A, Chhablani J, Fiore T, Singh
5. Prasad AG, Van Gelder RN. Presumed ocular his- SR, Cardillo Piccolino F, et al. Optical coherence
toplasmosis syndrome. Curr Opin Ophthalmol. tomography angiography in age-related macular
2005;16(6):364–8. degeneration: the game changer. Eur J Ophthalmol.
6. Mavrikakis E, Levin AV, Lam WC. Choroidal neovas- 2018;28(4):349–57.
cularization secondary to congenital toxoplasmosis in 21. Sodi A, Murro V, Caporossi O, Passerini I, Bacci GM,
an infant. Can J Ophthalmol. 2010;45(6):e11–2. Caputo R, et al. Long-term results of photodynamic
7. Ketata A, Ben Zina Z, Hajji D, Sellami D, Abdelkefi therapy for choroidal neovascularization in pediatric
A, Kharrat W, et al. Two cases of subretinal neovascu- patients with best Vitelliform Macular Dystrophy.
lar membrane in Vogt-Koyanagi-Harada syndrome. J Ophthalmic Genet. 2015;36(2):168–74.
Fr Ophtalmol. 2006;29(3):302–6. 22. Sears J, Capone A Jr, Aaberg T Sr, Lewis H,
8. Ozdek S, Ozmen MC, Tufan HA, Gurelik G, Grossniklaus H, Sternberg P Jr, et al. Surgical man-
Hasanreisoglu B. Photodynamic therapy for best agement of subfoveal neovascularization in children.
disease complicated by choroidal neovasculariza- Ophthalmology. 1999;106(5):920–4.
tion in children. J Pediatr Ophthalmol Strabismus. 23. Uemura A, Thomas MA. Visual outcome after sur-
2012;49(4):216–21. gical removal of choroidal neovascularization in
9. Secretan M, Sickenberg M, Zografos L, Piguet pediatric patients. Arch Ophthalmol. 2000;118(10):
B. Morphometric characteristics of traumatic cho- 1373–8.
roidal ruptures associated with neovascularization. 24. Kohly RD, Muni RH, Kertes PJ, Lam
Retina. 1998;18:62–6. WC. Management of pediatric choroidal neovascular
10. Patel MM, Chee YE, Eliott D. Choroidal rupture: a membranes with intravitreal anti-VEGF agents: a ret-
review. Int Ophthalmol Clin. 2013;53(4):69–78. rospective consecutive case series. Can J Ophthalmol.
11. Tilleul J, Mimoun G, Querques G, Puche N, Zerbib 2011;46:46–50.
J, Lalloum F, et al. Intravitreal ranibizumab for cho- 25. Rich R, Vanderveldt S, Berrocal AM, Mavrofrides
roidal neovascularization in angioid streaks: four year EC, Murray TG, Gregori N. Treatment of choroi-
follow-up. Retina. 2016;36(3):483–91. dal neovascularization associated with Best’s dis-
ease in children. J Pediatr Ophthalmol Strabismus. 29. Piermarocchi S, Benetti E, Fracasso G. Intravitreal
2009;46:306–11. bevacizumab for posttraumatic choroidal neovascu-
26. Baillif S, Nguyen E, Colleville-El Hayek A, Betis larization in a child. J AAPOS. 2011;15:314–6.
F. Long term follow-up after single intravitreal 30. Henry CR, Sisk RA, Tzu JH, Albini TA, Davis JL,
ranibizumab injection for choroidal neovascular- Murray TG, et al. Long-term follow-up of intravitreal
ization secondary to optic nerve head drusen in a bevacizumab for the treatment of pediatric retinal and
5-year old child. Graefes Arch Clin Exp Ophthalmol. choroidal diseases. J AAPOS. 2015;19(6):541–8.
2013;251:1657–9. 31. Kozak I, Mansour A, Diaz RI, Calzada JI, Pichi F,
27. Vayalambrone D, Misra A. Paediatric choroidal oste- Cruz-Villegas V, et al. Outcomes of treatment of pedi-
oma treated with ranibizumab. BMJ Case Rep. 2012; atric choroidal neovascularization with intravitreal
https://doi.org/10.1136/bcr-2012-007446. antiangiogenic agents. Retina. 2014;34:2044–52.
28. Benevento J, Jager RD, Noble AG, Latkany P, Mieler
WF, Sautter M, et al. Toxoplasmosis-associated
neovascular lesions treated successfully with ranibi-
zumab and antiparasitic therapy. Arch Ophthalmol.
2008;126:1152–6.
Polypoidal Choroidal
Vasculopathy: Diagnostic 18
and Therapeutic Considerations
Adrian H. C. Koh
18.1 Background Importantly, although PCV appears to be some-

what more refractory to monotherapy with anti-
Age-related macular degeneration (AMD) is the vascular endothelial growth factor (VEGF)
leading cause of blindness in elderly people glob- agents than typical wet AMD (i.e., polyps may
ally and it is estimated that 288 million people remain), clinical trials do not always distinguish
will be affected by the year 2040 [1]. Early-stage between these cases; thus, the efficacy of treat-
AMD is characterized by drusen and retinal pig- ments for PCV remains the subject of debate. The
ment epithelial (RPE) changes and typically does disparities in epidemiology, clinical presentation,
not cause vision loss. Late-stage AMD, generally and treatment response suggest that there may be
classified as geographic atrophy (“dry”) AMD or differences in the underlying pathophysiology of
neovascular (“wet”) AMD, accounts for 90% of PCV and typical wet AMD.
vision loss in AMD.
There is increasing evidence that wet AMD is
not a homogeneous entity and that a significant 18.2 Definition and Diagnosis
proportion of patients (8–13% of white
Caucasians and 20–54% of Asians) have a sub- PCV appears to be a subtype of wet AMD, pre-
type known as polypoidal choroidal vasculopathy senting differently to typical wet AMD. The cur-
(PCV) [2–5]. rent understanding that PCV is not a separate
Clinically, PCV appears to affect younger disease entity is based on several lines of evi-
individuals than typical wet AMD and frequently dence [6]. First, almost 30% of AMD cases
presents with large subretinal or intraretinal hem- develop polypoidal lesions over time; second, in
orrhage, but often with minimal scarring. many cases, ICGA identifies polyps in lesions
Definitive diagnosis is based on the use of indo- classified as wet AMD; third, some risk factors
cyanine green angiography (ICGA), which shows for PCV (e.g., smoking) appear to be similar to
characteristic polyp-like hyperfluorescent those for wet AMD [7]; fourth, histopathological
lesions, possibly associated with a branching vas- analyses have located the vascular network and
cular network (BVN). However, other features of polyp between Bruch’s membrane and the RPE;
PCV are not well defined, thus complicating the and finally, there are emerging data that show that
interpretation of the epidemiological data. the two conditions have largely similar genetic
profiles [8].
A. H. C. Koh (*) Clinically, PCV is more likely than typical
Eye & Retina Surgeons, Camden Medical Centre, AMD to present with recurrent serous or
Singapore, Singapore

218 A. H. C. Koh
h emorrhagic detachments of the RPE and/or vitre- made if focal hyperfluorescent lesions were
ous hemorrhage, but less likely to present with observed on confocal scanning laser ophthalmo-
fibrous scarring. In wet AMD, the classic appear- scope (cSLO)-based ICGA and at least one of the
ance of the choroidal neovascularization (CNV) is following angiographic or clinical criteria was
a dark gray lesion on clinical examination and a present: BVN; “pulsatility”; nodular appearance
“sea fan-like network” seen in the early phases of when viewed stereoscopically; hypofluorescent
fundus fluorescein angiography (FA). In contrast, halo; orange subretinal nodule on color photo-
PCV appears as orange-red, bulging polyp-like graph; or associated massive submacular hemor-
dilations with submacular hemorrhage in the fun- rhage. Initial studies suggest that the two systems
dus, and with an FA pattern more likely to be that are largely similar, with the EVEREST study cri-
of occult than classic CNV. While PCV was origi- teria having slightly higher specificity than the
nally believed to arise from the inner choroidal vas- use of subretinal focal hyperfluorescence alone
cular network, spectral-domain and en-face optical [12]. This is a critical area for future research.
coherence tomography (OCT) imaging has con-
firmed that the entire complex (consisting of the
BVN as well as the polypoidal nodules) arises 18.2.2 Indocyanine Green
above the Bruch’s membrane and below the base- Angiography (ICGA)
ment membrane of the RPE, consistent with type 1
or “occult” CNV [9]. This accounts for why the ICGA, either flash-based or cSLO-based, is con-
most common fluorescein angiographic pattern is sidered the gold standard for the diagnosis of
as “occult” lesions, with granular hyperfluores- PCV because it provides imaging of PCV-specific
cence in the early phase and multiple indistinct features that are predominantly sub-RPE and
leakage spots in the later phase. As FA is unable to therefore poorly visualized using FA. Detailed
distinguish between CNV and PCV, the gold-stan- choroidal vessel morphology can be visualized, as
dard diagnosis of PCV is based on the use of ICGA, indocyanine green binds plasma protein and so
which shows polyp-like hyperfluorescent lesions remains within the vascular compartment instead
possibly with a BVN and other less-well-defined of leaking from the choriocapillaris, hence pro-
supporting features (see the section on Imaging). viding later images with higher definition than FA
[13, 14]. Moreover, indocyanine green is imaged
using infrared light, which can effectively pene-
18.2.1 Diagnostic Criteria for PCV trate the RPE and, to some extent, hemorrhage;
this is advantageous because PCV primarily
Unfortunately, there is no consensus on ICGA involves the sub-RPE space just above the Bruch’s
diagnostic criteria, although there are currently two membrane [13]. Although the emitted light is of
widely used systems: the Japanese Study Group lower intensity than with fluorescein, the use of
guidelines and the criteria from the EVEREST trial the cSLO for ICGA yields images of high resolu-
[10, 11]. The Japanese Study Group guidelines tion [15]. Studies comparing digital fundus pho-
propose that a diagnosis of “definite” PCV should tography with cSLO showed both to be similarly
be made when the classic features of elevated effective at detecting PCV [16, 17]. A recent study
orange-red lesions on fundus examination and/or comparing the individual ICGA features reported
polypoidal lesions on ICGA are seen; a diagnosis that both fundus photography and cSLO-based
of “probable” PCV should be made if only the systems could detect PCV in the presence of typi-
abnormal BVN is seen on ICGA or if there are cal nodular lesions (seen in over 80% of cases);
recurrent hemorrhagic and/or serous RPE detach- however, cSLO is superior for detecting addi-
ments without the features of “definite” PCV. tional features, particularly the BVN [16].
A more elaborate set of criteria was used by Furthermore, other diagnostic features, such as
the central reading center in the EVEREST clini- pulsatility of nodules, can only be visualized
cal trial [11], where a diagnosis of PCV was using the dynamic cSLO videoangiography mode.
18 Polypoidal Choroidal Vasculopathy: Diagnostic and Therapeutic Considerations 219
PCV exhibits several distinctive angiographic 18.2.3 Fluorescein Angiography (FA)

features on ICGA. Polyps exhibit nodularity when
viewed stereoscopically and appear early in the Eyes with PCV generally do not display any spe-
ICGA (within the first 6 min) [13]; however, focal cific features on FA; therefore, FA alone is not
hyperfluorescence may also indicate “pseudo- usually adequate to confirm or exclude the diag-
polyp” lesions (such as macroaneurysms or a reti- nosis of PCV. Most eyes with PCV display an
nal pigment epithelium defect associated with occult leakage pattern, or avascularized and hem-
choroidal vascular knuckle). To add to the speci- orrhagic large pigment epithelial detachments
ficity for polyps, several additional features have (PEDs) on FA [21, 22], which is to be expected
been used under the EVEREST criteria [18]. A because most of the vascular network is located
distinctive hypofluorescent halo around the pol- under the RPE. Erosion of the overlying RPE
ypoidal lesions can be seen in about 45–70% of may lead to a variable degree of window defect.
cases [11, 16]. In many cases, polyps are associ- Occasionally, a “classic CNV” may coexist with
ated with a BVN with vascular dilation at the bor- PCV [3, 22]. While FA is not diagnostic of PCV,
ders, which appears as a hyperfluorescent dilated it is still useful for evaluating treated and recur-
complex of sub-RPE vessels that leaks in the lat- rent cases, where fluorescein leakage may come
ter stages of the angiogram [13]. However, visual- from the BVN, either alone or together with
ization of the BVN may not always be possible, residual polypoidal lesions. This distinction
especially using flash-based ICGA [16]. between BVN and polyp leakage has important
The Japanese Study Group of PCV describes therapeutic implications (see Sect. 18.7.5).
masses of grape-like clusters on ICGA that start
as small nodules in the early phases and gradu-
ally expand into the late phase [10]. However, 18.2.4 Spectral Domain
this feature is less common than the typical nodu- OCT, Enhanced-Depth OCT,
lar presentation (only 19% of PCV cases) [16]. In and Swept Source OCT
the late ICGA phases, the core of the lesion may Imaging
become hypofluorescent owing to the washout of
the dye, staining the polyps in a ring-like appear- The most distinctive OCT feature of the polypoi-
ance. Polyps that are leaking will be visible as a dal nodules in PCV is a sharply angulated
ring of hyperfluorescence, whereas polyps that inverted V-shaped elevation of the RPE, separat-
are not leaking will appear hypofluorescent [13]. ing the RPE from the underlying relatively intact
Pulsation of blood flow in polyps may be visual- Bruch’s membrane; this sign has also been
ized with dynamic cSLO ICGA; this is thought to dubbed “thumb-like protrusion” in literature. The
be a unique feature of PCV. BVN appears as irregular in the space between
Choroidal vascular hyperpermeability is a the RPE and Bruch’s membrane, giving rise to
commonly noted feature in PCV [19] and is visu- the “double-layer” sign. Kim et al. observed both
alized during the middle and late phases of ICGA geographic hyperfluorescence and the double-
as multifocal hyperfluorescence with blurred layer sign in approximately 80% of PCV eyes,
margins, distant from the main PCV lesion [19, indicating that these features are sensitive mark-
20]. In the very late stages of ICGA (45–60 min), ers for the diagnosis of PCV [19, 23]. Visualization
one may observe “ghost vessels,” where dye has of the cross-section of the choroid by OCT and
emptied out of the larger choroidal vessels into enhanced-depth imaging (EDI) suggests that the
the surrounding choroidal interstitial tissue caus- subfoveal choroid may be thicker in PCV than in
ing choroidal vessel silhouetting. Late geographic typical wet AMD, potentially because of the
hyperfluorescence is another characteristic find- greater incidence of choroidal vascular hyperper-
ing in PCV, distinct from choroidal vascular meability; furthermore, Maruko et al. reported
hyperpermeability. This represents the staining of that PCV eyes with ICGA choroidal hyperperme-
the entire lesion complex, comprising the BVN ability tend to have a greater subfoveal choroid
as well as the polypoidal lesions [19]. thickness than those without [24]. It is now
220 A. H. C. Koh
widely accepted that PCV is part of the pachy- persistent retinal detachment may lead to second-
choroid spectrum of diseases. Therefore, EDI- ary effects and fluorescein leakage will become
OCT may be a useful noninvasive tool to help more diffuse, resembling PCV and making dif-
differentiate between PCV and typical AMD sub- ferential diagnosis difficult [26]. Some atypical
types. Recently, swept-source OCT has also cases of PCV may also appear very similar to
shown value in demonstrating the polypoidal CSC, presenting with serous detachment of the
lesion and BVN, using both cross-sectional cuts retina and polypoidal lesions that resemble small
and en-face imaging [25]. PEDs on evaluation by FA [26]. In these cases
ICGA is useful to confirm the presence or absence
of multifocal areas of choroidal hyperpermeabil-
18.2.5 U
se of ICGA for Differential ity, which cannot be viewed using FA alone.
Diagnosis These areas are known to be indicative of CSC
but absent in PCV [26].
ICGA is important, not only to confirm a diagno- High-speed video ICGA with a cSLO is useful
sis of PCV but also to distinguish it from other for distinguishing between PCV and RAPs. RAP
lesions such as vascularized PED, central serous lesions can be distinguished by the presence of
chorioretinopathy (CSC), and retinal angioma- intraretinal small dilated blood vessels within the
tous proliferation (RAP). The sensitivity and macula, which are seen to turn sharply from the
specificity for PCV is further increased by gath- inner retina to the choroidal interface when stereo
ering all data from multimodal imaging, includ- pairs of images are viewed; these lesions are
ing FA and OCT (see Box 18.1). often associated with large PEDs and prominent
Vascularized PED appears on ICGA as late, intraretinal cystic fluid [15].
non-nodular hyperfluorescence (also called “hot
spots” or “plaques”) within a notch in the margin
of a serous PED. There is typically no associated 18.3 Epidemiology
BVN, and the lesion is never pulsatile.
Distinguishing between PCV and CSC can be Over the past three decades there have been sig-
difficult in atypical cases. Typically, CSC patients nificant advances in our understanding of the epi-
present with exudative retinal detachment, mul- demiology of AMD, both in white populations
tiple small serous PEDs, and patches or tracks of and other races [1]. However, epidemiological
pigment epithelial atrophy. An FA showing pin- data on PCV remain limited and those that exist
point leaks at the level of the RPE can confirm are complicated by inconsistent classification and
the diagnosis of CSC. However, in some instances diagnosis of the disease between populations. As
PCV can only be definitively diagnosed using
ICGA, most studies have not distinguished it
Box 18.1 Features Highly Suggestive of PCV from typical wet AMD [27–29]. Moreover, the
Features highly suggestive of PCV prevalence of PCV is likely to be significantly
• Orange-red polyp-like dilations (viewed by underestimated because ICGA is often not used
routine ophthalmoscopy and contact lens for routine assessment [30, 31].
slit-lamp biomicroscopy) PCV was first described more than two
• Large (>2000 μm in diameter) hemorrhagic decades ago predominantly in black females [32–
and exudative PED
37]. Although more common in those of African
• Chronic CSC
or Asian descent, PCV is now known to affect
CSC central serous chorioretinopathy, PCV polyp-
oidal choroidal vasculopathy, PED pigment epithe- individuals of all races [3, 6, 38, 39]. Studies of
lial detachment. Japanese, Korean, and Chinese patients report a
prevalence of 20–54% among patients with wet
AMD [3–5, 40, 41], while the corresponding fig- involvement in PCV or the associated risk fac-
ure in Caucasian populations is about 8–13% [21, tors. Ueta et al. followed up 125 Japanese patients
42, 43]; the prevalence may be higher in patients with PCV, while Kim et al. evaluated 47 Korean
presenting with large hemorrhagic and exudative patients; the cumulative incidence of second-eye
neurosensory detachments [37, 44, 45]. involvement was 11% at 3 years and 19% at
Very few studies have examined the incidence 2 years, respectively [54, 55].
and prevalence of PCV in the general, seemingly
healthy, population. In a study based on health
check-ups in 10,890 participants in Korea, only 18.3.4 Risk Factors
nine patients with exudative AMD received fluo-
rescein angiography or ICGA; of these, two eyes 18.3.4.1 Ocular
(22%) were found to have PCV [46]. While the presence of drusen is a significant risk
factor for wet AMD [53, 56], they appear to be
less common in Asians and in patients with PCV
18.3.1 Demographics (27–33%) [43, 57]. Thus, the role of drusen in the
pathogenesis of PCV is unclear.
In addition to differences in prevalence, the Other ocular risk factors for PCV include a
demographics of PCV in different races also history of CSC, presence of PED [45], and cho-
show notable variability [6, 38, 39]. In African roidal hyperpermeability [26, 58]. Ahuja et al.
and Caucasian patients, a female preponderance proposed that PCV might develop as a conse-
has been reported [33, 38], whereas in Asian pop- quence of chronic CSC [58], a theory supported
ulations, PCV patients are more likely to be male by advances in choroidal imaging [20, 59, 60]
[2–5, 40, 47, 48]. There are studies which suggest and the finding that patients with PCV are more
that PCV might affect younger people compared likely to have a history of CSC than those with
to those with wet AMD [3, 6]. typical AMD (15 vs. 3% of patients) [61].
18.3.4.2 Systemic
18.3.2 Lesion Location Hypertension has long been regarded as a poten-
tial risk factor for PCV [47, 62]; there have been
In African and Caucasian patients, polypoidal case reports of hypertensive retinopathy and
lesions appear to have a predilection for the peri- macroaneurysms with coexisting PCV lesions
papillary area [21, 33, 34, 37, 42, 43]. In Asian [63], and histopathological studies reporting hya-
patients, however, macular PCV is more preva- linization of choroidal vessels similar to that seen
lent, being found in more than 63% [2, 40, 49]. in arteriosclerosis [64]. However, not all studies
have shown an association [2].
Cigarette smoking, a significant risk factor for
18.3.3 Laterality wet AMD, has also been shown to increase the
risk of PCV with a similar effect size (odds ratio
Bilateral involvement is common in wet AMD 4.4 for PCV and 4.9 for AMD) [7].
(over 42% of patients by 5 years after presenta-
tion with unilateral disease) [50–53]. However,
the frequency of bilateral PCV is less well- 18.4 Pathology and Genetics
studied and reports so far have been inconsistent,
giving figures ranging from 7 to 80% [2–5, 21, 18.4.1 Histopathology
32–34, 37, 40, 47–49]. The incidence of bilateral
PCV appears to be higher in Caucasian patients In the mid-1990s, several clinicopathological
than Asian patients [39]. Few prospective studies studies of suspected PCV in enucleated eyes
have examined the incidence of second-eye described extensive fibrovascular proliferation
222 A. H. C. Koh
within the subretinal space and Bruch’s mem- [69] found strong VEGF positivity. Potential
brane [65, 66]. These studies were probably explanations for the disparity include possible
describing the end-stage of PCV. Okubo and col- differences in the amount of scarring, age-related
leagues noted a hyaline-like appearance of the atherosclerotic changes, or type of PCV (see
vessel walls in a surgically excised specimen Sect. 18.5.2). Tong et al. (2006) found higher lev-
while it was still active [67]. This finding was els of aqueous VEGF in eyes with PCV compared
confirmed by Kuroiwa and colleagues [68]. with normal controls, albeit at lower levels than
Nakashizuka and colleagues [64] further char- in wet AMD [70].
acterized these lesions in surgical specimens that
met the diagnosis criteria of PCV based on ICGA,
and documented extensive exudative change and 18.4.2 Genetics
hyalinization of vessels. Hyalinized vessels were
characterized by extravasation of plasma protein The underlying genetics of AMD and PCV are
and deposition of basement membrane-like mate- largely similar, although some discrepancies
rial, and the smooth muscle component was have been reported. For example, the strongest
replaced by amorphous pseudo-collagenous tis- association of AMD in the CFH gene was
sue of a poorly defined nature. The authors sug- through single nucleotide polymorphism
gested that these histopathological lesions were (SNP) rs1061170 (Y402H); however, its role in
similar to retinal arterial macroaneurysms and PCV is rather unclear. This discrepancy could
thus hypertension may be a common risk factor be due to ethnic differences in the frequency of
(see Sect. 18.3.4.2). the CFH risk allele, which is uncommon in
In a case that included both CNV and PCV Asian populations [71]. A recent meta-analysis
portions in the same specimen, the CNV portion found that another variant on CFH, rs800292
showed granulation tissue proliferation, support- (I62V), was strongly associated with PCV in
ing the concept that CNV represents a stereo- Asian patients [72].
typic, nonspecific wound-repair response. In A separate comprehensive meta-analysis
contrast, little (if any) fibrosis or granulation tis- found that variants at ARMS2/HTRA1
sue proliferation was observed in the PCV por- (rs10490924/rs11200638), CFH (rs1061170,
tion. Hyalinized vessels in the CNV portion rs3753394, rs1329428, rs1410996, and rs800292)
contained pericytes that were immunoreactive for and the complement pathway (C2; rs547154)
alpha-smooth muscle actin (SMA), while vessels were significantly associated with PCV, with the
in the PCV portion were negative for alpha-SMA highest odds ratio observed at variants in ARMS2/
expression, suggesting that the smooth muscle HTRA1. When comparing PCV and wet AMD,
cells of the choroidal vessels had disappeared in only variant rs10490924 of ARMS2/HTRA1
the PCV portion. Furthermore, immunoreactive potentially implicated a genetic and biological
expressions of CD34 revealed discontinuity in difference between these two disease entities.
the vascular endothelium in the PCV portion. The risk genotypes of rs10490924 were associ-
Both of these findings were indicative of vascular ated with larger lesion size, a greater chance of
damage similar to that seen in hypertension. The vitreous hemorrhage, and worse therapeutic
hyalinized vessels represented the choroidal vas- response in PCV [73].
culature with arteriosclerotic changes, similar to The non-synonymous SNPs rs10490924
the hyalinization seen in other parts of the body. (A69S) within ARMS2, and rs11200638 within
As such, the authors concluded that these vessels HTRA1, at 10q26 have been tested for disease
did not represent neovascularization [64]. susceptibility for both AMD and PCV. Due to a
The relative role of VEGF in the pathogenesis strong linkage disequilibrium across this region,
of PCV is variable: while Nakashizuka et al. [64] it is difficult to genetically distinguish these two
observed no VEGF expression in vascular endo- susceptible candidates and prioritize their impor-
thelial cells of PCV specimens, Matsuoka et al. tance toward disease pathogenesis [6].
Other studies have also examined the associa- 18.5.2 Classification of PCV Subtypes
tion of a number of genetic polymorphisms with
PCV in different ethnicities, but a recent meta- Just as the classification of PCV itself is chal-
analysis of these associations gave inconclusive lenging, so too is the lack of a universally
results [74]. accepted system of classification for PCV
subtypes.
One commonly proposed subtyping of PCV is
18.5 Classification by the presence or absence of a BVN. Type 1
PCV describes the disease in which ICGA shows
At present, there is no generally recognized clas- the polypoidal lesions to be fed and drained by an
sification scheme in which PCV is defined. The overt BVN and is characterized by CNV rapidly
lack of a clear classification for PCV—either expanding beneath the RPE with terminal dilata-
within or outside the wet AMD spectrum—con- tions at the end of the branching vascular net-
founds available epidemiological data, and com- work. The BVN typically displays one of two
plicates diagnosis and management. distinct patterns, either an “umbrella” pattern
where vessels radiate from the feeder vessel in all
directions, or a “rake” pattern in which the ves-
18.5.1 Classification Within CNV sels extend from the feeder vessel in one direc-
tion. In contrast, the type 2 PCV variant shows
Current opinion, based on published evidence either a faint BVN or none at all [75–77].
and clinical experience, is that rather than being a Type 1 PCV, thought to be the less common
discrete clinical entity, PCV is a subtype of wet type, has been shown to be associated with the
AMD. However, it is difficult to place PCV ARMS2 A69S genetic polymorphism, which cor-
within the current classification for CNV, in relates with a larger lesion size and more limited
which disease is categorized according to lesion effect of photodynamic therapy (PDT) [75, 76].
location (extrafoveal, juxtafoveal, or subfoveal) Type 1 PCV is associated with a smaller mean
and subfoveal leakage pattern (occult or classic) subfoveal thickness than type 2 PCV [78].
based on FA findings. This issue must be viewed The type 1/2 classification system described
alongside the wider limitations of the current above is not widely used, as interpretation can be
classification for CNV. While the present scheme subjective and interobserver agreement has not
is practical when considering laser photocoagula- been tested. Other groups have suggested classi-
tion—where one needs to be able to direct ther- fying PCV lesions by polyp size (largest diameter
apy accurately—it is of questionable value in the of polyp), location (peripapillary, subfoveal, jux-
era of anti-VEGF therapy, where lesion location tafoveal, or extrafoveal), formation (single, clus-
is not important for treatment decisions. ter, or string), and a number of discrete polyp
Accordingly, updating the CNV classification areas (single or multiple) [49]. It is also possible
system to improve its value for modern treatment to classify PCV based on its primary features, i.e.,
modalities would be a worthwhile undertaking. quiescent, exudative, or hemorrhagic [13, 79].
In doing so, it would be pertinent to account for However, despite the interest in PCV sub-
all possible variants of CNV, including PCV. types, evidence relating to the prognostic value of
One option for a new CNV classification sys- these schemes, and their value in guiding treat-
tem would be to define neovascularization as ment, remain limited.
sub-RPE (e.g., occult type 1 CNV or PCV), sub- Recent guidelines for the diagnosis and treat-
retinal (e.g., classic type 2 CNV), or intraretinal ment of PCV described a more pragmatic
(e.g., RAP), factors that have implications for approach that could be applied in the clinic. In an
prognosis and response to anti-VEGF treatment. algorithm for the diagnosis of PCV, the disease
Using this classification system, PCV would rep- was classified as either active or inactive, with
resent a polyp-exhibiting form of sub-RPE CNV. active disease defined by a list of features reflect-
224 A. H. C. Koh
ing a likely symptomatic impact and/or possible nant presenting pattern, either exudative or
need for treatment (see Sect. 18.7.3) [13]. hemorrhagic. In the exudative pattern, serous
As PDT and focal laser remain possible treat- RPE or retinal detachment and lipid deposits may
ment options in addition to anti-VEGF therapy continue for a long time in a stable condition, fol-
(see Sect. 18.7), lesion location (extrafoveal, jux- lowed by eventual degeneration and atrophy of
tafoveal, or subfoveal) also remains an important the RPE and deterioration in vision [83]. In the
consideration. hemorrhagic pattern, hemorrhagic detachments
of the RPE and/or submacular hemorrhage
(sometimes acute) can occur, leading to sudden
18.6 Natural History visual deterioration. In these cases, massive reti-
and Prognosis nal hemorrhage may even lead to breakthrough
vitreous hemorrhage, which may resolve sponta-
18.6.1 Natural History neously but can recur. Although these patterns
may change over the course of follow-up, ulti-
In contrast to wet AMD, the natural history of mately, recurrent hemorrhage leads to degenera-
PCV is highly variable and poorly understood tion of the RPE and outer retina, and loss of
[50, 51]. vision.
Yannuzzi et al. described a prolonged, relaps- Based on ICGA findings, polyps with a “clus-
ing, and remitting course including multiple ter of grapes” configuration (up to two-thirds of
serosanguinous detachments of the RPE [33]. PCV cases) are usually associated with an aggres-
Long-term vision may be preserved, and fibrous sive disease course, with or without treatment [2,
proliferation and scarring are uncommon com- 49, 80, 83, 85]. Bessho et al. followed up 42 eyes
pared with wet AMD; however, significant vision with PCV over 12 months and reported that eyes
loss can occur through the development of with “clustered” PCV experienced significant
chronic atrophy or vitreous hemorrhage [33, 35]. deterioration in mean visual acuity, while in those
Other studies have described PCV as a rela- showing “nonclustered” PCV, vision was largely
tively benign condition [47, 80–82], remaining maintained [80]. Polyps that show pulsations
stable in 50% of eyes for 2 or more years without (viewed by video ICGA) may also exhibit high
treatment [83]. In a study by Yamaoka et al., disease activity, potentially being associated with
despite persistent serous retinal detachment in extensive hemorrhage [86–88].
11% of eyes with PCV but no coexisting classic Reports have suggested that eyes with better
CNV, 89 and 83% had stable or improved vision initial vision (≥20/40) tend to maintain this
without treatment at 3 and 12 months, respec- vision at 1 year [47] and 2 years after diagnosis
tively [47]. Furthermore, although patients with [83]. In a retrospective review of 110 eyes by
PCV have been reported to present later than Okubo et al., smaller lesions and a shorter dura-
those with wet AMD (average 21.2 vs. tion of disturbed vision were associated with a
14.8 months from onset of symptoms to first lower rate of decline in visual acuity than larger
evaluation), they generally have better visual lesions and a longer history of disturbed vision
acuity [2]. Spontaneous involution of polyps may [89]. In another study of 88 eyes, those with
occur [33, 83, 84]. larger PCV complex size (≥1 disc area; n = 66)
showed significant progression in the mean area
of the lesion, with marked deterioration of vision
18.6.2 Predictive Factors during follow-up, whereas eyes with smaller
PCV complex size (<1 disc area; n = 22) often
18.6.2.1 Clinical and Angiographic showed minimal progression with limited exuda-
Features tive change and tended to maintain their initially
As discussed in the Sect. 18.5.2, many investiga- good vision [90]. Furthermore, severe vision-
tors have broadly divided PCV by its predomi- threatening complications, such as suprachoroi-
dal hemorrhage, vitreous hemorrhage, and RPE of leakage is the BVN, essentially behaving as
tears, were only seen in eyes with larger PCV typical occult CNV [96]. Lesion location should
complexes [90]. be based on ICGA findings, and the location of
Coexisting type 2 CNV has been described in the polypoidal lesions and the BVN should be
a small proportion of patients with PCV, and is determined separately. FA and ICGA also serve
associated with a worse prognosis than PCV to delineate the different components of the PCV
alone [3, 22, 47, 91]. lesion: FA provides better information on leakage
from the BVN, while ICGA is required to visual-
ize the polypoidal lesions themselves. This has a
18.7 Management significant bearing on the selection of treatment
modality in recurrent and persistent cases (see
18.7.1 Goals of Treatment Sect. 18.7.5).
The primary goal of treatment in PCV, as in

other forms of wet AMD, is the improvement 18.7.3 When to Treat
and subsequent maintenance of vision. This can
be achieved through the reduction and resolu- All active PCV should be treated as soon as pos-
tion of exudation from the two components of sible. PCV may be considered active if there is
PCV: the polyps themselves and/or the BVN clinical, OCT, FA, or ICGA evidence of any of
[92]. Due to the relapsing and remitting course, the following [13]:
which is common in eyes with PCV, an impor-
tant goal of PCV management that is different • Symptoms of visual loss and metamorphopsia
from typical wet AMD is the complete closure • Subretinal fluid and/or intraretinal fluid (with
of polyps, confirmed on ICGA. This second- or without PED)
ary objective is especially important in cases • Subretinal hemorrhage
with recurrent or persistent fluid at the macula • Fluorescein leakage on FA
despite multiple repeated anti-VEGF treatments • Polyps and other features on ICGA
for active polyps [93, 94]. Persistent polyps are
associated with a high risk of massive submacu- The fellow eye should also be monitored
lar hemorrhage, disciform fibrosis, and RPE carefully for PCV because a significant propor-
atrophy of the macula, leading to severe and tion of patients may have bilateral disease, even
irreversible visual loss. Breakthrough vitreous though bilaterality has been reported to be lower
hemorrhage in eyes with PCV may have devas- than in wet AMD. If polyps are present in the
tating complications, including secondary glau- fellow eye, the decision to treat should be based
coma and phthisis bulbi [95]. on disease activity and presence or absence of
symptoms [13].
18.7.2 Imaging Requirements

for Management 18.7.4 Treatment Options
In addition to diagnosis (see Sect. 18.2), OCT Visual morbidity from PCV arises from the activ-
(and to a lesser extent angiography) are invalu- ity of polyps and the BVN (bi-component dis-
able for guiding the management of PCV. It is ease), and as such there are several management
important to emphasize that FA alone is insuffi- options to be considered, each of which is dis-
cient and inaccurate in diagnosing PCV and in cussed below. A summary of the recommended
guiding its management. However, when ICGA treatment approach is provided in Box 18.2.
reveals leakage in the absence of visible polyps, Representative cases are shown in Figs. 18.1,
the FA is essential in confirming that the source 18.2, and 18.3.
226 A. H. C. Koh
Box 18.2 Summary of PCV Management

Summary of PCV management
• Visual morbidity from PCV arises from the activity of polyps and the BVN (bi-component disease).
• Polyp closure is more effectively achieved by PDT than with ranibizumab or bevacizumab, while
reduction in fluid, visual gains, and control of leakage from BVN are best achieved by anti-VEGF
treatment. Intravitreal aflibercept monotherapy may offer polyp closure rates similar to PDT.
• In cases with good initial visual acuity (e.g., 20/40 or better), anti-VEGF monotherapy may be initiated,
with PDT added only if there is a persistent activity from active polyps; otherwise, the current treatment of
choice is PDT–anti-VEGF combination treatment to the entire lesion, as delineated by ICGA in the first
instance.
• Recurrences should be treated based on the primary source of leakage: if polyps are still present, further
PDT applied selectively to only the polyps may be administered; if no polyps are present, and leakage is
ascertained to be originating only from the BVN, anti-VEGF monotherapy is the treatment of choice.
• Modified PDT protocols may reduce the potential complications of PDT such as submacular hemorrhage
and RPE atrophy.
• Ongoing research will demonstrate whether anti-VEGF monotherapy is adequate to achieve and maintain
visual acuity gains and polyp closure.
BVN branching vascular network, ICGA indocyanine green angiography, OCT optical coherence tomography,
PCV polypoidal choroidal vasculopathy, PDT photodynamic therapy, RPE retinal pigment epithelial, VEGF
vascular endothelial growth factor.
Fig. 18.1 Case 1: A 70-year-old Chinese male presented vasculopathy (upper row). The patient underwent combi-
with reduced central vision with metamorphopsia for nation therapy (PDT and ranibizumab) and responded
2 months. He had a history of three intravitreal aflibercept well (middle row—left OCT). However, he had a recur-
injections at monthly intervals in the same eye; however, rence after 4 months (middle row—middle OCT) and
he did not notice any significant improvement in vision. underwent three monthly ranibizumab injections with any
On examination, fundus examination showed pigment improvement (middle row—right OCT). Repeat ICG
epithelial detachment (PED). Angiography and OCT showed polyps and underwent combination therapy which
imaging confirmed the presence of polypoidal choroidal led to recovery (bottom row)
a Angiography (Superficial) Angiography (Deep) Angiography (Outer retina) Angiography (Chorioc apillaris)
OCT B-Scan OCT Projection Fundus
Comments: Signature: Date:
b Angiography (Superficial) Angiography (Deep) Angiography (Outer retina) Angiography (Chorioc apillaris)
OCT B-Scan OCT Projection Fundus
Comments: Signature: Date:
Fig. 18.2 (a) OCT angiography of case 1 at presentation raphy of case 1 at the last follow-up. Note the persistent and
shows polyp and BVN. (b) OCT angiography of case 1 at increased flow in BVN; therefore, the patient was asked to
the time of recurrence shows polyp and BVN. OCT angiog- continue with aflibercept on treat-and-extend regimen
18.7.4.1 Laser Photocoagulation BVN. The limitations of laser photocoagulation

Laser photocoagulation only has a role in ablat- include a high rate of recurrence, formation of
ing extrafoveal (preferably extramacular) pol- new lesions, enlargement of photocoagulation
yps [97]. Rarely, it may be effective in treating scars, tear or rip of the RPE, as well as iatro-
the entire complex of extrafoveal polyps and genic CNV. Accordingly, rates of visual loss
228 A. H. C. Koh
Fig. 18.3 Case 2: A 74-year-old female had a history of firmed the presence of polyps and she underwent combina-
intravitreal 3 aflibercept then 3 ranibizumab at monthly tion therapy and improved from 6/60 to 6/24 with resolution
intervals without any improvement. Imaging studies con- of disease activity (bottom row: color photo and OCT)
have been reported to be as high as 54% [98]. Studies comparing visual outcomes have
Adjunct anti-VEGF in addition to focal laser shown that anti-VEGF therapy gives better
has been shown to improve the outcome in results than PDT [104]. Potential complications
extrafoveal PCV [99]. of PDT include acute vision decrease, increased
subretinal hemorrhage, and RPE tears [105,
18.7.4.2 Verteporfin (Visudyne™) 106]. Longer-term limitations of PDT monother-
Photodynamic Therapy apy include reduction of vision, recurrence of
Treatment of PCV with PDT has been shown to polyps, persistence of BVN activity, secondary
be effective in the closure of polyps in PCV. One CNV, and progressive RPE atrophy [107]. In
prospective study demonstrated that, unlike 20–40% of cases, secondary CNV occurs within
CNV, treatment of PCV with PDT resulted in a 2 years [108–110]. Repeated full-fluence PDT
mean improvement in visual acuity of 8 letters may lead to persistent choroidal ischemia,
at 12 months, with 25% of patients experiencing upregulation of VEGF, and RPE atrophy, all of
significant gains in vision of at least 3 Snellen which contribute to unsatisfactory long-term
lines; only 8% of patients lost 3 or more Snellen visual outcomes [111]. A recent systematic
lines of vision [100]. In addition, the vast major- review and meta-analysis reported stable visual
ity of patients (86%) had cessation of fluores- outcomes for up to 2 years of PCV, but worsen-
cein leakage at 1 year [100, 101]. Similarly, in ing by 3 years, particularly in eyes that experi-
the EVEREST trial, the mean change in visual enced recurrence [112].
acuity in the PDT-treated group was +7.5 letters
at 6 months. The rate of complete closure of 18.7.4.3 Anti-VEGF Monotherapy
polyps was 71%, and that of partial but incom- Anti-VEGF therapy such as intravitreal afliber-
plete closure was 86% [11]. The treatment-free cept (EYLEA®), ranibizumab (Lucentis®), and
period in patients receiving PDT for PCV may off-label bevacizumab (Avastin®), is currently the
also be significantly longer than in those receiv- gold-standard treatment for wet AMD, with over
ing the same treatment for CNV in AMD [101]. 95% of patients avoiding moderate visual loss,
However, PDT is relatively ineffective in resolv- and about 30% showing a doubling of the visual
ing activity arising from the BVN, with the angle [113–115]; this effect can be maintained
majority of cases remaining unchanged despite over many years with adequate monitoring and
repeated treatment [102, 103]. retreatment [116]. Ranibizumab monotherapy
has been associated with similar visual outcomes action that may be synergistic. While PDT may
in PCV [11, 117, 118]. One indication for anti- be effective for polyp regression, anti-VEGF
VEGF therapy in PCV is for patients with good therapy improves vision by reducing both exuda-
presenting visual acuity [119]; indeed, current tion from polyps and BVN activity. Anti-VEGF
Japanese guidelines recommend anti-VEGF therapy also mitigates the post-PDT upregulation
monotherapy for patients presenting with rela- of VEGF that is caused by choroidal ischemia,
tively good visual acuity of 20/32 or better [120]. thereby reducing the risk of the acute vision
Despite providing improvements in visual acu- decrease sometimes observed with PDT mono-
ity, the polyp closure rate with bevacizumab and therapy; it might also minimize the risk of post-
ranibizumab monotherapy remains low. In the PDT subretinal hemorrhage [131, 132]. The
EVEREST trial, complete closure of polyps at EVEREST study demonstrated that the PDT–
6 months with ranibizumab therapy was 28% ranibizumab combination was superior to ranibi-
compared with over 70% with PDT, while the par- zumab monotherapy in terms of polyp closure
tial polyp closure rates were 43% in the ranibi- rate while achieving comparable short-term
zumab group and over 83% in the PDT-treated (6-month) visual outcomes [11]. Similar results
arms [11]. In a retrospective study of three initial have been observed in patients receiving PDT
monthly injections of bevacizumab, only 1 of 11 with bevacizumab [133]. Moderate vision gains
cases showed closure of polyps on the ICGA were achieved in the monotherapy arm of the
[121]. In addition, while bevacizumab was effec- recently presented EVEREST II study (n = 322)
tive in controlling exudation, the size of the BVN [134]. The ranibizumab and PDT combination
remained unchanged. What is undisputed, how- arm of EVEREST II achieved greater vision
ever, is that bevacizumab therapy is effective in gains and a rate of polyp regression twice that of
reducing the activity of the BVN [110, 122]. the ranibizumab monotherapy arm. However, in
Intravitreal aflibercept can be used for effec- studies and real-life evidence generated to date,
tive management of PCV, with evidence from intravitreal aflibercept as monotherapy has shown
VIEW 2 for probable PCV cases suggesting sim- similar vision gains and polyp regression rates to
ilar VA benefits to “standard” wet AMD [123]. the ranibizumab and PDT combination arm of
The recently presented results of the PLANET EVEREST II [125, 135, 136]. In the recently pre-
study (n = 318) of intravitreal aflibercept in PCV sented PLANET study of intravitreal aflibercept
demonstrate that intravitreal aflibercept mono- in PCV, similar results were achieved using both
therapy achieves significant visual acuity gains aflibercept monotherapy and combination ther-
over 1 year and resolves polyp activity in the apy (aflibercept + rescue PDT). The majority of
majority of patients when treated with the patients achieved good vision gains when treated
approved regimen for wet AMD [124]. Very few with the approved regimen for wet AMD, with
patients met the criteria for rescue PDT, suggest- very few patients requiring PDT rescue therapy
ing intravitreal aflibercept monotherapy is a safe [124]. Therefore, the use of intravitreal afliber-
and efficacious treatment in the vast majority of cept has the potential to negate the need for PDT
PCV cases. Furthermore, several case series have and associated costs and side effects.
reported superior polyp closure with intravitreal The long-term visual results of combination
aflibercept than is reported for ranibizumab, and therapy are yet to be established. Some studies
intravitreal aflibercept has also been used suc- indicate that while visual outcomes are good in
cessfully to treat ranibizumab tachyphylaxis in the first 6–12 months, there is a significant drop-
patients with PCV [125–130]. off into the second and third years [137, 138]. A
meta-analysis reported greater visual improve-
18.7.4.4 Combination Therapy: PDT ment in eyes treated with a combination of PDT
with Anti-VEGF Agents and anti-VEGF (ranibizumab or bevacizumab)
There are sound biologic and scientific rationales than in those treated with PDT monotherapy,
for combining PDT with anti-VEGF therapy, as although the beneficial effect may be lost with
both therapies employ different mechanisms of longer (2-year) follow-up [112]. Two-year results
230 A. H. C. Koh
from the ongoing EVEREST II and PLANET A systematic review reported recurrence rates
studies will provide more data on the long-term after PDT with or without anti-VEGF (any agent)
effects of PDT combination therapy in conjunc- of 6–50% after 1 year, 9–83% after 2 years, and
tion with a treat-and-extend regimen. Another 40–79% after 3 years of follow-up [112].
potential advantage of combination with PDT is
that the number of anti-VEGF injections may be
reduced without compromising visual acuity 18.8 Research Questions
gains [139].
Owing to concerns regarding potential compli- It is evident from the review of the current litera-
cations of PDT, several authors have used modi- ture that there remain numerous important ques-
fied PDT protocols (e.g., reduced-fluence PDT) tions surrounding PCV and its management.
alongside anti-VEGF agents, with the aim of
reducing the potential complications associated
with PDT [140–143]. These protocols are yet to 18.8.1 Questions Regarding
be tested against conventional full-fluence PDT– Epidemiology
anti-VEGF combination regimens. However, in and Pathophysiology
recurrent cases, it seems pertinent to minimize
the laser treatment size by targeting only polyps • What is the absolute definition of PCV and
(guided by ICGA), rather than the entire lesion what are the key features for diagnosis?
complex [144–146]. • What is the true prevalence and incidence of
PCV?
• What are the risk factors associated with PCV,
18.7.5 Treatment in Cases and how do they compare with those for wet
of Persistent Activity and/or AMD?
Recurrence • What are the underlying genetics in PCV
and how they relate to phenotypic pre-
Retreatment criteria should be based on disease sentation, type of PCV, and response to
activity as indicated by reduced visual acuity, treatment?
macular thickening from fluid, and leakage from • How should PCV be classified in order to
polyps and/or the BVN [13]. The treatment predict prognosis and inform management
modality will then depend on the predominant decisions?
cause. It is generally agreed that persistent polyps
should be selectively treated with PDT in combi-
nation with anti-VEGF therapy, while activity 18.8.2 Questions Regarding
from the BVN alone should be managed using Treatment
anti-VEGF monotherapy, since PDT has little
effect on the activity or size of the BVN and • What is the role of combination therapy in
might accelerate RPE atrophy and photoreceptor PCV and should the different treatment
cell loss. However, it is worth considering poten- modalities be administered simultaneously or
tial differences in anti-VEGF agents, as intravit- sequentially?
real aflibercept may be associated with a greater • Is it sufficient to initiate therapy with anti-
likelihood of polyp regression than ranibizumab; VEGF monotherapy or would first-line com-
indeed, intravitreal aflibercept monotherapy has bination with PDT be better?
recently been reported to induce polyp regression • Will combination therapy with PDT help
in cases that are refractory to ranibizumab treat- improve outcomes in patients whose vision is
ment. A key objective in retreatment is to mini- not improving with anti-VEGF monotherapy?
mize complications from PDT by reducing the • Are there significant differences in polyp clo-
frequency, spot size, and possibly fluence [146]. sure rates by the different anti-VEGF agents?
• Does reducing fluence, especially in repeat References

treatment, reduce long-term complications of
PDT (especially RPE atrophy, macular hem- 1. Wong WL, Su X, Li X, et al. Global prevalence
of age-related macular degeneration and disease
orrhage, and RPE rips)? burden projection for 2020 and 2040: a system-
• Should treatment-naive and recurrent cases be atic review and meta-analysis. Lancet Glob Health.
approached differently? 2014;2(2):e106–16.
• Does recurrence indicate undertreatment? 2. Sho K, Takahashi K, Yamada H, et al. Polypoidal
choroidal vasculopathy: incidence, demographic fea-
tures, and clinical characteristics. Arch Ophthalmol.
2003;121(10):1392–6.
18.8.3 Questions Regarding 3. Maruko I, Iida T, Saito M, et al. Clinical character-
Outcome istics of exudative age-related macular degeneration
in Japanese patients. Am J Ophthalmol. 2007;144:
15–22.
• Can we identify distinct types of PCV that are 4. Liu Y, Wen F, Huang S, et al. Subtype lesions of
associated with better or worse outcomes, and neovascular age-related macular degeneration in
should treatment be tailored to the prognosis Chinese patients. Graefes Arch Clin Exp Ophthalmol.
2007;245(10):1441–5.
(e.g., if the risk of persistent BVN activity is 5. Byeon SH, Lee SC, Oh HS, et al. Incidence and clini-
high, perhaps these patients should receive cal patterns of polypoidal choroidal vasculopathy in
proactive rather than reactive therapy)? Korean patients. Jpn J Ophthalmol. 2008;52(1):57–62.
• What is the expected long-term visual out- 6. Laude A, Cackett PD, Vithana EN, et al. Polypoidal
choroidal vasculopathy and neovascular age-related
come for PCV patients? macular degeneration: same or different disease? Prog
• How does angiographic polyp closure corre- Retin Eye Res. 2010;29(1):19–29.
late with visual outcome? 7. Cackett P, Yeo I, Cheung CM, et al. Relationship of
smoking and cardiovascular risk factors with polyp-
oidal choroidal vasculopathy and age-related macular
degeneration in Chinese persons. Ophthalmology.
18.9 Summary 2011;118(5):846–52.
8. Cheng CY, Yamashiro K, Jia Chen L, et al. New
PCV is increasingly recognized as a major sub- loci and coding variants confer risk for age-related
macular degeneration in East Asians. Nat Commun.
type of AMD, estimated to affect nearly 1 million 2015;6:6063.
people in China alone, and common elsewhere in 9. Miura M, Makita S, Iwasaki T, et al. Three-dimensional
Asian populations [147]. While it is historically visualization of ocular vascular pathology by optical
considered to be a concern only in Asian popula- coherence angiography in vivo. Invest Ophthalmol Vis
Sci. 2011;52(5):2689–95.
tions, the condition is now thought to be preva- 10. Japanese Study Group of PCV. Criteria for diagnosis
lent globally and may be grossly underestimated of polypoidal choroidal vasculopathy. Nippon Ganka
in non-Asian populations. The current lack of Gakkai Zasshi. 2005;109(7):417–27.
disease understanding limits optimal diagnosis 11. Koh A, Lee WK, Chen LJ, et al. EVEREST study:
efficacy and safety of verteporfin photodynamic ther-
and treatment. apy in combination with ranibizumab or alone versus
This chapter brings together the current evi- ranibizumab monotherapy in patients with symp-
dence for PCV and provides expert comment on tomatic macular polypoidal choroidal vasculopathy.
the available data. Moreover, the paper identifies Retina. 2012;32(8):1453–64.
12. Cheung GCM, Laude A, Wong W, et al. Improved
gaps in our knowledge, which further research is specificity of polypoidal choroidal vasculopathy diag-
required to address. nosis using a modified EVEREST criteria. Retina.
Because PCV is considered to be a subtype of 2015;35(7):1375–80.
wet AMD it should be managed using the cur- 13. Koh AH, Chen LJ, Chen SJ, et al. Polypoidal choroidal
vasculopathy: evidence-based guidelines for clinical
rently approved therapies; however, optimal diagnosis and treatment. Retina. 2013;33(4):686–716.
treatment regimens have yet to be determined. 14. Desmettre T, Devoisselle JM, Mordon S. Fluorescence
Further research is necessary to improve our properties and metabolic features of indocyanine green
understanding of PCV and inform management (ICG) as related to angiography. Surv Ophthalmol.
2000;45(1):15–27.
decisions.
232 A. H. C. Koh
15. Royal College of Ophthalmologists. Age-related in Japanese and French patients: multicenter diag-
macular degeneration: guidelines for management. nosis with multimodal imaging. Am J Ophthalmol.
London: Royal College of Ophthalmologists; 2013. 2014;158:309–18.e302.
16. Cheung CM, Lai TY, Chen SJ, et al. Understanding 32. Yannuzzi LA, Wong DW, Sforzolini BS, et al.
indocyanine green angiography in polypoidal choroi- Polypoidal choroidal vasculopathy and neovas-
dal vasculopathy: the group experience with digital cularized age-related macular degeneration. Arch
fundus photography and confocal scanning laser oph- Ophthalmol. 1999;117(11):1503–10.
thalmoscopy. Retina. 2014;34:2397–406. 33. Yannuzzi LA, Ciardella A, Spaide RF, et al. The
17. Gomi F, Sawa M, Mitarai K, et al. Angiographic lesion expanding clinical spectrum of idiopathic polyp-
of polypoidal choroidal vasculopathy on indocyanine oidal choroidal vasculopathy. Arch Ophthalmol.
green and fluorescein angiography. Graefes Arch Clin 1997;115:478–85.
Exp Ophthalmol. 2007;245(10):1421–7. 34. Spaide RF, Yannuzzi LA, Slakter JS, et al. Indocyanine
18. Lim TH, Laude A, Tan CS. Polypoidal choroidal vas- green videoangiography of idiopathic polypoidal cho-
culopathy: an angiographic discussion. Eye (Lond). roidal vasculopathy. Retina. 1995;15:100–10.
2010;24(3):483–90. 35. Perkovich BT, Zakov ZN, Berlin LA, et al. An update
19. Kim JH, Kang SW, Kim TH, et al. Structure of polyp- on multiple recurrent serosanguineous retinal pig-
oidal choroidal vasculopathy studied by colocalization ment epithelial detachments in black women. Retina.
between tomographic and angiographic lesions. Am J 1990;10:18–26.
Ophthalmol. 2013;156(5):974–80.e972. 36. Stern RM, Zakov ZN, Zegarra H, et al. Multiple
20. Koizumi H, Yamagishi T, Yamazaki T, et al. recurrent serosanguineous retinal pigment epithe-
Relationship between clinical characteristics of lial detachments in black women. Am J Ophthalmol.
polypoidal choroidal vasculopathy and choroi- 1985;100:560–9.
dal vascular hyperpermeability. Am J Ophthalmol. 37. Ahuja RM, Stanga PE, Vingerling JR, et al. Polypoidal
2013;155(2):305–13.e301. choroidal vasculopathy in exudative and haemorrhagic
21. Lafaut BA, Leys AM, Snyers B, et al. Polypoidal cho- pigment epithelial detachments. Br J Ophthalmol.
roidal vasculopathy in Caucasians. Graefes Arch Clin 2000;84(5):479–84.
Exp Ophthalmol. 2000;238(9):752–9. 38. Ciardella AP, Donsoff IM, Huang SJ, et al.
22. Maruko I, Iida T, Saito M, et al. Combined cases of Polypoidal choroidal vasculopathy. Surv Ophthalmol.
polypoidal choroidal vasculopathy and typical age- 2004;49(1):25–37.
related macular degeneration. Graefes Arch Clin Exp 39. Imamura Y, Engelbert M, Iida T, et al. Polypoidal
Ophthalmol. 2010;248:361–8. choroidal vasculopathy: a review. Surv Ophthalmol.
23. De Salvo G, Vaz-Pereira S, Keane PA, et al. Sensitivity 2010;55(6):501–15.
and specificity of spectral-domain optical coherence 40. Kwok AK, Lai TY, Chan CW, et al. Polypoidal choroi-
tomography in detecting idiopathic polypoidal choroidal vasculopathy in Chinese patients. Br J Ophthalmol.
dal vasculopathy. Am J Ophthalmol. 2014;158:1228– 2002;86:892–7.
38.e1221. 41. Wen F, Chen C, Wu D, et al. Polypoidal choroidal vas-
24. Maruko I, Iida T, Sugano Y, et al. Subfoveal retinal and culopathy in elderly Chinese patients. Graefes Arch
choroidal thickness after verteporfin photodynamic Clin Exp Ophthalmol. 2004;242(8):625–9.
therapy for polypoidal choroidal vasculopathy. Am J 42. Ladas ID, Rouvas AA, Moschos MM, et al. Polypoidal
Ophthalmol. 2011;151(4):594–603.e591. choroidal vasculopathy and exudative age-related
25. Sayanagi K, Gomi F, Akiba M, et al. En-face high- macular degeneration in Greek population. Eye
penetration optical coherence tomography imaging in (Lond). 2004;18(5):455–9.
polypoidal choroidal vasculopathy. Br J Ophthalmol. 43. Scassellati-Sforzolini B, Mariotti C, Bryan R, et al.
2015;99(1):29–35. Polypoidal choroidal vasculopathy in Italy. Retina.
26. Yannuzzi LA, Freund KB, Goldbaum M, et al. 2001;21(2):121–5.
Polypoidal choroidal vasculopathy masquerading as 44. Cackett P, Htoon H, Wong D, et al. Hemorrhagic pig-
central serous chorioretinopathy. Ophthalmology. ment epithelial detachment as a predictive feature of
2000;107(4):767–77. polypoidal choroidal vasculopathy in a Chinese popu-
27. Cheung CM, Wong TY. Treatment of age-related mac- lation. Eye. 2009;24:789–92.
ular degeneration. Lancet. 2013;382(9900):1230–2. 45. Hikichi T, Ohtsuka H, Higuchi M, et al. Causes of mac-
28. Cheung CM, Bhargava M, Laude A, et al. Asian age- ular serous retinal detachments in Japanese patients 40
related macular degeneration phenotyping study: years and older. Retina. 2009;29(3):395–404.
rationale, design and protocol of a prospective cohort 46. Song SJ, Youm DJ, Chang Y, et al. Age-related macular
study. Clin Exp Ophthalmol. 2012;40:727–35. degeneration in a screened south Korean population:
29. Jonas JB. Global prevalence of age-related macular prevalence, risk factors, and subtypes. Ophthalmic
degeneration. Lancet Glob Health. 2014;2:e65–6. Epidemiol. 2009;16:304–410.
30. Kokame GT. Polypoidal choroidal vasculopathy—an 47. Yamaoka S, Okada AA, Sugahara M, et al. Clinical
important diagnositic tool with therapeutic implica- features of polypoidal choroidal vasculopathy and
tions. Retina. 2012;32(8):1446–8. visual outcomes in the absence of classic choroidal
31. Coscas G, Yamashiro K, Coscas F, et al. Comparison of neovascularization. Ophthalmologica. 2010;224:
exudative age-related macular degeneration subtypes 147–52.
48. Uyama M, Matsubara T, Fukushima I, et al. Idiopathic arterial macroaneurysm and hypertensive retinopathy.
polypoidal choroidal vasculopathy in Japanese Retina. 1996;16(2):105–11.
patients. Arch Ophthalmol. 1999;117:1035–42. 64. Nakashizuka H, Mitsumata M, Okisaka S, et al.
49. Cackett P, Wong D, Yeo I. A classification sys- Clinicopathologic findings in polypoidal cho-
tem for polypoidal choroidal vasculopathy. Retina. roidal vasculopathy. Invest Ophthalmol Vis Sci.
2009;29(2):187–91. 2008;49(11):4729–37.
50. Chakravarthy U, Evans J, Rosenfeld PJ. Age related 65. MacCumber MW, Dastgheib K, Bressler NM, et al.
macular degeneration. BMJ. 2010;340:c981. Clinicopathologic correlation of the multiple recurrent
51. Wong TY, Chakravarthy U, Klein R, et al. The natu- serosanguineous retinal pigment epithelial detach-
ral history and prognosis of neovascular age-related ments syndrome. Retina. 1994;14(2):143–52.
macular degeneration: a systematic review of the lit- 66. Spraul CW, Grossniklaus HE, Lang GK. Idiopathic
erature and meta-analysis. Ophthalmology. 2008;115: polypoid choroid vasculopathy. Klin Monatsbl
116–26. Augenheilkd. 1997;210(6):405–6.
52. Kawasaki R, Wang JJ, Amirul FM, et al. Is bilateral 67. Okubo A, Sameshima M, Uemura A, et al.
age-related macular degeneration less common in Clinicopathological correlation of polypoidal choroi-
Asians than Caucasians? Ophthalmic Epidemiol. dal vasculopathy revealed by ultrastructural study. Br
2011;18(6):253–8. J Ophthalmol. 2002;86(10):1093–8.
53. Age-Related Eye Disease Study Research Group. 68. Kuroiwa S, Tateiwa H, Hisatomi T, et al. Pathological
Risk factors associated with age-related macu- features of surgically excised polypoidal choroidal
lar degeneration. A case-control study in the vasculopathy membranes. Clin Exp Ophthalmol.
age-related eye disease study: Age-Related Eye 2004;32(3):297–302.
Disease Study Report Number 3. Ophthalmology. 69. Matsuoka M, Ogata N, Otsuji T, et al. Expression of
2000;107(12):2224–32. pigment epithelium derived factor and vascular endo-
54. Kim YT, Kang SW, Chung SE, et al. Development of thelial growth factor in choroidal neovascular mem-
polypoidal choroidal vasculopathy in unaffected fel- branes and polypoidal choroidal vasculopathy. Br J
low eyes. Br J Ophthalmol. 2012;96:1217–21. Ophthalmol. 2004;88(6):809–15.
55. Ueta T, Iriyama A, Francis J, et al. Development of 70. Tong JP, Chan WM, Liu DT, et al. Aqueous humor lev-
typical age-related macular degeneration and polypoi- els of vascular endothelial growth factor and pigment
dal choroidal vasculopathy in fellow eyes of Japanese epithelium-derived factor in polypoidal choroidal
patients with exudative age-related macular degenera- vasculopathy and choroidal neovascularization. Am J
tion. Am J Ophthalmol. 2008;146:96–101. Ophthalmol. 2006;141(3):456–62.
56. Pauleikhoff D, Barondes MJ, Minassian D, et al. 71. Grassi MA, Fingert JH, Scheetz TE, et al. Ethnic
Drusen as risk factors in age-related macular disease. variation in AMD-associated complement fac-
Am J Ophthalmol. 1990;109:38–43. tor H polymorphism p.Tyr402His. Hum Mutat.
57. Iwama D, Tsujikawa A, Sasahara M, et al. Polypoidal 2006;27(9):921–5.
choroidal vasculopathy with drusen. Jpn J Ophthalmol. 72. Wang D, Zhou J, Hou X, et al. CETP gene may be
2008;52(2):116–21. associated with advanced age-related macular degen-
58. Ahuja RM, Downes SM, Stanga PE, et al. Polypoidal eration in the Chinese population. Ophthalmic Genet.
choroidal vasculopathy and central serous chorioreti- 2014;36(4):303–8.
nopathy. Ophthalmology. 2001;6:1009–10. 73. Chen H, Liu K, Chen LJ, et al. Genetic associations
59. Fung AT, Yannuzzi LA, Freund KB. Type 1 (sub- in polypoidal choroidal vasculopathy: a systematic
retinal pigment epithelial) neovascularization in review and meta-analysis. Mol Vis. 2012;18:816–29.
central serous chorioretinopathy masquerading as 74. Kuo JZ, Wong TY, Ong FS. Genetic risk, ethnic varia-
neovascular age-related macular degeneration. Retina. tions and pharmacogenetic biomarkers in age-related
2012;32:1829–37. macular degeneration and polypoidal choroidal vascu-
60. Koizumi H, Yamagishi T, Yamazaki T, et al. Subfoveal lopathy. Expert Rev Ophthalmol. 2013;8(2):127–40.
choroidal thickness in typical age-related macu- 75. Miki A, Honda S, Kondo N, et al. The association of
lar degeneration and polypoidal choroidal vas- age-related maculopathy susceptibility 2 (ARMS2)
culopathy. Graefes Arch Clin Exp Ophthalmol. and complement factor H (CFH) variants with two
2011;249(8):1123–8. angiographic subtypes of polypoidal choroidal vascu-
61. Ueta T, Obata R, Inoue Y, et al. Background com- lopathy. Ophthalmic Genet. 2013;34(3):146–50.
parison of typical age-related macular degeneration 76. Tanaka K, Nakayama T, Mori R, et al. Associations
and polypoidal choroidal vasculopathy in Japanese of complement factor H (CFH) and age-related macu-
patients. Ophthalmology. 2009;116:2400–6. lopathy susceptibility 2 (ARMS2) genotypes with
62. Kikuchi M, Nakamura M, Ishikawa K, et al. Elevated subtypes of polypoidal choroidal vasculopathy. Invest
C-reactive protein levels in patients with polypoidal Ophthalmol Vis Sci. 2011;52(10):7441–4.
choroidal vasculopathy and patients with neovascular 77. Kawamura A, Yuzawa M, Mori R, et al. Indocyanine
age-related macular degeneration. Ophthalmology. green angiographic and optical coherence tomographic
2007;114:1722–7. findings support classification of polypoidal choroi-
63. Ross RD, Gitter KA, Cohen G, et al. Idiopathic pol- dal vasculopathy into two types. Acta Ophthalmol.
ypoidal choroidal vasculopathy associated with retinal 2013;91(6):e474–81.
234 A. H. C. Koh
78. Chung SE, Kang SW, Lee JH, et al. Choroidal thick- 94. Hatz K, Prunte C. Polypoidal choroidal vascu-
ness in polypoidal choroidal vasculopathy and exuda- lopathy in Caucasian patients with presumed neo-
tive age-related macular degeneration. Ophthalmology. vascular age-related macular degeneration and
2011;118(5):840–5. poor ranibizumab response. Br J Ophthalmol.
79. Chan WM, Lam DS, Lai TY, et al. Photodynamic ther- 2014;98(2):188–94.
apy with verteporfin for symptomatic polypoidal cho- 95. Matsushita S, Naito T, Takebayashi M, et al. The
roidal vasculopathy: one-year results of a prospective prognosis of cases with massive subretinal hemor-
case series. Ophthalmology. 2004;111(8):1576–84. rhage after photodynamic therapy. J Med Investig.
80. Bessho H, Honda S, Imai H, et al. Natural course and 2008;55(3–4):231–5.
funduscopic findings of polypoidal choroidal vascu- 96. Windisch R, Windisch BK, Cruess AF. Use of
lopathy in a Japanese population over 1 year of follow- fluorescein and indocyanine green angiography
up. Retina. 2011;31(8):501–15. in polypoidal choroidal vasculopathy patients fol-
81. Moorthy RS, Lyon AT, Rabb MF, et al. Idiopathic lowing photodynamic therapy. Can J Ophthalmol.
polypoidal choroidal vasculopathy of the macula. 2008;43(6):678–82.
Ophthalmology. 1998;105:1380–5. 97. Lee MY, Lee WK, Baek J, et al. Photodynamic
82. Okubo A, Arimura N, Abematsu N, et al. Predictive therapy versus combination therapy in polypoidal
signs of benign course of polypoidal choroidal choroidal vasculopathy: changes of aqueous vas-
vasculopathy: based upon the long-term obser- cular endothelial growth factor. Am J Ophthalmol.
vation of non-treated eyes. Acta Ophthalmol. 2013;156(2):343–8.
2010;88(4):e107–14. 98. Yuzawa M, Mori R, Haruyama M. A study of
83. Uyama M, Wada M, Nagai Y, et al. Polypoidal choroi- laser photocoagulation for polypoidal choroi-
dal vasculopathy: natural history. Am J Ophthalmol. dal vasculopathy. Jpn J Ophthalmol. 2003;47(4):
2002;133(5):639–48. 379–84.
84. Okubo A, Sameshima M, Sakamoto T. Plasticity 99. Cheung CM, Yeo I, Li X, et al. Argon laser with and
of polypoidal lesions in polypoidal choroidal without anti-vascular endothelial growth factor ther-
vasculopathy. Graefes Arch Clin Exp Ophthalmol.
apy for extrafoveal polypoidal choroidal vasculopa-
2004;242(11):962–5. thy. Am J Ophthalmol. 2013;155(2):295–304.e291.
85. Lee WK, Kim KS, Kim W, et al. Responses to photo- 100. Gomi F, Ohji M, Sayanagi K, et al. One-year out-
dynamic therapy in patients with polypoidal choroidal comes of photodynamic therapy in age-related
vasculopathy consisting of polyps resembling grape macular degeneration and polypoidal choroidal
clusters. Am J Ophthalmol. 2012;154:355–65. vasculopathy in Japanese patients. Ophthalmology.
86. Byeon SH, Lew YJ, Lee SC, et al. Clinical features 2008;115(1):141–6.
and follow-up results of pulsating polypoidal choroi- 101. Honda S, Imai H, Yamashiro K, et al. Comparative
dal vasculopathy treated with photodynamic therapy. assessment of photodynamic therapy for typical
Acta Ophthalmol. 2010;88:660–8. age-related macular degeneration and polypoi-
87. Okubo A, Ito M, Sameshima M, et al. Pulsatile dal choroidal vasculopathy: a multicenter study
blood flow in the polypoidal choroidal vasculopathy. in Hyogo prefecture, Japan. Ophthalmologica.
Ophthalmology. 2005;112(8):1436–41. 2009;223(5):333–8.
88. Yuzawa M, Mori R, Kawamura A. The origins of 102. Akaza E, Mori R, Yuzawa M. Long-term results of
polypoidal choroidal vasculopathy. Br J Ophthalmol. photodynamic therapy of polypoidal choroidal vas-
2005;89:602–7. culopathy. Retina. 2008;28(5):717–22.
89. Okubo A, Hirakawa M, Ito M, et al. Clinical fea- 103. Akaza E, Yuzawa M, Mori R. Three-year follow-
tures of early and late stage polypoidal choroidal up results of photodynamic therapy for polypoi-

vasculopathy characterized by lesion size and dis- dal choroidal vasculopathy. Jpn J Ophthalmol.
ease duration. Graefes Arch Clin Exp Ophthalmol. 2011;55(1):39–44.
2008;246:491–9. 104. Oishi A, Kojima H, Mandai M, et al. Comparison
90. Tsujikawa A, Ojima Y, Yamashiro K, et al. of the effect of ranibizumab and verteporfin for
Association of lesion size and visual progno- polypoidal choroidal vasculopathy: 12-month
sis to polypoidal choroidal vasculopathy. Am J LAPTOP study results. Am J Ophthalmol.
Ophthalmol. 2011;151(6):961–72. 2013;156(4):644–51.
91. Tamura H, Tsujikawa A, Otani A, et al. Polypoidal 105. Hirami Y, Tsujikawa A, Otani A, et al. Hemorrhagic
choroidal vasculopathy appearing as classic choroi- complications after photodynamic therapy for
dal neovascularization on fluorescein angiography. polypoidal choroidal vasculopathy. Retina.
Br J Ophthalmol. 2007;91:1152–9. 2007;27(3):335–41.
92. Yannuzzi LA, Sorenson J, Spaide RF, et al. Idiopathic 106. Ojima Y, Tsujikawa A, Otani A, et al. Recurrent
polypoidal choroidal vasculopathy (IPCV). 1990. bleeding after photodynamic therapy in polyp-
Retina. 2012;32(Suppl):1–8. oidal choroidal vasculopathy. Am J Ophthalmol.
93. Cho M, Barbazetto IA, Freund KB. Refractory 2006;141(5):958–60.
neovascular age-related macular degeneration sec- 107. Tsuchihashi T, Mori K, Ueyama K, et al. Five-year
ondary to polypoidal choroidal vasculopathy. Am J results of photodynamic therapy with vertepor-
Ophthalmol. 2009;148(1):70–8.e71. fin for Japanese patients with neovascular age-
related macular degeneration. Clin Ophthalmol. works in polypoidal choroidal vasculopathy. Br J

2013;7:615–20. Ophthalmol. 2012;96(3):394–9.
108. Lee WK, Lee PY, Lee SK. Photodynamic therapy 123. Ogura Y, VIEW 2 Study Group. Effect of aflibercept
for polypoidal choroidal vasculopathy: vaso- on polypoidal choroidal vasculopathy. The results
occlusive effect on the branching vascular net- of re-analysis of Japanese patients. Oral presenta-
work and origin of recurrence. Jpn J Ophthalmol. tion at The 53rd Annual Meeting of Japanese Retina
2008;52(2):108–15. and Vitreous Society; 2014 Nov 28–30; Osaka,
109. Saito M, Iida T, Nagayama D. Photodynamic Japan.
therapy with verteporfin for age-related macular 124. Iida T. Efficacy and safety of intravitreal aflibercept
degeneration or polypoidal choroidal vasculopa- in polypoidal choroidal vasculopathy: the PLANET
thy: comparison of the presence of serous retinal study. Presentation at the 10th Congress of the Asia-
pigment epithelial detachment. Br J Ophthalmol. Pacific Vitreo-retina Society (APVRS); 2016 Dec
2008;92(12):1642–7. 8–10; Bangkok, Thailand.
110. Wakabayashi T, Gomi F, Sawa M, et al. Marked 125. Inoue M, Arakawa A, Yamane S, et al. Short-term
vascular changes of polypoidal choroidal vasculopa- efficacy of intravitreal aflibercept in treatment-naive
thy after photodynamic therapy. Br J Ophthalmol. patients with polypoidal choroidal vasculopathy.
2008;92(7):936–40. Retina. 2014;34(11):2178–84.
111. Dewi NA, Yuzawa M, Tochigi K, et al. Effects of 126. Ijiri S, Sugiyama K. Short-term efficacy of intravit-
photodynamic therapy on the choriocapillaris and real aflibercept for patients with treatment-naïve pol-
retinal pigment epithelium in the irradiated area. Jpn ypoidal choroidal vasculopathy. Graefes Arch Clin
J Ophthalmol. 2008;52(4):277–81. Exp Ophthalmol. 2014;253(3):351–7.
112. Wong CW, Cheung CM, Mathur R, et al. Three year 127. Yonekawa Y. Aflibercept for the treatment of refrac-
results of polypoidal choroidal vasculopathy treated tory polypoidal choroidal vasculopathy. Can J
with photodynamic therapy:retrospective study and Ophthalmol. 2013;48(3):e59–60.
systematic review. Retina. 2015;35(8):1577–93. 128. Miura M, Iwasaki T, Goto H. Intravitreal aflibercept
113. Heier JS, Brown DM, Chong V, et al. Intravitreal for polypoidal choroidal vasculopathy after develop-
aflibercept (VEGF trap-eye) in wet ageing ranibizumab tachyphylaxis. Clin Ophthalmol.
related macular degeneration. Ophthalmology. 2013;7:1591–5.
2012;119(12):2537–48. 129. Saito M, Kano M, Itagaki K, et al. Switching to
114. Rosenfeld PJ, Brown DM, Heier JS, et al. intravitreal aflibercept injection for polypoidal
Ranibizumab for neovascular age-related macular choroidal vasculopathy refractory to ranibizumab.
degeneration. N Engl J Med. 2006;355(14):1419–31. Retina. 2014;34(11):2192–201.
115. Brown DM, Kaiser PK, Michels M, et al. 130. Kawashima Y, Oishi A, Tsujikawa A, et al. Effects
Ranibizumab versus verteporfin for neovascular of aflibercept for ranibizumab-resistant neovascu-
age-related macular degeneration. N Engl J Med. lar age-related macular degeneration and polypoi-
2006;355(14):1432–44. dal choroidal vasculopathy. Graefes Arch Clin Exp
116. Schmidt-Erfurth U, Kaiser PK, Korobelnik JF, Ophthalmol. 2015;253(9):1471–7.
et al. Intravitreal aflibercept injection for neovas- 131. Gomi F, Sawa M, Wakabayashi T, et al. Efficacy of
cular age-related macular degeneration: ninety-six- intravitreal bevacizumab combined with photody-
week results of the VIEW studies. Ophthalmology. namic therapy for polypoidal choroidal vasculopa-
2014;121(1):193–201. thy. Am J Ophthalmol. 2010;150(1):48–54.e41.
117. Hikichi T, Higuchi M, Matsushita T, et al. Results 132. Saito M, Iida T, Kano M, et al. Two-year results
of 2 years of treatment with as-needed ranibizumab of combined intravitreal ranibizumab and pho-
reinjection for polypoidal choroidal vasculopathy. todynamic therapy for polypoidal choroidal vas-
Br J Ophthalmol. 2013;97(5):617–21. culopathy. Graefes Arch Clin Exp Ophthalmol.
118. Kokame GT, Yeung L, Teramoto K, et al. Polypoidal 2013;251(9):2099–110.
choroidal vasculopathy exudation and hemorrhage: 133. Kang HM, Koh HJ, Lee CS, et al. Combined pho-
results of monthly ranibizumab therapy at one year. todynamic therapy with intravitreal bevacizumab
Ophthalmologica. 2014;231(2):94–102. injections for polypoidal choroidal vasculopa-
119. Mori R, Yuzawa M, Akaza E, et al. Treatment results thy: long-term visual outcome. Am J Ophthalmol.
at 1 year of ranibizumab therapy for polypoidal cho- 2014;157(3):598–606.e591.
roidal vasculopathy in eyes with good visual acuity. 134. Koh A. Ranibizumab and vPDT combination ther-
Jpn J Ophthalmol. 2013;57(4):365–71. apy versus ranibizumab monotherapy for macu-
120. Takahashi K, Ogura Y, Ishibashi T, et al. Treatment lar PCV: 12-month results from the EVEREST II
guidelines for age-related macular degeneration. study. Presentation at the American Academy of
Nippon Ganka Gakkai Zasshi. 2012;116(12):1150–5. Ophthalmology (AAO) 2016 Annual Meeting; 2016
121. Gomi F, Sawa M, Sakaguchi H, et al. Efficacy of Oct 14; Chicago, IL, USA; 2016.
intravitreal bevacizumab for polypoidal choroidal 135. Hosokawa M, Shiraga F, Yamashita A, et al. Six-
vasculopathy. Br J Ophthalmol. 2008;92(1):70–3. month results of intravitreal aflibercept injections for
122. Wakabayashi T, Gomi F, Sawa M, et al. Intravitreal patients with polypoidal choroidal vasculopathy. Br
bevacizumab for exudative branching vascular net- J Ophthalmol. 2015;99(8):1087–91.
236 A. H. C. Koh
136. Oishi A, Tsujikawa A, Yamashiro K, et al. One-year 142. Yamashita A, Shiraga F, Shiragami C, et al. Two-
result of aflibercept treatment on age-related macu- year results of reduced-fluence photodynamic ther-
lar degeneration and predictive factors for visual out- apy for polypoidal choroidal vasculopathy. Am J
come. Am J Ophthalmol 2015;159(5):853–60.e851. Ophthalmol. 2013;155(1):96–102.e101.
137. Kim M, Kim K, Kim DG, et al. Two-year results of 143. Yoshida Y, Kohno T, Yamamoto M, et al. Two-year
photodynamic therapy combined with intravitreal results of reduced-fluence photodynamic therapy
anti-vascular endothelial growth factor for polyp- combined with intravitreal ranibizumab for typi-
oidal choroidal vasculopathy. Ophthalmologica. cal age-related macular degeneration and polyp-
2011;226(4):205–13. oidal choroidal vasculopathy. Jpn J Ophthalmol.
138. Nemoto R, Miura M, Iwasaki T, et al. Two-year 2013;57(3):283–93.
follow-up of ranibizumab combined with photody- 144. Eandi CM, Ober MD, Freund KB, et al. Selective
namic therapy for polypoidal choroidal vasculopa- photodynamic therapy for neovascular age-related
thy. Clin Ophthalmol. 2012;6:1633–8. macular degeneration with polypoidal choroidal
139. Kang HM, Koh HJ. Two-year outcome after combi- neovascularization. Retina. 2007;27(7):825–31.
nation therapy for polypoidal choroidal vasculopa- 145. Hikichi T, Ohtsuka H, Higuchi M, et al. Factors pre-
thy: comparison with photodynamic monotherapy dictive of visual acuity outcomes 1 year after photo-
and anti-vascular endothelial growth factor mono- dynamic therapy in Japanese patients with polypoidal
therapy. Ophthalmologica. 2014;231(2):86–93. choroidal vasculopathy. Retina. 2011;31(5):857–65.
140. Sagong M, Lim S, Chang W. Reduced-fluence pho- 146. Jeon S, Lee WK, Kim KS. Adjusted retreat-
todynamic therapy combined with intravitreal beva- ment of polypoidal choroidal vasculopathy after
cizumab for polypoidal choroidal vasculopathy. Am combination therapy: results at 3 years. Retina.
J Ophthalmol. 2012;153(5):873–82.e872. 2013;33(6):1193–200.
141. Sakurai M, Baba T, Kitahashi M, et al. One-year 147. Cheung CM, Wong TY. Ranibizumab and bevaci-
results of intravitreal ranibizumab combined with zumab for AMD. N Engl J Med. 2011;365(23):2237,
reduced-fluence photodynamic therapy for polyp- author reply 2237.
oidal choroidal vasculopathy. Clin Ophthalmol.
2014;8:235–41.
Council of the Asia Pacific Vitreo Retina Society (APVRS)

and was Regional Secretary of the Asia Pacific Academy
of Ophthalmology (APAO). He is an active member of
EURETINA, American Academy of Ophthalmology, Asia
Pacific Academy of Ophthalmology, Vision Academy, and
Asia Pacific Vitreoretina Society.
Prof Koh completed Fellowships in Medical Retina at
Moorfields Eye Hospital and the Jules Stein Eye Institute.
He trained under Professors Alan Bird, Susan Downer,
Graham Holder, Bradley Straatsma, and John
Heckenlively.
Professor Koh is the Founding Partner and Senior His areas of expertise include degenerative macular
Consultant at the Eye & Retina Surgeons, Camden disease such as polypoidal choroidal vasculopathy (PCV),
Medical Centre and Associate Professor at the National choroidal neovascularization, retinal dystrophy, visual
University of Singapore. He is the visiting consultant to electrophysiology, and retinal vascular disease. He is one
Singapore National Eye Centre and Tan Tock Seng of the key international opinion leaders in Retina. He was
Hospital. He is the past President of the Singapore Society the principal investigator of the EVEREST-1 and
of Ophthalmology, and a member of the Executive EVEREST-2 trials and Chairman of the PCV Roundtable
Council of the College of Ophthalmologists. He is the of Experts. He sits on several international advisory
Executive Director of ESASO (European School of boards and steering committees, including Vision
Advanced Studies in Ophthalmology) Asia. He was the Academy Steering Committee and Brolucizumab Global
Secretary-General and Vice President of the 2006 APAO Steering Committee. Prof Koh has published extensively
Congress in Singapore and Congress President of the 5th in peer-reviewed journals and ophthalmology textbooks,
APVRS in 2010. He is an active member of the Macula and lectures in numerous national and international meet-
Society and Club Jules Gonin. He serves on the Executive ings each year.
Associated with Rare Entities 19
Pasha Anvari, Masood Naseripour,
and Khalil Ghasemi Falavarjani
19.1 Choroidal Nevus thelial detachments, drusen, and photoreceptor

degenerations [4, 6].
Choroidal nevi are usually found incidentally Rarely, choroidal nevi may lead to choroidal
during routine ophthalmoscopic examination. neovascularization (CNV) with subsequent
The prevalence ranges from 4 to 8% in general visual loss, particularly if the fovea is involved.
white population [1, 2] making it the most com- Gass in 1967 [7] was the first to describe the
mon benign melanocytic intraocular tumor in association between choroidal nevus and CNV. In
adults [3]. Clinically, choroidal nevi appear as early reports, there was uncertainty regarding this
flat or slightly elevated, round or oval pigmented association [8]. Nonetheless, growing evidence
gray-brown (and sometimes amelanotic) lesions suggested that chronic mechanical, inflamma-
with well-defined margins. They are located deep tory, or degenerative effects of nevus cells on
to the retina, evenly distributed in all four quad- overlying RPE–Bruch’s membrane is presum-
rants of fundus and more common in post- ably responsible for CNV development. The
equatorial region [4]. nevus-associated CNV prevalence is estimated to
Although mostly asymptomatic, choroidal be less than 1% in all choroidal nevi [9]. CNV-
nevi should be observed for possible malignant complicated choroidal nevi are located in the
transformation [5]. They may result in posterior pole and frequently in the macular area
decreased visual acuity by inducing secondary [1, 10, 11]. The reason for this macular predilec-
changes in overlying retinal pigment epithe- tion remains unclear.
lium (RPE) and retina including RPE atrophy The patients are typically in their sixth decade
or secondary fibrous metaplasia, pigment epi- of life, presenting with blurry vision or metamor-
phopsia. On examination, subretinal fluid (SRF),
lipid exudates, or subretinal hemorrhage, sugges-
tive of presumed CNV is evident at the epicenter
P. Anvari · M. Naseripour
Eye Research Center and Eye Department, Rassoul or edge of preexisting choroidal nevus. Signs of
Akram Hospital, Iran University of Medical Sciences, nevus chronicity including drusen or RPE
Tehran, Iran changes are practically always present. Relatively
K. G. Falavarjani (*) common age-related macular changes in the
Eye Research Center and Eye Department, Rassoul same or contralateral eye may lead to an incorrect
Akram Hospital, Iran University of Medical Sciences, diagnosis of neovascular age-related macular
Tehran, Iran
degeneration (AMD) on cursory examination
Eye Research Center, Rassoul Akram Hospital, [12]. In addition, in the case of prominent exuda-
Sattarkhan-Niaiesh St, Tehran, Iran

238 P. Anvari et al.
tion from CNV, the underlying choroidal nevus inability to show activity and thus the necessity
may be overlooked, making the correct diagnosis for treatment.
challenging [13]. Generally, the presence of CNV associated
Owing to macular predilection of most CNVs with a pigmented choroidal lesion points out to
complicating nevi, prompt diagnosis of CNV and the benign nature of the lesion [19]. Reports
differentiating from simple nevus, and appropri- describing melanoma-associated CNV are
ate treatment is imperative to preserve vision. exceedingly rare [20]. However, the wise
More to the point is, the onset of symptoms; SRF approach would be to follow these nevi for signs
or increased height of preexisting choroidal of growth by serial follow-up examinations.
nevus may give the wrong impression of malig- It should be emphasized that choroidal nevus
nant transformation to the clinician. In case series may be associated with overlying subretinal fluid
of nevus associated CNV by Zografos et al. [10] without the evidence of CNV in multimodal
and Pellegrini et al. [14] suspected choroidal imaging evaluation. In the largest analysis of
melanoma and choroidal metastasis were the 3422 eyes with choroidal nevus, in 345 (10%)
most common initial diagnosis of referred eyes, nevi showed evidence of subretinal fluid.
patients. Therefore, an accurate diagnosis may However, these findings indicated the CNV pres-
obviate the need for unnecessary work-up or ence only in 20 (0.6%) eyes [9].
inappropriate treatments. Because of the lack of randomized clinical tri-
The diagnosis of choroidal nevus-associated als, treatment decisions for nevus-associated
CNV relies on clinical examination and appropri- CNV are mainly guided by small, retrospective
ate multimodal imaging. Type 1 CNV, Type 2 case series or reports, and should be tailored to
CNV, and polypoidal choroidal neovasculariza- the patient based on clinical findings. Although
tion (PCV) [15] have all been described in the there are reports of successful management of
context of choroidal nevus with varying inci- extrafoveal CNV by conventional argon laser
dence. Typical ophthalmoscopic signs of CNV photocoagulation, scar expansion into fovea is a
including subretinal hemorrhage, SRF, and exu- major concern [8].
dation may be seen. Fluorescein angiography For subfoveal or juxtafoveal CNV, trans-
(FA) and indocyanine green angiography (ICGA) pupillary thermotherapy [21] and photodynamic
are the standard imaging modalities to demon- therapy (PDT) [22] have been used with variable
strate dye leakage from new vessels. Spectral results. Some patients may require multiple PDT
domain-optical coherence tomography sessions with subsequent collateral damage to
(SD-OCT) is useful by revealing SRF, intra- RPE and choriocapillaris [22, 23]. In recent
retinal fluid, and PED. In the case of PCV, years, with the advent of anti-vascular endothe-
SD-OCT may show typical features including lial growth factors (VEGF), some authors have
sharp PED peak, multiple PEDs, and hypo- used intravitreal anti-VEGF injections success-
reflective lumen within hyper-reflective materials fully for the treatment of subfoveal nevus-
[16]. In recent years, optical coherence tomogra- associated CNV [11, 24, 25]. In a recent report,
phy angiography (OCTA) has gained popularity 73% of eyes with CNV associated with choroidal
among ophthalmologists to detect CNV in vari- nevus responded properly to intravitreal bevaci-
ous ocular pathologies. In a study by Pellegrini zumab injection and a combination of PDT, and
et al., OCTA was able to visualize CNV in all 11 bevacizumab led to further regression in unre-
cases of nevus associated CNV with “sea fan” or sponsive eyes [26].
“long filamentous linear vessel” patterns [14]. In our institution, we manage nevus associ-
Major advantages of OCTA to conventional dye- ated CNV lesions according to the CNV location
based angiography include its noninvasive nature, and patient’s symptoms. For the small
rapid acquisition time, and depth-resolved images asymptomatic, extra-macular CNV, we follow

[17]. Nevertheless, the main drawbacks of this the patients closely. If edema or hard exudates are
technique are high rates of artifacts [18] and increasing and threatening the fovea, we consider
19 Choroidal Neovascularization Associated with Rare Entities 239
intravitreal injection of anti-VEGF agents, with receptor loss and subsequently can lead to
or without argon laser photocoagulation treat- profound vision loss if occurs subfoveally [34].
ment based on FA findings. For subfoveal or jux- CNV can complicate CO and is considered a
tafoveal CNV, we treat the patients by intravitreal major cause of visual impairment in subjects with
anti-VEGF injection. We may consider adjunc- CO. In a study by Shields et al. [35] 31% of
tive PDT to reduce the frequency of intravitreal patients with CO developed CNV by 10 years.
injections in selected patients with anti-VEGF Similarly, Aylward et al. [33] reported that 10-year
resistance or in case of recurrent CNV. An probability for CNV development was 47%.
absence of response to anti-VEGF injection There is no identified current treatment to
should raise a suspicion of malignant lesions induce regression of CO and therapy mainly tar-
[27]. gets complications. Because the majority of the
In conclusion, CNV may rarely complicate patients with CO and CNV are relatively young,
choroidal nevi in the posterior pole of the fundus. the burden of visual impairment on quality of life
It is recommended to perform appropriate imag- is deemed to be high. This highlights the impor-
ing in subjects with choroidal nevi and recent tance of prompt diagnosis and treatment of CNV.
onset of SRF, exudate, or hemorrhage to consider The pathogenesis of CNV arising from CO is
this entity. It is not a harbinger of malignant yet to be determined. It has been speculated that
transformation and could be treated successfully thin and degenerated RPE–Bruch’s membrane
by intravitreal injection of anti-VEGF agents. overlying tumor allows growth of abnormal cho-
roidal vessel [35]. The presence of osteoclasts in
the surgical specimen obtained from a subject
19.2 Choroidal Osteoma with CO and CNV [36] has raised the possibility
that CNV is an extension of tumor vessels.
Choroidal osteoma (CO) is a rare, benign, intra- Moreover, inadequate transmural blood supply,
ocular tumor that typically occurs in otherwise high metabolic demand of tumor, and chronic
healthy young females in their second or third inflammation may result in retinal ischemia and
decades of life [28]. Clinically, it appears as a overexpression of VEGF that may contribute to
slightly elevated (<2.5 mm) yellow-orange sub- the pathogenesis of CO-associated CNV. Recent
retinal mass. The tumor is unilateral in 80% of reports of choroidal excavation and CNV associ-
cases and is usually located in the peripapillary or ated with CO in decalcified cases speculated a
macular area with well-defined, geographic, or common pathogenic pathway for the develop-
scalloped borders [29]. The exact etiology of ment of FCE and CNV in choroidal osteoma [37].
tumor remained still unclear; however, trauma, Patients commonly present with sudden onset
congenital, inflammatory, or hormonal theories of decreased visual acuity or metamorphopsia.
have been postulated [30–32]. Although patients On examination, the tumor along with subretinal
may initially be asymptomatic or have good hemorrhage (SRH) or subretinal fluid (SRF) is
visual acuity, the long-term visual prognosis is evident. CNV may occur over or at the edge of
poor as nearly half of the patients will experience the tumor and is frequently associated with tumor
severe vision loss (visual acuity less than 20/200) decalcification and secondary RPE abnormali-
after 10 years of follow-up [33]. ties. Subretinal pigmentation is usually found
Over the years, the tumor may grow slowly or around CNV [38]. Cases with CO and CNV
undergo decalcification. The decalcified portion is occurring at the bottom or slope of choroidal
characterized by a lighter color (yellow-white) excavation have also been reported [37, 39].
due to retinal pigment epithelium (RPE) depig- The diagnosis of CO-associated CNV is quite
mentation and increased visibility of underlying challenging. SRH and tumor surface irregularity
choroidal vessels. The tumor decalcification that are the two strongest predictors of CNV develop-
can occur spontaneously, or following laser treatment [35]. However, in subjects with CO, SRH
ment is associated with overlying RPE and photo- and SRF can occur in the absence of detectable
CNV. Aylward et al. [33] reported that only ing from CO has several challenges. First,
22.7% of eyes with CO and serous retinal detach- because of morphologic alterations secondary to
ment had concurrent CNV. Similarly, SRH can tumor, including elevation of the choroidal sur-
occur spontaneously or secondary to Valsalva face, choriocapillaris compression, RPE, and
maneuvers rather than from CNV. outer retinal disorganization, automated segmen-
FA has been the gold standard for the diagno- tation will likely result in artifacts; therefore,
sis of CNV. Both classic and occult subtypes manual correction of segmentation lines is essen-
have been reported, with the predominance of tial to correctly diagnose CNV in most cases.
classic components [40]. However, inherent Second, the distinction of the neovascular net-
patchy hyperfluorescence of the tumor and RPE work from intrinsic tumor vascularity is crucial.
abnormalities may either masquerade or obscure The CNV may appear as a sea fan or tangled net-
leakage from CNV. work surrounded by a dark halo, in contrast to the
OCT is useful in demonstrating SRF, intrareti- fine and thin vascular network of the tumor [44].
nal fluid (IRF), and subretinal hyper-reflective Third, the presence of CNV in OCTA images
material. The characteristics of tumor including does not necessarily indicate activity and the
sponge-like or lattice-like pattern, multiple intra- need for treatment. Despite these limitations,
lesional horizontal lamellar lines, and intrinsic OCTA may be very helpful in differentiating neo-
transparency in addition to a wide range of tumor vascular components from tumor vessels [44].
reflectivity from hypo-to hyperreflectivity are Various treatment modalities including laser
evident, particularly on EDI-OCT. [41] photocoagulation, photodynamic therapy (PDT),
OCT-angiography is helpful by revealing the transpupillary thermotherapy (TTT) [45], surgi-
neovascular network in the outer retina or chorio- cal resection [36], and intravitreal injection of
capillaris slabs (Fig. 19.1) [42–44]. However, the anti-VEGF agents have been attempted to man-
use of OCT-angiography to diagnose CNV aris- age CNV in patients with CO.
a b
Superficial (ILM - IPL) Deep (IPL - OPL) Outer Retina (OPL - BRM) Choriocapillaris (BRM - BRM+30µm)
Fig. 19.1 A 33-year-old female with a best-corrected reflective horizontal lines compatible with choroidal
visual acuity of 4/10 OD. (a) Color fundus photography osteoma. No subretinal or intraretinal fluid was evident.
showed a large peripapillary choroidal osteoma with mac- (c) En-face OCT-angiography demonstrated neovascular
ular extension. An area of decalcification and secondary tuft in choriocapillaris slab. Given the non-exudative
pigmentary alterations was noted. (b) OCT revealed sub- nature of CNV, no treatment was initiated
retinal mass with spongy-like appearance and hyper-
The results of laser photocoagulation of CNV 19.3 Idiopathic Intracranial

in CO are disappointing [46]. Insufficient mela- Hypertension
nin pigmentation of tumor will limit laser absorp-
tion. Multiple sessions may be required and the Idiopathic intracranial hypertension (IIH), for-
risk of complications including hemorrhage, dis- merly known as pseudotumor cerebri or benign
ciform scar, and decalcification is high. Similarly, intracranial hypertension, is a syndrome character-
surgical resection of CNV in an eye with CO ized by elevated intracranial pressure (ICP) in the
resulted in no significant functional improve- absence of clinical, laboratory, or radiological evi-
ment [36]. dence of intracranial pathology [59]. The typical
PDT has been used to successfully treat patient is a young obese woman presenting with
extrafoveal CNV. However, because of the risk headache and transitory blurred vision, commonly
of triggering decalcification or subretinal hem- lasting less than a minute. Some patients may
orrhage, it should be used with caution in sub- experience decreased visual acuity at presentation
foveal CNV [47]. Half-fluence PDT in or during the course of the disease. The causes of
combination with intravitreal anti-VEGF injec- reduced visual acuity in subjects with IIH can be
tion has also been used (Fig. 19.2) [48]. The categorized as optic neuropathy, macular changes,
rationale of combination therapy is to blunt or a combination of both. Optic neuropathy can be
post-PDT VEGF spike and reduce the number progressive and leads to permanent visual loss
of injections. However, in the absence of a ran- while macular changes including hard exudate,
domized clinical trial, this theoretical benefit is subretinal fluid, and chorioretinal folds are largely
difficult to ascertain. reversible by reducing ICP [60]. Rarely, peripapil-
Several studies have reported the efficacy of lary CNV can cause blurred vision.
intravitreal injection of bevacizumab and ranibi- CNVs complicating IIH are rare. Less than 30
zumab in the treatment of CO-associated CNV cases have been reported in the literature. The
[49, 50]. In one study, switching to aflibercept was prevalence of CNV in IIH has been estimated to be
successful in the management of CNV, nonre- 0.53% [61]. Although the exact pathogenesis
sponder to bevacizumab and ranibizumab [51]. remained unclear, Morse et al. [62] speculated that
Some authors suggested that the high efficiency of the pressure deformity of Bruch’s membrane’s
anti-VEGF agents is due to their enhanced pene- border caused by severe disc edema leads to dis-
trance through thin, degenerated RPE–Bruch’s continuity of chorioretinal layers. This anatomical
membrane complex [52]. The patients may require dehiscence coupled with local hypoxia induced by
multiple injections. There might be an increased axonal swelling may result in CNV development.
risk of RPE tear following intravitreal anti-VEGF In some cases, IIH-associated CNV was the
injections in eyes with CO, as their RPE–Bruch’s presenting feature of IIH [63]. Although bilateral
membrane complex might be brittle [53]. cases have been reported, classically, the patient
As mentioned earlier, some patients with CO presents with sudden onset of blurry vision in one
may develop serous retinal detachment in the eye. Previous history of headaches might exist.
absence of CNV. The cause of this condition is On examination, features of papilledema includ-
thought to be chronic RPE decompensation [54]. ing bilateral blurred disc margin and hyperemia
In most cases, subretinal fluid will resolve spon- are evident. On the affected side, subretinal hem-
taneously [55]. The role of anti-VEGF agents in orrhage, hard exudates, and edema suggestive of
this condition is unclear. However, successful CNV are present. The peripapillary CNV mem-
treatment by the use of intravitreal anti-VEGF brane itself might be obscured by severe disc
agents [56, 57] or TTT [58] has been reported. edema and becomes more clinically visible after
In summary, CNV is a common, late compli- the decrease of edema following ICP reduction.
cation of CO. Multimodal imaging including FA, The clinical diagnosis of IIH-associated CNV
OCT, and OCTA aids the diagnosis and intravit- is usually straightforward and is confirmed by
real injection of anti-VEGF leads to anatomic appropriate retinal imaging. FA reveals hyper-
improvement in the majority of cases. fluorescence secondary to CNV leakage, dispro-
a b
c d
Fig. 19.2 A 25-year-old woman presented with sudden secondary to tumor and focal hyperfluorescence due to
onset of blurry vision in her right eye (OD). Best-corrected leakage of choroidal neovascularization. (c) B-scan dem-
visual acuity (BCVA) was 3/10 OD. (a) Color fundus pho- onstrated hyper-echoic lesion with posterior shadowing,
tography showed a large yellow-orange peripapillary compatible with choroidal osteoma (CO). (d) OCT shows
lesion, deep to retina with extension to macula. Subretinal subretinal and intraretinal fluid. (e) Three months post
hemorrhage is visible. (b) Fluorescein angiography PDT, and IVB OCT depicted complete regression of fluid
revealed hypofluorescent area due to blockage of and BCVA improved to 8/10
subretinal hemorrhage; diffuse patchy hyperfluorescence
portionate to the disc edema. In addition, disc Differential diagnosis of peripapillary CNV
leakage due to papilledema is evident in the con- includes neovascular age-related macular
tralateral eye. OCT is useful in demonstrating degeneration, angioid streaks, excavated disc
SRF and intraretinal cystoid edema. Moreover, anomalies, choroidal osteoma, and inflamma-
OCT is an invaluable tool to evaluate the extent tory retinochoroidopathies [64]. These condi-
of foveal involvement and to monitor treatment tions can be differentiated by careful
response in serial examination. It is noteworthy examination of the fellow eye and features spe-
to mention that, although hard exudates, SRF, cific to each entity.
and retinal hemorrhages can be found in cases of Neurology consults for neuroimaging, mea-
severe papilledema, subretinal hemorrhage is suring CSF opening pressure and composition
highly suggestive for CNV. are warranted in newly diagnosed subjects.
The course of IIH-associated CNV is highly tations. Laser pointers are typically red (670 nm
variable. Although some authors reported sponta- wavelength) or green (532 nm wavelength) diode
neous resolution following ICP reduction [65], lasers with maximum power output less than 5
others described stable or progressive course [66]. mw (class 2 or 3a, according to American
Various treatment modalities including surgi- National Standard for Safe Use of Lasers classifi-
cal resection [67], argon laser photocoagulation cation [74]), and thus generally considered safe
[61], photodynamic therapy [61], and intravitreal during an accidental momentarily eye exposure
injection of anti-VEGFs have been used. Surgical [75]. However, the incidence of eye injuries sec-
resection of the CNV membrane has been disap- ondary to laser pointers appears to be increasing
pointing and resulted in choriocapillaris atrophy [76]. The reasons behind this rise are considered
[67]. Argon laser photocoagulation poses the risk to be public availability of high-powered lasers
of damage to papillomacular bundle [68]. (class 3b or 4) with a very similar appearance to
Recently, several case reports of successful man- low-powered lasers [77] and mislabeling of the
agement of IIH-associated CNV by intravitreal actual maximum power of lasers. These devices
anti-VEGFs have been published [69–72]. Some are perceived as fascinating toys by children that
patients may require multiple injections to are especially at risk for accidental laser pointer
achieve complete resolution of macular edema; injuries. The normal defensive reaction of adults
however, the long-term outcomes are promising. to laser beam (i.e., blinking and aversion) that
In conclusion, peripapillary CNV may rarely limits the exposure to less than 0.25 seconds is
occur in subjects with IIH. It should be considered probably less prominent in children as they may
in any patient with bilateral disc swelling and in not fully appreciate the serious hazards of laser
the presence of subretinal hemorrhage. Measures eye injuries.
to reduce ICP (medically or surgically) and intra- Various retinal pathologies including macular
vitreal injection of anti-VEGF are recommended. hole, subhyaloid hemorrhage, sub-internal limit-
ing membrane (sub-ILM) hemorrhage and outer
retinal disruption following laser pointer injury
19.4 Laser-Induced CNV have been described [78, 79]. However, reports of
laser pointer induced CNV are sporadic [80].
Optical radiations in the visible and near-infrared Fujinami et al. [81] reported on an intellectually
spectrum of light (380–1400 nm) can pass disabled child who suffered from CNV following
through ocular media and focus on the retina presumably frequent and repeated exposure to a
[73]. Lasers are monochromatic, coherent, unidi- laser pointer as a toy. Xu et al. [82] described
rectional sources of light with wide applications CNV development in a 12-year-old boy after star-
in medicine, industry, science, and military. ing into a laser pointer beam for about 15 seconds.
The photochemical, thermal, and mechanical His condition was successfully managed by a
effects of laser may lead to retinal damages. The single injection of intravitreal bevacizumab.
energy level, pulse duration, and the target loca- The beneficial role of corticosteroids in the
tion of the beam incidence are among the laser- acute stage of laser-induced maculopathy is not
related factors that determine the severity of well-established. As there is specific treatment
retinal injury [73]. for CNV, the correct diagnosis of this rare com-
plication is important. In this regard, utilizing
FA and OCT-angiography can be advantageous
19.4.1 Choroidal Neovascularization by demonstrating leakage and neovascular net-
Secondary to Accidental Laser work, respectively. Clear visualization of neo-
Exposure vascular membrane by OCT-angiography may
obviate the need for performing invasive, time-
19.4.1.1 Handheld Laser Pointers consuming FA that is particularly a challenge
Laser pointers are ubiquitous devices used in among pediatric age populations. Figure 19.3
classrooms and lecture halls to facilitate presen- shows the utility of the OCTA imaging for
a b
Fig. 19.3 An 8-year-old girl presented with blurry vision retinal fluid was detected. (b) Spectral-domain OCT
in her left eye. Three weeks earlier, she had starred into a revealed increased retinal thickness, subretinal hyper-
green laser pointer for few seconds at a distance of about reflective material (SHRM), and subretinal fluid (SRF) (c)
1 meter. The laser pointer notification on the device Macular OCTA (XR Avanti, Optovue Inc., Fremont CA,
showed a Class III laser production with a maximum out- USA) demonstrated a neovascular tuft in the outer retinal
put power of less than 1000 mW and a wavelength of slab. (d) She was treated with a single intravitreal injec-
523 nm (Changchun Realpoo Photoelectric CO., Ltd., tion of bevacizumab. The size of the neovascular mem-
China). Her best-corrected visual acuity (BCVA) was brane reduced significantly and SRF and SHRM resolved
20/20 OD and counting fingers OS. (a) On funduscopic in OCT at 1-month follow-up. Her vision improved to
examination of the left eye, an elevated yellowish lesion in 20/20 and no recurrence was detected until last visit,
fovea with subretinal hemorrhage and surrounding sub- 5 months after injection
detecting CNV in a patient with accidental laser occurrence [95]. PDT may induce choriocapil-
injury. The prognosis after intravitreal anti- laris occlusion, choroidal ischemia, and thus
VEGF therapy is generally favorable if VEGF overexpression. Noteworthy, all reported
patients are treated properly. cases of CNV secondary to PDT occurred
after full-fluence PDT rather than half-fluence
19.4.1.2 Other Laser Types PDT. Reduced-setting PDT seems to be a safer
Neodymium:yttrium–aluminum–garnet alternative to conventional PDT.
(Nd:YAG) laser has been associated with CNV The time interval between laser and CNV
development in a physicist from an accidental development ranges from 3 weeks to 10 years
exposure at work [83]. In addition, CNV devel- [95]. Careful assessment of pretreatment images
opment has been reported in a civilian person is essential, as preexisting CNV may confound
who unintentionally looked into the laser beam of the diagnosis of laser-associated CNV in CSCR
military range finder [84, 85] and in a dermatolo- patients [96].
gist who suffered CNV following the use of The visual prognosis of CNV secondary to
Q-switched Nd:YAG laser while treating a focal laser photocoagulation in CSCR is gener-
depressed skin scar [86]. ally favorable because focal laser is usually used
Using Alexandrite laser (750 nm) for hair to treat extrafoveal RPE leaks, away from the
removal has been reported to cause CNV follow- central fovea.
ing momentarily accidental exposure in an aes- Various treatment modalities including
thetician [87]. observation, surgical resection, additional laser
photocoagulation, and additional PDT have
been undertaken to manage laser-induced CNV
19.4.2 Choroidal Neovascularization in CSCR. Recently, Chhablani et al. [95]
Secondary to Therapeutic showed the efficacy and safety of intravitreal
Lasers injections of anti-VEGF agents for the treat-
ment of laser-
associated CNV in
19.4.2.1 Central Serous CSCR. Notably, PDT- associated CNVs were
Chorioretinopathy more resistant to anti-VEGF injections com-
Central serous chorioretinopathy (CSCR) is a pared to those secondary to thermal laser
major cause of vision loss that affects young and photocoagulations.
middle-aged men characterized by idiopathic
serous neurosensory detachment [88]. CNV can 19.4.2.2 Diabetic Macular Edema
develop during the natural course of the disease Intravitreal injections of anti-VEGF agents have
or rarely as a complication of laser therapy. The largely supplanted focal/grid laser as the standard
incidence of CSCR-associated CNV varies from of care for patients with center-involving DME,
2 to 9% depending on the age and length of fol- due to their superior anatomical and functional
low-up [89, 90]. outcome. However, focal/grid laser still holds its
CNV development has been observed follow- role in certain clinical situations including
ing focal laser photocoagulation [91] or PDT adjunctive treatment, focal DME associated with
[92–94] in CSCR patients. The underlying RPE microaneurysms, or in patients with good visual
and choroidal circulation abnormality may pre- acuity.
dispose CSCR patients to laser-induced There are several reports of CNV develop-
CNV. Thermal laser photocoagulation may lead ment following macular photocoagulation in
to Bruch’s membrane (BM) defect; however, DME [97–99]. Intense laser burns, repeated
clinically evident BM defect is usually absent. It application of laser over the same spot, and
is suggested that even the weakening of BM in smaller spot size appear to be associated with
susceptible individuals might contribute to CNV CNV development by damaging to
Fig. 19.4 A 54-year-old man with type 2 diabetes pre- cence due to blockage, compatible to subretinal
sented with acute visual loss in the left eye (OS). There hemorrhage and an area of hyperfluorescence related to
was a history of previous macular photocoagulation a few choroidal neovascularization. (c) Optical coherence
years ago. Best-corrected visual acuity (BCVA) was tomography (OCT) demonstrated increased macular
counting fingers at 3 meters in OS. (a) Hard exudates, thickness, intraretinal fluid, subretinal fluid, and subreti-
microaneurysms, and intraretinal hemorrhages corre- nal hyperreflective material. (d) OCT showed complete
sponding to non-proliferative diabetic retinopathy along resolution of subretinal and intraretinal fluids after intra-
with an area of subretinal hemorrhage (arrow) were evi- vitreal injections of bevacizumab. BCVA improved to
dent. (b) Fluorescein angiography revealed hypofluores- 2/10
BM. Additionally, underlying ischemia and 19.5 Trauma

elevated VEGF levels in diabetic patients may
contribute to the pathogenesis of laser-associ- Ocular trauma can lead to temporary or
ated CNV [100]. Furthermore, older subjects permanent visual impairment through several
with associated alterations in RPE and BM are mechanisms. The main causes of visual decline
particularly at risk for this complication [101]. are hyphema, dislocated lens, vitreous or retinal
The median interval between macular photo- hemorrhage, commotio retinae, and optic neu-
coagulation and CNV occurrence is 8 weeks [97, ropathy. Other causes of decreased vision follow-
98]. Thus, timely follow-up examination is essen- ing trauma include retinal detachment, glaucoma
tial for prompt diagnosis and treatment of this secondary to angle recession, or infrequently
rare complication to preserve vision. Intravitreal CNV. Blunt trauma is the main cause of posttrau-
injection of anti-VEGF agents leads to resolution matic CNV; however, CNV has been reported
of fluid and improvement in vision (Fig. 19.4). after penetrating trauma and alkali burn [106]. In
addition, CNV associated with accidental laser
19.4.2.3 Other Retinal Disorders burns was discussed in a separate section.
CNV may also occur as a complication of laser Following blunt trauma, choroidal rupture can
therapy in various retinal disorders including reti- occur at the site of impact, known as direct cho-
nal vein occlusion [102], sickle cell retinopathy roidal rupture, or more frequently at the opposite
[103], Eales disease [104], and age-related macu- site (counter-coup), which is classified as indirect
lar degeneration [105]. choroidal rupture (ICR) [107].
Direct choroidal rupture typically occurs ante- within 1 year of ocular injury with a median of
riorly, parallel to the limbus, and is associated 5–6 months [117].
with retinal detachment rather than CNV. The diagnosis of rupture associated with CNV
Bruch’s membrane (BM) has a lower tensile is quite straightforward. On examination, subreti-
strength, compared to the elastic retina or rigid nal hemorrhage or subretinal fluid is evident
sclera. Thus, during blunt trauma, anteroposte- along choroidal rupture. Subretinal hemorrhage
rior compression and equatorial expansion of the may also be present early in the course of trauma,
globe creates a shearing force that may cause a without evidence of neovascularization. FA is
break in BM, RPE, and choriocapillaris, resulting useful by showing leakage from CNV network.
in ICR [108]. OCT reveals SRF, IRF, and hyper-reflective sub-
ICR can be present in 5% of eyes with blunt retinal material. Recently, OCT-angiography has
ocular injuries [109, 110]. Clinically, they appear been found useful in demonstrating neovascular
as yellow-white crescentic or curvilinear streaks, tuft within choroidal rupture [118, 119]. Care
concentric to optic disc. ICRs are commonly should be taken not to mistake projection arti-
located temporal to disc. In acute stage, intrareti- facts from normal retinal or choroidal vessels
nal or subretinal hemorrhage can be found along- with neovascularization.
side them. Over weeks, with resolution of Various treatment modalities including obser-
hemorrhage and edema, the white streak with vation, laser photocoagulation [120], surgical
hyperpigmented edges (due to RPE hyperplasia) resection [121], and PDT [122] have been
remains [111]. attempted to manage CNV complicating choroi-
The incidence of choroidal rupture-associated dal rupture. Recently, several studies [113, 117,
CNV after blunt injury varies greatly according 123–126] have reported the efficacy of intravit-
to different studies [110, 112, 113]. Less than 5% real anti-VEGF injection in the treatment of cho-
to up to 30% incidence rates have been reported. roidal rupture-associated CNV in adults and
Eyes with angioid streaks are at special risk for children. Remarkably, most patients required
developing CNV after minor trauma. Choroidal only one or few injections and no case of recur-
ruptures are essential but insufficient for CNV rence has been reported. This good prognosis is
development following ocular injury. It is not attributed to the local nature of the disease entity
clear why some choroidal ruptures are compli- and the young age of patients with remaining
cated by CNV whereas others remain stable. healthy RPE.
Older age, macular locations of ruptures, and
long ruptures have been identified as predispos- Penetrating Injury
ing factors for CNV development [114, 115]. A Reports describing CNV development after pen-
diverse set of interrelated factors contributes to etrating injury are rare. Chen et al. [127] illus-
pathogenesis of CNV secondary to choroidal trated a case with CNV, developed within months
ruptures. Loss of barrier function of BM, loss of following primary repair and vitrectomy for open
RPE and its inhibitory actions against neovascu- globe injury and intraocular foreign body. In
larization, VEGF overexpression caused by another report [128], CNV occurred at the site of
trauma-related inflammation, or release of retinal perforation, 2 months after globe penetra-
elastin-
derived peptides from disrupted BM tion during retrobulbar anesthesia. Both cases
[116] have been proposed as mechanistic links were managed successfully by the use of anti-
between trauma and CNV. VEGF intravitreal injections.
Patients usually present with sudden onset In conclusion, CNV secondary to trauma is a
of decreased vision following a period of sta- major cause of delayed onset visual loss in sub-
ble visual acuity. The time interval between jects with a history of trauma and choroidal rup-
trauma and CNV development varies from few tures. Patients with older age, choroidal rupture in
weeks to several years [110]. However, the macula, and long ruptures are particularly at risk
majority of patients (nearly 80%) present and should be followed regularly, especially within
the first year of trauma. Considering the efficacy been reported [133, 135]. Preverbal children may
and safety of intravitreal anti-VEGF injection, it present with strabismus. Older patients com-
appears to be the best therapeutic approach. monly present with recent-onset metamorphop-
sia or reduced visual acuity.
On examination, the coloboma may involve
19.6 Coloboma the optic nerve head, macula, or iris. RCC typi-
cally occurs inferonasally. Subretinal grayish-
Abnormal closure of embryonic fissure during green mass, subretinal hemorrhage, SRF, and
5–7 weeks of gestation may lead to coloboma of lipid exudates at the margin of coloboma are sug-
optic nerve, retina, choroid, ciliary body, and iris. gestive of CNV development. Care must be taken
These ocular anomalies may occur in isolation or to differentiate CNV membrane from choroidal
together, depending on the extent of abnormality thickening and RPE hyperplasia that typically
[129]. Moreover, coloboma may be a part of sys- develops at the coloboma margin [134]. The
temic diseases such as CHARGE syndrome [130] diagnosis can be confirmed by FA that shows
(Coloboma, Heart defects, Atresia of nasal choa- leakage from the neovascular membrane. OCT is
nae, Retardation of growth, Genitourinary abnor- useful by demonstrating SRF, intraretinal fluid,
malities, and Ear abnormalities). and subretinal hyper-reflective material, and in
Visual acuity varies and is primarily affected monitoring treatment response. It is noteworthy
by the extent of foveal and papillomacular bundle to mention that retinoschisis and serous retinal
involvement. Apart from amblyopia management detachment can occur in subjects with optic nerve
in the affected children, there is no current ther- coloboma in the absence of CNV [136].
apy for coloboma. However, several complica- Several treatment modalities including obser-
tions including early cataract [131], retinal vation [135, 137–139], laser photocoagulation
detachment [132], and choroidal neovasculariza- [130, 140–144], PDT alone [145] or in combina-
tion may arise in the course of the disease that tion with intravitreal anti-VEGF injection, [146]
needs to be identified and treated promptly to intravitreal anti-VEGF injection monotherapy
preserve vision. [147, 148] and surgical resection [149] have been
CNVs complicating optic nerve coloboma used in coloboma-associated CNV. In the absence
(ONC) or retinochoroidal coloboma (RCC) are of a randomized controlled trial (which may not
rare and our knowledge is limited to the case be feasible, given the rarity of this entity), it is not
reports and small case series. possible to recommend the best treatment option
As practically all coloboma associated CNVs with certainty. However, intravitreal anti-VEGF
develop at the margin of coloboma [133], the injections are thought to be the first treatment
pathogenesis of this entity is probably related to option, especially in subfoveal or juxtafoveal
anatomical alterations at the margin. CNVs where laser photocoagulation will likely
Histopathologic study of choroidal coloboma has result in permanent collateral damage.
demonstrated discontinuity or disruption of Bruch’s Prophylactic barrier laser around the margin
membrane (BM) and RPE at the edge of coloboma of coloboma might reduce the risk of retinal
[131]. Loss of barrier function of BM at the margin detachment [150]. Gupta and colleagues [137]
of coloboma may permit the invasion of choroidal suggested that chorioretinal adhesion induced by
vessels into the subretinal space and CNV develop- laser photocoagulation at the margin of coloboma
ment. In addition, the occurrence of CNVs at the might prevent the growth of choroidal vessels
temporal edge of coloboma, in close relation to the and CNV development; however, the efficacy of
fovea, has led some authors to hypothesize the pos- this approach is not clear.
sible role of foveal angiogenic signaling in the In summary, CNV can rarely complicate optic
pathogenesis of coloboma-associated CNV [134]. nerve or retinochoroidal coloboma. Regular exam-
CNV can complicate coloboma at any age. A ination, particularly in children might be helpful in
wide range from 10.5 months to 70 years old has the timely diagnosis and treatment of this entity.
Subjects with coloboma should be instructed to 3. Shields JA, Shields CL. Intraocular tumors: a text and
atlas. Philadelphia: W.B. Saunders Company; 1992.
seek ophthalmologic examination upon occur- p. 85–100.
rence of sudden onset metamorphopsia, decreased 4. Shields CL, Furuta M, Mashayekhi A, Berman EL,
visual acuity, floaters, or flashing. Intravitreal anti- Zahler JD, Hoberman DM, et al. Clinical spectrum of
VEGF injections appear to be safe and efficient in choroidal nevi based on age at presentation in 3422
consecutive eyes. Ophthalmology [Internet]. 2008
the treatment of coloboma-associated CNV. Mar;115(3):546–552.e2. Available from: http://www.
ncbi.nlm.nih.gov/pubmed/18067966.
Key Learning Points 5. Shields CL, Furuta M, Berman EL, Zahler JD,
• Nevus-associated CNV should be included in Hoberman DM, Dinh DH, et al. Choroidal nevus
transformation into melanoma: analysis of 2514 con-
the differential diagnosis of any pigmented secutive cases. Arch Ophthalmol (Chicago, Ill 1960)
choroidal lesion with associated subretinal [Internet]. 2009 Aug;127(8):981–7. Available from:
fluid, subretinal hemorrhage, or exudates. http://www.ncbi.nlm.nih.gov/pubmed/19667334.
• CNV is a common, late complication of cho- 6. Gonder JR, Augsburger JJ, McCarthy EF, Shields
JA. Visual loss associated with choroidal nevi.
roidal osteoma and is a major cause of visual Ophthalmology [Internet]. 1982 Aug;89(8):961–5.
impairment. Available from: http://www.ncbi.nlm.nih.gov/
• CNV may rarely complicate idiopathic intra- pubmed/6182515.
cranial hypertension. 7. Gass JD. Pathogenesis of disciform detachment of the
neuroepithelium. Am J Ophthalmol [Internet]. 1967
• To reduce the risk of MPC-induced CNV, do Mar;63(3):Suppl:1–139. Available from: http://www.
not strive to intensely blanch microaneurysms, ncbi.nlm.nih.gov/pubmed/6019308.
and avoid repeated laser application over the 8. Waltman DD, Gitter KA, Yannuzzi L, Schatz
same spot. H. Choroidal neovascularization associated with
choroidal nevi. Am J Ophthalmol [Internet]. 1978
• High-powered handheld laser pointers are not May;85(5 Pt 1):704–10. Available from: http://www.
playthings as they may result in retinal inju- ncbi.nlm.nih.gov/pubmed/566035.
ries including CNV. 9. Shields CL, Furuta M, Mashayekhi A, Berman EL,
• Subjects with traumatic choroidal rupture in Zahler JD, Hoberman DM, et al. Visual acuity in
3422 consecutive eyes with choroidal nevus. Arch
macula should be followed regularly, particu- Ophthalmol (Chicago, Ill 1960) [Internet]. 2007
larly during the first year after trauma to detect Nov;125(11):1501–7. Available from: http://www.
early signs of CNV development. ncbi.nlm.nih.gov/pubmed/17998511.
• Patients with retinochoroidal coloboma should 10. Zografos L, Mantel I, Schalenbourg A. Subretinal
choroidal neovascularization associated with choroi-
be instructed to seek ophthalmologic examina- dal nevus. Eur J Ophthalmol [Internet]. 14(2):123–
tion upon the occurrence of sudden onset of 31. Available from: http://www.ncbi.nlm.nih.gov/
metamorphopsia, and decreased visual acuity, pubmed/15134109.
to detect possible CNV. 11. Chiang A, Bianciotto C, Maguire JI, Park CH, Baker
PS, Shields JA, et al. Intravitreal bevacizumab for
choroidal neovascularization associated with cho-
Competing Interest Statement None of the authors has roidal nevus. Retina [Internet]. 2012 Jan;32(1):60–
any financial interest in the subject matter of this chapter. 7. Available from: http://www.ncbi.nlm.nih.gov/
pubmed/21886019.
12. Namavari A, Zheng F, Motulsky EH, de Oliveira Dias
JR, Gregori G, Rosenfeld PJ. Swept-Source OCT
References Angiography identifies Choroidal neovascularization
arising from a Choroidal nevus. Ophthalmic Surg
1. Callanan DG, Lewis ML, Byrne SF, Gass lasers imaging Retina [Internet]. 2018;49(5):360–
JD. Choroidal neovascularization associated with 3. Available from: http://www.ncbi.nlm.nih.gov/
choroidal nevi. Arch Ophthalmol (Chicago, Ill 1960) pubmed/29772047.
[Internet]. 1993 Jun;111(6):789–94. Available from: 13. Tuncer S, Tugal-Tutkun I. Choroidal neovascular-
http://www.ncbi.nlm.nih.gov/pubmed/7685589. ization secondary to choroidal nevus simulating an
2. Sumich P, Mitchell P, Wang JJ. Choroidal nevi in a inflammatory lesion. Indian J Ophthalmol [Internet].
white population: the Blue Mountains Eye Study. 2013 Jun;61(6):305–6. Available from: http://www.
Arch Ophthalmol (Chicago, Ill 1960) [Internet]. 1998 ncbi.nlm.nih.gov/pubmed/23571241.
May;116(5):645–50. Available from: http://www. 14. Pellegrini M, Corvi F, Say EAT, Shields CL,
ncbi.nlm.nih.gov/pubmed/9596501. Staurenghi G. Optical coherence tomography angi-
ography features of choroidal neovascularization cated by choroidal neovascular membrane and outer
associated with choroidal nevus. Retina [Internet]. retinal tubulation. Br J Ophthalmol [Internet]. 2013
2018;38(7):1338–46. Available from: http://www. Aug;97(8):1014–9. Available from: http://www.ncbi.
ncbi.nlm.nih.gov/pubmed/28570484. nlm.nih.gov/pubmed/23686326.
15. Pascual-Camps I, López-Corell PM, Andreu-Fenoll 26. Munie MT, Demirci H. Management of Choroidal
M, Gallego-Pinazo R. Polypoidal choroidopathy Neovascular Membranes Associated with Choroidal
associated with choroidal nevus. Indian J Ophthalmol Nevi. Ophthalmol Retin [Internet]. 2018 Jan;2(1):53–
[Internet]. 2017;65(11):1213–4. Available from: 8. Available from: http://www.ncbi.nlm.nih.gov/
http://www.ncbi.nlm.nih.gov/pubmed/29133655. pubmed/31047303.
16. De Salvo G, Vaz-Pereira S, Sehmi KS, Andrews 27. Lee J, Kwon HJ, Kim M, Lee CS, Lee SC. Treatment
RM, Sagoo MS. Spectral-domain optical coherence response to intravitreal bevacizumab in small pig-
tomography of Polypoidal Choroidal vasculopathy mented choroidal lesions with subretinal fluid. BMC
associated with benign Choroidal nevus. Ophthalmic Ophthalmol [Internet]. 2019 May 3;19(1):103.
Surg lasers imaging Retina [Internet]. 46(10):1062– Available from: http://www.ncbi.nlm.nih.gov/
4. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/31053116.
pubmed/26599253. 28. Gass JDM, Guerry RK, Jack RL, Harris G. Choroidal
17. Falavarjani KG, Sarraf D. Optical coherence tomog- Osteoma. Arch Ophthalmol [Internet]. 1978 Mar
raphy angiography of the retina and choroid; current 1;96(3):428–35. Available from: http://archopht.jam-
applications and future directions. J Curr Ophthalmol. anetwork.com/article.aspx?articleid=632514
2017;29(1):1–4. 29. Shields CL, Shields JA, Augsburger JJ. Choroidal
18. Ghasemi Falavarjani K, Al-Sheikh M, Akil H, Sadda osteoma. Surv Ophthalmol [Internet] 1988
SR. Image artefacts in swept-source optical coher- Jul;33(1):17–27. Available from: https://linkinghub.
ence tomography angiography. Br J Ophthalmol elsevier.com/retrieve/pii/0039625788900690
[Internet]. 2017 May [cited 2017 Jul 13];101(5):564– 30. Trimble SN, Schatz H. Choroidal Osteoma after
8. Available from: http://www.ncbi.nlm.nih.gov/ intraocular inflammation. Am J Ophthalmol
pubmed/27439739. [Internet]. 1983 Dec;96(6):759–64. Available
19. Gass JD. Problems in the differential diagnosis of cho- from: https://linkinghub.elsevier.com/retrieve/pii/
roidal nevi and malignant melanomas. The XXXIII S000293941471921X
Edward Jackson Memorial Lecture. Am J Ophthalmol 31. Gass JD. New observations concerning choroi-
[Internet]. 1977 Mar;83(3):299–323. Available from: dal osteomas. Int Ophthalmol [Internet]. 1979
http://www.ncbi.nlm.nih.gov/pubmed/848534. Feb;1(2):71–84. Available from: http://www.ncbi.
20. Lubin JR, Gragoudas ES, Albert DM. Choroidal neo- nlm.nih.gov/pubmed/553048.
vascularization associated with malignant melanoma: 32. McLeod BK. Choroidal osteoma presenting in
a case report. Acta Ophthalmol [Internet]. 1982 pregnancy. Br J Ophthalmol [Internet] 1988 Aug
Jun;60(3):412–8. Available from: http://www.ncbi. 1;72(8):612–4. Available from: http://bjo.bmj.com/
nlm.nih.gov/pubmed/6182734. cgi/doi/10.1136/bjo.72.8.612
21. Parodi MB. Transpupillary thermotherapy for sub- 33. Aylward GW. A long-term follow-up of choroidal
foveal choroidal neovascularization associated with osteoma. Arch Ophthalmol [Internet]. 1998 Oct
choroidal nevus. Am J Ophthalmol [Internet]. 2004 1;116(10):1337. Available from: http://archopht.
Dec;138(6):1074–5. Available from: http://www.ncbi. jamanetwork.com/article.aspx?doi=10.1001/
nlm.nih.gov/pubmed/15629316. archopht.116.10.1337
22. Parodi MB, Boscia F, Piermarocchi S, Ferrari TM, 34. Shields CL, Perez B, Materin MA, Mehta S, Shields
Furino C, Sborgia C. Variable outcome of photody- JA. Optical coherence tomography of Choroidal
namic therapy for choroidal neovascularization asso- Osteoma in 22 cases. Ophthalmology [Internet]. 2007
ciated with choroidal nevus. Retina [Internet]. 2005 Dec;114(12):e53–8. Available from: https://linking-
Jun;25(4):438–42. Available from: http://www.ncbi. hub.elsevier.com/retrieve/pii/S0161642007008536
nlm.nih.gov/pubmed/15933589. 35. Shields CL, Sun H, Demirci H, Shields JA. Factors
23. Stanescu D, Wattenberg S, Cohen SY. Photodynamic predictive of tumor growth, tumor decalcification,
therapy for choroidal neovascularization secondary to choroidal neovascularization, and visual outcome in
choroidal nevus. Am J Ophthalmol [Internet]. 2003 74 eyes with choroidal osteoma. Arch Ophthalmol.
Sep;136(3):575–6. Available from: https://linkinghub. 2005;123(12):1658–66.
elsevier.com/retrieve/pii/S0002939403003210. 36. Foster BS. Surgical Removal and Histopathologic
24. Cavalcante ML, Villegas VM, Gold AS, Cavalcante Findings of a Subfoveal Neovascular Membrane
LL, Lonngi M, Shah N V, et al. Treatment of vascular Associated With Choroidal Osteoma. Arch
activity secondary to atypical choroidal nevus using Ophthalmol [Internet]. 2003 Feb 1;121(2):273.
intravitreal bevacizumab. Clin Ophthalmol [Internet]. Available from: http://archopht.jamanetwork.com/
2014;8:1377–82. Available from: http://www.ncbi. article.aspx?doi=10.1001/archopht.121.2.273
nlm.nih.gov/pubmed/25092961. 37. Pierro L, Marchese A, Gagliardi M, Introini U,
25. Papastefanou VP, Nogueira V, Hay G, Andrews RM, Battaglia Parodi M, Casalino G, et al. Choroidal
Harris M, Cohen VML, et al. Choroidal naevi compli- excavation in choroidal osteoma complicated by
choroidal neovascularization. Eye [Internet] 2017 47. Palamar M, Uretmen O, Gündüz K. Transient sub-
Dec 21;31(12):1740–3. Available from: http://www. retinal hemorrhage after photodynamic therapy of
nature.com/doifinder/10.1038/eye.2017.136 subfoveal choroidal osteoma. Retin cases brief rep
38. Yoshikawa T, Takahashi K. Long-term outcomes of [Internet]. 2012;6(2):166–8. Available from: http://
intravitreal injection of bevacizumab for choroidal www.ncbi.nlm.nih.gov/pubmed/25390953.
neovascularization associated with choroidal oste- 48. Morris R, Shah P, Prabhu V, Narendran V. Combination
oma. Clin Ophthalmol [Internet]. 2015 Mar;429. therapy of low-fluence photodynamic therapy and
Available from: http://www.dovepress.com/long- intravitreal ranibizumab for choroidal neovascular
term-outcomes-of-intravitreal-injection-ofnbspbeva- membrane in choroidal osteoma. Indian J Ophthalmol
cizumab-for-cho-peer-reviewed-article-OPTH [Internet]. 2011;59(5):394. Available from: http://
39. Chawla R, Azad SV, Takkar B, Sharma A, Kashyap www.ijo.in/text.asp?2011/59/5/394/83622
B. Nonconforming Deep Focal Choroidal Excavation 49. Mansour AM, Arevalo JF, Al Kahtani E, Zegarra
in a Patient With Choroidal Osteoma: A Diagnostic H, Abboud E, Anand R, et al. Role of Intravitreal
Dilemma. Ophthalmic Surgery, Lasers Imaging Antivascular endothelial growth factor injections
Retin [Internet]. 2017 Nov 1;48(11):944–7. for Choroidal neovascularization due to Choroidal
Available from: http://www.healio.com/doiresolver? Osteoma. J Ophthalmol [Internet] 2014;2014:1–8.
doi=10.3928/23258160-20171030-12 Available from: http://www.hindawi.com/journals/
40. Papastefanou VP, Pefkianaki M, Al Harby L, Arora joph/2014/210458/
AK, Cohen VML, Andrews RM, et al. Intravitreal 50. Khan MA, DeCroos FC, Storey PP, Shields JA, Garg
bevacizumab monotherapy for choroidal neovascular- SJ, Shields CL. Outcomes of anti-vascular endothe-
isation secondary to choroidal osteoma. Eye (Lond) lial growth factor therapy in the management of cho-
[Internet]. 2016 Jun;30(6):843–9. Available from: roidal neovascularization associated with choroidal
http://www.ncbi.nlm.nih.gov/pubmed/27034203. osteoma. Retina [Internet]. 2014 Sep;34(9):1750–6.
41. Mello PC de, Berensztejn P, Brasil OFM. Available from: http://www.ncbi.nlm.nih.gov/
Enhanced depth imaging optical coherence pubmed/24936941.
tomography of choroidal osteoma with second- 51. Saitta A, Nicolai M, Neri P, Reibaldi M, Giovannini
ary neovascular membranes: report of two cases. A, Mariotti C. Rescue therapy with intravitreal
Arq Bras Oftalmol [Internet] 2016;79(3):197–9. aflibercept for choroidal neovascularization sec-
Available from: http://www.gnresearch.org/ ondary to choroidal osteoma non-responder to
doi/10.5935/0004-2749.20160057 intravitreal bevacizumab and ranibizumab. Int
42. Shen C, Yan S, Du M, Zhao H, Shao L, Hu Ophthalmol [Internet]. 2015 Jun 12;35(3):441–4.
Y. Assessment of choroidal osteoma complicating Available from: http://link.springer.com/10.1007/
choroidal neovascularization by optical coherence s10792-015-0059-5
tomography angiography. Int Ophthalmol [Internet]. 52. Ahmadieh H, Vafi N. Dramatic response of choroidal
2018 Apr;38(2):787–92. Available from: http://link. neovascularization associated with choroidal osteoma
springer.com/10.1007/s10792-017-0503-9 to the intravitreal injection of bevacizumab (Avastin).
43. Olguin-Manríquez F, Enríquez AB, Crim N, Meraz- Graefe’s Arch Clin Exp Ophthalmol [Internet]. 2007
Gutierrez M, Soberón-Ventura V, Ávila I, et al. Sep 24;245(11):1731–3. Available from: http://link.
Multimodal imaging in choroidal osteoma. Int J Retin springer.com/10.1007/s00417-007-0636-z
Vitr [Internet]. 2018;4:30. Available from: http:// 53. Erol MK, Coban DT, Ceran BB, Bulut M. Retinal pig-
www.ncbi.nlm.nih.gov/pubmed/30128167. ment epithelium tear formation following intravitreal
44. Furino C, Di Antonio L, Grassi MO, Rispoli M, ranibizumab injection in choroidal neovascularization
Reibaldi M, Niro A, et al. Choroidal neovascu- secondary to choroidal osteoma. Cutan Ocul Toxicol
larization due to choroidal osteoma treated with [Internet] 2014 Sep 11;33(3):259–63. Available from:
anti–vascular endothelial growth factor therapy: http://www.tandfonline.com/doi/full/10.3109/155695
An optical coherence tomography angiography 27.2013.844702
study. Eur J Ophthalmol [Internet]. 2018 Aug 54. Chen J, Lee L, Gass JDM. Choroidal osteoma:
13;112067211879218. Available from: http://jour- evidence of progression and decalcification
nals.sagepub.com/doi/10.1177/1120672118792187 over 20 years. Clin Exp Optom [Internet] 2006
45. Sharma S, Sribhargava N, Shanmugam MP. Choroidal Mar;89(2):90–4. Available from: http://doi.wiley.
neovascular membrane associated with choroidal com/10.1111/j.1444-0938.2006.00012.x
osteoma (CO) treated with transpupillary thermo 55. Mansour A, Alameddine R, Kahtani E. Review of
therapy. Indian J Ophthalmol [Internet]. 2004 choroidal osteomas. Middle East Afr J Ophthalmol
Dec;52(4):329–30. Available from: http://www.ncbi. [Internet]. 2014;21(3):244. Available from: http://
nlm.nih.gov/pubmed/15693329. www.meajo.org/text.asp?2014/21/3/244/134686
46. Aylward GW, Chang TS, Pautler SE, Gass JD. A 56. Song JH, Bae JH, Rho MI, Lee SC. Intravitreal beva-
long-term follow-up of choroidal osteoma. Arch cizumab in the management of subretinal fluid associ-
Ophthalmol (Chicago, Ill 1960) [Internet]. 1998 ated with choroidal osteoma. Retina [Internet] 2010
Oct;116(10):1337–41. Available from: http://www. Jun;30(6):945–51. Available from: https://insights.
ncbi.nlm.nih.gov/pubmed/9790633. ovid.com/crossref?an=00006982-201006000-00015
57. Najafabadi F, Hendimarjan S, Zarrin Y, Najafabadi 67. Castellarin AA, Sugino IK, Nasir M’a., Zarbin
M. Intravitreal bevacizumab for management MA. Clinicopathological correlation of an excised
of choroidal osteoma without choroidal neovas- choroidal neovascular membrane in pseudotumour
cularization. J Ophthalmic Vis Res [Internet]. cerebri. Br J Ophthalmol [Internet]. 1997 Nov
2015;10(4):484. Available from: http://www.jovr.org/ 1;81(11):994–1000. Available from: http://bjo.bmj.
text.asp?2015/10/4/484/176905 com/cgi/doi/10.1136/bjo.81.11.994
58. Ben Yahia S, Zaouali S, Attia S, Messaoud R, 68. Jamison RR. Subretinal neovascularization and
Ladjimi A, Khairallah M. Serous retinal detach- papilledema associated with pseudotumor cerebri.
ment secondary to choroidal osteoma success- Am J Ophthalmol [Internet] 1978 Jan;85(1):78–81.
fully treated with transpupillary thermotherapy. Available from: https://linkinghub.elsevier.com/
Retin Cases Brief Rep [Internet]. 2008;2(2):126–7. retrieve/pii/S0002939414766693
Available from: https://insights.ovid.com/crossref 69. Jamerson SC, Arunagiri G, Ellis BD, Leys
?an=01271216-200800220-00014 MJ. Intravitreal bevacizumab for the treatment of
59. Skau M, Brennum J, Gjerris F, Jensen R. What is choroidal neovascularization secondary to pseudo-
new about idiopathic intracranial hypertension? tumor cerebri. Int Ophthalmol [Internet] 2009 Jun
An updated review of mechanism and treatment. 15;29(3):183–5. Available from: http://link.springer.
Cephalalgia [Internet]. 2006 Apr;26(4):384–99. com/10.1007/s10792-007-9186-y
Available from: http://www.ncbi.nlm.nih.gov/ 70. Más-Ramírez AM, Villegas VM, Garcia JM, Acevedo
pubmed/16556239. S, Serrano L. Intravitreal bevacizumab for peri-
60. Chen JJ, Thurtell MJ, Longmuir RA, Garvin MK, papillary subretinal neovascular membrane associ-
Wang J-K, Wall M, et al. Causes and Prognosis of ated to papilledema: a case report. P R Health Sci J.
Visual Acuity Loss at the Time of Initial Presentation 2012;31(3):148–150.
in Idiopathic Intracranial Hypertension. Invest 71. Lee I-J, Maccheron LJ, Kwan AS. Intravitreal
Ophthalmol Vis Sci [Internet]. 2015 Jun;56(6):3850– Bevacizumab in the treatment of Peripapillary
9. Available from: http://www.ncbi.nlm.nih.gov/ Choroidal Neovascular membrane secondary to
pubmed/26070058. idiopathic intracranial hypertension. J Neuro-
61. Wendel L, Lee AG, Boldt HC, Kardon RH, Wall Ophthalmology [Internet]. 2013 Jun;33(2):155–7.
M. Subretinal neovascular membrane in idio- Available from: http://content.wkhealth.com/
pathic intracranial hypertension. Am J Ophthalmol linkback/openurl?sid=WKPTLP:landingpage
[Internet]. 2006 Mar;141(3):573–4. Available from: &an=00041327-201306000-00014
http://www.ncbi.nlm.nih.gov/pubmed/16490514. 72. Kumar N, Tigari B, Dogra M, Singh R. Successful man-
62. Morse PH, Leveille AS, Antel JP, Burch J V. Bilateral agement of peripapillary choroidal neovascular mem-
juxtapapillary subretinal neovascularization associ- brane secondary to idiopathic intracranial hypertension
ated with pseudotumor cerebri. Am J Ophthalmol with intravitreal ranibizumab. Indian J Ophthalmol
[Internet]. 1981 Mar;91(3):312–7. Available from: [Internet]. 2018;66(9):1358. Available from: http://
http://www.ncbi.nlm.nih.gov/pubmed/6163360. www.ijo.in/text.asp?2018/66/9/1358/239359
63. Belliveau MJ, Xing L, Almeida DRP, Gale JS, 73. Barkana Y, Belkin M. Laser eye injuries. Surv
ten Hove MW. Peripapillary choroidal neo- Ophthalmol [Internet]. 2000 May;44(6):459–78.
vascular membrane in a teenage boy. J Neuro- Available from: http://linkinghub.elsevier.com/
Ophthalmology [Internet] 2013 Mar;33(1):48–50. retrieve/pii/S0039625700001120
Available from: https://insights.ovid.com/crossref 74. Laser Institute of America LIA. ANSI Z136.1:
?an=00041327-201303000-00012 American national standard for safe use of lasers.
64. Browning D, Fraser C. Ocular Conditions associ- SPIE Med Imaging. 2007;
ated with peripapillary subretinal neovascularization, 75. Ajudua S, Mello MJ. Shedding some light on
their relative frequencies, and associated outcomes. laser pointer eye injuries. Pediatr Emerg Care.
Ophthalmology [Internet]. 2005 Jun;112(6):1054– 2007;23(9):669–72.
1061.e2. Available from: https://linkinghub.elsevier. 76. Torp-Pedersen T, Welinder L, Justesen B, Christensen
com/retrieve/pii/S016164200500117X UC, Solborg Bjerrum S, La Cour M, et al. Laser
65. Kaeser P-F, Borruat F-X. Peripapillary neovascu- pointer maculopathy – on the rise? Acta Ophthalmol.
lar membrane: a rare cause of acute vision loss in 2018;96(7):749–54.
pediatric idiopathic intracranial hypertension. J Am 77. Wyrsch S, Baenninger PB, Schmid MK. Retinal injuries
Assoc Pediatr Ophthalmol Strabismus [Internet] 2011 from a handheld laser pointer. N Engl J Med [Internet]
Feb;15(1):83–6. Available from: https://linkinghub. 2010;363(11):1089–91. Available from: http://www.
elsevier.com/retrieve/pii/S1091853110005422 nejm.org/doi/abs/10.1056/NEJMc1005818
66. Sathornsumetee B, Webb A, Hill DL, Newman 78. Alsulaiman SM, Alrushood AA, Almasaud J, Alzaaidi
NJ, Biousse V. Subretinal hemorrhage from a peri- S, Alzahrani Y, Arevalo JF, et al. High-power hand-
papillary choroidal neovascular membrane in held blue laser-induced maculopathy: the results of
papilledema caused by idiopathic intracranial hyper- the king khaled eye specialist hospital collabora-
tension. J Neuro-Ophthalmology [Internet]. 2006 tive retina study group. Ophthalmology [Internet]
Sep;26(3):197–9. Available from: https://insights. 2014;121(2):566–72. Available from: https://doi.
ovid.com/crossref?an=00041327-200609000-00011 org/10.1016/j.ophtha.2013.09.006
79. Turaka K, Bryan JS, Gordon AJ, Reddy R, Kwong 93. Colucciello M. Choroidal neovascularization com-
HM, Sell CH. Laser pointer induced macular dam- plicating photodynamic therapy for central serous
age: case report and mini review. Int Ophthalmol. retinopathy. Retina. 2006;26(2):239–4.
2012;32(3):293–7. 94. Chan WM, Lam DSC, Lai TYY, Tam BSM, Liu
80. Ghasemi Falavarjani K, Au A, Anvari P, Mollaei S, DTL, Chan CKM. Choroidal vascular remodelling
Ghasemizadeh S, Verma A, et al. En face OCT of in central serous chorioretinopathy after indocyanine
type 2 neovascularization: a reappraisal of the pitch- green guided photodynamic therapy with vertepor-
fork sign. Ophthalmic Surg Lasers Imaging Retin. fin: a novel treatment at the primary disease level. Br
2019;50(11):719–25. J Ophthalmol. 2003;87(12):1453–8.
81. Fujinami K, Yokoi T, Hiraoka M, Nishina S, Azuma 95. Chhablani J, Pichi F, Silva R, Casella AM, Murthy
N. Choroidal neovascularization in a child follow- H, Banker A, et al. Antiangiogenics in Choroidal
ing laser pointer-induced macular injury. Jpn J Neovascularization associated with laser in central
Ophthalmol [Internet] 2010 Nov 28;54(6):631–3. serous Chorioretinopathy. Retina. 2016;36(5):901–8.
Available from: http://link.springer.com/10.1007/ 96. Matsunaga H, Nangoh K, Uyama M, Nanbu H,
s10384-010-0871-4 Fujiseki Y, Takahashi K. Occurrence of choroi-
82. Xu K, Chin EK, Quiram PA, Davies JB, Parke dal neovascularization following photocoagu-
DW, DRP A. Retinal injury secondary to lation treatment for central serous retinopathy.
laser pointers in pediatric patients. Pediatrics. Nihon Ganka Gakkai Zasshi [Internet]. 1995
2016;138(4):e20161188. Apr;99(4):460–8. Available from: http://www.ncbi.
83. Ying HS, Symons RCA, Lin KL, Solomon SD, nlm.nih.gov/pubmed/7537938.
Gehlbach PL. Accidental Nd:YAG laser-induced 97. Varley MP, Frank E, Purnell EW. Subretinal neovas-
choroidal neovascularization. Lasers Surg Med. cularization after focal argon laser for diabetic mac-
2008;40(4):240–2. ular edema. Ophthalmology. 1988;95(5):567–73.
84. Roider J, Buesgen P, Hoerauf H, Schmidt-Erfurth U, 98. Lewis H, Schachat AP, Haimann MH, Haller JA,
Laqua H, Birngruber R. Macular injury by a military Quinlan P, von Fricken MA, et al. Choroidal neo-
range finder. Retina. 1999;19:531–5. vascularization after laser photocoagulation for
85. Modarres-Zadeh M, Parvaresh MM, Pourbabak diabetic macular edema. Ophthalmology [Internet].
S, Peyman GA. Accidental parafoveal laser burn 1990 Apr;97(4):503–10; discussion 510–1.
from a standard military ruby range finder. Retina. Available from: http://www.ncbi.nlm.nih.gov/
1995;15(4):356–8. pubmed/1691477.
86. Lee YH, Kim YC. Foveal choroidal neovasculariza- 99. Shah GK. Photodynamic therapy for choroidal neo-
tion secondary to accidental laser exposure in a der- vascularization after thermal laser photocoagula-
matologist. Medicine (Baltimore) [Internet]. 2019 tion for diabetic macular edema. Am J Ophthalmol
May;98(18):e15429. Available from: http://insights. [Internet]. 2003 Jan;135(1):114–6. Available from:
ovid.com/crossref?an=00005792-201905030-00049 http://www.ncbi.nlm.nih.gov/pubmed/12504717.
87. Wang R, Wykoff CC, Christie L, Croft DE, Major 100. Deschler EK, Sun JK, Silva PS. Side-effects and
JC, Fish RH, et al. Choroidal neovascularization sec- complications of laser treatment in diabetic retinal
ondary to alexandrite laser exposure. Retin Cases Br disease. Semin Ophthalmol. 2014;29(5–6):290–300.
Rep. 2016;10(3):244–8. 101. Berger AR, Boniuk I. Bilateral subretinal neovas-
88. Daruich A, Matet A, Dirani A, Bousquet E, Zhao M, cularization after focal argon laser photocoagula-
Farman N, et al. Central serous chorioretinopathy: tion for diabetic macular edema. Am J Ophthalmol.
recent findings and new physiopathology hypoth- 1989;108(1):88–90.
esis. Prog Retin Eye Res. 2015;48:82–118. 102. Yarng SS, Hsieh CL. Choriovitreal neovasculariza-
89. Loo RH, Scott IU, Flynn HW, JDM G, Murray TG, tion following laser-induced chorioretinal venous
Lou LM, et al. Factors associated with reduced anastomosis. Ophthalmology. 1996;103(suppl):136.
visual acuity during long-term follow-up of patients 103. Jacobson MS, Gagliano DA, Cohen SB, Rabb MF,
with idiopathic central serous chorioretinopathy. Jampol LM, Farber MD, et al. A randomized clinical
Retina. 2002;22(1):19–2. trial of feeder vessel photocoagulation of sickle cell
90. Spaide RF, Campeas L, Haas A, Yannuzzi LA, Fisher retinopathy. A long-term follow-up. Ophthalmology
YL, Guyer DR, et al. Central serous chorioretinopa- [Internet]. 1991 May;98(5):581–5. Available from:
thy in younger and older adults. Ophthalmology. http://www.ncbi.nlm.nih.gov/pubmed/2062489.
1996;103(12):2070–9. 104. Theodossiadis G, Velissaropoulos P. Choroidal vascu-
91. Çakir M, Çekiç O, Yilmaz ÖF. Photodynamic ther- lar involvement in Eales’ disease. Ophthalmologica.
apy for iatrogenic CNV due to laser photocoagula- 1973;166(1):1–9.
tion in central serous Chorioretinopathy. Ophthalmic 105. Choroidal Neovascularization Prevention Trial
Surg Lasers Imaging. 2009;40(4):405–8. Research Group. Laser treatment in eyes with
92. Yaman A, Arikan G, Saatci AO, Cingil G. Choroidal large drusen. Ophthalmology [Internet]. 1998
neovascularization following photodynamic therapy Jan;105(1):11–23. Available from: http://www.ncbi.
in a patient with chronic central serous chorio- nlm.nih.gov/pubmed/9442774.
retinopathy. Bull Soc Belge Ophtalmol. 2007;303: 106. Al-Husainy S, Stavrou P, Hope-Ross MW. Alkali
69–73. injury complicated by choroidal neovascularisation.
Eye (Lond) [Internet]. 2000 Dec;14(Pt 6):904–5. neovascularization secondary to traumatic choroi-
Available from: http://www.ncbi.nlm.nih.gov/ dal rupture. Retin Cases Brief Rep [Internet]. 2018
pubmed/11584853. Mar;1. Available from: http://insights.ovid.com/cros
107. Aguilar JP, Green WR. Choroidal rupture. A his- sref?an=01271216-900000000-98840
topathologic study of 47 cases. Retina [Internet]. 119. Lorusso M, Micelli Ferrari L, Nikolopoulou E, Micelli
4(4):269–75. Available from: http://www.ncbi.nlm. Ferrari T. Optical coherence tomography angiogra-
nih.gov/pubmed/6531524. phy evolution of choroidal neovascular membrane
108. Pujari A, Chawla R, Agarwal D, Gagrani M, Kapoor in choroidal rupture managed by intravitreal beva-
S, Kumar A. Pathomechanism of traumatic indirect cizumab. Case Rep Ophthalmol Med [Internet].
choroidal rupture. Med Hypotheses [Internet]. 2019 2019 Jan 6;2019:1–4. Available from: https://www.
Mar;124:64–6. Available from: https://linkinghub. hindawi.com/journals/criopm/2019/5241573/
elsevier.com/retrieve/pii/S0306987718311617 120. Smith RE, Kelley JS, Harbin TS. Late macular com-
109. Bellows JG. Observations on 300 consecutive cases plications of Choroidal ruptures. Am J Ophthalmol
of ocular war injuries. Am J Ophthalmol [Internet]. [Internet]. 1974 May;77(5):650–8. Available
1947 Mar;30(3):309–23. Available from: http:// from: https://linkinghub.elsevier.com/retrieve/
www.ncbi.nlm.nih.gov/pubmed/20288592. pii/0002939474905261
110. Wood CM, Richardson J. Chorioretinal neovascular 121. Abri A, Binder S, Pavelka M, Tittl M, Neumuller
membranes complicating contusional eye injuries J. Choroidal neovascularization in a child with
with indirect choroidal ruptures. Br J Ophthalmol traumatic choroidal rupture: clinical and ultrastruc-
[Internet]. 1990 Feb;74(2):93–6. Available from: tural findings. Clin Exp Ophthalmol [Internet] 2006
http://www.ncbi.nlm.nih.gov/pubmed/1690025. Jul;34(5):460–3. Available from: http://doi.wiley.
111. Pierro L, Giuffrè C, Rabiolo A, Gagliardi M, Arrigo com/10.1111/j.1442-9071.2006.01248.x
A, Bandello F. Multimodal imaging in a patient with 122. Harissi-Dagher M, Sebag M, Gauthier D, Marcil
traumatic Choroidal ruptures. Eur J Ophthalmol G, Labelle P, Arbour JD. Photodynamic therapy
[Internet] 2017 Nov 8;27(6):e175–8. Available in young patients with choroidal neovasculariza-
from: http://journals.sagepub.com/doi/10.5301/ tion following traumatic choroidal rupture. Am J
ejo.5001005 Ophthalmol [Internet]. 2005 Apr;139(4):726–8.
112. Secrétan M, Sickenberg M, Zografos L, Piguet Available from: https://linkinghub.elsevier.com/
B. Morphometric characteristics of traumatic cho- retrieve/pii/S0002939404012498
roidal ruptures associated with neovascularization. 123. Liang F, Puche N, Soubrane G, Souied EH.
Retina [Internet]. 1998;18(1):62–6. Available from: Intravitreal ranibizumab for choroidal neovascular-
http://www.ncbi.nlm.nih.gov/pubmed/9502283. ization related to traumatic Bruch’s membrane rup-
113. Chanana B, Azad R V, Kumar N. Intravitreal beva- ture. Graefe’s Arch Clin Exp Ophthalmol [Internet]
cizumab for subfoveal choroidal neovasculariza- 2009 Sep 26;247(9):1285–8. Available from: http://
tion secondary to traumatic choroidal rupture. Eye link.springer.com/10.1007/s00417-009-1098-2
[Internet]. 2009 Nov 23;23(11):2125–6. Available 124. De Benedetto U, Battaglia Parodi M, Knutsson KA,
from: http://www.nature.com/articles/eye2008434 Librando A, Bandello F, Lanzetta P, et al. Intravitreal
114. Ament CS. Predictors of visual outcome and choroi- bevacizumab for extrafoveal choroidal neovascu-
dal neovascular membrane formation after traumatic larization after ocular trauma. J Ocul Pharmacol
choroidal rupture. Arch Ophthalmol [Internet]. 2006 Ther [Internet] 2012 Oct;28(5):550–2. Available
Jul 1;124(7):957. Available from: http://archopht. from: http://www.liebertpub.com/doi/10.1089/
jamanetwork.com/article.aspx?doi=10.1001/ jop.2012.0022
archopht.124.7.957 125. Artunay O, Rasier R, Yuzbasioglu E, Sengül A,
115. Secrétan M, Sickenberg M, Zografos L, Piguet Bahcecioglu H. Intravitreal bevacizumab injec-
B. Morphometric characteristics of traumatic cho- tion in patients with choroidal neovascularization
roidal ruptures associated with neovascularization. due to choroid rupture after blunt-head trauma. Int
Retina [Internet]. 1998;18(1):62–6. Available from: Ophthalmol [Internet]. 2009 Aug 30;29(4):289–91.
http://www.ncbi.nlm.nih.gov/pubmed/9502283. Available from: http://link.springer.com/10.1007/
116. Wang H, Ninomiya Y, Sugino IK, Zarbin MA. Retinal s10792-008-9226-2
pigment epithelium wound healing in human 126. Prasad A, Patel CC, Puklin JE. Intravitreal beva-
Bruch’s membrane explants. Invest Ophthalmol cizumab in the treatment of choroidal neovascu-
Vis Sci [Internet]. 2003 May;44(5):2199–210. larization from a traumatic choroidal rupture in a
Available from: http://www.ncbi.nlm.nih.gov/ 9-year-old child. Retin Cases Brief Rep [Internet]
pubmed/12714662. 2009;3(2):125–7. Available from: https://insights.
117. Joo K, Um T, Park SJ, Cho JH, Ahn J, Kim JT, et al. ovid.com/crossref?an=01271216-200900320-00006
Efficacy of bevacizumab for posttraumatic choroi- 127. Chen T-L, Sun M-H, Lin K-K, Lai C-C, Chen
dal neovascularization. Acta Ophthalmol [Internet] K-J. Intravitreal Bevacizumab With Regression of
2019 Mar;97(2):e324–6. Available from: http://doi. Subretinal Neovascularization After Intraocular
wiley.com/10.1111/aos.13896 Foreign Body Trauma. J Trauma Inj Infect
118. Preziosa C, Corvi F, Pellegrini M, Bochicchio S, Crit Care [Internet]. 2010 Mar;68(3):747.
Rosar AP, Staurenghi G. Optical coherence tomog- Available from: https://insights.ovid.com/crossref
raphy angiography findings in a case of choroidal ?an=00005373-201003000-00046
128. Dikci S, Yılmaz T, Gök Z, Demirel S, Genç O. 140. Maberley AL, Gottner MJ, Antworth M V. Subretinal
Choroidal neovascularization secondary to ocu- neovascularization associated with retinochoroi-
lar penetration during retrobulbar anesthesia and dal colobomas. Can J Ophthalmol [Internet]. 1989
its treatment. Oman J Ophthalmol [Internet]. Jun;24(4):172–4. Available from: http://www.ncbi.
2017;10(1):44. Available from: http://www.ojoon- nlm.nih.gov/pubmed/2472858.
line.org/text.asp?2017/10/1/44/200695 141. Brodsky MC, Ford RE, Bradford JD. Subretinal neo-
129. Maumenee IH, Mitchell TN. Colobomatous malforma- vascular membrane in an infant with a retinochoroidal
tions of the eye. Trans Am Ophthalmol Soc [Internet]. coloboma. Arch Ophthalmol (Chicago, Ill 1960)
1990;88:123–32; discussion 133-5. Available from: [Internet]. 1991 Dec;109(12):1650–1. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/2095017. http://www.ncbi.nlm.nih.gov/pubmed/1726773.
130. Guirgis MF, Lueder GT. Choroidal neovascular 142. Steahly LP. Laser treatment of a subretinal neovascu-
membrane associated with optic nerve coloboma in a lar membrane associated with retinochoroidal colo-
patient with CHARGE association. Am J Ophthalmol boma. Retina [Internet]. 6(3):154–6. Available from:
[Internet]. 2003 Jun;135(6):919–20. Available from: http://www.ncbi.nlm.nih.gov/pubmed/2432636.
http://www.ncbi.nlm.nih.gov/pubmed/12788148. 143. Dailey JR, Cantore WA, Gardner TW. Peripapillary
131. Onwochei BC, Simon JW, Bateman JB, Couture choroidal neovascular membrane associated with an
KC, Mir E. Ocular colobomata. Surv Ophthalmol optic nerve coloboma. Arch Ophthalmol (Chicago,
[Internet]. 45(3):175–94. Available from: http:// Ill 1960) [Internet]. 1993 Apr;111(4):441–2.
www.ncbi.nlm.nih.gov/pubmed/11094243. Available from: http://www.ncbi.nlm.nih.gov/
132. Jesberg DO, Schepens CL. Retinal detachment pubmed/7682407.
associated with Coloboma of the choroid. Arch 144. Spitzer M, Grisanti S, Bartz-Schmidt KU, Gelisken
Ophthalmol [Internet] 1961 Feb 1;65(2):163–73. F. Choroidal neovascularization in retinochoroi-
Available from: http://archopht.jamanetwork.com/ dal coloboma: thermal laser treatment achieves
article.aspx?articleid=626425 long-term stabilization of visual acuity. Eye (Lond)
133. Hussain RM, Abbey AM, Shah AR, Drenser KA, [Internet]. 2006 Aug;20(8):969–72. Available from:
Trese MT, Capone A. Chorioretinal Coloboma com- http://www.ncbi.nlm.nih.gov/pubmed/16151482.
plications: retinal detachment and choroidal neovas- 145. von Eicken J, Höh H, Rehfeldt K. [Photodynamic
cular membrane. J Ophthalmic Vis Res [Internet]. therapy for choroidal neovascularisation due to
12(1):3–10. Available from: http://www.ncbi.nlm. choroidal coloboma in a 5 1/2-year-old child].
nih.gov/pubmed/28299000. Klin Monbl Augenheilkd [Internet]. 2007
134. Grewal DS, Tran-Viet D, Vajzovic L, Mruthyunjaya Feb;224(2):140–5. Available from: http://www.ncbi.
P, Toth CA. Association of pediatric choroidal nlm.nih.gov/pubmed/17309012.
neovascular membranes at the temporal edge of 146. Bhende M, Suganeswari G, Gopal L, Bhende PS,
optic nerve and retinochoroidal Coloboma. Am Gopal L, Rao C. Choroidal neovascularization asso-
J Ophthalmol [Internet]. 2017 Feb;174:104–12. ciated with coloboma of the choroid: a series of three
Available from: https://linkinghub.elsevier.com/ cases. Indian J Ophthalmol [Internet]. 59(2):148–51.
retrieve/pii/S0002939416305220 Available from: http://www.ncbi.nlm.nih.gov/
135. Leff SR, Britton WA, Brown GC, Lucier AC, Brown pubmed/21350286.
JF. Retinochoroidal coloboma associated with 147. Rajendran A, Gupta SR, Brahadeesh S, Ramasamy
subretinal neovascularization. Retina [Internet]. K. Intravitreal bevacizumab for choroidal neo-
5(3):154–6. Available from: http://www.ncbi.nlm. vascularization associated with a retinocho-
nih.gov/pubmed/2416022. roidal coloboma. Eye (Lond) [Internet]. 2010
136. Jeng-Miller KW, Cestari DM, Gaier ED. Congenital May;24(5):933–4. Available from: http://www.ncbi.
anomalies of the optic disc: insights from opti- nlm.nih.gov/pubmed/19713977.
cal coherence tomography imaging. Curr Opin 148. Naithani P, Vashisht N, Mandal S, Sankaran P, Garg
Ophthalmol [Internet]. 2017 Nov;28(6):579–86. S. Intravitreal bevacizumab in choroidal neovas-
Available from: http://www.ncbi.nlm.nih.gov/ cularization associated with congenital choroidal
pubmed/28817389. and optic nerve coloboma in children: long-term
137. Gupta V, Gupta A, Dogra MR. Subretinal neovascu- improvement in visual acuity. J AAPOS Off Publ
larization associated with retinochoroidal coloboma. Am Assoc Pediatr Ophthalmol Strabismus [Internet].
Indian J Ophthalmol [Internet]. 1997 Jun;45(2):116– 2010 Jun;14(3):288–90. Available from: http://www.
7. Available from: http://www.ncbi.nlm.nih.gov/ ncbi.nlm.nih.gov/pubmed/20603066.
pubmed/9475031. 149. Uemura A. Visual outcome after surgical removal
138. Rouland JF, Constantinides G. Retinochoroidal of choroidal neovascularization in pediatric
coloboma and subretinal neovascularization. Ann patients. Arch Ophthalmol [Internet]. 2000 Oct
Ophthalmol [Internet]. 1991 Feb;23(2):61–2. Available 1;118(10):1373. Available from: http://archopht.
from: http://www.ncbi.nlm.nih.gov/pubmed/2029115. jamanetwork.com/article.aspx?doi=10.1001/
139. Takenaka J, Yamane K, Minamoto A, Mishima archopht.118.10.1373
HK, Hayashida H. Subretinal neovascularization 150. Uhumwangho OM, Jalali S. Chorioretinal coloboma
associated with retinochoroidal coloboma. Eur J in a paediatric population. Eye (Lond) [Internet].
Ophthalmol [Internet]. 15(6):815–7. Available from: 2014 Jun;28(6):728–33. Available from: http://www.
http://www.ncbi.nlm.nih.gov/pubmed/16329073. ncbi.nlm.nih.gov/pubmed/24675580.
Khalil Ghasemi Falavarjani MD graduated in

Ophthalmology from Iran University of Medical Sciences,
Tehran, Iran. He was trained in Surgical and Medical
Pasha Anvari, MD MPH is an Assistant Professor of Fellowship in Retina in the same institution. He com-
Ophthalmology at the Rassoul Akram Hospital, Iran pleted a Clinical and Research Fellowship in Retina at the
University of Medical Sciences. He received his medical University of California, Los Angeles (UCLA). Currently,
degree and Master of Public Health from the Tehran he is the Chair of the Department of the Ophthalmology,
University of Medical Sciences and completed his oph- Iran University of Medical Sciences. He has been elected
thalmology residency at the Farabi Eye Hospital, Tehran as the Editor in Chief of the Journal of Current
University of Medical Sciences, and his retina fellowship Ophthalmology since 2015. In addition, he is a member of
at the Rassoul Akram Hospital, Iran University of Medical the Editorial Board of Middle East African Journal of
Sciences. Ophthalmology (MEAJO) and Journal of Ophthalmic and
Vision Research.
Masood Naseripour MD is a Professor of

Ophthalmology and head of Eye Research Center, and
Ocular Oncology Unit at the Rassoul Akram Hospital,
affiliated with the Iran University of Medical Sciences,
Tehran, Iran. He is also the Head of the Iranian Board of
Ophthalmology. He finished his Ophthalmology training
at the Tabriz University of Medical Sciences in 1993 and
obtained his fellowship in Vitreoretinal Surgery in 1995.
He completed his fellowship in Ocular Oncology at the
Wills Eye Hospital, Philadelphia, USA in 2001.
Part III
Clinical Trials
Clinical Trials Related to Choroidal
Neovascularization Secondary to 20
Age-Related Macular
Degeneration
Paisan Ruamviboonsuk, Peranut Chotcomwonse,

Variya Nganthavee, Warissara Pattanapongpaiboon,
and Kornwipa Hemarat
Acronyms and Titles of DENALI Verteporfin plus

Selected Clinical Trials of CNV-AMD ranibizumab for choroidal
neovascularization in age-
ABC Avastin (bevacizumab) for related macular degeneration
choroidal neovascular age- EXCITE Efficacy and safety of
related macular degeneration monthly versus quarterly
ANCHOR Anti-VEGF antibody for the ranibizumab treatment in
treatment of predominantly neovascular age-related mac-
classic choroidal neovascu- ular degeneration
larization in age-related mac- FOCUS RhuFab V2 ocular treatment
ular degeneration combining the use of
BRAMD Comparing the effectiveness Visudyne to evaluate safety
of bevacizumab to ranibi- GEFAL The Groupe d’Etude Français
zumab in patients with exu- Avastin versus Lucentis dans
dative age-related macular la DMLA néovasculaire
degeneration HARBOR The pHase III, double-masked,
CATT The comparison of age- multicenter, randomized,
related macular degeneration active treatment-controlled
treatments trials study of the efficacy and safety
DAWN Dorzolamide-timolol in com- of 0.5 mg and 2.0 mg
bination with anti-vascular Ranibizumab administered
endothelial growth factor monthly or on an as-needed
injections for wet age-related basis (PRN) in patients with
macular degeneration subfoveal neOvasculaR age-
related macular degeneration
P. Ruamviboonsuk (*) · P. Chotcomwongse HAWK and
V. Nganthavee · W. Pattanapongpaiboon HARRIER Phase 3, multicenter, ran-
Department of Ophthalmology, Rajavithi Hospital,
domized, double-masked tri-
Bangkok, Thailand
als of brolucizumab for
K. Hemarat
neovascular age-related mac-
Department of Ophthalmology, Faculty of Medicine,
Vajira Hospital, Navamindradhiraj University, ular degeneration
Bangkok, Thailand

260 P. Ruamviboonsuk et al.
HORIZON Open-label extension trial of RIVAL The comparison of

Ranibizumab for Choroidal Ranibizumab and Aflibercept
neovascularization secondary for the development of geo-
to age-related macular graphic atrophy in (wet)
degeneration AMD patients
IVAN The inhibition of VEGF in SALUTE A randomized trial to com-
age-related choroidal neovas- pare the safety and efficacy of
cularization trial two ranibizumab dosing regi-
LUCAS The Lucentis compared to mens in a Turkish cohort of
Avastin study patients with choroidal neo-
MANTA A randomised double- vascularization secondary to
masked trial comparing the AMD
visual outcome after treat- SUSTAIN Safety and efficacy of a flexi-
ment with ranibizumab or ble dosing regimen of
bevacizumab in patients with Ranibizumab in Neovascular
neovascular age-related mac- age-related macular
ular degeneration degeneration
MARINA Minimally classic/occult trial TAP Photodynamic therapy of
of the anti-VEGF antibody subfoveal choroidal neovas-
Ranibizumab in the treatment cularization in age-related
of Neovascular age-related macular degeneration with
macular degeneration verteporfin
MONT BLANC Verteporfin plus ranibizumab TREND TReat and extend study
for choroidal neovasculariza- TREX-AMD Treat-and-extend versus
tion in age- related macular monthly dosing for
degeneration Neovascular age-related
PIER A phase IIIb, multicenter, macular degeneration
randomized, double masked, VIEW The VEGF trap-eye:
sham injection controlled Investigation of efficacy and
study of the efficacy and safety in wet AMD
safety of Ranibizumab in VIP Verteporfin therapy of subfo-
subjects with Subfoveal veal choroidal neovascular-
Choroidal neovascularization ization in age-related macular
[CNV] with or without clas- degeneration
sic CNV secondary to age-
related macular degeneration
PrONTO Prospective optical coher-
ence tomography (OCT)
imaging of patients with The number of clinical trials for age-related mac-
Neovascular age-related ular degeneration (AMD) increases dramatically
macular degeneration (AMD) in the past decades. There were at least 200 pub-
treated with intraOcular lished papers on the subject in the 1990s, 700 in
Ranibizumab the 2000s, and 1000 in the 2010s until mid-2019.
PROSPERO Conbercept for patients with The results from many phase III clinical trials of
age-related macular degener- treatment for choroidal neovascularization
ation: a systematic review (CNV) secondary to AMD have shifted the treat-
REACH Evaluation of AGN-150998 ment paradigm in favor of anti-vascular endothe-
in exudative age-related mac- lial growth factor (anti-VEGF) as the gold
ular degeneration standard treatment [1]. The clinical trials for
20 Clinical Trials Related to Choroidal Neovascularization Secondary to Age-Related Macular Degeneration 261
neovascular AMD (nAMD) or CNV secondary to as a treatment for CNV-AMD. The trials for
AMD (CNV-AMD) were reviewed extensively in ranibizumab were therefore split into a trial for
both systematic [2–4] and non-systematic meth- predominantly classic CNV or ANCHOR [15]
ods [5, 6]. This chapter aims to explore into the using PDT as control against ranibizumab and
design, analysis, and interpretation [7, 8] of the another trial for minimally classic and occult
results of the clinical trials of CNV-AMD. with no classic CNV or MARINA [16] using
sham as control since PDT did not provide visual
benefits in this group of patients. Ranibizumab
20.1 Superiority Versus Non- was found to be superior to sham by 18 letters
inferiority Trials (p < 0.001) and to PDT by 20 letters (p < 0.001)
with 95% power in both trials. It was the first
Every clinical trial should start with a research time that we could document with strong evi-
question [9]. A well-designed clinical trial is dence from clinical trials that a modality for
conducted to answer that particular question treatment of CNV-AMD could improve vision.
with the primary outcome. If the question is Intravitreal injection of ranibizumab since then
whether an intervention is better than another, has become the gold standard for treatment of
the trial is usually designed as prospective ran- CNV-AMD [17]. Newer anti-VEGF agents or
domized controlled trial (RCT) in which an newer kinds of treatments are investigated against
intervention is designated as a comparator or ranibizumab for their efficacy including safety.
control, with the outcome is the difference The clinical trials designed to compare the
between two interventions [9]. An RCT should other anti-VEGFs, such as bevacizumab and
have enough power to make sure that the statis- aflibercept, with ranibizumab, however, were not
tical significance of the outcome is actual by aimed to prove that these two medications were
having enough samples [10]. superior to ranibizumab but aimed to prove that
In the era when laser photocoagulation was they were not inferior to ranibizumab for improv-
becoming the gold standard treatment of subfo- ing vision, i.e., CATT [18] and VIEW [19] stud-
veal CNV-AMD, since there was no other treat- ies. Proving that bevacizumab was not inferior to
ment or comparator, the laser treatment was ranibizumab in terms of efficacy and safety was
compared with observation in the Macular the aim of CATT because treatment by the former
Photocoagulation Study (MPS) [11]. Although was economical [20] but off-label. It would be
the laser treatment caused immediate vision loss justified for clinical use if non-inferiority in effi-
after treatment, the MPS concluded that the final cacy and safety between the two agents were
vision was significantly superior to observation found. Proving that aflibercept was not inferior to
(p = 0.002) in the long-term follow-up [8]. In the ranibizumab would also be justified because
era when the treatment was shifted to photody- treatment with aflibercept might be less burden.
namic therapy (PDT) with verteporfin, major Generally, when a published RCT did not have
clinical trials, such as TAP [12] and VIP [13], a statement clarifying that it was designed as non-
demonstrated that PDT was superior to sham inferiority trial, we might assume that it was supe-
treatment for preventing visual loss. These stud- riority trial. This is always true to RCT comparing
ies used sham instead of laser photocoagulation between a drug and a placebo, such as TAP, VIP,
as control. The results from TAP and VIP showed and MARINA. Superiority and non-inferiority tri-
that only predominantly classic CNV identified als are different in terms of design, data analysis,
by fluorescein angiography (FA) had benefit from and interpretation of results [21]. It is plausible to
PDT based on the secondary outcome that 59% switch from interpretation of inferiority outcome
of PDT-treated patients compared with 31% of to superiority [22], however, this should be done
placebo-treated patients lost fewer than 15 letters under certain circumstances and a predesign of
at the month 24 examination (p < 0.001) [14]. the trial allowing the switch should be made.
PDT was then used as control in clinical trials On statistical point of view, the null hypothe-
for assessing an anti-VEGF agent, ranibizumab, sis of non-inferiority trial would be easier to be
Favor Ranibizumab Favor Aflibercept
Ranibizumab 0.5 mg monthly-Aflibercept 2 mg

monthly
Ranibizumab 0.5 mg monthly-Aflibercept 0.5 mg

monthly
Ranibizumab 0.5 mg monthly-Aflibercept 2 mg q
8 weeks
+10% +5% 0 -5% -10%
Fig. 20.1 Primary endpoint of VIEW studies. The 95% CI for each arm was within +5% margin to −5%
Difference in proportion of patients with best-corrected margin. This chart shows non-inferiority of aflibercept
visual acuity (BCVA) losing <15 letters. The horizontal comparing to ranibizumab in all treatment arms.
lines indicated 95% confidence interval (95% CI) of each (Reproduced from the published VIEW studies)
treatment arm of the studies without actual value shown.
rejected than superiority trial [21, 23]. Its sample for CNV-AMD at 52 weeks, the primary end-
size is generally smaller. It is important to specify point was similar to ANCHOR: the proportion of
the difference margin of the efficacy between patients who lost fewer than 15 letters from base-
interventions that were investigated in both supe- line BCVA. The difference margin between these
riority and non-inferiority trials [21]. The margin two agents for measuring this endpoint was pre-
is used primarily for sample size estimation in a specified at 10% for assessing non-inferiority and
superiority trial whereas it is used for both sam- the U.S. Food and Drug Administration recom-
ple size estimation and interpretation of trial mended a 5% margin for assessing clinical equiv-
results in a non-inferiority trial [24]. alence. In addition, the investigators stated that
In ANCHOR [15], for example, a superiority the non-inferiority margin of 7% was also set
trial comparing ranibizumab and PDT for treat- according to the European Medicines Agency/
ment of CNV-AMD at 52 weeks, the margin of Committee for Medicinal Products for Human
difference of 17% was pre-specified in the sam- Use for analysis of the integrated data from
ple size estimation. The results from ANCHOR VIEW 1 and VIEW 2. The investigators were
showed that, with respect to the primary efficacy able to claim non-inferiority in all arms of afliber-
endpoint (the proportion of patients who lost cept in VIEW 1 and VIEW 2 because the 95%
fewer than 15 letters from baseline best-corrected Confidence Interval (95% CI) of the difference
visual acuity [BCVA]), 94.3% of patients in the between each of the six aflibercept arms and
0.3-mg group and 96.4% in the 0.5-mg group of ranibizumab for the primary endpoint had the
ranibizumab injection reached the primary end- lower margin of the CI above the pre-specified
point compared with 64.3% in PDT group difference margin of 10%. This occurred when
(p < 0.001 for each comparison). This was both the mean difference was also in favor of afliber-
statistical significance and clinical importance at cept. In addition, the 95% CI for each arm was
the time when this study was conducted. The within +5% margin to −5% margin for claiming
P-value was important to show that the s uperiority clinical equivalence between aflibercept and
of approximately 30% from ranibizumab had a ranibizumab (Fig. 20.1).
small chance to be false positive and 95% power The investigators found further in VIEW 1
made sure that this was actual. The pre-specified that only the group of aflibercept 2 mg every
margin of difference, 17%, was not as important 4 weeks showed statistically superior to the group
on the trial analysis and interpretation. of ranibizumab with a gain of mean of +10.9
In VIEW 1 and VIEW 2 [19], for example, letters from baseline compared to +8.1 letters. It
which were designed as non-inferiority trials was shown in the published table of this study
comparing between aflibercept and ranibizumab (reproduced as Fig. 20.2) that only 95% CI of the
Aflibercept 2 mg q 4 weeks
Aflibercept 0.5 mg q 4 weeks
Aflibercept 2 mg q 8 weeks
Ranibizumab is better. 0 Aflibercept is better.
Fig. 20.2 Secondary endpoint of VIEW 1. Mean best- ment arm. Only the arm of 2 mg aflibercept every 4 weeks
corrected visual acuity (BCVA) difference between each shows superiority over ranibizumab interpreted by the
of the three aflibercept treatment arms and ranibizumab at lower limit of the 95% CI greater than zero. No superior-
52 weeks. The horizontal lines indicate 95% confidence ity was shown in the other 2 arms. (Recreated from VIEW
interval (95% CI) of least square mean difference of studies, only values without graph shown in the published
BCVA between aflibercept and ranibizumab of each treat- manuscript of VIEW)
least square mean difference between aflibercept demonstrating that the 95% CI of the difference
and ranibizumab in this arm of the VIEW 1 had between the two agents fell within the pre-
the lower margin above zero. This is the interpre- specified equivalence margin, both sides.
tation of a non-inferiority trial when the claim for
superiority is to be made [22]. However, this
mean difference of BCVA in letters was not the 20.2 Assessment of Primary
primary endpoint in these studies. Outcomes in Clinical Trials
For superiority trial, analysis of data from all of CNV-AMD
patients in each randomized arm, i.e., the full
analysis set, is recommended [23]. This is 20.2.1 Clinical Trials on Different
because it is more conservative to claim superior- Agents
ity when those with missing data, drop out, or
nonresponsive to treatment after randomization 20.2.1.1 Comparison Between
are included for analysis. For non-inferiority Bevacizumab
trial, however, an even more conservative and Ranibizumab: CATT
approach may be recommended. To decrease and IVAN
bias, both analyses on full dataset or “intention- IVAN was another major non-inferiority RCT
to-
treat” analysis, and “per-protocol” analysis, comparing between bevacizumab and ranibi-
which is an analysis of data from patients who zumab conducted in the United Kingdom whereas
only completed the treatment at the end exclud- CATT was in the United States. The other smaller
ing those with missing, dropout, nonresponsive, RCTs, included GEFAL, MANTA, BRAMD,
are recommended [25, 26]. The LUCAS Study and LUCAS, were conducted in other countries
[27] that was a non-inferiority randomized con- with a similar aim to show that bevacizumab was
trolled clinical trial for ranibizumab vs. bevaci- non-inferior to ranibizumab for visual improve-
zumab in treat and extended dosing regimen is a ment in CNV-AMD.
good example. This study showed a clear figure
of study groups with intention-to-treat (ITT) and One-Year Results of CATT
per-protocol (PP) analyses in each arm. Both ITT There were 1185 patients from 44 clinical centers
and PP analyses in LUCAS shown non-inferiority randomized into 4 groups defined by drugs,
between the two agents for achieving the primary ranibizumab or bevacizumab, and dosing regi-
outcome. In addition, the investigators claimed mens, monthly or as-needed, without initial load-
equivalence efficacy in achieving the endpoint by ing injections [18]. This study was pre-designed
with the ability to test for superiority if a treat- the two drugs at the end of the second year. There
ment was found to be non-inferior. The non- was non-inferiority for a comparison between
inferiority margin for the difference between PRN and monthly when both medications were
study groups in the mean change of BCVA at pooled together within each dosing regimen. The
1 year was 5 letters. results were in favor of monthly regimen and
At one year, among the six comparisons in the non-inferiority could be claimed with the mean
trial, only the comparisons of visual gain between change of −2.4 letters, 95% CI between −4.8 and
bevacizumab PRN and bevacizumab monthly −0.1 (p = 0.046) at the end of the second year.
and between bevacizumab PRN and ranibizumab The results were still similar after applying alter-
monthly could not reach the non-inferiority mar- native methods for handling missing visual acu-
gin. The lower margins of the 95% CIs of both ity data at 2 years (ITT analysis).
comparisons fell below the pre-specified −5.0
letters. The favor of ranibizumab over bevaci- Two-Year Results of IVAN
zumab was found in each arm of the compari- IVAN randomized 628 patients into four treat-
sons, except that bevacizumab monthly was ment groups similar to CATT year 1 but 3
favorable to ranibizumab PRN. BCVA gain from monthly loading injections were applied in the
baseline was 8.5 in the ranibizumab-monthly PRN regimens [29]. The investigators of IVAN
group, 8.0 in the bevacizumab-monthly group, used more conservative approach than CATT for
6.8 in the ranibizumab-as-needed group, and designing this trial. They specified a non-
5.9 in the bevacizumab as-needed group, respec- inferiority margin of 3.5 letters, instead of 5 let-
tively. This BCVA gain between 5 and 10 letters ters; the retreatment regimen of IVAN required 3
found in CATT is a benchmark for BCVA gain monthly injections instead of one injection as
from treatment of CNV-AMD with a new anti- CATT.
VEGF agent or a new kind of treatment in later At 2 years, 84% of patients from the first year
clinical trials. were still in the study. A total of 271 eyes received
ranibizumab and 254 eyes received bevacizumab
Two-Year Results of CATT whereas 261 received monthly and 264 PRN reg-
A total of 93.4% of participants from year 1 con- imens. The analysis on these 2-year data showed
tinued in CATT year 2 [28]. During the second inconclusive evidence that the mean change in
year, approximately half of patients in monthly BCVA from baseline between both drugs (beva-
groups were randomized to switch to PRN regi- cizumab vs. ranibizumab difference, −1.37 let-
men. This created 6 groups of treatments, instead ters; p = 0.26, 95% CI −3.75 to 1.01) and the
of four groups as in the first year. This re- mean change between both treating regimens
randomization reduced the sample size of groups (monthly vs. PRN difference, −1.63 letters;
originally treated monthly but yielded larger sam- p = 0.18, 95% CI −4.01 to 0.75) could reach non-
ple sizes, greater precision, and increased power inferiority. The lower margin of both 95% CIs
to assess the actual effects of PRN treatment. fell below −3.5 letters prespecified margin.
This approach provided an accurate description Although it seemed that the non-inferiority in
of the results when there was no interaction IVAN could have been reached if the non-
between drug and dosing regimen [28]. inferiority margin of 5 letters was pre-specified
On analysis, the investigators pooled data as in CATT; however, the sample size would be
from all arms to compare the differences between re-estimated and the results would be uncertain.
the two drugs: ranibizumab and bevacizumab and
between the two treatment regimens: monthly 20.2.1.2 Comparisons Between
and PRN. At the end of year 2, bevacizumab was Bevacizumab
non-inferior to ranibizumab for visual gain, 95% and Ranibizumab Other than
CI was −3.7–0.8, the mean BCVA change was CATT and IVAN
−1.4 letters in favor of ranibizumab (p = 0.21). In MANTA [30], the RCT in Austria, patients in
This held true as an overview assessment between each arm of ranibizumab and bevacizumab
received three initial monthly injections followed improvement (bevacizumab 7.9, ranibizumab
by monthly evaluation of BCVA and lesion activ- 8.2, p = 0.845, 95% CI −2.4 to 2.9 according to
ity for re-treatment (PRN regimen). At month 12, per-protocol analysis; bevacizumab 7.8, ranibi-
the mean gain of BCVA of bevacizumab and zumab 8.0, p = 0.550, 95% CI −2.2 to 2.5 accord-
ranibizumab group was 4.9 and 4.1 letters, ing to intention-to-treat analysis). The
respectively (p = 0.78) from intention-to-treat non-inferiority could also be claimed.
analysis. The investigators interpreted that this
mean change in BCVA of bevacizumab was 20.2.1.3 Systematic Review
equivalent to ranibizumab because the 95% CI of Comparisons Between
was within −7 to +7 letters (the pre-specified Bevacizumab
margin of 7 letters of non-inferiority) without and Ranibizumab
showing the actual value of the 95% CI in the In a systematic review from Cochrane Group that
article. The 7 letters margin may be too wide to included 16 randomized controlled trials that
claim non-inferiority in real clinical practice enrolled 6347 patients with CNV-AMD (the
since this is a change of more than a line on number of patients per trial ranged from 23 to
ETDRS visual acuity chart. 1208) for assessment of the efficacy and safety of
GEFAL [31], the RCT in France, had similar three anti-VEGF agents, ranibizumab, bevaci-
study design as MANTA comparing ranibizumab zumab, and pegatanib [2]. The overall certainty
and bevacizumab in PRN regimen. The margin of of the evidence among the included trials was
clinical non-inferiority was pre-specified at 5 let- moderate to high, and most trials had an overall
ters. At 1 year, bevacizumab was non-inferior to low risk of bias. There were 10 trials compared
ranibizumab (bevacizumab vs. ranibizumab between bevacizumab and ranibizumab.
+2.36 letters; 95% CI −0.72 to 5.44, p < 0.0001, According to this review, at 1 year, patients
from intention-to-treat analysis) and the mean who received intravitreal injections of any of the
number of injections was not significantly differ- three anti-VEGF agents had a significantly higher
ent between the two drugs (bevacizumab 6.8, chance of gaining 15 letters or more (risk ratio
ranibizumab 6.5, p = 0.39). The main difference [RR] 4.19, 95% CI 2.32 to 7.55; the evidence was
between GEFAL and other well-designed non- moderate-certainty), lost fewer than 15 letters
inferiority trials like IVAN and CATT might lay (RR 1.40, 95% CI 1.27 to 1.55; the evidence was
on the aspect of conducting the trial. The investi- high-certainty), and showed better mean improve-
gators stated that the effective double-masking ment in visual acuity (mean difference [MD] 6.7
was remained throughout the entire study of letters, 95% CI 4.4 to 9.0 in one pegaptanib trial;
GEFAL, unlike CATT and IVAN in which the MD 17.8 letters, 95% CI 16.0 to 19.7 in three
masking differed between centers. ranibizumab trials; the evidence was moderate-
BRAMD [32] was the RCT to compare certainty), compared to controls. Patients treated
between monthly regimen of both ranibizumab with anti-VEGF agents also showed improve-
and bevacizumab in the Netherlands. The non- ment in central retinal thickness (CRT) and size
inferiority margin was pre-specified at 4 letters. of CNV and other morphologic outcomes (the
AT 1 year, the mean gain in BCVA was 5.1 letters evidence was moderate-certainty).
in the bevacizumab group and 6.4 letters in the Also in this review, at 1 year when bevaci-
ranibizumab group (p = 0.37) in intention-to- zumab was compared with ranibizumab, visual
treat analysis. The lower limit of 95% CI of the acuity outcomes were similar in terms of gaining
difference in BCVA gain between the two groups 15 letters or more (RR 0.95, 95% CI 0.81 to 1.12)
was 3.72 which allowed the investigators to claim and loss of fewer than 15 letters (RR 1.00, 95%
non-inferiority. CI 0.98 to 1.02). The mean improvement in visual
LUCAS [27] was the RCT to compare bevaci- acuity in those received ranibizumab were from 3
zumab and ranibizumab in the Treat-and-Extend to 8 letters whereas those with bevacizumab
(T&E) regimen in Norway. At 1 year, there was improved 0.5 letters or fewer in average (95% CI
no difference between both drugs in mean BCVA −1.5 to 0.5). The mean reduction in CRT in those
received ranibizumab were from 30 to 182 μm NV-AMD with ziv-aflibercept [37]. There were
C
whereas those with bevacizumab were 13.97 μm two clinical studies reported on relatively long-
in average (95% CI −26.52 to −1.41). All of term outcomes. One had a follow-up of 52 weeks
these evidence were in high certainty [2]. [38] and another of 30 months [39].
In the 52-week study, ziv-aflibercept was
20.2.1.4 Comparisons Between given initially for three monthly injections of
Aflibercept 1.25 mg/0.05 ml (regular dose of aflibercept is
and Ranibizumab 2 mg/0.05 ml), followed by PRN, BCVA was
The first RCT that compared aflibercept with improved from 0.95 ± 0.41 (20/200) at baseline
ranibizumab was VIEW study, in which two par- to 0.75 ± 0.51 (20/125) logMAR units
allel phase III studies were conducted in different (p = 0.0066); the central retinal thickness
countries as VIEW 1 and VIEW 2. A total of decreased from 478.21 ± 153.48 mm at baseline
1217 patients were included in VIEW 1 and 1240 to 304.43 ± 98.59 mm (p = 0.0004). Full-field
patients were included in VIEW2. Patients electroretinography parameters were used to
enrolled were classified into four groups: (1) assess retinal toxicity after the injections (rod
0.5 mg aflibercept monthly injection for 3 doses response and oscillatory potentials). It was found
then every 4 weeks (A0.5q4); (2) 2 mg afliber- that both parameters remained unchanged during
cept monthly injection for 3 doses then every the follow-up. The average multifocal electroreti-
4 weeks (A2q4); (3) 2 mg aflibercept monthly nography macular response in 5° showed
injection for 3 doses then every 8 weeks (A2q8); increased N1-P1 amplitude and decreased P1
and (4) 0.5 mg ranibizumab every 4 weeks implicit time from baseline (p < 0.05). The inves-
(R0.5q4). The results of the primary endpoint tigators discussed that these significant changes
were stated previously [19]. could be from disease processes rather than from
the drug.
20.2.1.5 Clinical Trials In the 30-month study, the investigators pro-
of Ziv-Aflibercept spectively treated consecutive cases of macular
Ziv-aflibercept is a drug containing a fusion pro- diseases with ziv-aflibercept using the Treat and
tein, aflibercept, and approved by U.S. Food and Extend (T&E) regimen. Of a total of 53 eyes in
Drug Administration as a systemic chemothera- the study, 35 were CNV-AMD. Central macular
peutic agent for treatment of colorectal cancer thickness of these patients decreased by 107.8 μm
[33]. For treatment of CNV-AMD, ziv-aflibercept (p = 0.012) with BCVA gain of 0.52 (p = 0.001).
was used off-label similar to bevacizumab. Ziv- Initiation of T&E in CNV-AMD occurred after a
aflibercept, however, could potentially be more mean of 8.0 injections (range 4–16) at 9.9 months
harmful than bevacizumab due to the higher (range 5–18 months). No eye with CNV-AMD
osmolality of the drug, approximately had initiation of T&E after three injections. One
1000 mOsm/kg, compared with the vitreous. On eye in this case series developed a single episode
the other hand, aflibercept itself has undergone a of transient iritis, which recovered well to topical
different purification process and contained dif- treatment. No systemic complications were
ferent buffer solutions resulting in a lower osmo- noted. The high visual gain from T&E of ziv-
lality of 286 mOsm/kg [34]. Despite this safety aflibercept in this study, compared with T&E of
concern of ziv-aflibercept, in vitro studies showed aflibercept in another study [40], might come
safety of ziv-aflibercept for cellular viability of from selection bias of recruiting patients with
cultured retinal pigment epithelium cells (ARPE- poorer vision at baseline.
19 [35]) as well as the rabbit retina with intravit- There has not yet been a published RCT of
real injection [36]. Many clinical studies of ziv-aflibercept for CNV-AMD. The only publi-
intravitreal injection of ziv-aflibercept showed no cation of RCT of ziv-aflibercept to date was a
particular safety concern. This included a comparison between ziv-aflibercept and beva-
26-week case series on visual outcome and elec- cizumab for diabetic macular edema in T&E
troretinographic findings of treatment of regimen [41].
20.2.1.6 Clinical Trials of Conbercept improve BCVA slightly compared to triamcino-

Conbercept (KH902) is another anti-VEGF lone acetonide (mean difference = 0.11, 95% CI
agent, which was designed as a recombinant 0.08–0.15), and reduce CRT compared to the
fusion protein with high affinity to all VEGF iso- other four therapies (conservative treatment,
forms and placental growth factor (PIGF) similar ranibizumab, transpupillary thermotherapy, and
to aflibercept, developed in China. Currently, it is triamcinolone acetonide).
still approved only in China. Almost all trials of
conbercept were also conducted in China and 20.2.1.7 Clinical Trials
most of them were written in Chinese. of Brolucizumab
The only prospective RCT on conbercept pub- Brolucizumab is a new small (26 kDa), human-
lished in English is the PHOENIX study [42] ized, single-chain antibody fragment, which is
which was phase III trial. Patients with CNV- another potent anti-VEGF. Preclinical data dem-
AMD and PCV were included, but the definite onstrated a 2.2- and 1.7-fold higher exposure in
diagnosis of PCV was not clear in some cases the retina and in the retinal pigment epithelium
due to the unavailability of indocyanine green and choroid, respectively, compared with ranibi-
angiography in some periods of the study. A total zumab [44].
of 124 patients were randomized into conbercept HAWK and HARRIER studies were conducted
or sham. Conbercept 0.5 mg was injected monthly to assess the efficacy and safety between broluci-
with three loading doses then fixed at every zumab and aflibercept at the dosing regimens of
3 months until month 12. Sham injection was three-monthly injections followed by every
performed monthly for the first 3 months fol- 12 weeks (q12w) for the former and every 8 weeks
lowed by conbercept 0.5 mg monthly injection (q8w) for the latter. Any patients in brolucizumab
for three doses, then every 3 months until month arm who met the criteria defining disease activity
12. This study design was not defined as superi- at the assessment visits, assessed by masked inves-
ority or non-inferiority. The primary endpoint tigators, would be changed dosing to q8w. The
was mean BCVA change at 3 months. selection of q12w and q8w dosing was based on
At 3 months, BCVA improvement in conber- the results of the OSPREY study [45], in which a
cept group was statistically better than sham. This head-to-head comparison of the q8w treatment
was +9.2 letters compared to +2.02 (mean differ- regimen between brolucizumab 6 mg and afliber-
ence + 7.27 letters, 95% CI 3.36–11.18; p < 0.001). cept 2 mg showed anatomical advantages with
At 12 months, the mean changes from baseline brolucizumab while reaching non-inferiority in
BCVA were + 9.98 letters in the conbercept group BCVA. In the same study, brolucizumab-treated
and + 8.81 letters in the sham group (95% CI patients were subsequently challenged with q12w
4.47–13.15; p = 0.64). The statistical significance dosing interval, with an outcome suggesting that
was only shown when the data were analyzed with half of the patients were adequately treated.
per-protocol analysis (mean difference + 7.27 let- Non-inferiority of mean BCVA change from
ters: 95% CI 3.27–11.26; p < 0.001), not intention- baseline at week 48 was set as a primary objec-
to-treat analysis (mean difference + 1.24 letters: tive in both HAWK (6 mg and 3 mg of broluci-
95% CI −4.01 to 6.50; p < 0.64). zumab) and HARRIER (6 mg brolucizumab). At
There was a systematic review on conbercept Week 48, each brolucizumab arm demonstrated
entitled “PROSPERO” [43]. Of the 18 RCTs non-inferiority to aflibercept in BCVA change
selected from 780 trials of conbercept, there were from baseline as +6.6 letters, +6.1 letters, and
three trials with the longest follow up of + 6.8 letters in brolucizumab 6 mg group, brolu-
12 months. One compared with ranibizumab cizumab 3 mg group, and aflibercept group,
whereas the other two compared with triamcino- respectively, with 95% CI of −2.5 to 1.3 and
lone. The primary outcome of these trials was −2.7 to 1.3 in HAWK study; as +6.9 letters and
central retinal thickness (CRT), not BCVA. Nine + 7.6 letters in brolucizumab 6 mg and afliber-
trials had a follow-up of three months or less. cept, respectively, with 95% CI of −2.1 to 1.2 in
According to this review, conbercept might HARRIER. Greater than 50% of brolucizumab
6 mg-treated eyes were maintained on q12w dos- SUSTAIN [47] was also a case series to inves-
ing through Week 48 (56% in HAWK and 51% in tigate ranibizumab given in PRN regimen similar
HARRIER). to PrONTO but with much larger population: 513
treatment-naive patients. The re-treatment criteria
were focused on both visual acuity and OCT
20.2.2 Clinical Trials on Different change as equally important whereas PrONTO
Dosing Regimens of Anti- focused mainly on OCT findings. All included
VEGF Agents patients received three initial monthly injections
of ranibizumab (0.3 mg) then followed up monthly
20.2.2.1 ase Series on PRN
C for one year. The pre-specified re-treatment crite-
of Ranibizumab ria during the follow-up were BCVA better than
PrONTO [46] was a small study without control 79 letters or the CRT was lower than 225 μm
arm to evaluate the potential benefit of using find- together with either one or two of (1) a decrease
ings from Optical Coherence Tomography (OCT) of BCVA by more than 5 letters or (2) an increase
to adjust treatment regimen of ranibizumab. It in CRT by more than 100 μm compared with best
was one of the first studies assessing ranibizumab value at any prior visit.
in PRN treatment. A total of 40 patients were The best mean BCVA was achieved at month
enrolled to receive three monthly loadings of 3 at +5.8 letters. The mean final BCVA was +3.6
0.5 mg ranibizumab then followed monthly for a at month 12. The mean number of injections was
year. At each monthly visit, the patients would 5.6 which was unexpectedly a similar number as
have re-treatment if they had the following crite- PrONTO at one year. However, the primary out-
ria: (1) persistent subretinal fluid after a 1-month come of SUSTAIN was assessment of adverse
period post injection, (2) increase in CRT on events, not visual efficacy.
OCT of at least 100 μm, (3) new macular hemor-
rhage or new classic CNV, and (4) loss of 5 letters 20.2.2.2 ixed Monthly Versus PRN
F
on BCVA with OCT evidence of fluid in the mac- of Ranibizumab Other than
ula. The investigators did not have the option CATT
“not to treat” in case of good visual acuity, a flat HARBOR [48] was conducted to evaluate ranibi-
retina, or futility of treatment. zumab in the higher than usual dose, 2.0 mg, in
At month 12, the mean visual acuity improved superiority comparison with the usual dose,
by 9.3 letters compared with baseline (p < 0.001) 0.5 mg. This study also evaluated whether PRN
and the mean CRT on OCT decreased by 178 μm dosing were non-inferiority to monthly dosing.
compared with baseline (p < 0.001). The aver- The HARBOR was different from CATT for the
age number of injections was 5.6 over non-inferiority comparison in the way that
12 months. The mean injection-free interval was HARBOR have initial three loading doses in the
4.5 months. PRN arm; the non-inferiority margin, 4 letters,
During the second year, the re-treatment crite- was smaller than CATT. All four treatment arms
ria were amended to include re-treatment if any in HARBOR demonstrated clinically significant
qualitative increase in the amount of macular BCVA improvement at months 12 and 24, but the
fluid was detected using OCT. At month 24, the 2.0 mg monthly dose failed to show superiority to
mean visual acuity improved by 11.1 letters the 0.5 mg monthly dose (model-adjusted mean
(p < 0.001) and the CRT decreased by 212 μm difference, −1.1 letters; 95.1% CI, −3.4 to 1.3;
(p < 0.001). Visual acuity improved by 15 letters p = 0.8145). The mean change in BCVA at month
or more in 43% of patients. These vision and 24 were + 9.1, +7.9, +8.0, and + 7.6 for the
OCT outcomes were achieved with an average of 0.5 mg monthly, 0.5 mg PRN, 2 mg monthly and
9.9 injections over 24 months. The potential of 2 mg PRN, respectively. The non-inferiority
using OCT-based PRN regimen for treatment of comparison between 0.5 mg PRN and 0.5 mg
CNV-AMD found in PrONTO was carried over monthly had a model-adjusted mean difference
to be investigated in many studies. of −2.0 letters (97.5% CI, −4.5 to 0.6) and incon-
clusive results. The comparison between 2.0 mg If an eye showed recurrent disease activity three
PRN and 0.5 mg monthly had a model-adjusted times at any given intervals, treatment was contin-
mean difference of −1.6 letters (98.4% CI, −4.4 ued at the next shorter interval for three consecutive
to 1.1) and also showed inconclusive results for visits.
non-inferiority. TREX-AMD found that the mean maximum
extension interval in the T&E cohort was
20.2.2.3 Monthly Versus “Treat & 8.4 weeks; 22% of patients still required monthly
Extend” (T&E) Regimen dosing and visit. At month 12, it was found that
of Ranibizumab 37% required treatment at 9-week intervals or
The T&E regimen was introduced to decrease the longer. These proportions are coincidentally sim-
treatment burden from PRN regimen by decreas- ilar to the respective 19 and 34% reported in
ing patient visits and providing the potential to another study [50], which was from patients
treat patients proactively before disease activity treated with T&E regimen of ranibizumab in the
recurred. real world. TREX-AMD also showed compara-
TREX-AMD [49] was a relatively small pro- ble BCVA change from baseline between the two
spective, multicenter, randomized, controlled treatment groups, +9.2 letters in monthly and
clinical trial with 60 naive patients randomized + 10.5 letters in T&E group (p = 0.60) and sig-
to ranibizumab monthly or T&E in the propor- nificantly lower number of injections in the T&E
tion of 1:2. The T&E protocol in this study had a group (13 vs. 10.1, p < 0.001). However, there
requirement that the injections were no more was still a limitation in the study design of TREX-
frequently than every 4 weeks and no less fre- AMD in a way that we may not be able to strongly
quently than every 12 weeks which was similar confirm non-inferiority of the T&E to monthly
to other studies on T&E. However, the protocol protocol.
for extension of visiting intervals in this study A larger and recently published trial assessing
was relatively more conservative than the ranibizumab in T&E regimen was TREND study
others. [51]. TREND was designed as non-inferiority
At each visit, the enrolled eyes were treated trial comparing 323 eyes with T&E and 327 eyes
monthly until a dry macula was achieved. The with monthly ranibizumab. The T&E arm was
dry macula was defined as either resolution of treated with two initial monthly doses, not three
subretinal and intraretinal hemorrhage related to as in TREX-AMD or other studies. The primary
CNV or resolution of intraretinal and subretinal outcome was mean change of BCVA from base-
fluid on Spectral Domain-OCT. The dry macula line until one year and the non-inferiority margin
also included cases with a pigment epithelial was pre-specified at 5 letters.
detachment without intraretinal or subretinal The T&E regimen in TREND was slightly dif-
fluid. ferent from TREX-AMD in the way that the
When the dry macula was found, the interval treatment interval between visits, no matter
between treatment visits was extended by the extended or reduced interval, would be 2 weeks.
increments of 2 weeks. If recurrent disease activ- There was no 1-week interval as in TREX-
ity was found, the interval was reduced by the AMD. The regimen in TREND is commonly
increments of 2 weeks until the dry macula was used in real-world practice. T&E in TREND was
achieved. At this point, the interval then was shown to be non-inferior to monthly regimen.
extended by the increments of only 1 week until The least-square mean difference between the
the disease was recurred. If this recurrence was two treatments was −1.9 letters (95% CI −3.83 to
again found, the interval then was reduced by the 0.07; p < 0.001 for non-inferiority). The mean
increments of 1 week until the macular was dry. change in BCVA was + 7.6 and + 6.6 letters and
At this point, the most recent interval between mean change in CST were −169.2 μm and
treatment visits was maintained for an additional −173.3 μm for the monthly and T&E, respec-
visit; if the macula remained dry, the interval was tively. The average duration between the visits
extended by the increments of 1 week once again. was 40.1 days in T&E and 28.5 days in monthly
group. In the T&E group, 51.7% of the patients into three groups in the following two more
received 6–9 injections. years: (1) patients treated with ranibizumab
Despite the differences in T&E protocols, (n = 600), (2) patients randomized to control who
TREX-AMD and TREND had a comparable pro- crossed over to receive ranibizumab (n = 253),
portion of patients who required monthly dosing and (3) ranibizumab-naïve patients (n = 253).
(22.6% in TREND and 22% in TREX-AMD) and Ranibizumab was administered at the investiga-
proportion of patients who required treatment at tor’s discretion without pre-specified re-treatment
9-week intervals or longer at month 12 (41.8% in criteria in this extension period. Systemic and
TREND and 37% in TREX-AMD). ocular side effects, which were primary outcomes
in HORIZON, were uncommon and consistent
with other previous studies (one mild endophthal-
20.3 Interpretation of Secondary mitis in 3552 injections in all ranibizumab-
Outcome, Ad Hoc, and Post treated patients).
Hoc Analyses This study had limitations as post hoc analysis
and the assessment of safety might suffer from
Only on some occasions when clinical trials are inadequate sample size. The investigators also
powered to cover secondary outcomes or vari- found that the mean change in BCVA from the
ables other than the primary outcome [52]. It may initial study baseline (the secondary outcome)
not be justified to increase the sample size large was only +2.0 letters in the ranibizumab-treated
enough to cover other outcomes since more initial group vs. −11.8 in the other two groups.
resources and time are needed. Ad hoc or post These results might be difficult to interpret since
hoc analysis (by definition, the former is usually many confounders might not be controlled.
conducted additionally when the trial has not yet
been completed whereas the latter is conducted
after the trial has been completed) [53], on the 20.3.2 Additional Analyses
other hand, usually conducts on data already from CATT and IVAN
available in well-designed, prospective, RCTs to
determine variables that are not accounted for Analysis of pooled data [29] between CATT and
during the design of a clinical trial. These analy- IVAN could be considered as another post hoc
ses can be conducted because of the availability analysis since the data from both trials, albeit
of plenty of data in the trials. The data may be were from relatively similar inclusion and exclu-
used to explore some additional research ques- sion criteria, should not be combined together as
tions. The results of ad hoc, and post hoc analy- a group for analysis. These pooled data, however,
ses, including the results of the secondary might demonstrate some trends. Pooled bevaci-
outcomes, should be interpreted carefully [52]. zumab (n = 629) was shown to be non-inferior to
They may better be used for pointing out the pooled ranibizumab (n = 666) with the mean dif-
direction for future research. ference of BCVA of −1.15 letters (95% CI –2.82
to 0.51). In addition, the pooled PRN regimen
(n = 775) was found to be inferior to the pooled
20.3.1 P
ost Hoc Study from ANCHOR monthly regimen (n = 522) with the mean differ-
and MARINA ence of −2.23 letters (95% CI –3.93 to −0.53).
Based on data from CATT alone, there are at
Following the ANCHOR and MARINA, an least 40 published papers on additional post hoc
extension study called HORIZON [54] was con- analyses. Among these are analyses of baseline
ducted to further evaluate the safety of ranibi- characteristics, newly defined associated fea-
zumab in patients who had already completed tures, such as subretinal hyperreflective material
one of the three RCTs, ANCHOR, MARINA, or (SHRM) [56] and hyperreflective foci on OCT
FOCUS [55] for 2 years. They were classified [57], angiographic cystoid macular edema
(CME) [58], retinal angiomatous proliferation to investigate this relationship which has not yet
(RAP) [56], or genetic risk factors [59]. been concluded with strong evidence.
20.3.2.1 Development of Geographic 20.3.2.2 Macular Morphology

Atrophy (GA) in CATT and Visual Acuity in CATT
and IVAN In another study on macular morphology and
Among the 1011 participants who did not have visual acuity in CATT which were published
GA at baseline and had follow-up images grad- sequentially based on the data in year one, two,
able for GA in CATT, the cumulative incidence and five, it was found that associations between
was 12% at 1 year, 17% at 2 years, and 38% at VA and morphologic features previously identi-
5 years. Older age, hypercholesterolemia, fied in year 1 were maintained or strengthened at
worse initial visual acuity, larger CNV area, year 5 [63]. New foveal scar, CNV, intraretinal
retinal angiomatous proliferation (RAP) lesion, fluid, SHRM and retinal thinning, development
GA in the fellow eye, and intraretinal fluid or worsening of foveal GA, and increased lesion
were associated with a higher risk of incidence size are important contributors to VA decline
of GA. GA in the fellow eye, subretinal hemor- from years 2 to 5.
rhage, and absence of subretinal pigment epi-
thelium fluid at baseline were associated with a 20.3.2.3 Five-Year CATT
higher growth rate. Thicker subretinal tissue Another pivotal post hoc analysis from CATT
complex and presence of subretinal fluid were was a report on 5-year results [64]. This report
associated with less GA development. Eyes included 647 of 914 (70.8%) living patients with
treated with ranibizumab in the first 2 years an average follow-up of 5.5 years (the interval of
had a higher growth rate than bevacizumab 4.3–85 months) who were still in the trial until
(adjusted growth rate, 0.38 vs. 0.28 mm/year; year 5. However, if we include patients who did
p = 0.009) [60]. not participate for 5-year follow-up (n = 203) and
Both the use of ranibizumab and bevacizumab patients who died after the CATT trials ended
monthly treatment was associated with an (n = 203), 57.9% of patients from the original
increased rate of development of GA. At the end trial participated and most of them (91.3%) con-
of year 2, eyes treated with ranibizumab had a tinued to have eye care at a CATT center after
higher incidence of GA (21% vs. 17%; p = 0.02) release from the trial.
[61]. on contrary, post hoc of IVAN showed the Despite the limitation of high patient dropout,
incidence was similar to both agents (28% vs. which is common in long-term research, the
31%; p = 0.46) [29]. This seemed to decrease the investigators could report the overall results of
likelihood of ranibizumab on development of GA both ranibizumab and bevacizumab together,
more often than bevacizumab. The association of without comparison between drugs and dosing
monthly treatment with an increased rate of regimens. It was found that BCVA of the follow-
development of GA might be more consistent. At up patients decreased 3 letters from baseline and
the end of year 2 of CATT and IVAN, eyes that 11 letters from year 2. On the other hand, 50% of
received monthly treatment were more likely to these patients had VA 20/40 or better, whereas
demonstrate GA than those receiving PRN (24% only 20% had VA of 20/200 or worse and almost
vs. 15%; p = 0.003 in CATT and 34% vs. 26%; 10% had VA of 20/20. The investigators found
p = 0.03 in IVAN). In HARBOR, eyes received further that these patients were 2 years younger,
monthly ranibizumab also showed a higher inci- had VA that was 3 letters better at baseline, and
dence of GA compared with PRN (hazard ratio, had VA that was 5 letters better at 2 years com-
1.3; 95% CI 1.0–1.7) [62]. The association pared with patients who did not return. The
between GA and the number of anti-VEGF decrease vision was found to be accompanied
injections raised initially by CATT and IVAN by expansion of the size of the total
post hoc analyses prompted many investigators neovascular complex comprising neovasculariza-
tion, scarring, and atrophy and by persistence of persistent fluid was approximately 20% in each
fluid on OCT. arm. In these eyes, mean BCVA gain from base-
line to week 52 was greater in A2q4 compared
with Rq4 (p < 0.01) and A2q8 (p < 0.05), whereas
20.3.3 Additional Analyses the statistical difference was not found between
from VIEW 1 and VIEW 2 Rq4 and A2q8 (p = 0.294). In eyes without per-
Studies sistent fluid, BCVA changes were similar across
treatment groups.
20.3.3.1 econdary Outcome of Year
S
1 VIEW Studies: NEI VFQ-25 20.3.3.3 isual Outcome of Year 2
V
The assessment of vision-related quality of life VIEW Studies
using the 25-item National Eye Institute Visual During year 2, the dosing intervals in all treat-
Function Questionnaire (NEI VFQ-25) is comment arms were changed to a common regimen
mon in clinical trials of CNV-AMD. This assess- of modified quarterly dosing with respect to their
ment is a descriptive study and interpretation of originally randomized dose and drug (all patients
scores from NEI VFQ-25 is straightforward. were monitored monthly and received a mini-
In VIEW 1 and 2, the NEI VFQ-25 assessments mum of dosing every 12 weeks with interim PRN
[65] were conducted by trained interviewers who monthly injections) [67]. The results of the
were masked to treatment arm at the screening second- year data revealed 81.6–85.7% patient
visit, and at weeks 12, 24, 36, and 52. Among the retention in all groups. This retention rate might
four arms of treatment in VIEW 1 and 2, the inves- be high enough to uphold enough power to iden-
tigators chose to compare the NEI VFQ-25 sub- tify statistical significance of these second-year
scale scores between aflibercept A2q8 and data. Comparable visual acuity maintenance (91–
ranibizumab R0.5q4 as primary outcome. 92%) was found in each arm at week 96. The
Baseline NEI VFQ-25 scores were similar in total number of active injections (from baseline
both VIEW studies although the questionnaire to week 96) was 16.0–16.2 in the two original
was administered by phone in VIEW 1 and face monthly aflibercept groups, 16.5 in the original
to face in VIEW 2. Mean change from baseline to R0.5q4, and 11.2 in the original A2q8 arm. The
52 weeks was similar for ranibizumab R0.5q4 finding showed that visual acuity maintenance
and aflibercept A2q8 across all 12 subscales. could be achieved for up to 96 weeks in the A2q8
Improvement of 4 points or more was observed in arm with similar gains in BCVA compared with
both treatments, and in both studies for subscales ranibizumab despite more than five fewer
near vision, distance vision, role difficulties, and injections.
dependency.
20.3.3.4 ubgroup Analysis on Final
S
20.3.3.2 ost Hoc Analysis of Year 1
P Visual Outcomes of Year 2
VIEW Studies: Persistent VIEW Studies
Fluid at 3 Months The investigators classified the patients who
A total of 1815 eyes (75% of the original studies) completed week 96 follow-up into four groups
were analyzed to find an association between [68]. They were (1) patients who lost more than 5
early persistent retinal fluid after 3 initial injec- letters from week 52 to week 96, (2) patients who
tions and the effect of aflibercept or ranibizumab at week 96 lost more than 5 letters from the previ-
on visual outcomes [66]. The analysis included ous best BCVA time point, (3) patients who
patients with known retinal fluid status at base- received the scheduled treatment of 3 injections
line and weeks 4, 8, and 12 in the three treatment during week 52–96, and (4) patients who lost 5
arms: ranibizumab R0.5q4 (n = 595), aflibercept letters or more at two consecutive visits from
A2q4 (n = 613), and A2q8 (n = 607) after 3 week 52 to 96 and received active treatment at
monthly injections. The proportion of eyes with that second visit.
It was found that the patterns of response to in either ranibizumab or aflibercept, but no differ-
treatment of these four groups during the first ence of BCVA change from baseline between
year were indistinguishable. At week 96, patients both groups (letter score difference 2.3; 95% CI
in Group 1 had the worst final BCVA, fell to be −0.1 to 4.7; p = 0.06) and no difference in the
lower than baseline, among the four groups. number of injection (rate ratio = 1.00; 95% CI
Patients in Group 3 had the best final BCVA, 1.0–1.1; p = 0.86) at month 12.
which was maintained from week 52 despite hav-
ing the least number of injections. Patients in
Group 4, despite having the greatest number of 20.4 Safety Data
treatments among the four groups, did not regain
final BCVA up to the level at week 52 as Group 3. According to CATT 2-year data, the rates of
This post hoc showed that switching from death were 5.3% in ranibizumab and 6.1% in
monthly to PRN during weeks 52–96 might have bevacizumab (p = 0.60). The proportion of
created the different patterns of response among patients with one or more systemic serious
patients. Data from VIEW 1 and 2 themselves adverse events was higher with bevacizumab
might not explain these patterns. These post hoc than ranibizumab (39.9% vs. 31.7%; adjusted
analyses revealed certain groups of patients that risk ratio, 1.30; 95% CI 1.07–1.57; p = 0.009)
warranted future studies. [28]. According to IVAN 2-year data, the athero-
thrombotic events were not different among all
treatment arms. The mortality rate was found to
20.3.4 I nterim Analysis of Secondary be lower with monthly than as-needed treatment
Outcome of Ranibizumab (OR 0.47, 95% CI 0.22–1.03; p = 0.05) with no
Versus Aflibercept in T&E: difference between drug groups (OR 0.96, 95%
RIVAL Study CI 0.46–2.02; p = 0.91) [29]. Pooled estimates
from both IVAN and CATT of safety outcomes
RIVAL study [69] was a recent RCT designed to showed no difference by drugs for deaths and
compare treatment with ranibizumab and afliber- atherothrombotic events. However, bevacizumab
cept for the development of geographic atrophy treatment might increase the risk of any systemic
(GA) in nAMD patients. The primary objective adverse events (p = 0.008) [29].
was the mean change in area of macular atrophy With longer follow-up at 5 years, data from
from baseline to month 24 and safety results. The CATT showed that 7.6% of patients who com-
recently published paper showed interim analysis pleted the follow-up and originally assigned to
at 12 months for the secondary endpoints as ranibizumab had an atherothrombotic event com-
BCVA gain and number of injections. pared with 4.5% of bevacizumab (p = 0.04) [64].
The head-to-head comparison between ranibi- There was no new safety concern reported in
zumab and aflibercept in RIVAL was in T&E clinical trials of newer anti-VEGF agents, such as
regimen. The criteria to determine treatment inter- aflibercept [7] and brolucizumab [44].
val were a loss of visual acuity of 5 or more letters Based on the systematic review, available
from the baseline, a new retinal hemorrhage, and information on the adverse effects of each medi-
retinal fluid. Injection interval was every 4 weeks cation, ranibizumab, aflibercept, did not suggest
and it was extended by 2 weeks until maximum a higher incidence of potentially vision-
12 weeks if there was no disease activity defined threatening complications compared with con-
to determine treatment interval as above. The trols; however, clinical trial sample sizes were
interval was reduced for 2 weeks with any sign of not sufficient to estimate differences in rare
disease activity. If two or more disease activities safety outcomes [2]. The systematic review of
were found during the T&E period, the interval conbercept showed the incidence of adverse
was reset to 4 weeks. At month 12, the change of events in patients receiving conbercept was sig-
BCVA from baseline was statistically significant nificantly lower than those receiving triamcino-
lone acetonide (RR = 0.25, 95% CI 0.09–0.72), received 4.8 ranibizumab injections in the combi-
but was similar to the other therapies [43]. There nation arm vs. 5.1 injections in the monotherapy
were a few recent case reports on brolucizumab- arm over 12 months; the mean number of ranibi-
associated retinal vasculitis. zumab retreatments was 1.9 in the combination
arm and 2.2 in the monotherapy arm (p = 0.1373).
DENALI study [71] assessed the efficacy and
20.5 Failed Clinical Trials safety of PDT in either standard fluence (SF) or
of CNV-AMD reduced fluence (RF) combined with ranibi-
zumab PRN regimen vs. ranibizumab monother-
We may judge the word “failed clinical trials” in apy in monthly regimen. Mean BCVA change at
at least two ways. On a statistical standpoint, month 12 was +5.3 and + 4.4 letters in PDT-SF
some trials may fail because they cannot reject plus ranibizumab and PDT-RF plus ranibizumab,
the null hypothesis. On clinical standpoint, some respectively, compared with +8.1 letters with
trials may fail, even though they can reject the ranibizumab monotherapy. Non-inferiority of
null hypothesis and the alternate hypothesis can either combination regimen to monthly ranibi-
be accepted, because the statistical significance is zumab monotherapy (97.5% CI, 7.90 to infinity;
not clinically meaningful. p = 0.0666 for PDT-SF plus and 97.5% CI, 8.51
to infinity; p = 0.1178 for PDT-RF plus) was not
demonstrated.
20.5.1 Combination Therapy The lessons learned from these trials of com-
bination therapy was that MONT BLANC could
FOCUS [55] was conducted to assess the combi- reject the null hypothesis, but analysis of second-
nation of PDT with ranibizumab monthly injec- ary outcome did not provide clinically meaning-
tions in CNV-AMD compared with PDT alone. ful data. On the other hand, DELANI could not
At month 24, 25% of patients in combined treat- even reject the null hypothesis. The combination
ment group had gained >15 letters (vs 7% for therapy did not gain popularity in real-world
PDT alone; p = 0.006), and the two treatment treatment of CNV-AMD.
groups differed by 12.4 letters in mean VA change
(p < 0.001). Furthermore, the combined group
exhibited less lesion growth, greater reduction of 20.5.2 F
ixed Quarterly Dosing
CNV leakage, reduction of subretinal fluid accu- Regimen
mulation, and required fewer PDT re-treatments.
It was clear from FOCUS that the combination PIER study [72] was a 24-month RCT evaluating
therapy was superior to PDT alone. ranibizumab with fixed quarterly regimen for
MONT BLANC [70] assessed whether combi- patients with CNV-AMD. The two doses, 0.3 and
nation of PDT and ranibizumab in PRN regimen 0.5 mg, were compared with sham injection.
was non-inferior to ranibizumab PRN monother- Ranibizumab was initially administered monthly
apy. At 12 months, the mean change in BCVA for 3 doses, then switched to every 3 months
from baseline was +2.5 and + 4.4 letters in the (quarterly dosing). The results from the first
combination arm and monotherapy arm respec- 12 months revealed that the treatment effect
tively with 95% CI, −5.76 to 1.86; the lower limit declined during quarterly dosing.
was greater than the present −7 letters of non- The protocols were subsequently revised in the
inferiority margin. This study achieved the pri- following year to have eligible patients in the sham
mary endpoint. However, it was the secondary group cross over to 0.5 mg ranibizumab quarterly
endpoint, the number of ranibizumab injections, and all eligible randomized patients cross over to
which failed the combination therapy to be 0.5 mg ranibizumab monthly. At month 24, visual
applied for CNV-AMD in the real world. It was acuity decreased even further at the mean of 2.2,
shown that the combination might not decrease and 2.3 letters from baseline in the original 0.3 mg
treatment burden from the monotherapy. Patients and 0.5 mg groups and decreased at the mean of
21.4 letters from baseline in the sham group, ied in combination with anti-VEGF for the
(p < 0.0001 for each ranibizumab group vs sham). treatment of CNV-AMD [74, 75]. Another failed
BCVA of patients in 0.3 mg and 0.5 mg groups phase III trial was from the study of lampali-
who were crossed over to monthly from quarterly zumab for treatment of GA [76].
dosing increased at the mean of 2.2 and 4.1 letters, In the phase II study of fovista, a total of 449
respectively. PIER study showed the fixed quar- eyes with “classic-containing” CNV-AMD were
terly was superior to sham but provided visual loss. randomized 1:1:1 among 3 monthly treatment
EXCITE study [73] was the RCT assessing arms: 0.3 mg fovista combined with ranibizumab,
whether the fixed quarterly dosing was non- 1.5 mg fovista combined with ranibizumab, and
inferior to fixed monthly dosing of ranibizumab. ranibizumab alone. The primary outcome, change
Two quarterly dosing regimens, 0.3 mg and in visual acuity at 24 weeks, strongly supported a
0.5 mg were compared with 0.3 mg monthly regi- treatment benefit from combination of 1.5 mg
men. A non-inferiority limit of −6.8 was pre- fovista and ranibizumab over ranibizumab alone.
determined based on the results of MARINA in However, it was found that there was a substan-
which the value of 6.8 was approximately half of tial imbalance in clinical characteristics, such as
the minimum estimated difference (13.6 which lesion sizes, between the groups in phase II data.
was the lower limit of a 2-sided 95% CI) of the Despite this, instead of adherence to the phase II
mean change in BCVA from baseline to month 12 protocol in phase III trial of fovista, the investiga-
between ranibizumab 0.3 mg and sham injection. tors decided to change the inclusion criteria to
The results of EXCITE showed that in per- include eyes with subretinal hyper-reflective
protocol analysis, BCVA increased from baseline material (SHRM)-containing CNV, instead of
to month 12 by +4.9, +3.8, and + 8.3 letters in classic-containing CNV, in the phase III trial.
0.3 mg quarterly, 0.5 mg quarterly and the This decision was based on a subgroup analysis
monthly group with 95% CI −7.1 to −0.2 and of the phase II trial showing CNV associated
−7.9 to −0.7 when both quarterly dosing arms with SHRM, regardless of the presence of the
were compared with monthly arm. In the classic component, responded best to fovista
intention-to-treat analysis, mean BCVA improve- combined with ranibizumab. It was very likely
ment were + 4.0, +2.8, and + 8.0 letters and the that this protocol change might not allow the
95% CIs are −7.7 to −0.9 and −8.6 to −1.7, for results from the phase II trial to be carried over
the same comparisons in PP analysis, respec- into the phase III trial. This is a lesson learned for
tively. The non-inferiority of the quarterly dos- investigators of clinical trial [74].
ings was not reached in this study. In the phase II trial of topical squalamine, all
PIER and EXCITE have given quarterly dos- types of treatment-naïve CNV-AMD were random-
ing the same fate as MONT BLANC and ized to receive a ranibizumab injection at baseline
DELANI had given to combination therapy. followed by topical squalamine twice a day or
vehicle drops twice a day. All eyes then received
as-needed monthly injections of ranibizumab.
20.5.3 Failed Phase III Trials After 36 weeks, better visual acuity was found in
the arm receiving topical squalamine. However,
There were a few clinical trials on new drugs for according to pre-specified subgroup analysis, an
treatment of CNV-AMD which demonstrated even greater visual benefit was found in eyes with
promising results in phase I and II clinical trials classic-containing lesions and eyes with any occult
but failed to duplicate the results in the larger fol- CNV measuring less than 10 mm2. The protocol in
lowing phase III RCTs recently. the phase III trial was then changed, according to
Of the three recently failed phase III RCTs of these positive outcomes, to include eyes with any
new drugs for AMD over the past 2 years, there occult CNV measuring less than 10 mm2
were two for CNV-AMD: fovista and squalamine (ClinicalTrials.gov Identifier: NCT02727881).
eye drop. Fovista is an inhibitor of platelet- This phase III trial of topical squalamine followed
derived growth factor (PDGF), which was stud- the same fate as the trial of fovista. It failed to rep-
licate the results of the subgroup analysis of the line features. They also used intention-to-treat
phase II trial. Another lesson learned [74, 77]. analysis as in the RCT.
The investigators found significant BCVA
improvement of both treatment groups, +3.7 and
20.6 Real-World Clinical Trials + 4.3 letters for ranibizumab and aflibercept
respectively at 12 months with a similar number
Clinical trials are almost always conducted in an of injections. These results supported data from
ideal situation when most variables are controlled some observational case series that the efficacy in
to prevent bias and confounding factors. The situ- the real world was somehow lower than the effi-
ation in clinical trials is rarely found in the real cacy in RCTs in which improvement of more
world. In fact, it is the real world where we would than 5 letters was relatively common. The inves-
like to apply the results from clinical trials for tigators found further that there was a low rate of
treating our patients. A gap between real-world switching treatment between the two agents.
evidence and clinical trial evidence always exists However, once it occurred, there was more
[78]. The gap has increasingly caused one of the switching from ranibizumab to aflibercept than
frequently asked questions in clinical practice: vice versa. This might also reflect real-world
Can the drugs in clinical trials achieve similar impression that aflibercept might have higher
efficacy in the real world? This question may be efficacy than ranibizumab and the switching
answered by studies conducted in the real world. should be performed from the agent with a lower
There was a study claimed to be the first real- efficacy.
world study on head-to-head comparison between This latter study, although based on retrospec-
ranibizumab and aflibercept in fixed dosing regi- tive data as the former study, demonstrated how
men for both drugs (three loading monthly injec- we could conduct a clinical trial by using real-
tions followed by bimonthly injections) [79]. world data in more reliable ways to yield stronger
However, this study was not RCT but a single- evidence.
center, retrospective, comparative, and nonran-
domized study. The retrospective nature of data
revealed discrepancy in the number of patients 20.7 Future Directions
and age between the two groups. The results
showed that the aflibercept arm had better BCVA, The ongoing clinical trials of CNV-AMD are
approximately 1–4 letters, and also better CRT investigating on some new anti-VEGF agents
improvement than ranibizumab at every time [80, 81], port delivery system [82], gene therapy
point. Could this occur because of the suboptimal [83, 84], etc. Some of these trials aim to decrease
dosing (bimonthly) of ranibizumab group or treatment burden by lowering number of treat-
because of the older age of the patients in this ments, some aim to increase magnitude of visual
group? It might be difficult to answer from data in improvement or to be alternatives to the current
this study in which there were no actual controls. anti-VEGF agents. Proving these new modalities
In another study of real-world evidence, treat- effective in real-world practice may be a long
ment of CNV-AMD by ranibizumab and afliber- way to go. The first steps for achieving this goal,
cept was compared from observational database however, still count on well-designed, well-
[69]. This study, however, tried to overcome limi- conducted, and well-interpreted clinical trials.
tation of retrospective nature of real-world data
by matching baseline characteristics, such as VA,
lesion size, and lesion type, between the eyes
received the two agents. The investigators
References
adapted principles in the design of RCT into this 1. Özkiris A. Anti-VEGF agents for age-related
observational data, such as identifying zero time macular degeneration. Expert Opin Ther Pat.
for determination of patient eligibility and base- 2010;20(1):103–18.
2. Solomon SD, Lindsley K, Vedula SS, Krzystolik MG, 15. Brown DM, Kaiser PK, Michels M, Soubrane G,
Hawkins BS. Anti-vascular endothelial growth fac- Heier JS, Kim RY, et al. Ranibizumab versus verte-
tor for neovascular age-related macular degeneration. porfin for neovascular age-related macular degenera-
Cochrane Database Syst Rev. 2019;3:1–123. tion. N Engl J Med. 2006;355(14):1432–44.
3. Guo MY, Cheng J, Etminan M, Zafari Z, Maberley 16. Rosenfeld PJ, Brown DM, Heier JS, Boyer DS, Kaiser
D. One year effectiveness study of intravitreal PK, Chung CY, et al. Ranibizumab for neovascular
aflibercept in neovascular age-related macular age-related macular degeneration. N Engl J Med.
degeneration: a meta-analysis. Acta Ophthalmol. 2006;355(14):1419–31.
2019;97(1):e1–7. 17. Triantafylla M, Massa HF, Dardabounis D, Gatzioufas
4. Li J, Xu J, Chen Y, Zhang J, Cao Y, Lu P. Efficacy Z, Kozobolis V, Ioannakis K, et al. Ranibizumab for
comparison of intravitreal anti-VEGF therapy for the treatment of degenerative ocular conditions. Clin
three subtypes of neovascular age-related macular Ophthalmol. 2014;8:1187–98.
degeneration: a systematic review and meta-analysis. 18. The CATT Research Group. Ranibizumab and bevaci-
J Ophthalmol. 2018;2018:1–10. zumab for neovascular age-related macular degenera-
5. Kovach JL, Schwartz SG, Flynn HW, Scott tion. N Engl J Med. 2011;364(20):1897–908.
IU. Anti-VEGF treatment strategies for wet AMD. J 19. Heier JS, Brown DM, Chong V, Korobelnik JF, Kaiser
Ophthalmol. 2012;2012:1–7. PK, Nguyen QD, et al. Intravitreal aflibercept (VEGF
6. Wykoff CC, Clark WL, Nielsen JS, Brill JV, Greene trap-eye) in wet age-related macular degeneration.
LS, Heggen CL. Optimizing anti-VEGF treatment Ophthalmology. 2012;119(12):2537–48.
outcomes for patients with neovascular age-related 20. Smiddy WE. Economic implications of cur-
macular degeneration. J Manag Care Spec Pharm. rent age-related macular degeneration treatments.
2018;24(2-a):S3–15. Ophthalmology. 2009;116(3):481–7.
7. Schmidt-Erfurth UM, Richard G, Augustin A, 21. Wang B, Wang H, Tu XM, Feng C. Comparisons of
Aylward WG, Bandello F, Corcòstegui B, et al. superiority, non-inferiority, and equivalence trials.
Guidance for the treatment of neovascular age-related Shanghai Arch Psychiatry. 2017;29(6):385–8.
macular degeneration. Acta Ophthalmol Scand. 22. Murray GD. Switching between superiority and non-
2007;85(5):486–94. inferiority switching between superiority and non-
8. Macular Photocoagulation Study Group. Laser inferiority: an introductory note. Br J Clin Pharmacol.
photocoagulation of subfoveal neovascular lesions 2001;52(3):219.
of age-related macular degeneration updated find- 23. Christensen E. Methodology of superiority vs. equiv-
ings from two clinical trials. Arch Ophthalmol. alence trials and non-inferiority trials. J Hepatol.
1993;111:1200–9. 2007;46(5):947–54.
9. Al Jundi A, Sakka S. Protocol writing in clinical 24. Schumi J, Wittes JT. Through the looking glass:
research. J Clin Diagn Res. 2016;10(11):ZE10–3. understanding non-inferiority. Trials. 2011;12:106.
10. Dawn Teare M, Dimairo M, Shephard N, Hayman A, 25. Hahn S. Understanding noninferiority trials. Korean J
Whitehead A, Walters SJ. Sample size requirements Pediatr. 2012;55(11):403–7.
to estimate key design parameters from external pilot 26. Ferreira-Gonzalez I. Basis for the interpretation of
randomised controlled trials: a simulation study. noninferiority studies: considering the ROCKET–AF,
Trials. 2014;15:264. RE-LY, and ARISTOTLE studies. Rev Esp Cardiol
11. Macular Photocoagulation Study Group. Subfoveal (Engl Ed). 2014;67(6):432–5.
neovascular lesions in age-related macular degen- 27. Berg K, Pedersen TR, Sandvik L, Bragadóttir
eration. Guidelines for evaluation and treatment R. Comparison of ranibizumab and bevacizumab
in the macular photocoagulation study. Macular for neovascular age-related macular degeneration
photocoagulation study group. Arch Ophthalmol. according to LUCAS treat-and-extend protocol.
1991;109(9):1242–57. Ophthalmology. 2015;122(1):146–52.
12. TAP study group. Photodynamic therapy of subfoveal 28. The CATT Research Group, Martin DF, Maguire
choroidal neovascularization in age-related macu- MG, Fine SL, Ying GS, Jaffe GJ, et al. Ranibizumab
lar degeneration with verteporfin: two-year results and bevacizumab for treatment of neovascular age-
of 2 randomized clinical trials-tap report 2. Arch related macular degeneration: two-year results.
Ophthalmol. 1999;117(10):1329–45. Ophthalmology. 2012;119(7):1388–98.
13. VIP study group. Verteporfin therapy of subfoveal 29. Chakravarthy U, Harding SP, Rogers CA, Downes SM,
choroidal neovascularization in age-related macu- Lotery AJ, Culliford LA, et al. Alternative treatments
lar degeneration: two-year results of a randomized to inhibit VEGF in age-related choroidal neovascu-
clinical trial including lesions with occult with no larisation: 2-year findings of the IVAN randomised
classic choroidal neovascularization—vertepor- controlled trial. Lancet. 2013;382(9900):1258–67.
fin in photodynamic therapy. Am J Ophthalmol. 30. Krebs I, Schmetterer L, Boltz A, Told R, Vécsei-
2001;131:541–60. Marlovits V, Egger S, et al. A randomised double-
14. Brown GC, Landy J. Photodynamic therapy of masked trial comparing the visual outcome after
subfoveal choroidal neovascularization in patho- treatment with ranibizumab or bevacizumab in
logic myopia with verteporfin. Arch Ophthalmol. patients with neovascular age-related macular degen-
2001;119:198–207. eration. Br J Ophthalmol. 2013;97(3):266–71.
31. Kodjikian L, Souied EH, Mimoun G, Mauget-Faÿsse 44. Dugel PU, Koh A, Ogura Y, Jaffe G, Schmidt-Erfurth
M, Behar-Cohen F, Decullier E, et al. Ranibizumab U, Brown D, et al. HAWK and HARRIER: phase
versus bevacizumab for neovascular age-related 3, multicenter, randomized, double-masked trials of
macular degeneration: results from the GEFAL brolucizumab for neovascular age-related macular
noninferiority randomized trial. Ophthalmology. degeneration. Ophthalmology. 2020;127:72–84.
2013;120(11):2300–9. 45. Singerman LJ, Weichselberger A, Sallstig P. OSPREY
32. Schauwvlieghe AME, Dijkman G, Hooymans JM, trial: randomized, active-controlled, phase II study to
Verbraak FD, Hoyng CB, Dijkgraaf MGW, et al. evaluate safety and efficacy of RTH258, a human-
Comparing the effectiveness of bevacizumab to ized single-chain anti-VEGF antibody fragment, in
ranibizumab in patients with exudative age-related patients with neovascular AMD. Investig Ophthalmol
macular degeneration. The BRAMD study. PLoS Vis Sci. 2015;56
One. 2016;11(5):1–16. 46. Csaky KG. A variable-dosing regimen with intravit-
33. Chung C, Pherwani N. Ziv-aflibercept: a novel real ranibizumab for neovascular age-related macular
angiogenesis inhibitor for the treatment of met- degeneration: year 2 of the PrONTO study: comment.
astatic colorectal cancer. Am J Heal Pharm. Evidence-Based Ophthalmol. 2010;11(1):26–7.
2013;70(21):1887–96. 47. Holz FG, Amoaku W, Donate J, Guymer RH. Safety
34. Cheng YD, Yang H, Chen GQ, Zhang ZC. and efficacy of a flexible dosing regimen of ranibi-
Molecularly targeted drugs for metastatic colorectal zumab in neovascular age-related macular degenera-
cancer. Drug Des Devel Ther. 2013;7:1315–22. tion. Ophthalmology. 2010;118(4):663–71.
35. Malik D, Tarek M, Del Carpio JC, Ramirez C, Boyer 48. Busbee BG, Ho AC, Brown DM, Heier JS, Suñer
D, Kenney MC, et al. Safety profiles of anti-VEGF IJ, Li Z, et al. Twelve-month efficacy and safety of
drugs: Bevacizumab, ranibizumab, aflibercept and 0.5 mg or 2.0 mg ranibizumab in patients with sub-
ziv-aflibercept on human retinal pigment epithelium foveal neovascular age-related macular degeneration.
cells in culture. Br J Ophthalmol. 2014;98:11–6. Ophthalmology. 2013;120(5):1046–56.
36. de Oliveira DJ, Badaró E, Novais E, Colicchio D, 49. Wykoff CC, Croft DE, Brown DM, Wang R, Payne
Chiarantin G, Matioli M, et al. Preclinical investiga- JF, Clark L, et al. Prospective trial of treat-and-extend
tions of intravitreal ziv-aflibercept. Ophthalmic Surg versus monthly dosing for neovascular age-related
Lasers Imaging Retina. 2014;45(6):577–84. macular degeneration elated macular degenera-
37. de Oliveira DJ, de Andrade G, Kniggendorf V, Novais tion TREX-AMD 1-year results. Ophthalmology.
E, Maia A, Meyer C, et al. Clinical and electrophysi- 2015;122:2514–22.
ological evaluation after intravitreal ziv-aflibercept 50. Gillies MC, Nguyen V, Daien V, Arnold JJ, Morlet
for exudative age-related macular degeneration. N, Barthelmes D. Twelve-month outcomes of ranibi-
Ophthalmologica. 2017;37(8):1499–507. zumab vs. aflibercept for neovascular age-related
38. de Oliveira D, de Andrade G, Kniggendorf V, Novais macular degeneration: data from an observational
E, Takahashi VKL, Maia A, et al. Intravitreal Ziv- study. Ophthalmology. 2016;123(12):2545–53.
aflibercept for neovascular age-related macular degen- 51. Silva R, Berta A, Larsen M, Macfadden W, Feller C,
eration: 52-week results. Retina. 2019;39:648–55. Monés J. Treat-and-extend versus monthly regimen
39. Mansour AM, Charbaji A, Farah ME, Mansour HA, in neovascular age-related macular degeneration:
Chhablani J. Long-term outcome of treat and extend results with ranibizumab from the TREND study.
intravitreal ziv-aflibercept therapy. Br J Ophthalmol. Ophthalmology. 2018;125(1):57–65.
2018;103:1–4. 52. Freemantle N. Interpreting the results of secondary
40. Jørstad ØK, Faber RT, Moe MC. Two-year func- end points and subgroup analyses in clinical trials:
tional and anatomical results after converting treat- should we lock the crazy aunt in the attic? Br Med J.
ment resistant eyes with exudative age-related 2001;322(7292):989–91.
macular degeneration to aflibercept in accordance 53. Fridman M, Erder MH. Interpreting the results of a
with a treat and extend protocol. Acta Ophthalmol. retrospective comparison of test and reference treat-
2017;95(5):460–3. ments in a randomized clinical trial setting. Clin Drug
41. Baghi A, Jabbarpoor Bonyadi MH, Ramezani A, Investig. 2014;35(2):133–40.
Azarmina M, Moradian S, Dehghan MH, et al. Two 54. Singer MA, Awh CC, Sadda S, Freeman
doses of intravitreal Ziv-Aflibercept versus beva- WR. HORIZON: an open-label extension trial
cizumab in treatment of diabetic macular edema: of ranibizumab for choroidal neovascularization
a three-armed. Double-blind randomized trial. secondary to age-related macular degeneration.
Ophthalmol Retina. 2016;1(2):103–10. Ophthalmology. 2012;119(6):1175–83.
42. Liu K, Song Y, Xu G, Ye J, Wu Z, Liu X, et al. 55. Heier JS, Boyer DS, Ciulla TA, Ferrone PJ, Jumper
Conbercept for treatment of neovascular age-related JM, Gentile RC, et al. Ranibizumab combined
macular degeneration: results of the random- with verteporfin photodynamic therapy in neo-
ized phase 3 PHOENIX study. Am J Ophthalmol. vascular age-related macular degeneration: year
2019;197:156–67. 2 results of the FOCUS study. Arch Ophthalmol.
43. Zhang J, Liang Y, Xie J, Li D, Hu Q, Li X, et al. 2006;124(11):1532–42.
Conbercept for patients with age-related macular 56. Daniel E, Shaffer J, Ying GS, Grunwald JE, Martin
degeneration: a systematic review. BMC Ophthalmol. DF, Jaffe GJ, et al. Outcomes in eyes with retinal
2018;18(1):1–12. angiomatous proliferation in the comparison of
age-related macular degeneration treatments trials 69. Gillies MC, Hunyor AP, Arnold JJ, Guymer RH, Wolf
(CATT). Ophthalmology. 2016;123(3):609–16. S, Ng P, et al. Effect of ranibizumab and aflibercept
57. Willoughby AS, Ying GS, Toth CA, Maguire MG, on best-corrected visual acuity in treat-and-extend
Burns RE, Grunwald JE, et al. Subretinal hyperreflec- for neovascular age-related macular degeneration:
tive material in the comparison of age-related macu- a randomized clinical trial. JAMA Ophthalmol.
lar degeneration treatments trials. Ophthalmology. 2019;137(4):372–9.
2015;122(9):1846–1853.e5. 70. Larsen M, Schmidt-Erfurth U, Lanzetta P, Wolf S,
58. Shah N, Maguir MG, Martin DF, Shaffer J, Ying G-S, Simader C, Tokaji E, et al. MONT BLANC study
Grunwald JE, et al. Angiographic cystoid macular group. Verteporfin plus ranibizumab for choroidal
edema and outcomes in the comparison of age-related neovascularization in age-related macular degen-
macular degeneration treatments trials (CATT). J eration: twelve-month MONT BLANC study results.
Autism Dev Disord. 2017;47(3):549–62. Ophthalmology. 2012;119(5):992–1000.
59. Hagstrom SA, Ying GS, Pauer GJT, Sturgill-Short 71. Kaiser PK, Boyer DS, Cruess AF, Slakter JS, Pilz S,
GM, Huang J, Callanan DG, et al. Pharmacogenetics Weisberger A, DENALI Study Group. Verteporfin
for genes associated with age-related macular degen- plus ranibizumab for choroidal neovascularization
eration in the comparison of AMD treatments trials in age-related macular degeneration: twelve-month
(CATT). Ophthalmology. 2013;120(3):593–9. results of the DENALI study. Ophthalmology.
60. Grunwald JE, Daniel E, Huang J, Ying GS, 2012;119(5):1001–10.
Maguire MG, Toth CA, et al. Risk of geographic 72. Abraham P, Yue H, Wilson L. Randomized, double-
atrophy in the comparison of age-related macu- masked, sham-controlled trial of ranibizumab for
lar degeneration treatments trials. Ophthalmology. neovascular age-related macular degeneration: PIER
2014;121(1):150–61. study year 2. Am J Ophthalmol. 2010;150(3):239–48.
61. Grunwald JE, Pistilli M, Daniel E, Ying GS, Pan 73. Schmidt-Erfurth U, Eldem B, Guymer R, Korobelnik
W, Jaffe GJ, et al. Incidence and growth of geo- J-F, Schlingemann RO, Axer-Siegel R, et al. Efficacy
graphic atrophy during 5 years of comparison of and safety of monthly versus quarterly ranibi-
age-related macular degeneration treatments trials. zumab treatment in neovascular age-related macular
Ophthalmology. 2017;124(1):97–104. degeneration: the EXCITE study. Ophthalmology.
62. Ho AC, Busbee BG, Regillo CD, Wieland MR, Van 2010;118(5):831–9.
Everen SA, Li Z, et al. Twenty-four-month efficacy 74. Rosenfeld PJ, Feuer WJ. Lessons from recent phase
and safety of 0.5 mg or 2.0 mg ranibizumab in patients III trial failures: don’t design phase III trials based on
with subfoveal neovascular age-related macular degen- retrospective subgroup analyses from phase II trials.
eration. Ophthalmology. 2014;121(11):2181–92. Ophthalmology. 2018;125(10):1488–91.
63. Sharma S, Toth CA, Daniel E, Grunwald JE, Maguire 75. Dunn EN, Hariprasad SM, Sheth VS. An overview of
MG, Ying GS, et al. Macular morphology and visual the fovista and rinucumab trials and the fate of anti-
acuity in the second year of the comparison of age- PDGF medications. Ophthalmic Surg Lasers Imaging
related macular degeneration treatments trials. Retina. 2017;48(2):100–4.
Ophthalmology. 2016;123(4):865–75. 76. Holz FG, Sadda SR, Busbee B, Chew EY, Mitchell P,
64. Maguire MG, Martin DF, Shuang YG, Jaffe GJ, Tufail A, et al. Efficacy and safety of lampalizumab for
Daniel E, et al. Five-year outcomes with anti–vascu- geographic atrophy due to age-related macular degen-
lar endothelial growth factor treatment of neovascular eration: Chroma and spectri phase 3 randomized clini-
age-related macular degeneration: the comparison of cal trials. JAMA Ophthalmol. 2018;136(6):666–77.
age-related macular degeneration treatments trials. 77. Wroblewski JJ, Hu AY. Topical squalamine 0.2%
Ophthalmology. 2016;123(8):1751–61. and intravitreal ranibizumab 0.5 mg as combination
65. Yuzawa M, Fujita K, Wittrup-Jensen KU, Norenberg therapy for macular edema due to branch and cen-
C, Zeitz O, Adachi K, et al. Improvement in vision- tral retinal vein occlusion: an open-label, random-
related function with intravitreal aflibercept: data ized study. Ophthalmic Surg Lasers Imaging Retina.
from phase 3 studies in wet age-related macular 2016;47(10):914–23.
degeneration. Ophthalmology. 2015;122(3):571–8. 78. Kim HS, Lee S, Kim JH. Real-world evidence versus
66. Jaffe GJ, Kaiser PK, Thompson D, Gibson A, Saroj randomized controlled trial: clinical research based
N, Vitti R, et al. Differential response to anti-VEGF on electronic medical records. J Korean Med Sci.
regimens in age-related macular degeneration patients 2018;33(34):1–7.
with early persistent retinal fluid. Ophthalmology. 79. Providência J, Rodrigues TM, Oliveira M, Bernardes
2016;123(9):1856–64. J, Marques JP, Murta J, et al. Real-world results of
67. Schmidt-Erfurth U, Kaiser PK, Korobelnik J-F, Brown aflibercept versus ranibizumab for the treatment of
DM, Chong V, Nguyen QD, et al. Intravitreal afliber- exudative AMD using a fixed regimen. Biomed Res
cept injection for neovascular age-related macular Int. 2018;2018:1–7.
degeneration: ninety-six-week results of the VIEW 80. Callanan D, Kunimoto D, Maturi RK, Patel SS,
studies. Ophthalmology. 2014;121(1):193–201. Staurenghi G, Wolf S, et al. Double-masked, ran-
68. Richard G, Monés J, Wolf S, Korobelnik JF, Guymer domized, phase 2 evaluation of abicipar pegol (an
R, Goldstein M, et al. Scheduled versus pro re anti-VEGF DARPin therapeutic) in neovascular age-
Nata dosing in the VIEW trials. Ophthalmology. related macular degeneration. J Ocul Pharmacol Ther.
2015;122(12):2497–503. 2018;34(10):700–9.
81. Graybug Vision I. GB-102 and GB-103 [Internet]. Positive Interim Phase I/IIa Trial Update for RGX-
2019. Available from: https://graybug.com/pipeline/ 314 for the Treatment of Wet AMD [Internet].
gb-102-gb-103/ 2019. Available from: https://www.pharmiweb.com/
82. Campochiaro PA, Marcus DM, Awh CC, Regillo press-release/2019-05-07/regenxbio-reports-first-
C, Adamis AP, Bantseev V, Chiang Y, Ehrlich JS, quarter-2019-financial-and-operating-results-and-
Erickson S, Hanley WD, Horvath J, Maass KF, Singh additional-positive-interim
N, Tang F, Barteselli G. The port delivery system 84. Kirkner RM. Wet AMD gene therapy shows prom-
with ranibizumab for neovascular age-related macular ise Phase I/IIa results of RGX-314 confirm pro-
degeneration: results from the randomized phase 2 lad- tein uptake over months [Internet]. 2019. Available
der clinical trial. Ophthalmology. 2019;126:1141–54. from: https://www.retina-specialist.com/article/
83. Rockville M. REGENXBIO Reports First Quarter wet-amd-gene-therapy-shows-promise
2019 Financial and Operating Results and Additional
Dr. Paisan Ruamviboonsuk is the Scientific Secretary

of Asia-Pacific Vitreo-Retina Society and a council mem-
ber of the Asia-Pacific Academy of Ophthalmology
(APAO). He is the Immediate Past President of the Royal
College of Ophthalmologists of Thailand and the Past
President of Thai Retina Society. He is well recognized Dr. Variya Nganthavee got her First Class Honor for
internationally and received many important international her bachelor degree from Faculty of Medicine, Siriraj
awards, including the United Nations Public Service Hospital, Mahidol University, Bangkok. She has been
Award for Prevention of Diabetic Blindness Project in working as a research fellow at the Department of
Thailand, APAO Arthur Lim Award, American Academy Ophthalmology, Rajavithi Hospital, Bangkok. Her pub-
of Ophthalmology Achievement Award. lished papers are on artificial intelligence for early detec-
tion of diabetic retinopathy and diabetic macular edema;
her ongoing works are clinical trials of diabetic macular
edema and age-related macular degeneration.
Dr. Peranut Chotcomwongse is currently a resident of

ophthalmology at the Department of Ophthalmology,
Rajavithi Hospital, Bangkok. He worked in EVEREST
Dr. Warisara Pattanapongpaiboon works as a research
Study, a randomized controlled trial for polypoidal cho-
fellow at the Departement of Ophthalmology, Rajavithi
roidal vasculopathy. His published studies are on artificial
Hospital, Bangkok. She is currently working on clinical
intelligence for detection of optical coherent tomography-
trials for diabetic macular edema, age-related macular
based centered-involved diabetic macular edema from
degeneration, and deploying artificial intelligence in real-
color fundus photography and for classifying diabetic
world diabetic retinopathy screening.
retinopathy severity levels.
Dr. Kornwipa Hemarat is currently the attending

physician at the Ophthalmology department,
Navamindradhiraj University. She completed her resi-
dency and Vitreoretinal fellowship program from the
Faculty of Medicine, Siriraj hospital, Mahidol University.
She also completed the post-doctoral research fellowship
at the University of California, San Francisco. Her pub-
lished studies are on artificial intelligence for early detec-
tion of diabetic retinopathy and diabetic macular edema,
dexamethasone implant for macular edema. She has been
working on clinical trials for wet macular degeneration
and deploying artificial intelligence in real-world dia-
betic retinopathy screening.
Clinical Trials Related to Myopic
Choroidal Neovascularization 21
Christine P. S. Ho and Timothy Y. Y. Lai
21.1 Introduction and Natural who are often part of the working population, and
History of Myopic CNV thus it has an immense impact on the patients’
quality of life [4]. Several factors are associated
Pathologic myopia affects up to 3% of the world’s with worse visual prognosis of myopic CNV,
population and is a major cause of blindness including the age of onset being >40 years, sub-
worldwide. The prevalence of visual impairment foveal location of CNV, larger size of CNV
resulting from pathologic myopia ranges from (>400 μm), and lower best-corrected visual acu-
0.1% to 0.5% in European studies and from 0.2% ity (BCVA) at baseline [1].
to 1.4% in Asian studies [1]. A number of com- Various extent of visual loss can develop in
plications relating to the posterior segment can the three stages during the natural course of
develop in patients with pathologic myopia, and myopic CNV. The initial active phase involves
the development of choroidal neovascularization direct photoreceptor damage caused by the CNV,
(CNV) is one of the most serious complications, resulting in central vision loss. CNV then
often leading to sudden onset but later progres- regresses with the formation of localized fibrous
sive central vision loss [2]. Myopic CNV has scars known as Fuchs’ spot, and macular chorio-
been estimated to develop in around 5% to 11% retinal atrophy develops around the regressed
of pathologic myopia patients, [1] with 35% of CNV as a late complication [5]. A long-term
patients with myopic CNV also developing CNV follow-up study by Yoshida et al. demonstrated
in the fellow eye within 8 years, which could dis- that patients with untreated myopic CNV gener-
turb the patients’ binocular visual function [3]. ally have poor visual outcomes due to develop-
Myopic CNV is also the most common cause of ment and enlargement of macular chorioretinal
CNV in individuals younger than 50 years of age atrophy around the regressed CNV. A reduction
in visual acuity to <20/200 was seen in 88.9%
C. P. S. Ho and 96.3% of patients at 5 and 10 years after the
Faculty of Medicine, The University of Hong Kong, onset of CNV, respectively [6].
Pok Fu Lam, Hong Kong
Various treatment modalities have been pro-
T. Y. Y. Lai (*) posed for the treatment of myopic CNV, includ-
Department of Ophthalmology and Visual Sciences,
ing thermal focal laser photocoagulation,
The Chinese University of Hong Kong, Hong Kong
Eye Hospital, Kowloon, Hong Kong verteporfin photodynamic therapy (vPDT), surgi-
cal excision of CNV, and macular translocation.
2010 Retina and Macula Centre, Tsim Sha Tsui,
Kowloon, Hong Kong However, these modalities demonstrated subopti-
e-mail: tyylai@cuhk.edu.hk mal long-term visual outcomes due to potential

284 C. P. S. Ho and T. Y. Y. Lai
CNV recurrence or complications such as month 24, and received retreatment with vPDT or
macular scar, macular atrophy, and postoperative placebo if leakage from CNV was detected on
retinal detachment [7]. Vascular endothelial fluorescein angiogram (FA). The primary out-
growth factor (VEGF) is an important cytokine come of the study was the proportion of eyes that
involved in angiogenesis and vascular permeabil- lost fewer than 8 letters at month 12 compared
ity, and has been implicated in the development with baseline, and the secondary outcome was
of myopic CNV. Elevated levels of VEGF have the proportion of eyes that lost <15 letters or <30
been found in the aqueous humor of eyes with letters at month 12 compared with baseline.
myopic CNV [8]. In recent years, intravitreal For the primary outcomes of the study, patients
anti-VEGF therapies have been introduced for treated with vPDT were more likely to maintain
the treatment of patients with myopic CNV. Large stable BCVA compared with the placebo group,
randomized controlled clinical trials have dem- with 72% of eyes in the vPDT group lost fewer
onstrated good efficacy and safety of anti-VEGF than 8 letters at month 12 compared with 44% in
therapy, and anti-VEGF therapy has subsequently the placebo group (P < 0.01). A smaller propor-
become the standard of care treatment for myo- tion of patients in the vPDT group had a loss of
pic CNV. Agents including ranibizumab, afliber- BCVA ≥15 letters (moderate visual loss) or 30
cept, and conbercept are currently approved for letters (severe visual loss) at month 12, with 14%
treatment of myopic CNV in various countries. and 7% of eyes in the vPDT group reported to
This chapter summarizes the design and major have moderate or severe loss of visual acuity,
results from key clinical trials in the past two respectively, compared with 33% and 8% of eyes
decades concerning vPDT and anti-VEGF thera- in the group receiving placebo (P = 0.01).
pies, demonstrating the efficacies of different In terms of BCVA gains, a greater proportion
therapeutic agents available for the treatment of of patients who had vPDT experienced improve-
patients with myopic CNV. ment in visual acuity. 32% and 6% eyes in the
vPDT group reported to have slight (≥5 letters)
and large (≥15 letters) improvements in visual
21.2 Verteporfin Photodynamic acuity at month 12, respectively, compared with
Therapy Clinical Trial: VIP 15% and 3% in the placebo group. Moreover, the
Study two treatment groups differed by 10 letters in
terms of median change in visual acuity, with a
The Verteporfin in Photodynamic Therapy (VIP) gain of 1 letter in the vPDT group and a loss of 9
study was a multicenter, double-masked, placebo- letters in the placebo group. The median BCVA
controlled, randomized controlled trial evaluat- remained stable in the vPDT group at month 12,
ing the efficacy of vPDT compared with placebo, but decreased to 20/80–2 in the placebo group.
in stabilizing or improving vision in patients with Patients in the vPDT group also had higher mean
subfoveal myopic CNV over a period of contrast sensitivity scores at month 12, with
12 months (Table 21.1) [9]. Patients with subfo- patients in the vPDT and placebo groups experi-
veal CNV caused by pathologic myopia (defined enced a mean gain of 0.1 letters and a mean loss
as less than or equal to −6.0 diopters, or less of 2 letters, respectively. Furthermore, patients
myopic than −6.0 diopters with retinal abnormal- who received vPDT demonstrated better anatom-
ities consistent with pathologic myopia and axial ical outcomes at month 12 compared with
length of ≥26.5 mm), with lesion greatest linear patients in the placebo group. Examination at
dimension ≤5400 μm and BCVA ≥50 Early month 12 revealed progression of classic CNV
Treatment for Diabetic Retinopathy Study beyond the area of lesion identified at baseline in
(ETDRS) letters were included in the study. 36% of patients in the vPDT group and 54% of
Eligible patients (n = 120) were randomized in a patients in the placebo group. 35% of patients in
2:1 ratio to receive vPDT or placebo at baseline. the vPDT group also demonstrated resolved clas-
Patients were followed up every 3 months until sic CNV leakage at month 12, compared with
21
Table 21.1 Summary of main findings of clinical trials on anti-VEGF agents for treating myopic CNV
Number of Mean age Mean BCVA Mean axial
Study Treatment groups eyes (years) (ETDRS letters) length (mm) CNV location Main visual outcomes
VIP [9, 10] vPDT 81 51 62 (median) N/A • Subfoveal 62% Mean BCVA gain of 1 letter at
(median) • Probably subfoveal 19% 12 months
• Not subfoveal 14%
• No CNV or cannot grade 6%
Sham injection 39 46 58 (median) N/A • Subfoveal 69% Mean BCVA loss of 9 letters at
(median) • Probably subfoveal 18% 12 months
• Not subfoveal 8%
• No CNV/cannot grade 5%
REPAIR [12, 13] Ranibizumab 65 55.5 59.5 N/A • Subfoveal 66.2% Mean BCVA gain of 13.8
0.5 mg • Juxtafoveal 26.2% letters at 12 months
• Probably subfoveal/juxtafoveal
7.7%
RADIANCE— Ranibizumab 106 54 55.4 29.3 • Subfoveal 67.0% Mean BCVA gain of 10.5
main study [15] 0.5 mg guided by • Juxtafoveal 24.5% letters at 3 months and 13.8
stabilization of • Extrafoveal 6.6% letters at 12 months
BCVA
Ranibizumab 116 56.1 55.8 28.8 • Subfoveal 69.8% Mean BCVA gain of 10.6
0.5 mg guided by • Juxtafoveal 20.7% letters at 3 months and 14.4
disease activity • Extrafoveal 2.6% letters at 12 months
Clinical Trials Related to Myopic Choroidal Neovascularization
vPDT 55 57.4 54.7 29.4 • Subfoveal 69.1% Mean BCVA gain of 2.2 letters
• Juxtafoveal 29.1% at 3 months and 9.3 letters at
• Extrafoveal 1.8% 12 months
RADIANCE— Caucasian patients 70 58.0 55.9 28.9 • Subfoveal 65.7% Mean BCVA gain of 14.1
Subgroup analysis • Juxtafoveal 20.0% letters at 12 months
[16] • Extrafoveal 4.3%
East-Asian patients 35 54.2 54.2 28.4 • Subfoveal 74.3% Mean BCVA gain of 17.0
• Juxtafoveal 22.9% letters at 12 months
• Extrafoveal 0%
(Continued)
285
Table 21.1 (Continued)
286
Number of Mean age Mean BCVA Mean axial

Study Treatment groups eyes (years) (ETDRS letters) length (mm) CNV location Main visual outcomes
RADIANCE— Additional 7 55.3 59.7 29.7 • Subfoveal 57.1% Mean BCVA gain of 19.4
long-term anti-VEGF • Juxtafoveal 42.9% letters at 48 months
outcomes [17] required • Extrafoveal 0%
No further 34 55.0 55.8 28.9 • Subfoveal 67.6% Mean BCVA gain of 11.0
treatment required • Juxtafoveal 26.5% letters at 48 months
• Extrafoveal 2.9%
• Cannot grade 2.9%
BRILLIANCE Ranibizumab 182 52.0 53.6 28.9 • Subfoveal 73.1% Mean BCVA gain of 9.5 letters
[18] 0.5 mg guided by • Juxtafoveal 8.8% at 3 months and 12.0 letters at
stabilization of • Extrafoveal 16.5% 12 months
BCVA
Ranibizumab 184 51.5 54.2 29.0 • Subfoveal 70.1% Mean BCVA gain of 9.8 letters
0.5 mg guided by • Juxtafoveal 12.0% at 3 months and 13.1 letters at
disease activity • Extrafoveal 17.4% 12 months
Verteporfin PDT 91 49.1 52.1 28.8 • Subfoveal 68.1% Mean BCVA gain of 4.5 letters
• Juxtafoveal 15.4% at 3 months and 10.3 letters at
• Extrafoveal 13.2% 12 months
MYRROR [19] Aflibercept 2 mg 90 58.5 56.4 28.8 • Subfoveal 60.0% Mean BCVA gain of 12.1
• Juxtafoveal 38.9% letters at 6 months and 13.5
• Extrafoveal 1.1% letters at 12 months
Sham/aflibercept 31 57.5 56.6 28.6 • Subfoveal 64.5% Mean BCVA loss of 2.0 letters
2 mg • Juxtafoveal 35.5% at 6 months and gain of 3.9
• Extrafoveal 0% letters at 12 months
SHINY [20] Conbercept 0.5 mg 132 49.30 54.05 N/A N/A Mean BCVA gain of 12.0
letters at 3 months and 13.3
letters at 9 months
Sham/conbercept 44 51.93 49.77 N/A N/A Mean BCVA gain of 0.6 letters
0.5 mg at 3 months and 11.3 letters at
9 months
C. P. S. Ho and T. Y. Y. Lai
21 Clinical Trials Related to Myopic Choroidal Neovascularization 287
28% of patients in the placebo group. For the ters. Patients (n = 65) received an initial intravit-
lesion size at month 12, patients who had vPDT real injection of 0.5 mg ranibizumab, with
were more likely to have lesions ≤1 disc area in subsequent PRN injections if subretinal or intra-
size, while patients who had placebo were more retinal fluid was detected on optical coherence
likely to have lesions size of >3 disc areas. tomography (OCT), or CNV leakage was
Patients in the vPDT group were also reported to detected on fluorescein angiography during
have decreased greatest linear dimension of CNV monthly monitoring until month 12. The primary
leakage, compared with an increase in the pla- outcome of the study was the mean change in
cebo group. No serious ocular or systemic BCVA from baseline to month 12. Following
adverse events of vPDT were reported in the ranibizumab treatment, patients demonstrated a
study. mean BCVA improvement of 12.2 letters at
The VIP study demonstrated the efficacy of month 6 (P < 0.001), and the gain was sustained
vPDT compared with placebo in stabilizing or up to month 12. At month 12, a mean BCVA gain
improving vision, and improving anatomical out- of 13.8 letters was observed, with the greatest
comes in patients with subfoveal myopic CNV mean visual gain of 8.4 letters seen in the first
over a period of 12 months. vPDT subsequently month of treatment. 86% of patients demon-
became the first therapy approved for the treat- strated improvement in BCVA over 12 months,
ment of mCNV. However, in the 2-year follow-up with 36.9% of patients gaining ≥15 letters and
of the study, although vPDT was able to stabilize 50.8% of patients gaining ≥10 letters. 1.5% of
vision, no significant improvement in BCVA was patients reported to have a loss of ≥15 letters.
observed [10]. Prior to receiving treatment, 73.8% of study eyes
were the worse-seeing eyes and 14.4% of eyes
became the better-seeing eyes upon receiving
21.3 Ranibizumab Clinical Trials treatment. Over 12 months, patients were given a
median number of 3 injections, with retreatment
Ranibizumab (Lucentis/Accentrix, Novartis, rates of 21.5%, 18.5%, 16.9%, and 15.5% for 0,
Basel, Switzerland) is a monoclonal anti-VEGF 1, 2, and 3 additional injections, respectively. The
antibody fragment that binds to all VEGF-A iso- study demonstrated that a minimal number of
forms specifically designed for intraocular use. It retreatments with intravitreal ranibizumab 0.5 mg
is currently approved for the treatment of myopic were required to achieve and maintain visual
CNV, with a recommended dosing regimen of a gains in patients with myopic CNV.
single injection of intravitreal ranibizumab Regarding anatomical outcomes, a substantial
0.5 mg followed by retreatment with as a pro re mean reduction in OCT central macular thick-
nata (PRN) approach [11]. ness (CMT) of 108 and 135 μm was observed
from baseline to month 6 and month 12, respec-
tively (P < 0.001), with the greatest mean reduc-
21.3.1 REPAIR Study tion of 109 μm seen in the first month of treatment,
which was consistent with the trend observed in
The REPAIR (Ranibizumab for the treatment of BCVA improvement. A mean decrease in lesion
Choroidal Neovascularization secondary to size was also observed on FA, with a mean
Pathological Myopia: an Individualized decrease of 0.54 mm2 by month 6, and a mean
Regimen) study was a phase II, prospective, mul- decrease of 0.37 mm2 by month 12 (P = 0.0287).
ticenter, open-label study exploring the efficacy Patients were also evaluated with OCT for the
of intravitreal ranibizumab 0.5 mg in patients presence of subretinal fluid, intraretinal cysts, or
with myopic CNV over a period of 12 months edema. At month 12, there were significant
(Table 21.1) [12, 13]. The study included patients reductions in the proportion of patients with sub-
aged ≥18 years with myopia less than or equal to retinal fluid, intraretinal cysts, and edema com-
−6 diopters, with active subfoveal or juxtafoveal pared with baseline, with reductions of 60%,
myopic CNV, and BCVA of 24 to 78 ETDRS let- 38.5%, and 80%, respectively (all P < 0.001). No
new safety issues were identified with the use of −6 diopters, active CNV, and BCVA of 24 to 78
intravitreal ranibizumab treatment in the study. ETDRS letters. Eligible patients (n = 277) were
One ocular and three non-ocular serious adverse randomized in a 2:2:1 ratio into three treatment
events (SAEs) were reported in the study period, groups. Patients in group 1 received intravitreal
which were considered unrelated to intravitreal ranibizumab 0.5 mg on day 1 and month 1, fol-
ranibizumab treatment [12, 13]. lowed by a PRN regimen, with retreatment deter-
The study also evaluated two patient-reported mined by visual acuity (VA) stabilization criteria,
outcomes, including treatment satisfaction mea- defined as no change in BCVA compared with
sured using the Macular Treatment Satisfaction the two previous monthly assessments. Patients
Questionnaire (MacTSQ), and patient wellbeing in group 2 received intravitreal ranibizumab
measured using the 12-item Wellbeing 0.5 mg on day 1 followed by PRN regimen, with
Questionnaire (W-BQ12). Regarding treatment retreatment determined by disease-activity
satisfaction, the mean MacTSQ scores increased guided criteria, which was defined as an impair-
from 55.0 at month 1 to 64.9 at month 12 ment of vision due to intraretinal or subretinal
(P = 0.0001), with the greatest increase in mean fluid, or active CNV leakage. Patients in group 3
MacTSQ scores observed in patients aged were treated with vPDT on day 1, and received
68 years or older. For the W-BQ12 general well- vPDT, intravitreal ranibizumab 0.5 mg, or a com-
being score, the mean general wellbeing score bination of both, guided by disease activity fol-
was 25.6 at baseline. It remained similar at lowing a PRN regimen from month 3. All patients
months 1 and 6, and increased significantly at also received either sham injections or sham PDT
month 12 to 27.3 (P = 0.03). Patients aged 40 depending on treatment groups they were allo-
years or younger had the greatest improvement in cated to. Baseline demographics were similar
general well-being. Overall, no significant corre- among the 3 study groups. The primary objective
lations were found between patients’ changes in of the study was to demonstrate the superiority of
BCVA, MacTSQ score and W-BQ12 general intravitreal ranibizumab treatment over vPDT in
well-being score [14]. mean BCVA gain at month 3, while the second-
In conclusion, the REPAIR study demon- ary objective was to show that ranibizumab
strated the efficacy of intravitreal ranibizumab retreatment guided by disease activity was non-
0.5 mg injection in producing substantial visual inferior to ranibizumab retreatment guided by
and anatomical improvements in patients with visual acuity stabilization criteria at month 6.
myopic CNV. A minimal number of retreatments Regarding visual outcomes, patients who
were required after the initial dose of intravitreal received intravitreal ranibizumab 0.5 mg, irre-
ranibizumab, and good patient satisfaction and spective of retreatment criteria, demonstrated
maintenance of well-being were reported. superiority over vPDT based on mean BCVA
changes from baseline to month 3. Patients in
groups 1 and 2 showed similar mean BCVA
21.3.2 RADIANCE Study improvements, with a mean gain of 10.5 letters
and 10.6 letters, respectively, at month 3, and a
The RADIANCE (Ranibizumab and PDT (verte- mean gain of 13.8 and 14.4 letters, respectively,
porfin) evaluation in myopic choroidal neovascu- at month 12 (P < 0.00001). Patients who had
larization) study was a 12-month, phase III, vPDT demonstrated lower mean BCVA gain of
randomized, double-masked, multicenter study, 2.2 letters at month 3. However, upon switching
which compared the efficacy and safety of intra- to intravitreal ranibizumab with or without vPDT
vitreal ranibizumab 0.5 mg, as guided by visual at month 3, patients showed a significant mean
acuity stabilization criteria (group 1) or disease- improvement of 9.3 letters at month 12, although
activity guided criteria (group 2), with vPDT the mean visual gain was less prominent than
(group 3) in patients with myopic CNV patients who received intravitreal ranibizumab
(Table 21.1) [15]. The study included patients from baseline. Moreover, the two different indi-
aged ≥18 years with myopia less than or equal to vidualized ranibizumab dosing regimens demon-
strated similar efficacies in stabilizing vision. protocol. The number of retreatment was low in
Ranibizumab retreatment guided by disease the ranibizumab groups, with 62% of patients
activity was non-inferior to ranibizumab retreat- who did not require additional ranibizumab in
ment guided by VA stability based on mean groups 1 and 2. From months 6 to 11, 30.9% of
changes in BCVA at month 6, with a mean gain patients in group 3 received intravitreal ranibi-
of 11.7 letters and 11.9 letters, respectively zumab injections, with a median of 2.0 injec-
(P < 0.00001). The study proposed the use of dis- tions. Regarding the safety of treatment, no new
ease activity as a guide to retreatment, as it could safety concerns were identified with ranibizumab
possibly control disease progression with fewer and vPDT treatment, with a low incidence of
intravitreal ranibizumab injections than using VA ocular and non-ocular serious adverse effects in
stabilization criteria. Furthermore, at month 3, a the study groups [15].
greater proportion of patients in the ranibizumab Additional subgroup analyses of the
groups gained ≥10 letters and ≥15 letters com- RADIANCE study were performed to evaluate the
pared to patients in the vPDT group. With intra- influence of age, ethnicity, and various ocular
vitreal ranibizumab introduced to patients in the characteristics such as axial length, refractive
vPDT group, an increased proportion of patients error, and CNV size on the efficacy of intravitreal
gained ≥10 letters and ≥15 letters at month 12. ranibizumab in the treatment of myopic CNV
However, the proportion was still lower than that [16]. Patients from group 2 of the RADIANCE
those who had ranibizumab from baseline. study who had intravitreal ranibizumab retreat-
Hence, the study suggested early initiation of ment guided by disease activity were included in
ranibizumab treatment in patients diagnosed with this post hoc analysis (n = 116), as this retreatment
myopic CNV. protocol based on disease activity was used for the
For anatomical outcomes, a decreased propor- approval of ranibizumab in the treatment of myo-
tion of patients with subretinal fluid, intraretinal pic CNV. Regarding the impact of ethnicity on
edema, and intraretinal cysts were observed in all visual outcomes, East-Asian and Caucasian
treatment groups at month 12. Patients in groups patients had similar baseline BCVA. However,
1 and 2 demonstrated a decrease in mean central East-Asians demonstrated a greater mean BCVA
retinal thickness (CRT) (66.6 and 71.3 μm, improvement of 17.0 letters compared to a mean
respectively) at month 12, which was comparable gain of 14.1 letters in Caucasians at month 12.
to the decrease in CRT of 60.8 μm seen in patients Differences in baseline ocular characteristics also
in the group receiving vPDT followed by PRN influenced patients’ mean BCVA gains at month
intravitreal ranibizumab treatment. At month 12, 12. Patients with the highest level of baseline
resolution of CNV leakage was also observed in BCVA (≥73 letters) demonstrated the lowest mean
63.8% to 65.7% patients. Moreover, patients in gain in BCVA of 8.2 letters, while patients with
groups 1 and 2 who had intravitreal ranibizumab lowest BCVA at baseline (<45 letters) demon-
demonstrated a mean ± standard deviation (SD) strated the greatest mean BCVA gain of 20.3 let-
decrease in lesion size of 0.31 ± 1.65 mm2 and ters. This is likely due to the ceiling effect in
0.57 ± 1.94 mm2, compared with a mean ± SD patients with good visual acuity as observed in
increase in lesions size of 0.28 ± 2.96 mm2 in other ophthalmology clinical trials. For treatment
patients in group 3. exposure, East-Asians required a median number
In terms of treatment exposure, a relatively of 2 injections compared with 3 injections in
low number of intravitreal ranibizumab injec- Caucasians over 12 months, with 74.3% East-
tions was required in the study to stabilize vision. Asians not requiring reinjections from month 6.
Patients in group 1 received a median number of Patients in the category with the lowest baseline
4.0 intravitreal ranibizumab injections over BCVA had a higher median number of 4 ranibi-
12 months, compared with a median of 2.0 injec- zumab injections, compared with patients in the
tions for group 2. The difference in treatment category with the highest baseline BCVA, which
exposure was attributed to different loading dos- requires only one single injection. A higher median
age regimens of the two groups as required by the number of injections was also required in patients
with larger baseline CNV lesion area. In summary, per patient per year. No serious ocular adverse
the subgroup analysis of RADIANCE demon- events related to anti-VEGF treatment was
strated the efficacy of ranibizumab in treating reported during the study period. Overall, patients
myopic CNV patients with different baseline char- with myopic CNV demonstrated positive visual
acteristics, and recommended the use individual- outcomes up to 48 months, with low incidence of
ized treatment in using ranibizumab as guided by myopic CNV recurrence, and few reinjections
disease activity criteria for treating myopic required to maintain or improve BCVA [17].
CNV. In addition, regardless of the baseline char-
acteristics, ranibizumab resulted in good visual
acuity improvement across all subgroups [16]. 21.3.3 BRILLIANCE Study
After the completion of the RADIANCE
study, a post-RADIANCE retrospective cohort The BRILLIANCE study was a phase III,
study was also conducted to investigate the long- 12-month, randomized, active-controlled, double-
term efficacy and safety in the use of intravitreal masked study, which compared the efficacy and
ranibizumab in the treatment of myopic CNV in safety of intravitreal ranibizumab 0.5 mg, as
East-Asian patients from the RADIANCE study guided by criteria based on visual acuity stabiliza-
[17]. Patients (n = 41) were followed for up to tion or disease activity, with vPDT in patients
48 months in order to evaluate the mean BCVA with myopic CNV (Table 21.1) [18]. The
changes, recurrence of myopic CNV, and ocular BRILLIANCE study design was identical to the
adverse events. During the period, patients with RADIANCE study, except that the BRILLIANCE
myopic CNV could be treated without restriction study involved predominately Asian patients,
on anti-VEGF agents or treatment regimens. while the RADIANCE study involved Asians,
Patients were categorized into two groups, with 7 Caucasians, and other races. A total of 431
(17.1%) patients received additional anti-VEGF patients completed the study, with a majority
injections, and 34 (82.9%) patients did not being female (68.1%) and Chinese (84.0%)
require any further treatment during the observa- patients. Patients were also randomized in a 2:2:1
tion period. Out of the 7 patients who had retreat- ratio to receive intravitreal ranibizumab 0.5 mg
ment, 4 patients were treated with intravitreal guided by visual acuity stabilization (group 1),
ranibizumab, while the other 3 patients were intravitreal ranibizumab 0.5 mg guided by disease
treated with other anti-VEGF agents. A mean activity (group 2), and vPDT (group 3). The base-
number of 5.0 injections were required in patients line demographics were comparable among the 3
who had additional treatment; with a mean dura- study groups. The primary objective of the study
tion of 29.4 months of follow-up in the post- was to demonstrate the superiority of intravitreal
RADIANCE observation period. ranibizumab treatment over vPDT at month 3,
The primary outcome of the study was the while the secondary objective was to show that
long-term efficacy of ranibizumab in terms of ranibizumab retreatment guided by disease activ-
mean BCVA change from baseline to each follow- ity was non-inferior to ranibizumab retreatment
up. During the follow-up, patients had significant guided by visual acuity stabilization criteria at
improvement in mean ± SD visual acuity from month 6.
baseline, with +14.3 ± 11.4 letters at month 12 Regarding visual outcomes of the study, intra-
(P < 0.0001), +10.4 ± 22.3 letters at month 24 vitreal ranibizumab 0.5 mg, irrespective of retreat-
(P = 0.0143), +11.0 ± 22.4 letters at month 30 ment criteria, was superior over vPDT treatment
(P = 0.0134), +12.9 ± 20.9 letters at month 42 based on mean BCVA changes from baseline to
(P = 0.0051) and +16.3 ± 18.7 letters at month 48 month 3, with patients in groups 1, 2, and 3 gained
(P = 0.0034). During the post-RADIANCE obser- 9.5 letters, 9.8 letters, and 4.5 letters at month 3,
vation period, a low rate of myopic CNV recur- respectively (P < 0.001). Ranibizumab retreatment
rences was observed, with recurrences reported in guided by disease activity was also non-inferior to
9.8% of patients, and the overall annual rate of ranibizumab retreatment guided by VA stability
myopic CNV recurrence being 0.06 recurrence based on mean changes in BCVA at month 6, with
a mean gain of 10.4 letters and 10.7 letters, respec- 21.4 Aflibercept Clinical Trial
tively (P < 0.001). At month 12, all three treatment
groups demonstrated improvement in BCVA, with Aflibercept (Eylea, Bayer, Leverkuensen,
patients in groups 1, 2, and 3 gained a mean of 12.0 Germany) is a recombinant protein with activities
letters, 13.1 letters, and 10.3 letters, respectively. inhibiting VEGF-A, VEGF-B, and placental
However, patients in group 3 who received vPDT/ growth factor, and has also been approved for the
intravitreal ranibizumab had reduced visual gains treatment of myopic CNV in various countries.
compared with patients who received intravitreal
ranibizumab from baseline.
For anatomical outcomes, reduced proportions 21.4.1 MYRROR Study
of patients with subretinal fluid, intraretinal edema,
and intraretinal cysts were observed in all treatment The MYRROR study was an international, phase
groups at month 12. Reductions in mean CNV leak- III, multicenter, randomized controlled trial eval-
age and lesion areas were also observed. Patients in uating the efficacy of intravitreal aflibercept
groups 1 and 2 who received intravitreal ranibi- 2.0 mg compared with sham treatment in patients
zumab demonstrated a decrease in mean CRT up to with myopic CNV, with the primary endpoint
month 3, with stabilization of CRT up to month 12. measured at week 24 (Table 21.1) [19]. Patients
Patients in group 3 demonstrated a decrease in aged ≥18 years with high myopia (defined as less
mean CRT, with stabilization of CRT up to month 3. than or equal to −6.0 diopters or axial length of
With the introduction of intravitreal ranibizumab to ≥26.5 mm), with active subfoveal or juxtafoveal
patients in the vPDT group at month 3, a further myopic CNV and BCVA of 35 to 73 ETDRS let-
reduction in mean CRT was observed. For treat- ters were included in the study. Patients (n = 121)
ment exposure, a median number of 4.0 and 3.0 were randomized in a 3:1 ratio into intravitreal
intravitreal ranibizumab injections were required in aflibercept group or sham/intravitreal aflibercept
groups 1 and 2, respectively, up to month 12, while treatment group. Patients in the intravitreal
patients in group 3 received a median number of 3.2 aflibercept group received one single injection of
intravitreal ranibizumab injections from months 3 2.0 mg aflibercept at baseline, followed by addi-
to 12. Regarding the safety of treatment, low fre- tional PRN injections in case of CNV persistence
quencies of adverse events and serious adverse or recurrence up to week 44 if the retreatment
events (SAE) were recorded, with 3 patients pre- criteria were met. The retreatment criteria were
senting with ocular SAEs and 24 patients with non- defined as reduction in ≥5 letters from the previ-
ocular SAEs, which were considered to be unrelated ous visit, increase in CRT by >50 μm, presence of
to the study drug. cystic retinal changes, subretinal fluid, or pigment
Overall, the results from the BRILLIANCE epithelial detachment with CNV or hemorrhage.
study were comparable to that of the RADIANCE Patients in the sham group received one sham
study, showing the superiority of intravitreal injection at baseline, followed by sham injections
ranibizumab, regardless of retreatment criteria, every 4 weeks up to week 20. From week 24
over vPDT up to 3 months in the treatment of onwards, patients in the sham group were eligible
patients with myopic CNV. to receive intravitreal aflibercept 2.0 mg injection
In summary, the phase II REPAIR study, the on a PRN basis in case of CNV persistence or
phase III RADIANCE, and BRILLIANCE stud- recurrence according to the retreatment criteria.
ies have shown consistent results, demonstrating The primary outcome of the study was the mean
that intravitreal ranibizumab is efficacious and change in BCVA from baseline to week 24, and the
safe in stabilizing and improving vision in secondary outcome was the proportion of patients in
patients with myopic CNV. It has been proposed the two treatment groups who gained ≥15 letters at
that the slight differences in BCVA improvement week 24. For treatment exposure, the median num-
reported in different studies could be due to the ber of aflibercept injections in the aflibercept group
differences in baseline demographics of different was 2.0 in the first 12 weeks and 0 in the remaining
patient groups [18]. study period. The median number of aflibercept
injections in the sham/intravitreal aflibercept group the aflibercept group, and an increase in 0.31 DA in
was 3.0 during the study period, with a median num- patients in the sham group (P < 0.0001). At week
ber of 2.0 and 1.0 injections in the third and fourth 48, patients in both groups demonstrated a decrease
study quarters, respectively. The study demonstrated in mean CNV size, with a reduced difference
that a minimal number of reinjections were required between the two groups. Patients in the aflibercept
following initial intravitreal aflibercept injection to group also showed a significant reduction in the
stop the myopic CNV disease process. area of FA leakage at week 24, with a mean decrease
Intravitreal aflibercept demonstrated superior of 0.48 DA, compared with an increase of 0.19 DA
efficacy compared with sham treatment in both pri- in patients in the sham group (P < 0.0001). At week
mary and secondary endpoints of the study. For the 48, a greater proportion of patients in the aflibercept
primary endpoint, the mean BCVA change from group had resolution of CNV leakage compared to
baseline to week 24 was a gain of 12.1 letters and a the sham/aflibercept group.
loss of 2.0 letters in the intravitreal aflibercept and In terms of quality of life indicator, patients in the
sham groups, respectively (P < 0.0001). For the intravitreal aflibercept group demonstrated a better
confirmatory key secondary endpoint, a greater pro- vision-related quality of life than patients in the
portion of patients in the intravitreal aflibercept sham/aflibercept group, as measured by the National
group (38.9%) gained ≥15 letters at week 24 com- Eye Institute Visual Function Questionnaire 25 at
pared to the sham group (9.7%) (P = 0.0001). For weeks 24 and 48. Regarding the safety of treatment,
other visual acuity endpoints, a greater proportion intravitreal aflibercept was generally well tolerated,
of patients who were treated with aflibercept gained and no new safety concerns emerged during the
≥10 letters and ≥5 letters compared to patients in study. Patients in the aflibercept and sham/intravit-
the sham group at both 24 and 48 weeks. The real aflibercept groups had a similar incidence of
improvement in BCVA in the intravitreal aflibercept ocular treatment–emergent adverse events in the
group was sustained to week 48, with a mean gain study period, with most being mild intensity events.
of 13.5 letters and 50% gain in ≥15 letters. Patients In summary, the MYRROR study demon-
in the sham/intravitreal aflibercept group received strated the efficacy of intravitreal aflibercept
intravitreal aflibercept at week 24 with a PRN regi- 2.0 mg compared with sham/intravitreal afliber-
men thereafter, and demonstrated a mean improve- cept in the treatment of patients with myopic
ment of 5.9 letters from week 24 to week 48. Since CNV, producing significant visual and anatomi-
the visual outcomes in the first 24 weeks of afliber- cal outcomes over a period of 48 weeks, with a
cept treatment were significantly better in patients limited number of injections required in the early
in the aflibercept group compared with patients in course of treatment [19].
the sham/intravitreal aflibercept group, the study
recommended early evaluation of visual change in
patients with pathologic myopia for detection of 21.5 Conbercept Clinical Trial
myopic CNV so that anti-VEGF treatment can be
initiated as soon as possible. Conbercept is a fusion protein that inhibits the
Regarding the anatomical outcomes, patients activity of VEGF-A, VEGF-B, and placental
who received intravitreal aflibercept injections had growth factor, and is currently approved in China
a significantly higher mean reduction in CRT from for the treatment of myopic CNV.
baseline to week 24 compared with patients in the
sham group (−80.7 μm vs. −13.9 μm) (P < 0.0001).
However, the difference between the two groups 21.5.1 SHINY Study
was not statistically significant at week 48. A differ-
ence in mean CNV size was also observed between The SHINY study was a 9-month, multicentered,
the two groups at week 24 compared with baseline, double-masked, sham-controlled, randomized
with a decrease in 0.24 disc areas (DA) in patients in controlled trial, which compared the efficacy and
safety of intravitreal conbercept 0.5 mg with 21.6 Conclusions

sham treatment in patients with myopic CNV
(Table 21.1) [20]. Patients (n = 176) were ran- The above major clinical trials demonstrated the
domized in a 3:1 ratio to receive intravitreal con- efficacy of vPDT and anti-VEGF therapy in
bercept 0.5 mg or sham injection. Patients in the improving vision in patients with myopic CNV,
intravitreal conbercept group received monthly with anti-VEGF therapy being superior over
intravitreal injections of conbercept 0.5 mg for vPDT regardless of the retreatment criteria
3 months, followed by conbercept injections with adopted. The phase 3 randomized controlled trials
a PRN regimen up to month 9. Patients in the including RADIANCE, BRILLIACE, MYRROR,
sham treatment group received monthly sham and SHINY studies have all demonstrated the
injections for 3 months, and at 3 months, they efficacy and safety of ranibizumab, aflibercept,
were eligible to receive intravitreal conbercept and conbercept in improving visual and anatomi-
0.5 mg on a PRN basis. cal outcomes in patients with myopic CNV.
The baseline demographics were similar As early diagnosis and initiation of treatment
between the two study groups. The primary objec- for myopic CNV are crucial, patients with high
tive of the study was to assess the mean change in myopia presenting with blurred vision, central
BCVA from baseline to month 3. A final assess- scotoma, and/or metamorphopsia, or signs of
ment of BCVA was then conducted at month 9. At CNV should be promptly referred to an ophthal-
month 3, patients in the conbercept group demon- mologist or a retinal specialist with expertise in
strated a significant mean BCVA gain of 12.00 let- treating myopic CNV [21]. Licensed anti-VEGF
ters (P < 0.001), while patients who received sham agents are currently the first-line therapy for
injections showed a mean gain of 0.56 letters treating myopic CNV, and the use of initial anti-
(P = 0.656). At month 9, patients in both groups VEGF injections followed by a PRN regimen is
demonstrated substantial improvement in BCVA, supported by results from the above mentioned
with patients in the conbercept and sham/intravit- pivotal clinical trials. In all these RCTs in the use
real conbercept groups achieved a mean gain of of anti-VEGF therapy for myopic CNV, patients
13.28 letters and 11.26 letters, respectively. demonstrated improvement in visual outcomes
For anatomical outcomes, at month 3, a mean regardless of the time of initiation of treatment.
reduction of 61.97 μm in CRT was observed in However, compared with prompt treatment, any
patients who had intravitreal conbercept 0.5 mg delay in applying anti-VEGF therapy resulted in
(P < 0.001), while the change in mean CRT was not reduced visual gains. Hence, early treatment dur-
statistically significant in patients who had sham ing the active stage of the disease is critical in
injections. However, significant reductions in mean limiting the disease activity. Following initial
CRT were observed in both study groups at month treatment, patients should be monitored regularly
9, with patients in the conbercept and sham/intra- for disease activity or visual changes, and receive
vitreal conbercept groups demonstrated a mean prompt retreatment if required. The above RCTs
reduction of 73.7 and 55.9 μm, respectively also demonstrated that the majority of patients
(P < 0.001). with myopic CNV required much fewer intravit-
Regarding treatment exposure, patients who real anti-VEGF injections to prevent disease pro-
had conbercept injections from baseline received gression, compared with CNV due to neovascular
a mean of 4.78 conbercept injections over age-related macular degeneration. If intravitreal
9 months, with 27% of patients required no fur- anti-VEGF therapy is contraindicated in patients,
ther injections following the initial 3 loading vPDT can be still adopted in the treatment of
doses. Patients in the sham/intravitreal conber- myopic CNV, though the visual gains are likely
cept group required a mean number of 3.49 con- to be substantially lower than that of using anti-
bercept injections from months 3 to 9. VEGF therapy [11, 21].
Key Learning Points choroidal neovascularization due to pathologic myo-

pia. Prog Retin Eye Res. 2018 Mar;63:92–106.
• Choroidal neovascularization is a major com- 8. Tong JP, Chan WM, Liu DT, Lai TY, Choy KW,
plication in patients with pathologic myopia Pang CP, Lam DS. Aqueous humor levels of vascu-
causing central vision loss. lar endothelial growth factor and pigment epithelium
• The long-term visual outcomes of myopic derived factor in polypoidal choroidal vasculopathy
and choroidal neovascularization. Am J Ophthalmol.
CNV without treatment are extremely poor. 2006;141:456–62.
• The VIP study demonstrated that vPDT was 9. Verteporfin in Photodynamic Therapy Study Group.
effective in improving vision in patients with Photodynamic therapy of subfoveal choroidal neo-
myopic CNV in the short-term, but vPDT was vascularization in pathologic myopia with vertepor-
fin. 1-year results of a randomized clinical trial – VIP
unable to result in significant visual improve- report no. 1. Ophthalmology. 2001;108:841–52.
ment at 24 months. 10. Blinder KJ, Blumenkranz MS, Bressler NM,
• Anti-VEGF therapy is currently the standard Bressler SB, Donato G, Lewis H, et al. Verteporfin
of care treatment for myopic CNV. therapy of subfoveal choroidal neovascularization
in pathologic myopia: 2-year results of a random-
• The phase 2 REPAIR study and phase 3 large- ized clinical trial – VIP report no. 3. Ophthalmology.
scaled randomized controlled trails including 2003;110:667–73.
RADIANCE, BRILLIANCE, MYRROR, and 11. Cheung CMG, Arnold JJ, Holz FG, Park KH, Lai
SHINY studies have demonstrated the favor- TYY, Larsen M, et al. Myopic choroidal neovascular-
ization: review, guidance, and consensus statement on
able efficacy and good safety of anti-VEGF management. Ophthalmology. 2017;124:1690–711.
agents including ranibizumab, aflibercept, and 12. Tufail A, Patel PJ, Sivaprasad S, Amoaku W, Brown-
conbercept in the treatment of myopic CNV. ing AC, Cole M, et al. Ranibizumab for the treatment
• Early diagnosis and prompt treatment of myo- of choroidal neovascularisation secondary to patho-
logical myopia: interim analysis of the REPAIR study.
pic CNV is crucial to prevent severe irrevers- Eye (Lond). 2013;27:709–15.
ible vision loss. 13. Tufail A, Narendran N, Patel PJ, Sivaprasad S, Amoaku
• A low number of anti-VEGF injections are W, Browning AC, et al. Ranibizumab in myopic choroi-
required in the treatment of myopic CNV. dal neovascularization: the 12-month results from the
REPAIR study. Ophthalmology. 2013;120:1944–5.e1.
14. Amoaku WM, Gale RP, Lotery AJ, Menon G,
Sivaprasad S, Petrillo J, et al. Treatment satisfaction
References and Well-being in patients with myopic choroidal
neovascularization treated with ranibizumab in the
1. Wong TY, Ferreira A, Hughes R, Carter G, Mitchell REPAIR study. PLoS One. 2015;10(6):e0128403.
P. Epidemiology and disease burden of pathologic 15. Wolf S, Balciuniene VJ, Laganovska G, Menchini
myopia and myopic choroidal neovascularization: an U, Ohno-Matsui K, Sharma T, et al. RADIANCE:
evidence-based systematic review. Am J Ophthalmol. a randomized controlled study of ranibizumab in
2014;157:9–25.e12. patients with choroidal neovascularization sec-
2. Hotchkiss ML, Fine SL. Pathologic myopia and ondary to pathologic myopia. Ophthalmology.
choroidal neovascularization. Am J Ophthalmol. 2014;121:682–92.e2.
1981;91:177–83. 16. Holz FG, Tufail A, Leveziel N, Lai TY, Lanzetta
3. Ohno-Matsui K, Yoshida T, Futagami S, Yasuzumi P, Wong TY, et al. Ranibizumab in myopic cho-
K, Shimada N, Kojima A, et al. Patchy atrophy and roidal neovascularization: a subgroup analysis
lacquer cracks predispose to the development of cho- by ethnicity, age, and ocular characteristics in
roidal neovascularisation in pathological myopia. Br J RADIANCE. Ophthalmologica. 2016;236:19–28.
Ophthalmol. 2003;87:570–3. 17. Tan NW, Ohno-Matsui K, Koh HJ, Nagai Y, Pedros
4. Cohen SY, Laroche A, Leguen Y, Soubrane G, Coscas M, Freitas RL, et al. Long-term outcomes of ranibi-
GJ. Etiology of choroidal neovascularization in young zumab treatment of myopic choroidal neovascular-
patients. Ophthalmology. 1996;103:1241–4. ization in east-Asian patients from the RADIANCE
5. Neelam K, Cheung CM, Ohno-Matsui K, Lai TY, Wong study. Retina. 2018;38:2228–38.
TY. Choroidal neovascularization in pathological myo- 18. Chen Y, Sharma T, Li X, Song Y, Chang Q, Lin R,
pia. Prog Retin Eye Res. 2012 Sep;31:495–525. et al. Ranibizumab versus verteporfin photodynamic
6. Yoshida T, Ohno-Matsui K, Yasuzumi K, Kojima A, therapy in Asian patients with myopic choroidal neo-
Shimada N, Futagami S, et al. Myopic choroidal neovascularization: BRILLIANCE, a 12-month, random-
vascularization: a 10-year follow-up. Ophthalmology. ized, Double-Masked Study. Retina. 2018; https://doi.
2003;110:1297–305. org/10.1097/IAE.0000000000002292.
7. Ohno-Matsui K, Ikuno Y, Lai TYY, Gemmy Cheung 19. Ikuno Y, Ohno-Matsui K, Wong TY, Korobelnik JF,
CM. Diagnosis and treatment guideline for myopic Vitti R, Li T, et al. Intravitreal Aflibercept injection
in patients with myopic choroidal neovascular- Presented in angiogenesis, exudation and degenera-
ization: the MYRROR Study. Ophthalmology. tion meeting 2016, Miami, FL.
2015;122:1220–7. 21. Wong TY, Ohno-Matsui K, Leveziel N, Holz FG, Lai
20. Chen Y. Efficacy and safety of conbercept ophthalmic TY, Yu HG, et al. Myopic choroidal neovascularisa-
injection in treatment of choroidal neovascularization: current concepts and update on clinical manage-
tion secondary to pathological myopia (Shiny trial). ment. Br J Ophthalmol. 2015;99:289–96.
Christine P. S. Ho is a fifth-year medical student cur-

rently studying in the Faculty of Medicine at the University Timothy Y. Y. Lai is currently a Clinical Professor
of Hong Kong. She has developed a strong interest in oph- (Honorary) of the Department of Ophthalmology &
thalmology and has previously published review articles Visual Sciences, The Chinese University of Hong Kong;
in macular diseases including polypoidal choroidal vascu- and Director of the 2010 Retina and Macula Centre, Hong
lopathy and choroidal neovascularization due to uncom- Kong. His clinical and research interests lie in the medical
mon causes. She has been awarded various scholarships and surgical management of retinal diseases, particularly
including the HKU Entrance Scholarship and the Dean’s polypoidal choroidal vasculopathy, myopic maculopathy,
Scholarship. choroidal neovascularization, central serous chorioreti-
nopathy, visual electrophysiology, and genetics of retinal
diseases and uveitis. Prof. Lai has published over 230
papers in international peer-reviewed journals and has a
Scopus h-index of 46. His academic achievements are
well-recognized internationally and has received many
awards including the Nakajima Award of the Asia-Pacific
Academy of Ophthalmology (2008), Achievement Award
of the American Academy of Ophthalmology (2010),
Achievement Award of the Asia-Pacific Academy of
Ophthalmology (2011), Victor Yong Lecture of the NHG
Eye Institute in Singapore (2013), Senior Achievement
Award of the American Academy of Ophthalmology
(2017), Constable Lecture Award of the Asia-Pacific
Vitreo-Retina Society (2017), and Senior Achievement
Award of the Asia-Pacific Academy of Ophthalmology
(2019).
Clinical Trials Related to Non-AMD
Choroidal Neovascularization 22
Tariq Alzahem, Nayef Alswaina,
and Marwan A. Abouammoh
22.1 Introduction The objective of this chapter is to summarize

the key clinical trials that have made the most
The condition that most commonly leads to the noticeable transformations in the management of
development of choroidal neovascularization patients affected with non-AMD causes of
(CNV) in industrialized nations is age-related CNV. We will focus on clinical trials addressing
macular degeneration (AMD) [1–3]. However, CNV associated with POHS, angioid streaks,
especially in younger patients, there are other uveitis, central serous chorioretinopathy (CSCR),
causes for CNVs. These include high myopia, pattern dystrophy, and idiopathic CNV’s.
presumed ocular histoplasmosis syndrome
(POHS), hereditary, traumatic, and inflammatory
ocular disorders [4]. Polypoidal choroidal vascu- 22.2 Methodology
lopathy (PCV) is another entity in which type 1
CNV has a characteristic abnormal branching We conducted a literature review on all clinical tri-
network of vessels and aneurysmal dilations [5]. als focusing on non-AMD, non-myopia, and non-
Thus, some believe that PCV is a variant of neo- PCV related choroidal neovascularizations. We
vascular AMD. Idiopathic CNV is a term used to searched MEDLINE and EMBASE for all pub-
describe CNV’s with no specific cause. lished phase III clinical trials written in the English
Before the introduction of intravitreal anti- language. Search terms used included: “Choroidal
vascular endothelial growth factor (VEGF) neovascularization” OR “Choroidal neovascular
agents to treat CNV, laser photocoagulation, membrane” OR “CNV.” No limits were placed on
verteporfin with photodynamic therapy (PDT), the year of publication. We excluded animal stud-
and surgical intervention were performed to treat ies. The search results, using the previously men-
CNV. In the last few years, revolutionary tioned terms and filters, returned 653 articles
advancements in patient outcomes have been (MEDLINE: 604, EMBASE: 49).
observed after the institution of anti-VEGF In the initial screening, abstracts were
therapy. reviewed by a single reviewer for obvious exclu-
sion criteria (AMD, myopia, and PCV trials).
T. Alzahem · M. A. Abouammoh (*) Then, the full publications were retrieved for all
Department of Ophthalmology, College of Medicine, clinical trials recognized at the initial screening.
King Saud University, Riyadh, Saudi Arabia They were then rescreened by two investigators.
N. Alswaina This resulted in 48 articles being used to develop
Department of Ophthalmology, College of Medicine, this chapter. Then, the articles were categorized
Qassim University, Buraydah, Saudi Arabia

298 T. Alzahem et al.
according to the disease entity. The clinical trials hemorrhagic disciform degeneration of the mac-
categories are CNV’s associated with ocular his- ula as measured by visual acuity and the activity
toplasmosis syndrome (17 articles), idiopathic of the chorioretinal scar in patients with POHS. A
CNV’s (9 articles), angioid streaks (4 articles), total of 29 patients were randomized to either
inflammatory CNV’s (6 articles), CSCR (2 arti- treatment (15 patients) or control (14 patients).
cles), and pattern dystrophy (1 article). Retinal All patients were followed for a minimum of 1
angiomatous proliferation (9 articles) are covered year. In the treatment group, the final visual acu-
under AMD trials. ity improved 1 Snellen line in 33.3%, remained
the same in 6.66%, and worsened in 60% of
patients. In the control group, the final visual acu-
22.3 Clinical Trials ity improved 1 Snellen line in 42.8% and wors-
ened in 57.1% of patients. The intervention did
We will detail the clinical trials for non-AMD not significantly change the activity of the lesions
causes of choroidal neovascular membranes in compared to the control group. Thus, short-term
the following order: improvement was observed in some patients, but
the Xenon-arc photocoagulation was not effec-
22.3.1 Presumed Ocular Histoplasmosis Syn- tive in the long-term management of CNV in
drome (POHS). patients with POHS [10].
22.3.2 Idiopathic choroidal neovascularization.
22.3.3 Angioid streaks. Macular Photocoagulation Study in POHS
22.3.4 Inflammatory choroidal neovasculariza- [11–17]
tion. The Macular Photocoagulation Study (MPS)
22.3.5 Central serous chorioretinopathy (CSCR). Group conducted a number of randomized, con-
22.3.6 Pattern dystrophy. trolled clinical trials to assess the laser treatment
of CNV in patients with AMD, POHS, and idio-
pathic CNV’s. The effect of Argon and Krypton
22.3.1 Presumed Ocular laser photocoagulation in the treatment of extra-
Histoplasmosis Syndrome foveal and juxtafoveal CNV secondary to POHS
was reported in the MPS. In all MPS trials, the
Various historical treatment modalities were pro- primary outcome was the change in best-
posed for patients with POHS. They include corrected visual acuity from baseline. The clini-
stress management, discontinuing aspirin, avoid- cal trials of Argon laser followed by Krypton
ing valsalva maneuvers, use of corticosteroids, laser are described below.
and others [6–8]. Those interventions were not
proven effective with no known treatment for Argon Laser Photocoagulation for Extrafoveal
inactive lesions. In the presence of active chorio- CNV’s in POHS [11–13]
retinal lesions and subfoveal CNV, corticoste- In 1983, the MPS Group published the first trial
roids also have been suggested [9]. However, addressing the effect of argon laser photocoagu-
since the advent of laser photocoagulation, PDT, lation in preventing severe visual loss in eyes
and anti-VEGF agents, the previous modalities with extrafoveal CNV’s and a clinical evidence
are generally of historic interest. of POHS [11]. Severe visual loss was defined as
loss of 6 or more lines from the baseline. A
22.3.1.1 Laser Treatment baseline best-corrected visual acuity of 20/100 or
better was required in the study eye. Eligible
Light Coagulation in Presumed patients (no. of eyes = 245) were randomly
Histoplasmic Choroiditis [10] assigned to either blue-green laser photocoagula-
The objective of this trial was to assess the effect tion (n = 124) or observation only (n = 121).
of light coagulation (Xenon-arc) on macular Supplementary treatment was performed when
22 Clinical Trials Related to Non-AMD Choroidal Neovascularization 299
new vessels recurred 200 μm away from the cen- 5 years, the relative risk of losing 6 or more lines
ter of the FAZ. The data were based on 231 eyes of visual acuity among untreated vs. treated eyes
(117 treated vs. 114 untreated) followed for at was 3.6 (P < 0.0001). Similar to the 3-year results,
least 6 months with a median follow-up for 68% of the treated group had a visual acuity of
18 months. In the most recent follow-up, 39 out 20/40. The percentage of having a 20/40 visual
of 114 (34.2%) untreated eyes lost 6 lines com- acuity or better in the untreated group increased
pared to 11 out of 117 (9.4%) treated eyes. At the from 38% at 3 years to 43% at 5 years. In the
same visit, 36% of untreated eyes had visual acu- laser-treated group, recurrent CNV had been
ity worse than 20/100 compared to 12.8% in the observed in 26% of eyes by 5 years after the start
treated group. Two years after starting the study, of the trial. In conclusion, the 5 years’ follow-up
half of the untreated eyes and 22% of the treated results emphasized the superiority of argon laser
eyes had lost 6 or more lines of visual acuity. The photocoagulation in the treatment of well-defined
risk of losing 6 lines was estimated to be 2.3 extrafoveal CNV’s secondary to POHS compared
times greater for untreated eyes compared to to no treatment [13].
treated eyes. The most common complications of
laser photocoagulation were hemorrhage and Krypton Laser Photocoagulation
internal limiting membrane wrinkles. This trial for Juxtafoveal CNV’s in POHS [14–16]
has demonstrated that argon laser photocoagula- Extrafoveal CNV’s represent 46.2% of CNV’s
tion of CNV related to POHS is superior to no occurring in patients with POHS. The remaining
treatment [11]. Additionally, patients’ recruit- CNV’s, i.e. 53.7%, are extending under the foveal
ment has been terminated in 1983 because of the avascular zone [18]. Those patients were not eli-
short-term treatment benefit described. However, gible for the previously mentioned trial because
follow-up of all patients continued to establish of proximity of the neovascular lesion to the
the long-term outcomes (described below). FAZ. Argon laser was not used on those lesions
The 3 and 5 years’ results of argon laser treat- because of the theoretical harm it would induce
ment were reported in 1986 and 1991, respec- in the area within 200 μm of the center of the
tively [12, 13]. At the end of summer, 1985, a FAZ [19]. Therefore, a second randomized clini-
minimum of three or more years of planned fol- cal trial using Krypton laser was initiated [14].
low-up examinations had been completed for 203 In 1987, the results of Krypton laser photoco-
(77%) of 262 eyes enrolled from the commence- agulation for juxtafoveal CNV’s in patients with
ment date of the study. The data were based on POHS were published [14]. This trial differs
194 eyes (104 treated vs. 90 untreated). At the from the first one described above in 3 points: the
3-year visit, 47.8% of untreated eyes had a visual location of CNV’s, the wavelength of the laser
acuity worse than 20/100 compared to 14.4% in used for photocoagulation, and that treatment
the treated group. In addition, 47.8% of the was permitted up to the center of the FAZ. The
untreated eyes and 8.6% of the treated eyes had aim of this study is to determine whether krypton
lost 6 or more lines of visual acuity. The relative laser treatment would be beneficial in halting the
risk of experiencing severe loss of vision in the progressive visual loss in eyes with POHS and
untreated eyes after 3 years was 5.5 (95% CI: 3.9 juxtafoveal CNV’s. Eligible patients should have
to 10.8). Two-thirds of the treated eyes had a an angiographic evidence of CNV located
visual acuity of 20/40 or better in comparison between 1 and 199 μm from the center of the
with approximately a third of the untreated group FAZ, or between 200 and 2500 μm from the cen-
[12]. At 5 years, the data of 232 eyes (116 in each ter if associated with an adjacent blood or pig-
group) were analyzed. Visual acuity of 20/100 or ment ring involving the FAZ. The best-corrected
worse was observed in 45% of untreated and 17% visual acuity should be equal or better than
of the treated eyes. A loss of 6 or more lines of 20/400 in the study eye. A total of 288 eyes were
visual acuity was detected in 44% of the untreated included in the study. 143 and 145 eyes were ran-
eyes compared to 9% of the treated eyes. In domly assigned to treatment and observation
groups, respectively. The data were based on 125 2.5 times more common in patients with systemic
eyes completing the 36-month visit. In both hypertension than in patients without [16].
groups, an almost even percentage of eyes
attained or preserved a visual acuity of 20/20 or Laser Photocoagulation for Peripapillary CNV’s
better (20.3% in the treatment group, 21.3% in and CNV’s Located Nasal to the Fovea in POHS
the observation group). However, 26.2% of the [17]
observation group had a visual acuity worse than There were some concerns regarding the treat-
20/200 compared to 4.7% in the treatment group. ment of two types of CNV’s: peripapillary
Similarly, 15 out of 61 untreated eyes (24.6%) CNV’s, and large CNV’s located nasal to the
lost 6 lines compared to 3 out of 64 treated eyes fovea in which the longest diameter is greater
(4.7%). A worse baseline visual acuity was asso- than 750 μm. Laser treatment of CNV’s in these
ciated with a greater proportion of visual two areas was thought to carry the risk of damag-
improvement. It has been concluded that treating the optic nerve and the papillomacular bun-
ment with Krypton laser photocoagulation was dle, respectively. Thus, the specific results of
associated with better visual acuity outcomes laser treatment of these CNV’s in the eyes
regardless of baseline variables [14]. included in the previous MPS trials were pub-
The 5 years’ results were reported in 1994. A lished in a separate report [17]. The objective was
total of 236 eyes, 118 in each group, eyes were to determine whether laser treatment of these
evaluated [15]. After 5 years, the median visual CNV’s was better than observation. The same
acuity was significantly better in the treated group protocol for laser treatment described above was
(Snellen visual acuity; 20/40) compared to in applied in which extrafoveal CNV’s were treated
observation group (Snellen visual acuity; 20/64). by argon laser and juxtafoveal CNV’s were
Almost similar to the earlier data described above, treated by krypton laser.
23 treated eyes (19%) and 24 untreated eyes The etiologies of CNV’s in this report were
(20%) had visual acuity that is equal or better than POHS and idiopathic causes with a total of 113
20/20. On the other hand, 0.8% and 13.6% had a eyes included. The results published were based
visual acuity of 20/400 or worse in the treated and on CNV type and not on etiology. However, 51 out
observation groups, respectively. 10 treated eyes of 58 treated eyes (88%) and 46 out of 55 untreated
(8%) had lost 6 or more lines of visual acuity eyes (84%) were diagnosed to have CNV’s sec-
compared to 33 untreated eyes (28%). Thus, early ondary to POHS. The data were based on 105 eyes
beneficial effects of Krypton laser photocoagula- (54 treated vs 51 untreated) followed for at least
tion continued for at least 5 years in eyes with 3 years. Regardless of the lesion type, treated eyes
POHS-related CNV [15]. had better visual acuity (P = 0.03), with 72% of
The occurrence of persistent and recurrent treated eyes and 56% untreated eyes having a
CNV after initial treatment with laser photocoag- visual acuity of 20/100 or better. Compared to the
ulation has been published in a separate report baseline visual acuity, 11% of the treated eyes vs.
[16]. The eyes assigned to krypton red laser pho- 41% of the untreated eyes lost 6 or more lines
tocoagulation (n = 144) were examined. Eyes in (P < 0.001). In the peripapillary CNV subgroup,
which fluorescein angiograms showing fluores- treated eyes had a slightly better visual acuity with
cein dye leakage taken during the first 6 weeks 64% vs. 60% in the untreated group having a
after initial treatment were classified as having a visual acuity of 20/100 or better. Approximately
persistent CNV. Recurrence of CNV was defined 14% of the treated eyes in this subgroup had lost 6
as the observation of leakage on fluorescein after or more lines of visual acuity compared to 26% of
the 6-week posttreatment visit without previous the untreated eyes. In the nasal CNV subgroup, the
persistence. Persistent CNV’s were documented differences were more marked with 78% treated
in 23% of treated eyes while recurrent CNV’s eyes vs. 50% untreated eyes achieving a visual
were detected among an additional 8%. An acuity of 20/100 or better. Among patients in this
increased frequency of severe visual loss was seen subgroup, 9% of the treated eyes lost 6 or more
in those eyes. Persistent CNV was approximately lines of visual acuity compared with 54% of the
untreated eyes (P < 0.001). The study findings The sizes of the new membranes and the recur-
confirmed beneficial effects of laser photocoagula- rent membranes should be equal or less than 3.5
tion for peripapillary and papillomacular CNV’s. and 4 disc areas, respectively [21].
It is important to note that the beneficial outcome A total of 25 patients (12 in the treatment
was less pronounced in the peripapillary CNV’s group, 13 in the observation group) were evalu-
subgroup [17]. ated. Among the treatment group, there were 4
new CNV’s and 8 recurrent CNV’s. Among the
Argon Green Versus Krypton Red Laser observation group, there were 6 new CNV’s and
Photocoagulation for Extrafoveal Choroidal 7 recurrent CNV’s. The mean visual acuity for
Neovascularization: One-Year Results both groups at baseline was 20/125. The results
in Ocular Histoplasmosis—The Canadian were based on 15 patients (7 in the treatment
Ophthalmology Study Group [20] group and 8 in the observation group) at
The Canadian Ophthalmology Study Group car- 12 months after treatment. The mean visual acu-
ried out a randomized clinical trial to determine ity at 12 months was 20/200 in both groups. The
whether krypton laser is superior to argon laser in results were inconclusive and do not suggest any
the treatment of extrafoveal CNV’s secondary to benefit or harm to the eyes receiving laser
POHS in terms of visual acuity at 1 year. Eligible photocoagulation.
patients should have an angiographic evidence of
an active CNV in which the posterior edge is 200 22.3.1.2 Photodynamic Therapy
to 2500 μm from the center of the FAZ, a clinical (PDT) in POHS
exam suggesting POHS, and best-corrected
visual acuity of 20/200 or better in a previously Verteporfin in Ocular Histoplasmosis
untreated eye. Study [22–24]
A total of 127 eyes were eligible (63 in the The Verteporfin in Ocular Histoplasmosis (VOH)
argon laser group and 64 in the krypton laser study is an open-label, uncontrolled, multicenter
group). At 1 year, 6% of the eyes in the argon clinical trial to evaluate the effect of PDT on sub-
laser group had lost 6 or more lines of visual acu- foveal CNV secondary to POHS. For patients to
ity compared with 14% of eyes in the krypton red be eligible, there should be an angiographic evi-
laser group. The difference was not statistically dence of classic or occult CNV involving the cen-
significant between the two groups. In the argon ter of the FAZ. Best-corrected Snellen visual
laser group, 21% of eyes required more than one acuity at baseline ranged from 20/40 to 20/200.
treatment, compared with 39% of the eyes in the The treatment protocol starts with a 10-minute
krypton red laser group. The Canadian infusion of Verteporfin at a dose of 6 mg/m2 of
Ophthalmology Study Group concluded that body surface area. Then, 15 min after the start of
krypton laser photocoagulation treatment is not infusion, a light application (689 nm) of 600 mW/
superior to argon laser photocoagulation therapy cm2 for 83 s to reach a total of 50 J/cm2 of energy.
for well-defined extrafoveal CNV’s in patients Outcomes to be measured were visual acuity
with POHS [20]. changes from baseline and contrast sensitivity
scores.
Laser Treatment for Subfoveal Neovascular A total of 26 patients were eligible. The mean
Membranes in Ocular Histoplasmosis visual acuity was 20/80 and the mean contrast
Syndrome: Results of a Pilot Randomized sensitivity score was 29 letters. At 12 months, 14
Clinical Trial [21] patients (56%) had gained 1 or more lines of
This was a pilot clinical trial published in 1993 to visual acuity, half of them (7 patients, 28%)
evaluate the effect of laser photocoagulation in gained 3 or more lines. Four patients (16%) lost 1
patients with POHS and subfoveal CNV’s, new or more lines of visual acuity from baseline with
and recurrent. New CNV’s and recurrent previ- two patients (8%) losing 3 or more lines. Seven
ously treated CNV’s were enrolled in the study. patients (28%) had a stable visual acuity. Only
one patient had a visual acuity of 20/200 or worse POHS will be discussed here. Inclusion criteria
at the 12-month visit. Contrast sensitivity included angiographic proof of a CNV beneath
increased by 3 or more letters in 12 patients (48%) the center of the FAZ. In all cases, a small reti-
and decreased by 3 or more lines in 3 patients notomy (about 200 μm) eccentric to the fovea
(12%). Ten patients (40%) had stable contrast was made. Then, two variations to remove or
sensitivity. At the 12-month visit, 43% and 100% manipulate the neovascular complex through the
of patients did not show fluorescein leakage from small retinotomy were performed in POHS
classic CNV and occult CNV, respectively. The patients. The first technique involved CNV
authors concluded that PDT with verteporfin is extraction using an angled subretinal pick or for-
safe and well tolerated in patients with subfoveal ceps. The second technique involved CNV dis-
CNV due to POHS for at least 1 year. connection without removal.
Approximately half of the patients experienced Out of 58 consecutive patients, 20 eyes of 20
some improvement in visual acuity [22]. patients had CNV secondary to POHS. In 85% of
After 2 years, 22 out of the 26 patients (85%) the cases, the CNV had been present from 6 months
were followed [23]. At 24 months, there was an to 1 year. Preoperative visual acuities ranged from
improvement of 1 or more lines from baseline in 20/30 to counting fingers at 2 feet. In eyes where
14 patients, similar to the 12-month results. Eight the subfoveal CNV was removed (n = 16), all but
patients (36%) gained 3 or more lines of visual one of the eyes had preoperative visual acuities of
acuity compared to 7 patients (28%) at the 20/200 or worse. In eyes where the subfoveal CNV
12-month examination. Visual acuity scores were was disconnected without removal (n = 4), the ini-
stable in 4 patients (18%) at the 24-month visit. tial visual acuities were 20/100 or better. In eyes
Similar to the 1-year results visual acuity where CNV was removed, with a mean follow-up
decreased by 1 or more lines from baseline in 4 of 4.8 months, the final visual acuity significantly
patients, with 1 patient having a visual acuity of improved in 6 eyes, worsened in one, and remained
20/200 or worse. Fluorescein angiographic leak- the same in nine eyes. A visual acuity of 20/70 or
age was absent in 17 out of 20 (85%) having clas- better was achieved in 4 eyes. In 37% of eyes, per-
sic CNV. No leakage was observed in the occult sistent or recurrent neovascularization were
CNV cases. The same conclusion was reached at observed in which the final visual acuity was
the 2-year follow-up [23]. negatively affected. In the four eyes where the
Fifteen patients completed the 48-month fol- subfoveal CNV was disconnected without
low-up examination [24]. A mean of 4.4 treat- removal, the mean follow-up duration was
ments was given to the patients during the 3.8 months. All four patients experienced a post-
48 months. The median visual acuity improvement operative significant reduction in visual acuity.
from the baseline was 15 letters. The median con- Complications include subretinal hemorrhage,
trast sensitivity improved by 3 letters compared to rhegmatogenous retinal detachment, and RPE
baseline. Through 48 months of PDT therapy for injury. Surgical management of subfoveal CNV’s
subfoveal CNV’s due to POHS, the visual acuity in patients with POHS showed better results in
and contrast sensitivity were maintained or some of the cases. Recurrence of CNV following
improved in the majority of patients [24]. surgery is an important cause of visual decline and
warrants careful follow-ups [25].
22.3.1.3 Surgical Management
of CNV’s in POHS Submacular Surgery Trials Research
Group: Surgical Removal vs Observation
Management of Subfoveal Choroidal for Subfoveal Choroidal
Neovascularization [25] Neovascularization, Either Associated
This interventional case series studied the use of with the Ocular Histoplasmosis Syndrome
vitreoretinal surgical techniques in subfoveal or Idiopathic [26]
CNV’s secondary to AMD, POHS, and miscella- The Submacular Surgery Trials (SST) Research
neous causes. The results of CNV’s secondary to Group conducted randomized trials of surgical
removal of subfoveal CNV’s secondary to POHS, examination, 38% of operated eyes had 1 or
and idiopathic CNV’s. The aim was to assess the more additional treatments. It was concluded
effectiveness of this intervention on visual acuity. from this trial that submacular surgery was not
Eligible patients had a subfoveal CNV that had a beneficial in eyes that had best-corrected visual
classic component. The best-corrected visual acuity of 20/100 or better. On the other hand, if
acuity scores of eligible patients ranged from the best-corrected visual acuity is worse than
20/50 to 20/800. A retinotomy was made by the 20/100, submacular surgery is considered to
surgeon to provide access to the CNV. The fibro- improve the chances of preserving or improving
vascular complex was disconnected and then visual acuity for at least 2 years [26]. An
removed manually with a micro forceps. The pri- improved vision targeted quality of life was
mary outcome was a success if there was an observed more after submacular surgery com-
improvement or stabilization of visual acuity at pared with observation [27].
the 2-year visit. Stabilization was defined as no
change of vision or a change of no more than 1 22.3.1.4 Anti-Vascular Endothelial
line worse than the baseline at 2 years. Growth Factors Agents
Two hundred twenty-five patients (112 in the in POHS
surgery arm and 113 in the observation arm) were
enrolled in the trial. Out of 225 patients, 192 Treatment of CNV Secondary to Presumed
(85%) patients were having CNV secondary to Ocular Histoplasmosis with Intravitreal
POHS (95 patients in the surgery arm and 97 in Aflibercept 2.0 mg Injection [28]
the observation arm). The Median visual acuity This was an open-label, prospective, randomized,
of study eyes at baseline was 20/100 for both phase II/III clinical trial investigating the safety
arms [26]. and efficacy of aflibercept 2 mg injections for
At the 24-month visit, 200 patients (includ- CNV in patients with POHS. The CNV should be
ing POHS and idiopathic) were evaluated. The less than 1 year in duration, subfoveal, or juxtafo-
median visual acuity was 20/160 in the surgery veal in location, with best-corrected visual acuity
arm compared to 20/250 in the observation arm ranging between 20/25 and 20/400.
(P = 0.07). The median visual acuity changes at A total of 5 patients were included. The mean
the 24-month follow-up were a loss of 1 line in baseline visual acuity was 62, measured in Early
the surgery arm and 2 lines in the observation Treatment Diabetic Retinopathy (ETDRS) let-
arm (P = 0.14). The SST definition of successful ters. The mean baseline central foveal thickness
outcomes at 24 months was achieved in 53% of and volume were 293 μm and 10.1 mm3,
the patients in the surgery arm and 45% in the respectively. Subretinal fluid was observed in all
observation arm (P = 0.26). The mean visual patients by OCT at baseline. All patients
acuity improvement was more prominent when received a total of 7 intravitreal aflibercept
the baseline visual acuity was worse than injections in 12 months. No adverse systemic or
20/100. For this subgroup of patients (n = 92), ocular events were reported during the study
the mean visual acuity in the surgery group was period. One year following the initiation of ther-
20/200 compared to 20/320 in the observation apy, the mean visual acuity improved by 12.4
group. At 2 years, the primary successful out- letters. The mean central subfoveal thickness
come was achieved in 76% in the surgery group and OCT volume reduced by 34.6 μm and
vs. 50% in the observation group. Complications 0.58 mm3. Four of 5 patients had resolved fluid
encountered in the surgery group include cata- on OCT at the 12-month visit. It was concluded
ract (39%) and rhegmatogenous retinal detach- that intravitreal aflibercept 2 mg given monthly
ments (4.5%). The 24-month cumulative for 3 months and then every 8 weeks is safe and
incidence of fluorescein leakage from the CNV effective for the treatment of neovascular POHS
was 58% in the treated eyes. By the 24-month for at least 1 year. [28].
22.3.2 Idiopathic Choroidal 5 years in the untreated group was 2.3. Five years
Neovascularization after entry, 57% of the treated eyes had a visual
acuity of 20/40 or better compared to 41% in the
22.3.2.1 Laser Treatment untreated eyes. Thirty-four percent of laser-treated
eyes had recurrent CNV throughout the 5 years.
Laser Photocoagulation for Idiopathic The argon laser treatment for idiopathic extrafo-
Neovascularization: Results veal CNV’s had a favorable outcome. However,
of a Randomized Clinical Trial [29] the beneficial outcome was less pronounced than
This was the first randomized clinical trial pub- that in the POHS trial [13].
lished by the MPS group addressing the effect of
argon laser photocoagulation in the treatment of Laser Photocoagulation for Idiopathic
extrafoveal CNV’s without an apparent cause, Neovascular Lesions: Results
i.e., idiopathic. The aim of this trial was to deter- of a Randomized Clinical Trial [30]
mine whether laser treatment would be of benefit The aim of this clinical trial was to evaluate
in preventing severe visual loss. whether krypton laser treatment would be of ben-
A total of 67 patients (33 patients in the laser efit in preventing visual acuity loss in eyes with
group and 34 in the observation group) were idiopathic juxtafoveal CNV’s. As mentioned ear-
included. At 1 year, 51 patients were evaluated lier in the POHS study, the posterior edge of the
(26 in the treatment group and 25 in the observa- CNV should be located 1 to 199 μm from the
tion group). At the most recent follow-up, 86.7% center of the FAZ or from 200 to 2500 μm from
vs. 63.6% of the patients in the treatment and the the center of FAZ if there was associated blood or
observation groups were having a visual acuity of blockage of fluorescence extending to the central
20/100 or better, respectively. Sixteen percent of 200-micron circle. There should be no drusen or
patients in the treatment group had lost six or atrophic scars suggestive of AMD or POHS,
more lines of visual acuity compared to 33.3% in respectively.
the observation group. The most common compli- A total of 49 eyes were enrolled at the start of
cation of treatment was the development of hem- the study (24 eyes in the laser group and 25 eyes
orrhage. Although the sample size was not large were observed). Ninety-two percent of the eyes
enough to derive a definitive conclusion, the pat- in the laser group vs. 76% of eyes in the observa-
tern of treatment benefit is similar to that observed tion group had a baseline visual acuity of 20/100
in the POHS patients mentioned above [29]. or better. At the 36-month visit, 40% of eyes in
Fifty-one patients completed the 3-year fol- the laser group vs. 42% of eyes in the observation
low-up (27 in the treatment group and 24 in the group had a visual acuity of 20/40 or better. On
observation group). Approximately 85% of the other hand, 10% of treated eyes had a visual
patients in the laser group had a visual acuity of acuity of 20/200 or worse vs. 40% of untreated
20/100 or better vs. 61% in the observation group. eyes. Regarding the change of visual acuity from
Severe visual loss was seen in 15% of patients in baseline, 10% of treated eyes sustained severe
the laser group compared to 35% of patients in the visual loss compared to 37% of observed eyes.
observation group. If left untreated, the relative Thus large visual acuity losses in eyes affected
risk of losing six or more lines after 3 years is 2.3 with juxtafoveal idiopathic CNV’s were seen less
[12]. Five years after enrollment, 57 patients were frequently after krypton laser treatment at the
evaluated (30 in the treatment group and 27 in the 3-year visit [30].
observation group). The mean visual acuities in Five years following patients’ enrollment (20
eyes assigned to treatment were 20/64 vs. 20/80 in eyes in the treatment group and 18 eyes in the
the observed eyes. The mean lines of visual acuity observation group), the median visual acuity in
lost after 5 years were 3 lines in the treatment eyes that received krypton laser was 20/50 com-
group and 4 lines in the observation group pared to 20/80 in the observed eyes (P = 0.27).
(P = 0.22). Similar to the three-year results, the Twenty over forty vision was achieved by 45% of
relative risk of having a severe visual loss at the treated eyes and 39% of the untreated eyes.
The benefit of laser treatment was less defined in Photodynamic Treatment Versus
the 5-year results. In addition, the number of Photodynamic Treatment Associated
patients was small for reliable estimation of the with Systemic Steroids for Idiopathic
treatment benefit [15]. Choroidal Neovascularization [32]
That was a prospective, open-label, consecutive,
22.3.2.2 Photodynamic Therapy and randomized controlled trial. The purpose was
(PDT) to investigate the efficacy of PDT vs. systemic
steroids with PDT in patients with idiopathic
Photodynamic Therapy with Verteporfin CNV’s. Patients were randomized into two
for Subfoveal Idiopathic Choroidal groups. The first group received PDT according
Neovascularization: One-Year Results to the standard protocol (group 1). The second
from a Prospective Case Series [31] group received intravenous methylprednisolone 1
That was a prospective, open-label, consecutive, gram daily for 3 days, followed by oral predni-
interventional case series that investigated the sone 1 mg/kg daily, and then tapered thereafter
outcome of PDT with verteporfin on eyes with (median duration was 5 months), in addition to
idiopathic CNV involving the subfoveal zone in PDT (group 2). The median time between the
Chinese patients. The patients included were start of steroids and PDT was 1.5 months.
aged 50 years or younger. The location of the A total of 20 patients were included (10
CNV or any associated blood or blocked fluores- patients in each group). The median follow-up for
cence should involve the foveal center. The best- both groups was 23 months. The baseline mean
corrected visual acuity was required to be 20/200 visual acuity was 0.34 and 0.26 in group 1 and 2,
or better. There should be evidence of leakage respectively (decimal visual acuity). The mean
within the lesion on fluorescein angiography. numbers of PDT treatments were 2.3 for group 1
A total of 17 patients received PDT with verte- and 1.5 for group 2. At the last follow-up, the
porfin according to the standardized protocol mean best-corrected visual acuity was 0.31 in
described above. The patients were followed-up group 1 compared to 0.41 in group 2. At the last
every 3 months for 1 year. The primary outcome follow-up, 40% of group 1 vs. 90% of group 2
was improved (2 or more lines), stable (within 1 patients had stable/improved visual acuity. With
line), or decreased (2 or more lines) visual acuity 1 regard to the CNV size at the last follow-up, 40%
year after enrollment. The secondary outcomes of group 1 vs. 90% of group 2 patients had stable/
included a change in the mean and median best- reduced lesions. It was concluded from this small
corrected visual acuity following treatment at 1 trial that PDT preceded by systemic steroids was
year. All cases had predominantly classic subfoveal better than PDT alone in the visual acuity out-
CNV. The median baseline best-corrected visual come, the number of PDT treatments, and the size
acuity was 20/70 with a range of 20/50 to 20/200. of the scar in patients with idiopathic CNV’s [32].
At the end of 1 year, 10 eyes (58%) showed
improvement of visual acuity, 6 eyes (35%) were Small Laser Spot Versus Standard Laser
stable, and one eye (6%) demonstrated worsening. Spot Photodynamic Therapy for Idiopathic
The median visual acuity after treatment was 20/50 Choroidal Neovascularization:
with a mean of 2.3 lines of improvement. Fourteen A Randomized Controlled Study [33]
CNV’s (82%) converted into fibrotic scars without This was a randomized controlled trial designed
any leakage in fluorescein angiography. The to compare PDT with small laser spot (spot diam-
remaining CNV’s (18%) showed mild leakage. A eter size is equal to CNV’s greatest linear dimen-
mean of 1.8 PDT treatments was performed per sion (GLD) + 500 μm) and PDT with standard
eye within the study period. The retreatments, if laser spot (CNV’s GLD + 1000 μm) in terms of
required, were given 3 months after the initial treat- visual outcome and RPE damage in patients with
ment. It was concluded that PDT with verteporfin idiopathic CNV.
demonstrated a beneficial outcome in the treatment A total of 52 patients were randomly assigned
of idiopathic subfoveal CNV’s[31]. to a study group (27 patients, small laser spot)
and a control group (25 patients, standard laser through the use of 810-nm infrared diode laser
spot). Approximately half of the patients com- with subsequent closure of the vascular complex.
pleted the 1-year follow-up. All patients received Patients with idiopathic subfoveal CNV with
intravenous infusion of verteporfin (6 mg/m2 of visual acuity better than 10/200 in previously
body surface area) over 10 minutes. Then, a 689- untreated eyes were included. Laser spot sizes
nm laser source (spot size is equal to and power settings ranged between 1.2–3 mm in
GLD + 500 μm in the study group vs. diameter and 300–600 mW, respectively.
GLD + 1000 μm in the control group) was used A total of 21 patients were treated. Twelve
to deliver 50 J/cm2 over 83 s. The mean baseline CNV’s were predominantly classic and 9 were
best-corrected visual acuity was 52.5 and 55.2 predominantly occult. The mean follow-up dura-
ETDRS letters in the study and control groups, tion was 5.1 months with a minimum follow-up
respectively. Subfoveal CNV was present in of 3 months. The mean number of treatments
44.4% of eyes in the study group and 40% of eyes given was 1.14. Out of 12 predominantly classic
in the control group. At 6- and 9-months follow- CNV’s, 11 CNV’s (92%) regressed. Eight out of
ups, the mean improvement in the best-corrected 9 predominantly occult CNV’s (89%) showed
visual acuity in the study group (25.5 and 27.5 regression. Persistence of subretinal fluid was
ETDRS letters, respectively) was significantly noted in one predominantly classic CNV and 3
better than that in the control group (14.7 and predominantly occult CNV’s. None of the eyes
15.5 ETDRS letters, respectively). The ETDRS showed recurrence. Overall, 17 out of 21 eyes
scores at 12 months were better in the study (81%) had either a stable or improved visual acu-
group compared to the control group. At 3 and ity. Out of 17 eyes, 11 (52%) had improvement of
6 months of follow-ups, the study group had 1.5 one or more lines of visual acuity and the remain-
and 2.5 quadrants involved compared to 3 and 4 ing 6 eyes were stable. Four eyes (19%) showed
quadrants affected in the control group, respec- worsening visual acuity compared to baseline.
tively. At 12 months, the difference was mini- Three of the 21 eyes (14%) required retreatment
mized to 3 and 4 quadrants affected in the study with TTT. The study findings suggested that TTT
and control groups, respectively. There was no has a potential role in the treatment of idiopathic
statistically significant difference between the subfoveal CNV’s [34].
two groups in the proportion of cases with
decreased or stabilized CNV leakage size and the 22.3.2.4 Anti-Vascular Endothelial
height of subretinal fluids. Thirty-seven percent Growth Factor Agents
of the study group and 32% of the control group
had recurrent CNV during the 12-month follow- Intravitreal Bevacizumab for Subfoveal
up. In summary, a trend toward a better visual Idiopathic Choroidal Neovascularization [35]
improvement and a less RPE alteration was noted This was a prospective, non-comparative, consec-
in the group with small laser spot PDT compared utive, interventional case series aimed to evaluate
to the standard laser spot PDT over 1 year [33]. the short-term outcomes and safety of intravitreal
bevacizumab in eyes with idiopathic subfoveal
22.3.2.3 Transpupillary CNV. Patients included were ≤50 years with an
Thermotherapy (TTT) OCT demonstrating intraretinal edema, subretinal
fluid, or pigment epithelial detachment (PED).
Transpupillary Thermotherapy Thirty-two consecutive patients were eligible.
for Idiopathic Subfoveal Choroidal All patients received a 1.25 mg/0.05 ml of intravit-
Neovascularization [34] real bevacizumab. The injection was repeated at a
This was a prospective, noncontrolled case series 4-week interval if OCT was still showing intrareti-
of patients with idiopathic subfoveal CNV treated nal edema, subretinal fluid, and/or PED. The
with transpupillary thermotherapy (TTT). It is a median best-corrected visual acuity at baseline was
treatment modality in which the choroid and the 20/200. The means of baseline OCT central macu-
overlying retinal pigment epithelium are heated lar thickness and volume were 314.37 μm and
6.9 μL, respectively. The mean duration of follow- Efficacy and Safety of Ranibizumab
up was 4.2 months with a minimum of 3 months of for the Treatment of Choroidal
follow-up. At 12 weeks, 19 eyes (59%) improved 3 Neovascularization Due to Uncommon
lines or more of visual acuity, 11 eyes (34%) were Cause: Twelve-Month Results
stable, and 2 eyes (6%) lost 3 or more lines. The of the MINERVA Study [37] (Please Refer
median posttreatment best-corrected visual acuity to the Previous Chapter for More Details)
was 20/50 with a mean improvement of 4.9 lines. The MINERVA study was a phase III, prospec-
The means of central macular thickness and macu- tive, double-masked, controlled, and randomized
lar volume at 12 weeks were 236.84 μm and 6.7 μL, clinical trial. The objective was to assess the
respectively. The mean number of intravitreal effectiveness and safety of intravitreal ranibi-
injections was 1.7 injections per eye. The short- zumab 0.5 mg/0.05 ml in patients with CNV due
term results suggested that intravitreal bevaci- to uncommon causes (idiopathic, angioid streak,
zumab is safe and effective in patients with post inflammatory, central serous chorioretinopa-
subfoveal idiopathic CNV [35]. thy, and miscellaneous) over 1 year.
The study enrolled 178 patients who were
Intravitreal Bevacizumab for Treatment randomized 2:1 to take either intravitreal
of Subfoveal Idiopathic Choroidal ranibizumab injection (119 patients) or sham
Neovascularization: Results of a 1-Year (59 patients) at baseline, and if required, month
Prospective Trial [36] one. An open-label tailored ranibizumab injec-
This was a prospective, consecutive, nonrandom- tion started from month two until the twelfth
ized, noncomparative, and interventional case month for both arms. Out of 178 patients, 63
series that investigated the visual, anatomic, and patients were having idiopathic CNV (37
safety outcomes of intravitreal bevacizumab in patients receiving ranibizumab and 26 patients
patients with subfoveal idiopathic CNV. receiving sham), and 27 patients were having
A total of 40 consecutive Chinese patients CNV related to angioid streaks (18 patients
with subfoveal idiopathic CNV were included. receiving ranibizumab and 9 patients receiving
All patients received an intravitreal injection of sham). The mean baseline visual acuity for all
bevacizumab (1.25 mg/0.05 mL) at baseline. All subgroups was 62.2 ETDRS letters. At month
patients were followed monthly for 12 months. 2, the ranibizumab group patients gained 9.5
Reinjection was done for persistent or recurrent letters vs. loss of 0.4 letters in the sham group.
activity as indicated by OCT. The median best- At month 12 (10 months after the open-label
corrected visual acuity at baseline was 20/63. individualized ranibizumab injection), the dif-
The mean OCT central macular thickness at ference was minimized to +11 letters and
baseline was 321 μm. At 12 months, the median +9.3 in the ranibizumab and the sham groups,
best-corrected visual acuity was 20/32 with the respectively. The least square means demon-
mean of 2.4 lines of improvement. Seventy per- strated a treatment effect of 10 letters gained
cent of eyes showed improved visual acuity, 30% compared to sham in patients with idiopathic
were stable, and no eye had deterioration. After CNV, CNV associated with angioid streaks, and
1 year, the mean central macular thickness was CNV in the miscellaneous groups. The underly-
237 μm, and all eyes were in the cicatricial stage ing etiology of the CNV did not contribute to
with no recurrence during the study period. A the generally positive result. Over 1 year, the
mean of 2 injections per eye were given, all mean number of ranibizumab injections was
retreatments were in the first 3 months. No severe 5.8 in the ranibizumab group and 5.4 injections
ocular or systemic adverse event was observed. in the sham group. Overall, intravitreal ranibi-
The results of this study demonstrated the benefi- zumab was safe and effective in treating CNV’s
cial effects of intravitreal bevacizumab in the and improving visual acuity regardless of the
treatment of subfoveal idiopathic CNV’s [36]. etiology [37].
22.3.3 C
NV’s Associated with Angioid management of subfoveal CNV associated with
Streaks angioid streaks. The study enrolled 11 patients
who received an intravitreal injection of bevaci-
22.3.3.1 Photodynamic Therapy zumab 1.25 mg/0.05 ml. The mean duration of
follow-up was 23.8 months. The mean best-
Verteporfin Photodynamic Therapy corrected visual acuity improved from 0.28
of Choroidal Neovascularization (ETDRS letters) at baseline to 0.56 at 20 months.
in Angioid Streaks: One-Year Results All patients had a stable/improved visual acuity
of a Prospective Case Series [38] and 90.9% had improved vision at the 20-month
This was a prospective, open-label, multicenter, visit. In addition, all patients had a stable/reduced
interventional case series of patients treated with CNV size at the same visit. The mean number of
PDT for subfoveal/juxtafoveal CNV secondary injections was 3.5 with no documented systemic
to angioid streaks. PDT was carried out accord- or local adverse events. In summary, intravitreal
ing to the standard protocol (see above). Patients bevacizumab was a safe and effective treatment
were reassessed every 3 months and retreatment modality for the long-term management of CNV
was done if there was evidence of fluorescein secondary to angioid streaks [39].
leakage on angiography. All patients were fol-
lowed for 12 months. Intravitreal Bevacizumab for Nonsubfoveal
A total of 23 eyes of 22 patients were included. Choroidal Neovascularization Associated
Out of 23 eyes, 16 eyes were having subfoveal with Angioid Streaks: 3-Year Follow-Up
CNV’s and 7 eyes had juxtafoveal CNV’s. All Study [40]
CNV’s were classic with no occult component at This was a prospective interventional case series
presentation. Over 1 year, 38% of the patients in aiming to study the effects of intravitreal bevaci-
the subfoveal group gained 8 or more letters and zumab for juxtafoveal/extrafoveal CNV second-
88% lost fewer than 15 letters. Six eyes (38%) ary to angioid streaks. The CNV’s were treatment
showed fluorescein leakage from the CNV at the naïve. Primary end-points were the change in the
12-month visit. The mean number of PDT per- best-corrected visual acuity and the proportion of
formed was 2.9. Two of 7 eyes (29%) gained 8 or eyes gaining 10 letters or more over 3 years.
more letters and 4 eyes (57%) lost more than 15 Secondary end-points included the change in the
letters at 12 months. The mean number of PDT central macular thickness (CMT) and progres-
treatments done on eyes in this group was 3.4. sion of the lesion to the fovea at the conclusion of
This case series suggested that PDT can mini- the study.
mize visual loss patients with CNV’s secondary Eighteen patients were eligible to be studied
to angioid streaks. Subfoveal CNV’s showed bet- with 15 patients completing the 3-year follow-up
ter results than juxtafoveal lesions in which PDT on which the results were based. Juxtafoveal
did not limit the progression of CNV into the CNV was present in 8 patients (53.3%) while the
subfoveal area [38]. remaining 7 patients (46.7%) presented with
extrafoveal CNV. The mean best-corrected visual
22.3.3.2 Anti-Vascular Endothelial acuity showed a nonsignificant positive change
Growth Factor Agents from 77.9 ETDRS letters (20/28 Snellen equiva-
lent) at baseline to 80.1 letters at the 1-year
Long-Term Control of Choroidal examination. At the 24- and 36-months’ exami-
Neovascularization Secondary to Angioid nations, a statistically significant worsening
Streaks Treated with Intravitreal (when compared to the 1-year results) of visual
Bevacizumab [39] acuities were noted, with the mean visual acuities
This study was a prospective, consecutive, open- being 72.8 and 65.8 (20/50 Snellen equivalent)
label, interventional case series to study the effi- ETDRS letters, respectively. However, the visual
cacy of intravitreal bevacizumab in the long-term acuities were stable between the time points in
the second and third year and between the time outcome (stable or improved visual acuity) was
points at baseline and the third year. On the other achieved in 93.3% at the end of follow-up. The
hand, 5 patients in the juxtafoveal group and 4 best-corrected visual acuity improved in 53.3%
patients in the extrafoveal group lost 10 or more and stabilized in 40.0% of the eyes. A worse best-
letters at the 3-year visit. The CMT changes were corrected visual acuity was observed in 1 eye
statistically insignificant in both groups. (6.7%). The mean greatest OCT lesion height
Six CNV’s (3 juxtafoveal and 3 extrafoveal) was reduced and the lesion size on fluorescein
extended to the subfoveal area at the second-year angiography decreased/stabilized in 14 eyes
follow-up with no more eyes observed to prog- (93.3%). The mean number of injections was 7.1
ress at the third-year visit. The mean number of per eye. It was concluded that repeat intravitreal
intravitreal bevacizumab injections over 3 years ranibizumab injections had beneficial visual and
was 5.2 (6.2 injections in the juxtafoveal CNV anatomical outcomes in patients with CNV’s sec-
group and 4.1 injections in the extrafoveal CNV ondary to angioid streaks over 1 year [41].
group). In conclusion, intravitreal bevacizumab
was beneficial for nonsubfoveal CNV’s second-
ary to angioid streaks over a 3-year follow-up and 22.3.4 Inflammatory Choroidal
monitoring recurrence/progression was recom- Neovascularization
mended on a monthly basis [40].
22.3.4.1 Corticosteroids Therapy
Intravitreal Ranibizumab Treatment for Inflammatory CNV’s
of Macular Choroidal Neovascularization
Secondary to Angioid Streaks: One-Year The Use of Corticosteroids for Choroidal
Results of a Prospective Study [41] Neovascularization in Young Patients [42]
This was a prospective interventional case series The aim of this prospective study was to evaluate
that investigated the efficacy of intravitreal the effect of systemic corticosteroids in the treat-
ranibizumab in the treatment of macular CNV’s ment of sight-threatening choroidal neovascular-
associated with angioid streaks. ization (CNV) in patients with punctate inner
A total of 15 eyes of 14 consecutive patients choroidopathy (PIC) and multifocal inner choroi-
were included and observed for 1 year. Five eyes ditis (MIC). It included all cases where the CNV
were having a subfoveal CNV, five eyes with jux- was not suitable for laser photocoagulation.
tafoveal CNV, and 5 eyes with extrafoveal Systemic oral prednisolone was started with a
CNV. All eyes were injected with intravitreal dose of 1 mg/kg and reduced gradually every
ranibizumab 0.5 mg/0.05 ml monthly for 3–5 days over the course of 6–8 weeks. This cor-
4 months. Then, patients were asked to come ticosteroids course was repeated if retreatment
back in 6 weeks for re-evaluation. If the CNV deemed necessary. Main outcome measures were
was inactive (no hemorrhage, retinal edema, or best-corrected visual acuity changes, and leakage
subretinal fluid), the eye was injected and the on FFA during a mean follow-up period of
patient was evaluated in 8 weeks. The same was 13 months.
done in this 8-week extension visit for inactive Twelve eyes of 10 patients with symptomatic
lesions except that the patients were requested to CNV due to PIC or MIC were identified. Mean
come back after 3, 4, and 5 months for injections initial visual acuity was 20/60 (range: 20/15–
until completing a total of 24 months. If the CNV 3/200). Ten of 12 eyes showed either improve-
was active at any visit after the 4 initial injec- ment in Snellen visual acuity or stabilization of
tions, the eye was injected and the patients were vision after the corticosteroids course. Mean final
followed up at the same time interval as that visual acuity was 20/40 (range: 20/20–20/200).
before the preceding injection. The mean best- Resolution of leakage on FFA was achieved in 9
corrected visual acuity at baseline was 20/100 eyes, and reduction of leakage on FFA was seen
which improved to 20/50 at 1 year. The primary in the remaining 3 eyes. Four patients required
more than one course of treatment. No systemic injection at baseline, then reassess every month
complications were observed during the study for a total of 12 months. Repeat injection was
period. The study concluded that oral corticoste- done during the monthly follow-up visits if fluid
roids course in healthy young patients with sub- on optical coherence tomography is seen and/or
foveal CNV secondary to PIC or MIC may reduce leakage on fluorescein angiography is detected.
vascular leakage as seen on FFA, and stabilize Fourteen eyes of 14 patients were included.
vision when no other treatment option is avail- Mean BCVA improved from 0.41 ± 0.23 LogMAR
able [42]. at baseline (20/50 Snellen equivalent) to
0.16 ± 0.13 LogMAR at 12-month follow-up
22.3.4.2 Photodynamic Therapy (20/28 Snellen equivalent) (P = 0.002). Mean cen-
tral macular thickness decreased to 239 μm at the
Photodynamic Therapy for Juxtafoveal 12-month examination compared to 318 μm at
Choroidal Neovascularization Associated baseline (P < 0.001). Mean number of injections
with Multifocal Choroiditis [43] was 2 ± 0.6 (range, 1–3) over the 12-month fol-
This study was an open-label, prospective, inter- low-up. No active CNV or extension of the CNV
ventional case series to evaluate the effect of to the fovea at the 12-month follow-up was noted.
PDT on visual acuity outcomes in patients with No systemic or ocular side effects were observed.
juxtafoveal CNV secondary to multifocal choroi- The study concluded that intravitreal bevacizumab
ditis (MC). is a useful treatment for juxtafoveal choroidal neo-
Seven eyes of seven patients were treated in vascularization associated with MC [44].
accordance with the treatment of age-related
macular degeneration with photodynamic ther- Choroidal Neovascularization Associated
apy (TAP) study. The median baseline visual acu- with Multiple Evanescent White Dot
ity score increased from 75 letters (mean: Syndrome Treated with Intravitreal
70 ± 12.5 SD) to 76 letters (mean: 74.5 ± 6.2 SD), Ranibizumab [45]
and 79 letters (mean: 76.2 ± 7.6 SD) at 12- and This prospective, interventional, case series study
24-month follow-up, respectively. Visual acuity was conducted to evaluate the outcomes of subfo-
decreased by 1.5 or more lines in only one patient, veal CNV associated with multiple evanescent
but three patients (43%) gained at least 1.5 lines white dot syndrome (MEWDS) after intravitreal
of visual acuity, and three patients (43%) main- ranibizumab injection. Patients were evaluated
tained their baseline visual acuity at the 24-month monthly for 12 months after the first ranibizumab
follow-up. There were no side effects recorded injection and re-treatment was performed if
during follow up. The study concluded that PDT activity was detected in the form of fluid on opti-
can be considered as a safe and practical treat- cal coherence tomography and/or leakage on flu-
ment option for juxtafoveal CNV secondary to orescein angiography.
MC [43]. Four eyes of four female patients were
included in the study. All eyes (100%) showed
22.3.4.3 Anti-Vascular Endothelial functional improvement of at least 15 ETDRS
Growth Factor Agents letters. At 12-month examination, mean CMT
decreased to 228 ± 14 μm compared to
Intravitreal Bevacizumab for Juxtafoveal 330 ± 32 μm at baseline. The mean number of
Choroidal Neovascularization Secondary ranibizumab injections administered was 2.2 at
to Multifocal Choroiditis [44] the end of 12 months. No systemic complications
This study was a prospective interventional case or ocular side-effects were noted during follow-
series to evaluate the effect of intravitreal bevaci- up. The study concluded that intravitreal ranibi-
zumab injection on juxtafoveal CNV in multifo- zumab treatment is a valuable therapeutic option
cal choroiditis (MC) patients. The treatment for the management of CNV associated with
protocol was as follows; intravitreal bevacizumab MEWDS [45].
22.3.4.4 Combined Therapy 3.8 (range, 3–7) intravitreal bevacizumab injec-

for Inflammatory CNV’s tions at the final 12-month visit. The study con-
cluded that better functional outcome can be
Photodynamic Therapy Combined achieved using intravitreal bevacizumab injec-
with Systemic Corticosteroids tion rather than PDT for the treatment of subfo-
for Choroidal Neovascularization veal CNV secondary to MC [47].
Secondary to Punctate Inner
Choroidopathy (PIC) [46]
This study was a prospective consecutive case 22.3.5 C
NV’s Associated with Central
series to evaluate functional and angiographic Serous Chorioretinopathy
outcomes of subfoveal CNV secondary to PIC
after a combined PDT/systemic corticosteroid 22.3.5.1 Photodynamic Therapy
treatment regimen. All patients received oral
prednisolone (1 mg/kg bodyweight/day) which Photodynamic Therapy with Verteporfin
was followed 5 days later by PDT. The dose of in Subfoveal Choroidal Neovascularization
prednisolone was then tapered by 10 mg weekly. Secondary to Central Serous
During follow-up, patients would receive another Chorioretinopathy [48]
PDT treatment if there was any sign of CNV This was a prospective interventional, non-
activity on fluorescein angiography. comparative case series of patients with subfoveal
Only 5 eyes of 5 female patients were treated CNV secondary to central serous chorioretinopa-
over the 12-month study period. Mean BCVA thy (CSC) treated with PDT. The diagnosis of CSC
improved from 28.4 ETDRS letters at baseline to was based on the patient’s history and fluorescein
38.8 ETDRS letters at the final visit. Mean num- and indocyanine green angiographic findings.
ber of PDT treatments was only two (range 1–3). Photodynamic therapy was administered to 26
All cases showed signs of decreased CNV activ- eyes of 24 patients. The decision for additional
ity both clinically and angiographically at the PDT treatments was based on follow-up fluores-
final visit. The study concluded that PDT com- cein angiography which was done on each visit.
bined with systemic oral prednisolone appears to The patients were followed-up for a mean of
be safe and effective as a primary treatment for 22.2 months (range, 6–36 months). No dropouts
subfoveal CNV secondary to PIC [46]. were encountered at the 6-month follow-up, 24
(92%) of 26 eyes were available for the 1-year
Bevacizumab Versus Photodynamic follow-up, and only 19 (73%) eyes were avail-
Therapy for Choroidal Neovascularization able at the 2-year follow-up. There was an obvi-
in Multifocal Choroiditis [47] ous treatment benefit in visual acuity gains at
This study was a prospective, pilot, randomized 6 months after PDT. After 24 months, 9 (47%) of
trial comparing the effect of PDT vs intravitreal 19 eyes had 3 lines or more visual improvement,
bevacizumab therapy in patients with multifocal 8 (42%) eyes had stable visual acuity, and 2
choroiditis (MC) complicated by subfoveal CNV. (11%) eyes had lost 3 lines or more. Change in
Twenty-seven eyes of 27 patients were ran- visual acuity from baseline to 12 and 24 months
domized to PDT treatment (13 eyes) or bevaci- was statistically significant. The mean number of
zumab treatment (14 eyes). At 12 months PDT treatments needed per patient was 2.6
examination, 5 eyes (36%) treated with bevaci- (range, 1–7). None of the patients had systemic
zumab gained more than 3 lines compared to or ocular adverse effects. This case series con-
none in the PDT group. Final central retinal cluded that PDT with verteporfin is a safe and
thickness dropped significantly in both groups effective treatment for subfoveal CNV secondary
compared to baseline values. A mean of 1.7 treat- to CSC and more than 75% of patients will have
ments (range, 1.0–3.0) were performed in the stable or improved visual acuity 2 years after
PDT subgroup compared to a mean number of treatment [48].
Treatment of Choroidal patient (8%) who lost one line compared to base-
Neovascularization in Central Serous line value. The CMT decreased from 276 ± 95 μm
Chorioretinopathy by Photodynamic at baseline to 199 ± 34 μm at the 24-month follow-
Therapy with Verteporfin [49] up (P = 0.016). No ocular or systemic side effects
This was a prospective, consecutive, nonrandom- were noted. The study concluded that intravitreal
ized, and interventional study on patients with bevacizumab injection for subfoveal CNV associ-
subfoveal or juxtafoveal CNV secondary to CSC ated with PD is a valuable treatment [50].
to evaluate the safety and efficacy of PDT with
verteporfin as a treatment modality. Key Learning Points
Ten eyes were included in the study with a • Laser photocoagulation treatment for juxtafo-
mean follow-up of 12.6 months (range, veal, extrafoveal, and peripapillary CNV’s
6–21 months). At the last follow-up, six (60%) secondary to POHS is superior to no
eyes gained 2 or more lines in visual acuity test- treatment.
ing, 4 (40%) eyes had the final BCVA within 1 • Krypton laser photocoagulation is not supe-
line of their initial BCVA, but none of the patients rior to Argon laser photocoagulation in the
suffered a loss of 2 or more lines in vision. The setting of extrafoveal CNV’s secondary to
mean LogMAR BCVA improvement after PDT POHS.
was 2.4 lines. The study concluded that PDT with • Argon laser photocoagulation for idiopathic
verteporfin therapy for CNV’s secondary to CSC CNV’s was beneficial but benefit was less pro-
is a safe, well-tolerated treatment option and nounced than that in the POHS trial.
achieves a stable or improved vision [49]. • Systemic steroids prior to PDT was better than
PDT alone in patients with idiopathic and
inflammatory CNV’s.
22.3.6 CNV’s Associated with Pattern • Intravitreal bevacizumab is safe and effective
Dystrophy (PD) in patients with idiopathic CNV’s.
• Intravitreal ranibizumab was safe and effec-
22.3.6.1 Anti-Vascular Endothelial tive in treating CNV’s and improving visual
Growth Factor Agents acuity regardless of the etiology.
• PDT appears safe and effective in treating
Intravitreal Bevacizumab for Subfoveal inflammatory CNV’s and CNV’s secondary to
Choroidal Neovascularization Associated CSCR.
with Pattern Dystrophy [50] • Intravitreal anti-VEGF agents appear safe and
This was a prospective, nonrandomized, open- effective in treating inflammatory CNV’s and
label, interventional study on patients with subfo- CNV’s secondary to CSCR.
veal CNV secondary to pattern dystrophy (PD). • Intravitreal anti-VEGF agents had greater
The aim was to assess the effects of treatment beneficial effects than PDT therapy in patients
with intravitreal bevacizumab. The treatment with inflammatory CNV’s.
protocol was to administer a loading dose of
three consecutive injections monthly, then
patients were followed up monthly and injections References
were administered “as needed” during a 24-month
1. Neri P, Lettieri M, Fortuna C, Manoni M, Giovannini
follow-up period. A. Inflammatory choroidal neovascularization.
Twelve patients were included in the study. At Middle East Afr J Ophthalmol. 2009;16(4):245–51.
the 12-month follow-up, 7 (58%) patients gained 2. Dhingra N, Kelly S, Majid MA, Bailey CB, Dick
at least 3 lines, 11 (92%) patients gained at least 1 AD. Inflammatory choroidal neovascular membrane
in posterior uveitis-pathogenesis and treatment.
line, and 1 patient showed stabilization of Indian J Ophthalmol. 2010;58(1):3–10.
BCVA. At the 24-month follow-up, data regarding 3. Bansal R, Bansal P, Gupta A, Gupta V, Dogra MR,
BCA remained unchanged except for a single Singh R, et al. Diagnostic challenges in inflammatory
choroidal neovascular membranes. Ocul Immunol 21. Fine SL, Wood WJ, Isernhagen RD, Singerman LJ,
Inflamm. 2017;25(4):554–62. Bressler NM, Folk JC, et al. Laser treatment for sub-
4. Cohen SY, Laroche A, Leguen Y, Soubrane G, Coscas foveal neovascular membranes in ocular histoplasmo-
GJ. Etiology of choroidal neovascularization in young sis syndrome: results of a pilot randomized clinical
patients. Ophthalmology. 1996;103(8):1241–4. trial. Arch Ophthalmol. 1993, 111(1):19–20.
5. Cheung CMG, Lai TYY, Ruamviboonsuk P, Chen 22. Saperstein DA, Rosenfeld PJ, Bressler NM, Rosa RH
SJ, Chen Y, Freund KB, et al. Polypoidal choroidal Jr, Sickenberg M, Sternberg P Jr, et al. Photodynamic
vasculopathy: definition, pathogenesis, diagnosis, and therapy of subfoveal choroidal neovascularization with
management. Ophthalmology. 2018;125(5):708–24. verteporfin in the ocular histoplasmosis syndrome:
6. Schlaegel TF. Ocular histoplasmosis. New York: one-year results of an uncontrolled, prospective case
Grune & Stratton; 1977. series. Ophthalmology. 2002;109(8):1499–505.
7. Makley TA Jr, Long JW, Suie T. Therapy of chorio- 23. Rosenfeld PJ, Saperstein DA, Bressler NM, Reaves
retinitis presumed to be caused by histoplasmosis. Int TA, Sickenberg M, Rosa RH Jr, et al. Photodynamic
Ophthalmol Clin. 1975;15(3):181–96. therapy with verteporfin in ocular histoplasmosis:
8. Schlaegel TF Jr. Corticosteroids in the treatment uncontrolled, open-label 2-year study. Ophthalmology.
of ocular histoplasmosis. Int Ophthalmol Clin. 2004;111(9):1725–33.
1983;23(2):111–23. 24. Saperstein DA, Rosenfeld PJ, Bressler NM, Rosa RH
9. Gass J. Stereoscopic atlas of macular diseases. St. Jr, Sternberg P Jr, Aaberg TM Sr, et al. Verteporfin
Louis, MO: Mosby; 1987. therapy for CNV secondary to OHS. Ophthalmology.
10. Watzke RC, Leaverton PE. Light coagulation in pre- 2006;113(12):2371.e1–3.
sumed histoplasmic choroiditis. Controlled clinical 25. Thomas MA, Grand MG, Williams DF, Lee CM,
study. Arch Ophthalmol. 1971;86(2):127–32. Pesin SR, Lowe MA. Surgical management of sub-
11. Argon laser photocoagulation for ocular histoplas- foveal choroidal neovascularization. Ophthalmology
mosis. Results of a randomized clinical trial. Arch 1992;99(6):952–68; discussion 75-6.
Ophthalmol. 1983;101(9):1347–57. 26. Hawkins BS, Bressler NM, Bressler SB, Davidorf
12. Argon laser photocoagulation for neovascular macu- FH, Hoskins JC, Marsh MJ, et al. Surgical removal
lopathy. Three-year results from randomized clinical vs observation for subfoveal choroidal neovascular-
trials. Macular Photocoagulation Study Group. Arch ization, either associated with the ocular histoplas-
Ophthalmol. 1986;104(5):694–701. mosis syndrome or idiopathic: I. Ophthalmic findings
13. Argon laser photocoagulation for neovascular macu- from a randomized clinical trial: Submacular Surgery
lopathy. Five-year results from randomized clinical Trials (SST) Group H Trial: SST Report No. 9. Arch
trials. Macular Photocoagulation Study Group. Arch Ophthalmol. 2004;122(11):1597–611.
Ophthalmol. 1991;109(8):1109–14. 27. Hawkins BS, Miskala PH, Bass EB, Bressler NM,
14. Krypton laser photocoagulation for neovascular Childs AL, Mangione CM, et al. Surgical removal
lesions of ocular histoplasmosis. Results of a random- vs observation for subfoveal choroidal neovascular-
ized clinical trial. Macular Photocoagulation Study ization, either associated with the ocular histoplas-
Group. Arch Ophthalmol. 1987;105(11):1499–507. mosis syndrome or idiopathic: II. Quality-of-life
15. Laser photocoagulation for juxtafoveal choroidal findings from a randomized clinical trial: SST Group
neovascularization. Five-year results from random- H Trial: SST Report No. 10. Arch Ophthalmol.
ized clinical trials. Macular Photocoagulation Study 2004;122(11):1616–28.
Group. Arch Ophthalmol. 1994;112(4):500–9. 28. Walia HS, Shah GK, Blinder KJ. Treatment of
16. Persistent and recurrent neovascularization after kryp- CNV secondary to presumed ocular histoplasmosis
ton laser photocoagulation for neovascular lesions of with intravitreal aflibercept 2.0 mg injection. Can J
ocular histoplasmosis. Macular Photocoagulation Ophthalmol. 2016;51(2):91–6.
Study Group. Arch Ophthalmol. 1989;107(3):344–52. 29. Argon laser photocoagulation for idiopathic neovas-
17. Laser photocoagulation for neovascular lesions nasal cularization. Results of a randomized clinical trial.
to the fovea. Results from clinical trials for lesions Arch Ophthalmol. 1983;101(9):1358–61.
secondary to ocular histoplasmosis or idiopathic 30. Krypton laser photocoagulation for idiopathic neo-
causes. Macular Photocoagulation Study Group. Arch vascular lesions. Results of a randomized clinical
Ophthalmol. 1995;113(1):56–61. trial. Macular Photocoagulation Study Group. Arch
18. Gutman FA. The natural course of active choroidal Ophthalmol. 1990;108(6):832–7.
lesions in the presumed ocular histoplasmosis syn- 31. Chan WM, Lam DS, Wong TH, Lai TY, Kwok AK,
drome. Trans Am Ophthalmol Soc. 1979;77:515–41. Tam BS, et al. Photodynamic therapy with vertepor-
19. Marshall J, Bird AC. A comparative histopathological fin for subfoveal idiopathic choroidal neovasculariza-
study of argon and krypton laser irradiations of the tion: one-year results from a prospective case series.
human retina. Br J Ophthalmol. 1979;63(10):657–68. Ophthalmology. 2003;110(12):2395–402.
20. Argon green vs krypton red laser photocoagulation 32. Giovannini A, Neri P, Mercanti L, Brue
for extrafoveal choroidal neovascularization. One- C. Photodynamic treatment versus photodynamic
year results in ocular histoplasmosis. The Canadian treatment associated with systemic steroids for idio-
Ophthalmology Study Group. Arch Ophthalmol. pathic choroidal neovascularisation. Br J Ophthalmol.
1994;112(9):1166–73. 2007;91(5):620–3.
33. Li XX, Tao Y. Small laser spot versus standard laser 42. Flaxel CJ, Owens SL, Mulholland B, Schwartz SD,
spot photodynamic therapy for idiopathic choroidal Gregor ZJ. The use of corticosteroids for choroidal
neovascularization: a randomized controlled study. neovascularisation in young patients. Eye (Lond).
Chin Med J. 2012;125(24):4424–8. 1998;12(Pt 2):266–72.
34. Kumar A, Prakash G, Singh RP. Transpupillary 43. Parodi MB, Iacono P, Spasse S, Ravalico
thermotherapy for idiopathic subfoveal choroi- G. Photodynamic therapy for juxtafoveal choroidal
dal neovascularization. Acta Ophthalmol Scand. neovascularization associated with multifocal choroi-
2004;82(2):205–8. ditis. Am J Ophthalmol. 2006;141(1):123–8.
35. Mandal S, Garg S, Venkatesh P, Mithal C, Vohra R, 44. Parodi MB, Iacono P, Mansour A, Benedetto U,
Mehrotra A. Intravitreal bevacizumab for subfo- Knutsson KA, Bandello F, et al. Intravitreal beva-
veal idiopathic choroidal neovascularization. Arch cizumab for juxtafoveal choroidal neovasculariza-
Ophthalmol. 2007;125(11):1487–92. tion secondary to multifocal choroiditis. Retina.
36. Zhang H, Liu ZL, Sun P, Gu F. Intravitreal bevaci- 2013;33(5):953–6.
zumab for treatment of subfoveal idiopathic choroi- 45. Battaglia Parodi M, Iacono P, Zucchiatti I, Bandello
dal neovascularization: results of a 1-year prospective F. Choroidal neovascularization associated with
trial. Am J Ophthalmol. 2012;153(2):300–6.e1. multiple evanescent white dot syndrome treated with
37. Lai TYY, Staurenghi G, Lanzetta P, Holz FG, intravitreal ranibizumab. Ocul Immunol Inflamm.
Melissa Liew SH, Desset-Brethes S, et al. Efficacy 2018;26(4):608–11.
and safety of ranibizumab for the treatment of cho- 46. Fong KC, Thomas D, Amin K, Inzerillo D, Horgan
roidal neovascularization due to uncommon cause: SE. Photodynamic therapy combined with systemic
twelve-month results of the Minerva study. Retina. corticosteroids for choroidal neovascularisation sec-
2018;38(8):1464–77. ondary to punctate inner choroidopathy. Eye (Lond).
38. Browning AC, Chung AK, Ghanchi F, Harding SP, 2008;22(4):528–33.
Musadiq M, Talks SJ, et al. Verteporfin photodynamic 47. Parodi MB, Iacono P, Kontadakis DS, Zucchiatti I,
therapy of choroidal neovascularization in angioid Cascavilla ML, Bandello F. Bevacizumab vs pho-
streaks: one-year results of a prospective case series. todynamic therapy for choroidal neovasculariza-
Ophthalmology. 2005;112(7):1227–31. tion in multifocal choroiditis. Arch Ophthalmol.
39. Neri P, Salvolini S, Mariotti C, Mercanti L, Celani S, 2010;128(9):1100–3.
Giovannini A. Long-term control of choroidal neovas- 48. Ergun E, Tittl M, Stur M. Photodynamic therapy with
cularisation secondary to angioid streaks treated with verteporfin in subfoveal choroidal neovascularization
intravitreal bevacizumab (Avastin). Br J Ophthalmol. secondary to central serous chorioretinopathy. Arch
2009;93(2):155–8. Ophthalmol. 2004;122(1):37–41.
40. Iacono P, Battaglia Parodi M, La Spina C, Bandello 49. Chan WM, Lam DS, Lai TY, Yuen KS, Liu DT, Chan
F. Intravitreal Bevacizumab for nonsubfoveal cho- CK, et al. Treatment of choroidal neovasculariza-
roidal neovascularization associated with angioid tion in central serous chorioretinopathy by photody-
streaks: 3-year follow-up study. Am J Ophthalmol. namic therapy with verteporfin. Am J Ophthalmol.
2016;165:174–8. 2003;136(5):836–45.
41. Ladas ID, Kotsolis AI, Ladas DS, Niskopoulou M, 50. Parodi MB, Iacono P, Cascavilla M, Zucchiatti I,
Georgalas I, Papakonstantinou D, et al. Intravitreal Kontadakis DS, Bandello F. Intravitreal bevacizumab
ranibizumab treatment of macular choroidal neo- for subfoveal choroidal neovascularization associated
vascularization secondary to angioid streaks: with pattern dystrophy. Invest Ophthalmol Vis Sci.
one-year results of a prospective study. Retina. 2010;51(9):4358–61.
2010;30(8):1185–9.
Tariq Alzahem is currently the Chief Resident in his last

year in King Saud University Ophthalmology Training
Program. He was awarded as The Resident of The Year in
2017 and 2018. He was also awarded the Ideal Student Nayef Alswaina is an Ophthalmology consultant with a
Prize at the level of King Saud University in 2014. Tariq is sub-specialization and interest in surgical retina. He fin-
planning to be a clinical scientist in the field of Surgical ished his residency and subspecialty program from Riyadh
Retina where his professional experience and education ophthalmology program (King Khaled Eye Specialist
will allow him to make an immediate contribution as an Hospital and King Abdulaziz University Hospital).
integral part of a progressive institution. Currently, he is working as an Assistant Professor at the
Department of Ophthalmology, College of Medicine,
Qassim University, Saudi Arabia.
Marwan A. Abouammoh is a full Professor of

Ophthalmology at King Saud University, Riyadh, Saudi
Arabia. He is a graduate of Riyadh’s Joint Ophthalmology
residency program and has completed a 2-year vitreoreti-
nal fellowship at King Khaled Eye Specialist Hospital. He
did a further year of fellowship training at Queen’s
University, Kingston, Ontario where he also studied and
practiced medical/scientific reasoning, health outcomes,
and safety of novel treatments for macular degeneration
and diabetic retinopathy. He is currently involved in clini-
cal-based research related to Retina and Ophthalmology
and actively contributes to original clinical trials, system-
atic reviews, and meta-analyses.
Reading Centers for Clinical Trials
of Choroidal Neovascularization 23
(CNV): Present Role and Future
Opportunities
René Rückert, Lala Ceklic, and Marion R. Munk
23.1 Introduction until recently. With (experimental) Enhanced

Deep Imaging (EDI)—OCT and Swept-Source
Photographic reading centers play an important (SS)—OCTs we are now able to even detect and
role in the objective, independent, and scientific quantify structures that were out of reach for non-
justified development of novel diagnostic and invasive in vivo diagnosis and follow-up just a
therapeutic modalities for various eye patholo- few years ago.
gies, and in particular for new therapies of retinal Pharmaceutical companies—big pharma and
diseases including choroidal neovascularization a growing number of small biotechs—develop
(CNV). In the recent decade, we all were wit- (-ed) various drugs and devices which can pre-
nesses of the development and launch of various vent blindness and improve vision of patients
improved and enhanced diagnostic tools. affected by CNV such as the approved anti-
TD-OCT was a breakthrough in retina imaging, VEGF drugs ranibizumab and aflibercept.
introduction of SD-OCT exponentially increased Besides the current domination of CNV therapy
the spatial resolution of retinal images. OCT-A as by those two approved (and one off-label) VEGF
a new imaging tool nowadays allows the nonin- blockers, new treatment modalities are under
vasive assessment of vascular structures in the research in clinical trials. As global clinical trials
retina that needed invasive procedures during FA for the approval or the label-extension of
including risk-associated injection of fluorescein approved drugs require prospective and double-
masked trials, and independent, expert process-
ing and analyzing of retina images, retina imaging
R. Rückert
Eyegnos Consulting for Ophthalmic Drug reading centers are a mainstay of current ophthal-
Development, Bern, Switzerland mic drug development. Experienced reading cen-
L. Ceklic ters use masked analysis which prevents research
Bern Photographic Reading Center, Inselspital, outcomes from being influenced by various
University Hospital Bern, Bern, Switzerland biases such as patient expectations (placebo
M. R. Munk (*) effect), investigators expectancy (observer bias),
Bern Photographic Reading Center, Inselspital, and sponsor expectations. Professional reading
University Hospital Bern, Bern, Switzerland centers can support protocol writing and defini-
Department of Ophthalmology, Inselspital, University tion of clinical endpoints for clinical trials.
Hospital Bern, Bern, Switzerland A significant number of reading centers exist
Department of Ophthalmology, Feinberg School of around the world, many are smaller with a focus
Medicine, Northwestern University, on academic studies, plus several larger and more
Chicago, IL, USA

318 R. Rückert et al.
professionalized reading centers that have suffi- standard operating procedures. Such image anal-
cient capacity for large, global, multicenter trials ysis data are the basis for the evaluation of mor-
with 4-digit patient numbers. Such large studies phological changes and the statistical analysis of
are needed for late-stage trials to assess the effi- imaging endpoints and/ or can be used to identify
cacy of new drugs to treat retinal diseases such as biomarker at baseline or during the course of a
exudative age-related macular degeneration therapeutic study—to assess and quantify bio-
(AMD). marker of high/low-responders and many more
Photographic Reading Centers apply profes- aspects of a patient’s data.
sional administrative measures and standard Photographic Reading Centers must base their
operational procedures (SOP’s) to assure GCP/ work on principles of Good Clinical Practice.
ICH guidelines are followed and highest ethical Development of numerous retinal imaging
standards are applied, to generate data which are techniques and need for more precise and more
the basis for regulatory submission. They apply detailed imaging of CNV and possible morpho-
measures to assure that images are properly logic risk factors require imaging experts that
labeled, that personal information are removed in receive permanent training and managing of a
order to de-identify patients, that image acquisi- reading center requires cooperation of retina
tion is complete, and that the images are of ade- experts, biomedicine engineers, and IT. Excellent
quate quality for analysis. IT experts and training of image readers in the pathogenesis and
engineers are the basis for development and sup- detection of CNV leads to optimal results of the
port of the digital image transfer and long-term evaluation of retinal images.
storage. Furthermore, reading centers usually The role of reading centers has expanded
offer certification and real-time teaching for vastly in the past years—beyond simple analysis
investigators and photographers in clinical stud- of images as service provider to pharma compa-
ies, thereby contributing to the education and nies—to a real partnership model in drug devel-
general improvement in the quality of day-by- opment including the following service and
day diagnostic and of clinical research in oph- expert support:
thalmology. During the clinical studies, the
digital images are assessed and analyzed by
trained graders. 23.2.1 Pre-trial Support
Ophthalmic reading centers collect data in a
harmonized fashion and thereby produce data • Clinical trial design and imaging endpoint
sets of high quality that were the basis for regula- consulting and development
tory filings and countless post hoc analyses of • Development of clinical site manual for imag-
global clinical studies. Such high-quality image ing procedures
data sets are furthermore the basis for new big • Training and certification of clinical sites
data-driven research and support the develop-
ment of automated, artificial intelligence (AI)-
based analysis tools. 23.2.2 Trial Support
• Internal validation of key grading criteria

23.2 Roles and Responsibilities across modalities/different imaging devices.
of Imaging Reading Centers • Clinical site support for quality assurance.
• Real-time sponsor and CRO access to clinical
The purpose of reading centers is to assure that trial reads and reports.
all steps in collecting, processing, and analyzing • Detection of safety signals in ongoing studies
data and images are independent and non-biased including support for independent data safety
by the sponsor of clinical studies following strict monitoring committees (IDMCs).
23 Reading Centers for Clinical Trials of Choroidal Neovascularization (CNV): Present Role and Future… 319
23.2.3 Post-trial Support The Age-Related Eye Disease Study (AREDS)

was a multicenter prospective cohort study of the
• Sponsor-tailored data outputs. clinical course, prognosis, and risk factors for
• Support of statistical analysis. AMD. An important goal of AREDS has been the
• Support for writing study report and regula- development of a severity scale for AMD, to pro-
tory filing documents. vide baseline risk categories, to allow tracking of
• Research and post hoc analyses for additional progression along the scale, and to define surro-
data generation. gate outcomes for progression to advanced AMD
[1, 2]. AREDS started in 1992 and ran until 2005
For evaluating CNV related to AMD, essential and enrolled 4613 study participants. The
retinal imaging methods are 3 field fundus color AREDS Report No.17 from 2005 established an
photography (CF), fluorescein angiography (FA), AREDS severity scale and risk factors were eval-
indocyanine green angiography (ICGA; optional, uated by human graders and reading of stereo-
in order to identify Type 3 neovascularization, scopic 3-field fundus photography’s and recent
polypoidal choroidal vasculopathy, and subclini- publication by Burlina and co-authors are based
cal neovascularization, respectively), optical on deep machine learning. This study used
coherence tomography (OCT) and optical coher- 67,401 color fundus images from the 4613 study
ence tomography-angiography (OCT-A). participants. The weighted kappa scores were
0.77 for the 4-step and 0.74 for the 9-step AMD
severity scales, which confirms that the concor-
23.3 NV Image Reading
C dance and reproducibility of artificial intelligence
Modalities image grading are relatively high and reliable [2].
Retinal Imaging reading centers were in the past

initially established to read color fundus and FA 23.5 Fluorescein Angiography
images. Recent technical development of OCT in CNV
devices has expanded the spectrum of analyses as
OCT has become a mainstay of retinal disease Fluorescein angiography (FA) has been used to
diagnostic in particular also for CNV. Retina define lesion size and type of CNV. It allows the
image reading centers accept and assess most of distinction between well-demarcated classic CNV,
the current standard image modalities and devices and ill-defined occult CNV. The location and exact
for the analysis of morphologic outcome mea- boundaries of occult CNV often are difficult to
sures. Below, we describe in more detail some of determine precisely on FA, i.e., due to obscuration
the standard assessment and parameters analyzed of the neovascular membrane by overlying turbid
by reading centers. exudates, blood and/or pigment, and rapid fluores-
cein pooling beneath a serous PED.
According to angiographic patterns, as ana-
23.4 olor Fundus Photography
C lyzed by photographic reading centers, CNVs are
for CNV in AMD classified and described as below [3, 4]:
Most prospective clinical trials require detailed 1. Classic CNV: Hyperfluorescence in the early
and objective analyses of fundus photographs. phases of the angiogram, maintains well-
Thus, standard requirements for data processing demarcated borders, and leaks late thus
for color fundus (CF) are stereoscopic three (3-) obscuring its borders.
fields images with focus on posterior pole and 2. Occult CNV: A lesion whose borders cannot
optic disk (M1), macula (M2), and temporal to be determined and relatively late hyperfluo-
the macula field (M3) with angle of 30° and 55° rescence or lesion that increases by time.
or based on latest developments in imaging 3. Fibrovascular pigment epithelial detachment
devices: ultra widefield imaging (105° and 180°). (PED); a form of occult CNV: A lesion, well-
demarcated or poorly demarcated, that is element and prognosis of the disease.
vated solidly and hyperfluoresces irregularly Computer-assisted methods for CNV segmenta-
to different degrees. tion can be supportive not only to reduce the bur-
4. Late leakage of undetermined source (LLUS); den of manual segmentation but also to reduce
a form of occult CNV: A lesion that demon- inter- and intra-observer variability. The set of
strates irregular, indistinct, late, and sub-RPE model parameters at each pixel are used to seg-
leakage [4]. ment the image into regions of homogeneous
parameters. Thus, development of computer-
CNV could be located relative to the center of assisted systems for determining CNV lesion is
fovea [4]: important to avoid inter- and intra-observer vari-
ability among human graders [11]. For nAMD
1. Extrafoveal (the border of the CNV is 200– trials, FAs are usually performed at baseline,
1500 μm from the center of the fovea) 1 year and 2 years, which is in most of the nAMD
2. Juxtafoveal (the border of the CNV is trials at the end of study visit.
1–199 μm from the center of the fovea)
3. Subfoveal
23.6 Optical Coherence
Prior to the anti-VEGF era, the location and Tomography in CNV
the type of CNV lesion were of uppermost impor- Due to AMD
tance, as lesion type and location determined
appropriate treatment such as laser treatment, Traditionally, FA has been considered the refer-
assessed in the macular photocoagulation study ence standard to detect CNV activity, but FA is
(MPS), or photodynamic therapy, assessed in the costly and invasive with associated risks due to
TAP, VIP, and VIM trials [5, 6]. Nowadays, with the fluorescein injection. Invention and wide-
the anti-VEGF treatments established as the stan- spread introduction of OCT as noninvasive
dard of care, lesion type, and location may impact method have revolutionized retinal diagnostic,
treatment response and long-term outcomes. The treatment assessment and follow-up and
location of CNV lesion (subfoveal and juxtafo- improved analysis of CNV and definition of pos-
veal vs. extrafoveal) may determine inclusion sible biomarkers as prognostic factors.
and exclusion in AMD clinical trials. The impor- Unfortunately, there is substantial disagreement
tance of reading centers is supported by research between OCT and FA findings in detecting active
performed early in the new decade of retina drug disease in patients with CNV due to AMD during
development, demonstrating only moderate inter- disease monitoring [12]. Recently published data
observer agreement for CNV categorization show that OCT was useful in quantitatively eval-
between clinicians and poor agreement among uating subretinal and intraretinal fluid, assessing
diverse retina specialists for categorization and the possible subfoveal location of neovascular-
localization of CNV on fluorescein angiograms. ization, and in monitoring CNV before and after
There was only a moderate intra-observer agree- treatment. OCT was unable to detect CNV
ment for both treatment decision and lesion cat- beneath serous pigment epithelial detachments.
egorization in past studies. And, there was only OCT may have the potential to accurately defin-
moderate agreement between observers and the ing the boundaries in a subset of angiographi-
reading center for angiographic CNV categoriza- cally occult CNV [3, 13]. Unlike FA, which is
tion [7–10]. Those data supported and build the two-dimensional but is a dynamic assessment
basis for funding retina reading centers in order (i.e., includes the time after injection as addi-
to reduce intra-observer variability and to harmo- tional information/parameter) and detects leak-
nize image analysis in large, multicenter, and age, lesion size, and localization, OCT provides a
multicounty clinical studies. three-dimensional but non-dynamic snapshot of
Determining the CNV size and type in fluores- structural information on chorioretinal layers and
cein angiograms are required for proper treat- the presence of fluid in the retina, and in subreti-
nal pigment epithelial spaces, which are consid- reading centers and maybe important biomarkers
ered a surrogate for leakage. OCT is used to for treatment response and (long-) term outcome,
determine both the presence and activity of CNV but are usually in the best case exploratory out-
and the need for anti-VEGF treatment/retreat- come parameters or often just used for post hoc
ment in clinical trials and routinely in clinical analyses. The evaluation of CRT as one of the
practice [14–16]. key secondary outcome parameters in nAMD tri-
New choroidal blood vessels may extend into als was challenging in the earlier days, when
the sub-RPE space (type 1) or into the subretinal TD-OCTs, as well as SD-OCTs, were used. The
space (type 2) [4, 17]. Bleeding and exudation segmentation software of TD-OCTs devices used
which may occur with further expansion of the the internal limiting membrane (ILM) and the
lesion, lead to visual impairment and other func- external limiting membrane (ELM) to obtain
tional symptoms. Alternatively, abnormal blood CRT values, while SD-OCT devices automati-
vessels may originate from an intraretinal loca- cally segmented the ILM and the RPE or Bruch’s
tion and grow into the subretinal space [4]. This membrane for CRT and retinal volume values,
pattern of growth has been named retinal angio- which resulted in an about 50–70 μm lower CRT
matous proliferation (RAP), or type 3 neovascu- measurements of TD-OCT’s compared to
larization [17]. Reading center usually assesses SD-OCTs [20–22]. Nowadays only SD-OCTs or
and defines the extent of CNV in the clinical SS-OCTs are used for nAMD trials; however,
images. CNV, when analyzed with OCT, could be also the current state-of-the-art OCT devices
described as thickening and fragmentation of the from different manufacturers still employ differ-
highly reflective RPE-choriocapillaris band [4]. ent segmentation borders, segmenting, i.e., the
Serous, hemorrhagic, or fibrovascular detach- RPE or the Bruch’s membrane, which also results
ments of RPE represent RPE band elevations in slightly different measurements [23, 24].
with shadowing of the structures beneath the These mean differences range of 14–37 μm,
elevated area [4, 18]. Neurosensory retinal while CRT is measured with SS-OCTs is, in gen-
detachment appears as elevation of moderately eral, lower than with SD-OCT. The two dominat-
reflective band above the RPE band [4]. ing OCT devices in clinical studies use their own
OCT imaging is standard procedure for man- software for CRT measurements, and so the
aging CNV and inter-grader and intra-grader Heidelberg Spectralis® software provides thicker
reproducibility are very high [19]. SD OCT CRT values than Zeiss Cirrus® software. One
showed high sensitivity (85.7–98.3%) and speci- way how to overcome these differences is
ficity (84.2–100%) compared to FA in the diag- custom-made software applying the same seg-
nosis of the CNV subtype [18]. Usually, in mentation for all imported OCTs scans, which
clinical nAMD trials, OCTs are performed at are regularly used in a reading center setting.
each study visit throughout the study, to assess In the era of emerging “Big data” machine
disease activity and treatment needs. In some tri- deep learning, machine learning, and AI attracts
als it is up to the investigator’s discretion to many investigators to develop efficient and accu-
decide whether CNV is active, in other trials dis- rate system for screening, monitoring, manage-
ease activity will be assessed by the reading cen- ment, and prediction systems for various
ter and the (re-) treatment plan will be adapted treatment options. Most of those developments
accordingly. Although many predictive biomark- are based on the large image databases of retinal
ers can be assessed by OCT, the main surrogate reading centers. It is very important to detect risk
markers on OCT are still central retinal thickness factors and early signs of CNV disease. Given
(CRT), presence of subretinal fluid, and presence that OCT imaging is the single most common
of intraretinal fluid. Further parameters such as diagnostic test performed on a daily basis in reti-
the presence of epiretinal membranes, vitreomac- nal clinics, its advantages are the features as non-
ular adhesion, or—traction, subretinal hyperre- invasive, quick, and highly sensitive method.
flective material, macular atrophy, reticular Those features potentially lend itself to automa-
pseudodrusen et al. will be in fact assessed by the tion with deep-learning techniques. Based on
many investigations, it is possible to detect small [33]. OCT-A provides noninvasive measurement
constraints like hard drusen automatically with of the area of neovascular lesions in
relatively high accuracy [25, 26]. Several groups AMD. Sustained growth of type 1 CNV can be
have successfully utilized deep learning in seg- identified in the majority of lesions (80%) that
mentation of OCT scans for the detection of mor- display characteristic patterns of progression
phological features such as intraretinal fluid despite ongoing anti-VEGF therapy [34].
(IRF) or subretinal fluid (SRF) from various reti- According to FA, CNV was classified as classic
novascular diseases [27, 28]. Having a system predominantly classic, minimally classic, and
that can reliably produce diagnostic evaluations occult. In a recently published study, correspond-
that match the physician’s standards may help to ing to FA in OCT-A scans, 46.4% (26/56 eyes)
increase direct face-time with patients, alleviate had well-circumscribed vessels, and 53.6%
administrative burden, reduce administrative (30/56 eyes) showed poorly circumscribed
practice costs and, ultimately, improve day-to- vessels. There were 11 false positives and 7 false
day clinic efficiency. For developing such sys- negatives using OCT-A. The specificity of
tems, reading centers play a crucial role in OCT-A for the detection of CNV was 67.6%,
acquiring and analyzing large pools of imaging with sensitivity of 86.5%. OCT-A may help in the
data, in order to precisely build a foundation of noninvasive diagnosis of CNV and may provide a
AI software’s ability to become a more effective, method for monitoring the evolution of CNV
accurate and reliable tool for screening, monitor- [35]. OCT-A has its rationale as a fast, noninva-
ing, and analyzing CNV [29, 30]. Automated sive screening tool for CNV in asymptomatic
classifiers may be also used in reading centers as patients [36]. There is a notable prevalence of
additional graders or ultimately to replace human subclinical CNV in fellow eyes with unilateral
graders in the near future. exudative CNV, and significantly greater chorio-
capillaris nonperfusion adjacent to all CNV
lesions. There is a trend for increased choriocap-
23.7 Optical Coherence illaris nonperfusion in exudative AMD eyes as
Tomography Angiography compared with their fellow subclinical CNV eyes
for CNV [37]. OCT-A was generally less successful in
detecting CNV than ICGA in patients who were
Optical coherence tomography angiography included in this study based on FA and
(OCT-A) is a recently introduced, noninvasive OCT. Types 1 and 2 CNV area were significantly
imaging technique that generates volumetric smaller in OCT-A than in ICGA. However,
angiography images in a matter of seconds. Using OCT-A detected all type 1 lesions except for one,
OCT-A allows the clinician to visualize CNV indicating that the SD-OCT-A signal is limited
noninvasively and may provide a method for by detection limits of blood flow velocity rather
identifying and guiding the treatment of than lesion type [38]. Treatment-naive eyes and
CNV. The specificity of CNV detection on treated eyes with CNV secondary to neovascular
OCT-A compared with FA seems to be high [31]. AMD respond differently to anti-VEGF therapy.
OCT-A is a noninvasive imaging technique that This should be taken into account when perform-
can be used to visualize blood flow comprising ing OCT-A image analysis in clinical studies and
CNV. Although OCT-A can reliably detect CNV, when using OCT-A for CNV follow-up or plan-
the morphologic appearance of CNV on OCT-A ning therapeutic strategies [39]. There are a num-
does not reliably correlate with clinical activity ber of promising OCT-A parameters that can be
of CNV [32]. OCT-A allows identification of dis- used to diagnose the presence of CNV and to
tinct CNV-specific vascular patterns at the level monitor the activity and progression of the lesion,
of the outer retinal layer and choriocapillaris. pre- and post-treatment morphological character-
Correlation with clinical and functional parame- istics, CNV dimensions, and other quantitative,
ters may be useful to better understand the pathol- computed parameters such as vessel density.
ogy and guide efficient therapeutic strategies However, as this methodology has been intro-
duced only recently to the market and clinic, all 23.8 Increased Complexity
OCT-A data and analysis require further valida- and Challenges in CNV
tion before they can be widely used [40]. Studies
Manufacturers provide OCT-A software with
their devices. Unfortunately, recent studies have The complexity of retina image analysis has
shown that the inter-device agreement and com- increased significantly over the last years. New
parability is rather low. This is contributed to, i.e., morphologic features are usually initially identi-
divergent segmentations, different algorithms to fied in small academic studies using new technol-
generate flow motion, and due to different post- ogy or advanced detection methods. Subsequently,
acquisition processing software [41, 42]. such new possible imaging endpoints and patho-
Therefore, comparability and pooling of acquired logic morphologic features of retinal diseases
data, in particular quantitative metrics for clinical need to be verified—and that is where reading
trials, is a difficult task and currently a key chal- centers can (and do) play a very significant role
lenge for reading centers in clinical trials. OCT-A with post hoc analyses of existing studies using
is still in the focus of many investigators world- data from their large imaging database. Once
wide and knowledge from OCT-A imaging is still verified as significant or interesting to evaluate
developing. Again though, image reading center disease progress or treatment outcome, such
will play a key role in collecting and analyzing marker can be incorporated in a prospective way
large pools of OCT-A data in order to improve in new clinical trials for CNV treatments as
knowledge and familiarity with this new addition exploratory endpoints.
to the retina (CNV) imaging tools. Current ongo- One impressive example of the increased com-
ing nAMD trials have included OCT-A parame- plexity of image analysis by reading centers is
ters as exploratory outcome parameters in their given in Table 23.1, with the listed outcome
protocols. parameter in two selected Phase 3/3b/4 studies as
Table 23.1 Secondary imaging/morphologic endpoints in pivotal AMD/CNV trials—2008 versus 2018 (from: www.
clinicaltrials.gov)
VIEW2 Study MERLIN Study
Aflibercept in AMD, Brolucizumab in AMD,
Study start: April 2008 Study start: October 2018
• Mean change from • Change in Central Subfield Thickness (CST) from baseline to each post-baseline
baseline in Choroidal visit; CST will be assessed using Spectral Domain Optical Coherence Tomography
Neovascularization (SD-OCT) images.
(CNV) area at week • Intraretinal fluid (IRF) at each post-baseline visit; IRF will be assessed using
52—LOCF; CNV area SD-OCT images. The number of subjects with IRF (present, absent) will be
values measured in reported for each post-baseline visit.
square millimeters; • Subretinal fluid (SRF) at each post-baseline visit; SRF will be assessed using
lower values represent SD-OCT images. The number of subjects with SRF (present, absent) will be
better outcomes. reported for each post-baseline visit.
• Sub-retinal pigment epithelium (sub-RPE) fluid in patients with sub-RPE fluid at
baseline; sub-RPE fluid will be assessed using SD-OCT images. The number of
subjects with sub-RPE fluid in subjects with sub-RPE fluid at baseline (present,
absent) will be reported for each post-baseline visit.
• Fluid-free status (no IRF, SRF, or sub-RPE fluid) at each post-baseline treatment
visit; IRF, SRF, and sub-RPE fluid will assessed using SD-OCT images. The
number of subjects with fluid-free status (no IRF, SRF, or sub-RPE) will be
reported for each post-baseline visit.
• Time to first dry retina (no IRF or SRF) finding; IRF and SRF will be assessed
using SD-OCT images. A dry retina is defined as no IRF or SRF at the visit.
• Time to first sustained dry retina (no IRF or SRF at ≥2 consecutive visits) finding;
IRF and SRF will be assessed using SD-OCT images. A sustained dry retina is
defined as no IRF or SRF at 2 or more consecutive visits.
revealed on the www.clinicaltrials.gov website. In Reading centers can play an essential role in
2008, the VIEW studies for approval of aflibercept educating regulators about the importance of
in AMD contained exactly one imaging endpoint imaging also on documenting the efficacy of new
as a secondary outcome, the MERLIN study, treatments. New developments, as a recent (April
which just started—10 years later—to evaluate the 08, 2019) conducted workshop at the FDA to dis-
efficacy of brolucizumab, a new generation anti- cuss (beyond other modalities)—the importance
VEGF for the treatment of nAMD, lists seven (!) of OCT-based imaging as valid and essential end-
imaging/OCT endpoints as secondary outcome points in ophthalmic clinical studies, can be seen
measures, all of which have to be evaluated and as a move in the right direction.
assessed by a central reading center (Table 23.1).
Both studies assess the therapeutic efficacy in
CNV/neovascular AMD: 23.9 Artificial Intelligence
in Image Analysis
• Vascular Endothelial Growth Factor (VEGF)
Trap-Eye: Investigation of Efficacy and Safety Artificial intelligence (AI) is a general term that
in Wet Age-Related Macular Degeneration implies the use of a computer to model intelli-
(AMD) (VIEW 2); https://clinicaltrials.gov/ gent behavior with minimal human interven-
ct2/show/NCT00637377 tion. Recent developments in the ophthalmology
• Study of Safety and Efficacy of Broluci- community and in reading centers are auto-
zumab 6 mg Dosed Every 4 Weeks Com- mated algorithms, which use AI to analyze and
pared to Aflibercept 2 mg Dosed Every quantify morphologic characteristics of the
4 Weeks in Patients With Retinal Fluid (retina-) images. Those AI-based software
Despite Frequent Anti-VEGF Injections assessments might be the future for analyzing
(MERLIN) in AMD; https://clinicaltrials. many diagnostic imaging methods. Whether
gov/ct2/show/NCT03710564 and when such automated segmentation and
assessment of images will be accepted by regu-
Such examples demonstrate the successful latory agencies such as EMA or FDA for piv-
development and introduction of high-resolution otal trials remain to be seen, though. However,
imaging devices and the acceptance and contri- automation could ease the work of reading cen-
bution of research, often led by reading centers; ter experts and the growing demand for analy-
in improving the knowledge and recognition of ses (see above). AI could support the
relevant parameters for diagnosis and treatment analysis—and maybe 1 day replace the human
of CNV. Nevertheless, it also shows that the num- trained experts as the final decision maker of
ber of the morphological secondary surrogate diagnosis, pathology, or treatment need or treat-
outcome parameters may have increased but are ment success. Especially, the fact that well-
still solely based on the presence of sub-RPE, trained algorithms already outperform single
intra- or subretinal fluid. human graders, raise hope that soon very accu-
One key issue remains for drug developers rate automated grading software will be avail-
and the retina community: the very conservative able, at least for CF imaging and OCT.
view of regulatory agencies on imaging outcomes Implementation of automated algorithms
in clinical studies. FDA and EMA usually accept would save time and support fast screening of
functional parameter (BCVA) as primary out- images. AI might also be helpful for improving
come in AMD/CNV studies, all imaging and vthe reproducibility of human graders and pre-
morphologic data are seen by regulators as just vent inconsistencies between image analysis for
secondary surrogate or rather supportive explor- different patients and/or patients’ follow-up
atory outcome parameter. through the course of a clinical trial.
Nowadays, AI-assisted medical screening 23.10 Conclusion

and diagnosis based on retinal images are
emerging, especially in retinal diseases, Photographic reading centers established pools
including morphologic feature identification of retinal images in CNV and the purpose of
and auxiliary diagnosis support. AI has broad those institutions is the proper analysis of mor-
application across many medical fields, it phologic data and development of modern, effi-
holds significant promises for use in use in cient, accurate, and reproducible analysis tools.
ophthalmology and will dramatically change Deep learning, machine learning, artificial intel-
the diagnostic and treatment pathways for ligence are projects that are based on large pools
many eye conditions such as AMD and dia- of digital retina images and serve to develop a
betic retinopathy. For example, detection of more efficient and precise system of early detec-
CNV on stereo color fundus (CF) photogra- tion, early treatment, and objective final out-
phy’s could be easily missed. CF alone could come evaluation. Reading Centers have a crucial
identify reliably pigmentary changes, drusen, role in the development of such systems and
and hemorrhage which are highly correlated to tools. A growing trend and demand in clinical
CNV [43]. Based on those highly correlated medicine are to develop standards which could
morphologic characteristics, AI systems could be practically applicable worldwide. Today
be very efficiently implemented. There are a globally there is intensive investment and devel-
variety of approaches to program AI systems opment of various devices for the screening
to automatically detect and measure patho- (including home-based OCTs) and clinical use
logic features in images of the eye. All of them in ophthalmology. For evaluating and assessing
analyze pixels or group of pixels in CF photo- real applicability, there is a need for objective,
graphs, FA images, and in OCT / OCT-A reproducible, and highly accurate data. For
images. The simplest form of AI is simple high-quality clinical trials, reading centers are
automated detectors, but recent advancements important players for initiating trials, obtaining
in the field led to basic machine learning, precious results and directions, and helping in
advanced machine learning, and deep machine submissions and regulatory filling. The future of
learning algorithms [44]. High-quality images, large, professional reading centers is to be a
based on homogenous and reliable image driver of AI-based analysis and development of
acquisition, and the high expertise of photo- more precise and more predictive diagnostic and
graphic reading centers (with their unique analysis systems for various diseases, including
asset of large image database and world-lead- but not limited to CNV. The final goal of all
ing human image analysis know-how) are an developments and investigations led by reading
essential basis for developing AI systems for center experts is to improve health care facili-
high reliability and reproducibility of such ties, support effective education of various pro-
automated output. Such systems and algo- fessional individuals involved in the process of
rithms can be included in the respective imag- screening, diagnosis, and treatment of blinding
ing devices to support decision making and to diseases and finally the prevention of blindness
guide treatment—also for not so experienced worldwide.
ophthalmologists. Recommendations for treat-
ment needs can be based on AI big data analy- Key Learning Points
sis built-in directly in the imaging device to 1. Reading centers are an essential part of suc-
support ophthalmologists in the future in their cessful drug development and regulatory
proper and efficient decision-making process approval in retinal diseases.
for further treatment and monitoring needs in 2. Large Phase 3 studies require professional
patients with, i.e., CNV and high-risk AMD. reading center structure and management.
3. Complexity in image acquisition and image apy eligibility (FLAP- study). Ophthalmology.
2003;110(2):400–5.
analysis in retinal diseases has significantly 8. Zayit-Soudry S, Alfasi M, et al. Variability among ret-
increased over the last decade. ina specialists in evaluating fluorescein angiograms of
4. Comparability of images from different patients with neovascular age-related macular degen-
devices/manufacturers is poor and need to be eration. Retina. 2007;27(6):798–803.
9. Jefferys JL, Alexander J, et al. Reproducibility of
improved with the help of large data sets from gradings of retinal photographs in eyes with subfoveal
reading centers. choroidal neovascularization and age-related macular
5. Reading centers can drive and support harmo- degeneration in the macular photocoagulation study.
nization of nomenclature and analysis of Ophthalmic Epidemiol. 2008;15(3):191–201.
10. Muni RH, Altaweel M, et al. Agreement among
newly developed imaging devices such as Canadian retina specialists in the determination
OCT-A. of treatment eligibility for photodynamic ther-
6. Reading centers can play an essential role in apy in age-related macular degeneration. Retina.
the development of automated, AI-based 2008;28(10):1421–6.
11. Abdelmoula WM, Shah SM, et al. Segmentation
image analysis algorithms. of choroidal neovascularization in fundus fluo-
7. Overall, imaging experts at reading centers rescein angiograms. IEEE Trans Biomed Eng.
support widespread use and progress in regu- 2013;60(5):1439–45.
latory acceptance and scientific advancement 12. Castillo MM, Mowatt G, et al. Optical coherence
tomography for the monitoring of neovascular age-
of retinal imaging. related macular degeneration: a systematic review.
Ophthalmology. 2015;122(2):399–406.
13. Hee MR, Baumal CR, et al. Optical coherence
tomography of age-related macular degeneration
References and choroidal neovascularization. Ophthalmology.
1996;103(8):1260–70.
1. Davis MD, Gangnon RE, Lee LY, Hubbard LD, Klein 14. Fung AE, Lalwani GA, et al. An optical coherence
BE, Klein R, Ferris FL, Bressler SB, Milton RC, tomography-guided, variable dosing regimen with
Age-Related Eye Disease Study Group. The age- intravitreal ranibizumab (Lucentis) for neovascular
related eye disease study severity scale for age-related age-related macular degeneration. Am J Ophthalmol.
macular degeneration: AREDS report no. 17. Arch 2007;143(4):566–83.
Ophthalmol. 2005;123(11):1484–94. 15. Kaiser PK, Blodi BA, et al. Angiographic and
2. Burlina PM, Joshi N, et al. Use of deep learn- optical coherence tomographic results of the
ing for detailed severity characterization and MARINA study of ranibizumab in neovascular
estimation of 5-year risk among patients with age- age-related macular degeneration. Ophthalmology.
related macular degeneration. JAMA Ophthalmol. 2007;114(10):1868–75.
2018;136(12):1359–66. 16. Wilde C, Patel M, et al. The diagnostic accuracy of
3. Malamos P, Sacu S, et al. Correlation of high-definition spectral-domain optical coherence tomography for
optical coherence tomography and fluorescein angiog- neovascular age-related macular degeneration: a com-
raphy imaging in neovascular macular degeneration. parison with fundus fluorescein angiography. Eye
Invest Ophthalmol Vis Sci. 2009;50(10):4926–33. (Lond). 2015;29(5):602–9. quiz 610
4. Wu L, Dahl AA. Choroidal Neovascularization 17. Freund KB, Ho IV, et al. Type 3 neovascularization:
(CNV) Workup. https://emedicine.medscape.com/ the expanded spectrum of retinal angiomatous prolif-
article/1190818-workup#c6 eration. Retina. 2008;28(2):201–11.
5. Moisseiev J, Alhalel A, Masuri R, Treister 18. Mathew R, Pefkianaki M, et al. Correlation of fundus
G. The impact of the macular photocoagulation fluorescein angiography and spectral-domain optical
study results on the treatment of exudative age- coherence tomography in identification of membrane
related macular degeneration. Arch Ophthalmol. subtypes in neovascular age-related macular degen-
1995;113(2):185–9. eration. Ophthalmologica. 2014;231(3):153–9.
6. Argon laser photocoagulation for neovascular macu- 19. Joeres S, Tsong JW, et al. Reproducibility of quan-
lopathy. Three-year results from randomized clinical titative optical coherence tomography subanalysis in
trials. Macular Photocoagulation Study Group. Arch neovascular age-related macular degeneration. Invest
Ophthalmol. 1986;104:694–701. Ophthalmol Vis Sci. 2007;48(9):4300–7.
7. Holz FG, Jorzik J, et al. Agreement among oph- 20. Mylonas G, Ahlers C, Malamos P, Golbaz I, Deak G,
thalmologists in evaluating fluorescein angio- Schuetze C, Sacu S, Schmidt-Erfurth U. Comparison
grams in patients with neovascular age-related of retinal thickness measurements and segmenta-
macular degeneration for photodynamic ther- tion performance of four different spectral and time
domain OCT devices in neovascular age-related 33. Sulzbacher F, Pollreisz A, et al. Identification
macular degeneration. Br J Ophthalmol. 2009 and clinical role of choroidal neovasculariza-
Nov;93(11):1453–60. tion characteristics based on optical coherence
21. Grover S, Murthy RK, Brar VS, Chalam tomography angiography. Acta Ophthalmol.
KV. Comparison of retinal thickness in normal 2017;95(4):414–20.
eyes using stratus and Spectralis optical coher- 34. Xu D, Davila JP, et al. Long-term progression of type
ence tomography. Invest Ophthalmol Vis Sci. 2010 1 neovascularization in age-related macular degenera-
May;51(5):2644–7. tion using optical coherence tomography angiogra-
22. Kakinoki M, Sawada O, Sawada T, Kawamura H, phy. Am J Ophthalmol. 2018;187:10–20.
Ohji M. Comparison of macular thickness between 35. Gong J, Yu S, et al. The diagnostic accuracy of optical
cirrus HD-OCT and stratus OCT. Ophthalmic Surg coherence tomography angiography for neovascular
Lasers Imaging. 2009;40(2):135–40. age-related macular degeneration: a comparison with
23. Sander B, Al-Abiji HA, Kofod M, Jørgensen TM. Do fundus fluorescein angiography. J Ophthalmol. 2016;
different spectral domain OCT hardwares measure the https://doi.org/10.1155/2016/7521478.
same? Comparison of retinal thickness using third- 36. Park MM, Rebhun CB, et al. Diagnosing cho-
party software. Graefes Arch Clin Exp Ophthalmol. roidal neovascularization in asymptomatic
2015 Nov;253(11):1915–21. individuals using optical coherence tomography angi-
24. Tan CS, Chan JC, Cheong KX, Ngo WK, Sadda ography. Ophthalmic Surg Lasers Imaging Retina.
SR. Comparison of retinal thicknesses measured 2017;48(7):596–8.
using swept-source and spectral-domain optical 37. Treister AD, Nesper PL, et al. Prevalence of subclini-
coherence tomography devices. Br J Ophthalmol. cal CNV and choriocapillaris nonperfusion in fellow
2015 Mar;99(3):354–8. eyes of unilateral exudative AMD on OCT angiogra-
25. Dufour PA, Abdillahi H, et al. Pathology hinting as phy. Transl Vis Sci Technol. 2018;7(5):19.
the combination of automatic segmentation with a 38. Told R, Sacu S, et al. Comparison of SD-optical
statistical shape model. Med Image Comput Comput coherence tomography angiography and Indocyanine
Assist Interv. 2012;15(Pt 3):599–606. green angiography in type 1 and 2 neovascular age-
26. Dufour PA, Ceklic L, et al. Graph-based multi- related macular degeneration. Invest Ophthalmol Vis
surface segmentation of OCT data using trained Sci. 2018;59(6):2393–400.
hard and soft constraints. IEEE Trans Med Imaging. 39. Miere A, Oubraham H, et al. Optical coherence
2013;32(3):531–43. tomography angiography to distinguish changes
27. Prahs P, Radeck V, et al. OCT-based deep learning of choroidal neovascularization after anti-VEGF
algorithm for the evaluation of treatment indica- therapy: monthly loading dose versus Pro Re Nata
tion with anti-vascular endothelial growth factor Regimen. J Ophthalmol. 2018;2018:3751702.
medications. Graefes Arch Clin Exp Ophthalmol. 40. Cole ED, Ferrara D, et al. Clinical trial endpoints for
2018;256(1):91–8. optical coherence tomography angiography in neo-
28. Schlegl T, Waldstein SM, et al. Fully automated detec- vascular age-related macular degeneration. Retina.
tion and quantification of macular fluid in OCT using 2016;36(Suppl 1):S83–92.
deep learning. Ophthalmology. 2018;125(4):549–58. 41. Munk MR, Giannakaki-Zimmermann H, et al.
29. Bogunovic H, Waldstein SM, et al. Prediction of OCT-angiography: a qualitative and quantita-
anti-VEGF treatment requirements in neovascular tive comparison of 4 OCT-A devices. PLoS One.
AMD using a machine learning approach. Invest 2017;12(5):e0177059.
Ophthalmol Vis Sci. 2017;58(7):3240–8. 42. Corvi F, Pellegrini M, Erba S, Cozzi M, Staurenghi
30. Schmidt-Erfurth U, Waldstein SM, et al. Prediction G, Giani A. Reproducibility of vessel density, fractal
of individual disease conversion in early AMD using dimension, and foveal avascular zone using 7 dif-
artificial intelligence. Invest Ophthalmol Vis Sci. ferent optical coherence tomography angiography
2018;59(8):3199–208. devices. Am J Ophthalmol. 2018;186:25–31.
31. de Carlo TE, Bonini Filho MA, et al. Spectral- 43. Mokwa NF, Ristau T, et al. Grading of age-related
domain optical coherence tomography angiography macular degeneration: comparison between
of choroidal neovascularization. Ophthalmology. color fundus photography, fluorescein angiog-
2015;122(6):1228–38. raphy, and spectral domain optical coherence
32. Liang MC, de Carlo TE, et al. Correlation of spectral tomography. J Ophthalmol. 2013; https://doi.
domain optical coherence tomography angiography org/10.1155/2013/385915.
and clinical activity in neovascular age-related macu- 44. Roach L. Artificial intelligence. EyeNet Mag 2017;
lar degeneration. Retina. 2016;36(12):2265–73. 77–83.
René Rückert earned his MD degree with distinction

from Charité Medical School, Berlin, Germany and MBA Marion R. Munk is a trained MD, PhD, Assoc. Prof.
from Warwick Business school, United Kingdom. He is a and Uveitis and Medical Retina specialist at the University
trained immunologist, and board-certified in biochemis- Clinic Bern in Switzerland and Managing Director at the
try. After many years in academic basic research he joined Bern Photographic Reading Center. She is adjunct
Bayer Healthcare in Berlin to establish the Ophthalmology Lecturer at the Northwestern University, Chicago,
franchise and to lead global development programs for USA. She did her residency and fellowships at Department
Eylea, later he led the global development clinical proof Ophthalmology, Univ. Clinic Vienna, Northwestern
grams for Lucentis at Novartis in Switzerland. Dr. Rückert University, Chicago and Univ. Clinic Bern and at
is now an independent consultant for ophthalmic drug Department of Rheumatology, Charite, Univ. Clinic of
development with his own company—eyegnos consult- Berlin. Marion is author of >90 scientific articles and
ing, based in Switzerland. book chapters. Her major research interests cover image
processing, image analyses, macular diseases, and poste-
rior uveitis. She is a member of the Editorial Board of
IOVS, Acta Ophthalmologica, Ocular Inflammation and
Infection, and BMC Ophthalmology and is providing
peer-review for a long list of scientific journals within
Ophthalmology.
Lala Ceklic MD is grader at Bern Photographer Reading

Center university of Bern, Bern Switzerland. She also
serves as consultant ophthalmologist at own private prac-
tice at Pale-E. Sarajevo, Bosnia, and Herzegovina. Her
primary research interest is primarily in retina. She has
published 18 peer-reviewed papers (https://www.ncbi.
nlm.nih.gov/pubmed/?term=lala+ceklic). She has been
awarded by Helmerich Foundation, American Academy
of Ophthalmology—Rotary foundation, International
Council of Ophthalmology and Society of Ophthalmology
Europeana. International member of American Academy
of Ophthalmology, Euretina and Association of medical
doctors of Serbia and Republic of Srpska.
Part IV
Therapy and Rehabilitation
Anti-Vascular Endothelial
Growth Molecules 24
Eduardo Tomazoni
and Eduardo Buchelle Rodrigues
24.1 Introduction retina vein occlusion (CRVO), infections and

inflammatory ocular diseases, trauma, angioid
Pathological angiogenesis and increased vascular streaks (AS), prematurity retinopathy (PR), neo-
permeability are involved with important eye dis- vascular glaucoma, high myopia, idiopathic mac-
eases like diabetic retinopathy (DR) and age- ular telangiectasia (IMT), and hereditary retinal
related macular degeneration (AMD) [1]. The diseases [4].
vascular endothelial growth factor (VEGF) fam- Choroidal neovascularization begins with reg-
ily coordinates these tissue modifications [2]. ulations of VEGF gene expression that is acti-
This chapter will cover angiogenesis mechanism, vated mostly by tissue hypoxia, cytokines, and
evolution of therapy with monoclonal antibody, cell differentiation and transformation [5]. These
monoclonal antibodies (mAbs), antibody frag- proangiogenic environments occur in association
ments, recombinant fusion protein, novel drugs, with fibroblasts, myofibroblasts, lymphocytes,
and future perspectives. and macrophages [6]. The breaks through Bruch’s
membrane lead to ingrowth of new vessels from
the choriocapillaris layer into the subretinal pig-
24.2 Angiogenesis Mechanism ment epithelial space [7].
Angiogenesis regard to a bridging new vessel
In 1971, Judah Folkman wrote that angiogenesis from existing vessels. This growth depends on
mechanism is a highly integrated ecosystem VEGF that acts signaling vessel formation and
started by pro-angiogenic factors with the aim to creating a mature and interconnected support net-
maintain a healthy organism or recover an injured work in vascular periphery [8]. A sequential acti-
tissue trough a preexisting vessel [3]. After some vation of receptors Tie1, Tie2, and platelet-derived
years, abnormal vessels growth-up was recog- growth factor (PDGF) receptor-β (PDGRF-β) by
nized as the common denominator to several sys- numerous binders in endothelial and mural cells
temic and eye diseases like: AMD, DR, central are required in the process of angiogenesis, with
VEGF representing the critical rate-limiting step
E. Tomazoni in physiological cases [9]. VEGF-A (usually
Department of Ophthalmology, Visum Retina Clinic, called only by VEGF) is a 45-kDa homodimeric
Florianópolis, SC, Brazil glycoprotein belonging to a family of cytokines
E. B. Rodrigues (*) that also includes: VEGF-B, VEGF-C, VEGF-D,
Department of Ophthalmology, Saint Louis VEGF-E, VEGF-F, and placental growth factor
University School of Medicine, (PGF) [10]. VEGF-A exists as six mRNA splice
Saint Louis, MO, USA

332 E. Tomazoni and E. B. Rodrigues
variants coded in gene 6q21.3—isoforms 121, forming growth factor (TNF)-α and β, angio-
145, 165, 183, 189, and 206 (Fig. 24.1), depend- genin, and several others that can be founded in
ing upon the number of amino acids present on Angiogenesis Foundation website: angio.org/
exons 6 and 7 [11, 12]. about-angiogenesis).
Besides VEGF, other molecules involved in The FGF family involves 20 factors catego-
angiogenesis bioassays (either directly promot- rized into six subfamilies, and display affinity to
ing endothelial cell proliferation or indirectly via surface heparan sulfate proteoglycans (HSPGs)
recruitment of inflammatory cells) include acting via receptors tyrosine kinases (FGFR 1, 2,
(Fig. 24.2): fibroblast growth factor (FGF), 3, and 4) through alternative splicing [13].
angiopoietins, epidermal growth factor, trans- Angiopoietins are a group of three structurally
linked proteins (Ang-1, 2, and 3) that acts on vas-
cular differentiation through angiogenesis and
1 2 3 4 5 6a 6b 7a 7b 8b 3’UTR maintenance of blood and lymphatic vessels [13].
VEGF206 In vivo injection of Ang-2 neutralizing antibody
(named αvβ5-integrin) inhibited astrocyte loss in
VEGF189 early diabetic retinopathy patients [14]. The
major signalling pathways in angiogenesis are
VEGF183 presented schematically in Fig. 24.3.
VEGF165
24.3 Evolution of Therapy
VEGF145
with Monoclonal Antibody
Paul Ehrilich in the early XX wrote, “By inject-

VEGF121 ing one animal with the cells of another, we can
produce substances in the serum of the first,
Fig. 24.1 The main difference between VEGF forms
resides on exons 6a, 6b, 7a, and 7b, which consists of
which have a specific damaging or destructive
heparan sulfate subunit (HSPGs) (Adapted from Ferrara influence on these cells. This possibility has
N [4]) within a short time extended the theoretical doc-
Fig. 24.2 Schematic

view of neovasculariza- Tissue hypoxia, release of cytokines and cell differentiation and transformation
tion steps
Pro angiogenic environment increasing VEGF, epidermal growth factor,

TNF-α and β, angiogenin, fibroblasts, myofibroblasts, lymphocytes,
and macrophages
Activation or receptors Tie1, Tie2, and platelet-derived growth factor (PDGF)

receptor-β (PDGRF-β) by several specific binders in endothelial and mural cells
Neovascularization
24 Anti-Vascular Endothelial Growth Molecules 333
VEGF Ang FGF IGF TG

Fβ
PDGFR-β VEGFR Tie FGFR IGFR TGFBR
p38
P13K Src ERK MAPK
SMAD PLCγ
PDGF-BB
ERK1/2 Pericyte
AKT eNOS FAK MAPK
Proliferation Cell survival Permeability Migration Proliferation
Angiogenesis
Fig. 24.3 Schematic diagram of major signalling pathways in angiogenesis. Special thanks to Patriek NM
trines of immunity in various directions.” [15]. Finally, Köhler and Milstein in 1975 described a
These models of therapy were recognized later method for the hybridoma project: cloned cell
by “passive serotherapy” that initially produced a lines producing monoclonal antibodies [18].
good response, however, after a relatively short Table 24.1 presents some examples of antigens
period of time, the patient experienced severe and their respective expressed tumors. Only after
side effects [16]. identifying the expressed antigen, the antibodies
During the 1960s and 1970s, Lloyd Old and can be produced to reach the control of the
Ted Boyse revolutionized the understanding of disease.
the immune system introducing the concept of The first therapy using antibody produced by
surface cell differentiation antigens (cluster of hybridoma project was Murine antibody, with
differentiation: CD) using cytotoxic tests and dis- target CEA and CD3 in patients with solid tumors
tinguish lineage and functional subsets of leuko- and hematologic tumors [17]. To avoid the
cytes in mice [17]. Assorted discoveries were immune response Winter and cols changed the
possible after this contribution, including the regions Gc and Fv of Murine antibodies to human
major histocompatibility complex (MHC), leuke- germline amino acids [19]. Currently, new strate-
mia, and Ly series (markers to distinguish gies are used to decrease the immune response,
between helper and cytotoxic T cells) [18]. like smaller antibodies, fusion proteins, and
Table 24.1 Antigens and their respective expressed tumors

Antigen class Antigens Tumor types expressing antigen
Cluster of CD20 Non-Hodkin lymphoma
differentiation (CD) CD30 Hodgkin lymphoma
antigens CD33 Acute myelogenous leukemia
CD52 Chronic lymphocytic leukemia
Glycolipids Gangliosides (GM2, GD2, GD3) Neuroectodermal tumors and other epithelial tumors
Carbohydrates Lewis-Y2 Epithelial tumors (breast, colon, lung, and prostate)
Glycoproteins TAG-72, EpCAM, and CEA Epithelial tumors (lung, colon, and breast)
gpA33 Colorectal carcinoma
Mucins Epithelial tumors (lung, colon, breast, and ovarian)
Carbonic anhydrase IX Renal cell carcinoma
PSMA Prostate carcinoma
Folate-binding protein Ovarian tumors
Vascular targets VEGF Tumour vascularization
VEGFR Epithelium-derived solid tumors
αVβ3 Tumor vascularization
Α5β1 Tumor vascularization
Growth factors ErbB1/EGFR Glioma, lung, breast, colon, head, and neck tumors
RrbB2/HER2 and ErbB3 Breast, colon, lung, ovarian, prostate tumors
c-MET Epithelial tumors (breast, ovary, and lung)
IGF1R Lung, breast, head and neck, prostate, thyroid, and
glioma
EphA3 Lung, kidney, colon, melanoma, glioma, and
hematological malignancies
TRAIL-R1, TRAIL-R2 Solid tumors (colon, lung, pancreas) and
hematological malignancies
RANKL Prostate cancer and bone metastases
Stromal and FAP Epithelial tumors (colon, breast, lung, head and neck,
extracellular matrix and pancreas)
antigens Tenascin Glioma and epithelial tumors (breast and prostate)
Adapted from: Scott AM, James PA, and Wolchok JD. Monoclonal antibodies in cancer therapy [17]
bispecific antibody [17]. Figure 24.4 presents the VEGFRs 1 and 2 [21, 22]. Until the publication
pathways used by antibodies against overex- of this chapter, the current pharmacological drugs
pressed cell antigens. approved by FDA to ophthalmologic use are
The first drug (or antigen) that acts against pegaptanib sodium (MACUGEN®), ranibizumab
overexpressed VEGF approved by Food and (Lucentis®), and aflibercept (Eylea®), with
Drug Administration (FDA) was pegaptanib bevacizumab (Avastin®) having his ophthalmo-
Sodium (MACUGEN®), and due to its low effec- logical uses off label for choroidal neovascular-
tiveness (acts only by blocking the forms VEGF165 ization and macular edema [22] (Fig. 24.5). As of
and VEGF188), this drug was supplanted [20, 21]. mid-2019, the company’s Biologics Licences
Ranibizumab and bevacizumab blocks six forms Application (BLA) for brolucizumab (RTH258)
of VEGF-A (VEGF121, VEGF145, VEGF165, was accepted for treatment of wet AMD by the
VEGF183, VEGF189, and VEGF206) [21]. US Food and Drug Administration (FDA).
Aflibercept blocks six types of VEGF-A, Systemic anti-VEGF agents are presented in
VEGF-B, and PGF, inhibited the activation of Table 24.2.
Fig. 24.4 Pathways

used by antibodies
against overexpressed Direct cell killing
cells’ antigens. Adapted -receptor
from Scott et al. [17] blockade or
induction of
apoptosis
Specific effects
Internalization of
on tumor
drug, radiation
vasculature Antibodies
or cytotoxic
(VEGF, VEGFR,
agent
αvβ3)
Lymphocytes T
activation
Bevacizumab and
Pegaptanib Ranibizumab Aflibercept
VEGF165 and VEGF121, VEGF145, VEGF121, VEGF145,
VEGF188 VEGF165, VEGF183, VEGF165, VEGF183,
VEGF189 and VEGF189 and
VEGF206 VEGF206, VEGF-B,
and placenta
growth factor
Fig. 24.5 Drugs and their respective site of action. acts blocking six forms of VEGF-A (121, 145, 165, 183,
Pegaptamib acts against two forms of VEGF (165 and 189, and 206), VEGF-B, and placental growth factor.
188); Bevacizumab and Ranibizumab block six forms of Adapted from Nicholas Papadopoulos et al. [21] and Anu
VEGF-A (121, 145, 165, 183, 189, and 206); Aflibercept Joseph [22]
Table 24.2 Available anti-VEGF treatment for systemic conditions

Systemic anti-VEGF
• Bevacizumab • Cabozantinib • Pazopanib • Regorafenib
• Ziv-Aflibercept • Sorafenib • Sutinib • Vandetanib
Adapted from Angiogenesis Foundation website
24.4 Monoclonal Antibodies The studies MARINA and ANCHOR prove

(mAb’s): Bevacizumab the superiority of ranibizumab under photody-
Humanized IgG1 (Avastin®) namic therapy (PDT) and placebo in patients
with DMRI [28, 29]. The EXCITE study con-
The bevacizumab (Avastin®) humanized IgG1 firms the safety and efficiency of ranibizumab
is a recombinant monoclonal antibody of [30]. The study SUSTAIN evaluates the sustained
149 kDa that binds to receptors Flt-1 and KDR improvement by ranibizumab on gains in visual
of VEGF. Bevacizumab binds to the six forms acuity [31]. Finally, the study PrONTO suggests
of VEGF-A (121, 145, 165, 183, 189, and 206). that monthly OCT reduces the necessity of injec-
Composed by human structure and complemen- tions in “treat as needed” scheme with ranibi-
tary structures of Murine antibody, the mole- zumab [32].
cule of bevacizumab is produced on ovarian
mammary cells of Chinese Hamsters in an
expression system containing the antibiotic 24.6 ecombinant Fusion Protein:
R
Gentamicin [23, 24]. Aflibercept (Eylea®)
The FDA approves bevacizumab to several
types of cancers, like colorectal metastatic carci- The aflibercept (Eylea®) is a recombinant fusion
noma, cervical cancer, peritoneal primary cancer, protein constituted by portions of human
fallopian tubes cancer, and epithelial ovarian can- VEGFR1 (Ig domain 2), VEGFR2 (Ig domain 3),
cer [23]. Although several studies have shown and extracellular domains fused to the Fc portion
satisfactory results in the treatment of ocular dis- of human IgG1. The aflibercept is formed by
eases, intravitreal injection of bevacizumab dimeric glycoprotein associated with glycosides
remains off label [24]. that result in a total molecular weight of 115 kDa
The CATT study (comparison of age-related that binds to VEGF-A, VEGF-B, and placental
macular degeneration treatment trials) suggests growth factor (PGF). The molecule is produced
the efficiency and safety of bevacizumab when in recombinant Chinese hamster ovary cells. The
compared to ranibizumab, with the similar gain drug is administered 2 mg per eye and achieves
of letters (15 letters) on both groups, and the regi- the high plasma concentration at day 2
mens “as needed” or “extended” have the same (0.081 mcg/mL), with total plasma washout after
results in one year of follow-up [25]. 2 weeks of administration. There are no signs of
the drug in the plasma after 4 weeks of intravit-
real injection. Currently, FDA approved afliber-
24.5 Antibodies Fragments: cept for the treatment of wet AMD, DME, dosing
Ranibizumab (Lucentis®) schedule (doses are given at least every 12 weeks,
and additional doses as needed in patients with
The ranibizumab (Lucentis®) is a fragment of wet AMD and DME), and diabetic retinopathy
bevacizumab with 48 kDa of weight produced by without macular edema [33].
Escherichia coli bacteria into a recipient con-
tending the antibiotic Tetracycline by the portion
Fc of bevacizumab catted out. The Fab part binds 24.7 Novel Agents
to VEGF through the VEGFR1 and VEGFR2 on
the endothelial cells’ surface. Ranibizumab also 24.7.1 Ziv-aflibercept
inhibiting the degradation products of VEGF-A
[26]. Currently, FDA approves Ranibizumab for Ziv-aflibercept (Zaltrap, Sanofi-Aventis US,
DMRI, CRVO, CME, DR, and macular choroidal LLC, Bridgewater, NewJersey, USA and
neovascularization [27]. Regeneron Pharmaceuticals, Inc., Tarrytown,
New York, USA) was approved by FDA in 2012 received intravitreal brolucizumab 3 mg
for treatment of metastatic colorectal carcinoma (HAWK) or 6 mg or aflibercept 2 mg. After 3
in specific conditions. Ziv-aflibercept is a VEGF- monthly doses, patients with brolucizumab-
trap (like the aflibercept) that acts selectively on treated eyes received an injection every
VEGF-A, VEGF-B, and PGF. The clinical trials 12 weeks if disease was not present and
using rabbits showed no complications after the 8 weeks if present. Aflibercept-treated eyes
intravitreal injection of 0.05 ml (25 mg/ml) of received a dose every 8 weeks. Patients that
ziv-aflibercept [34]. Using twice of clinical received brolucizumab 6 mg maintained on
doses, ziv-aflibercept reduces Müller cells func- 12-week dosing interval through week 48, and
tion on electroretinogram exam. This abnormal- anatomic outcomes favored brolucizumab also.
ity was not present for bevacizumab, ranibizumab, Overall safety was similar on both drugs [39].
or aflibercept [35]. In 2019 was announced by the US Food and
The first human use of ziv-aflibercept was in Drug Administration (FDA) that the compa-
a patient with refractory neovascular AMD. The ny’s Biologics License Application (BLA) for
patients received 1.25 mg (0.05 ml) of the brolucizumab (RTH258) was accepted for the
drug, and after 30 days of the injection was not treatment of wet age-related macular degenera-
any damage to the retina and visual acuity tion (AMD).
improvements were observed [36]. In 2019, a
review of prospective, retrospective, case
study, and case series trials, suggest that ziv- 24.7.3 Faricimab
aflibercept have the same safety profile, and
effectiveness in patients with diabetic macular Faricimab (developed by Roche, Basel,
edema and neovascular AMD, regarding to (1) Switzerland and Genentech, South San
reduction retinal fluid or central foveal thick- Francisco, CA, USA) is a bispecific antibody
ness and (2) maintenance or improvement of that acts simultaneously binding both VEGF-A
visual acuity [37]. and Ang- 2. Patients with diabetic macular
edema (DME) have a multifactorial angiogenic
factors and inflammatory pathways, that can
24.7.2 Brolucizumab respond to other agents besides anti-VEGF
drugs. The Ang-2 receptor stimulation increases
Brolucizumab (RTH258, Alcon Research Ltd., permeability and migration of the endothelial
a Novartis Company, Fort Worth, TX) is an cells that contribute to angiogenesis in patients
anti-
VEGF single-chain antibody fragment with DME. Blocking those two receptors
(scFv) with 26.3 kDa of weight and consisting (VEGF-A and Ang-2) possibly improves the
of 252 amino acids with activeness against all efficacy and durability during treatment of dia-
VEGF-A isoforms. In nonhuman primates, betic macular edema. The BOULEVARD study
after intravitreal infection of brolucizumab the compared faricimab with ranibizumab and dem-
drug readily reaches the RPE and choroid onstrated a statistically significant gain of 3.6
layer with minimal subsequent systemic expo- letters (P = 0.03) and central subfield thickness
sure [38]. reduction in treatment-naïve patients with far-
The study HAWK and HARRIER [39] com- icimab [40]. The clinical trials YOSEMITE and
pared brolucizumab with aflibercept to treat RHINE, until the publication of this chapter, are
neovascular AMD in patients without previous investigating the efficacy, durability and safety
treatment. The patients were randomized and of the faricimab [40].
24.7.4 Conbercept Table 24.3 Ranibizumab biosimilars under development

Ranibizumab biosimilars
The conbercept (Lumitin; Chengdu Kanghong Samsung Pfenex
Biotech CO, Ltd., Chengdu, China) is an anti- Formycon Bioepis BioCND
German Pfizer The
VEGF molecule produced in China with 141-
company company Republic
kDa of weight. Conbercept is a fusion protein of
that binds to the same receptors that aflibercept Korea-
binds (VEGF-A, VEGF-B, and Placental Growth based
company
factor), plus the fourth binding domain of
Phase 3 Phase 3 Phases 1 Phase 3
VEGFR2. The capacity of binding in the fourth clinical trial clinical and 2 clinical
domain of VEGFR2 theoreticaly extends the for trial for conducted trial of
half-life of the conbercept, decreasing VEGF lev- FYB201 in SB11. in 2016 for BCD
els over 60 days on animals’ experiments [41]. 2019. Probable PF582. 300 in
Probable release 2019.
Conbercept proves to have the same efficacy of release date: date: Probable
ranibizumab for the treatment of diabetic macu- 2020. 2020. release
lar edema [42], retinopathy of prematurity [43], date: 2020.
and macular edema secondary to branch retinal Source: https://www.formycon.com/en/press-release/for-
vein occlusion [41]. mycon-and-bioeq-achieve-important-milestone-biosimi-
lar-ranibizumab-candidate-fyb201-shows-efficacy-com-
parable-to-the-reference-product-in-phase-iii-study/,
https://clinicaltrials.gov/ct2/show/NCT03150589?term=s
24.8 Future Perspectives b11&rank=1, https://www.pfenex.com/products/, and
https://adisinsight.springer.com/drugs/800048500
In conclusion, new drugs have aloud much
progress in choroidal neovascularization.
Nowadays, there are several monoclonal anti- Key Learning Points
bodies and fusion proteins available for the A pro-angiogenic environment increases not only
management of naïve and complex cases. VEGF, but also angiopoietins and other mole-
Future drugs should further elucidate and cules involved in angiogenesis. The anti-VEGF
improve clinical outcomes. drugs are one way to reduce the choroidal neo-
In the next years, generic biosimilars could be vascularization, but more drugs with different
the mostly used anti-VEGF drugs due to a high targets are in development with the same
potential cost-efficient choice [44]. Biosimilars purpose.
are medicines manufactured by or extracted from Nowadays, overexpressed VEGF is the main
a biological source. The first patent to be expired target of the drugs. The FDA approved drugs
is for ranibizumab in 2020 for the United States anti-VEGF are ranibizumab and aflibercept, and
(US) and 2023 for Europe. There are many com- brolucizumab will probably also be approved in
panies developing biosimilars for ranibizumab the coming years.
(Table 24.3). Aflibercept has its patent expiration The Faricimab is the first available drug that
date set for the year 2020 in the United States, binds to VEGF and Angiopoietin-2, theoretically
however, there are indications that the patent may reducing the effects of choroidal neovasculariza-
be extended until the year of 2023 in the United tion with more intensity.
States and 2025 in Europe.
The biosimilars are coming in the next few αvβ5-integrin signaling in diabetic retinopathy. Cell
years with the end of the patents of ranibizumab Death Dis. 2016;7:e2101.
15. Ehrlich P. Studies in immunity. New York: Wiley;
and aflibercept, probably resulting on a lower 1910. p. p23.
price with generics. 16. Ritz J, Schlossman SF. Utilization of monoclonal anti-
bodies in the treatment of leukemia and lymphoma.
Blood. 1982;59(1):1–11.
17. Scott AM, James PA, Wolchok JD. Monoclonal
References antibodies in cancer therapy. Cancer Immun.
2012;12:p14.
1. Witmer AN, et al. Vascular endothelial growth factors 18. Moticka EJ. Learn about Ly antigen – T lympho-
and angiogenesis in eye disease. Prog Retin Eye Res. cyte subpopulations. In: Moticka EJ, editor. A his-
2003;22:1–29. torical perspective on evidence-based immunology.
2. Adamis AP, David TS, Kiang-Tech Y, Tet-Kin Y, Amsterdam: Elsevier; 2016. p. p197–205.
Lawrence FB, Brygida B, Patricia ADA, Folkman 19. Riechmann L, Clark M, Waldmann H, Winter
J. Synthesis and secretion of vascular permeability G. Reshaping human antibodies for therapy. Nature.
factor/vascular endothelial growth factor by human 1988;332:323–7.
retinal pigment epithelial cells. Biochem Biophys Res 20. Goldbaum M, Cunningham ET Jr. Aptamers and int-
Commun. 1993;193(2):631–8. ramers: pegaptanib. In: Nguyen QD, editor. Retinal
3. Folkman J. Tumor angiogenesis: therapeutic implica- pharmacotherapy, vol. 38. Maryland Heights, MO:
tions. N Engl J Med. 1971;285(21):1182–6. Elsevier; 2010. p. 265–72.
4. Garner A. Vascular diseases. In: Garner A, Klintworth 21. Ppadopoulos N, Martin J, Ruan Q, Rafique A,
GK, editors. Pathobiology of ocular diseases. Rosconi MP, Shi E, Pyles EA, Yancopoulos GD, Stahl
New York: Dekker; 1994. p. 1625–710. N, Wiegand SJ. Binding and neutralization of vas-
5. Ferrara N, Davis-Smyth T. The biology of vascular endothelial growth factor (VEGF) and related
cular endothelial growth factor. Endocr Rev. ligands by VEGF trap, ranibizumab and bevacizumab.
1997;18(1):4–25. Angiogenesis. 2012;15:171–85.
6. Killingsworth MC, Sarks JP, Sarks SH. Macrophages 22. Joseph A. Intravitreal injections: a brief note. Kerala J
related to Bruch’s membrane in age-related macular Opthalmol. 2018;30(1):63–6.
degeneration. Eye (Lond). 1990;4:613–21. 23. Avastin (bevacizumab) Information. Update 29
7. Heriot WJ, Henkind P, Bellhorn RW, Burns June 2011. https://www.fda.gov/drugs/postmarket-
MS. Choroidal neovascularization can digest Bruch’s drug-safety-information-patients-and-providers/
membrane. Opthalmology. 1984;91(12):1603–8. avastin-bevacizumab-information
8. Jain Rakesh K. Molecular regulation of vessel matu- 24. Avastin (bevacizumab) injection, for intravenous use.
ration. Nat Med. 2003;9(6):685–93. Initial US approval. 2004. Revised June 2019. https://
9. Ferrara N. Vascular endothelial growth factor: www.gene.com/download/pdf/avastin_prescribing.pdf
basic science and clinical progress. Endocr Soc. 25. The CATT Research Group. Ranibizumab and beva-
2004;25(4):581–611. cizumab for neovascular age-related maucular degen-
10. Adamis AP. Retinal Pharmacotheraphy, vol. 4. Basel: eration. N Engl J Med. 2011;364:1897–908.
Karger AG; 2010. p. 23–36. 26. Ursula S-E. Therapeutic monoclonal antibodies and
11. Harper SJ, Bates DO. VEGF-A splicing: the key fragments: ranibizumab. In: Nguyen QD, editor.
to anti-angiogenic therapeutics? Nat Rev Cancer. Retinal Pharmacotherapy, vol. 33. Maryland Heights,
2008;8:880–7. MO: Elsevier; 2010. p. 226–9.
12. Robinson CJ, Stringer SE. The splice variants of 27. LUCENTIS® (ranibizumab), for intravitreous injec-
vascular endothelial growth factor (VEGF) and their tion. Initial U.S. approval. 2006. Revised Jan 2017
receptors. J Cell Sci. 2001;114:853–65. https://www.accessdata.fda.gov/drugsatfda_docs/
13. Patriek NM, Ramya A, Howard LP, Krishna label/2017/125156s111lbl.pdf
KS. Friends turned foes: angiogenic growth factors 28. Rosenfeld PJ, Brown DM, Heier JS, Boyer DS, Kaiser
beyond angiogenesis. Biomol Ther. 2017;7:74. PK, Chung CY, Kim RY. Ranibizumab for neovascu-
14. Yun JH, Park SW, Kim JH, Park YJ, Cho CH, Kim lar age-related macular degeneration. N Engl J Med.
JH. Angiopoietin 2 induces astrocyte apoptosis via 2006;355:1419–31.
29. Brown DM, Michaels M, Kaiser PK, Heirer JS, Sy JP, 37. Barmas-Alamdari D, D’Souza HS, Kapoor
Ianchulev T, ANCHOR Study Group. Ranibizumab KG, Wagner AL. Intravitreal Ziv-Aflibercept:
versus verteporfin photodynamic therapy for neo- a comprehensive review. Semin Ophthalmol.
vascular age-related macular degeneration: two-year 2019;34(6):420–45.
results of the ANCHOR study. Ophthalmology. 2009 38. Nimz EL, Vanà Land CW, Yàñez JA, Chastain
Jan;116(1):57–65.e5. JE. Intraocular and systemic pharmacokinetics of bro-
30. Schmidt-Erfurth U, Eldem B, Guymer R, Korobelnik lucizumab (RTH258) in nonhuman primates. Program
JF, Schlingemann RO, Axer-Siegel R, Wiedmann P, number: 4996. Poster Board Number: B0202. ARVO
Simader C, Gekkieva M, Weichselberger A, EXCITE 2016 Annual Meetings Abstracts.
Study Group. Efficacy and safety of monthly versus 39. Dugel PU, Koh A, Ogura Y, Jaffe GJ, Schmidt-
quarterly ranibizumab treatment in neovascular age- Erfurth U, Brown DM, Gomes AV, Warburton J,
related macular degeneration: the EXCITE study. Weichselberger A, Holz FG. HAWK and HARRIER:
Ophthalmology. 2011 May;118(5):831–9. Phase 3, multicenter, randomized, double-masked tri-
31. Holz FG, Amoaku W, Donate J, Guymer RH, Kellner als of brolucizumab for neovascular age-related macu-
U, Schlingemann RO, Weichselber A, Staurenghi lar degeneration. Presented at: American Academy of
G, SUSTAIN Study Group. Safety and efficacy of a Ophthalmology Annual Meeting, November 11–14,
flexible dosing regimen of ranibizumab in neovascu- 2017, New Orleans, Louisiana.
lar age-related macular degeneration: the SUSTAIN 40. Sahn J, Patel SS, Dugel PU, Khanani AM, Jhaveri
study. Ophthalmology. 2011 Apr;118(4):663–71. CD, Wykoff CC, Hershberger VS, Pauly-Evers M,
32. Lalwani GA, Rosenfeld PJ, Fung AE, Dubovy Sadikhov S, Szczesny P, Schwab D, Nogoceke E,
SR, Michels S, Feuer W, Davis JL, Flynn HW Jr, Osborne A, Weikert R, Fauser S. Simultaneous inhi-
Esquiabro M. A variable-dosing regimen with intra- bition of angiopoietin-2 and vascular endothelial
vitreal ranibizumab for neovascular age-related macu- growth factor-A with faricimab in diabetic macular
lar degeneration: year 2 of the PrONTO Study. Am J edema: BOULEVARD phase 2 randomized trial.
Ophthalmol. 2009 Jul;148(1):43–58.e1. Ophthalmology. 2019 Aug;126(8):1155–70.
33. EYLEA® (aflibercept) approval history. Initial U.S. 41. Li F, Sun M, Guo J. Comparison of conbercept with
approval. 2011. https://www.drugs.com/history/eylea. ranibizumab for the treatment of macular edema sec-
html ondary to branch retinal vein occlusion. Curr Eye Res.
34. de Oliveira Dias JR, Badaró E, Novais EA, Colicchio 2017;42(8):1174–8.
D, Chiarantin GM, Matioli M. Preclinical investiga- 42. Xu Y, Rong A, Xu W, Niu Y, Wang Z. Comparison
tions of intravitreal ziv-aflibercept. Ophthalmic Surg of 12-month therapeutic effect of conbercept and
Lasers Imaging Retina. 2014;45:577–84. ranibizumab for diabetic macular edema: a real-
35. del Carpio JC, Moustafa MTM, Ramirez CA, et al. life clinical practice study. BMC Ophthalmol.
Effects of ra-nibizumab, bevacizumab, Aflibercept 2017;17:158.
and ziv-aflibercept on cultured human retinal Müller 43. Jin E, Yin H, Li X, Zhao M. Short-term outcomes
cells. Report presented at: The Annual Meeting after intravitreal injections of conberceptversus
of the Association for Research in Vision and ranibizumab for the treatment of retinopathy of pre-
Ophthalmology (ARVO) 2014–Orlando, Fla, May maturity. Retina. 2018;38(8):1595–604.
4–8, 2014. 44. Sharma A, Reddy P, Kuppermann BD, Bandello F,
36. de Oliveira Dias JR, Xavier CO, Maia A, de Moraes Lowenstein A. Biosimilars in ophthalmology: “is
NS, Meyer C, Farah ME, et al. Intravitreal injection of there a big change on the horizon?”. Clin Ophthalmol.
ziv-aflibercept in patient with refractory age-related 2018;12(12):2137–43.
macular degeneration. Ophthalmic Surg Lasers
Imaging Retina. 2015;46:91–4.
Eduardo Buchelle Rodrigues is an Assistant Professor

of Ophthalmology, Saint Louis University, Saint Louis,
USA. He graduated in medicine at the Federal University
of Santa Catarina in 1998. Prof. Buchele Rodrigues fin-
ished his doctoral and postdoctoral degrees at the Federal
University of São Paulo. He has published around 160
scientific articles in international journals of impact such
as Archives of Ophthalmology, American Journal of
Eduardo Tomazoni is a medical doctor who graduated Ophthalmology, and Retina. He is one of the editors of the
from the University of Joinville Region in 2013. He con- major international book on the use of medications in the
cluded his ophthalmology residency in 2019 at Governador treatment of retinal diseases, Retinal Pharmacotherapy,
Celso Ramos Hospital, in Florianópolis-SC, Brazil. He published in 2010. He is Co-Editor In-Chief of the
became a member of the Brazilian Ophthalmology International Journal of Retina and Vitreous. He has con-
Society in 2019. Currently, he is doing Retina and Vitreous ducted more than 250 paper presentations in national and
and Ocular Oncology fellowship at Lumine Institute, international conferences, received 19 awards in congress
Florianópolis-SC, Brazil. He is also a reviewer in the by presenting work including the prestigious Paul Kayser
International Journal of Retina and Vitreous. Award of the Retina Research Foundation through the
Pan-American Association of Ophthalmology. He is a
member of the medical societies of the area as BRAVS,
PAAO, ASRS, Club Jules Gonin, CBO, and AAO.
Surgical Interventions
25
Elizabeth D. Marlow and Tamer H. Mahmoud
25.1 Introduction visual function. Neither treatment option restored

vision nor prevented recurrent CNV with its
Age-related macular degeneration (AMD) was attendant complications.
the leading cause of irreversible severe vision Amidst a paucity of therapeutic options,
loss among adults aged more than 50 years in the efforts were devoted to developing surgical
United States prior to the emergence of anti- approaches for AMD. This chapter reviews the
vascular endothelial growth factor (VEGF) ther- evolution of vitreoretinal surgical techniques and
apy, and it remains one of the most important principles in the management of exudative AMD
causes of vision loss among Caucasians in devel- with a focus on subretinal hemorrhage, excision
oped countries [1, 2]. Nonexudative or dry macu- of submacular CNV, macular translocation, and
lar degeneration causes localized RPE transplantation of retinal pigment epithelium
dysfunction, sub-RPE drusenoid deposits, and (RPE), choroid, and retina to restore or improve
geographic atrophy [1]. The majority of severe vision. The role of adjuvants such as tPA, anti-
vision loss occurs due to the exudative or wet VEGF, and gas will also be discussed.
form of the disease, which accounts for two- Although some techniques have been aban-
thirds of AMD cases and is typified by submacu- doned with time due to limited benefit and inva-
lar choroidal neovascular (CNV) networks siveness, surgical management remains an option
leading to subretinal fluid, bleeding, and fibrotic for select AMD patients with advanced disease or
scar formation [3]. No treatment for nonexuda- subretinal hemorrhage. An understanding of the
tive AMD exists, though vitamin supplementa- historic progression toward our current surgical
tion has shown to be effective at decreasing rates practice lends valuable perspective to the con-
of progression and conversion to exudation. temporary vitreoretinal surgeon with an eye to
Prior to anti-VEGF, the primary treatment future advances such as stem cell and gene ther-
options for exudative AMD were retinal photoco- apy, which will once again emphasize subretinal
agulation for non-fovea involving CNV and pho- procedures.
todynamic therapy with verteporfin for subfoveal
CNV in eyes with good vision. Laser photoco-
agulation carried an immediate loss of vision in 25.2 Submacular Hemorrhage
exchange for the reduced long-term decline in
Submacular hemorrhage (SMH) is a rare compli-
cation of CNV that can markedly diminish vision
E. D. Marlow · T. H. Mahmoud (*) and has limited potential for spontaneous
Associated Retinal Consultants, Royal Oak, MI, USA

344 E. D. Marlow and T. H. Mahmoud
recovery [4]. The degree of visual impairment produced improvement in visual acuity [15].
caused by SMH is proportionate to the initial size Despite the value of prompt evacuation of SRH
and thickness of the bleed, and typically deterio- suggested by histologic studies, an initial case
rates over time [5]. Bennet et al. [6] showed that series of surgical removal of massive SMH and
patients with SMH due to AMD had an average associated fibrosis without tPA resulted in poor
VA of 20/1700 at final follow-up. Moreover, visual function and high rates of complication,
hemorrhage impairs absorption of laser energy including retinal detachment. The greatest visual
and thus precluded treatment of the causative acuity gains were observed in those patients
CNV in the era of laser photocoagulation [7–9]. undergoing clearance within 1 week of hemor-
All studies support that the clot volume and time rhage [16]. Given the visual recovery noted by de
to evacuation have prognostic significance and, Juan and Machemer with CNV removal, these
in general, clots should be evacuated within results highlighted that functional outcome were
1 week. dependent on minimizing surgical trauma yet
Animal models of subretinal hemorrhage limited by the underlying macular degenerative
(SRH) have elucidated multiple mechanisms changes. As methods for the management of
inflicting permanent damage to retinal photore- SMH evolved, tandem efforts were underway to
ceptors, which is the principle cause of vision advance surgical techniques for CNV excision,
loss: hemoglobin-derived iron toxicity that which is discussed later in this chapter.
worsens with the amount of hemorrhage and
duration of exposure, separation of the photore-
ceptors from underlying RPE, impaired diffu- 25.2.1 TPA-Assisted Evacuation
sion of oxygen from the choroid, shearing
effects from clot contraction [10]. Anatomic Initial techniques utilizing fluid–air exchange
changes following subretinal hemorrhage and active suction through a retinotomy to evacu-
include fibrous tissue proliferation, atrophic ate clots frequently caused inadvertent enlarge-
scar formation, and RPE tear [5]. ment of the retinotomy. The introduction of
Toth et al. [11] demonstrated in a cat model subretinal tPA in the clearance of SRH facilitated
that retinal damage from SRH is partially medi- aspiration through retinotomies; although
ated by fibrin formation leading to tearing of pho- smaller, less invasive retinotomies were
toreceptor inner and outer segments. Fibrin attempted, they required prolonged waiting time
formation and photoreceptor trauma were dimin- for tPA-induced clot liquefaction and were gener-
ished with the delivery of subretinal tissue plas- ally less successful than the larger retinotomies
minogen activator (tPA), even without that were ultimately utilized [17, 18]. Still, a
hemorrhage evacuation. In feline eyes, the use of comparison of outcomes among patients with
tPA in conjunction with a small retinotomy for large SMH due to multiple etiologies undergoing
clot evaluation showed better preservation of prompt pars plana vitrectomy (PPV) with
outer retinal architecture compared to eyes that mechanical clot extraction versus tPA-assisted
were untreated or had clots evacuated without (10–40 mcg dose) drainage showed that AMD-
adjuvant tPA [12, 13]. Tissue plasminogen acti- related hemorrhages portend a worse visual prog-
vator is well-tolerated in the subretinal space, and nosis compared to other causes, and that the use
was shown in rabbits to provide rapid clearance of tPA did not improve visual outcomes over
of SRH via a mechanism beyond simple dilu- mechanical extraction [19]. Limitations in visual
tional effects, principally by minimizing clot outcomes for all patients were proposed to be
contraction [14]. from photoreceptor shearing during clot evacua-
The origin of submacular surgery in humans tion, as elucidated in animal models, with under-
began in 1988 with Eugene de Juan and lying macular degenerative changes leading to
Machemer accessing the subretinal space to worse visual outcomes in AMD-associated hem-
remove fibrous complications of CNV, which orrhages compared to other causes.
25 Surgical Interventions 345
Intravitreal perfluorocarbon liquid (PFCL) predictably limited by the progression of under-

was applied in combination with tPA to provide lying disease and recurrent hemorrhages [28].
passive, mechanical extrusion of hemorrhage A retrospective comparison of PPV combined
through a smaller retinotomy into the vitreous with either intravitreal or subretinal administra-
cavity thereby reducing iatrogenic retinal dam- tion of tPA showed greater displacement rates
age [20]. The majority of eyes achieved when tPA was delivered directly into the subreti-
improved vision (88%) with adjuvant PFCL, nal space with equivalent VA outcomes [29]. A
although the procedure was complicated by reti- similar retrospective, nonrandomized study of 18
nal detachment (14%), epiretinal membrane for- patients by Bell et al. [30] comparing PPV with
mation (14%), and recurrent submacular subretinal tPA versus intravitreal tPA with pneu-
hemorrhage (18%). matic showed no significant difference in visual
Less invasive techniques for large SMH acuity outcomes at 1 year.
include intravitreal tPA to induce clot liquefac- A comparison of outcomes between intravit-
tion in the absence of a direct communication real tPA versus vitrectomy combined with sub-
into the subretinal space followed by pneumatic retinal tPA is limited by patient selection as many
displacement, thereby minimizing outer retinal physicians deliver clinic-based therapy for less
damage caused by clot extraction. Heriot et al. severe hemorrhages. Still, the literature to date
[21] first introduced this technique for use in an supports intravitreal tPA combined with pneu-
outpatient clinic setting in 1996. Although ani- matic as an effective and less invasive treatment
mal data did not support the capacity for tPA to option for non-massive SRH.
diffuse into the subretinal space [22], multiple It should be noted that clot liquefaction
studies in humans have shown visual acuity gains induced by intravitreal tPA is more variable and
in a majority of patients [23–25]. dose-dependent than with subretinal administra-
Subsequent studies by Haupert et al. [26] tion. Intravitreal tPA has also been shown to
showed that subretinal tPA followed by intravit- cause vitreous opacification, and at high doses
real gas provided good, and in some cases supe- (100 μg) may provoke large, self-resolving exu-
rior, displacement compared to intravitreal dative retinal detachments, which were not
tPA. This was a retrospective study of 11 eyes observed with 25 μg or 50 μg dose [31]. Recently
with thick SMH in the setting of exudative AMD tenecteplase has been explored as an alternative
who underwent vitrectomy and subretinal injec- to tPA for its superior fibrin specificity, and has
tion of 25 μg/0.1 mL or 50 μg/0.1 mL of tPA fol- shown promising early results in SMH due to
lowed by fluid gas exchange. The hemorrhage exudative AMD [32]. Future studies will be
was successfully displaced in all 11 eyes. required to compare this agent to the well-
Baseline VA was 20/200 to HM, and best postop- established efficacy of tPA.
erative VA showed improvement in eyes eyes
with four eyes being 20/80 or better. However,
final postoperative VA was not significantly dif- 25.2.2 Anti-VEGF as Monotherapy
ferent from preoperative VA, likely due to recur- and Adjuvant
rent hemorrhage (27%) and disease progression.
Olivier et al. [27] applied the same technique In the era preceding anti-VEGF, the capacity for
with lower concentrations of tPA (125 μg/mL) surgical techniques to improve visual acuity was
and larger subretinal volumes to achieve greater stunted by a high incidence of recurrent hemor-
hemorrhage displacement. With this technique, rhage and progression of underlying exudative
subfoveal blood displacement was achieved in 25 AMD. When anti-VEGF became available for
of 29 eyes with significant improvement in post- intraocular applications, its role as monotherapy
operative visual acuity at 3 months. Further stud- or in combination with pneumatic displacement
ies showed long-term visual acuity gains were were established as effective treatment options
achieved in a majority of patients, but remained for submacular hemorrhage [33]. Some studies
have reported that anti-VEGF in combination Guthoff et al. [41] compared intravitreal tPA and
with pneumatic displacement delivers superior gas versus intravitreal bevacizumab, tPA, and gas
visual outcomes compared to monotherapy for in 38 patients with recent (within 31 days) SRH
thicker hemorrhages [34]. and showed significantly better visual acuity at
Anti-VEGF has also been applied as an adju- 4 weeks among eyes that received intravitreal
vant to subretinal tPA and pneumatic tamponades anti-VEGF. This benefit persisted at 7 months
with the aim of preventing repeat bleeding and with 50% of eyes receiving anti-VEGF being
maintaining visual acuity gains from hemorrhage restored to reading vision compared to none in
displacement. A study by Chang et al. [35] the tPA and gas-only group.
showed that among 101 eyes with exudative Comparison of outcomes in retrospective
AMD undergoing vitrectomy with subretinal tPA studies of intravitreal versus subretinal tPA injec-
and gas tamponade, those eyes that received post- tion with gas when either technique is combined
operative intravitreal anti-VEGF injections with intravitreal anti-VEGF are hampered by the
showed greater visual acuity improvement at limitations previously discussed. It remains com-
6 months postoperatively compared to those eyes mon clinical practice to triage eyes with more
that did not. The rate of displacement away from severe hemorrhages to undergo vitrectomy while
the fovea was 82%, and both groups achieved less substantial bleeds are frequently managed in
significant improvement in postoperative visual clinic. A prospective study by de Jong et al. [42]
acuity with a minimum follow-up of 3 months. randomized 24 patients with recent (<14 days)
The compatibility of tPA with anti-VEGF SMH (250 μm to 1250 μm thickness at the fovea)
agents varies. Bevacizumab has proven efficacy to receive either intravitreal injection tPA, gas,
and stability when co-administered with tPA into and bevacizumab, versus vitrectomy with sub-
the subretinal space [36–38]. Klettner et al. [39] retinal tPA, gas and intravitreal bevacizumab.
studied the impact of recombinant tPA and plas- The median reduction in subretinal blood volume
min on aflibercept and ranibizumab when the at 6 weeks postoperatively was not significantly
agents were incubated together by evaluating different between the two groups: 97% in the
resultant antibody fragments with electrophore- intravitreal tPA group versus 100% in the subreti-
sis, and quantifying antiangiogenic activity with nal tPA group. Larger prospective, randomized
VEGF-ELISA. Ranibizumab was not cleaved or clinical trials will be needed to further evaluate
functionally compromised by either tPA or plas- the difference between subretinal versus intravit-
min, supporting its stability when co-applied real tPA in the era of anti-VEGF.
with tPA in the presence of plasmin for submacu-
lar hemorrhage. While aflibercept was not
cleaved or inhibited by tPA, additional fragments 25.2.3 PFCL
appeared after incubation with plasmin and its
VEGF-binding ability was inhibited by plasmin As previously mentioned, PFCL can mechani-
at clinical doses. A retrospective case series of 45 cally displace SRH into the vitreous cavity
eyes by González-López et al. [40] showed that through a small retinotomy [20], but it can also
small gauge vitrectomy with subretinal tPA and be left within the vitreous cavity postoperatively
ranibizumab for SMH secondary to AMD fol- for a period of 7–17 days to provide prolonged
lowed by pro re nata monthly intravitreal ranibi- displacement [43]. Concerns regarding the use of
zumab injections resulted in visual acuity PFCL as an intraocular tamponade for a pro-
improvement in 73% of patients, though follow- longed period include risk of inflammation and
up was limited in some cases (12.9 ± 10.8 months). increased intraocular pressure, but a limited dura-
In addition, anti-VEGF can be combined with tion of intraocular exposure can provide the ben-
tPA and pneumatic displacement without per- efits of prolonged mechanical displacement
forming vitrectomy. A retrospective analysis by without significant complications.
25.2.4 Positioning 25.2.5 Subretinal Air
With the application of expansile gas and heavy A recent innovation has been the use of subretinal
liquids for displacement of SMH, greater consid- air to provide greater SMH displacement and ear-
eration was given to optimizing the forces acting lier visual acuity improvement compared to intra-
on the hemorrhage through postoperative patient vitreal gas alone. In this technique developed by
positioning. Initial studies utilizing pneumatics Martel and Mahmoud [45], pars plana vitrectomy
universally applied prone positioning with the is followed by simultaneous injection of subreti-
thought that intravitreal gas would exert maximal nal rTPA (12.5 ug/0.1 mL, total 50 ug), 0.1 mL
upward force on the macula leading to hemor- bevacizumab (2.5 mg), and filtered air. The injec-
rhage displacement away from the fovea. tion is delivered via a 41-gauge extendable sub-
However, an analysis of the biophysical forces of retinal cannula that is inserted at the superior
intravitreal gas provided by Stopa and Lincoff margin of the SMH, and is connected to a syringe
[44] led to the adoption of upright positioning attached to the fluid control unit of the vitrectomy
with improved outcomes. machine. While the rTPA enzymatically lysis the
To summarize this analysis, one can start clot to reduce friction of RBCs on the photore-
with a fundamental understanding of ceptors, the total injected subretinal volume also
Archimedes’ principle as it applies to both liq- serves to reduce mechanical friction. The injected
uids and gases, which states that a body 0.5 mL of fluid creates space for the liquefied clot
immersed in fluid experiences a buoyant to be displaced inferiorly. The introduction of air
(upward) force equal to the weight of the fluid (typically 0.2 mL) into the subretinal space dra-
it has displaced. The relative weight of the body matically reduces the buoyancy of the hemor-
and fluid determine if the body floats, remains rhage within the subretinal space and allows
in equilibrium, or sinks. Hemorrhage and vitre- sequestration of air into the central macular
ous have nearly equal weights, and thus no dis- region. As the ratio of gravitational force to buoy-
placement occurs in the natural course of ancy is increased, so is the degree and rapidity of
SMH. When the hemorrhage is fully immersed hemorrhage displacement inferiorly. A partial
in gas (albeit separated by the neurosensory fluid–air exchange is then followed by nonex-
retina), then there is a strong driving force to pansile concentration of sulfur hexafluoride (SF6
displace the hemorrhage. In the upright posi- 20%), which will serve to keep the subretinal air
tion, gravity is acting maximally in the subreti- within the macula rather than tracking superiorly.
nal space to drive the hemorrhage downward. The patient is then positioned upright. Risks of
In contrast, there is no gravitational force act- the procedure include iatrogenic hole creation
ing to displace the hemorrhage in the prone and inadvertent displacement of SMH into the
position. The effects of intraocular pressure and fovea. In one case, a 55-year-old with baseline
gas surface tension are distributed equally VA of 20/80 who suffered at large SMH due to
across the surface of the hemorrhage and thus exudative AMD that reduced vision to count fin-
do not induce displacement. gers (CF), the above protocol effectively dis-
Of key importance is the concept that the hem- placed the SMH and vision recovered to 20/100
orrhage must be fully immersed in gas. As the within 1 month postoperatively.
volume of gas injection is limited in the absence Subsequent elaboration of this technique was
of vitrectomy, the patient’s head must be posi- provided in a prospective study by Kadonosono
tioned to provide adequate coverage of the hem- [46] in which a 47-gauge microneedle (50 μm
orrhage, while continuing to maximize outer diameter) connected to a 10-mL syringe
gravitational force parallel to the desired direc- attached to the viscous fluid control unit of the
tion of displacement (i.e., inferiorly). vitrectomy machine was used to inject a 0.4-mL
volume of TPA (62.5 μg/0.1 mL; total dose

250 μg) into the submacular space at a pressure
of 15 mmHg. In contrast to Martel and
Mahmoud’s [45] technique that simultaneously
injected a therapeutic cocktail of TPA, air, and
anti-VEGF, Kadosono techniques allow the tPA
to sit for one minute, after which 0.4 mL filtered
air is injected with 4–6 mmHg pressure. No anti-
VEGF was used, and a fluid air exchange was not
performed. Patients maintained prone position-
ing overnight. This technique allowed subfoveal
blood displacement in all eyes, and visual acuity
improved by over two lines in 11 of 13 eyes with
average gains being significantly improved com- Fig. 25.1 Postoperative day 1 image from a patient with
pared to baseline at 1 and 3 months subretinal air, intravitreal gas, inferior subretinal fluid,
and displaced subretinal hemorrhage. The intravitreal gas
postoperatively. prevents superior migration of the subretinal air while the
A retrospective study examined ten eyes with patient is in the upright position, allowing for maximum
SMH due to AMD that had undergone vitrectomy displacement of the subretinal hemorrhage. Text quoted
followed by submacular injection of TPA from original publication [48]
(12.5 ug/0.1 mL), bevacizumab (2.5 mg/0.1 mL
and air (0.3 mL) followed by gas tamponade with In the same study by Sharma et al. [48], none
20% SF6 and postoperative upright positioning. of the surgeons had prior experience injecting
At 6 month follow-up, visual acuity improved in subretinal air. Macular hole formation is a poten-
80% of patients, and bleed occurred in two eyes, tial complication of any subretinal injection near
which was treated with intravitreal tPA, anti- the macula, and patients with SMH are predis-
VEGF, and 20% SF6 [47]. posed due to the effect of SMH on foveal tissue.
In 2018, the initial findings of a multicenter One of the 24 eyes included developed an intra-
retrospective study of 24 eyes undergoing dis- operative macular hole where the subretinal air
placement of SMH due to exudative AMD of pol- was delivered superior to the macula.
ypoidal choroidal vasculopathy with vitrectomy, Complications on subsequent cases were avoided
subretinal injections of air and tPA (125 mg/mL) by administering the subretinal injection closer to
followed by partial fluid–air exchange and gas the inferior arcade and directing the injection
tamponade coupled with either preoperative, toward the inferior retina (Fig. 25.2). This allows
intraoperative, or postoperative intravitreal injec- the subretinal air and tPA to move into the infe-
tion of anti-VEGF were published [48]. Complete rior periphery and does not expose the fovea to
displacement of SMH from the foveal center was the mechanical force of injection, when the risk
achieved in all eyes, with 75% having displace- of macular hole formation is greatest. The air
ment outside the arcades and additional 20% then moves passively to displace the hemorrhage
beyond the equator. Visual functional signifi- when the patient is upright.
cantly improved in 95.8% of eyes postopera-
tively. In addition, 74% of sub-RPE hemorrhages
were also displaced, which was noted to be more 25.2.6 Emerging Techniques
efficient in cases of smaller sub-RPE and sub-
macular hemorrhage. In massive SRH, subretinal Recent advances in subretinal surgery include the
air does not adequately localize to the height of creation of increasingly small retinotomies, auto-
the sub-RPE hemorrhage to provide mechanical mated injection techniques, and new small gauge
displacement and flattening in prone positioning intravitreal devices that can deliver subretinal
(Fig. 25.1). therapy without the need for vitrectomy.
a b
c d
Fig. 25.2 Depiction of subretinal air injection from sur- mechanical trauma to the fovea during injection, which
geon’s view. (a) Submacular hemorrhage. (b) Subretinal poses the greatest risk for iatrogenic macular hole forma-
delivery device is oriented toward the inferior periphery. tion. The air then moves passively to displace the hemor-
(c, d) The subretinal air and pharmacologic agents are rhage when the patient is upright
delivered near the inferotemporal arcade, avoiding
Injection of pharmacologic agents or gas into The most recent advancement to facilitate
the subretinal space may be performed manually delivery of pharmacologic agents and air into the
with an assistant pushing an insulin syringe subretinal space is the Nano Subretinal Gateway
plunger while the surgeon holds the syringe body. Device developed by Mahmoud et al. [50]. It
Flexible tubing can be attached to the subretinal consists of an external very thin needle-like tip
cannula to minimize the translation of surgeon and an inner extendable, flexible cannula 25%
tremor to instruments in the subretinal space, but smaller than the 41G (Fig. 25.3). The outer nee-
great care must still be taken to avoid iatrogenic dle has a sharp tip capable of penetrating the
trauma or induction of a macular hole with rapid sclera through pars plana to establish intraocular
subretinal injection. Recent studies have shown access without a traditional sclerotomy with a
better control of automated administration via cannula. Once inside the vitreous cavity, the
41-gauge cannula and insulin syringe coupled to internal cannula is extended when the needle is
the viscous fluid control of the standard vitrec- close to the retinal surface, and its beveled tip
tomy systems, typically with 6–10 PSI [48, 49]. creates a retinotomy for access to the subretinal
The controlled introduction of liquids and gas space. Pharmacologic agents are delivered
into the subretinal space may minimize photore- through automated injection as described above,
ceptor trauma. and the retinotomy is self-sealing. Visualization
The surgical excision of CNVM has been per-

formed for presumed ocular histoplasmosis
(POHS), AMD, and other causes of CNV. The
removal of subfoveal CNV provided substantial
visual acuity gains in POHS and non-AMD
related causes, but more typically leads to stabi-
lization or mild visual improvement in AMD
[53–55]. The relative visual acuity gains in
POHS as compared to AMD were attributed to
Fig. 25.3 Nano Subretinal Gateway Device. A thin differences in CNV architecture and localiza-
needle-like tip capable of penetrating sclera through the tion. The CNV in POHS is typically type 2,
pars plana. Once inside the eye, an inner, flexible cannula existing anterior to the RPE. In contrast, AMD
is extended and penetrates into the subretinal space for
delivery of pharmacologic agents. The inner cannula is often has diffuse changes with type 1 choroidal
25% smaller than the 41G and provides a self-sealing reti- neovascular membranes (CNVM) residing pos-
notomy [50] terior to the RPE [55, 56]. Surgical removal of
CNVM is invariably associated with traumatic
for intraocular maneuvers is provided by chande- loss of the RPE and disruption of the photore-
lier illumination. ceptor–RPE complex. Visual recovery following
CNV excision in AMD is dependent on the
integrity of the subfoveal RPE, with diminished
25.3 Excision of Choroidal functional improvements observed in cases
Neovascular Membranes where the RPE has been incorporated into
mature CNVM. The use of TPA before surgical
As previously noted, the functional gains in the excision of subfoveal membranes has shown no
management of CNV-related complications are benefit in mitigating this limitation [14].
often limited by the progression of underlying Maintenance of long-term visual improve-
macular degeneration. Photocoagulation of sub- ments in AMD remained limited by CNVM
foveal CNV diminished visual acuity loss over recurrence with a rate of 20% over a 2-year fol-
time with the distinct drawback of immediately low- up period, and rarely produced reading
decreased vision at the time of treatment, as vision [57]. Consequently, criteria for patient
shown in the Macular Photocoagulation Study selection for surgical excision versus laser photo-
Group [7]. Excision of subfoveal CNV in an coagulation were considered, and a surgical
effort to restore vision and stop disease progres- approach was primarily utilized in more advanced
sion was first reported by Thomas et al. [51] in disease with low baseline visual acuity (generally
1992. The surgical technique utilized in Thomas’ under 20/200), large CNVM occupying the
and similar studies published in the same year foveal avascular zone, and exudative lesions [56].
involved a pars plana vitrectomy, induction of Following CNVM excision, decreased choroi-
posterior vitreal detachment, and the creation of dal perfusion observed on fluorescein and indo-
200 μm retinotomy with diathermy, typically cyanine green angiography was indicative of
temporal to the fovea. A blunt 30G needle was choriocapillaris atrophy that may have contrib-
used to inject subretinal balanced salt solution to uted to limited visual acuity outcomes. In con-
create a localized neurosensory retinal detach- trast, large and medium choroidal vessels
ment. Once in the subretinal space, an angled remained perfused [58]. The cause of choriocap-
pick facilitated dissection of the CNV from the illaris atrophy was unknown, but may have been
RPE and was then extracted through the retinot- related to the underlying disease process or surgi-
omy with forceps. Following fluid–air exchange, cal trauma. Histologic analysis of surgically
the retinotomy was sealed with diathermy or removed tissue showed the universal presence of
laser photocoagulation [51, 52]. RPE, but an infrequent appearance of choriocap-
illaris [58, 59]. It was proposed that the removal the degree of macular rotation [63, 64]. This
of overlying RPE may cause subsequent choroi- modification, which was termed “Limited
dal atrophy with proposed mechanisms including Macular Translocation” involved only 15–25
inadequate RPE repopulation, abnormalities of degree rotation of the retina, and was associated
the repopulated RPE cells, or permanent changes with fewer postoperative complications such as
to the denuded Bruch’s membrane [58, 59] retinal detachment and PVR [65]. In this proce-
The submacular surgery trials were the semi- dure, a partial thickness scleral resection was cre-
nal study that compared complications and visual ated in the supratemporal or infratemporal
acuity changes in eyes undergoing surgical equatorial region, followed by a retinotomy to
removal versus laser photocoagulation of recur- induce a subtotal retinal detachment, before
rent extrafoveal or juxtafoveal CNV due to neo- finally suturing the edges of the resected sclera.
vascular AMD. In this prospective, randomized, Others performed scleral imbrication or infolding
multicenter study by Bressler et al. [60], no sig- rather than resection [65–67]. Either approach
nificant advantage to submacular surgery in terms served to shorten the length of the globe in one
of visual acuity or size of the central macular meridian so that the fovea no longer rested on the
lesions was found. A total of 70 patients were CNVM when the retina was reattached [68].
enrolled and among these, 65% (31 eyes) in the Visual acuity gains were seen in 40% of eyes, but
laser arm and 50% (14 eyes) of the surgical group persistence and recurrence of CNVM remained
had visual acuity at 2 years after enrollment that common causes of vision loss. A review of a
was better or no more than one line worse than at visual acuity gains following limited macular
baseline. The similarity in ophthalmic outcomes translocation showed that the foveal sensory ret-
in the two treatment arms was also reflected in ina does recover when overlaid onto healthy RPE
health-related quality of life metrics [61] up to 5 years after surgery when combined with
CNVM excision or photocoagulation [69]. As
with full macular translocation, metamorphopsia,
25.4 Foveal Translocation and image tilt remained common complaints and
complications included PVR, complex RD,
25.4.1 Exudative AMD epiretinal membrane, macular hole, corneal
astigmatism, and constricted visual field [70, 71].
Given the limitations in visual acuity gains fol- The macular translocation with 360-degree
lowing CNVM removal in exudative AMD that retinectomy (MTS360) carried the advantage of
was attributed in part to loss of foveal RPE avoiding scleral dissection while providing
function, an alternative therapeutic intervention access to the subretinal space for CNVM exci-
was to relocate the foveal retina to overlie sion and evacuation of submacular hemorrhage.
healthy RPE. In other studies, the CNV was treated with macu-
The initial strategy for macular translocation lar photocoagulation rather than excision [67].
involved a PPV, the creation of a total retinal Considering the profound changes in RPE–
detachment with 360-degree retinectomy and retina anatomy and frequent torsional diplopia
excision of the CNVM, followed by rotation of the following macular translocation, the functional
foveal neural retina onto adjacent, intact RPE, and outcomes were surprisingly good. Long-term
subsequent silicone oil tamponade. The first dem- follow-up of MTS360 retinectomy and silicone
onstration of this technique by Machemer in 1993 oil tamponade showed stabilization or improve-
[62] involved three patients in whom the fovea was ment in visual acuity and reading speed for most
rotated between 30 to 80 degrees to overlay healthy patients. At 12 months after MTS360, a majority
RPE. One patient demonstrated improvement in of patients maintained their preoperative vision,
vision, albeit with excyclotorsion of images. and in a study by Abdel et al. [72], 33% of
Future elaborations on this surgical technique patients showed improvement (three or more
varied the extent of the retinectomy and reduced lines logMar) [73]). Another study by Aisenbrey
et al. [71] included patients with submacular VEGF. Among the small number of patients
hemorrhage and showed similar outcomes at 1 recruited, no significant difference was seen in
year. Among eyes undergoing MTS360 with the surgical versus observation or PDT arm [82].
compensatory muscle surgery followed for an Another small study comparing full macular
average of 21 months, two-thirds of patients translocation vs PDT in the treatment of neovas-
experienced visual acuity gains consistent with cular AMD also showed lack of difference
reading vision and nearly one-third showed sta- between the two methods [83]. With the intro-
ble vision (within 1 Snellen line); deterioration of duction of anti-VEGF, macular translocation was
vision was observed infrequently (6%) [74]. used with decreasing frequency and only for
Overall, a majority of patients with 20/80 vision select patients.
or better at baseline were likely to retain this level
of vision 12 months after surgery [75]. Central
visual fields also improved following macular 25.4.2 Nonexudative AMD
translocation [76].
Cyclotropia and diplopia affected approxi- In the evolution of surgical techniques to man-
mately 50% of patients and was effectively miti- age AMD, few therapeutic options had emerged
gated by extraocular muscle surgery, either at the for the nonexudative form of AMD. Geographic
time of surgery or early in the postoperative atrophy (GA) accounts for severe vision loss in
period [72, 77]. Infrared eye tracking showed that 40% of patients with AMD [84]. Vitamin sup-
the majority of eyes had improvement in central plementation in the form of AREDS was intro-
scotoma following macular translocation and duced in 1991, and no other therapeutic
adequate foveal function to produce a single, intervention has shown the ability to slow or
stable focus for fixation. Saccadic function was stabilize disease progression [85]. The concept
impaired, but oculomotor limitations did not sig- of foveal translocation away from the area of
nificantly impede reading behavior [78]. RPE atrophy held the attractive possibility of
In all of these techniques, recurrence of improving visual function for patients with non-
CNVM remained a common complication that exudative AMD.
curtailed long-term gains in visual function. Macular translocation was applied to nonexu-
Recurrences were noted to have a predisposition dative AMD and GA and showed visual acuity
to involve the newly repositioned fovea, which gains [76, 86]. As GA has been reported to have
led to the hypothesis that the development of an average growth rate of 139 microns per year in
CNV may occur via foveal signaling [79], though one direction [86, 87], the macula had to be trans-
this was not supported by later studies. As located a sufficient distance to avoid the spread of
expected, choroidal perfusion in individuals who atrophy to involve the fovea at its new site. The
underwent macular translocation without CNV posterior retina was typically rotated about 40
excision showed no change in choroidal perfu- degrees upward until the fovea was positioned
sion following surgery [80]. Nonetheless, long- approximately two-thirds of a disk diameter
term follow-up showed that half of the patients away from the border of the RPE lesion.
who underwent MTS360 had a stable or improved Geographic atrophy was shown to recur subfove-
vision at 3 years [81]. ally in a variable portion of patients with nonexu-
The Macular Relocation in Age-related dative AMD who underwent macular
Neovascular disease (MARAN) trial sought to translocation [88]. A comparison of outcomes for
assess the efficacy of macular relocation in neo- MTS360 for the management of exudative versus
vascular AMD as compared to observation or nonexudative AMD showed that the prevalence
photodynamic therapy (PDT). Limited recruit- of postoperative subfoveal RPE atrophy was
ment was attributed to inherent risks of the surgi- higher in the nonexudative group compared to
cal arm as well as the emergence of new the exudative group due to recurrence of GA
pharmacologic treatment options, namely anti- lesions [89]. There was not a significant d ifference
in mean preoperative or postoperative VA comparison of grafts consisting of RPE-choroid

between nonexudative versus exudative AMD sheets versus RPE cell suspension overlaying
groups. healthy choroid showed comparable anatomic
and functional outcomes [95].
25.5 Transplantation
25.5.2 Transplantation of Alternative
25.5.1 RPE Transplantation Tissue
As techniques to excise choroidal neovascular Alternative sources of tissue to replace RPE were
complexes elaborated in the early 1990s, the also considered. Transplantation of iris pigment
importance of RPE integrity to visual recovery epithelium (IPE) after removal of CNV demon-
became increasingly apparent, thus directing strated that autologous IPE cells are tolerated in
efforts toward the simultaneous transplantation the subretinal space and could be considered in
of RPE. The fundamental rationale is to transport cases with diffuse RPE degeneration [96, 97].
healthy RPE and choroid from the periphery to Human fetal RPE (15–17 week gestational) was
the fovea. As the technique has evolved, multiple also cultured and transplanted as a monolayer
studies have shown its capacity to stabilize vision into the subretinal space with viability for up to 3
loss, improve visual acuity, and increase reading months [98]. Still, the majority of RPE transplan-
speed for several years following surgery. tations for the management of neovascular AMD
The initial techniques created a retinal flap to utilize an autologous graft harvested from the
access the subfoveal RPE then used an autolo- periphery.
gous pedicle RPE flap that was rotated into place
or homologous RPE cells and Bruch’s mem-
brane, which was ultimately covered with the 25.5.3 RPE–Choroid Transplantation
same retinal flap [90]. Alternatively, RPE grafts in Wet AMD
harvested from the edge of the RPE defect where
CNV was excised were transplanted onto the The observation of choriocapillaris atrophy in
denuded Bruch’s membrane with the neurosen- eyes that underwent CNV excision, and the dem-
sory retina relaid on top; these RPE grafts exhib- onstrated visual improvements with RPE trans-
ited autofluorescence and patients demonstrated plantation, paved the way for the creation of
improved central visual function with the ability RPE–choroid transplants. The first subfoveal
to fixate [91]. Larger series showed that best- RPE-choroid transplantation of a graft harvested
corrected visual acuity remained improved in from the nasal midperiphery was performed in
over half of eyes, and was stable in 35% over an 2003. An average of two Snellen line visual acu-
average of 17 months follow-up [92]. ity increase and graft survival was noted 1 year
However, the encouraging results of studies postoperatively [99]. Although harvest sites for
with a follow-up of 2 years or less were not per- RPE–choroid grafts were generally from the
petuated at longer intervals. Although macular nasal periphery, RPE–choroidal translocation
translocation of the RPE demonstrated anatomi- with pedicle flap to maintain a connection with
cal longevity up to 5–6 years, there was ultimate adjacent choroidal blood supply was also suc-
loss of foveal fixation and graft autofluorescence cessfully demonstrated [100]. There is no corre-
[93]. lation between superior or inferior periphery
Transplantation of suspended peripheral RPE donor sites and the incidence of PVR after trans-
cells using poly-L-lysine to promote adhesion in plantation [101].
the bed of the choroid after CNV extraction was One large series of 84 eyes undergoing autolo-
technically feasible but associated with high rates gous free RPE–choroid grafts showed modest
of PVR and no functional improvement [94]. A visual acuity gains that continued to slightly
improve up to 4 years after surgery, with fixation of eyes, including those with GA, and that visual
located on the graft [102]. Long-term follow-up outcomes were not clearly related to the type
showed that RPE–choroid grafts in eyes with AMD [109]. A separate study of ten patients
exudative AMD were sustained at 7 years with demonstrated no recurrence of RPE atrophy with
minimal complication rates and CNVM recur- graft survival at 3 years, but limited visual benefit
rence under 10% [103]. Full-thickness RPE– due to postoperative complications and patient
choroid grafts were also successfully transplanted selection. Notably, individuals with extrafoveal
[104]. However, functional outcomes in eyes that fixation preoperatively were not able to convert
underwent successful autologous RPE–choroid to foveal fixation following transplantation
transplantation appeared limited by overlying despite graft viability [110].
neuroretinal damage [105].
A smaller series in exudative AMD showed
more heterogeneous results with frequent postop- 25.5.5 Retina–RPE–Choroid
erative complications [106]. The variable results Transplants
between studies highlight the lack of standardiza-
tion for patient selection to undergo RPE–cho- Subretinal surgery for exudative AMD leads to
roid transplants, and the uncertainty of what damage to RPE and Bruch’s membrane and vari-
preoperative features of exudative AMD (e.g., able retinal trauma, which may compound preex-
submacular hemorrhage) may predispose to isting outer retinal atrophy from AMD. Visual
favorable postoperative outcomes. Preoperative acuity following successful RPE-choroid trans-
factors shown to carry better postsurgical prog- plantation is limited by this neurosensory dam-
nosis were predominantly classic and occult age [104]. Autologous transplantation of
lesions; in contrast, minimally classic or large neurosensory retina has demonstrated efficacy in
(>50%) hemorrhagic lesions showed less visual the management of large, refractory macular
acuity improvement [102]. holes with the capacity to integrate with the adja-
Histologic evidence of angiogenesis leading cent neurosensory retina and improve vision
to revascularization of autologous RPE–choroid [111–113].
grafts has been demonstrated in porcine models The combination of this technique with the
with bridging vessels between the recipient layer RPE–choroid graft in advanced exudative AMD
and graft identified from 1 week to 3 months after was first presented by Parolini et al. [114] who
surgery [107]. This integration occurred regard- reported on outcomes in nine eyes where the
less of the integrity of Bruch’s membrane. RPE–choroid and retinal grafts were placed sub-
Angiographic studies assessing revascularization retinal (n = 5), intraretinal to cover an area of an
of the RPE–choroid graft in patients showed that iatrogenically induced macular hole (n = 1), and
the vast majority of grafts revascularized as early preretinal (n = 1). In this study, the average fol-
as 1 week after surgery and persisted up to 3 low-up was 7 ± 5.5 months and the visual acuity
years [108]. changed from being an average of CF preopera-
tively to approximately 20/800 postoperatively
(Fig. 25.4). Vision was stable in five eyes and
25.5.4 RPE–Choroid Transplantation improved in four eyes. Complications included
in Dry AMD graft atrophy (n = 1) and dislocation (n = 1).
Future studies to establish the long-term survival
Similar work in nonexudative AMD showed that of these grafts the visual function improvements
RPE–choroid grafts vascularized in the majority they offer remain to be established [114].
a b
c d
Fig. 25.4 Patient with advanced exudative AMD (a) with round spaces because of the dilatation of choroidal vessels
RPE atrophy in the macular region (white arrow). OCT (c) inside the graft (white arrows). The neurosensory retinal
showing lamellar MH configuration with a very thin resid- graft (yellow arrow) seems well integrated into the MH
ual foveal floor, significant loss of outer retinal layers and and is partly lying over the original retina, separated by a
RPE atrophy (red arrow), overlying intraretinal fluid, and hyperreflective line. Postoperative OCT angiography (e,
subretinal hyporeflective spaces. Postoperative photo- scan area 66 mm) 3 months after surgery. The choroidal
graph 1 month after surgery (b) showing the choroidal– graft (yellow arrows) is visualized on the en face images
RPE graft in the center of the macular region and the at the level of the outer retina and choriocapillaris. The
neurosensory retinal graft covering the MH (yellow superficial and deep retinal vasculature seems grossly pre-
arrow). The temporal harvest site of the neurosensory reti- served in the area of the original overlying neurosensory
nal graft is shown with the white asterisk. Postoperative retina (white arrows); however, the scan is limited by arti-
OCT 3 months after surgery (d) showing RPE–choroidal fact and segmentation errors. Text quoted from original
patch in the macula area, with rounded hyporeflective publication [114]
25.6 Conclusion lopathy: the Beaver Dam Eye Study. Ophthalmology.

1997;104:7–21.
3. la Cour M, Kiilgaard JF, Nissen MH. Age-related
Efforts directed at the surgical management of macular degeneration: epidemiology and optimal
exudative AMD have spurred profound advances treatment. Drugs Aging. 2002;19:101–33.
in vitreoretinal surgery technology and tech- 4. Avery RL, Fekrat S, Hawkins BS, Bressler
NM. Natural history of subfoveal subretinal hemor-
nique. With the emergence of anti-VEGF therapy rhage in age-related macular degeneration. Retina.
for the treatment of exudative AMD, surgical 1996;16:183–9.
management has become limited to select 5. Scupola A, Coscas G, Soubrane G, Balestrazzi
patients. Fortunately, the medical management of E. Natural history of macular subretinal hemorrhage in
age-related macular degeneration. Ophthalmologica.
AMD in our current era is minimally invasive and 1999;213:97–102.
provides stable, improved vision for a majority of 6. Bennett SR, Folk JC, Blodi CF, Klugman M. Factors
patients. Thick large submacular hemorrhage dis- prognostic of visual outcome in patients with subreti-
placement remains the most common surgical nal hemorrhage. Am J Ophthalmol. 1990;109:33–7.
7. Macular Photocoagulation Study Group. Laser pho-
indication for complications of neovascular tocoagulation of subfoveal neovascular lesions of
AMD, with better anatomical and functional out- age-related macular degeneration. Updated find-
comes using newer techniques. In patients with ings from two clinical trials. Arch Ophthalmol.
advanced disease, retina–RPE–choroid trans- 1993;111:1200–9.
8. Virgili G, Bini A Laser photocoagulation for neovas-
plantation remains a treatment option that will cular age-related macular degeneration. Cochrane
continue to be refined through surgical device Database Syst Rev CD004763. 2007.
innovation and less invasive techniques. 9. Macular Photocoagulation Study Group. Subfoveal
neovascular lesions in age-related macular degenera-
tion. Guidelines for evaluation and treatment in the
Key Learning Points macular photocoagulation study. Arch Ophthalmol.
• Displacement rather than removal of submac- 1991;109:1242–57.
ular hemorrhage improves visual recovery and 10. Glatt H, Machemer R. Experimental subretinal hemor-
is facilitated by delivery of subretinal TPA and rhage in rabbits. Am J Ophthalmol. 1982;94:762–73.
11. Toth CA, Morse LS, Hjelmeland LM, Landers MB
gas. 3rd. Fibrin directs early retinal damage after experi-
• Newer techniques of displacement using sub- mental subretinal hemorrhage. Arch Ophthalmol.
retinal TPA, subretinal air, with and without 1991;109:723–9.
antiVEGF results in 100% displacement far 12. Benner JD, Hay A, Landers MB 3rd, Hjelmeland LM,
Morse LS. Fibrinolytic-assisted removal of experi-
away from the fovea, preserving the central mental subretinal hemorrhage within seven days
visual field. reduces outer retinal degeneration. Ophthalmology.
• Minimally invasive, nanovitreoretinal gate- 1994;101:672–81.
way devices allow subretinal access for deliv- 13. Toth CA, Benner JD, Hjelmeland LM, Landers
MB 3rd, Morse LS. Ultramicrosurgical removal of
ery of pharmacologic agents with less subretinal hemorrhage in cats. Am J Ophthalmol.
iatrogenic trauma and may circumvent the 1992;113:175–82.
need for vitrectomy. 14. Lewis H, Resnick SC, Flannery JG, Straatsma
• Transplantation of retina, RPE, and choroidal BR. Tissue plasminogen activator treatment of experi-
mental subretinal hemorrhage. Am J Ophthalmol.
grafts for improved central visual acuity is an 1991;111:197–204.
area of active development in the surgical 15. de Juan E Jr, Machemer R. Vitreous surgery for hemor-
management of exudative AMD. rhagic and fibrous complications of age-related macu-
lar degeneration. Am J Ophthalmol. 1988;105:25–9.
16. Vander JF, Federman JL, Greven C, Slusher MM,
Gabel VP. Surgical removal of massive subretinal
References hemorrhage associated with age-related macular
degeneration. Ophthalmology. 1991;98:23–7.
1. Bressler NM, Bressler SB, Fine SL. Age- related 17. Claes C, Zivojnovic R. Efficacy of tissue plasmino-
macular degeneration. Surv Ophthalmol. 1988;32: gen activator (t-PA) in subretinal hemorrhage removal.
375–413. Bull Soc Belge Ophtalmol. 1996;261:115–8.
2. Klein R, Klein BE, Jensen SC, Meuer SM. The five- 18. Peyman GA, Nelson NC Jr, Alturki W, Blinder KJ,
year incidence and progression of age-related macu- Paris CL, Desai UR, Harper CA 3rd. Tissue plasmino-
gen activating factor assisted removal of subretinal plasminogen activator and gas. Graefes Arch Clin Exp
hemorrhage. Ophthalmic Surg. 1991;22:575–82. Ophthalmol. 1999;237:273–7.
19. Ibanez HE, Williams DF, Thomas MA, Ruby AJ, 32. Lee JP, Park JS, Kwon OW, You YS, Kim
Meredith TA, Boniuk I, Grand MG. Surgical manage- SH. Management of Acute Submacular Hemorrhage
ment of submacular hemorrhage. A series of 47 con- with intravitreal injection of Tenecteplase, anti-
secutive cases. Arch Ophthalmol. 1995;113:62–9. vascular endothelial growth factor and gas. Korean J
20. Kamei M, Tano Y, Maeno T, Ikuno Y, Mitsuda H, Ophthalmol. 2016;30:192–7.
Yuasa T. Surgical removal of submacular hemorrhage 33. Cho HJ, Koh KM, Kim JH, Kim HS, Han JI, Lew
using tissue plasminogen activator and perfluorocar- YJ, Lee TG, Kim JW. Intravitreal ranibizumab injec-
bon liquid. Am J Ophthalmol. 1996;121:267–75. tions with and without pneumatic displacement for
21. Heriot WJ. Intravitreal gas and TPA: an out- treating submacular hemorrhage secondary to neo-
patient procedure for submacular hemorrhage. vascular age-related macular degeneration. Retina.
Case Rep Ophthalmol Med. 1996:4. https://doi. 2015;35:205–12.
org/10.1155/2016/9809583. 34. Shin JY, Lee J-M, Byeon SH. Anti-vascular endothe-
22. Kamei M, Misono K, Lewis H. A study of the ability lial growth factor with or without pneumatic displace-
of tissue plasminogen activator to diffuse into the sub- ment for submacular hemorrhage. Am J Ophthalmol.
retinal space after intravitreal injection in rabbits. Am 2015;159:904–14.e1.
J Ophthalmol. 1999;128:739–46. 35. Chang W, Garg SJ, Maturi R, Hsu J, Sivalingam A,
23. Hassan AS, Johnson MW, Schneiderman TE, Regillo Gupta SA, Regillo CD, Ho AC. Management of thick
CD, Tornambe PE, Poliner LS, Blodi BA, Elner submacular hemorrhage with subretinal tissue plas-
SG. Management of submacular hemorrhage with minogen activator and pneumatic displacement for
intravitreous tissue plasminogen activator injec- age-related macular degeneration. Am J Ophthalmol.
tion and pneumatic displacement. Ophthalmology. 2014;157:1250–7.
1999;106:1900–6. discussion 1906–7 36. Shah SP, Hubschman JP, Gonzales CR, Schwartz
24. Chen CY, Hooper C, Chiu D, Chamberlain M, Karia SD. Submacular combination treatment for manage-
N, Heriot WJ. Management of submacular hemorrhage ment of acute, massive submacular hemorrhage in
with intravitreal injection of tissue plasminogen activa- age-related macular degeneration. Ophthalmic Surg
tor and expansile gas. Retina. 2007;27:321–8. Lasers Imaging. 2009;40:308–15.
25. Hattenbach LO, Klais C, Koch FH, Gümbel 37. Klettner A, Puls S, Treumer F, Roider J, Hillenkamp
HO. Intravitreous injection of tissue plasminogen J. Compatibility of recombinant tissue plasminogen
activator and gas in the treatment of submacular hem- activator and bevacizumab co-applied for neovascu-
orrhage under various conditions. Ophthalmology. lar age-related macular degeneration with submacular
2001;108:1485–92. hemorrhage. Arch Ophthalmol. 2012;130:875–81.
26. Haupert CL, McCuen BW 2nd, Jaffe GJ, Steuer 38. Hillenkamp J, Klettner A, Puls S, Treumer F, Roider
ER, Cox TA, Toth CA, Fekrat S, Postel EA. Pars J. Subretinal co-application of rtPA and bevaci-
plana vitrectomy, subretinal injection of tissue zumab for exudative AMD with submacular hemor-
plasminogen activator, and fluid-gas exchange rhage. Compatibility and clinical long-term results.
for displacement of thick submacular hemor- Ophthalmologe. 2012;109:648–56.
rhage in age-related macular degeneration. Am J 39. Klettner A, Grotelüschen S, Treumer F, Roider J,
Ophthalmol. 2001;131:208–15. Hillenkamp J. Compatibility of recombinant tissue
27. Olivier S, Chow DR, Packo KH, MacCumber MW, plasminogen activator (rtPA) and aflibercept or ranibi-
Awh CC. Subretinal recombinant tissue plasminogen zumab coapplied for neovascular age-related macu-
activator injection and pneumatic displacement of lar degeneration with submacular haemorrhage. Br J
thick submacular hemorrhage in age-related macular Ophthalmol. 2015;99:864–9.
degeneration. Ophthalmology. 2004;111:1201–8. 40. González-López JJ, McGowan G, Chapman E, Yorston
28. Kamei M, Tano Y. Tissue plasminogen activator- D. Vitrectomy with subretinal tissue plasminogen acti-
assisted vitrectomy: surgical drainage of submacular vator and ranibizumab for submacular haemorrhages
hemorrhage. Dev Ophthalmol. 2009;44:82–8. secondary to age-related macular degeneration: ret-
29. Hillenkamp J, Surguch V, Framme C, Gabel V-P, rospective case series of 45 consecutive cases. Eye.
Sachs HG. Management of submacular hemorrhage 2016;30:929–35.
with intravitreal versus subretinal injection of recom- 41. Guthoff R, Guthoff T, Meigen T, Goebel
binant tissue plasminogen activator. Graefes Arch Clin W. Intravitreous injection of bevacizumab, tissue plas-
Exp Ophthalmol. 2010;248:5–11. minogen activator, and gas in the treatment of submac-
30. Bell JE, Shulman JP, Swan RJ, Teske MP, Bernstein ular hemorrhage in age-related macular degeneration.
PS. Intravitreal versus subretinal tissue plasmino- Retina. 2011;31:36–40.
gen activator injection for submacular hemor- 42. de Jong JH, van Zeeburg EJT, Cereda MG, van
rhage. Ophthalmic Surg Lasers Imaging Retina. Velthoven MEJ, Faridpooya K, Vermeer KA, van
2017;48:26–32. Meurs JC. Intravitreal versus subretinal administra-
31. Hesse L, Schmidt J, Kroll P. Management of acute tion of recombinant tissue plasminogen activator com-
submacular hemorrhage using recombinant tissue bined with gas for acute submacular hemorrhages due
to age-related macular degeneration: an exploratory veal neovascular membranes in age-related macular

prospective study. Retina. 2016;36:914–25. degeneration. Ophthalmology. 1994;101:1201–10.
43. Fleissig E, Barak A, Goldstein M, Loewenstein A, 58. Thach AB, Marx JL, Frambach DA, LaBree LD, Lopez
Schwartz S. Massive subretinal and subretinal pig- PF. Choroidal hypoperfusion after surgical excision of
ment epithelial hemorrhage displacement with per- subfoveal neovascular membranes in age-related mac-
fluorocarbon liquid using a two-step vitrectomy ular degeneration. Int Ophthalmol. 1996;20:205–13.
technique. Graefes Arch Clin Exp Ophthalmol. 59. Nasir MA, Sugino I, Zarbin MA. Decreased chorio-
2017;255:1341–7. capillaris perfusion following surgical excision of cho-
44. Stopa M, Lincoff A, Lincoff H. Analysis of forces act- roidal neovascular membranes in age-related macular
ing upon submacular hemorrhage in pneumatic dis- degeneration. Br J Ophthalmol. 1997;81:481–9.
placement. Retina. 2007;27:370–4. 60. Bressler NM, Bressler SB, Hawkins BS, Marsh MJ,
45. Martel JN, Mahmoud TH. Subretinal pneumatic Sternberg P Jr, Thomas MA, Submacular Surgery
displacement of subretinal hemorrhage. JAMA Trials Pilot Study Investigators. Submacular surgery
Ophthalmol. 2013;131:1632–5. trials randomized pilot trial of laser photocoagulation
46. Kadonosono K, Arakawa A, Yamane S, Inoue M, versus surgery for recurrent choroidal neovasculariza-
Yamakawa T, Uchio E, Yanagi Y. Displacement of sub- tion secondary to age-related macular degeneration:
macular hemorrhages in age-related macular degener- I. Ophthalmic outcomes submacular surgery tri-
ation with subretinal tissue plasminogen activator and als pilot study report number 1. Am J Ophthalmol.
air. Ophthalmology. 2015;122:123–8. 2000;130:387–407.
47. Kumar A, Roy S, Bansal M, Tinwala S, Aron N, 61. Bressler NM, Bressler SB, Hawkins BS, Marsh MJ,
Temkar S, Pujari A. Modified approach in manage- Sternberg P Jr, Thomas MA, Submacular Surgery
ment of submacular hemorrhage secondary to wet age- Trials Pilot Study Investigators. Submacular surgery
related macular degeneration. Asia Pac J Ophthalmol. trials randomized pilot trial of laser photocoagulation
2016;5:143–6. versus surgery for recurrent choroidal neovasculariza-
48. Sharma S, Kumar JB, Kim JE, Thordsen J, Dayani P, tion secondary to age-related macular degeneration:
Ober M, Mahmoud TH. Pneumatic displacement of II. Quality of Life Outcomes Submacular Surgery
submacular hemorrhage with subretinal air and tissue Trials Pilot Study report number 2. Am J Ophthalmol.
plasminogen activator: initial United States experi- 2000;130:408–18.
ence. Oph Retina. 2018;2:180–6. 62. Machemer R, Steinhorst UH. Retinal separation,
49. Novelli FJD, Preti RC, Monteiro MLR, Nóbrega MJ, retinotomy, and macular relocation: II. A surgi-
Takahashi WY. A new method of subretinal injection cal approach for age-related macular degenera-
of tissue plasminogen activator and air in patients with tion? Graefes Arch Clin Exp Ophthalmol. 1993;231:
submacular hemorrhage. Retina. 2017;37:1607–11. 635–41.
50. Mahmoud T. Nanovitreoretinal subretinal gate- 63. Fujikado T, Ohji M, Kusaka S, Hayashi A, Kamei
way device to displace submacular hemor- M, Okada AA, Oda K, Tano Y. Visual function after
rhage: access to the subretinal space without foveal translocation with 360-degree retinotomy and
vitrectomy. Retina. 02/2019; https://doi.org/10.1097/ simultaneous torsional muscle surgery in patients with
IAE.0000000000002669. myopic neovascular maculopathy. Am J Ophthalmol.
51. Thomas MA, Grand MG, Williams DF, Lee CM, 2001;131:101–10.
Pesin SR, Lowe MA. Surgical management of sub- 64. Ninomiya Y, Lewis JM, Hasegawa T, Tano
foveal choroidal neovascularization. Ophthalmology. Y. Retinotomy and foveal translocation for surgi-
1992;99:952–68. discussion 975–6 cal management of subfoveal choroidal neovascular
52. Skaf AR, Mahmoud T. Surgical treatment of age- membranes. Am J Ophthalmol. 1996;122:613–21.
related macular degeneration. Semin Ophthalmol. 65. Fujii GY, de Juan E Jr, Pieramici DJ, Humayun
2011;26:181–91. MS, Phillips S, Reynolds SM, Melia M, Schachat
53. Berger AS, Kaplan HJ. Clinical experience with the AP. Inferior limited macular translocation for sub-
surgical removal of subfoveal neovascular membranes. foveal choroidal neovascularization secondary to
Short-term postoperative results. Ophthalmology. age-related macular degeneration: 1-year visual
1992;99:969–75. discussion 975–6 outcome and recurrence report. Am J Ophthalmol.
54. Benson MT, Callear A, Tsaloumas M, Chhina J, 2002;134:69–74.
Beatty S. Surgical excision of subfoveal neovascular 66. Mateo C, Moreno J, Rosales G, Lechuga M, Castillo
membranes. Eye. 1998;12(Pt 5):768–74. R, Vaz F, Corcóstegui B. Two-year results of macu-
55. Thomas MA, Dickinson JD, Melberg NS, Ibanez HE, lar translocation with scleral infolding in myopic
Dhaliwal RS. Visual results after surgical removal choroidal neovascularisation. Semin Ophthalmol.
of subfoveal choroidal neovascular membranes. 2004;19:29–42.
Ophthalmology. 1994;101:1384–96. 67. Pieramici DJ, De Juan E Jr, Fujii GY, Reynolds SM,
56. Sabates FN, Fletcher DC. Surgical excision of subfo- Melia M, Humayun MS, Schachat AP, Hartranft
veal neovascular membranes in age-related macular CD. Limited inferior macular translocation for the
degeneration. Am J Ophthalmol. 1992;114:241–2. treatment of subfoveal choroidal neovascularization
57. Ormerod LD, Puklin JE, Frank RN. Long-term out- secondary to age-related macular degeneration. Am J
comes after the surgical removal of advanced subfo- Ophthalmol. 2000;130:419–28.
68. de Juan E Jr, Loewenstein A, Bressler NM, Alexander 82. Joussen AM, Wong D, Walter P, et al. Surgical manage-
J. Translocation of the retina for m
anagement of sub- ment of subfoveal choroidal neovascular membranes
foveal choroidal neovascularization II: a preliminary in age-related macular degeneration by macular relo-
report in humans. Am J Ophthalmol. 1998;125:635–46. cation: experiences of an early-stopped randomised
69. Oshima H, Iwase T, Ishikawa K, Yamamoto K, clinical trial (MARAN Study). Eye. 2010;24:284–9.
Terasaki H. Long-term results after limited macu- 83. Gelisken F, Voelker M, Schwabe R, Besch D,
lar translocation surgery for wet age-related macular Aisenbrey S, Szurman P, Grisanti S, Herzau V, Bartz-
degeneration. PLoS One. 2017;12:e0177241. Schmidt KU. Full macular translocation versus photo-
70. Ohji M, Fujikado T, Saito Y, Hosohata J, Hayashi A, dynamic therapy with verteporfin in the treatment of
Tano Y. Foveal translocation: a comparison of two neovascular age-related macular degeneration: 1-year
techniques. Semin Ophthalmol. 1998;13:52–62. results of a prospective, controlled, randomised pilot
71. Aisenbrey S, Lafaut BA, Szurman P, et al. Macular trial (FMT-PDT). Graefes Arch Clin Exp Ophthalmol.
translocation with 360 degrees retinotomy for exu- 2007;245:1085–95.
dative age-related macular degeneration. Arch 84. Owen CG, Fletcher AE, Donoghue M, Rudnicka
Ophthalmol. 2002;120:451–9. AR. How big is the burden of visual loss caused by age
72. Abdel-Meguid A, Lappas A, Hartmann K, Auer F, related macular degeneration in the United Kingdom?
Schrage N, Thumann G, Kirchhof B. One year follow Br J Ophthalmol. 2003;87:312–7.
up of macular translocation with 360 degree retinot- 85. Age-Related Eye Disease Study Research Group.
omy in patients with age related macular degeneration. The age-related eye disease study (AREDS): design
Br J Ophthalmol. 2003;87:615–21. implications. AREDS report no. 1. Control Clin Trials.
73. Toth CA, Freedman SF. Macular translocation with 1999;20:573–600.
360-degree peripheral retinectomy impact of tech- 86. Sunness JS, Gonzalez-Baron J, Applegate CA,
nique and surgical experience on visual outcomes. Bressler NM, Tian Y, Hawkins B, Barron Y, Bergman
Retina. 2001;21:293–303. A. Enlargement of atrophy and visual acuity loss in
74. Pertile G, Claes C. Macular translocation with 360 the geographic atrophy form of age-related macular
degree retinotomy for management of age-related degeneration. Ophthalmology. 1999;106:1768–79.
macular degeneration with subfoveal choroidal neo- 87. Schatz H, McDonald HR. Atrophic macular degenera-
vascularization. Am J Ophthalmol. 2002;134:560–5. tion: rate of spread of geographic atrophy and visual
75. Mruthyunjaya P, Stinnett SS, Toth CA. Change in loss. Ophthalmology. 1989;96:1541–51.
visual function after macular translocation with 360 88. Khurana RN, Fujii GY, Walsh AC, Humayun MS,
degrees retinectomy for neovascular age-related macu- de Juan E Jr, Sadda SR. Rapid recurrence of geo-
lar degeneration. Ophthalmology. 2004;111:1715–24. graphic atrophy after full macular translocation for
76. Eckardt C, Eckardt U. Macular translocation in non- nonexudative age-related macular degeneration.
exudative age-related macular degeneration. Retina. Ophthalmology. 2005;112:1586–91.
2002;22:786–94. 89. Cahill MT, Mruthyunjaya P, Bowes Rickman C,
77. Eckardt C, Eckardt U, Conrad HG. Macular rota- Toth CA. Recurrence of retinal pigment epithelial
tion with and without counter-rotation of the globe changes after macular translocation with 360 degrees
in patients with age-related macular degeneration. peripheral retinectomy for geographic atrophy. Arch
Graefes Arch Clin Exp Ophthalmol. 1999;237:313–25. Ophthalmol. 2005;123:935–8.
78. Uppal G, Feely MP, Crossland MD, Membrey L, Lee 90. Peyman GA, Blinder KJ, Paris CL, Alturki W, Nelson
J, da Cruz L, Rubin GS. Assessment of reading behav- NC Jr, Desai U. A technique for retinal pigment epi-
ior with an infrared eye tracker after 360° macular thelium transplantation for age-related macular degen-
translocation for age-related macular degeneration. eration secondary to extensive subfoveal scarring.
Invest Ophthalmol Vis Sci. 2011;52:6486–96. Ophthalmic Surg. 1991;22:102–8.
79. Baer CA, Rickman CB, Srivastava S, Malek G, Stinnett 91. Stanga PE, Kychenthal A, Fitzke FW, Halfyard AS,
S, Toth CA. Recurrent choroidal neovascularization Chan R, Bird AC, Aylward GW. Retinal pigment
after macular translocation surgery with 360-degree epithelium translocation after choroidal neovascular
peripheral retinectomy. Retina. 2008;28:1221–7. membrane removal in age-related macular degenera-
80. Cekiç O, Ohji M, KesercI B, Sawa M, Zheng Y, Hayashi tion. Ophthalmology. 2002;109:1492–8.
A, Ikuno Y, Nakata K, Gomi F, Tano Y. Evaluation of 92. Binder S, Stolba U, Krebs I, et al. Transplantation
choroidal perfusion of the new central macular area of autologous retinal pigment epithelium in eyes
by dilution analysis of indocyanine green angiog- with foveal neovascularization resulting from age-
raphy after macular translocation. Retina. 2004;24: related macular degeneration: a pilot study. Am J
210–4. Ophthalmol. 2002;133:215–25.
81. Aisenbrey S, Bartz-Schmidt KU, Walter P, Hilgers 93. MacLaren RE, Bird AC, Sathia PJ, Aylward
RD, Ayertey H, Szurman P, Thumann G. Long-term GW. Long-term results of submacular surgery
follow-up of macular translocation with 360 degrees combined with macular translocation of the retinal
retinotomy for exudative age-related macular degen- pigment epithelium in neovascular age-related macu-
eration. Arch Ophthalmol. 2007;125:1367–72. lar degeneration. Ophthalmology. 2005;112:2081–7.
94. van Meurs JC, ter Averst E, Hofland LJ, van Hagen lar space. Graefes Arch Clin Exp Ophthalmol.
PM, Mooy CM, Baarsma GS, Kuijpers RW, Boks T, 2010;248:37–47.
Stalmans P. Autologous peripheral retinal pigment epi- 105. Parolini B, Alkabes M, Baldi A, Pinackatt S. Visual
thelium translocation in patients with subfoveal neo- recovery after autologous retinal pigment epithe-
vascular membranes. Br J Ophthalmol. 2004;88:110–3. lium and choroidal patch in a patient with choroi-
95. Falkner-Radler CI, Krebs I, Glittenberg C, Povazay dal neovascularization secondary to ANGIOID
B, Drexler W, Graf A, Binder S. Human retinal pig- streaks: long-term results. Retin Cases Brief Rep.
ment epithelium (RPE) transplantation: outcome 2016;10:368–72.
after autologous RPE-choroid sheet and RPE cell- 106. Degenring RF, Cordes A, Schrage NF. Autologous
suspension in a randomised clinical study. Br J translocation of the retinal pigment epithelium
Ophthalmol. 2011;95:370–5. and choroid in the treatment of neovascular age-
96. Thumann G, Aisenbrey S, Schraermeyer U, related macular degeneration. Acta Ophthalmol.
Lafaut B, Esser P, Walter P, Bartz-Schmidt 2011;89:654–9.
KU. Transplantation of autologous iris pigment 107. Maaijwee KJM, van Meurs JC, Kirchhof B,
epithelium after removal of choroidal neovascular Mooij CM, Fischer JH, Mackiewicz J, Kobuch K,
membranes. Arch Ophthalmol. 2000;118:1350–5. Joussen AM. Histological evidence for revascu-
97. Lappas A, Foerster AMH, Weinberger AWA, larisation of an autologous retinal pigment epithe-
Coburger S, Schrage NF, Kirchhof B. Transloca- lium--choroid graft in the pig. Br J Ophthalmol.
tion of iris pigment epithelium in patients with 2007;91:546–50.
exudative age-related macular degeneration: long- 108. Maaijwee K, Van Den Biesen PR, Missotten T,
term results. Graefes Arch Clin Exp Ophthalmol. Van Meurs JC. Angiographic evidence for revas-
2004;242:638–47. cularization of an rpe-choroid graft in patients
98. Algvere PV, Berglin L, Gouras P, Sheng with age-related macular degeneration. Retina.
Y. Transplantation of fetal retinal pigment epithe- 2008;28:498–503.
lium in age-related macular degeneration with sub- 109. Joussen AM, Heussen FMA, Joeres S, Llacer H,
foveal neovascularization. Graefes Arch Clin Exp Prinz B, Rohrschneider K, Maaijwee KJM, van
Ophthalmol. 1994;232:707–16. Meurs J, Kirchhof B. Autologous translocation of
99. van Meurs JC, Van Den Biesen PR. Autologous reti- the choroid and retinal pigment epithelium in age-
nal pigment epithelium and choroid translocation in related macular degeneration. Am J Ophthalmol.
patients with exudative age-related macular degen- 2006;142:17–30.
eration: short-term follow-up. Am J Ophthalmol. 110. Caramoy A, Liakopoulos S, Menrath E, Kirchhof
2003;136:688–95. B. Autologous translocation of choroid and reti-
100. Angunawela RI, Williamson TH, Khan MAJ, Chong nal pigment epithelium in geographic atrophy:
V. Choroidal translocation with a pedicle following long-term functional and anatomical outcome. Br J
excision of a type 1 choroidal neovascular mem- Ophthalmol. 2010;94:1040–4.
brane. Br J Ophthalmol. 2005;89:386. 111. Grewal DS, Mahmoud TH. Autologous neuro-
101. van Zeeburg EJT, Maaijwee K, van Meurs JC. There sensory retinal free flap for closure of refrac-
is no relation between the occurrence of prolifera- tory myopic macular holes. JAMA Ophthalmol.
tive vitreoretinopathy and the location of the donor 2016;134:229–30.
site after transplantation of a free autologous retinal 112. Thomas AS, Mahmoud TH. Subretinal transplanta-
pigment epithelium-choroid graft. Acta Ophthalmol. tion of an autologous retinal free flap for chronic
2014;92:228–31. retinal detachment with proliferative vitreoreti-
102. Maaijwee K, Heimann H, Missotten T, Mulder P, nopathy with and without macular hole. Retina.
Joussen A, van Meurs J. Retinal pigment epithe- 2018;38(Suppl 1):S121–4.
lium and choroid translocation in patients with 113. Grewal DS, Charles S, Parolini B, Kadonosono
exudative age-related macular degeneration: long- K, Mahmoud TH. Autologous retinal transplant
term results. Graefes Arch Clin Exp Ophthalmol. for refractory macular holes: multicenter interna-
2007;245:1681–9. tional collaborative study group. Ophthalmology.
103. van EJT Z, KJM M, TOAR M, Heimann H, van 2019;126:1399. https://doi.org/10.1016/j.
Meurs JC. A free retinal pigment epithelium- ophtha.2019.01.027.
choroid graft in patients with exudative age-related 114. Parolini B, Grewal DS, Pinackatt SJ, Baldi A,
macular degeneration: results up to 7 years. Am J Di Salvatore A, Besozzi G, Finzi A, Cardillo D,
Ophthalmol. 2012;153:120–7.e2. Mahmoud TH. Combined autologous transplanta-
104. Cereda MG, Parolini B, Bellesini E, Pertile tion of neurosensory retina, retinal pigment epithe-
G. Surgery for CNV and autologous choroidal lium and choroid free grafts. Retina. 2018;38(Suppl
RPE patch transplantation: exposing the submacu- 1):S12–22.
many surgical techniques including the use of subretinal

air to displace submacular hemorrhage and performed the
first human autologous retinal transplantation. He also
introduced chandelier buckles to the United States and
developed the ILM retracting door technique for macular
holes. In 2019, he developed the Nanovitreoretinal
Surgery Platform.
Mahmoud finished his internship, residency, and vit-
reoretinal fellowship at Duke. He is a member of the
AAO, ARVO, ASRS, the Club Jules Gonin, the Macula
Society, and the Retina Society. He received the Edward
K. Isbey, Jr., M.D. award for “Excellence in Clinical Care,
Ethics, and Research” from Duke, the Retina Research
Elizabeth D. Marlow M.D. is completing her fellow- Foundation Award from ARVO, the “Senior Honor
ship in vitreoretinal surgery at Associated Retinal Award” from ASRS, The “Achievement Award” from the
Consultants/William Beaumont Hospital with plans to AAO, the prestigious Robert A. Machemer Research
practice in the San Francisco Bay Area. She is originally Award from Duke, the “Distinguished Teacher of the Year
from Washington State and completed her ophthalmology Award” from the Kresge Eye Institute and the “Golden
residency at Weill Cornell Medical Center/New York– Globe Award” for residents’ education from Duke.
Presbyterian Hospital. Interests include diabetic eye dis-
ease and complex retinal detachment repair.
Tamer H. Mahmoud is a professor of Ophthalmology

at Oakland University William Beaumont School of
Medicine. He was the program director of the vitreoretinal
surgery fellowship at the Duke University Eye Center and
a tenured associate professor at Duke. He has developed
Home Monitoring for Age-Related
Macular Degeneration 26
Voraporn Chaikitmongkol
26.1 Introduction ity at one year [6]. Therefore, early recognition of

CNV development when visual acuity is least
Age-related macular degeneration remains the affected is crucial for AMD treatment. Delay in
leading cause of central visual loss and blindness treatment may lead to irreversible damage and
in individuals aged 50 or older in North America poor visual outcome [9].
and other regions worldwide [1]. The primary eti- Patients at risk to develop neovascular AMD
ology of visual deterioration in neovascular AMD are those with intermediate stage of AMD, either
eyes associated with the development of choroi- bilateral large drusen (125 μm or larger in diam-
dal neovascularization (CNV). Patients with neo- eter) or large drusen in one eye and advanced
vascular AMD could experience various visual AMD (neovascular AMD or central geographic
symptoms in their central visual field, i.e. blurry atrophy) in the fellow eye, whom typically are
vision, metamorphopsia, and scotoma. scheduled for fundus evaluations every 6–12
Intravitreous anti-VEGF injection has been dra- months or only 1–2 times per year. Due to the
matically improved visual functions of eyes with longer durations between clinic visits in these
neovascular AMD [2–5]. Ninety percent of neo- patients, recognition of visual function changes
vascular AMD patients receiving anti-VEGF by the patients themselves is very important as
therapy could avoid moderate or more severe such recognition would urge them to be evaluated
visual loss over a 2-year treatment with anti- by an ophthalmologist sooner than regular sched-
VEGF. However, not more than one-third of ules. Then, they will be able to receive prompt
treated patients achieved moderate visual gain treatment if there is a presence of CNV and result
[2–5]. The main predictors for favorable treat- in excellent visual acuity. The home-monitoring
ment outcomes include the presenting level of device for AMD, therefore, plays an essential
visual acuity and size of the CNV lesion at the role in assisting patients in detecting subtle
time of treatment initiation [6–8]. According to changes in their visual functions at an early stage
the CATT subgroup analysis, eyes presenting of CNV development.
with good baseline visual acuity (20/25 to 20/40) The idea to detect an early stage of macular
were eyes that achieved the best mean visual acu- abnormalities has been established since 1940s
when the Amsler grid was developed by Mark
Amsler. Subsequently, the Amsler grid has
become the only self-monitoring tool used world-
V. Chaikitmongkol (*) wide for more than 70 years despite an absence
Retina Division, Department of Ophthalmology, of effective treatment for AMD. Until the past
Chiang Mai University, Chiang Mai, Thailand

364 V. Chaikitmongkol
decade when there are advancements in

therapeutic options for neovascular AMD, the

importance of an early detection of CNV has
been re-emphasized and many home-monitoring
devices for AMD have been invented.
This chapter will describe the testing princi-
ples and scientific evidence on the validity of
each home-monitoring device for AMD includ-
ing the Amsler grid, preferential hyperacuity
perimeter (ForeseeHome device), shape discrim-
ination hyperacuity test (myVisionTrack), and
other self-measured visual function tests designed
for personal electronic devices, e.g., smartphone
or tablet device, which are recently available.
26.2 The Amsler Grid

Fig. 26.1 The original Amsler grid (white lines on a
The Amsler grid was first introduced by Marc black background)
Amsler, a Swiss ophthalmologist, in 1947 and
subsequently publicized around 1950 as a method
for macular function testing [10]. Amsler had line, blurry or missing square, or greyish or
been interested in developing methods for blackish spot.
patients to detect early symptoms of maculopa- The Amsler grid testing is a simple, inexpen-
thy, i.e., metamorphopsia and relative scotoma, sive, and practical testing method that patients
which typically fails to be revealed by the perim- can be used easily at home. However, there
eter and the Bjerrum screen [11]. Amsler designed were studies suggesting poor validity of the
a square-shaped grid as his testing tool “since it is Amsler grid testing for detection macular
a matter of making apparent some kind of defor- pathologies [12, 13]. Regarding scotoma detec-
mation, we must present to the patient a precise tion, Schuchart et al. evaluated the Amsler grid
and regular contour which forms as simple a fig- testing in 55 patients with vision loss in the
ure as possible, the alterations of which he can macular region and ten normally sighted sub-
access without difficulty” [11]. jects. Results revealed nearly half of the stan-
The original Amsler grid was drawn as a dard and threshold scotomas were not detected
10-cm. square white grids on a black background by the Amsler grid testing. In addition, for sco-
(Fig. 26.1). The square was subdivided every tomas of six degrees or less in diameter, 77% of
5 mm. by vertical and horizontal parallel lines. standard and 87% of threshold scotomas were
Each small 5-mm. square subtends an angle of not detected by the Amsler grid [13]. The poor
one degree at 28–30 cm. distance. The entire grid validity of the Amsler Grid may be a result of
is thus 20 degrees high and 20 degrees wide [11]. difficulty with fixation and the perceptual com-
The field spans ten degrees from the fixation cen- pletion phenomenon. The perceptual comple-
ter. To start testing, simply place the Amsler grid tion phenomenon is an ability of the brain to
28–30 cm. from an eye under good illumination. “fill in” a scotoma with the same color and tex-
Refractive errors, if presents, should be corrected ture as the surrounding background, as similar
with reading glasses or contact lens. The patient to visual perceptions when part of an image
is instructed to close one eye, look at the center of falls on a blind spot of the visual field.
the grid, and asked if they notice any distorted Therefore, it appears that the Amsler grid is less
26 Home Monitoring for Age-Related Macular Degeneration 365
efficient in assessing the extent of scotomas, 26.3 Home-Based Preferential

while may still be useful for early detection of Hyperacuity Perimeter (PHP)
metamorphopsia [14].
Modified versions of the original Amsler grid The controversial role of the Amsler grid in
have been developed, including a version with detecting CNV in AMD leads to development of
white lines on a black background. However, the PHP. Hyperacuity (or vernier acuity) is an
Augustin et al. evaluated a role of this modified ability to perceive a difference in the spatial
Amsler grid in 182 patients with dry and exuda- localization of two or more stimuli [17]. This
tive age-related macular degeneration and a his- property would permit the detection of a meta-
tory of metamorphopsia, compared with the morphopsia caused by the displacement of pho-
original Amsler grid, and found that for overall toreceptors by the CNV or by the separation of
study population, the original Amsler grid pro- the neurosensory retina or the retinal pigment
vided significantly better results than the modi- epithelium related to neovascular AMD [18]. The
fied version (p < 0.05). For patients with visual PHP device presents artificial images with differ-
acuity ranging from 0.1 to 0.3 (20/63 to 20/200 ent degrees of distortion. Due to brain preferen-
Snellen equivalent), there was no significant dif- tial awareness, the device user can detect more
ference across groups. However, for patients noticeable distortions and overlook the smaller
with visual acuity of 0.5 (20/40 Snellen equiva- ones. Regarding this brain mechanism, a quanti-
lent) or better, the original Amsler grid was sig- fied visual field map of the metamorphopsia for
nificantly more informative and reliable than the each eye was created along with longitudinal
modified version [15]. Therefore, in case that the detections of visual field changes as well as devi-
Amsler grid is still considered as part of home- ation from the normative data. The first genera-
monitoring tools in clinical practice, the original tion of devices using PHP technology was
Amsler grid (white lines on a black background) designed for supervised use in a clinical setting
might be preferred. and previously shown to have high sensitivity
Recently, a systematic review and meta- and specificity in detecting CNV [19].
analysis by Faes et al. had focused on the diag- Following several modifications of the origi-
nostic accuracy of the Amsler grid and the nal PHP device, including decreasing in its phys-
preferential hyperacuity perimeter (PHP) in the ical size; enclosure of the screen viewer in a
screening of patients with AMD. Results closed hood in order to control distance from the
revealed the pooled sensitivity of studies assess- display, ambient light conditions, and occlusion
ing the Amsler grid was 0.78 (95% confidence of the non-tested eye; and addition of infrared
intervals (CI); 0.64–0.87), and the pooled speci- sensors to ensure correct positioning of the head,
ficity was 0.97 (95%CI; 0.91–0.99). The pooled a home device or ForeseeHome was developed
sensitivity of studies assessing the PHP was 0.85 [19] for use in the detection and characterization
(95%CI; 0.80-0.89), and specificity was 0.87 of central and paracentral metamorphopsia in
(95%CI; 0.82-0.91). These results suggest prom- patients with AMD, as an aid in monitoring pro-
ising test performance characteristics of both the gression of disease factors causing metamor-
Amsler grid and the PHP in terms of specificity. phopsia including but not limited to CNV. The
However, the Amsler grid has a relatively lower ForeseeHome device has been approved by the
sensitivity in detecting neovascular AMD when US FDA in 2009 and becomes the first FDA
compared with the PHP [16]. Also, the Amsler approved home-monitoring device for AMD.
grid does not provide quantifiable measures of The ForeseeHome device tests the central area
scotomas or visual field defects. Therefore, the of the visual field of 14 degrees, with approxi-
Amsler grid cannot be used to monitor disease mately 500 data points sampled rapidly 3–5 times
progression over time. (Fig. 26.2). The stimulus is a dot deviation signal
a b c
Fig. 26.2 The ForeseeHome device. The patient tests one ted line will appear, and the test continues for approxi-
eye at a time with the device. (a) The testing screen shows mately 3 min. Daily testing results are sent to the central
a dotted line with a wave or a bump in it, which appears monitoring center. (b) The device tests the central area of
briefly then disappears. The patient is instructed to use the the visual field of 14 degrees with approximately 500 data
device’s mouse to click where the bump appeared on the points sampled rapidly 3–5 times. (c) (Reference: www.
screen and return the cursor to the center dot. A new dot- foreseehome.com; accessed June 24, 2019)
flashed in a pseudo-random predetermined order. to recall the patient for an exam [20]. The sensi-
During the course of the testing, the patient is tivity and specificity of the ForeseeHome moni-
presented with artificial distortions similar to toring device were shown to be similar to that of
those seen by individuals with neovascular a clinical device using PHP technology that was
AMD. The patient indicates where he/she per- used to determine visual field defects associated
ceives distortion along the dotted line using a with AMD progression in an office setting in a
computer mouse. When a signal is presented to cross-sectional study [12]. The ForeseeHome
the retina over a CNV lesion, the pathological monitoring device plus standard care was evalu-
distortion and the artificial distortion compete for ated in the HOME study (the randomized trial of
the participant’s attention. Preferential looking a home monitoring system for early detection of
will subconsciously direct the user to select the CNV-Home Monitoring of the Eye (HOME)
more extreme distortion. Varying degree of the study) [21].
artificial distortions serves as control stimuli to The HOME study was the unmasked, con-
quantify the extent of possible pathologic distor- trolled, randomized clinical trial aimed to deter-
tions originating from retinal pathology. A gray- mine whether home monitoring with
scale map of the visual field is generated, and the ForeseeHome device, using macular visual field
intensity of the visual field defect is determined. testing with hyperacuity techniques and telemon-
Based on a comparison of this intensity with a itoring, results in earlier detection of CNV asso-
normative database and the participant’s baseline ciated with neovascular AMD, reflected in better
performance with the ForeseeHome device, the vision when compared with standard care. The
system determines whether there are visual study was conducted in 44 clinical sites of the
responses that may be associated with the pro- Age-Related Eye Disease Study2 (AREDS2), a
gression of AMD to CNV [20]. Participants are clinical trial of nutritional supplements for the
instructed to use the ForeseeHome device ideally treatment of AMD [22]. The enrolled participants
every day to monitor each eye separately, which were individuals at risk for developing CNV,
takes approximately 3 min per eye. Results of with large drusen in both eyes (two study eyes) or
each home-testing session are transmitted via a large drusen in one eye (study eye) and advanced
cellular modem to a data monitoring center, when AMD in the fellow (non-study) eye, and best-
the monitoring center detects a significant change corrected visual acuity of 20/60 or better in the
compared with baseline, the physician is alerted study eye(s). Each participant was screened using
a brief tutorial on the home monitoring device in on average (SD, 1.7), and less than two times/
the clinic to make sure that they could use the week in 57 participants. Results from the HOME
device and did not have any abnormalities pre- study suggest that individuals with higher risk to
venting device monitoring in the future. Of 1970 develop CNV would receive advantage from this
participants who were screened, 1520 (77%) home monitoring strategy to detect CNV devel-
were enrolled and randomized to one of the two opment at an early stage and be able to achieve
arms: 763 were randomized to device arm and favorable visual outcomes after intravitreous
757 to standard care. Participants in the standard anti-VEGF injections [21].
care arm received investigator-specific instruc- For clinical characteristics of eyes that newly
tions to monitor any changes in their vision at developed CNV detected and alerted by the
home (Amsler grid may be recommended). ForeseeHome device, a report of two cases dem-
Participants randomized to the device arm onstrated that the device alerts occurred when the
received the same standard care instruction, and patients had minimal or no visual symptoms and
also received a home monitoring device, with visual acuities were 20/32 or better; subtle cys-
instructions for installation and use. Participants toid abnormalities appeared on optical coherence
were encouraged to use the device on a daily tomography of both cases with minimal or no
basis, and results were transmitted automatically fluorescein angiogram leakage [23].
via cellular modem to a central data monitoring Although the ForeseeHome device has been
center. When the device testing suggested a proven in its efficacy in detecting early develop-
change compared with the baseline measurement of neovascular AMD, not every AMD
ment, an alert was sent from the monitoring cen- patient can successfully use the device. According
ter to the participant’s clinical center prompting to the HOME study, eyes with preexisting visual
the staff to schedule a visit with the study oph- field defects resulted in an approximate 20%
thalmologist within 72 h. The development of screen failure rate. An additional 8% of the eyes
CNV was evaluated and confirmed by an investi- could not establish baseline measurements after
gator, based on standardized ocular imaging with randomization, and 20% stopped monitoring at
fundus photography, OCT, and FA [21]. different time points during the study [21].
Results showed the mean (SD) study follow- Individuals with difficulty using a computer
up period of 1.4 (0.6) years. Fifty-one patients in mouse would not be able to use the ForeseeHome
the device arm and 31 in the standard care arm device as well.
developed CNV. Participants in the device arm The ForeseeHome device is currently avail-
demonstrating less impaired visual acuity or able only in the United States by physician’s pre-
fewer letters lost from baseline at the time of scription on a rental basis for individuals who are
CNV detection (median (interquartile range at risk to develop a neovascular form of AMD.
[IQR]):−4 [−11.0, −1.0] letters) compared with
standard care arm (median [IQR]:−9 [−14.0,
−4.0] letters) (p = 0.021). Also, as a secondary 26.4 Handheld Shape
visual acuity outcome, a higher percentage of Discrimination Hyperacuity
eyes maintained 20/40 or better visual acuity at Test on Smart Phone
the time of CNV diagnosis and initiation of treat-
ment in the device arm (87%) compared with the The shape discrimination hyperacuity (SDH) test
standard care arm (62%) (p = 0.014). Among par- was developed by Wang et al. using perfect and
ticipants who used the device at the recom- distorted circular contour called radial frequency
mended frequency, the proportion of eyes that patterns [24] as visual stimuli, based on a hypoth-
maintained BCVA of 20/40 or better was 94%. In esis that patients with maculopathy have more
the device arm, among 609 subjects who con- difficulty performing visual tasks that require
tinue using the home device, the weekly usage global integration of visual stimuli over a large
throughout the study period was 4.4 times/week retinal area than performing a localized task like
visual acuity [25]. The optimal performance of sion, and automatically alerts a health care pro-
SDH requires global visual integration [24]. By vider if a significant deterioration of visual
measuring SDH, a patient’s ability to detect function is suspected.
visual distortion at his/her own ability to inte- Wang et al. compared the handheld SDH
grate visual information can be quantified. SDH (hSDH) (MyVisionTrack app) on a mobile device
is significantly reduced in AMD [25], while less with a previously established desktop SDH
affected by normal aging [26]. (dSDH) in 100 subjects, results suggested that
Recently, a handheld SDH test (myVision- the hSDH test on a mobile device was highly cor-
Track; Vital Art and Science, Inc., Richardson, related with PC-based testing methods (r = 0.88,
TX) has been implemented on a mobile platform p < 0.0001). The mobile app is also sensitive to
(iOS; Apple Inc., Cupertino, CA). It received the severity of maculopathy. The mean hSDH
clearance by the US FDA in 2013 to be marketed measurement of eyes with advanced AMD or
as an at-home method for monitoring the pro- with severe to very severe nonproliferative dia-
gression of degenerative eye diseases, such as betic retinopathy (NPDR) was significantly
macular degeneration and diabetic retinopathy. worse than that of the eyes with intermediate
The myVisionTrack app displays side-by-side AMD or with mild-to-moderate NPDR
circles on the screen (Fig. 26.3); firstly, one circle (p < 0.0001). Ninety-eight percent of 46 patients
is significantly different from the others (for with maculopathy who completed the usability
example, with a wavy or jagged edge). To per- survey reported that the hSDH test was easy to
form a self-check, the patient covers one eye and use [27].
then touches/selects what he or she determines to A prospective study had evaluated the feasi-
be the “differently-shaped” circle. Each time the bility of the handheld SDH (myVisionTrack) in
patient makes a selection, subsequent screens 160 patients with neovascular AMD in at least
display additional circles with increasingly subtle one eye from 24 centers in the United States.
differences among all circles. The test is then Approximately two-thirds of participants aged 75
repeated with the other eye. The device stores the years or older. Results demonstrated that 84.7%
self-check test results, tracks eye disease progres- on average complied with daily myVisionTrack
Fig. 26.3 The myVisionTrack (handheld shape discrimi- subtle differences among all circles. The test is then
nation hyperacuity test) on an iOS mobile platform. To repeated with the other eye. The device stores the self-
start testing, the patient covers one eye and then selects check test results, tracks eye disease progression, and
what he/she determines to be the “differently-shaped” automatically alerts a health care provider if a significant
circle. Each time the patient makes a selection, subse- deterioration of visual function is suspected (Reference:
quent screens display additional circles with increasingly www.myvisiontrack.com; accessed June 24, 2019)
testing, and 98.9% complied with at least weekly mine whether changes in HAC score compared to
myVisionTrack testing [28]. However, reports on baseline correlates with the changes in OCT.
the validity of myVisionTrack in detecting
changes associated with AMD from this study
have not yet been published. 26.6 Self-Measured Retinal
Sensitivity on A Tablet
Device
26.5 Self-Measured Hyperacuity
Test on A Tablet Device The PsyPad platform is a central retinal sensitiv-
ity testing developed for use on a tablet device
The Hyperacuity App (HAC) is a test developed (iPad; Apple, Cupertino, California).
for use on an iPad (Apple, Cupertino, California). Adams et al. determined its feasibility in 38
The aim of the app is to be intuitive and easy to participants with intermediate AMD to monitor
use for individuals without cognitive impairment retinal sensitivity, the benefits of weekly remind-
up to the age of 90 years while quantitatively ers, and the comparison with clinic-based results.
assessing the individual’s vision. The sensitivity results obtained in the home mir-
The app introduces challenges as a presence rored those obtained in formal clinic-based
of distortion in the line. The distortion is shown perimetry. However, there was a high level of
for 200 milliseconds before the distortion inactivity (45%) mainly due to noncompatible
becomes straight. The subject then selects on the devices and other technology access issues [30].
line where he/she believes the distortion was. The Another study by Wu et al. compared the cen-
selections are required to be within a certain per- tral retinal sensitivity measured by the Psypad
pendicular distance of the line to be considered. platform on a tablet device with standardized
Audio feedback is given for each selection to microperimetry in 30 participants with AMD. The
confirm that a valid selection was accomplished. average test duration on PsyPad was 53.9 ± 7.5 s.
After a valid selection, the screen clears to black. The mean central retinal sensitivity was not sig-
Subjects are then required to select the center of nificantly different between PsyPad
the empty screen for the next challenge in order (25.7 ± 0.4 dB) and microperimetry (26.1 ± 0.4
to refocus the fovea on the center. Challenges are dB, P = 0.094), and the 95% limits of agreement
randomized during each test and can be either between the two measures were between −4.12
horizontal or vertical. Some challenges are ran- and 4.92 dB [31].
domized to have no distortions. Subjects are Ho et al. evaluated the feasibility of the Psypad
asked to select the center of the screen if no dis- platform to identify alterations in retinal sensitiv-
tortion is seen. A score is then calculated from the ity and correlations with underlying pathology in
data collected based on a grading system created 53 neovascular AMD eyes receiving treatment
prior to the study that was optimized to detect and in 21 at-risk fellow eyes. Results demon-
errors in distortion perception. strated that the decreased retinal sensitivity asso-
In 2016, a cross-sectional study performed to ciated with the presence of atrophy (p < 0.01),
explore the HAC as a screen for the progression retinal pigment epithelium disruption (p < 0.01),
of disease in 33 AMD subjects. Results showed and absent ellipsoid zone (p < 0.01), but not with
that the HAC had high sensitivity (92.3%) and the presence of subretinal fluid (p = 0.94) nor
moderate specificity (61.5%) in distinguishing intraretinal fluid (p = 0.52). There were differ-
between patients who required treatment and ences in retinal sensitivity between locations of
those who did not require treatment. The the intact retina to that of retinal pathology asso-
researcher determined whether each subject ciated with AMD [32].
would receive an anti-VEGF treatment based on Future studies with a more significant number
the OCT findings [29]. of sample sizes might be warranted to evaluate
Future studies with a larger sample size or a the validity of the Psypad platform in detecting
longitudinal study may be warranted to deter- changes in central retinal sensitivity associated
with an early development of CNV or reactiva- the visual acuity remains unaffected is consid-
tion of CNV during treatments. ered the ultimate goal for home-monitoring
devices for AMD patients.
In the past decade, a number of home-
26.7 Home-Based monitoring devices for AMD have been researched
Self-Measuring OCT and developed. However, only two (the
ForeseeHome and the MyVisionTrack) have been
In an office-setting, optical coherence tomogra- approved by the US FDA, and only one (the
phy (OCT) imaging has become the mainstay in ForeseeHome monitoring program) was proven in
AMD management, as OCT assists physicians in a large randomized clinical trial to be beneficial
being able to identify even subtle disease activity, for individuals at high risk to develop neovascular
i.e., subretinal or intraretinal fluid both in quanti- AMD (large drusen in both eyes, or large drusen
tative and in qualitative methods, and far better in one eye and advanced AMD in the fellow eye)
than fundus examination alone. The use of OCT to detect CNV at an earlier stage and would
device as part of home monitoring for AMD increase the patients’ opportunities to achieve bet-
would increase the sensitivity and specificity of ter outcomes following intravitreous anti-VEGF
early CNV diagnosis. treatment, compared with standard care alone
Although the home-based OCT device is cur- (with aids include the Amsler grid) [21].
rently not available in the market, some investi- Given the increasing popularity of personal
gators have explored the possibility of using a electronic devices, e.g., smartphone, tablet usage
home-based self OCT scan. Mansour et al. evalu- among individuals of any age group including
ated the image quality of self-OCT by the oph- those aged 50 years or older who may be at risk
thalmologist and optometrists. Results for neovascular AMD, many applications for
demonstrated that self-OCT scans of the macula iPhone or tablet device have been developed for
could be performed with excellent image quality, people to monitor visual functions at home.
but image decentration is a common finding [33]. Nevertheless, substantial evidence to support
Recently, Maloca et al. have reported the safety their reliability or efficacy in detecting newly
and feasibility of a novel sparse OCT device for developed CNV or changes in the macula is very
patient-delivered retina home monitoring in limited at this time. Home-based self-measuring
AMD patients and to compare the device to an OCT may be a potential home-monitoring device
OCT device that is commercially available. in the future but required further developments
Results demonstrated the feasibility and safety of and validations.
this home-based OCT monitoring under real-life In the meantime, while the efficacy-proven
settings and suggesting the potential clinical ben- ForeseeHome AMD monitoring program is avail-
efit of the device. However, the device tested in able only in the United States and other devices are
this study was just the prototype, which required still under investigation, the Amsler grid may still
further modification and evaluation as a portable play a role in addition to regular eye care as a sim-
self-measuring OCT device in the future [34]. ple and inexpensive home-monitoring tool for
AMD. Though its efficacy in detecting early CNV
is relatively lower when compared with recent
26.8 Summary technologies, the Amsler grid may still be benefi-
cial for AMD patients living in countries or regions
Main predictors for favorable final vision in eyes where the efficacy-proven home-monitoring pro-
with neovascular AMD receiving intravitreous gram is not available or those with limited access
anti-VEGF therapy include a good visual acuity to smartphones or tablet devices. As a physician,
(20/40 or better) and smaller size of CNV lesions in addition to prescribing home- monitoring
at the time of treatment initiation [6, 8]. Therefore, devices to patients at risk for neovascular AMD, it
the earliest detection of CNV developments when is also crucial to raise the patients’ awareness on
the changes of their visual functions, emphasize lar degeneration: two-year results of the ANCHOR
study. Ophthalmology. 2009 Jan;116(1):57–65.e5.
importance of early detection of CNV with timely 4. Schmidt-Erfurth U, Kaiser PK, Korobelnik J-F, Brown
anti-VEGF treatments, and the importance of reg- DM, Chong V, Nguyen QD, et al. Intravitreal afliber-
ular usage of any available home-monitoring cept injection for neovascular age-related macular
devices (daily or at least once weekly), to be able degeneration: ninety-six-week results of the VIEW
studies. Ophthalmology. 2014 Jan;121(1):193–201.
to detect the CNV as early as possible and be able 5. Comparison of Age-related Macular Degeneration
to stabilize or maximize their visual functions after Treatments Trials (CATT) Research Group, Martin
receiving anti-VEGF treatments. DF, Maguire MG, Fine SL, Ying G, Jaffe GJ, et al.
Ranibizumab and bevacizumab for treatment of neo-
vascular age-related macular degeneration: two-year
Key Learning Points results. Ophthalmology. 2012 Jul;119(7):1388–98.
• Early detection of CNV when vision remains 6. Ying G, Huang J, Maguire MG, Jaffe GJ, Grunwald
unaffected is crucial for neovascular AMD eye JE, Toth C, et al. Baseline predictors for one-year
to achieve favorable visual outcomes after visual outcomes with ranibizumab or bevacizumab
for neovascular age-related macular degeneration.
anti-VEGF treatments. Ophthalmology. 2013 Jan;120(1):122–9.
• Scotoma sometimes not detected when using 7. Ying G, Maguire MG, Daniel E, Ferris FL, Jaffe
the Amsler grid due to the perceptual comple- GJ, Grunwald JE, et al. Association of baseline
tion phenomenon of the brain. Characteristics and early vision response with
2-year vision outcomes in the comparison of AMD
• According to the HOME study, participants in treatments trials (CATT). Ophthalmology. 2015
the device arm (ForeseeHome device plus Dec;122(12):2523–2531.e1.
standard care) had a significantly better vision 8. Blinder KJ, Bradley S, Bressler NM, Bressler SB,
when CNV was detected compared with the Donati G, Hao Y, et al. Effect of lesion size, visual
acuity, and lesion composition on visual acuity
standard care arm. change with and without verteporfin therapy for cho-
• A number of self-measured visual function roidal neovascularization secondary to age-related
tests designed for personal electronic devices, macular degeneration: TAP and VIP report no. 1. Am
i.e., smartphone and tablet devices are avail- J Ophthalmol. 2003 Sep;136(3):407–18.
9. Lim JH, Wickremasinghe SS, Xie J, Chauhan DS,
able, but strong evidence to support their Baird PN, Robman LD, et al. Delay to treatment
validity in detecting changes in AMD eyes is and visual outcomes in patients treated with anti-
very limited. vascular endothelial growth factor for age-related
• Patients’ awareness and regular usage of any macular degeneration. Am J Ophthalmol. 2012
Apr;153(4):678–686.e2.
home-monitoring device (daily or at least 10. Marmor MF. A brief history of macular grids: from
once weekly) would increase an opportunity Thomas Reid to Edvard Munch and Marc Amsler.
for early detection of CNV. Surv Ophthalmol. 2000 Feb;44(4):343–53.
11. Amsler M. Earliest symptoms of diseases of the mac-
ula. Br J Ophthalmol. 1953 Sep;37(9):521–37.
12. Fine AM, Elman MJ, Ebert JE, Prestia PA, Starr JS,
References Fine SL. Earliest symptoms caused by neovascular
membranes in the macula. Arch Ophthalmol. 1986
1. Wong WL, Su X, Li X, Cheung CMG, Klein R, Apr;104(4):513–4.
Cheng C-Y, et al. Global prevalence of age-related 13. Schuchard RA. Validity and interpretation of
macular degeneration and disease burden projec- Amsler grid reports. Arch Ophthalmol. 1993
tion for 2020 and 2040: a systematic review and Jun;111(6):776–80.
meta-analysis. Lancet Glob Health. 2014 Feb;2(2): 14. Achard OA, Safran AB, Duret FC, Ragama E. Role
e106–16. of the completion phenomenon in the evaluation
2. Rosenfeld PJ, Brown DM, Heier JS, Boyer DS, Kaiser of Amsler grid results. Am J Ophthalmol. 1995
PK, Chung CY, et al. Ranibizumab for neovascular Sep;120(3):322–9.
age-related macular degeneration. N Engl J Med. 15. Augustin AJ, Offermann I, Lutz J, Schmidt-Erfurth U,
2006 Oct 5;355(14):1419–31. Tornambe P. Comparison of the original Amsler grid
3. Brown DM, Michels M, Kaiser PK, Heier JS, Sy JP, with the modified Amsler grid: result for patients with
Ianchulev T, et al. Ranibizumab versus verteporfin pho- age-related macular degeneration. Retina Phila Pa.
todynamic therapy for neovascular age-related macu- 2005 Jun;25(4):443–5.
16. Faes L, Bodmer NS, Bachmann LM, Thiel MA, 25. Wang Y-Z, Wilson E, Locke KG, Edwards
Schmid MK. Diagnostic accuracy of the Amsler grid AO. Shape discrimination in age-related macu-
and the preferential hyperacuity perimetry in the lar degeneration. Invest Ophthalmol Vis Sci. 2002
screening of patients with age-related macular degen- Jun;43(6):2055–62.
eration: systematic review and meta-analysis. Eye. 26. Wang YZ. Effects of aging on shape discrimina-
2014 Jul;28(7):788–96. tion. Optom Vis Sci Off Publ Am Acad Optom. 2001
17. Goldstein M, Loewenstein A, Barak A, Pollack A, Jun;78(6):447–54.
Bukelman A, Katz H, et al. Results of a multicenter 27. Wang Y-Z, He Y-G, Mitzel G, Zhang S, Bartlett
clinical trial to evaluate the preferential hyperacuity M. Handheld shape discrimination hyperacuity test
perimeter for detection of age-related macular degen- on a mobile device for remote monitoring of visual
eration. Retina Phila Pa. 2005 May;25(3):296–303. function in maculopathy. Invest Ophthalmol Vis Sci.
18. Preferential Hyperacuity Perimetry Research Group. 2013 Aug 13;54(8):5497–505.
Preferential hyperacuity perimeter (PreView PHP) 28. Kaiser PK, Wang Y-Z, He Y-G, Weisberger A,
for detecting choroidal neovascularization study. Wolf S, Smith CH. Feasibility of a novel remote
Ophthalmology. 2005 Oct;112(10):1758–65. daily monitoring system for age-related macu-
19. Loewenstein A, Ferencz JR, Lang Y, Yeshurun I, lar degeneration using mobile handheld devices:
Pollack A, Siegal R, et al. Toward earlier detection of results of a pilot study. Retina Phila. 2013
choroidal neovascularization secondary to age-related Oct;33(9):1863–70.
macular degeneration: multicenter evaluation of a 29. Chen JS, Adelman RA. Hyperacuity exam screens for
preferential hyperacuity perimeter designed as a home choroidal neovascularization in age-related macular
device. Retina Phila Pa. 2010 Aug;30(7):1058–64. degeneration on a mobile device. Ophthalmic Surg
20. Chew EY, Clemons TE, Bressler SB, Elman MJ, Lasers Imaging Retina. 2016;47(8):708–15.
Danis RP, Domalpally A, et al. Randomized trial of 30. Adams M, Ho CYD, Baglin E, Sharangan P, Wu Z,
the ForeseeHome monitoring device for early detec- Lawson DJ, et al. Home monitoring of retinal sensi-
tion of neovascular age-related macular degenera- tivity on a tablet device in intermediate age-related
tion. The home monitoring of the eye (HOME) study macular degeneration. Transl Vis Sci Technol. 2018
design – HOME study report number 1. Contemp Clin Sep;7(5):32.
Trials. 2014 Mar;37(2):294–300. 31. Wu Z, Guymer RH, Jung CJ, Goh JK, Ayton LN, Luu
21. Chew EY, Clemons TE, Bressler SB, Elman MJ, Danis CD, et al. Measurement of retinal sensitivity on tablet
RP, Domalpally A, et al. Randomized trial of a home devices in age-related macular degeneration. Transl
monitoring system for early detection of choroidal neo- Vis Sci Technol. 2015 Jun;4(3):13.
vascularization home monitoring of the eye (HOME) 32. Ho CYD, Wu Z, Turpin A, Lawson DJ, Luu CD,
study. Ophthalmology. 2014 Feb;121(2):535–44. McKendrick AM, et al. A tablet-based retinal function
22. Age-Related Eye Disease Study 2 Research Group. test in Neovascular age-related macular degeneration
Lutein + zeaxanthin and omega-3 fatty acids for age- eyes and at-risk fellow eye. Transl Vis Sci Technol.
related macular degeneration: the age-related eye 2018 Mar;7(2):2.
disease study 2 (AREDS2) randomized clinical trial. 33. Mansour AM. Self-optical coherence tomogra-
JAMA. 2013 May 15;309(19):2005–15. phy and angiography. Case Rep Ophthalmol. 2017
23. Chaikitmongkol V, Bressler NM, Bressler SB. Early Apr;8(1):108–15.
detection of choroidal neovascularization facilitated 34. Maloca P, Hasler PW, Barthelmes D, Arnold P,
with a home monitoring program in age-related mac- Matthias M, Scholl HPN, et al. Safety and feasibil-
ular degeneration. Retin Cases Brief Rep. 2015;9(1): ity of a novel sparse optical coherence tomography
33–7. device for patient-delivered retina home monitoring.
24. Wilkinson F, Wilson HR, Habak C. Detection and rec- Transl Vis Sci Technol. 2018 Jul;7(4):8.
ognition of radial frequency patterns. Vis Res. 1998
Nov;38(22):3555–68.
joined the faculty at the Department of Ophthalmology,

Chiang Mai University. In 2012, Dr. Chaikitmongkol did
her post-doctoral clinical research fellowship at the Retina
Division, Wilmer Eye Institute, Johns Hopkins University
School of Medicine (USA); medical retina training with
Professor Neil M. Bressler and Professor Susan
B. Bressler (2012–2014), and visual neurophysiology
training with Professor Hendrik P.N. Scholl (2013–2014).
Her research interest involves the diagnosis and treat-
ments of polypoidal choroidal vasculopathy, age-related
macular degeneration, diabetic macular edema, diabetic
retinopathy, and novel retinal imaging modalities. Dr.
Chaikitmongkol is currently an Assistant Professor in
Ophthalmology at the Faculty of Medicine, Chiang Mai
Voraporn Chaikitmongkol received her medical degree
University, Chiang Mai, Thailand. Dr. Chaikitmongkol
in 2005 and an ophthalmology degree in 2010 from
has recently received the Japanese Ophthalmological
Chiang Mai University, Thailand. Following her vitreo-
Society International Young Investigator Award 2019.
retinal fellowship training in 2011, she subsequently
Rehabilitation
27
Yogeshwari Bansal
Age-related macular degeneration (AMD) is in the better eye, or a total visual field less than 10
the leading cause of visual impairment in the degrees from the point of fixation, but who wants
Western worId [1, 2] and presents a substantial to and is potentially able to use, vision for the
healthcare burden currently after Diabetic planning and/or execution of a task. In AMD, it is
Retinopathy in the adult age group. AMD the central visual acuity that is reduced while
causes visual impairments of reduced visual peripheral visual fields are intact until and unless
acuity, central scotoma, and reduced contrast it is combined with some other disease such as
sensitivity which cause multiple disabilities [3] glaucoma, etc.
of walking difficulty, difficulty in recognizing The treating ophthalmologist needs to iden-
faces, and near tasking problems of difficulty in tify and refer patients with AMD to LVR depart-
reading and writing, mobile operations and ment for LVR counseling, rather than referring
other activities of daily living (ADL) which do only patients with less than 6/18 vision. As
affect the quality of life [4, 5]. Despite medical patients with AMD have functional difficulties
advancements of anti-VEGF therapy, there are even when the vision is 6/18 or better. At this
limitations. Patients need low vision rehabilita- stage, the patients need counseling about under-
tion (LVR) during orter the therapy. LVR aims standing the disease process and prognosis,
at minimizing visual disabilities by helping visual requirements need to be discussed to set
them to make use of their dual vision that can the goals, contrast acuity needs to be checked to
be used with the help of certain magnifiers and work on contrast enhancement by improving
techniques to make the best use of remaining the lighting conditions, environment modifica-
eyesight. tions if need so and simple higher plus near
A person with low vision exists with visual addition in near glasses to help them in reading.
impairment even after treatment and /or standard This is called LVR counseling or conventional
refractive correction, and has a best-corrected LVR that could be done by all the clinical prac-
visual acuity of less than 6/18 to light perception ticing Optometrists.
Patients by Low vision definition need
“Enhanced LVR” that includes prescription of
Low vision devices or aids (LVD/LVA). LVA
Y. Bansal (*) includes various optical magnifiers and nonop-
Heading Department of Optometry, ICARE Eye
tical aids, projection magnifiers, different com-
Hospital & Postgraduate Institute, Noida,
Uttar Pradesh, India puter Software program, and mobile
e-mail: yogeshwari@icarehospital.org applications. It will include environment modi-

376 Y. Bansal
fications and rehabilitators’ counseling to help Though different magnification formulas calcu-
them in social issues such as depression, loss of late X required that depend upon best-corrected
independence, reduced confidence level, and visual acuity(BCVA).One easy Kestenbaum for-
dependency issues. Mild-to-Moderate cases of mula will calculate magnification (X), which is
low vision, that is Snellen acuity of 3/60 to reciprocal of BCVA like 6/24 visual acuity means
6/24, can be helped a lot with the help of LVA 4x. Optical magnifiers can provide up to 12–20X
and patients can function independently but but higher magnification will reduce the field of
severe cases of vision impairment, with visual vision. So, projection magnifiers are to be used
acuity less than 3/60, need some help from their which can provide higher magnification with bet-
family members as well to perform their daily ter field of vision.
routine activities. Nonoptical magnifiers work on contrast
enhancement. Maximum contrast is like jet black
on white gives maximum contrast and is more
27.1 LVA in AMD visible and watching the white screen in a dark
room is more sharper. The reading acuity can be
Functional vision problems for distance and increased by keep reading text as bigger, bolder,
near need to be identified and tabulated at the and brighter. I always prefer Big B (Photos 27.1,
time of LVR. Most commonly for distance, the 27.2, 27.3, 27.4, and 27.5).
patient is going to have problems in recogniz-
ing faces at home and outside. Common chal-
lenges include watching television at home,
navigation within the house, and inability to see
the clock. So, for distance functional vision
improvement, magnification needs to be
explained which means closer the object is to
the patient, it becomes bigger. In this magnifi-
cation, the closer object subtends a larger visual
angle and larger image at the retina. Keeping a
larger TV screen, clock with larger numbers or Photo 27.1 Central scotoma
talking watches, nail trimmers with a magnifier
attached to it are a few examples of nonoptical
magnifiers that are useful in AMD.
Optical magnifiers for distance will include
telescope and watch television.
Spectacle mounted telescope, head-mounted
telescope, and head-mounted magnifier are better
carried in old age group [1]. Head-mounted tele-
scope like Ocutech is US FDA approved for driv-
ing as well.
Optical magnifiers work on linear and angular
magnification and create a virtual and enlarged
image that is defined in X. 2X means the image is
twice the object size. All the magnifiers are
marked with X according to the power of the sys-
tem and X amount required by the patient is Photo 27.2 Distance optical magnifier- telescope-4x,
decided by the trial of the optical magnifier. 6x, see tv and Ocutech
27 Rehabilitation 377
27.2 Central Scotoma
AMD patients are the most unhappy in doing

even simple near tasks like reading [6] newspa-
per, signing a cheque, writing a letter, mobile
operation, threading a needle, etc. As there is
macular involvement causing central scotoma,
near tasking is the most affected. But rest of the
visual field is spared so magnifiers work well as
they enlarge the image and image falls on the see-
ing area now, away from the scotomatous area.
Near optical magnifiers like spectacle
Photo 27.3 Near magnifiers—optical and projection magnifiers(SM), handheld illuminated magnifiers
magnifiers (HHM), stand and dome magnifiers, pocket mag-
nifiers, bar magnifiers along with chest magnifier
work well. Nonoptical aids include reading stand
or riser, typoscope, letter writer, cheque guide, and
signature guide. Near projection magnifiers and
desktop magnifiers like Merlyn also are very use-
ful in severe visual impairment. Close circuit tele-
visions (CCTV) are of immense help in moderate
to severe VI*^. When optical magnifiers do not
work then projection magnifiers definitely help in
reading. There is SARA reader also which is talk-
ative assistance technology based, which converts
text into audio and the patient can listen to it.
Kindle books and e-books are also available,
which are more useful than textbooks as size and
Photo 27.4 Reading stand with table lamp contrast can be increased on the screen.
Socioeconomic and behavioral issues are
taken care of through counseling by the
Optometrist or Rehabilitator that plays a very
important role in LVR. The patient needs to con-
tinue his or her hobbies and profession not to feel
left out. This assurance has to be provided to the
patient along with the family member through
successful case examples and testimonials.
Eccentric viewing is a good technique to see
better in AMD. This makes use of extrafoveal
peripheral area called preferred retinal locus
(PRL) [7] to see by turning your head as and
when the central area is affected. So all patients
of AMD should have Amsler grid charting done
Photo 27.5 Non optical aids- typoscope, notex, cheque if possible to localize the extent of damage and
guide the technique followed.
378 Y. Bansal
Walking cane with sensors and light and audi- accept the remedies so as to spread the basic mes-
ble clues is an excellent support for walking in sage of early referral to low vision specialists.
old age group. Besides, peer group discussions
among AMD, Facebook groups can be a booster
for them as some people come out with great References
ideas and innovations and can be the resource
providers for the other patients. Larger screen 1. Evans J, Wormald R. Is the incidence of registrable
age-related macular degeneration increasing? Br J
size mobiles with magnifier option in smart Ophthalmol. 1996;80:9–14.
phones along with flashlight features and various 2. Harvey PT. Common eye diseases of elderly peo-
apps in Android and Apple give a lot of assistance ple: identifying and treating causes of vision loss.
to AMD patients. Gerontology. 2003;49:1–11.
3. Rubin GS, Bandee-Roche K, Huang GH, et al. The
To reduce the complaints of glare and photo- association of multiple visual impairments with
phobia, filters and photochromatic glasses are self-reported visual disability: SEE project. Invest
advised. Usually, yellow filter reduce glare Ophthalmol Vis Sci. 2001;42:64–72.
indoors in AMD. 4. Williams RA, Brody BL, Thomas RG, et al. The
psychosocial impact of macular degeneration. Arch
My personal experience has seen a lot of satis- Ophthalmol. 1998;116:514–20.
factory clinical improvement in AMD as most of 5. Mangione CM, Gutierrez PR, Lowe C, et al. Influence
the patients can read newspaper N10–N8 print of age-related maculopathy on visual functioning
with the help of optical LVA or Projection magni- and health-related quality of life. Am J Ophthalmol.
1999;128:45–53.
fier [5] but the depression is high and confidence 6. Crossland MD, Culham LE, Rubin GS. Predicting
level is so low sometimes that they are unable to reading fluency in macular disease. Optom Vis Sci.
make use of that device. So the identification and 2005;82(1):11–7.
referral of AMD patient have to be earlier to avoid 7. CrossIand MD, CuIham LE, et al. Fixation stabil-
ity and reading speed in patients with newly devel-
that burnout stage. The ARMD patients need to oped macular disease. Ophthalmic Physiol Opt.
accept their disabilities and we need to make them 2004;24(4):327–33.
27 Rehabilitation 379
Yogeshwari Bansal has a Master’s degree in Optometry

from Vinayaka University, India and Bachelor’s degree in
Optometry from AIIMS, New Delhi, India. She completed
her Fellowship in optometry from L. V. Prasad Eye
Institute, Hyderabad, India in 1993–1994. She is also a
Fellow of IACLE, Australia. Her areas of interest and
excellence are contact lenses and low visual aids besides
comprehensive optometry. Currently she is heading the
Optometry Department at ICARE Eye Hospital and Post
Graduate Institute, Noida, India with 35 full-time optom-
etrists in the team. She has co-authored many thesis works
at ICARE besides a few publications. She is an academi-
cian and active key speaker and chairperson in many
national-level optometry CMEs. She has organized LVA
workshops in-house and other tertiary hospitals.
Gene Therapy for Choroidal
Joo Yong Lee and Joon Hyung Yeo
28.1 Introduction lar degeneration (AMD), which is extremely

common and has a considerable social and eco-
The United States Food and Drug Administration nomic impact. Neovascular AMD (nAMD) has a
(FDA) has recently approved gene therapy for complex pathogenesis and CNV develops due to
treating RPE65 mutation-associated retinal dys- imbalances between pro-angiogenesis and anti-
trophy (a.k.a. Leber’s congenital amaurosis). angiogenesis factors. Many factors may be
This has garnered considerable attention for its employed as therapeutic targets to treat CNV,
possibilities for treating inherited retinal disease including VEGF receptor 1, soluble fms-like
(IRD) such as choroideremia, Usher’s disease, tyrosine kinase-1 (sFLT-1), pigment epithelium-
retinitis pigmentosa (RP), Stargardt disease, and derived factor (PEDF), and vascular endothelial
X-linked retinoschisis [1, 2]. growth factor (VEGF). This section will encom-
In gene therapy, mutated disease-causing pass a discussion of the latest advances in AMD
genes are replaced with healthy genes, mutated gene therapy and ongoing clinical testing.
genes that do not function properly are deacti-
vated, and/or new genes are introduced into the
system for the treatment of a particular condition. 28.2 Viral Vectors
Of the various methods of gene therapy, gene
augmentation therapy may be employed for treat- Several means of delivering genes, both viral and
ing IRD through inserting a new gene which can nonviral, have been created in the last 20 years.
produce missing functioning products; it can also Because target cells have close associations with
treat choroidal neovascularization (CNV) through nearby nontarget cells, and in some instances can
producing a number of proteins including anti- be problematic to access, viral vectors have a
angiogenic proteins, amongst others that have a number of advantages in comparison to nonviral
therapeutic effect, in sufficient quantity follow- treatments. These include electroporation and
ing genes being transferred to targeted cells. transfection, which can allow cellular tropism to
Successful gene therapies for LCA have boosted be altered via pseudotyping and which incorpo-
interest in employing it to treat age-related macu- rates enhanced safety. Additionally, when tissue-
specific promoters are incorporated, transgene
J. Y. Lee (*) · J. H. Yeo expression may be kept solely within the targeted
Department of Ophthalmology, Asan Medical Center, cells.
University of Ulsan College of Medicine, Some of the most widely used viral vectors are
Seoul, South Korea lentivirus, gamma-retrovirus, adeno-associated
e-mail: ophthalmo@amc.seoul.kr

382 J. Y. Lee and J. H. Yeo
virus (AAV), and adenovirus. Viral vectors are capacities (4.5–5.0 kb) and problems with elimi-
chosen specifically for individual applications; nating risks of humoral immune responses with
the type chosen is dependent on several elements, subjects with a viral exposure history [11]. A pair
including the vector’s cloning capacity, tissue tro- of clinical trials employing AAV2 vectors for
pism, and safety profiles, including insertional human rational gene therapy offered contradic-
oncogenesis and genotoxicity. Adenovirus vec- tory results with regard to serum neutralizing
tors, especially third-generation/helper-dependent antibody (NAb) levels and viral vector efficacy
vectors, possess high cloning capacities (around [12, 13]. In one of these trials AAV2-sFLT01 was
35 kb) and are capable of transducing a broad tested to see how tolerable and safe it was for
variety of cell types, which includes quiescent tis- subjects with advanced nAMD; the second trial
sues. The adenoviral vector genome generally tested the same treatment for subjects with
stays within the nucleus as an episome, rarely Leber’s hereditary optic neuropathy (LHON).
integrating. Nevertheless, the employment of The AMD patient study found that there was a
first-generation vectors has been hampered negative correlation between the presence of
because they can activate the inflammatory NAbs to AAV2 and AAV2-sFLT01’s capacity for
response [3, 4]. Second-generation vectors are the production of sFLT01 protein. However, the
not as immunogenic, and helper-dependent ade- research on subjects suffering LHON found that
noviral vectors are the safest by a considerable high NAbs serum levels do not necessarily limit
margin [5, 6]. the capacity of AAV2 vectors to effectively
At present, the most commonly employed deliver genes. Looking at the findings of Lee
gene delivery vehicles in ocular gene therapy are et al., such contradictory outcomes may be
AAV vectors. AAV vectors usually exhibit low explainable through the variations in the mainte-
immunogenicity, which makes the vector system nance status of the blood–retinal barrier (BRB)
highly suitable in the gene therapy treatment of depending upon the disease targeted [14]. These
numerous human conditions. As with adenoviral researchers demonstrated that NAbs levels in the
vectors, the genome of AAV vectors is generally vitreous humor may be elevated in instances of
maintained as an episome, and they transduce BLB disruption; they proposed that a number of
quiescent tissues [7]. In addition, multiple sero- retinal diseases that have associations with BRB
types are available, with each one offering breakdown could be disruptive to the retina’s
enhanced tropism related to a specific tissue, e.g., immune-privileged characteristics and could
AAV2 transduces the central nervous system affect absolute values of NAbs within the vitre-
(CNS), the liver, the retina, and skeletal muscle; ous humor.
AAV8 transduces the retina, the liver, the heart, Either equine infectious anemia virus (EIAV)
the pancreas, and the CNS [8]. There are numer- or human immunodeficiency virus 1 (HIV1) can
ous other features displayed by AAV vectors that be used to derive lentiviral factors [15], having
make it particularly suitable for ocular gene ther- cloning capacities reaching 10 kb [16]. Lentiviral
apies. Among such features are abilities to trans- factors can be employed when an application
duce nondividing cells, low levels of retinal requires the vector genome to be integrated
toxicity, the fact that it is generally nonintegrat- within the host genome; it has seen extensive use
ing, and that in human trials it has shown to have in transduction for hematopoietic stem cells [17].
good levels of safety and tolerability [9, 10]. The Lentiviruses have increased in popularity in
possible value of using AAV vectors in AMD recent times due to the fact that they are capable
gene therapy is clear from the success of past of transducing quiescent tissues, dividing cells,
clinical trials with IRDs. It is notable that the and have a superior safety profile to gamma retro-
FDA granted recent approval for Luxturna™, a viruses because vector genomes do not exhibit
novel gene therapy for subjects with genetic eye preferences between promoter or enhancer
diseases. Nevertheless, there are limitations to regions and therefore have a lower likelihood of
AAV vectors, including restricted transgene being genotoxic [18, 19].
28 Gene Therapy for Choroidal Neovascularization 383
28.3 Route of Administration as less effective in delivering viral vectors

straight to target tissue [21, 22]. Viral vector dif-
There are numerous features of the eye that make fusion can be limited by a number of physical
it an extremely useful model for AMD gene ther- blocks between the vitreous and the targeted
apy, including: retinal cells, including the inner retina (espe-
cially the dense interphotoreceptor matrix) and
(a) There is a tight blood–ocular barrier that the inner limiting membrane (ILM). A number
imposes limitations on the new genetic mate- of attempts have been made to mitigate this
rial being disseminated throughout the problem: Petrs-Silva et al. found that intravit-
system. real injections of capsid-mutated AAV2 in
(b) The eye is immune-privileged, meaning mouse retinas promoted the efficiency of trans-
inflammatory responses to new genetic mate- duction [23]; Dalkara et al. demonstrated that
rial will be limited. transduction could be enhanced in a number of
(c) The targeted retinal cells are easily accessi- different retinal cell types by mildly digesting
ble for direct delivery of genetic materials ILM and a nonspecific protease in combination
and for noninvasive monitoring of the pro- with AAV intravitreal injections [24]. Lee et al.
gression of disease and response to therapy. found that AAV vectors’ transduction efficiency
(d) When the other eye remains untreated it can could be enhanced using prior laser photocoag-
act as a control against which the therapeutic ulation after intravitreal injection [25]. This
response can be measured [20]. positive result of laser pretreatment could be
caused by the fact that basic thermal disruption
At present, there are two ways in which in the ILM opens up access to every retinal layer
genetic material can be introduced to targeted surrounding the burned area, or it may be due to
rational tissue, both by injection, either subreti- Müller cell endfeet movement increasing due to
nal or intravitreal. The most highly researched Müller cell stress responses.
method is pars plana vitrectomy (PPV), then Because variations in inflammatory responses
retinotomy and subretinal injections of viral have been demonstrated, depending on which
vectors. There is no standardization regarding ocular compartment the injection targets, subreti-
the exact retinal location in the eye (generally nal injection has been considered safe 1 year after
the injection is made in the subretinal gap readministration [26]. Contrastingly, intravitreal
between the retinal pigment epithelium and the injection with animal models is responsible for
photoreceptors) for injection, nor the dosage or stimulating NAbs formation [27].
volume of viral vector, nor the quantity of reti-
notomies used to deliver the vector or the loca-
tion of the injection. This method is more 28.4 nAMD Gene Therapy
invasive and causes temporary detachment of
the retina, but it does provide a means of direct nAMD has a complicated pathogenesis involving
delivery of viral vectors in the targeted tissue. a number of pathways making contributions to
Following this, the target cell (e.g., RPE or pho- pathologic endothelial proliferation, especially in
toreceptor cells) becomes infected with the causing imbalances between pro-angiogenesis
virus, which makes the host cells undertake and anti-angiogenesis elements. A number of
transcription and translation of the genetic widely recognized elements include relative
material carried with the virus into therapeutic PEDF deficiencies, VEGF overexpression, and
proteins. The alternative means of intravitreal the secreted extracellular domain of sFLT-1, the
injection of viral vectors has been tried by some VEGF receptor 1, being downregulated [28].
researchers. Although this technique is not as Many other elements also have involvement with
invasive and has the potential to cause fewer AMD pathogenesis, e.g., angiostatin and anti-
undesirable complications, it is generally seen angiogenic endostatin (Table 28.1).
Table 28.1 List of ongoing or recently completed human gene therapy trials for age-related macular degeneration
Mechanism of gene Targeted Clinicaltrials.gov Study
Therapy and delivery method therapy disease Number Sponsor phase
Intravitreal AdGVPEDF.11D PEDF expression nAMD NCT00109499 GenVec Phase I
with
anti-angiogenesis
Subretinal rAAV. sFLT-1 Cellular expression nAMD NCT01494805 Lions Eye Phase I
of VEGF-binding Institute Phase II
receptor FLT1 Avalanche
Biotechnologies,
Inc.
Intravitreal AAV2-sFLT01 Cellular expression nAMD NCT01024998 Sanofi Genzyme Phase I
of VEGF-binding
receptor FLT1
Subretinal AAV-8-based Monoclonal antibody nAMD NCT03066258 Regenxbio Inc. Phase I
anti-VEGF (RGX-314) binds VEGF
Subretinal lentiviral vector Angiogenesis nAMD NCT01301443 Oxford BioMedica Phase I
expressing endostatin and inhibition NCT01678872
angiostatin (RetinoStat)
Intravitreal Bevasiranib Downregulation of nAMD NCT00722384 OPKO Health Phase I
VEGF production NCT00259753 Phase II
Intravitreal Sirna-027 Knockdown of the nAMD NCT00363714 Allergan Phase I
Flt-1
Intravitreal PF-04523655 Hypoxia-inducible nAMD NCT00713518 Quark Phase II
RTP801 gene Pharmaceuticals
inhibition
28.4.1 PEDF 28.4.2 Anti-VEGFs
An early use of intraocular gene therapy for AMD Following on from this, attention switched in
took place in a phase I trial in 2006 (NCT00109499) terms of gene therapy for AMD two investigating
with advanced AMD 1 being treated with the sFLT-1, which is regarded as having anti-
PEDF gene delivered by AAV [29]. PEDF is an angiogenic effects through two means: the first is
anti-angiogenic peptide that occurs naturally the way in which inactive heterodimers are
which has been shown to have a positive effect in formed with membrane-bound VEGF receptor 1
the prevention and regression of neovasculariza- (Flt-1) and VEGF receptor 2 (Flk-1); the second
tion with AMD models in mice [30, 31]. For a is the direct binding and sequestration of VEGF
phase I clinical trial, 28 subjects suffering severe [32–34]. Additionally, because the expression of
nAMD were given single intravitreal injections of sFLT-1 is endogenous, the system should not be
adenoviral vector expressing human PEDF identified as being a harmful antigen in the target
(AdPEDF.11). No serious side effects were tissue and so when expressed from an AAV vec-
reported, although 25% of subjects suffered tran- tor no inflammatory response should occur. CNV
sient intraocular inflammation. Because there was formation was successfully prevented in preclini-
no control group and the sample size was limited cal primate and rodent testing using a subretinal
it was not possible to draw conclusions about recombinant AAV (rAAV) with sFLT-1 injection
visual outcomes, although subjects receiving 108 [35–38]. Avalanche Biotechnologies, collaborat-
particle units or more generally displayed stabil- ing with the Australian Lions Eye Institute
ity, while those receiving less than this demon- (NCT01494805) undertook investigations of
strated decreases in visual acuity and CNV lesions both the safety and efficacy of subretinal rAAV
increased in size. This research suggests that the sFLT-1 using clinical trials involving humans. In
size of dose may be influential [29]. 2015 they published the outcomes of the phase I
trial, following up with a report on 3-year out- median number of 7.0 [43]. The group receiving
comes in 2017 [39, 40]. This experimentation treatment and the control group both showed sig-
incorporated administering PPV and following nificant underperformance in this trial in com-
up with subretinal injections of low-dose (1 × 1010 parison to the benchmark MARINA (7.2 ETDRS
vector genomes (vg)) or high-dose (1 × 1011 vg) letter improvement) and ANCHOR (11.3 ETDRS
vector. Every subject was given ranibizumab letter improvement) [41, 43].
injections at the start of the program and four The intravitreal delivery of AAV2-sFLT01
weeks post-surgery, with further injections being was investigated by Sanofi Genzyme
provided when necessary in accordance with pre- (NCT01024998). The sFLT1 employed was sim-
set active nAMD criteria. The report on 3-year ilar to the one used by the Avalanche research,
follow-up stated that the site of the subretinal apart from it only encoding the VEGF-binding
injection showed no scarring, chorioretinal atro- domain of sFlt1 with links to a human IgG1
phy, or RPE cell proliferation [41]. It was not heavy chain Fc domain and the expression
possible to undertake visual/anatomical outcome employed the chicken β-actin (CBA) promoter
statistical analysis because of the small size of [44]. This research employed a CBA promoter
the sample. Nevertheless, subjects receiving the that fused the cytomegalovirus (CMV)
low dose, and the control group, needed more immediate-early enhancer and the chicken-actin
intravitreal ranibizumab injections than those promoter; when macaques were given intravitreal
receiving the high dose. As the safety elements of injections, they demonstrated increases in gene
the phase I trial were within acceptable parame- expression in ganglion and Muller cells [45]. For
ters, further investigations were pursued. In line a phase I trial, 19 subjects having serious nAMD
with previous research, subjects in the phase IIa were given one intravitreal AAV2-SFLT01 injec-
clinical trial were randomly assigned subretinal tion of one of four doses (2 × 108 vg, 2 × 109 vg,
rAAV.sFLT-1 gene therapy (1 × 1011 vg) or no 6 × 109 vg, and 2 × 1010 vg); one group was given
treatment, with both groups being given injec- the maximum tolerated dose (2 × 1010 vg) [13]. In
tions of ranibizumab as identical points in time. total, there were 10 adverse events involving five
Avalanche Biotechnologies released preliminary patients, including a retinal tear, retinal hemor-
data in a press release in June 2015 reporting that rhage, and one subject aged 91 years dying
the group receiving a high dosage of rAAV. 12 months following the end of the research.
sFLT-1 gained just 2.2 letters in their mean However, it is important to note that none of these
change for best-corrected visual acuity (BCVA); events are believed to have any relationship to
control group results, in comparison with phase I vector administration. Adverse events were expe-
results, were significantly lower, with 9.3 letters rienced by two of the maximum tolerated dosage
being lost in comparison with the baseline. groups that were classified as being caused by the
Publication of further results occurred in research drug: one subject developed pyrexia that
December 2016, and in March 2019 the 3-year resolved itself within 180 min; a second subject
outcomes of the combined phases I and IIa were developed an intraocular inflammation within a
published [42, 43]. As with phase I, no systemic month of administration which underwent suc-
side effects were demonstrated and no significant cessful treatment with topical steroids.
ocular damage was found from the treatment. Reginexbio also has a new AAV8 vector in
For controls, there was a 61.5–41.0 change in development, RGX-314, which can express a sol-
median BCVA for ETDRS letters; for the group uble anti-VEGF monoclonal antibody within
receiving treatment the change was from 58.5 to transduced retinal cells. Publicly available infor-
47.0. At no point were any statistically significant mation states that in primate eyes that underwent
variations between data points revealed. The treatment with RGX-314 the anterior chamber
number of ranibizumab treatments ranged from showed a maximal expression of therapeutic pro-
5.0 to 15.2, with the median number being 10.5, tein of 4992 ng/mL (Avalanche therapy showed
for those receiving gene therapy; for the control maximum expression of 0.217 ng/mL, and
group the number ranged from 1.0 to 22.0, with a Genzyme 528 ng/mL) [35, 46]. In March 2017 a
phase I open-label dose-escalation trial began with a direct action anti-VEGF agent prior to ini-
(NCT03066258). This clinical trial, which cur- tiating the treatment with bevasiranib.
rently continues, is undertaking the assessment of Sirna-027 (a.k.a. AGN 211745) is an RNA
three doses of RGX-314 with 18 patients who had duplex, 21 nucleotides in length, capable of target-
received previous treatment for nAMD. The pri- ing a conserved region of Flt-1 mRNA molecules
mary endpoint incorporates a safety assessment at in humans, mice, rats, and cynomolgus monkeys
26 weeks. The secondary endpoints, to be assessed [48]. As Flt-1 is a receptor for P1GF and VEGF,
at 106 weeks, include central retinal thickness, receptor knockdown may be more influential than
changes to BCVA, areas of CNV and leakage, and inhibiting single ligands. Pathological neovascu-
mean numbers of anti-VEGF injections. larization was reduced in a laser-induced CNV
RNA interference (RNAi) offers an alterna- model with mice by Sirna-027 [49]. Successful
tive to inhibiting VEGF on a protein basis. A trio completion of a phase I clinical trial has already
of research studies are investigating intravitreal occurred (NCT00363714). Nevertheless, when
injection of chemically modified short-Ambati et al. demonstrated that CNV in mice
interfering RNA (siRNA) molecules. Bevasiranib models could be suppressed by a number of siR-
(a.k.a. Cand5) the earliest nAMD treatment NAs, which included Sirna-027 and bevasiranib,
using siRNA agents, is an RNA duplex, 21 through specific RNAi effects and the activation of
nucleotides in length, intended for targeting the cell surface Toll-like receptor-3, the phase II
VEGF messenger RNA (mRNA). With intravit- study was abandoned (NCT00395057) [50].
real injection, this agent induces catalytic The hypoxia-inducible RTP801 gene is tar-
destruction of mRNA, which silences gene geted by a different RNAi (PF-04523655, an
expression. It has no effect on extant VEGF pro- RNA duplex 19 nucleotides long); there is clear
tein, which suggests that it may provide a syner- target validation for the RNAi molecules men-
gistic effect in administration combined with tioned above. Inhibiting RTP801 causes suppres-
anti-VEGF therapies. The early results of the sion of the rapamycin (m-TOR) pathway in
phase I and II bevasiranib clinical trials mammalian targets, which may reduce synthesis
(NCT00722384/NCT00259753) have shown VEGF-A. Inhibiting m-TOR can additionally
considerable potential in treating nAMD [47]. cause the downstream response of endothelial
However, as yet there are no data in the literature cells regarding VEGF activation to be decreased
presenting the results in detail. A phase III clini- due to intracellular signaling being downregu-
cal trial was not completed due to the unlikeli- lated [51–53]. There were promising outcomes
hood of it meeting its primary endpoint. For that from a phase II clinical trial (NCT00713518),
trial, a comparison was made between how safe demonstrating that monotherapy with combina-
and efficacious bevasiranib, administered at 8- tion therapy with ranibizumab was superior to
or 12-week intervals following a triple injection both ranibizumab and PF-04523655 monother-
for initial pretreatment, was in comparison to apy. There were no identifiable safety issues [54].
ranibizumab-only treatment (NCT00499590).
The preliminary outcomes of this clinical trial
offer indications that bevasiranib shows no effect 28.4.3 Angiostatin/Endostatin
until 6 weeks after treatment has commenced,
suggesting that it could be useful to employ Preclinical safety research employing RetinoStat
combination therapies with bevasiranib as an (Oxford Biomedica) has supported initial clinical
adjunct [47]. It may be that bevasiranib’s effects trials of LV gene therapy vectors for subjects suf-
appear so late due to the way in which it works. fering nAMD (NCT01301443) [55, 56].
Because it is not involved in the elimination of RetinoStat employs the nonhuman, nonreplicat-
extant VEGF but is responsible for the inhibition ing recombinant equine infectious anemia lentivi-
of fresh VEGF molecules synthesizing, it may be rus (EIAV-LV). This treatment is given to humans
necessary to neutralize extant VEGF in the eye subretinally, chiefly transducing target RPE cells
and leading to strong angiostatic proteins, human teins over a long period spurred by just one round
endostatin, and angiostatin, being expressed [57, of therapy.
58]. RetinoStat has been created for expressing
both of these proteins because each has a syner- Key Learning Points
gistic effect on the formation of new blood vessels • The usage of gene therapy in a clinical context
[59, 60]. Phase 1 trials took 21 subjects with has recently started to be applied to neovascu-
severe refractory nAMD. The subject was given a lar age-related macular degeneration (nAMD).
subretinal viral dose after PPV of either 2.4 × 104, Such therapies could mitigate the heavy treat-
2.4 × 105, or 8.0 × 105 transduction units. During ment loads experienced with chronic intravit-
the study, the eyes that had been treated exhibited real therapies and may offer improvements to
continuously high levels of expression of end- the unsatisfactory visual results of “real-
ostatin and angiostatin. 71% of patients experi- world” undertreatment with anti-vascular
enced leakage reductions on FA. Nevertheless, endothelial growth factor (anti-VEGF).
significant reductions against baseline levels for • Adeno-associated virus (AAV) vectors are fre-
intra-/subretinal fluid only occurred with one quently employed in gene therapy to deliver
patient [61]. These results were not particularly genes due to the fact that they are nonpatho-
promising, but because patients with advanced genic, offer low risk of inflammation, can
nAMD were enrolled in this trial, the improve- transduce nondividing cells, have low levels
ments might have been limited. The surgery of retinal toxicity (when dosed appropriately),
caused patients to develop retinal tears and one to and trials on human subjects have an extremely
develop a macular hole; these were managed good safety record.
without further complications. No adverse effects • At present, there are two ways in which a viral
related to the lentiviral vector were noted. vector can be administered for targeting of
retinal cells, either intravitreal injections or
subretinal injections.
28.5 Conclusions • In relation to nAMD, assessments are being
made of gene therapy as a means of chroni-
Gene therapy demonstrates considerable promise cally expressing anti-angiogenic proteins, e.g.,
as a therapy for several diseases, including genetic fms-like tyrosine kinase-1 and pigment epithe-
disorders and also chronic and refractory retinal lium-derived factor, and also endostatin,
disease. Nevertheless, a number of issues still exist aflibercept, angiostatin, and ranibizumab.
in relation to using gene therapy as an approved • If gene therapy is to be approved for the treat-
treatment. The first important issue relates to the ment of nAMD than a number of concerns
ability to deliver a gene to the correct cell. Once must be resolved, including the triggering of
the gene is delivered in the target cell, it must be adverse immune responses, target cell deliv-
activated to function properly. The second issue is ery, and off-target incidents.
that when a gene has been delivered it may have
off-target influences, with the new gene being
inserted into other genes and disrupting their activ- References
ities, harming the patient. The final issue is how to
1. Bennett J. Taking stock of retinal gene therapy:
avoid immune responses: the host may regard
looking back and moving forward. Mol Ther.
newly introduced genes as being harmful antigens 2017;25(5):1076–94.
and so initiate harmful immune responses. 2. Hafler BP. Clinical progress in inherited retinal degen-
Should gene therapy prove itself to be signifi- erations: gene therapy clinical trials and advances in
cantly more effective and safer than the treat- genetic sequencing. Retina. 2017;37(3):417–23.
3. Morral N, O'Neal W, Zhou H, Langston C, Beaudet
ments currently available, it has the potential to A. Immune responses to reporter proteins and
provide therapy for nAMD and IRD through the high viral dose limit duration of expression with
induction of the expression of therapeutic pro- adenoviral vectors: comparison of E2a wild
type and E2a deleted vectors. Hum Gene Ther. 20. Streilein JW. Ocular immune privilege: therapeutic
1997;8(10):1275–86. opportunities from an experiment of nature. Nat Rev
4. Gao GP, Yang Y, Wilson JM. Biology of adenovirus Immunol. 2003;3(11):879–89.
vectors with E1 and E4 deletions for liver-directed 21. Yu-Wai-Man P. Genetic manipulation for inherited
gene therapy. J Virol. 1996;70(12):8934–43. neurodegenerative diseases: myth or reality? Br J
5. Morral N, O’Neal W, Rice K, Leland M, Kaplan J, Ophthalmol. 2016;100(10):1322–31.
Piedra PA, et al. Administration of helper-dependent 22. Cheng L, Toyoguchi M, Looney DJ, Lee J, Davidson
adenoviral vectors and sequential delivery of dif- MC, Freeman WR. Efficient gene transfer to retinal
ferent vector serotype for long-term liver-directed pigment epithelium cells with long-term expression.
gene transfer in baboons. Proc Natl Acad Sci USA. Retina. 2005;25(2):193–201.
1999;96(22):12816–21. 23. Petrs-Silva H, Dinculescu A, Li Q, Min SH, Chiodo
6. Morral N, Parks RJ, Zhou H, Langston C, Schiedner V, Pang JJ, et al. High-efficiency transduction of the
G, Quinones J, et al. High doses of a helper-dependent mouse retina by tyrosine-mutant AAV serotype vec-
adenoviral vector yield supraphysiological levels of tors. Mol Ther. 2009;17(3):463–71.
alpha1-antitrypsin with negligible toxicity. Hum Gene 24. Dalkara D, Kolstad KD, Caporale N, Visel M,
Ther. 1998;9(18):2709–16. Klimczak RR, Schaffer DV, et al. Inner limiting mem-
7. Kay MA, Nakai H. Looking into the safety of AAV brane barriers to AAV-mediated retinal transduction
vectors. Nature. 2003;424(6946):251. from the vitreous. Mol Ther. 2009;17(12):2096–102.
8. Lisowski L, Tay SS, Alexander IE. Adeno-associated 25. Lee SH, Colosi P, Lee H, Ohn YH, Kim SW, Kwak
virus serotypes for gene therapeutics. Curr Opin HW, et al. Laser photocoagulation enhances adeno-
Pharmacol. 2015;24:59–67. associated viral vector transduction of mouse retina.
9. Daya S, Berns KI. Gene therapy using adeno- Hum Gene Ther Methods. 2014;25(1):83–91.
associated virus vectors. Clin Microbiol Rev. 26. Bennett J, Ashtari M, Wellman J, Marshall KA,
2008;21(4):583–93. Cyckowski LL, Chung DC, et al. AAV2 gene therapy
10. Stieger K, Cronin T, Bennett J, Rolling F. Adeno- readministration in three adults with congenital blind-
associated virus mediated gene therapy for reti- ness. Sci Transl Med. 2012;4(120):120ra15.
nal degenerative diseases. Methods Mol Biol. 27. Li Q, Miller R, Han PY, Pang J, Dinculescu A, Chiodo
2011;807:179–218. V, et al. Intraocular route of AAV2 vector administra-
11. Surace EM, Auricchio A. Versatility of AAV vectors tion defines humoral immune response and therapeu-
for retinal gene transfer. Vis Res. 2008;48(3):353–9. tic potential. Mol Vis. 2008;14:1760–9.
12. Guy J, Feuer WJ, Davis JL, Porciatti V, Gonzalez PJ, 28. Campochiaro PA. Gene transfer for ocular neo-
Koilkonda RD, et al. Gene therapy for leber heredi- vascularization and macular edema. Gene Ther.
tary optic neuropathy: low- and medium-dose visual 2012;19(2):121–6.
results. Ophthalmology. 2017;124(11):1621–34. 29. Campochiaro PA, Nguyen QD, Shah SM, Klein ML,
13. Heier JS, Kherani S, Desai S, Dugel P, Kaushal S, Holz E, Frank RN, et al. Adenoviral vector-delivered
Cheng SH, et al. Intravitreous injection of AAV2- pigment epithelium-derived factor for neovascular
sFLT01 in patients with advanced neovascular age- age-related macular degeneration: results of a phase
related macular degeneration: a phase 1, open-label I clinical trial. Hum Gene Ther. 2006;17(2):167–76.
trial. Lancet. 2017;390(10089):50–61. 30. Mori K, Duh E, Gehlbach P, Ando A, Takahashi K,
14. Lee S, Kang IK, Kim JH, Jung BK, Park K, Chang Pearlman J, et al. Pigment epithelium-derived factor
H, et al. Relationship between neutralizing antibod- inhibits retinal and choroidal neovascularization. J
ies against adeno-associated virus in the vitreous and Cell Physiol. 2001;188(2):253–63.
serum: effects on retinal gene therapy. Transl Vis Sci 31. Mori K, Gehlbach P, Ando A, McVey D, Wei L,
Technol. 2019;8(2):14. Campochiaro PA. Regression of ocular neovascular-
15. Everson EM, Trobridge GD. Retroviral vector inter- ization in response to increased expression of pigment
actions with hematopoietic cells. Curr Opin Virol. epithelium-derived factor. Invest Ophthalmol Vis Sci.
2016;21:41–6. 2002;43(7):2428–34.
16. Kumar M, Keller B, Makalou N, Sutton RE. Systematic 32. Bouck N. PEDF: anti-angiogenic guardian of ocular
determination of the packaging limit of lentiviral vec- function. Trends Mol Med. 2002;8(7):330–4.
tors. Hum Gene Ther. 2001;12(15):1893–905. 33. Ohno-Matsui K, Morita I, Tombran-Tink J, Mrazek
17. Booth C, Gaspar HB, Thrasher AJ. Treating immu- D, Onodera M, Uetama T, et al. Novel mechanism
nodeficiency through HSC gene therapy. Trends Mol for age-related macular degeneration: an equilibrium
Med. 2016;22(4):317–27. shift between the angiogenesis factors VEGF and
18. Zufferey R, Dull T, Mandel RJ, Bukovsky A, Quiroz PEDF. J Cell Physiol. 2001;189(3):323–33.
D, Naldini L, et al. Self-inactivating lentivirus vector 34. Luo L, Uehara H, Zhang X, Das SK, Olsen T, Holt D,
for safe and efficient in vivo gene delivery. J Virol. et al. Photoreceptor avascular privilege is shielded by
1998;72(12):9873–80. soluble VEGF receptor-1. elife. 2013;2:e00324.
19. Shaw A, Cornetta K. Design and potential of 35. Lai CM, Estcourt MJ, Himbeck RP, Lee SY, Yew-San
non-integrating lentiviral vectors. Biomedicine. Yeo I, Luu C, et al. Preclinical safety evaluation of
2014;2(1):14–35. subretinal AAV2.sFlt-1 in non-human primates. Gene
Ther. 2012;19(10):999–1009.
36. Lai CM, Brankov M, Zaknich T, Lai YK, Shen WY, 49. Krzystolik MG, Afshari MA, Adamis AP,
Constable IJ, et al. Inhibition of angiogenesis by Gaudreault J, Gragoudas ES, Michaud NA, et al.
adenovirus-mediated sFlt-1 expression in a rat model Prevention of experimental choroidal neovascular-
of corneal neovascularization. Hum Gene Ther. ization with intravitreal anti-vascular endothelial
2001;12(10):1299–310. growth factor antibody fragment. Arch Ophthalmol.
37. Lai CM, Estcourt MJ, Wikstrom M, Himbeck RP, 2002;120(3):338–46.
Barnett NL, Brankov M, et al. rAAV.sFlt-1 gene 50. Kleinman ME, Yamada K, Takeda A, Chandrasekaran
therapy achieves lasting reversal of retinal neo- V, Nozaki M, Baffi JZ, et al. Sequence- and target-
vascularization in the absence of a strong immune independent angiogenesis suppression by siRNA via
response to the viral vector. Invest Ophthalmol Vis TLR3. Nature. 2008;452(7187):591–7.
Sci. 2009;50(9):4279–87. 51. Sabers CJ, Martin MM, Brunn GJ, Williams
38. Drga V, Plack CJ, Yasin I. Frequency tuning of the JM, Dumont FJ, Wiederrecht G, et al. Isolation
efferent effect on cochlear gain in humans. Adv Exp of a protein target of the FKBP12-rapamycin
Med Biol. 2016;894:477–84. complex in mammalian cells. J Biol Chem.
39. Rakoczy EP, Lai CM, Magno AL, Wikstrom ME, 1995;270(2):815–22.
French MA, Pierce CM, et al. Gene therapy with 52. Corradetti MN, Inoki K, Guan KL. The stress-
recombinant adeno-associated vectors for neovascu- inducted proteins RTP801 and RTP801L are nega-
lar age-related macular degeneration: 1 year follow- tive regulators of the mammalian target of rapamycin
up of a phase 1 randomised clinical trial. Lancet. pathway. J Biol Chem. 2005;280(11):9769–72.
2015;386(10011):2395–403. 53. Dumont FJ, Su Q. Mechanism of action of the immuno-
40. Constable IJ, Lai CM, Magno AL, French MA, suppressant rapamycin. Life Sci. 1996;58(5):373–95.
Barone SB, Schwartz SD, et al. Gene therapy in neo- 54. Nguyen QD, Schachar RA, Nduaka CI, Sperling
vascular age-related macular degeneration: three-year M, Klamerus KJ, Chi-Burris K, et al. Evaluation of
follow-up of a phase 1 randomized dose escalation the siRNA PF-04523655 versus ranibizumab for
trial. Am J Ophthalmol. 2017;177:150–8. the treatment of neovascular age-related macular
41. Brown DM, Kaiser PK, Michels M, Soubrane G, degeneration (MONET Study). Ophthalmology.
Heier JS, Kim RY, et al. Ranibizumab versus verte- 2012;119(9):1867–73.
porfin for neovascular age-related macular degenera- 55. Igarashi T, Miyake K, Kato K, Watanabe A, Ishizaki
tion. N Engl J Med. 2006;355(14):1432–44. M, Ohara K, et al. Lentivirus-mediated expression of
42. Constable IJ, Pierce CM, Lai CM, Magno AL, Degli- angiostatin efficiently inhibits neovascularization in
Esposti MA, French MA, et al. Phase 2a randomized a murine proliferative retinopathy model. Gene Ther.
clinical trial: safety and post hoc analysis of subretinal 2003;10(3):219–26.
rAAV.sFLT-1 for wet age-related macular degenera- 56. Lai LJ, Xiao X, Wu JH. Inhibition of corneal neo-
tion. EBioMedicine. 2016;14:168–75. vascularization with endostatin delivered by adeno-
43. Rakoczy EP, Magno AL, Lai CM, Pierce CM, Degli- associated viral (AAV) vector in a mouse corneal
Esposti MA, Blumenkranz MS, et al. Three-year injury model. J Biomed Sci. 2007;14(3):313–22.
follow-up of phase 1 and 2a rAAV.sFLT-1 subretinal 57. Binley K, Widdowson PS, Kelleher M, de Belin J,
gene therapy trials for exudative age related macular Loader J, Ferrige G, et al. Safety and biodistribution
degeneration. Am J Ophthalmol. 2019;204:113–23. of an equine infectious anemia virus-based gene ther-
44. Pechan P, Rubin H, Lukason M, Ardinger J, DuFresne apy, RetinoStat((R)), for age-related macular degen-
E, Hauswirth WW, et al. Novel anti-VEGF chime- eration. Hum Gene Ther. 2012;23(9):980–91.
ric molecules delivered by AAV vectors for inhi- 58. Balaggan KS, Binley K, Esapa M, Iqball S, Askham
bition of retinal neovascularization. Gene Ther. Z, Kan O, et al. Stable and efficient intraocular gene
2009;16(1):10–6. transfer using pseudotyped EIAV lentiviral vectors. J
45. Yin L, Greenberg K, Hunter JJ, Dalkara D, Kolstad Gene Med. 2006;8(3):275–85.
KD, Masella BD, et al. Intravitreal injection of AAV2 59. Scappaticci FA, Smith R, Pathak A, Schloss D,
transduces macaque inner retina. Invest Ophthalmol Lum B, Cao Y, et al. Combination angiostatin and
Vis Sci. 2011;52(5):2775–83. endostatin gene transfer induces synergistic anti-
46. Maclachlan TK, Lukason M, Collins M, Munger R, angiogenic activity in vitro and antitumor efficacy
Isenberger E, Rogers C, et al. Preclinical safety evalua- in leukemia and solid tumors in mice. Mol Ther.
tion of AAV2-sFLT01—a gene therapy for age-related 2001;3(2):186–96.
macular degeneration. Mol Ther. 2011;19(2):326–34. 60. Yokoyama Y, Dhanabal M, Griffioen AW, Sukhatme
47. Singerman L. Combination therapy using the small VP, Ramakrishnan S. Synergy between angiostatin
interfering RNA bevasiranib. Retina. 2009;29(6 and endostatin: inhibition of ovarian cancer growth.
Suppl):S49–50. Cancer Res. 2000;60(8):2190–6.
48. Shen J, Samul R, Silva RL, Akiyama H, Liu H, 61. Campochiaro PA, Lauer AK, Sohn EH, Mir TA, Naylor
Saishin Y, et al. Suppression of ocular neovascular- S, Anderton MC, et al. Lentiviral vector gene transfer
ization with siRNA targeting VEGF receptor 1. Gene of endostatin/angiostatin for macular degeneration
Ther. 2006;13(3):225–34. (GEM) study. Hum Gene Ther. 2017;28(1):99–111.
Dr. Joo Yong Lee is the Associate Professor of

Ophthalmology and a member of the vitreoretinal faculty
Joon Hyung Yeo graduated from Chung-Ang University,
at Asan Medical Center and University of Ulsan College
South Korea. He subsequently completed his ophthalmol-
of medicine. Dr. Lee received his medical degree at Seoul
ogy residency at the Chung-Ang University Hospital in
National University and his ophthalmology residency
Seoul. He then concluded a vitreoretinal fellowship at the
training at Asan Medical Center in Seoul, Korea. He com-
Asan Medical Center under the mentorship of Drs. Young
pleted a 2-year clinical vitreoretinal fellowship at Asan
Hee Yoon and Joo Yong Lee.
Medical Center. He joined the faculty at Asan Medical
His research and clinical interests involve diabetic reti-
Center in 2009.
nopathy, novel medical and surgical therapies of age-
He has published over 50 articles in peer-reviewed
related macular degeneration, hereditary retinal dystrophy
journals and has an active clinical and basic science
including retinitis pigmentosa, and gene therapy. He has
research program. Dr. Lee treats patients with a variety of
participated in numerous clinical trials of new therapies
medical and surgical vitreoretinal diseases. His clinical
for vitreoretinal diseases.
research interests include novel medical and surgical ther-
apies of age-related macular degeneration, retinitis pig-
mentosa, and other posterior segment disorders. He has
been a pioneer in the development of stem cell-based cell
therapy and gene delivery to the retina to treat ocular dis-
ease. He has participated in numerous clinical trials of
new therapies for vitreoretinal diseases. He directs a basic
research program to develop a new treatment option for
age-related macular degeneration and retinitis pigmentosa
and investigate the mechanisms responsible for uveitis.
Dr. Lee serves on the Editorial Board of the journal of
Retina and reviews manuscripts for a variety of ophthal-
mology journals.
Radiotherapy for Choroidal
David Pérez González, Matias Iglicki,
and Dinah Zur
29.1 Introduction multiple medical visits and injections over years,

as well as an increasing heavy economic burden
Age-related macular degeneration (AMD) is the on healthcare systems and providers. In this
leading cause of blindness in people over 65-years sense, reducing the number of injections without
old [1]. Currently, the main objective to treat cho- compromising outcomes would be desirable.
roidal neovascularization (CNV) in neovascular These are part of the objectives aimed to reach
AMD is focused on reducing the amount of vaso- with the use of radiotherapy [5].
proliferative factors, mainly vascular endothelial
growth factor (VEGF) [2].
The prognosis of patients with neovascular 29.2 Basic Principles
AMD was radically favored with the onset of in Radiotherapy
anti-VEGF intravitreal therapy. Ranibizumab for Neovascular AMD
and aflibercept are approved for intravitreal use
in the treatment of CNV secondary to When energy of sufficient level is emitted to a
AMD. Bevacizumab has been widely used off- cell surface or tissue, electrons separate from
label in clinical practice, showing comparable their respective atoms; this process is known as
visual outcomes [3, 4]. However, treatment with ionization. The ionization of oxygen atoms pro-
anti-VEGF agents does not cure the disease; duces reactive oxygen species with consequent
nonetheless, it reduces the rate of progression damage to DNA molecules and eventually cell
and establishes good visual outcomes in most death [6, 7]. Albeit all the cells in the area to be
patients. With an aging population, projections radiated are at risk to be damaged, the effect of
indicate that there will be 200 million AMD radiation is specific and relatively localized as
patients globally by 2020. This situation is caus- only cells with active replication retain the capac-
ing a tremendous burden on patients needing ity for DNA repair. Endothelial cells in particular
are much more radiosensitive than any other
mesenchymal cell in the body [8, 9].
D. Pérez González · D. Zur (*) Ionizing radiation is especially promising
Ophthalmology Division, Tel Aviv Medical Center, as a synergic treatment in neovascular AMD
Affiliated to the Sackler Faculty of Medicine, due to its anti-angiogenic properties by
Tel Aviv University, Tel Aviv, Israel
decreasing capillary permeability, increasing
M. Iglicki vascular flow rate, decreasing cellular stasis,
Private Retina Practice, University of Buenos Aires,
influencing capillary closure, and also due to
Buenos Aires, Argentina

392 D. Pérez González et al.
its important anti- inflammatory as well as

anti-fibrotic effects [7, 10].
29.3 Current Approaches

a
for Radiotherapy in AMD
b
In medical practice, two types of radiotherapy
can be differentiated: Brachytherapy, where the
radiation source is placed directly or as close as
possible to the treatment site, and teletherapy in
which the radiation is emitted from a distant
source and is projected toward the tissue or area
of interest [5, 10, 11].
The use of radiation in other ocular patholo-
Fig. 29.1 Once the vitrectomy is performed, the probe
gies has been noted since many years ago. It was (a) is held by the surgeon above the lesion (b) for approxi-
until 1993 when its implementation directly asso- mately 4–5 min. The distance and the location of the
ciated with the treatment of choroidal neovascu- probe are hands skills dependent and crucial for the suc-
larization in AMD began. Early clinical studies cess of the treatment. Image courtesy of Estephania Tirado
Torrero
showed contradictory results [12–15].
The use of anti-VEGF has a rapid mechanism
of action, however poor durability. On the other removed, the surgeon places the source of emis-
hand, radiotherapy has a slow-acting effect but sion above the area to be treated, using a dedi-
provides a long time of action. This builds the cated device. The probe with the emission source
scientific rationale to consider a synergistic is held in place during 4–5 min, transmitting 24
response when combining radiotherapy and anti- Gray (Gy) to the macular lesion [6, 10]. The suc-
VEGF treatment [6]. cess of the procedure mainly depends on the hand
There are currently two systems to approach skills of the surgeon, i.e., accurate and stable
radiation therapy in AMD: the epimacular localization of the emission probe, which is not
brachytherapy (Vidion Anti-Neovascular easy at all, and minimal distance changes or time
Epimacular Brachytherapy (EMBT), System variance can rely on poor or less effective out-
NeoVista Inc. Fremont, CA), and stereotactic comes (Fig. 29.1).
radiation (IRay Radiotherapy System Oraya While this approach bares the known compli-
Therapeutics Inc. Newark, CA) [10]. cations of vitrectomy, it should be noted that
there is good evidence that the removal of the vit-
reous ameliorates oxygenation of the inner reti-
29.3.1 Epimacular Brachytherapy nal layers [16]. In addition, the simultaneous
for Neovascular AMD application of brachytherapy provides a synergis-
tic action in the production of reactive oxygen
This approach aims to provide treatment in a species with double-stranded DNA and hereby
localized manner, i.e., adjacent to the vascular prevention of further formation of neovascular
proliferation site. This technique uses beta radia- membranes [6, 17].
tion, whose dose declines with increasing dis-
tance from the source of emission, limiting 29.3.1.1 Major Trials
exposure, and allaying damage to adjacent ocular in Brachytherapy
structures [7, 10]. Epimacular brachytherapy is for Neovascular AMD
an invasive procedure since it is delivered through One of the pivotal studies demonstrating the
pars plana vitrectomy. Once the vitreous has been safety and efficacy of brachytherapy for AMD
29 Radiotherapy for Choroidal Neovascularization 393
arose in 2009 [18]. This was a non-randomized, 15 letters. Although EMBT is an invasive proce-
multicenter feasibility study. Thirty-four dure, patients in this study needed only a small
treatment-naïve patients with subfoveal predom- number of subsequent bevacizumab injections
inantly classic, minimally classic, or occult CNV over 3 years. The authors hypothesize that if not
lesions due to AMD were included. They treated by EMBT, these patients would have
received a single intraocular treatment of either needed a high number of reinjections in order to
15-Gy or 24-Gy beta radiation, using stron- reach those visual outcomes [19].
tium-90 with an intraocular delivery device.
Patients were evaluated for 12 months; at the end The MERITAGE Study
of the study, there were no adverse effects asso- Shortly after these two pivotal studies, the role of
ciated with radiation exposure. A higher percent- brachytherapy for patients who have been previ-
age of patients who received 24-Gy of radiation ously treated with multiple anti-VEGF injections
(76%) improved or maintained their visual acu- for neovascular AMD was investigated. The
ity. The authors concluded proper short-term MERITAGE study was a prospective, multi-
safety and efficacy of intraocular epiretinal center, interventional, noncontrolled clinical trial
delivery of beta radiation [18]. which was initiated in the United Kingdom and
A 3-year follow-up study evaluated the long- then subsequently expanded to the United States
term safety and visual acuity outcomes of brachy- and Israel [20].
therapy combined with intravitreal bevacizumab Fifty-three patients with AMD requiring fre-
[19]. A total of 34 subjects were included and fol- quent anti-VEGF injections were enrolled.
lowed for 24 months. Based on the beneficial Participants received a 24-Gy dose of beta radi-
effects of higher radiation doses shown in the ear- ation delivered via pars plana vitrectomy, fol-
lier trial, each subject received 24-Gy beta radia- lowed by ranibizumab injections administered
tion treatment with a simultaneous intravitreal monthly if needed (PRN). Optical coherence
bevacizumab injection, and a second dose 1 tomography (OCT) monitoring was performed
month after brachytherapy. Further treatment monthly to evaluate anatomical response. After
with bevacizumab was bestowed as a pro-re-nata a single treatment with radiation therapy, 81%
(PRN) regimen. maintained stable vision with an average of 3.49
No radiation retinopathy events were identi- ranibizumab injections over 12 months. This
fied over the follow-up period. However, com- finding was of particular interest since vitrec-
pared to the prior study, there was a higher tomy is believed to reduce the clearance of some
percentage of cataract formation. At 24 months, intravitreal drugs [21]. The MERITAGE study
50% of the patients exhibited two grades of pro- proved a reduction in CNV activity by the need
gression according to LOCS II classification sys- for less frequent anti-VEGF injections.
tem and a total of five eyes underwent cataract However, it was not possible to determine if this
extraction. could be translated to fewer hospital visits, as all
Mean best-corrected visual acuity (BCVA) patients in this prospective trial were monitored
demonstrated an average gain of +15.0 letters at monthly by OCT. [20] Twenty-four months’
12 months and drop to −4.9 letters from baseline results showed that 68.1% of participants main-
at 24 months. The drop in mean gain at month 24 tained stable visual acuity and received a mean
was mainly attributable to cataract formation. At of 8.7 ranibizumab retreatment injections over
month 36, mean VA improved to +3.9 letters, 24 months [22].
which can be explained by cataract extraction
performed in part of the phakic eyes during the The CABERNET Study
third year of follow-up. Moreover, at 36 months, This was a multicenter, randomized, active-
90% of the patients showed stability, 53% gained controlled, phase III clinical trial that evaluated
more than one letter, and 21% gained more than the safety and efficacy of epimacular brachyther-
apy (EMBT) in conjunction with anti-VEGF was 84% in the EMBT arm compared to 92% in
therapy in naïve neovascular AMD patients [23]. the ranibizumab arm. In the EMBT arm, the
Four hundred ninety-four patients with clas- mean total lesion and CNV size increased signifi-
sic, minimally classic, and occult lesions were cantly more than in the ranibizumab group.
enrolled and then randomized (2:1) to EMBT or Hence, the results of the MERLOT study did not
a ranibizumab monotherapy control arm. The support the hypothesis established previously. On
EMBT group received two monthly loading the contrary, patients who underwent EMBT
injections of 0.5 mg ranibizumab. The control received more intravitreal injections, and had a
arm received three mandated, monthly loading remarkably worse visual acuity than those on
injections of ranibizumab, and then fixed quar- ranibizumab monotherapy.
terly injections. These results differ in some proportion from
The results of this study were controversial: what was shown in the MERITAGE study, where
after 24 months of follow-up, 77% of the EMBT a reduction in the frequency of injections was
group and 90% of the control group lost fewer found for patients undergoing adjuvant treatment
than 15 letters. The EMBT group did not meet with EMBT. The reasons for these differences
the superiority margin, predefined as a gain of may be associated with the surgical technique,
≥15 letters in vision (16% for the EMBT group probe accuracy, surgeon’s hand skills, or for
vs. 26% for the control group). While there was a some other unidentified factors.
statistically significant difference for patients The MERLOT study group concluded a lack
with occult lesions, none was found predomi- of support for EMBT in chronic, active CNV
nantly classic and minimally classic lesions. patients with AMD [24].
Mean visual acuity change was −2.5 letters in the
EMBT group and +4.4 letters in the control arm. 29.3.1.2 Other Trials Involving
Although the EMBT group received less injec- Brachytherapy
tions (6.2 ranibizumab injections vs. 10.4 in the for Neovascular AMD
control arm), the data in this trial did not support A retrospective study evaluated the clinical feasi-
the routine use of EMBT in patients with naïve bility, safety, and efficacy of EMBT for CNV in
neovascular AMD despite the good safety profile AMD for unresponsive eyes after repeated anti-
[23]. VEGF injections in real-life conditions [1]. A
total of 22 patients were treated with EMBT and
The MERLOT Study retreated with anti-VEGF injections on an as-
This study evaluated the safety and efficacy of needed basis if subretinal or intraretinal fluid was
EMBT in patients with chronic neovascular detected on OCT.
AMD previously treated with ranibizumab with At 12 months, half of the patients maintained
the main hypothesis that the use of EMBT will stable visual acuity, 40% showed improvement,
reduce the number of intravitreal injections and and only 10% presented decrease in vision with
will maintain a non-inferior visual acuity [24]. A loss of more than 3 Snellen lines. The average
total of 363 patients were enrolled in order to number of injections before treatment with
receive either pars plana vitrectomy with 24-Gy EMBT was 13.8, compared to 5.5 ± 4.4 during
EMBT and continued ranibizumab injections as 12 months after. It is important to mention that
needed or monotherapy with PRN ranibizumab there was lacking data about the exact number of
injections. anti-VEGF injections patients received during
After 12 months, the mean number of PRN the 1-year period prior EMBT treatment, and
injections was 4.8 in the EMBT arm and 4.1 in therefore, there was no standard analysis in rela-
the ranibizumab monotherapy arm. The proportion to the reduction of injection frequency
tion of participants losing fewer than 15 letters 12 months before and after EBMT. The authors
concluded that EMBT was safe and feasible for

use in clinical practice, and could help in cases of
resistant disease with a poor response even after
multiple anti-VEGF injections [1].
In 2013, a novel brachytherapy device called
SMD-1 was proposed using a less invasive sub-
Tenon retrobulbar approach. A small study
a
included 6 patients who received 24-Gy radiation
delivered over 5.5 min using the SMD-1 brachy- b
therapy probe [25]. After removal of the probe,
patients received concomitant anti-VEGF injec-
tions with further PRN regimen. After 12 months
of follow-up, three out of the six patients showed
Fig. 29.2 The IRay technique uses a robotically con-
stabilization and improvement of visual acuity. trolled delivery system which is connected to the patient
Although this approach seems less invasive, the through a contact lens (a) with a suction force of approxi-
technique was used only in a very small number mately 25 mmHg. Once placed on the patient, 2–3 sepa-
of patients, and results so far are inconclusive. rate beams of radiation are emitted entering the globe
through the lower portion of the pars plana (b), overlap-
ping the predicted foveal center. Image courtesy of
Estephania Tirado Torrero
29.3.2 Stereotactic Radiotherapy
for Neovascular AMD
patient through a contact lens with a suction force
Consideration for teletherapy in neovascular of approximately 25 mmHg. Once placed on the
AMD began around the 1990s. Initial studies patient, 2–3 separate beams of radiation are emit-
generally used a large amount of radiation with ted entering the globe through the lower portion
little precision range over the area to be treated, of the pars plana, overlapping the predicted
so the results were not favorable, especially com- foveal center. When the beam of radiation enters
pared to those achieved since the introduction of the eye, it disperses, reducing exposure and dam-
anti-VEGF injections. A Cochrane Database age to adjacent ocular structures, mainly the lens
Systematic review that analyzed around 30 stud- and the optic nerve. During the procedure, the
ies using teletherapy concluded that at that time, patient can be monitored through a small win-
the use of this approach did not provide enough dow, and the eye is continuously tracked. If an
evidence to be recommended as a treatment in unexpected movement occurs an inbuilt safety
patients with neovascular AMD [26, 27]. feature interrupts the process allowing the physi-
Recently, the way in which teletherapy is used cian to avoid misdirected beams into the eye [10,
has evolved. Nowadays specifically with the IRay 28] (Fig. 29.2).
system (formerly Oraya Therapeutics Inc.,
Newark, CA), the radiation source is precisely 29.3.2.1 Main Trials in Stereotactic
direct to the site of interest from different angles, Radiotherapy
minimizing damage to healthy surrounding tis- for Neovascular AMD
sues [5]. This device uses a low-voltage X-ray The INTREPID Study
source which does not require the same degree of This was a large randomized, double-masked,
radiation shielding as the first linear accelerators sham-controlled, multicenter, clinical trial with
that were investigated in the past. the main objective to determine the safety and
The IRay technique uses a robotically con- efficacy of stereotactic radiotherapy (SRT) for
trolled delivery system which is connected to the patients with neovascular AMD [9].
A total of 230 participants who received more are randomized to receive a single session of SRT
than three anti-VEGF injections (ranibizumab or with 16-Gy or sham, with a concomitant baseline
bevacizumab) during the year prior to recruit- ranibizumab injection, and then continue under
ment and still requiring injections at the time of the PRN regimen.
the study were considered. The primary outcome will be the number of
Patients received 16-Gy, 24-Gy, or sham injections needed over the 2-year follow-up
SRT. All received injections of ranibizumab period. Accordingly, additional visits will be
using a PRN regimen during the first 12 months, considered at 36 and 48 months to evaluate the
then continued with either bevacizumab or ranibi- development of radiation retinopathy. The esti-
zumab also with PRN regimen with a total fol- mated primary completion date will be estimated
low-up of 24 months. Interestingly, the number in October 2021, with the final results for March
of injections was significantly reduced with the 2024 [33].
use of SRT during the first 12 months. Results
during the second year were markedly favorable
showing that 15% of patients who received SRT 29.3.3 Main Complications
did not require any injection, and 43% received Associated with Radiotherapy
three or fewer injections. in Neovascular AMD
During the second year, 3% of the participants
showed microvascular abnormalities; however, Complications due to radiotherapy vary
the vision was not affected because most of these depending on the method and the amount of
lesions did not involve the fovea. energy emitted to each ocular structure. The use
Albeit 2-year results showed promising of radiotherapy in patients with neovascular
results, potential long-term complications of AMD differs from other modalities, and then,
radiation exposure remain a major concern with the rate and severity of complications are off-
SRT and need to be further investigated [29–31]. beat. Also, with the advance of new techniques
Although both, the CABERNET and that sharpened the precision and diminished
INTREPID studies showed improvement in extension to unwanted structures, these compli-
vision, it is important to point out that the tech- cations are becoming less frequent. By far, the
nique of radiation transmission is different and most affected structures are the lens, the optic
might impact the outcomes. EBMT used in the nerve, and the retina.
CABERNET study bares the limitations of The reported thresholds for clinically
surgeon-dependent procedures as detailed above observable radiation damage in the lens and
[32]. On the contrary, stereotactic radiation elim- progression of cataracts are around 2-Gy.
inates the human error factor during the Usually, brachytherapy and stereotactic radio-
treatment. therapy will not exceed the amount of 0.13-Gy
emitted to the lens. Radiation retinopathy is
The STAR Study usually observed with levels around 33 to
This is a multicenter, double-masked, random- 55-Gy. Generally, with the new treatment
ized, sham-controlled 24-months clinical trial modalities the amount of emission to the retinal
and one of the most recent studies, with an ongo- surface outside the fovea will not exceed 24-Gy.
ing recruitment phase at the time of writing this Recent studies report mostly microvascular
chapter. Patients with chronic and active neovas- abnormalities without showing decreased
cular AMD are enrolled with the aim to evaluate visual acuity and even remission of these
the efficacy of SRT in this population. Participants changes during the follow-up is possible [31].
Optic neuropathy is usually observed after References

exposure to a much greater amount of radiation
considering thresholds above 55-Gy. The 1. Zur D, Loewenstein A, Barak A. One-year results from
clinical practice of epimacular strontium-90 brachy-
amount of radiation with new modalities will therapy for the treatment of subfoveal choroidal neo-
not exceed 2.4-Gy [7, 19, 34]. vascularization secondary to AMD. Ophthalmic Surg
Although there is evidence that the use of Lasers Imaging Retina. 2015 Mar;46(3):338–43.
radiotherapy seems safe, caution is advised as 2. Kaiser PK. Emerging therapies for neovascular age-
related macular degeneration. Ophthalmology. 2013
clinically visible structural damage caused by May;120(5):S11–5.
radiation can appear several years after exposure. 3. Villegas VM, Aranguren LA, Kovach JL, Schwartz
Further studies evaluating the rate of long-term SG, Flynn HW. Current advances in the treatment of
complications are needed in order to establish a neovascular age-related macular degeneration. Expert
Opin Drug Deliv. 2017 Feb;14(2):273–82.
better understanding of the potential and risks for 4. Michels S, Rosenfeld P, Puliafito C, Marcus E,
radiotherapy as an adjuvant treatment in neovas- Venkatraman A. Systemic bevacizumab (avastin)
cular AMD. therapy for neovascular age-related macular degen-
eration twelve-week results of an uncontrolled
open-label clinical study. Ophthalmology. 2005
Key Learning Points and Conclusions Jun;112(6):1035–1047.e9.
1. The adjuvant use of radiation (i.e., brachyther- 5. Ranjbar M, Kurz M, Holzhey A, Melchert C, Rades
D, Grisanti S. Stereotactic radiotherapy in neovas-
apy and stereotactic radiotherapy) in the treat- cular age-related macular degeneration. Medicine
ment of choroidal neovascularization due to (Baltimore). 2016 Dec;95(52):e5729.
neovascular AMD is to minimize the number 6. Casaroli-Marano R, Alforja S, Giralt J, Farah
of anti-VEGF injections, without reducing ME. Epimacular brachytherapy for wet AMD: current
perspectives. Clin Ophthalmol. 2014 Aug;8:1661.
their effectiveness. 7. Kishan AU, Modjtahedi BS, Morse LS, Lee
2. Epimacular brachytherapy requires the use of P. Radiation therapy for neovascular age-related
vitrectomy. The emission source is placed just macular degeneration. Int J Radiat Oncol. 2013
above the macular lesion for 4–5 min. Mar;85(3):583–97.
8. Johnson LK, Longenecker JP, Fajardo LF. Differential
3. Stereotactic radiotherapy is a less invasive radiation response of cultured endothelial cells
procedure, although more microvascular and smooth myocytes. Anal Quant Cytol. 1982
changes have been seen after its use. Sep;4(3):188–98.
4. The main studies investigating brachytherapy 9. Kirwan JF, Constable PH, Murdoch IE, Khaw
PT. Beta irradiation: new uses for an old treatment: a
are MERITAGE, CABERNET, and review. Eye. 2003 Mar;17(2):207–15.
MERLOT. While injection frequency was 10. Petrarca R, Jackson TL. Radiation therapy for neo-
reduced after epimacular brachytherapy in the vascular age-related macular degeneration. Clin
MERITAGE and CABERNET trials, the Ophthalmol. 2011 Jan;5:57.
11. Taddei PJ, Chell E, Hansen S, Gertner M, Newhauser
MERLOT study revealed contradictory results WD. Assessment of targeting accuracy of a low-
with less improvement using this modality. energy stereotactic radiosurgery treatment for age-
5. The main studies for stereotactic radiotherapy related macular degeneration. Phys Med Biol. 2010
are INTREPID, which showed favorable Dec;55(23):7037–54.
12. Journal B. Editorials Treatment of age-related sub-
results, and STAR, the most recent one that foveal neovascular membranes by teletherapy.
still ongoing. 1993;261–2.
6. More studies are needed, not only to prove 13. Bergink GJ, Hoyng CB, van der Maazen RW,
effectiveness but also to investigate long-term Vingerling JR, van Daal WA, Deutman AF. A ran-
domized controlled clinical trial on the efficacy of
safety. radiation therapy in the control of subfoveal choroidal
neovascularization in age-related macular degenera-
Acknowledgment To Estephania Tirado Torrero for tion: radiation versus observation. Graefes Arch Clin
graphical illustrations. Exp Ophthalmol. 1998 May;236(5):321–5.
14. A prospective, randomized, double-masked trial on 24. Jackson TL, Desai R, Simpson A, Neffendorf JE,
radiation therapy for neovascular age-related macu- Petrarca R, Smith K, et al. Epimacular brachytherapy
lar degeneration (RAD study). Ophthalmology. 1999 for previously treated neovascular age-related macu-
Dec;106(12):2239–47. lar degeneration (MERLOT). Ophthalmology. 2016
15. Sivagnanavel V, Evans JR, Ockrim Z, Chong Jun;123(6):1287–96.
V. Radiotherapy for neovascular age-related macular 25. Balaggan K, Schindler R, Joffe L, Patel P, Stea B,
degeneration. In: Sivagnanavel V, editor. Cochrane Adnan Tufail LM. Novel minimally-invasive epi-
database of systematic reviews. Chichester: Wiley; scleral brachytherapy for the treatment of neovascular
2004. age-related macular degeneration (nAMD): results
16. Stefansson E, Landers MB 3rd, Wolbarsht of a twelve month prospective phase i safety and
ML. Increased retinal oxygen supply following pan- tolerability evaluation. Invest Ophthalmol Vis Sci.
retinal photocoagulation and vitrectomy and lensec- 2013;54(15):3787.
tomy. Trans Am Ophthalmol Soc. 1981;79:307–34. 26. Holz FG, Engenhart R, Bellmann C, Debus J, Völcker
17. Nordsmark M, Overgaard M, Overgaard HE. Stereotactic radiation therapy for subfoveal cho-
J. Pretreatment oxygenation predicts radiation roidal neovascularization secondary to age-related
response in advanced squamous cell carcinoma of the macular degeneration. Front Radiat Ther Oncol.
head and neck. Radiother Oncol. 1996 Oct;41(1):31–9. 1997;30:238–46.
18. Avila MP, Farah ME, Santos A, Kapran Z, Duprat JP, 27. Evans JR, Sivagnanavel V. Chong V. Cochrane
Woodward BW, et al. Twelve-month safety and visual Database Syst Rev: Radiotherapy for neovascular
acuity results from a feasibility study of intraocular, age-related macular degeneration; 2010 May.
epiretinal radiation therapy for the treatment of subfo- 28. Lee C, Chell E, Gertner M, Hansen S, Howell
veal CNV secondary to AMD. Retina [Internet]. 2009 RW, Hanlon J, et al. Dosimetry characterization
Feb;29(2):157–69. Available from http://www.ncbi. of a multibeam radiotherapy treatment for age-
nlm.nih.gov/pubmed/19202425 related macular degeneration. Med Phys. 2008
19. Ávila MP, Farah ME, Santos A, Carla L, Fuji G, Rossi Oct;35(11):5151–60.
J, et al. Three-year safety and visual acuity results 29. Jackson TL, Chakravarthy U, Slakter JS, Muldrew
of epimacular 90strontium/90yttrium brachytherapy A, Shusterman EM, O’Shaughnessy D, et al.
with bevacizumab for the treatment of subfoveal cho- Stereotactic radiotherapy for neovascular age-
roidal neovascularization secondary to age-related related macular degeneration. Ophthalmology. 2015
macular degeneration. Retina. 2012 Jan;32(1):10–8. Jan;122(1):138–45.
20. Dugel PUU, Petrarca R, Bennett M, Barak A, 30. Jackson TL, Chakravarthy U, Kaiser PK, Slakter JS,
Weinberger D, Nau J, et al. Macular epiretinal brachy- Jan E, Bandello F, et al. Stereotactic radiotherapy
therapy in treated age-related macular degeneration: for neovascular age-related macular degeneration.
MERITAGE study: twelve-month safety and efficacy Ophthalmology. 2013 Sep;120(9):1893–900.
results. Ophthalmology [Internet]. 2012 Jul [cited 31. Jackson TL, Shusterman EM, Arnoldussen M,
2014 Nov 13];119(7):1425–31. Available from http:// Chell E, Wang K, Moshfeghi DM. Stereotactic
www.ncbi.nlm.nih.gov/pubmed/22465819 radiotherapy for wet age-related macu-
21. Chin H-S, Park T-S, Moon Y-S, Oh J-H. Difference lar degeneration (INTREPID). Retina. 2015
in clearance of intravitreal triamcinolone acetonide Feb;35(2):194–204.
between vitrectomized and nonvitrectomized eyes. 32. Dugel PU, Petrarca R, Bennett M, Barak A,
Retina. 2005;25(5):556–60. Weinberger D, Nau J, et al. Macular epiretinal brachy-
22. Petrarca R, Dugel PU, Bennett M, Barak A, therapy in treated age-related macular degeneration.
Weinberger D, Nau J, et al. Macular epiretinal brachy- Ophthalmology. 2012 Jul;119(7):1425–31.
therapy in treated age-related macular degeneration 33. Neffendorf JE, Desai R, Wang Y, Kelly J, Murphy C,
(meritage). Retina. 2014 May;34(5):874–9. Reeves BC, et al. StereoTactic radiotherapy for wet
23. Dugel PU, Bebchuk JD, Nau J, Reichel E, Singer M, age-related macular degeneration (STAR): study pro-
Barak A, et al. Epimacular brachytherapy for neovas- tocol for a randomised controlled clinical trial. Trials.
cular age-related macular degeneration: a random- 2016;17(1):560.
ized, controlled trial (CABERNET). Ophthalmology 34. Parsons JT, Bova FJ, Fitzgerald CR, Mendenhall WM,
[Internet]. 2013 Feb [cited 2014 Nov 13];120(2):317– Million RR. Radiation retinopathy after external-
27. Available from http://www.ncbi.nlm.nih.gov/ beam irradiation: analysis of time-dose factors. Int J
pubmed/23174399 Radiat Oncol. 1994 Nov;30(4):765–73.
and postgraduate courses, and he is currently doing research

about Diabetic Retinopathy at the university of Buenos
Aires. He won an award from the university of Buenos
Aires because his Project of Diagnosis of Diabetic
Retinopathy by ‘Telemedicine’” He has attended and deliv-
ered lectures in Buenos Aires and abroad. He have won the
“best Resident award” when he was in his 1st year of his
residency programme. This award was given by his col-
leagues. He has sat for every International Council of
Ophthalmology standardized test, that is why he has got an
Scholarship to did part of his Fellowship in Bern,
Switzerland Nowadays he teaches at the University of
Buenos Aires, performs surgeries and does research. He has
won several awards about his diabetic macular edema proj-
ects, the latest in May 2017 for the IRG project
“Dexamethasone implant for diabetic macular edema in
David Pérez González is a doctor from Mexico where he naïve compared to refractory cases: a retrospective 24
was raised and where he studied medicine at the Autonomous months study”, this award was given by the Pan-American
Popular University of the State of Puebla (UPAEP). At the Association of Ophthalmology at the Pan American
end of the medical school, he continued his pre-graduate Research Day in Baltimore, USA where ARVO 2017 took
medical internship at the National Institute of Medical place.
Sciences and Nutrition “Salvador Zubiran” (INCMNSZ) in
Mexico City and then continued another year as an active
member in the research area at his university. He completed
his ophthalmology residency program at the Twenty-First
Century National Medical Center in Mexico City, endorsed
and certified by the National Autonomous University of
Mexico (UNAM) for later being recognized as a certified
active member of the Mexican Council of Ophthalmology
and International Member in Training Candidate by the
American Academy of Ophthalmology. He is currently con-
ducting the Fellowship in Medical and Surgical Retina at Tel
Aviv Medical Center affiliated with the Sackler Faculty of
Medicine, Tel Aviv University.
Dinah Zur is a Retina specialist at the Tel Aviv Medical

Center and a Senior Lecturer at the Sackler Faculty of
Medicine, Tel Aviv University, Tel Aviv, Israel. She was
born and raised in Berlin, Germany, where she studied
medicine at the Charite University Hospital. After her
medical studies, she moved to Israel and completed resi-
dency in the Ophthalmology Division at the Tel Aviv
Medical Center in Tel Aviv, and fellowship in Medical
Retina at the Sankt Gertrauden Hospital, Charite, Berlin,
Germany. Following this, she returned to Israel and since
2005 she is part of the Faculty at the Department of
Matias Iglicki is a certified Teacher of Ophthalmology at Ophthalmology at the Tel Aviv Medical Center.
the University of Buenos Aires. He was born in Buenos She published more than 40 original papers; she is
Aires, he went to Medical School at the university of the author of reviews and book chapters and a reviewer
Buenos Aires and he did his residency at the Hospital de in high impact journals. Her main field of interest is in
Clinicas which depends on the university of Buenos Aires retinal and choroidal imaging, retinal vascular dis-
He has obtained a Degree Ophthalmology Specialist eases, and hereditary retinal diseases with focus on the
Physician Higher Course at the university of Buenos Aires identification of biomarkers in these conditions.
His current activities involve, research, clinical as well as She is one of the initiators of the International Retina
surgical practices, He went to the to conservartory of music Group, which is an independent research group.
where he lerned how to play the violin, he loves playing the Moreover, she is coordinating the Continuing Medical
violin and listening to classical music. He teaches graduate Education in Ophthalmology at the Tel Aviv University.
Laser for Prevention of Choroidal
Jeffrey K. Luttrull and David Kent
30.1 Introduction pointing as the treatment of CNV itself. Recent

studies indicate a change in this regard [8–11].
For most of the modern history of ophthalmol- Spurred by a new understanding of retinal laser
ogy, laser treatment has been the primary treat- treatment, new information from both the labora-
ment for macular disease [1]. Until the relatively tory and clinic suggests that laser for prevention
recent advent of vascular endothelial growth fac- of CNV should be effective, and is [11]. To
tor (VEGF) inhibitors, ablative photocoagulation understand the past failure, and future promise,
was the only treatment option for most eyes with of laser treatment to prevent CNV it is essential
choroidal neovascularization (CNV) [1–4]. Due to understand exactly what is meant by “laser”
to the risks of treatment-associated visual loss, treatment, and all that follows from it [8]. Our
location, and often poor definition of macular discussion of laser CNV prevention will focus on
neovascular lesions, few patients with macular CNV complicating age-related macular degener-
CNV were candidates for treatment. Of those, ation (AMD). This is for two reasons: first,
fewer still enjoyed any long-term treatment ben- because AMD is the most important cause of
efits [5]. visual loss due to CNV; and second, because only
While anti-VEGF medications have revolu- in AMD have notable efforts been made to pre-
tionized the treatment of CNV, both due to their vent CNV by any means, and in particular, with
effectiveness and minimal risks of treatment- laser [12, 13].
associated visual loss, other than Age-Related For over 50 years, photocoagulation was uni-
Eye Disease Study (AREDS) vitamin supple- versally presumed to be the necessary and suffi-
ments, there remains no effective treatment, cient cause of all therapeutic benefits of retina
pharmacologic or otherwise, to prevent macular laser treatment [14–16]. Low-intensity/high-
CNV [6, 7]. The history of macular laser treat- density subthreshold diode micropulse laser
ment to prevent CNV is as long and as disap- (SDM) proved this presumption to be false by
demonstrating therapeutically effective treatment
in the total absence of laser-induced retinal dam-
age (LIRD) [17–21]. As a watershed develop-
J. K. Luttrull (*) ment in the history of retinal laser treatment, this
Ventura County Retina Vitreous Medical Group, discovery led to two key advances in our under-
Ventura, California, USA standing of retinal laser treatment. These bear
e-mail: office@venturacountyretina.com
heavily on the following examination of the
D. Kent potential for the laser to prevent CNV. First, as
The Vision Clinic, Kilkenny, Ireland

402 J. K. Luttrull and D. Kent
both unnecessary and the sole source of all risks, cally useful indicator of the degree of age-related
adverse treatment effects, and treatment macular dysfunction. The risk of age-related
limitations, we now understand photocoagulation visual loss generally parallels the number, size,
(for other than cautery) to be a complication of and proximity of drusen to the fovea [25–30].
retinal laser treatment. Second, abandonment of Thus, drusen reduction has been a natural target
photocoagulation as the necessary precondition for therapeutic intervention in dry AMD [31, 32].
for therapy has led to, for the first time, a satisfy- Inflammatory disturbances of the macula in
ing and useful understanding of the mechanism eyes with drusen, such as the development of
of action of retinal laser treatment as a physio- CNV, other exudations, and focal chorioretinitis,
logic “reset” phenomenon [22, 23]. Unlike prior including the iatrogenic chorioretinitis of macu-
unsuccessful attempts to explain the effects of lar photocoagulation, have long been noted to
retinal laser arising from photocoagulation, reset result in local disappearance of drusen in and
theory accounts for all clinically observed retinal around the inflammatory lesion [27, 30, 33, 34].
laser effects. Further, in a fundamental test of any This suggested macular photocoagulation might
theory, reset theory has accurately predicted a be used therapeutically to reduce drusen and the
new retinal laser application never conceived of risks of visual loss due to AMD [33, 35].
in the photocoagulation era. These include rever- At the time of the first attempts to use photo-
sal of anti-VEGF drug tolerance in neovascular coagulation to reduce drusen, the standard for
AMD (NAMD); improved retinal and visual photocoagulation intensity was established by
function following panmacular laser treatment in the Early Treatment of Diabetic Retinopathy
dry AMD, inherited retinopathies, and open- Study (ETDRS) and Macular Photocoagulation
angle glaucoma; and neuroprotective effects in Study Group (MPSG) reports, which employed
glaucomatous optic neuropathy [22–25]. intense suprathreshold treatment resulting in
Critical to a clear understanding of the retinal clinically obvious white, full-thickness retinal
laser literature is a clear understanding of various burns [36, 37]. Recognition that drusen occurred
retinal laser modes and their effects [8]. at the level of the RPE suggested that less
Complicating this task is historical imprecision intense and thus less obviously clinically visible
and frequent misuse, unintentional and other- photocoagulation lesions (limited to the RPE
wise, of key clinical terminology that can obfus- and outer retina, or “threshold” intensity),
cate and mislead readers not keenly aware of barely visible, or even initially non-ophthalmo-
these issues. The history of laser prevention of scopically “subthreshold” photocoagulation
NAMD is particularly illustrative in this regard. lesions, might be effective and reduce adverse
Essential differences in the conception, design, treatment effects by reducing neurosensory reti-
expectations, claimed and actual effects of vari- nal damage [32]. These less-severe lower inten-
ous laser modes account for both the past failures sity retinal burns were described, in comparison
and future promise of retinal laser for the preven- to the starkly white ETDRS and MPSG lesions,
tion of CNV. To draw out the key distinctions, the as “invisible” [38]. However, photocoagulation
following discussion will divide the topic into lesions acutely invisible or difficult to see oph-
two main headings: Laser for drusen; and laser thalmoscopically are virtually always immedi-
not for drusen. ately visible by fundus fluorescein angiography
(FFA) and become clinically visible minutes to
weeks after treatment. The contrast in lesion
30.2 Laser for Drusen severity with conventional ETDRS and MPSG
suprathreshold photocoagulation, however, led
As the hallmark of dry AMD, drusen are an investigators to incorrectly assume a difference
important risk factor for progression and visual in substance, rather than style. This was (1)
loss, particularly due to the development of CNV because LIRD was assumed to be essential to
and NAMD [25–30]. As such, drusen are a clini- the therapeutic benefits of retinal laser treatment
30 Laser for Prevention of Choroidal Neovascularization 403
in general, and drusen reduction in particular nosomes. In this volatile, literally explosive set-
could not be done without; and (2) over-estima- ting, excess laser energy or RPE pigment density
tion of the risk and adverse effect reduction will necessarily result in collateral damage
associated with lower intensity retinal photoco- extending to the retina and Bruch’s membrane
agulation [31, 32, 35, 38]. [8]. Again, next to the severity of the suprathresh-
The initial reports of macular laser treatment old full-thickness retinal burns from conventional
to reduce drusen were positive and encouraging standard photocoagulation in the ETDRS and
[31, 32]. Treatment, directed at the drusen them- MPSG reports, however, the LIRD resulting from
selves, was effective in reducing drusen numbers short-pulse CW lasers, generally confined to the
and density, and improving visual acuity [31, 32, RPE and outer retina, still considered essential,
35, 38]. With time, however, treated eyes began seemed also negligible in comparison. By confin-
to demonstrate an increased propensity to develop ing LIRD to the RPE and outer retina, it was
new CNV thus converting to NAMD; worsening, hoped that drusen could be reduced without
rather than improving, the long-term visual prog- incurring either neurosensory retinal damage or
nosis [39–41]. Subsequent studies found that increasing the risk of CNV, led to studies of both
eyes demonstrating both the greatest drusen microsecond and nanosecond lasers, more selec-
reduction and the greatest likelihood of CNV tive for the RPE, as preventative treatments for
were those receiving the most intense macular AMD, and thus age-related CNV.
photocoagulation [42, 43]. This realization In 1999, Roider and coworkers reported two
brought about the effective end of interest in con- patients with soft drusen treated with
ventional (millisecond) continuous wave (CW) 1.7-microsecond exposures of 527 nm laser with
laser treatment for drusen [44]. identical treatment parameters (“selective retinal
It was recognized that LIRD to retinal archi- laser therapy”, or SRT, Lutronic, Billerica MA,
tecture, particularly the Bruch’s membrane/RPE USA), both clinically subthreshold, without visi-
complex, was a likely key predisposing factor to ble laser spots at the time of treatment [51]. Both
CNV, but possibly not necessary to achieve effec- produced LIRD visible by FFA. However, in one
tive drusen reduction [9]. What to do? By short- eye the LIRD became visible clinically following
ening laser pulse duration the thermal effects of treatment. In this eye, the drusen disappeared. In
retinal laser treatment could be more selectively the other patient, no clinically visible LIRD
concentrated in the RPE [45, 46]. Microsecond developed later and there was no drusen reduc-
CW exposures demonstrated the ability to selection. The authors duly note that these eyes dem-
tively damage the outer retina and the homoge- onstrate the clinical variability of short pulse
neously pigmented RPE of animals in laboratory laser effects, a clinical manifestation of the nar-
studies [47–50]. Further shortening of the laser row CW laser TR.
pulse to the nanosecond range can result in selec- As noted above, NSL takes the quest for selec-
tive heating of the RPE melanosomes themselves tive RPE destruction a step further. By shortening
[50]. However, shortening of pulse duration the pulse further, NSL maximizes the thermal
results in narrowing of the therapeutic range energy uptake to the RPE melanosomes them-
(TR), such that at nanosecond exposures, there is selves, causing explosive vaporization of the
no TR. Instead, the threshold for RPE cell death melanosomes with internal cavitation of the RPE
is exceeded before the threshold for HSP activa- causing cell death (2RT®; Ellex Pty Ltd., Adelaide,
tion (the cell dying before the HSPs can activate), Australia). In its idealized form, 2RT NSL seeks
the prime mediator of therapeutic retinal laser to limit LIRD to selective killing of the RPE,
effects [8, 49]. In practice, this means that the which they term “rejuvenation,” without causing
effect of nanosecond laser (NSL) exposure is damage to adjacent structures such as photorecep-
either no effect at all (below the TR threshold); tors and Bruch’s membrane. Pilot studies of NSL
or—at minimum—photodisruptive killing of the demonstrated both drusen reduction and improve-
RPE by intracellular vaporization of RPE mela- ment in visual acuity without a notable increase in
CNV [9, 10]. At 24 months follow-up in a ran- high-risk eyes. As with other studies of laser for
domized clinical trial of c omparing NSL to sham drusen reduction, the LEAD study challenges the
for dry AMD, the LEAD study found no overall presumption that drusen reduction is a desirable
effect of treatment [52]. Subgroup analyses end. Rather, the acceleration of visual loss due to
showed drusen reduction in eyes with early, low- age-related geographic atrophy (ARGA) in the
risk AMD, without an increase in CNV incidence, LEAD study suggests 2RT appears to cause rapid
suggesting the success of the trial hypothesis. decompensation of the most vulnerable eyes with
However, the 2RT NSL treatment response in high-risk AMD. This may be due to added physi-
eyes with high-risk AMD, those with the greatest ologic stresses associated with the inflammation
risk of visual loss and with thus the most to gain and healing response to the LIRD caused by NSL
from effective preventive treatment, was negative in these already tenuous eyes. Transmacular
[52]. In these eyes, often identified by the pres- shock-waves produced by explosive photodisrup-
ence of reticular pseudodrusen (RPD), NSL tion of the RPE, unique to NSL, may also con-
caused rapid progression of AMD and visual loss, tribute to RPE decompensation as far as 2 mm
particularly due to the development and/or pro- from an NLS application site (Chang DB, Luttrull
gression of geographic atrophy. Of note is that JK; unpublished data, March 2019). The worsen-
LIRD was not considered an adverse treatment ing of AMD, and particularly the progression of
effect in the LEAD study. This is despite the fact ARGA following laser for drusen in the LEAD
that LEAD trial was performed to assess the effect study is not entirely unexpected. In the Cochrane
of “non-damaging” 2RT NSL in AMD, a stated meta-analysis of laser for drusen studies, Virgili
differentiator of the LEAD study from prior stud- and associates noted that, while little information
ies of laser for drusen (https://www.ellex.com/us/ could be gleaned from prior studies in this regard,
products/2rt/). Unfortunately, no fundus images available data suggested that ARGA worsened
accompanied the published report [52]. However, after laser for drusen [44]. In addition, a prospec-
in a meeting presentation of LEAD study results tive clinical trial of microsecond SRT laser hop-
by the investigators, all posttreatment fundus ing to slow ARGA progression was abandoned
photographs presented demonstrated laser dam- early, due to the finding that treatment, designed
age due to the NSL applications, clinically indis- to produce LIRD at the margins of the ARGA
tinguishable from conventional CW lesions, increased the rate of ARGA progression
photocoagulation lesions (Guymer R, Marshall J, by 50% compared to untreated controls [53].
et al. The LEAD Study, presentation to the In sum, efforts to reduce the risks of visual
European Society of Retina specialists, Sept. 22, loss in AMD with laser treatment have been
2018, Vienna, Austria). The LEAD study authors driven by the recognized necessity of LIRD to
did not explicitly correlate worsening of high- achieve drusen reduction and the presumption of
risk AMD with LIRD, but concluded that 2RT a resulting benefit. The failure of these efforts
NSL was contraindicated in eyes with high-risk results from this very same LIRD, and casts
AMD [52]. doubt on the presumption of benefits from drusen
Despite disappointing results, the LEAD reduction.
study findings offer important insights. While it
supports the concept that LIRD is necessary for
drusen reduction (without establishing a benefit) 30.3 Laser Not for Drusen
it also suggests that avoidance or minimization of
damage to Bruch’s membrane may reduce the The only treatment as yet proven to reduce the
risk of CNV as an adverse treatment effect. risk of developing age-related CNV in a multi-
However, it also demonstrates that that NSL is center prospective clinical trial is AREDS vita-
unreliable for this purpose, resulting in signifi- min supplementation [6]. Interestingly,
cant collateral damage in many, if not most, challenging the presumption of the need for dru-
treated eyes leading to significant worsening of sen reduction, the success of AREDS treatment
was not associated with any notable treatment vascular arcades including the fovea; and “pan-
effects on drusen or other morphologic features retinal” treatment, consisting of confluent and
of AMD. The treatment benefits were consistent complete treatment of all the retina outside the
over time, reducing progression to advanced major vascular arcades [11, 20, 22–25, 46, 56,
AMD by about 4% per year over the course of the 57]. Panmacular treatment is used for all treat-
study. Unlike short pulse lasers, the greatest ben- ment indications. Panretinal treatment is added
efit from AREDS supplements was seen in eyes for generalized retinopathies, such as diabetic
at the greatest risk for visual loss [6]. As 90% of retinopathy or early retinitis pigmentosa, result-
eyes with advanced AMD have CNV, it was a ing in treatment of the entire retina, and thus all
reduction in new CNV that accounted for most of of the retinopathy, just as would be treated by
the AREDS treatment benefits as the progression medication (Figs. 30.1 and 30.4).
of ARGA was not affected. Thus, it was nutri- Initial studies of SDM demonstrated effective
tionally derived improvement in retinal physiol- treatment for complications of diabetic retinopa-
ogy alone that accounted for the benefits of thy in the complete absence of LIRD [18–20, 46].
AREDS study, and it did so, via CNV reduction, Like AREDS supplements, the absence of LIRD
without reducing drusen or effecting any other with SDM indicates, by exclusion, that SDM
alteration in macular anatomy [6]. treatment benefits arise entirely from laser-
All laser modes are capable of causing induced improvements in retinal physiology elic-
LIRD. Some, like NSL, cause it necessarily. For ited by sublethal thermal laser stimulation of the
others, such as CW lasers, LIRD is prohibitively RPE [58]. Whereas the TR of CW laser is only
difficult to avoid and still accomplish effective 10× ANSI MPE, the same ANSI data show the
treatment. Only a micropulsed laser (MPL) oper- TR of micropulse lasers, like SDM, to be much
ated at a low duty-cycle (DC) can reliably and wider, 100×MPE or more [47]. SDM is thus safe
predictably preclude LIRD while maintaining and without any known adverse treatment effects
therapeutic effectiveness [9]. The epitome of [45, 47, 58]. Because of the broad SDM TR,
low-DC MPL reliably sublethal to the RPE is treatment intensity need not be subjectively
“low-intensity/high-density subthreshold diode titrated on a per eye basis (effectively turning
micropulse laser” (SDM). SDM both established each treated eye into a dosimetry experiment), as
and defined the fundamental principles of mod- decades of clinical experience have identified
ern retinal laser therapy. First; by employing a safe and effective SDM parameters that can be
low-DC (5%) 810 nm near-infrared laser, SDM is used effectively in all eyes of all patients without
reliably sublethal to the RPE (“low-intensity”), regard for pathology or pigment variation [8, 18–
which it selectively targets. Despite this safety, it 20, 45, 47, 57, 58]. Employment of “fixed” SDM
is an effective activator of RPE HSPs, the essen- laser parameters enhances treatment safety by
tial mediators of the therapeutic laser response. eliminating the possibility for surgeon misjudg-
Second; the therapeutic response is amplified, ment, the most common cause of visual loss from
maximized, and made clinically effective by con- modern retinal laser treatment [8].
fluent and complete treatment aimed at recruiting As noted, SDM led to a new understanding of
large areas of diseased retina (“high-density”) to the therapeutic benefits of retinal laser treatment
the normalizing process [9, 45, 46, 49, 50, 54, stemming from a physiologic “reset” effect on
55]. This is in contrast to the traditional focal or the RPE, normalizing retinal function largely
local application strategy used in conventional independent of the underlying cause of dysfunc-
photocoagulation, which are abandoned in mod- tion [23–25]. Making no unique contribution to
ern retinal laser therapy [37, 38]. Over time, the treatment efficacy and solely responsible for all
SDM treatment paradigm has evolved to the risks, limitations, and adverse effects of retinal
employment of just two treatment fields: “pan- laser treatment, photocoagulation, and all other
macular” treatment, consisting of confluent treat- forms of LIRD are now considered necessary
ment of the entire retina between the major complications of treatment [8, 21–24, 59].
Fig. 30.1 Intravenous fundus fluorescein angiograms medical treatment for DR was used. Left photo, before
(FFA) of diabetic retinopathy (DR) following total retinal treatment; right photo after treatment. Note reduction in
SDM laser (panmacular and panretinal) in three patients macro and microvascular leakage along with reversal/
(a), (b) and (c). No VEGF inhibitors, steroids, or other diminution of retinopathy severity in each case
Reset theory indicated that SDM should 93–98% lower than expected compared to the
improve retinal function not just in the conven- AREDS and other natural history studies of AMD
tional retinal laser indications such as diabetic [6, 11]. In a subsequent study of 111 eyes of 70
macular edema, proliferative diabetic retinopa- patients with ARGA from the same dry AMD
thy, central serous chorioretinopathy, and retina cohort, SDM reduced the radial velocity of
vein occlusions; but in any chronic progressive ARGA progression 55% per year compared to
retinopathy (CPR) of any cause [23–25]. In short, untreated controls (p = 0.0002) (pretreatment
this is because all CRPs are neurodegenerations, observation period avg. 2.7 years, posttreatment
and thus share many key features that are avg. 1.9 years) [56]. Echoing the AREDS, other
addressed by the reset effects of laser treatment than reducing the incidence of new CNV and
[23, 24, 54, 55]. By addressing those fundamen- slowing ARGA progression, panmacular SDM
tal commonalities, the reset phenomenon acts as prophylaxis of dry AMD was not associated with
a “non-specific trigger of disease-specific repair,” any notable short-term effects on drusen or other
improving both retinal and visual function in all macular morphology [11, 56]. The findings of
CPRs, including AMD [23, 24]. these studies will inform the design of future con-
In a pilot study of a large cohort of 547 eyes of firmatory studies. However, the robust results of
363 patients with dry AMD, panmacular SDM these pilot studies of panmacular SDM for dry
treatment was followed by a much lower than AMD indicate that retinal laser treatment
expected rate of new CNV [11]. In the absence of designed to preclude, rather than cause, LIRD
LIRD or any other adverse treatment effects, the may become an important tool in the prevention
annual incidence of new CNV following pan- of visual loss due to AMD and age-related CNV
macular SDM in dry AMD was 0.87%, compared in particular [11].
to 4% in the AREDS [6]. This 80% reduction in In sum, while retina-damaging laser increases
expected CNV was noted despite significantly the risk of CNV, non-damaging laser sublethal to
higher risk factors for CNV in the panmacular RPE reduces that risk. While damaging laser
SDM-treated group compared to the AREDS, appears to cause rapid progression of high-risk
particularly age (median 84 vs. 69 years) [6]. AMD; non-damaging laser, sublethal to the RPE,
Adjusting only for the risk factor age, SDM appears to slow progression significantly. Thus,
treated eyes had an incidence of new CNV contrary to traditional thinking (but as predicted
a b
c d
Fig. 30.2 Pseudophakic cystoid macular edema (CME). coherence tomography (OCT) prior to treatment with
Unresponsive to combination of topical steroid and non- severe CME. Visual acuity (VA) 20/200. (d) IR and (e)
steroidal anti-inflammatory drops for 1 year. Drops dis- OCT 1 month following panmacular SDM. CME
continued and panmacular SDM performed. (a) Infrared resolved. VA 20/30
(IR), (b) late phase FFA and (c) Spectral-domain optical
by reset theory) it is successful avoidance of LIRD, retinal integrity and RPE function; third, by recog-
and thus avoidance of drusen reduction, that nizing how the classical dynamics of wound repair
appears key to the success or failure of laser for are engaged by treatment; and fourth, by under-
prevention of CNV in AMD. How do we account standing the ability of laser to modulate both acute
for this? First, by understanding AMD as a chronic and chronic inflammation—one the healer, the
progressive neurodegenerative disease; second, by other the driver, of the chronic disease process
taking into account the effects of retinal laser on (Figs. 30.1, 30.2, 30.3, 30.4, 30.5).
Fig. 30.3 Sarcoid

uveitis with cystoid
macular edema. History
of severe steroid
response in the fellow
previously.
Unresponsive to
bevacizumab. Other
VEGF inhibitors refused
by insurance. Top: Prior
to panmacular SDM
treatment VA 20/30.
Bottom, 1 month
following panmacular
SDM. VA 20/25
Fig. 30.4 Idiopathic retinal vasculitis, optic neuritis, and medical treatment given. Note the decrease in inflamma-
vitriitis. (a) FFA before treatment. (b) FFA 1 month fol- tory dye leakage from retinal vessels and optic nerve. VA
lowing total retinal SDM laser. No other local or systemic prior to treatment 20/70; after, 20/50
Fig. 30.5
Bevacizumab- and
ranibizumab-resistant
serous macular
detachment associated
with diabetic macular
edema. No steroid
therapy was given. Top:
Before panmacular
SDM. VA 20/200.
Bottom: Three months
after panmacular
SDM. Note the decrease
in macular edema and
resolution of
inflammatory serous
macular detachment. VA
20/70
30.4 ry AMD as a Model

D ally all chronic, age-related diseases, including
of Chronic Disease AMD [60]. Unlike acute inflammation,
and Inflammaging inflammaging is low grade, chronic, persistent,
and self-perpetuating, and leads to tissue degen-
The association of aging with chronic inflamma- eration [61]. To understand the mechanisms by
tion, often referred to as “inflammaging,” is now which inflammaging is generated, we must first
recognized as a significant component of virtu- understand the essential role of the immune sys-
tem in the maintenance of normal tissue function A classic example of a disease with self-debris
and homeostasis. is AMD. In dry AMD, this debris is the pheno-
In healthy cells, there is constant surveillance typic marker of disease and a clinically useful
and repair to maintain normal cell function and biomarker of the heightened inflammatory state
homeostasis [49, 62, 63]. However, in disease within the retina, classically represented by dru-
intracellular abnormalities often either escape sen. Ultrastructurally, this is reflected by priming
repair and/or exceed the cell’s ability to manage or activation of the NLRP3 inflammasome and
them successfully leading to a loss of normal the specific proteins known to be associated with
function that is generally characteristic of the pri- inflammasome assembly [65]. Experimentally,
mary underlying disease process. Further, both identified proteins include C1q complement
normal and diseased tissues produce waste, or component in drusen extracts from AMD donor
“self-debris,” that includes damaged cells and tissue, carboxyethylpyrrole-adducted proteins
macromolecules. In disease, the accumulation of from the aging retina, and Alu RNA transcripts
this waste is excessive and thus progressive, ulti- from the RPE of patients with ARGA [67–71].
mately compromising tissue structure and func- This inflammasome-mediated damage to the ret-
tion [60]. At the tissue level, the mechanism ina in AMD may be further aggravated by inflam-
employed to repair this damage and remove this matory pyroptotic and apoptotic effects resulting
waste is inflammatory-mediated (inflammation from IL-1β and IL-18 expression [70]. In those
being a prerequisite of repair) and is dependent on individuals who have the genetic risk variant for
resident macrophages and mast cells [64]. With complement factor H (CFH) or other comple-
aging this “housekeeping” function becomes less ment risk variants, regulation of these factors is
efficient, due mainly to the combination of abnormal and excessive quantities of inflamma-
increased generation of self-debris and inefficient tory complement components are generated and
removal, requiring additional inflammatory input deposited in the retina. Thus, the chronic inflam-
to maintain the tissue in a physiologic (normal) or matory state we recognize as AMD is driven by a
near physiologic working state, a process that is number of factors, including complement dys-
mediated but ultimately compromised by assem- function and inflammasome-mediated inflamma-
bly of the “inflammasome.” The inflammasome is tion, with genetic and environmental (such as
a multiprotein stimulus-dependent oligomer that smoking) modifiers [72–85].
activates the inflammatory process by promoting
secretion of pro-inflammatory cytokines and
interleukins with dysregulation of inflamma- 30.5 Acute Inflammation—
somes being a feature of all chronic diseases [65]. Prelude to Repair
This heightened inflammatory state, between
basal physiologic inflammation and pathologic It is a basic tenet of biologic repair that an acute
inflammation, is referred to as “para-inflamma- inflammatory response is necessary to activate
tion” [64]. With further aging the inflammatory the correct cascade of molecular events to gener-
stakes continue to rise, eventually escalating to ate successful and complete repair [81, 82]. In
require mobilization of a systemic immune contrast, the molecular signature of all acquired
response that includes the recruitment of addi- age-related diseases, such as AMD, is the pres-
tional leukocytes and expression of systemic pro- ence of chronic inflammation. Chronic inflam-
inflammatory cytokines [66]. Thus, maintenance mation, by definition, indicates that normal repair
of tissue homeostasis in the aging human requires is not proceeding [61, 65, 81, 82, 86, 87].
an increasing inflammatory response to address Unimpeded, this can lead to tissue degeneration
increased reparative demands that eventually and vision loss via the end-stage AMD pheno-
moves beyond para-inflammation to a self-perpet- types we recognize clinically as ARGA and dis-
uating and degenerative chronic inflammatory ciform scarring due to CNV. Thus, a desired goal
state referred to as “inflammaging” [60, 66]. of any therapy would be to intervene preventa-
tively, prior to anatomic derangement and visual that will result in activation of a classical tissue
loss (end-stage disease markers), when the dis- repair response [89]. As an extreme example of
ease process is manifest only by the earliest sorts, it also offers an important insight into the
detectable physiologic dysfunction and most prior failures of laser treatment associated with
amenable to repair [23]. If the chronic inflamma- LIRD to prevent CNV.
tory disease bed that is AMD could be converted Classical tissue repair involves a sequence of
to an acute inflammatory lesion, would the result- key interactions that can be divided into the three
ing corrective molecular cascades allow for heal- overlapping phases of (1) inflammation, (2) tis-
ing and repair sufficient to permit restoration and sue formation, and (3) tissue remodeling; involv-
normalization of retinal function? ing cells, cytokines, and the extracellular matrix
(ECM) [82, 90–92].
Ultimately, the degree of tissue remodeling
30.6 Laser—Prelude to Acute and resultant scar tissue deposition is directly
Inflammation proportional to the severity of the initial injury:
the greater the injury, the greater the degree of
As discussed previously, clinical retinal laser acute inflammation induced and the greater the
effects can be generally divided into two types: degree of irreversible tissue damage and scar
damaging (lethal to at least the RPE) and non- tissue formation. Scar tissue formation repre-
damaging (sublethal to the RPE). These divisions sents the restoration of tissue integrity, but
reflect progression in our understanding of the without restoration of normal tissue function.
mechanism of retinal laser treatment. As dis- Thus, we can ultimately say that, based on the
cussed, while photocoagulation was once the goal canons of wound healing, a therapeutically
of retinal laser treatment, it is now clear that pho- ideal laser treatment causes acute injury with-
tocoagulation and indeed all forms and degrees of out structural damage. This in turn generates
LIRD are unnecessary and detrimental; complica- the desirable endpoint of complete repair with-
tions of treatment rather than the necessary and out scar tissue formation, thus restoration and
sufficient cause of therapeutic retinal laser effects. normalization of tissue function. LIRD violates
Within the context of LIRD there is a spectrum of this maxim.
damage that is also instructive. An extreme exam- This new understanding holds that any degree
ple of LIRD is the inadvertent rupture of Bruch’s of LIRD lethal to the target RPE is not only unde-
membrane. This can be seen with excessively sirable but a complication counterproductive and
intense photocoagulation, or even as the result of detrimental to the goal of therapeutically effec-
micro- or nanosecond laser damage limited to the tive retinal laser treatment. “Non-damaging”
outer retina, such as that reported in the LEAD laser can thus only properly refer to the complete
study [52]. Clinically acute rupture of Bruch’s absence of LIRD at the histopathologic level at
membrane is heralded by the instantly recogniz- any point postoperatively, as the RPE is affected
able audible “pop” and subretinal gas-bubble for- but not killed by laser exposure. Functional retina
mation, with the simultaneous localized is lost, RPE cells, the mediators of laser response,
appearance of subretinal hemorrhage. Clinically are destroyed rather than revitalized and the ther-
undesirable as promoting the development of apeutic response thus diminished. Finally, as the
CNV, this observation has led to the use of laser previously discussed clinical studies illustrate,
rupture of Bruch’s membrane to experimentally violation of tissue integrity, no matter how small,
promote CNV animal models [88]. can “light the fuse” for the eventual development
By any definition, whether clinically inadver- of CNV by promoting angiogenesis and/or com-
tent or experimentally deliberate, laser-induced promising the natural barrier to vascularization
rupture of Bruch’s membrane is a good example presented by the healthy and intact RPE/Bruch’s
of an acute, brief, and once-off traumatic event membrane complex [39–42, 44].
30.7 Laser: Mechanism of Action and further accumulation of damage, debris,

chronic inflammation, cell death, and tissue fail-
As stated, laser efficacy derives from its ability to ure [49, 99–103].
generate a sublethal injury to the RPE leading to As noted above, thermal laser exposure is an
repair [8]. At the cellular level, this results in nor- effective trigger of HSP activation [49, 97, 104].
malization of RPE function, or “homeotrophy.” The attributes of SDM which are ideal in this
By improving RPE function, retinal function and regard arise from the facility of SDM to control-
autoregulation are improved via normalized lably photocoagulate only a small fraction of
expression of, and response to, RPE-derived fac- intracellular proteins sufficient to activate HSP-
tors such as cytokines and interleukins. mediated cellular repair, but insufficient to cause
Photoreceptor toiletry and waste processing are cell death [8, 105]. Most importantly, this laser-
improved and thus debris accumulation dimin- induced HSP activation response (salvific and
ished [26, 29, 30, 33–35, 93–96]. At the tissue homeotrophic, rather than baseline and homeo-
level, this leads to improved retinal and visual static) results in the repair of not only the acute
function via activation of reparative laser-induced laser-induced damage but also proceeds to indis-
acute inflammation which is inherently antago- criminate repair of the accumulated damage from
nistic to disease-driving chronic inflammation. the underlying chronic disease process that has
Therefore, for the laser to be therapeutically escaped HSP surveillance and would otherwise
effective, it must cause an acute, brief, and once- lead to progressive cellular dysfunction and ulti-
off “injury” to the RPE to activate tissue repair mate death. Independence of the salvific HSP
and to be maximally beneficial, the laser must be repair process from the cause(s) of the accumu-
“tissue-sparing” and thus sublethal to the lated cell damage and resultant dysfunction par-
RPE. MPL such as panmacular SDM stands ticular to the underlying disease process makes
alone in its ability to predictably and reliably ful- the reset phenomenon both powerful and elegant,
fil both criteria [8]. So, in the absence of LIRD, and the basis for the description of retinal laser in
how does laser work to elicit retinal repair and chronic retinopathies as a “non-specific trigger of
functional restoration? disease-specific repair” [49, 100–103].
Heat shock proteins (HSPs) are a family of
proteins that are constitutively expressed in all
cells, having critical roles in maintaining homeo- 30.8 Laser and AMD-Biologic
stasis and normal cell function, but, as noted Effects
above, are also significantly upregulated in
response to acute insults, such as heat and oxida- Laser activation of RPE HSPs triggers a cascade
tive stress, perceived by the cell to be existential of reparative and modulated inflammatory
threats [97]. HSPs are typically grouped into dif- effects, factors, and processes that improve cell
ferent subfamilies according to their molecular and retinal function. These include decreased
weights in kilo Daltons (hsp100, hsp90, hsp70, expression of angiogenesis promoters VEGF,
hsp60, hsp40, and small HSPs (sHSP), which TGF-β, and bFGF; increased angiogenesis
includes α-crystallins). Each has a particular role. inhibitor pigment epithelial-derived factor
The hsp 70 family is of particular relevance to the expression; improved mitochondrial function;
current discussion [98]. HSP activation is espe- increased retinal nitrous oxide levels; inhibition
cially sensitive to the acuity and severity of a cel- of apoptosis; modulation of tissue matrix metal-
lular threat. Thus, the insidiously progressive loproteinases; reduced free radical species and
dysfunctions of chronic disease are poor triggers increased superoxide dismutase activity;
of HSP-mediated repair [18, 22, 23, 49, 57, 58, increased mRNA expression of cytokine mark-
62, 63]. In chronic disease, the homeostatic func- ers of reparative acute inflammation and
tions of HSPs may become taxed to the point of decreased markers of chronic inflammation;
failure, leading to failure of the HSP system itself local and systemic immunomodulation includ-
ing local stem cell activation and monocyte and tor that plays a critical role in diverse cellular
hematopoietic progenitor cell recruitment to the processes associated with proliferation, cell
retina; and improved retinal macro- and microg- death, and development; as well as innate and
lial function [8, 22, 23, 47, 58, 63, 106–116]. adaptive immune responses, is normally seques-
Absent compromise of the RPE/Bruch’s mem- tered in the cytoplasm by a family of inhibitory
brane complex integrity and tissue scarring from proteins known as inhibitors of
LIRD that increase the risk of CNV in AMD, NF-kappaB. Activation of NF-kappaB leads to
any or all of the above responses to laser treat- the assembly of the NLRP3 inflammasome and
ment sublethal to the RPE such as SDM may activation of a tissue repair response [119].
contribute to a reduced risk of age-related CNV. Clinically, we recognize this stage of the
As noted, a characteristic abnormality in immune response to aging as the typical
AMD is the accumulation of metabolic by- inflammaging- associated AMD phenotype of
products such as lipofuscin, damaged organelles, soft drusen and pigmentary alterations at the
and nonfunctioning or toxic proteins [117]. This macula. In other words, these clinical character-
informs us that the RPE in AMD is functioning istics are key biomarkers of what is happening
in an environment of chronic oxidative stress, at the cellular and molecular level. As biomark-
while still managing to maintain relatively nor- ers they are important because they inform us
mal visual cycle metabolism and upkeep of that the RPE is chronically stressed and func-
Bruch’s membrane. With aging the normal level tionally compromised, but still viable. From a
of homeostatic HSP function is insufficient to natural history perspective, the disappearance of
maintain repair. This oxidative stress-induced these lesions is often associated with the death
cellular damage also contributes to increased of RPE leading to ARGA and visual loss [125].
protein misfolding and formation of detrimental Once begun, the process of ARGA is progres-
protein aggregates within the cytosol, further sive [96].
compromising cell function [118, 119]. Here, Although drusen are a sign of impaired
activation of additional compensatory processes autophagy and inflammaging (and indirectly of
may mitigate HSPs failure. Two that are particu- impaired proteasomal degradation and HSP func-
larly important are the proteasome and autoph- tion), their presence nonetheless indicates that
agy pathways. The first involves tagging these pathways are still functional [93]. SDM
misfolded proteins with ubiquitin and transfer- appears to reduce the risks of visual loss in high-
ring them to the proteasome degradation path- risk AMD from both ARGA and CNV by improv-
way, which is a multicatalytic proteolytic ing, rather than further stressing and
complex that recognizes and selectively degrades decompensating RPE function; and by preserv-
oxidatively damaged and ubiquitinated proteins. ing the integrity of the RPE/Bruch’s membrane
The second, the autophagy pathway is in itself complex [11, 26, 56].
an inbuilt intracellular waste disposal system As noted above, the NF kappaB pathway,
[120–124]. resulting in the assembly of the NLRP3 inflam-
masome, appears to be crucial to the pathogene-
sis of AMD [126]. The biological significance of
30.9 hy Drusen Elimination
W NLRP3 inflammasome activation is the release of
Might Be Undesirable active IL-1β and IL-18 into the extracellular
space. The secreted IL-1β facilitates the chronic
HSPs, the proteasome, and cellular autophagy inflammatory response in the tissues while IL-18
are the three key intracellular pathways that promotes caspase-3 dependent RPE apoptosis,
maintain RPE homeostasis. In aging, all are both hallmarks of AMD [65]. Inhibition of IL-1β
increasingly compromised leading to activation and IL-18 expression via RPE HSP activation is
of the nuclear factor-kappaB (NF-kappaB) sig- thus yet another point in the disease process
nalling pathway. NF-kappaB, a transcription fac- where retinal laser sublethal to the RPE may
reduce the risk of CNV in AMD by inhibiting the RPE. The resultant reversal of the chronic
chronic inflammation [127]. disease process reduces the risks of visual loss—
Finally, at the tissue level, retinal laser-induced including the development of CNV. Absent
inflammatory recalibration in the direction of res- LIRD, the benefits of modern laser are unop-
toration and repair and away from progressive posed by adverse treatment effects and are great-
degeneration has local microglial and systemic est in eyes with the highest risks of age-related
components. These include activation of resident visual loss [11, 56]. Conversely, LIRD is adverse
retinal stem cells and recruitment of bone to retinal integrity and retinal and visual function.
marrow-derived stem cells, a potent combination In AMD, LIRD may accelerate retinal degenera-
that can foster repair, regeneration, and function, further increasing the risk of visual loss
tional restoration [54, 55, 111]. from ARGA and CNV, especially in the sickest,
While difficult to demonstrate clinically in dry most compromised and vulnerable eyes [44, 52,
AMD, the images in Figs. 30.1, 30.2, 30.3, 30.4, 53]. Thus drusen elimination requiring LIRD
30.5 illustrate the anti-inflammatory effects of appears undesirable, especially in the most func-
panmacular and total retinal SDM as monother- tionally compromised high-risk eyes, and of
apy in various other clinical settings described uncertain long-term benefit in eyes with early
above. Note also the complete absence of LIRD AMD. In contrast, retinal laser treatment sub-
in each case. lethal to the RPE, and thus wholly therapeutic,
appears to hold great promise in the prevention of
visual loss from chronic progressive retinopa-
30.10 Summary thies in general, and from CNV in AMD in par-
ticular. Further study will be illuminating.
In summary, as the most important cause of CNV
and related visual loss, AMD is a complex web of Key Learning Points
dysfunctional intracellular and extracellular 1. While CNV may result from focal damage to
events involving hundreds, if not thousands, of the macula from a number of causes, the main
signalling molecules in innumerable interrela- cause is chronic progressive disease, princi-
tions. At our current level of understanding it is pally age-related macular degeneration.
hard to envisage a targeted therapy, acting on a 2. All chronic progressive retinopathies (CPRs)
specific molecule(s) at a particular point in a par- are neurodegenerations, and as such have
ticular pathway(s), that might singularly modu- much in common despite disparate etiologies
late a disease process of such complexity. The and phenotypes.
complexity of targeted therapeutic alteration of 3. Dysfunction of the retinal pigment epithelium
cell physiology dwarfs the comparative simplic- is a key commonality of all CRPs, leading to
ity of binding VEGF in the extracellular space the second key commonality of chronic, self-
accounting for the failure of targeted drug therapy perpetuating, degenerative inflammation.
to prevent AMD—and thus CNV—to date. Yet These are the key predisposing factors to
the judicious application of light seems to do CNV in AMD.
exactly that. Rather than attempting to selectively 4. The therapeutic effects of retinal laser treat-
manipulate cell chemistry, retinal laser awakens ment improve RPE and thus retinal function,
powerful and fundamental mechanisms of cellu- and are antagonistic to chronic inflammation,
lar repair and functional restoration harnessing resulting in repair and functional restoration.
this same biologic complexity to advantage. 5. Laser-induced retinal damage (LIRD) is the
Acting on a fraction of proteins in a single cell cause of all adverse treatment effects and
via HSP activation, laser is a catalyst launching a compromises the therapeutic benefits of laser
multitude of cascading effects within and far treatment while offering no unique benefits
beyond the cell, resulting in physiologic—and over modern retinal laser therapy which is
thus ideal—functional normalization, “resetting” reliably sublethal to the RPE. LIRD increases
AMD progression and the risks of visual loss, 6. A randomized, placebo-controlled, clinical trial of
high-dose supplementation with vitamins C and
including CNV, particularly in high-risk eyes. E, beta carotene, and zinc for age-related macular
6. By eliminating LIRD and maximizing thera- degeneration and vision loss: AREDS report no. 8.
peutic laser effects, modern retinal laser, epit- Arch Ophthalmol. 2001;119(10):1417–36. Epub
omized by low-intensity/high-density 2001/10/23. PubMed PMID: 11594942; PubMed
Central PMCID: PMC1462955.
subthreshold diode micropulse laser (SDM), 7. Heier JS, Antoszyk AN, Pavan PR, Leff SR,
safely slows disease progression, reducing the Rosenfeld PJ, Ciulla TA, et al. Ranibizumab for
risks of visual loss and CNV in AMD, espe- treatment of neovascular age-related macular degen-
cially in the highest risk eyes. eration: a phase I/II multicenter, controlled, multi-
dose study. Ophthalmology. 2006;113(4):633.e1–4.
7. By addressing the commonalities of retinal Epub 2006/02/18. https://doi.org/10.1016/j.oph-
neurodegenerations via non-targeted, pathose- tha.2005.10.052. PubMed PMID: 16483659.
lective disease-specific repair, SDM may offer 8. Chhablani J, Roh YJ, Jobling AI, Fletcher EL, Lek
similar benefits in other retinopathies that pre- JJ, Bansal P, et al. Restorative retinal laser therapy:
present state and future directions. Surv Ophthalmol.
dispose to macular CNV. 2017. Epub 2017/10/11. https://doi.org/10.1016/j.sur-
vophthal.2017.09.008. PubMed PMID: 28987614.
Financial Disclosures 9. Guymer RH, Brassington KH, Dimitrov P, Makeyeva
Dr. Luttrull: G, Plunkett M, Xia W, et al. Nanosecond-laser appli-
cation in intermediate AMD: 12-month results of
Ojai Retinal Technologies, LLC— fundus appearance and macular function. Clin Exp
Management, Patent, Equity. Ophthalmol 2014;42(5):466–479. Epub 2013/10/15.
Retinal Protection Sciences, LLC— https://doi.org/10.1111/ceo.12247. PubMed PMID:
Management, Equity. 24118741.
10. Jobling AI, Guymer RH, Vessey KA, Greferath U,
Replenish, Inc.—Patent, Equity. Mills SA, Brassington KH, et al. Nanosecond laser
Dr. Kent: therapy reverses pathologic and molecular changes in
None. age-related macular degeneration without retinal dam-
age. FASEB J 2015;29(2):696–710. Epub 2014/11/14.
https://doi.org/10.1096/fj.14-262444. PubMed PMID:
25392267.
References 11. Luttrull JK, Sinclair SH, Elmann S, Glaser BM. Low
incidence of choroidal neovascularization follow-
1. Jepson CN, Wetzig PC. Photocoagulation in dis- ing subthreshold diode micropulse laser (SDM) in
ciform macular degeneration. Am J Ophthalmol high-risk AMD. PLoS One 2018;13(8):e0202097.
1969;67(6):920–930. Epub 1969/06/01. PubMed Epub 2018/08/24. https://doi.org/10.1371/journal.
PMID: 5785852. pone.0202097. PubMed PMID: 30138455.
2. Argon laser photocoagulation for senile macu- 12. Friedman DS, O'Colmain BJ, Munoz B, Tomany SC,
lar degeneration. Results of a randomized clinical McCarty C, de Jong PT, et al. Prevalence of age-related
trial. Arch Ophthalmol 1982;100(6):912–918. Epub macular degeneration in the United States. Arch
1982/06/01. PubMed PMID: 7046707. Ophthalmol 2004;122(4):564–572. Epub 2004/04/14.
3. Avila MP, Weiter JJ, Jalkh AE, Trempe CL, Pruett RC, https://doi.org/10.1001/archopht.122.4.564. PubMed
Schepens CL. Natural history of choroidal neovas- PMID: 15078675.
cularization in degenerative myopia. Ophthalmology 13. Wong WL, Su X, Li X, Cheung CM, Klein R,
1984;91(12):1573–1581. Epub 1984/12/01. PubMed Cheng CY, et al. Global prevalence of age-related
PMID: 6084222. macular degeneration and disease burden projection
4. Francois J, De Laey JJ, Cambie E, Hanssens M, for 2020 and 2040: a systematic review and meta-
Victoria-Troncoso V. Neovascularization after argon analysis. Lancet Glob Health 2014;2(2):e106–e116.
laser photocoagulation of macular lesions. Am J Epub 2014/08/12. https://doi.org/10.1016/s2214-
Ophthalmol 1975;79(2):206–210. Epub 1975/02/01. 109x(13)70145-1. PubMed PMID: 25104651.
PubMed PMID: 1167736. 14. Molnar I, Pournaras CJ, Tsacopoulos M, Gilodi N,
5. Recurrent choroidal neovascularization after argon Leuenberger PM. [Panphotocoagulation. Mechanism
laser photocoagulation for neovascular maculopa- of action and early indications]. Ophtalmologie:
thy. Macular Photocoagulation Study Group. Arch organe de la Societe francaise d’ophtalmologie
Ophthalmol 1986;104(4):503–512. Epub 1986/04/01. 1987;1(4):449–452. Epub 1987/10/01. PubMed
PubMed PMID: 2420315. PMID: 3153915.
15. Shimizu K. Light coagulation in diabetic retinopathy protective therapy in primary open-angle glaucoma. In:
and the mechanism of its effect. Ganka Ophthalmol Samples JR, Kneppper PA, editors. Glaucoma research
1971;13(11):1001–1006. Epub 1971/11/01. PubMed 2018–2020. Amsterdam: Kugler Publications; 2018.
PMID: 5169175. p. 281–94.
16. Stefansson E. Oxygen and diabetic eye disease. 26. Abdelsalam A, Del Priore L, Zarbin MA. Drusen
Graefes Arch Clin Exp Ophthalmol 1990;228(2):120– in age-related macular degeneration: pathogenesis,
123. Epub 1990/01/01. PubMed PMID: 2186971. natural course, and laser photocoagulation-induced
17. Brader HS, Young LH. Subthreshold diode micropulse regression. Surv Ophthalmol 1999;44(1):1–29. Epub
laser: a review. Semin Ophthalmol 2016;31(1–2):30– 1999/08/31. PubMed PMID: 10466585.
39. Epub 2016/03/10. https://doi.org/10.3109/088205 27. Bressler NM, Munoz B, Maguire MG, Vitale SE,
38.2015.1114837. PubMed PMID: 26959127. Schein OD, Taylor HR, et al. Five-year incidence and
18. Chen G, Tzekov R, Li W, Jiang F, Mao S, Tong disappearance of drusen and retinal pigment epithe-
Y. Subthreshold micropulse diode laser versus con- lial abnormalities. Waterman Stud Arch Ophthalmol
ventional laser photocoagulation for diabetic macular 1995;113(3):301–308. Epub 1995/03/01. PubMed
Edema: a meta-analysis of randomized controlled tri- PMID: 7534060.
als. Retina 2016;36(11):2059–2065. Epub 2016/10/25. 28. Bressler SB, Maguire MG, Bressler NM, Fine
https://doi.org/10.1097/iae.0000000000001053. SL. Relationship of drusen and abnormalities of
PubMed PMID: 27096529. the retinal pigment epithelium to the prognosis of
19. Luttrull JK, Musch DC, Mainster MA. Subthreshold neovascular macular degeneration. The Macular
diode micropulse photocoagulation for the treatment Photocoagulation Study Group. Arch Ophthalmol
of clinically significant diabetic macular oedema. Br 1990;108(10):1442–1447. Epub 1990/10/01. PubMed
J Ophthalmol 2005;89(1):74–80. Epub 2004/12/24. PMID: 1699513.
https://doi.org/10.1136/bjo.2004.051540. PubMed 29. Chamberlin JA, Bressler NM, Bressler SB, Elman
PMID: 15615751; PubMed Central PMCID: MJ, Murphy RP, Flood TP, et al. The use of fundus
PMC1772486. photographs and fluorescein angiograms in the identi-
20. Luttrull JK, Musch DC, Spink CA. Subthreshold diode fication and treatment of choroidal neovascularization
micropulse panretinal photocoagulation for prolifera- in the Macular Photocoagulation Study. The Macular
tive diabetic retinopathy. Eye (Lond) 2008;22(5):607– Photocoagulation Study Group Ophthalmology
612. Epub 2007/02/13. https://doi.org/10.1038/ 1989;96(10):1526–1534. Epub 1989/10/01. PubMed
sj.eye.6702725. PubMed PMID: 17293791. PMID: 2479899.
21. Luttrull JK, Spink CJ. Serial optical coherence tomog- 30. Smiddy WE, Fine SL. Prognosis of patients with bilat-
raphy of subthreshold diode laser micropulse photo- eral macular drusen. Ophthalmology 1984;91(3):271–
coagulation for diabetic macular edema. Ophthalmic 277. Epub 1984/03/01. PubMed PMID: 6201789.
Surg Lasers Imaging 2006;37(5):370–377. Epub 31. Ho AC, Maguire MG, Yoken J, Lee MS, Shin DS,
2006/10/05. PubMed PMID: 17017196. Javornik NB, et al. Laser-induced drusen reduc-
22. Luttrull JK, Chang DB, Margolis BW, Dorin G, tion improves visual function at 1 year. Choroidal
Luttrull DK. Laser resesitization of medically Neovascularization Prevention Trial Research Group.
unresponsive neovascular age-related macular Ophthalmology 1999;106(7):1367–1373.; discussion
degeneration: efficacy and Implications. Retina 74. Epub 1999/07/16. PubMed PMID: 10406624.
2015;35(6):1184–1194. Epub 2015/02/05. https://doi. 32. Olk RJ, Friberg TR, Stickney KL, Akduman L,
org/10.1097/iae.0000000000000458. PubMed PMID: Wong KL, Chen MC, et al. Therapeutic benefits of
25650711. infrared (810-nm) diode laser macular grid photo-
23. Luttrull JK, Margolis BW. Functionally guided reti- coagulation in prophylactic treatment of nonexu-
nal protective therapy for dry age-related macular dative age-related macular degeneration: two-year
and inherited retinal degenerations: a pilot study. results of a randomized pilot study. Ophthalmology
Invest Ophthalmol Vis Sci 2016;57(1):265–275. Epub 1999;106(11):2082–2090. Epub 1999/11/26. https://
2016/01/29. https://doi.org/10.1167/iovs.15-18163. doi.org/10.1016/s0161-6420(99)90487-6. PubMed
PubMed PMID: 26818793. PMID: 10571341.
24. Luttrull JK. Improved retinal and visual function 33. Dithmar S, Pollithy S, Ach T. Disappearance of cen-
following panmacular subthreshold diode micro- tral confluent soft drusen following vitrectomy and
pulse laser for retinitis pigmentosa. Eye (Lond). ILM peeling. Eye (Lond). 2013;27(6):779–81. Epub
2018;32(6):1099–110. Epub 2018/02/17. https:// 2013/04/06. https://doi.org/10.1038/eye.2013.35.
doi.org/10.1038/s41433-018-0017-3. PubMed PubMed PMID: 23558209; PubMed Central PMCID:
PMID: 29449615; PubMed Central PMCID: PMCPMC3682349.
PMCPMC5997672. 34. Lim LT, Holicki J, Walker JD. Disappearance of
25. Luttrull JK, Samples JR, Kent D, Lum BJ. Panmacular extramacular drusen after chronic retinal detach-
subthreshold diode micropulse laser (SDM) as neuro- ment. Oman journal of ophthalmology. 2012;5(1):67.
Epub 2012/05/05. https://doi.org/10.4103/0974- 2015;(10):Cd006537. Epub 2015/10/24. https://doi.

620x.94793. PubMed PMID: 22557885; PubMed org/10.1002/14651858.CD006537.pub3. PubMed
Central PMCID: PMCPMC3339683. PMID: 26493180; PubMed Central PMCID:
35. Duvall J, Tso MO. Cellular mechanisms of resolution PMCPmc4733883.
of drusen after laser coagulation. An experimental 45. Dorin G. Subthreshold and micropulse diode laser pho-
study. Arch Ophthalmol 1985;103(5):694–703. Epub tocoagulation. Semin Ophthalmol 2003;18(3):147–
1985/05/01. PubMed PMID: 4039560. 153. Epub 2004/10/30. https://doi.org/10.1076/
36. Argon laser photocoagulation for neovascular macu- soph.18.3.147.29812. PubMed PMID: 15513476.
lopathy. Three-year results from randomized clinical 46. Luttrull JK, Sramek C, Palanker D, Spink CJ, Musch
trials. Macular Photocoagulation Study Group. Arch DC. Long-term safety, high-resolution imaging, and
Ophthalmol. 1986;104(5):694–701. Epub 1986/05/01. tissue temperature modeling of subvisible diode micro-
PubMed PMID: 2423061. pulse photocoagulation for retinovascular macular
37. Treatment techniques and clinical guidelines for edema. Retina 2012;32(2):375–386. Epub 2011/10/06.
photocoagulation of diabetic macular edema. Early https://doi.org/10.1097/IAE.0b013e3182206f6c.
Treatment Diabetic Retinopathy Study Report Number PubMed PMID: 21971077.
2. Early Treatment Diabetic Retinopathy Study 47. Lavinsky D, Wang J, Huie P, Dalal R, Lee SJ, Lee
Research Group. Ophthalmology. 1987;94(7):761–74. DY, et al. Nondamaging Retinal Laser Therapy:
Epub 1987/07/01. PubMed PMID: 3658348. Rationale and Applications to the Macula. Invest
38. Akduman L, Olk RJ. Subthreshold (invisible) modi- Ophthalmol Vis Sci. 2016;57(6):2488–500. Epub
fied grid diode laser photocoagulation in diffuse 2016/05/10. https://doi.org/10.1167/iovs.15-18981.
diabetic macular edema (DDME). Ophthalmic Surg PubMed PMID: 27159441; PubMed Central PMCID:
Lasers 1999;30(9):706–714. Epub 1999/11/26. PMCPMC5995023.
PubMed PMID: 10574491. 48. Roider J, Brinkmann R, Wirbelauer C, Laqua H,
39. Laser treatment in patients with bilateral large drusen: Birngruber R. Subthreshold (retinal pigment epithe-
the complications of age-related macular degeneration lium) photocoagulation in macular diseases: a pilot
prevention trial. Ophthalmology. 2006;113(11):1974– study. Br J Ophthalmol. 2000;84(1):40–7. Epub
86. Epub 2006/11/01. https://doi.org/10.1016/j.oph- 1999/12/28. PubMed PMID: 10611098; PubMed
tha.2006.08.015. PubMed PMID: 17074563. Central PMCID: PMCPMC1723228.
40. Owens SL, Bunce C, Brannon AJ, Wormald R, Bird 49. Sramek C, Mackanos M, Spitler R, Leung LS, Nomoto
AC. Prophylactic laser treatment appears to promote H, Contag CH, et al. Non-damaging retinal photother-
choroidal neovascularisation in high-risk ARM: results apy: dynamic range of heat shock protein expression.
of an interim analysis. Eye (Lond) 2003;17(5):623– Invest Ophthalmol Vis Sci 2011;52(3):1780–1787.
627. Epub 2003/07/12. https://doi.org/10.1038/ Epub 2010/11/20. https://doi.org/10.1167/iovs.10-
sj.eye.6700442. PubMed PMID: 12855972. 5917. PubMed PMID: 21087969.
41. Owens SL, Bunce C, Brannon AJ, Xing W, Chisholm 50. Sramek CK, Leung LS, Paulus YM, Palanker
IH, Gross M, et al. Prophylactic laser treatment DV. Therapeutic window of retinal photocoagulation
hastens choroidal neovascularization in unilateral with green (532-nm) and yellow (577- nm) lasers.
age-related maculopathy: final results of the drusen Ophthalmic Surg Lasers Imaging 2012;43(4):341–347.
laser study. Am J Ophthalmol 2006;141(2):276– Epub 2012/05/17. https://doi.org/10.3928/15428877-
281. Epub 2006/02/07. https://doi.org/10.1016/j. 20120426-05. PubMed PMID: 22589338.
ajo.2005.08.019. PubMed PMID: 16458680. 51. Roider J, Brinkmann R, Wirbelauer C, Birngruber
42. Kaiser RS, Berger JW, Maguire MG, Ho AC, Javornik R, Laqua H. Variability of RPE reaction in two cases
NB. Laser burn intensity and the risk for choroi- after selective RPE laser effects in prophylactic treat-
dal neovascularization in the CNVPT Fellow Eye ment of drusen. Graefes Arch Clin Exp Ophthalmol
Study. Arch Ophthalmol 2001;119(6):826–832. Epub 1999;237(1):45–50. Epub 1999/02/10. PubMed
2001/06/19. PubMed PMID: 11405833. PMID: 9951641.
43. Rodanant N, Friberg TR, Cheng L, Aurora A, Bartsch 52. Guymer RH, Wu Z, Hodgson LAB, Caruso E,
D, Toyoguchi M, et al. Predictors of drusen reduc- Brassington KH, Tindill N, et al. Subthreshold nano-
tion after subthreshold infrared (810 nm) diode laser second laser intervention in age-related macular degen-
macular grid photocoagulation for nonexudative eration: the LEAD randomized controlled clinical trial
age-related macular degeneration. Am J Ophthalmol ophthalmology 2018. Epub 2018/09/24. https://doi.
2002;134(4):577–585. Epub 2002/10/18. PubMed org/10.1016/j.ophtha.2018.09.015. PubMed PMID:
PMID: 12383815. 30244144.
44. Virgili G, Michelessi M, Parodi MB, Bacherini D, 53. Prahs P, Walter A, Regler R, Theisen-Kunde D,
Evans JR. Laser treatment of drusen to prevent pro- Birngruber R, Brinkmann R, et al. Selective retina
gression to advanced age-related macular degenera- therapy (SRT) in patients with geographic atrophy due
tion. The Cochrane database of systematic reviews. to age-related macular degeneration. Graefes Arch
Clin Exp Ophthalmol 2010;248(5):651–658. Epub 64. Medzhitov R. Origin and physiological roles of
2009/12/22. https://doi.org/10.1007/s00417-009- inflammation. Nature 2008;454(7203):428–435. Epub
1208-1. PubMed PMID: 20024687. 2008/07/25. https://doi.org/10.1038/nature07201.
54. Midena E, Pilotto E. Emerging Insights into PubMed PMID: 18650913.
Pathogenesis. Dev Ophthalmol 2017;60:16–27. Epub 65. Ambati J, Atkinson JP, Gelfand BD. Immunology of
2017/04/21. https://doi.org/10.1159/000459687. age-related macular degeneration. Nat Rev Immunol
PubMed PMID: 28427061. 2013;13(6):438–451. Epub 2013/05/25. https://doi.
55. Vujosevic S, Micera A, Bini S, Berton M, Esposito org/10.1038/nri3459. PubMed PMID: 23702979;
G, Midena E. Proteome analysis of retinal glia cells- PubMed Central PMCID: PMC3941009.
related inflammatory cytokines in the aqueous humour 66. Xu H, Chen M, Forrester JV. Para-inflammation in the
of diabetic patients. Acta Ophthalmol 2016;94(1):56– aging retina. Prog Retin Eye Res 2009;28(5):348–368.
64. Epub 2015/08/14. https://doi.org/10.1111/ Epub 2009/06/30. https://doi.org/10.1016/j.pretey-
aos.12812. PubMed PMID: 26268591. eres.2009.06.001. PubMed PMID: 19560552.
56. Luttrull JK. Slowed progression of age-related geo- 67. Doyle SL, Campbell M, Ozaki E, Salomon RG,
graphic atrophy following subthreshold diode micro- Mori A, Kenna PF, et al. NLRP3 has a protective
pulse laser—in review. 2019. role in age-related macular degeneration through the
57. Luttrull JK, Sinclair SH. Safety of transfoveal sub- induction of IL-18 by drusen components. Nat Med
threshold diode micropulse laser for fovea-involving 2012;18(5):791–798. Epub 2012/04/10. https://doi.
diabetic macular edema in eyes with good visual acu- org/10.1038/nm.2717. PubMed PMID: 22484808;
ity. Retina 2014;34(10):2010–2020. Epub 2014/05/20. PubMed Central PMCID: PMC3984677.
https://doi.org/10.1097/iae.0000000000000177. 68. Hollyfield JG, Bonilha VL, Rayborn ME, Yang X,
PubMed PMID: 24837050. Shadrach KG, Lu L, et al. Oxidative damage-induced
58. Luttrull JK, Dorin G. Subthreshold diode micropulse inflammation initiates age-related macular degenera-
laser photocoagulation (SDM) as invisible retinal pho- tion. Nat Med 2008;14(2):194–198. Epub 2008/01/29.
totherapy for diabetic macular edema: a review. Curr https://doi.org/10.1038/nm1709. PubMed PMID:
Diabetes Rev 2012;8(4):274–284. Epub 2012/05/17. 18223656; PubMed Central PMCID: PMC2748836.
PubMed PMID: 22587512; PubMed Central PMCID: 69. Kaneko H, Dridi S, Tarallo V, Gelfand BD, Fowler
PMC3412206. BJ, Cho WG, et al. DICER1 deficit induces Alu
59. Lavinsky D, Cardillo JA, Melo LA, Jr., Dare A, Farah RNA toxicity in age-related macular degeneration.
ME, Belfort R, Jr. Randomized clinical trial evaluating Nature 2011;471(7338):325–330. Epub 2011/02/08.
mETDRS versus normal or high-density micropulse https://doi.org/10.1038/nature09830. PubMed PMID:
photocoagulation for diabetic macular edema. Invest 21297615; PubMed Central PMCID: PMC3077055.
Ophthalmol Vis Sci 2011;52(7):4314–4323. Epub 70. Tarallo V, Hirano Y, Gelfand BD, Dridi S, Kerur
2011/02/25. https://doi.org/10.1167/iovs.10-6828. N, Kim Y, et al. DICER1 loss and Alu RNA induce
PubMed PMID: 21345996. age-related macular degeneration via the NLRP3
60. Franceschi C, Campisi J. Chronic inflammation inflammasome and MyD88. Cell 2012;149(4):847–
(inflammaging) and its potential contribution to age- 859. Epub 2012/05/01. https://doi.org/10.1016/j.
associated diseases. J Gerontol A Biol Sci Med Sci cell.2012.03.036. PubMed PMID: 22541070; PubMed
2014;69 Suppl 1:S4–S9. Epub 2014/05/17. https:// Central PMCID: PMC3351582.
doi.org/10.1093/gerona/glu057. PubMed PMID: 71. Tseng WA, Thein T, Kinnunen K, Lashkari K,
24833586. Gregory MS, D’Amore PA, et al. NLRP3 inflam-
61. Howcroft TK, Campisi J, Louis GB, Smith MT, Wise masome activation in retinal pigment epithelial cells
B, Wyss-Coray T, et al. The role of i nflammation in age- by lysosomal destabilization: implications for age-
related disease. Aging (Albany NY). 2013;5(1):84– related macular degeneration. Invest Ophthalmol Vis
93. Epub 2013/03/12. https://doi.org/10.18632/ Sci 2013;54(1):110–120. Epub 2012/12/12. https://
aging.100531. PubMed PMID: 23474627; PubMed doi.org/10.1167/iovs.12-10655. PubMed PMID:
Central PMCID: PMC3616233. 23221073; PubMed Central PMCID: PMC3544415.
62. Baldwin RL. Energetics of protein folding. J Mol 72. Anderson DH, Radeke MJ, Gallo NB, Chapin EA,
Biol 2007;371(2):283–301. Epub 2007/06/22. https:// Johnson PT, Curletti CR, et al. The pivotal role of the
doi.org/10.1016/j.jmb.2007.05.078. PubMed PMID: complement system in aging and age-related macular
17582437. degeneration: hypothesis re-visited. Prog Retin Eye
63. Kregel KC. Heat shock proteins: modifying factors in Res 2010;29(2):95–112. Epub 2009/12/08. https://
physiological stress responses and acquired thermo- doi.org/10.1016/j.preteyeres.2009.11.003. PubMed
tolerance. J Appl Physiol (1985). 2002;92(5):2177– PMID: 19961953.
86. Epub 2002/04/19. https://doi.org/10.1152/ 73. Clark SJ, Perveen R, Hakobyan S, Morgan BP, Sim
japplphysiol.01267.2001. PubMed PMID: 11960972. RB, Bishop PN, et al. Impaired binding of the age-
related macular degeneration-associated comple- https://doi.org/10.1056/nejm199909023411006.

ment factor H 402H allotype to Bruch's membrane PubMed PMID: 10471461.
in human retina. J Biol Chem 2010;285(39):30192– 83. Hageman GS, Anderson DH, Johnson LV, Hancox
30202. Epub 2010/07/28. https://doi.org/10.1074/jbc. LS, Taiber AJ, Hardisty LI, et al. A common hap-
M110.103986. PubMed PMID: 20660596; PubMed lotype in the complement regulatory gene factor H
Central PMCID: PMC2943316. (HF1/CFH) predisposes individuals to age-related
74. Johnson PT, Betts KE, Radeke MJ, Hageman GS, macular degeneration. Proc Natl Acad Sci USA
Anderson DH, Johnson LV. Individuals homo- 2005;102(20):7227–7232. Epub 2005/05/05. https://
zygous for the age-related macular degenera- doi.org/10.1073/pnas.0501536102. PubMed PMID:
tion risk-conferring variant of complement factor 15870199; PubMed Central PMCID: PMC1088171.
H have elevated levels of CRP in the choroid. 84. Haines JL, Hauser MA, Schmidt S, Scott WK, Olson
Proc Natl Acad Sci USA 2006;103(46):17456– LM, Gallins P, et al. Complement factor H variant
17461. Epub 2006/11/03. https://doi.org/10.1073/ increases the risk of age-related macular degeneration.
pnas.0606234103. PubMed PMID: 17079491; Science 2005;308(5720):419–421. Epub 2005/03/12.
PubMed Central PMCID: PMC1859950. https://doi.org/10.1126/science.1110359. PubMed
75. Khandhadia S, Cipriani V, Yates JR, Lotery AJ. Age- PMID: 15761120.
related macular degeneration and the complement 85. Klein RJ, Zeiss C, Chew EY, Tsai JY, Sackler RS,
system. Immunobiology 2012;217(2):127–146. Haynes C, et al. Complement factor H polymor-
Epub 2011/08/27. https://doi.org/10.1016/j. phism in age-related macular degeneration. Science
imbio.2011.07.019. PubMed PMID: 21868123. 2005;308(5720):385–389. Epub 2005/03/12.
76. Lachmann PJ. The amplification loop of the com- https://doi.org/10.1126/science.1109557. PubMed
plement pathways. Adv Immunol 2009;104:115– PMID: 15761122; PubMed Central PMCID:
149. Epub 2009/01/01. https://doi.org/10.1016/ PMC1512523.
s0065-2776(08)04004-2. PubMed PMID: 20457117. 86. Bonafe M, Storci G, Franceschi C. Inflamm-aging
77. Laine M, Jarva H, Seitsonen S, Haapasalo K, Lehtinen of the stem cell niche: breast cancer as a paradig-
MJ, Lindeman N, et al. Y402H polymorphism of com- matic example: breakdown of the multi-shell cyto-
plement factor H affects binding affinity to C-reactive kine network fuels cancer in aged people. BioEssays
protein. J Immunol 2007;178(6):3831–3836. Epub 2012;34(1):40–49. Epub 2011/11/17. https://doi.
2007/03/07. PubMed PMID: 17339482. org/10.1002/bies.201100104. PubMed PMID:
78. Richards A, Kavanagh D, Atkinson JP. Inherited 22086861.
complement regulatory protein deficiency predisposes 87. Varadhan R, Yao W, Matteini A, Beamer BA, Xue QL,
to human disease in acute injury and chronic inflam- Yang H, et al. Simple biologically informed inflam-
matory states the examples of vascular damage in matory index of two serum cytokines predicts 10
atypical hemolytic uremic syndrome and debris accu- year all-cause mortality in older adults. J Gerontol
mulation in age-related macular degeneration. Adv A Biol Sci Med Sci 2014;69(2):165–173. Epub
Immunol 2007;96:141–177. Epub 2007/11/06. https:// 2013/05/22. https://doi.org/10.1093/gerona/glt023.
doi.org/10.1016/s0065-2776(07)96004-6. PubMed PubMed PMID: 23689826; PubMed Central PMCID:
PMID: 17981206. PMC4038244.
79. Sjoberg AP, Trouw LA, Clark SJ, Sjolander J, Hein- 88. Tobe T, Ortega S, Luna JD, Ozaki H, Okamoto N,
egard D, Sim RB, et al. The factor H variant associated Derevjanik NL, et al. Targeted disruption of the FGF2
with age-related macular degeneration (His-384) and gene does not prevent choroidal neovascularization in
the non-disease-associated form bind differentially to a murine model. Am J Pathol 1998;153(5):1641–1646.
C-reactive protein, fibromodulin, DNA, and necrotic Epub 1998/11/12. https://doi.org/10.1016/s0002-
cells. J Biol Chem 2007;282(15):10894–10900. Epub 9440(10)65753-7. PubMed PMID: 9811357; PubMed
2007/02/13. https://doi.org/10.1074/jbc.M610256200. Central PMCID: PMC1853405.
PubMed PMID: 17293598. 89. Kent D, Sheridan C. Choroidal neovascular-
80. Tuo J, Grob S, Zhang K, Chan CC. Genetics of ization: a wound healing perspective. Mol Vis.
immunological and inflammatory components in age- 2003;9:747–55.
related macular degeneration. Ocul Immunol Inflamm 90. Clark RAF. Wound repair: overview and general
2012;20(1):27–36. Epub 2012/02/14. https://doi.org considerations. In: Clark RAF, editor. The molecu-
/10.3109/09273948.2011.628432. PubMed PMID: lar and cellular biology of wound repair. Second ed.
22324898; PubMed Central PMCID: PMC3292785. New York: Plenum; 1996. p. 1–50.
81. Clark RA. Basics of cutaneous wound repair. J 91. Hunt TK. Basic principles of wound healing. J Trauma
Dermatol Surg Oncol 1993;19(8):693–706. Epub 1990;30(12 Suppl):S122–S128. Epub 1990/12/01.
1993/08/01. PubMed PMID: 8349909. PubMed PMID: 2254971.
82. Singer AJ, Clark RA. Cutaneous wound healing. N 92. Nathan C, Sporn M. Cytokines in context. J Cell Biol.
Engl J Med 1999;341(10):738–746. Epub 1999/09/02. 1991;113(5):981–6.
93. Blasiak J, Pawlowska E, Szczepanska J, Kaarniranta 103. Xu Q, Metzler B, Jahangiri M, Mandal

K. Interplay between Autophagy and the Ubiquitin- K. Molecular chaperones and heat shock proteins
Proteasome System and Its Role in the Pathogenesis in atherosclerosis. Am J Physiol Heart Circ Physiol.
of Age-Related Macular Degeneration. International 2012;302(3):H506–14. Epub 2011/11/08. https://
journal of molecular sciences. 2019;20(1). Epub doi.org/10.1152/ajpheart.00646.2011. PubMed
2019/01/11. https://doi.org/10.3390/ijms20010210. PMID: 22058161; PubMed Central PMCID:
PubMed PMID: 30626110; PubMed Central PMCPMC3353778.
PMCID: PMCPMC6337628. 104. Santoro MG. Heat shock factors and the control of the
94. Bornstein P. Diversity of function is inherent in stress response. Biochem Pharmacol 2000;59(1):55–
matricellular proteins: an appraisal of thrombospon- 63. Epub 1999/12/22. PubMed PMID: 10605935.
din 1. J Cell Biol. 1995;130(3):503–6. 105. Lanneau D, de Thonel A, Maurel S, Didelot C,
95. Mones J, Garcia M, Biarnes M, Lakkaraju A, Ferraro Garrido C. Apoptosis versus cell differentiation: role
L. Drusen Ooze: a novel hypothesis in geographic of heat shock proteins HSP90, HSP70 and HSP27.
atrophy. Ophthalmol Retina 2017;1(6):461–473. Prion. 2007;1(1):53–60. Epub 2007/01/01. PubMed
Epub 2017/01/01. https://doi.org/10.1016/j. PMID: 19164900; PubMed Central PMCID:
oret.2017.02.006. PubMed PMID: 31047436. PMCPmc2633709.
96. Schmitz-Valckenberg S, Sadda S, Staurenghi G, 106. De Cilla S, Vezzola D, Farruggio S, Vujosevic
Chew EY, Fleckenstein M, Holz FG. Geographic atro- S, Clemente N, Raina G, et al. The subthreshold
phy: semantic considerations and literature review. micropulse laser treatment of the retina restores the
Retina. 2016;36(12):2250–64. Epub 2016/08/24. oxidant/antioxidant balance and counteracts pro-
https://doi.org/10.1097/iae.0000000000001258. grammed forms of cell death in the mice eyes. Acta
PubMed PMID: 27552292; PubMed Central Ophthalmol. 2018. Epub 2018/12/27. https://doi.
PMCID: PMCPMC5115977. org/10.1111/aos.13995. PubMed PMID: 30585429.
97. Pockley AG. Heat shock proteins, inflamma- 107. Gao X, Xing D. Molecular mechanisms of cell pro-
tion, and cardiovascular disease. Circulation liferation induced by low power laser irradiation.
2002;105(8):1012–1017. Epub 2002/02/28. PubMed Journal of biomedical science. 2009;16:4. Epub
PMID: 11864934. 2009/03/11. https://doi.org/10.1186/1423-0127-
98. Fort PE, Lampi KJ. New focus on alpha-crystallins 16-4. PubMed PMID: 19272168; PubMed Central
in retinal neurodegenerative diseases. Exp Eye PMCID: PMCPMC2644974.
Res. 2011;92(2):98–103. Epub 2010/12/01. https:// 108. Iwami H, Pruessner J, Shiraki K, Brinkmann R, Miura
doi.org/10.1016/j.exer.2010.11.008. PubMed Y. Protective effect of a laser-induced sub-lethal
PMID: 21115004; PubMed Central PMCID: temperature rise on RPE cells from oxidative stress.
PMCPmc4454605. Exp Eye Res 2014;124:37–47. Epub 2014/05/08.
99. Dudeja V, Vickers SM, Saluja AK. The role of https://doi.org/10.1016/j.exer.2014.04.014. PubMed
heat shock proteins in gastrointestinal diseases. PMID: 24800654.
Gut. 2009;58(7):1000–9. Epub 2009/06/13. 109. Karu TI, Kolyakov SF. Exact action spectra for cel-
https://doi.org/10.1136/gut.2007.140194. PubMed lular responses relevant to phototherapy. Photomed
PMID: 19520890; PubMed Central PMCID: Laser Surg 2005;23(4):355–361. Epub 2005/09/08.
PMCPMC2896385. https://doi.org/10.1089/pho.2005.23.355. PubMed
100. Lu RC, Tan MS, Wang H, Xie AM, Yu JT, Tan L. Heat PMID: 16144476.
shock protein 70 in Alzheimer's disease. BioMed 110. Lund DJ, Sliney DH. A new understanding of
research international. 2014;2014:435203. Epub multiple-pulsed laser-induced retinal injury
2014/11/29. https://doi.org/10.1155/2014/435203. thresholds. Health Phys 2014;106(4):505–515.
PubMed PMID: 25431764; PubMed Central Epub 2014/02/25. https://doi.org/10.1097/
PMCID: PMCPMC4241292. HP.0b013e3182a2a837. PubMed PMID: 24562071.
101. Morimoto RI, Santoro MG. Stress-inducible 111. Caballero S, Kent DL, Sengupta N, Li Calzi S,
responses and heat shock proteins: new pharma- Shaw L, Beli E, et al. Bone marrow-derived cell
cologic targets for cytoprotection. Nat Biotechnol recruitment to the neurosensory retina and retinal
1998;16(9):833–838. Epub 1998/09/22. https://doi. pigment epithelial cell layer following subthresh-
org/10.1038/nbt0998-833. PubMed PMID: 9743115. old retinal phototherapy. Invest Ophthalmol Vis
102. Srivastava K, Narang R, Bhatia J, Saluja Sci. 2017.
D. Expression of heat shock protein 70 gene and its 112. Flaxel C, Bradle J, Acott T, Samples JR. Retinal
correlation with inflammatory markers in essential pigment epithelium produces matrix metalloprotein-
hypertension. PLoS One. 2016;11(3):e0151060. ases after laser treatment. Retina 2007;27(5):629–
Epub 2016/03/19. https://doi.org/10.1371/journal. 634. Epub 2007/06/15. https://doi.org/10.1097/01.
pone.0151060. PubMed PMID: 26989902; PubMed iae.0000249561.02567.fd. PubMed PMID:
Central PMCID: PMCPMC4798713. 17558327.
113. Hattenbach LO, Beck KF, Pfeilschifter J, Koch F, 121. Johnson LV, Leitner WP, Rivest AJ, Staples MK,
Ohrloff C, Schacke W. Pigment-epithelium-derived Radeke MJ, Anderson DH. The Alzheimer’s a beta-
factor is upregulated in photocoagulated human peptide is deposited at sites of complement activation
retinal pigment epithelial cells. Ophthalmic Res in pathologic deposits associated with aging and age-
2005;37(6):341–346. Epub 2005/09/15. https://doi. related macular degeneration. Proc Natl Acad Sci
org/10.1159/000088263. PubMed PMID: 16158012. USA 2002;99(18):11830–11835. Epub 2002/08/22.
114. Inagaki K, Shuo T, Katakura K, Ebihara N, Murakami https://doi.org/10.1073/pnas.192203399. PubMed
A, Ohkoshi K. Sublethal photothermal stimulation PMID: 12189211; PubMed Central PMCID:
with a micropulse laser induces heat shock pro- PMC129354.
tein expression in ARPE-19 cells. J Ophthalmol 122. Kopito RR. Aggresomes, inclusion bod-
2015;2015:729792. Epub 2015/12/24. https:// ies and protein aggregation. Trends Cell Biol
doi.org/10.1155/2015/729792. PubMed PMID: 2000;10(12):524–530. Epub 2000/12/21. PubMed
26697211; PubMed Central PMCID: PMC4677213. PMID: 11121744.
115. Karu T. Photobiology of low-power laser effects. 123. Ryhanen T, Hyttinen JM, Kopitz J, Rilla K, Kuusisto
Health Phys 1989;56(5):691–704. Epub 1989/05/01. E, Mannermaa E, et al. Crosstalk between Hsp70
PubMed PMID: 2651364. molecular chaperone, lysosomes and proteasomes
116. Midena E, Bini S, Martini F, Enrica C, Pilotto E, in autophagy-mediated proteolysis in human
Micera A, et al. Changes of aqueous humor muller retinal pigment epithelial cells. J Cell Mol Med.
cells’ biomarkers in human patients affected by dia- 2009;13(9b):3616–31. Epub 2008/11/20. https://doi.
betic macular edema after subthreshold micropulse org/10.1111/j.1582-4934.2008.00577.x. PubMed
laser treatment. Retina. 2018. Epub 2018/10/10. PMID: 19017362; PubMed Central PMCID:
https://doi.org/10.1097/iae.0000000000002356. PMCPMC4516511.
PubMed PMID: 30300267. 124. Ryhanen T, Viiri J, Hyttinen JM, Uusitalo H,
117. Kaarniranta K, Tokarz P, Koskela A, Paterno J, Salminen A, Kaarniranta K. Influence of Hsp90
Blasiak J. Autophagy regulates death of retinal and HDAC inhibition and tubulin acetylation on
pigment epithelium cells in age-related macu- perinuclear protein aggregation in human reti-
lar degeneration. Cell biology and toxicology. nal pigment epithelial cells. J Biomed Biotechnol.
2017;33(2):113–28. Epub 2016/12/03. https:// 2011;2011:798052. Epub 2010/10/29. https://
doi.org/10.1007/s10565-016-9371-8. PubMed doi.org/10.1155/2011/798052. PubMed
PMID: 27900566; PubMed Central PMCID: PMID: 20981255; PubMed Central PMCID:
PMCPMC5325845. PMCPMC2963810.
118. Kaarniranta K, Elo M, Sironen R, Lammi MJ, 125. Mones J, Garcia M, Biarnes M, Lakkaraju A, Ferraro
Goldring MB, Eriksson JE, et al. Hsp70 accumu- L. Drusen Ooze: a novel hypothesis in geographic
lation in chondrocytic cells exposed to high con- atrophy. Ophthalmol Retina. 2017;1:461–73.
tinuous hydrostatic pressure coincides with mRNA 126. Gao J, Liu RT, Cao S, Cui JZ, Wang A, To E, et al.
stabilization rather than transcriptional activation. NLRP3 inflammasome: activation and regulation
Proc Natl Acad Sci USA. 1998;95(5):2319–24. Epub in age-related macular degeneration. Mediators
1998/04/16. PubMed PMID: 9482883; PubMed Inflamm. 2015;2015:690243. Epub 2015/02/24.
Central PMCID: PMCPMC19331. https://doi.org/10.1155/2015/690243. PubMed
119. Kaarniranta K, Salminen A, Eskelinen EL, Kopitz PMID: 25698849; PubMed Central PMCID:
J. Heat shock proteins as gatekeepers of pro- PMCPMC4324923.
teolytic pathways-Implications for age-related 127. Martine P, Chevriaux A, Derangere V, Apetoh L,
macular degeneration (AMD). Ageing Res Rev Garrido C, Ghiringhelli F, et al. HSP70 is a nega-
2009;8(2):128–139. Epub 2009/03/12. PubMed tive regulator of NLRP3 inflammasome activa-
PMID: 19274853. tion. Cell Death Disease. 2019;10(4):256. Epub
120. Buchner J. Supervising the fold: functional principles 2019/03/16. https://doi.org/10.1038/s41419-019-
of molecular chaperones. FASEB J 1996;10(1):10– 1491-7. PubMed PMID: 30874540; PubMed Central
19. Epub 1996/01/01. PubMed PMID: 8566529. PMCID: PMCPMC6420651.
David Kent FRCOphth is a vitreoretinal surgeon prac-

ticing in The Vision Clinic, Kilkenny, Ireland. His research
Jeffrey K. Luttrull, MD, is a vitreoretinal surgeon and interests include the experimental investigation and thera-
clinical researcher practicing in Ventura, California. His peutic use of light to modulate and promote repair in the
interests and publications include medical and surgical aging retina. He has been a member of the RPE Cell biol-
retina, with a special interest in retinal laser therapy. He is ogy section of ARVO since 1996 and currently serves as a
the founder and director of LIGHT: The International director of LIGHT: The International Retinal Laser
Retinal Laser Society. Society.
Choroidal Neovascularization:
Newer Molecules 31
Aamir A. Aziz, Ibrahim Khanani,
Fawwaz A. Siddiqui, Ryan N. Constantine,
and Arshad M. Khanani
31.1 Introduction bevacizumab, a humanized anti-VEGF antibody

aimed at blocking VEGF-A. Bevacizumab ini-
Over the past two decades the treatment for neo- tially entered Phase I clinical trials for cancer.
vascular age-related macular degeneration With results indicating its efficacy and minimal
(NVAMD) has undergone significant improve- toxicity profile, it was subsequently approved for
ments. No longer are patients destined for blind- the co-treatment with chemotherapy of colon
ness, but rather, efficacious therapies are now cancer by the FDA in 2004 [3]. However, increas-
enabling patients to maintain vision as well as ing evidence showing the role of VEGF in ocular
improve visual outcomes following the onset of angiogenesis as noted above led to the off-label
symptoms. By highlighting the development of use of intravenous bevacizumab for the treatment
current therapies for NVAMD, one can gain of NVAMD [4]. Shortly thereafter, off-label
insight into the direction the field is headed in the intravitreal injections of bevacizumab for the
future [1]. treatment of NVAMD were shown to be effective
With the elucidation of direct scientific evi- with minimal systemic side effects [5].
dence supporting the role of vascular endothelial Ranibizumab is also developed by Genentech
growth factor (VEGF) in ocular angiogenesis, (South San Francisco, CA) and received FDA
development and administration of anti-VEGF approval in 2006. Ranibizumab is a monoclonal
therapies aimed at treating NVAMD, was estab- antibody fragment that inhibits angiogenesis by
lished as the primary therapeutic avenue. The inhibiting vascular endothelial growth factor
first antiangiogenic molecule approved by the A. Two pivotal Phase III trials; ANCHOR and
U.S. Food and Drug Administration (FDA) for MARINA established efficacy and safety of
the treatment of ocular neovascularization was ranibizumab in patients with NVAMD [6, 7].
pegaptanib. It is an RNA aptamer that binds A third molecule, aflibercept, developed by
VEGF165 and was shown in Phase II and III clin- Regeneron Pharmaceuticals (Tarrytown, NY)
ical trials to decrease visual loss associated with employs a different anti-VEGF strategy.
NVAMD [2]. Aflibercept functions as a VEGF-trap, acting as a
Prior to the development of pegaptanib, receptor decoy to sequester VEGF. Aflibercept is
Genentech (South San Francisco, CA) developed a recombinant fusion protein consisting of vascu-
lar endothelial growth factor (VEGF)—binding
portions from the extracellular domains of human
A. A. Aziz · I. Khanani · F. A. Siddiqui VEGF receptors 1 and 2 that are fused to the Fc
R. N. Constantine · A. M. Khanani (*) portion of human IgG1 [8]. Aflibercept was
Sierra Eye Associates, Reno, NV, USA

426 A. A. Aziz et al.
approved by the FDA in 2011 following the rates of fluid resolution [11]. This led to the ini-
Phase III results from the two pivotal trials: tiation of pivotal Phase III HAWK and HARRIER
VIEW 1 and VIEW 2. These trials indicated that studies which looked at 1825 patients with
aflibercept dosed every two months was not infe- treatment-naïve NVAMD that were either treated
rior to monthly ranibizumab dosing [3]. with brolucizumab or aflibercept. In HAWK and
Despite the effectiveness of the current thera- HARRIER, patients received three monthly
pies available for patients with NVAMD, there doses of brolucizumab and then received either a
continues to be a need for treatment options with dose every 8 weeks or every 12 weeks based on
increased durability. Achieving this goal is not disease activity assessment. Patients in the
easy and significant effort has been invested in aflibercept arm received three monthly doses and
discovering new efficacious targets and mole- then a dose every 8 weeks per the label. The stud-
cules aimed at decreasing the treatment burden ies met the primary endpoint of non-inferiority of
that patients with NVAMD face. The remainder brolucizumab compared to aflibercept in mean
of this chapter focuses on new molecules that are change in best-corrected visual acuity from base-
in clinical development for the treatment of line to week 48 [12]. Patients gained +6.6 (6 mg)
NVAMD. and +6.1 (3 mg) letters with brolucizumab versus
+6.8 letters with aflibercept in HAWK and +6.9
(6 mg) letters with brolucizumab versus +7.6 let-
31.2 Brolucizumab ters with aflibercept in HARRIER. The other key
endpoints were patients maintained on every
Brolucizumab is a single-chain antibody frag- 12-week dosing till week 48 as well as anatomic
ment with a molecular weight of 26 kDa. This outcomes at week 16 and week 48. Fifty-six per-
molecule is considerably smaller than any other cent of eyes in HAWK and 51% in HARRIER
available anti-VEGF agent [9]. This results in were maintained on every 12-week dosing till
concentrated molar dosing and potent VEGF-A week 48. As far as anatomic outcomes, patients
inhibition. The 6 mg dose of brolucizumab is treated with brolucizumab 6 mg had significantly
equivalent to approximately 11–12 times higher less intraretinal fluid (IRF), subretinal fluid
molar dose than 2 mg aflibercept. Preclinical (SRF), and subretinal pigment epithelium (Sub-
studies have shown that the smaller molecular RPE) fluid at week 16 and week 48 [13].
size results in increased ocular tissue penetration, Brolucizumab treatment resulted in an additional
increasing localized concentrations of the mole- 30–40% of patients with no fluid compared to
cule in deeper levels of the retina. aflibercept at those time points. The overall safety
During animal studies, brolucizumab showed of brolucizumab was comparable to aflibercept.
a fourfold lower systemic exposure in compari- Based on the data from HAWK and HARRIER,
son to ranibizumab while also showing tolerabil- Brolucizumab received FDA approval in the
ity to higher doses and a high affinity to VEGF United States in October 2019.
[10]. Early human studies also exhibited positive
results, where a single dose of brolucizumab was
shown to be more potent and longer lasting than 31.3 Abicipar
a single dose of ranibizumab.
The Phase II (OSPREY) study met its primary Abicipar is a novel designed ankyrin repeat pro-
endpoint and confirmed non-inferiority of brolu- tein (DARPin) which functions by binding to tar-
cizumab to aflibercept in terms of mean change geted proteins with increased precision and
in best-corrected visual acuity (BCVA) from affinity [14]. A new agent functioning against
baseline to weeks 12 and 16. The study also choroidal neovascularization, abicipar pegol tar-
showed that eyes treated with brolucizumab had gets the VEGF-A165 isoform, which is primarily
more stable central subfield reductions, received expressed in humans and associated with patho-
fewer unscheduled treatments, and had higher logic angiogenesis [14]. Abicipar has a small
31 Choroidal Neovascularization: Newer Molecules 427
molecular weight of 32 kDa with a polyethylene ogy to bind VEGF-A on one arm and angiopoi-
tail, which effectively increases the drug’s intra- etin-2 (Ang2) on the other. Preclinical studies
vitreal half-life. Abicipar 2 mg dose is equal to have shown that Ang-2 levels are elevated in
approximately 3.4 times the 2 mg aflibercept patients with AMD [19–22] and blocking Ang-2
dose and demonstrates an increased period of reduces VEGF-induced endothelial barrier break-
effectiveness when compared to ranibizumab at down [23]. Animal studies have also shown that
equal molar doses. When comparing abicipar to combined VEGF-A/Ang-2 inhibition reduces
ranibizumab or aflibercept, abicipar has a 100- CNV lesion number and area, inhibits retinal leu-
fold increased affinity for VEGF-A165. kocyte infiltration, and prolongs anti-leakage
In the Phase II REACH study, abicipar dem- effect. Phase I study results confirmed faricimab
onstrated BCVA and CRT improvements that was safe and well-tolerated with improvements
were similar between abicipar and ranibizumab in BCVA and anatomic parameters for patients
at weeks 16 and 20. This was 8 and 12 weeks with difficult-to-treat neovascular AMD [24],
after the last abicipar injection and 4 weeks after leading to Phase II AVENUE and STAIRWAY
the last ranibizumab injection, confirming a dura- studies.
bility effect [15]. AVENUE confirmed the efficacy and safety of
In the Phase III CEDAR and SEQUOIA stud- faricimab every 4 and 8 weeks in patients with
ies, 1885 treatment-naïve NVAMD eyes were treatment-naïve NVAMD compared to monthly
randomized to abicipar or ranibizumab. Patients doses of ranibizumab. STAIRWAY was designed
received either 0.5 mg ranibizumab monthly, to evaluate the efficacy and durability of farici-
2.0 mg abicipar pegol every 8 weeks after 3 mab in treat-naïve patients with NVAMD. Patients
monthly doses, or 2.0 mg abicipar pegol every were treated with four monthly doses of farici-
12 weeks after doses at baseline, week 4, and mab and then treated every 16 week or every
week 12. The primary endpoint at week 52 of 12 weeks based on disease activity compared to
non-inferiority to ranibizumab was met. 91.2% ranibizumab monthly. The primary endpoint was
and 96% of each abicipar arm lost less than 15 efficacy of faricimab given every 16 weeks and
letters at week 52 compared to baseline [16]. every 12 weeks as assessed by BCVA [24].
Looking at the ocular safety, approximately 15% Prespecified disease activity was performed at
of patients in the abicipar treatment groups expe- week 24, which was 12 weeks after the last farici-
rienced inflammatory events [17]. More informa- mab dose. Sixty-five percent of patients treated
tion is needed to understand these events better, with faricimab had no disease activity 12 weeks
but the current hypothesis is that this could be after the last injection. Vision gains were similar
related to the manufacturing process. The com- between all three treatment groups and were fully
pany is working on a better purification and man- maintained through week 52. Patients treated
ufacturing process. The open-label safety Phase with every 16-week faricimab gained +11.4 chart
III MAPLE study was conducted with the letters from baseline, compared to +10.1 letters
improved manufacturing process in an effort to in 12-week faricimab and +9.6 letters in every
increase safety. Results from MAPLE has shown 4-week ranibizumab group. All three treatment
improvements in these inflammatory event rates regimens showed a similar proportion of patients
at 8.9%, a reduction from the 15% experienced gaining +15 chart letters and avoiding a loss of
by patients in CEDAR and SEQUOIA [18]. more than −15 letters. Anatomic outcomes were
also similar among three groups as evaluated by
change in central subfield thickness as well as
31.4 Faricimab reduction in CNV lesion size from baseline to
week 52. There were no new safety signals and
Faricimab (formerly RG7716) is the first bispe- the overall safety profile of faricimab was similar
cific antibody that is designed for intravitreal use. to ranibizumab. Based on the data from the
It has been developed using CrossMAb technol- AVENUE and STAIRWAY studies, faricimab
holds the potential of extended durability in mean improvement of acuity using squalamine
patients with NVAMD. Therefore, Phase 3 failed expected efficacy results, ending the hope
TENAYA/LUCERNE studies were initiated in for squalamine as a topical therapy for NVAMD.
2019 and are currently ongoing.
31.7 Fovista
31.5 Nesvacumab
Fovista is an aptamer directed against platelet-
Nesvacumab is a fully human monoclonal anti- derived growth factor (anti-PDGF). PDGF func-
body and effectively inactivates Ang2. When tions by recruiting pericytes to neovascular
combining nesvacumab with an anti-VEGF complexes and once there, they supply VEGF
agent, the potential for preventing the pathologi- and additional growth factors necessary for endo-
cal processes of angiogenesis in neovascular thelial cells to build new blood vessels [29]. Both
AMD is greatly increased. cell culture and animal models showed that over-
The Phase I study combined nesvacumab and expression of PDGF resulted in an increased pro-
aflibercept, while evaluating safety and tolerance in liferation of pericytes and tumor growth [30–32].
20 patients with neovascular AMD and DME. The By inhibiting PDGF, Fovista indirectly decreases
reported results exhibited that visual and anatomi- the VEGF load, thus inhibiting neovasculariza-
cal improvement at all dose levels, with no reported tion within the retina.
dose-limiting toxicities, ocular inflammation, or After positive Phase II trials, Fovista was
any unexpected systemic effects [25]. tested in combination with ranibizumab com-
The Phase II study, ONYX, evaluated patients pared to ranibizumab monotherapy in two sepa-
with wet AMD, using a combination of nes- rate Phase III trials. No significant improvements
vacumab and aflibercept. The results of this study in visual acuity were reported with the combina-
did not provide sufficient differentiation between tion therapy when compared to patients treated
the combination therapy and independent afliber- with ranibizumab monotherapy [33]. Therefore,
cept to support a Phase III study [26]. the Fovista development program has been
terminated.
31.6 Squalamine Drops

31.8 KSI-301
Developed by OHR Pharmaceutical, squalamine
is a topical antiangiogenic ophthalmic eye drop KSI-301 is an antibody biopolymer conjugate
that functions by inhibiting multiple growth fac- developed by Kodiak Sciences [34]. KSI-301
tors, such as VEGF, platelet-derived growth fac- also targets VEGF and in vitro assays demon-
tor (PDGF) and basic fibroblast growth factor strate that the molecule has increased potency
(bFGF) [27, 28]. when compared to bevacizumab, ranibizumab,
In the Phase III study, squalamine drops were and aflibercept. The molecular weight is 950 kDa,
given topically in conjunction with ranibizumab as a result of its phosphorylcholine biopolymer.
injections (n = 119), compared to an independent This biopolymer leads to increased ocular tissue
treatment of just ranibizumab (n = 118). Patients bioavailability in both the retina and the choroid
with the squalamine/ranibizumab combination when compared to aflibercept (approximately a
experienced an average of an +8.33 chart letter 30-day increase in both cases).
improvement from baseline visual acuity, while The Phase 1a trial met safety outcomes and
the ranibizumab monotherapy patients experi- improvements in vision and retinal thickness in
enced a +10.58 chart letter improvement. heavily previously treated patients with diabetic
Although there was no difference in the safety macular edema. All dose levels were
profile of both therapy groups, the difference in well-tolerated without any drug-related adverse
events. Twelve weeks after the single dose, cohorts showed that sunitinib maintained BCVA
median vision gain from baseline was +9 letters, and CST that was equivalent to patients receiving
and median OCT improvement of −121 microns standard of care intravitreal anti-VEGF agents
were observed across all three dose levels. with 6-week dosing. The majority of the adverse
Phase 1b trials recently began in 2019 with events noted in the Phase 1/2a study were the
safety, efficacy and durability as the key end- result of the migration of the bio-absorbable par-
points. The 1b trials tested multiple doses of KSI- ticles into the anterior chamber due to incomplete
301 in treatment-naïve patients with NVAMD, aggregation into a depot. These events were
diabetic macular edema, and retinal vein occlu- described as self-limited and reversible as the
sions. Interim data from the trial revealed that the formulation biodegrades within the eye and at
wAMD cohort gained an average of +8 letters month 6 there were no observed sequelae.
from baseline and showed a −96 micron improve- A new optimized version of sunitinib has been
ment in median retinal thickness [35]. The posi- manufactured to eliminate particle dispersion for
tive results influenced both an extension of the 1b Phase 2b NVAMD study, which was initiated in
trial and the Phase 2 trial DAZZLE, which began Q3 of 2019. The study ALTISSIMO compares
in Q3 of 2019. DAZZLE is a randomized trial 1 mg and 2 mg doses of sunitinib biannually to a
comparing KSI-301 to aflibercept and seeks to 2 mg dose of aflibercept administered every 2
extend injection intervals for KSI-301 after 3 months. Sunitinib is also currently being tested
loading doses to up to 20 weeks, greatly reducing for its safety and efficacy in RVO and DME,
the treatment burden of afflicted patients. along with NVAMD.
31.9 Sunitinib 31.10 Conbercept
Sunitinib malate is a multiple receptor tyrosine Conbercept is a recombinant fusion protein tar-
kinase inhibitor that is approved as an oral agent geting VEGF-R-1 and -2 [37]. Conbercept is
for solid tumors including renal cell carcinoma, composed of the second IgG domain of
imatinib-resistant gastrointestinal stromal tumor, VEGF-R-1 and the third and fourth domains of
and pancreatic neuroendocrine tumors. Whereas VEGF-R-2 to the Fc region of human IgG. The
bevacizumab, ranibizumab, and aflibercept all agent has a molecular weight of 143 kDa and in-
function to inhibit VEGF-receptors 1 and 2 vitro assays revealed that the drug has a 30-fold
(VEGFR), sunitinib selectively inhibits increased affinity for VEGF than ranibizumab or
VEGFR-1, -2, -3, platelet-derived growth factor bevacizumab. The biochemical functionality of
receptors (PDGFR-α, -β), stem cell growth factor conbercept allows it to target both VEGF-A, -B,
receptor (KIT), colony-stimulating factor recep- and -C, along with placental derived growth fac-
tor (CSFR-1), and Fms-related tyrosine kinase tors (PIGF).
receptor (FLT3). Sunitinib is formulated as In the trial AURORA, patients were random-
injectable intravitreal poly-(lactic-co-glycolic ized 1:1 to either a 0.5 mg dose or 2.0 dose. After
acid) (PLGA) molecules that aggregate to form a three monthly doses, patients were randomized to
depot in the inferior vitreous out of the visual either a prn treatment schedule or continued
axis. The depot allows sunitinib to be regarded monthly, with no changes to their initial dosage
with the potential of only two injections per year, randomization. At month 3, mean BCVA
a massive decrease for patients receiving monthly improvement in the 0.5 mg cohort was +8.97 let-
injections. ters, while the 2.0 mg cohort had an increase of
Preclinical animal studies support biannual +10.43 letters. At the month 12 evaluation, mean
dosing following a single intravitreal injection of BCVA improvement was +14.31 letters for the
sunitinib [36]. A Phase 1/2a study of a single 0.5 mg prn cohort and +9.31 for the 0.5 mg
injection of sunitinib with dose-escalating monthly cohort. At the same checkpoint, BCVA
improvements for the 2.0 mg cohort were +12.42 31.11 OPT-302

for the 2.0 mg prn cohort, and +15.43 for the
2.0 mg monthly cohort, respectively. Statistically OPT-302 is a soluble form of VEGF-R-3 and
significant reductions in central retinal thickness functions as a “trap” molecule that blocks
(CRT) were also found. At month 12, CRT reduc- VEGF-C and VEGF-D [39]. When an anti-VEGF
tion was −119.8 microns for the 0.5 mg prn agent such as ranibizumab blocks VEGF-A, the
cohort, −129.7 microns for the 0.5 mg monthly proteins VEGF-C and VEGF-D are noted to be
cohort, −152.1 microns for the 2.0 mg prn cohort, upregulated. OPT-302 is designed to work in
and −170.8 microns for the 2.0 mg monthly conjunction with an anti-VEGF-A agent and pro-
cohort. Two serious adverse events were recorded vides a more complete inhibition of the four
in this trial, a case of endophthalmitis associated classes of VEGFs by also inhibiting
to the injection procedure, and cataract develop- VEGF-C/-D.
ment associated to the study drug. OPT-302 recently completed its Phase 2b trial
The subsequent trial PHOENIX randomized and met its primary endpoints [40]. Combined
patients 2:1 to two cohorts: treatment and delayed OPT-302 and ranibizumab treatment had
treatment [38]. In the treatment cohort, patients increased vision gains compared to ranibizumab
were given monthly injections of conbercept monotherapy. Along with these results, OPT-302/
0.5 mg for 3 months, then extended to quarterly ranibizumab also showed anatomical improve-
injections of the same dosage. In the delayed ments, including a reduced central subfield thick-
treatment cohort, patients received sham injec- ness (CST), a decrease in choroidal
tions monthly for 3 months, then moved to con- neovascularization (CNV) lesion size and
bercept 0.5 mg injections monthly for 3 months, decreases in intra-/subretinal fluid. Patients in the
then extended to quarterly injections at the 2.0 mg OPT-302/ranibizumab therapy gained
0.5 mg dosage. By month 12, the treatment group +14.2 letters, while the low-dosage cohort of
showed a +9.9 letter gain in BCVA, while the 0.5 mg OPT-302/ranibizumab gained +9.4 let-
delayed treatment group showed a +8.8 letter ters. The ranibizumab monotherapy group
improvement. Both groups showed reductions in showed an improvement of +10.8 letters. By
CRT with no statistical significance between the week 24, the 2.0 mg OPT-302/ranibizumab
two cohorts. No ocular serious adverse events cohort showed a mean CST reduction of −147
were noted in relation to the study drug. microns, from baseline. The ranibizumab mono-
Currently, conbercept is being studied at the therapy showed a −134 micron reduction at the
global level in two masked, randomized studies same checkpoint. OPT-302 is currently in prepa-
known as PANDA-1 and PANDA-2. Both trials ration for Phase 3 studies as the Phase 2a study
have three cohorts in a 1:1:1 randomization, of for DME nears completion.
0.5 mg conbercept every 8 weeks, 1.0 mg conber-
cept every 12 weeks, and 2.0 mg aflibercept every
8 weeks. All three cohorts received 3 monthly 31.12 Conclusions
loading doses and were then extended to their
respective intervals. PANDA-2 differs at week Despite currently available efficacious vision
40, where all patients then begin a capped (max saving therapies for NVAMD, including bevaci-
interval of 16 weeks) prn treatment of their zumab, ranibizumab, and aflibercept, significant
cohort. Both studies are ongoing and follow efforts continue to be made toward developing
patients through 96 weeks, with the hopes that novel therapies for patients suffering from
conbercept proves itself as non-inferior to NVAMD. The current treatment regime for those
aflibercept. afflicted with NVAMD, typically the aging popu-
lation, is burdensome. To maintain visual acuity, related macular degeneration [Internet]. Alcon. Alcon;
2015 [cited 2019Feb13]. Available from https://www.
Q4 weekly to Q8 weekly office visits are the alcon.com/media-release/positive-phase-ii-data-high-
norm and for an elderly patient population, can lights-benefits-alcons-rth258-patients-neovascular-
pose several challenges including cost, transpor- wet-age
tation, and postinjection pain and discomfort. As 10. Dugel P. Results of ESBA 1008, a single-chain
antibody fragment for the treatment of neovascular
outlined in this chapter, the focus is to develop AMD. Paper presented at: Annual meeting of the
new molecules with increased durability as com- American Society of Retina Specialists; San Diego,
pared to current therapies. These longer lasting CA; August 9–13, 2014.
treatments for NVAMD carry the hope of decreas- 11. Dugel PU, Jaffe GJ, Sallstig P, Warburton J,
Weichselberger A, Wieland M, Singerman
ing treatment burden and improving visual acuity L. Brolucizumab versus aflibercept in partici-
outcomes in the real world. pants with neovascular age-related macular degen-
eration: a randomized trial. Ophthalmology. 2017
Sep;124(9):1296–1304. https://doi.org/10.1016/j.
ophtha.2017.03.057. Epub 2017 May 24.
References 12. Dugel P.U., Koh A., Ogura Y., Jaffe G., Schmidt-
Erfurth U., Brown D. HAWK and HARRIER: phase
1. Holz FG, Tadayoni R, Beatty S. Multi-country real- 3, multicenter, randomized, double-masked trials
life experience of anti-vascular endothelial growth of brolucizumab for neovascular age-related macu-
factor therapy for wet age-related macular degenera- lar degeneration. Ophthalmology. 2019 Apr 12. pii:
tion. Br J Ophthalmol. 2015;99:220–6. S0161-6420(18)33018-5. https://doi.org/10.1016/j.
2. Gragoudas ES, Adamis AP, Cunningham ET Jr, ophtha.2019.04.017. [Epub ahead of print].
Feinsod M, Guyer DR. VEGF inhibition study in ocu- 13. Novartis. Two-year data for Novartis broluci-
lar neovascularization clinical trial group pegaptanib zumab reaffirm superiority versus aflibercept
for neovascular age-related macular degeneration. N in reducing retinal fluid in patients with nAMD
Engl J Med. 2004;351:2805–16. [Press Release] (2018 Oct 27) [cited 2019Feb13].
3. Kim LA, D’Amore PA. A brief history of anti-VEGF Available from https://novartis.gcs-web.com/
for the treatment of ocular angiogenesis. Am J Pathol. Two - y e a r- d a t a - f o r- N ova r t i s - b r o l u c i z u m a b -
2012 Aug;181(2):376–9. reaffirm-superiority-versus-aflibercept-in-reducing-
4. Michels S, Rosenfeld PJ, Puliafito CA, Marcus retinal-fluid-in-patients-with-nAMD
EN, Venkatraman AS. Systemic bevacizumab 14. Gerard A. Rodrigues, Matthew Mason, Lori-Ann
(Avastin) therapy for neovascular age-related macu- Christie, Candice Hansen, Lisa M. Hernandez,
lar degeneration: twelve-week results of an uncon- James Burke, Keith A. Luhrs, Thomas C. Hohman;
trolled open-label clinical study. Ophthalmology. Functional Characterization of Abicipar-Pegol, an
2005;112:1035–47. Anti-VEGF DARPin Therapeutic That Potently
5. Rosenfeld PJ, Moshfeghi AA, Puliafito CA. Optical Inhibits Angiogenesis and Vascular Permeability.
coherence tomography findings after an intravitreal Invest Ophthalmol Vis Sci. 2018;59(15):5836–5846.
injection of bevacizumab (Avastin) for neovascular https://doi.org/10.1167/iovs.18-25307
age-related macular degeneration. Ophthalmic Surg 15. Callanan D, Kunimoto D, Maturi RK, Patel SS,
Lasers Imaging. 2005;36:331–5. Staurenghi G, Wolf S, Cheetham JK, Hohman TC,
6. Rosenfeld PJ, Brown DM, Heier JS, Boyer DS, Kaiser Kim K, López FJ, Schneider S, and for the REACH
PK, Chung CY, Kim RY. MARINA study group Study Group. Double-masked, randomized, phase 2
ranibizumab for neovascular age-related macular evaluation of abicipar pegol (an anti-VEGF DARPin
degeneration. N Engl J Med. 2006;355:1419–31. therapeutic) in neovascular age-related macular
7. Brown DM, Kaiser PK, Michels M, Soubrane G, degeneration. J Ocular Pharmacol Therapeutics. Dec
Heier JS, Kim RY, Sy JP, Schneider S. ANCHOR 2018. https://doi.org/10.1089/jop.2018.0062 [Epub
study group ranibizumab versus verteporfin for neo- ahead of print].
vascular age-related macular degeneration. N Engl J 16. Khurana R. Phase 3 CEDAR and SEQUOIA clini-
Med. 2006;355:1432–44. cal trial results. Presentation at; 2018; American
8. Semeraro F, Morescalchi F, Duse S, Parmeggiani Academy of Ophthalmology, Chicago, IL.
F, Gambicorti E, Costagliola C. Aflibercept in wet 17. Allergan and molecular partners announce two
AMD: specific role and optimal use. Drug Des Devel positive phase 3 clinical trials for abicipar pegol 8 a
Ther. 2013;7:711–722. Published 2013 Aug 5. https:// [Internet]. Allergan. Allergan; 2018 [cited 2019 Feb
doi.org/10.2147/DDDT.S40215 13]. Available from https://www.allergan.com/news/
9. Positive phase II data highlights benefits of Alcon’s news/thomson-reuters/allergan-and-molecular-part-
RTH258 for patients with neovascular (wet) age- ners-announce-two-posit.aspx
18. Molecular Partners. Allergan and Molecular Partners 28. OHR-102 Squalamine [Internet]. OHR
Announce Topline Safety Results from MAPLE Pharmaceutical, Inc. (OHRP). OHR Pharmaceutical,
Study of Abicipar pegol [Press Release] (2019 Apr Inc.; 2018 [cited 2019 Mar 12]. Available from https://
2) [Cited 2019 June 02]. Available from https://www. www.ohrpharmaceutical.com/product-portfolio/
molecularpartners.com/allergan-and-molecular-part- squalamine
ners-announce-topline-safety-results-from-maple- 29. Reinmuth N, Liu W, Jung YD, et al. Induction of
study-of-abicipar-pegol/ VEGF in perivascular cells defines a potential para-
19. Thurston G, Daly C. The complex role of angiopoi- crine mechanism for endothelial cell survival. FASEB
etin-2 in the angiopoietin-tie signaling pathway. Cold J. 2001;15(7):1239–41.
Spring Harbor Perspectives in Medicine. 2012;2(9). 30. Benjamin LE, Hemo I, Keshet E. A plasticity window
20. Khanani AM. The Tie2 activator AKB-9778, used for blood vessel remodelling is defined by pericyte
in combination with ranibizumab, enhances reduc- coverage of the preformed endothelial network and
tion of diabetic macular edema compared to ranibi- is regulated by PDGF-B and VEGF. Development.
zumab monotherapy. Invest Ophthalmol Vis Sci. 1998;125(9):1591–8.
2016;57(12); no pagination specified. http://iovs. 31. Lindahl P, Johansson BR, Levéen P, Betsholtz
arvojournals.org/article.aspx?articleid=2560785 C. Pericyte loss and microaneurysm forma-
Accessed May 15, 2017. tion in PDGF-B-deficient mice. Science.
21. Campochiaro PA, Khanani AM. Enhanced ben- 1997;277(5323):242–5.
efit in diabetic macular edema from AKB-9778 32. Furuhashi M, Sjöblom T, Abramsson A, et al. Platelet-
Tie2 activation combined with vascular endothe- derived growth factor production by B16 melanoma
lial growth factor suppression. Ophthalmology. cells leads to increased pericyte abundance in tumors
2016;123:1722–30. and an associated increase in tumor growth rate.
22. Chakravarthy U, Bailey C, Brown D, Campchiaro P, Cancer Res. 2004;64(8):2725–33.
et al. Phase I trial of anti-vascular endothelial growth 33. Elvidge S. Opthotech’s Fovista crashes out in wet
factor/anti-angiopoietin 2 bispecific antibody RG7716 AMD [Internet]. BioPharma Dive. Industry Dive;
for neovascular age-related macular degeneration. 2017 [cited 2019 Mar 14]. Available from https://
Ophthalmol Retina [Internet]. 2017 Nov–Dec [cited www.biopharmadive.com/news/opthotech-fovista-
2019 Jun 1]: 1(6); 474–485. Available from https:// phase-3-failure-setback-novartis/449248/
doi.org/10.1016/j.oret.2017.03.003. 34. Do D. Phase 1 first-in-human study of KSI-301: a
23. Regula JT, Lundh von Leithner P, Foxton R, Barathi novel anti-VEGF antibody biopolymer conjugate
VA, Cheung CM. Targeting key angiogenic path- with extended durability. Presentation presented at:
ways with a bispecific CrossMAb optimized for 2019; Angiogenesis, Exudation, and Degeneration
neovascular eye diseases. EMBO Mol Med. 2016 Symposium, University of Miami – Miller School of
Nov 2;8(11):1265–88. https://doi.org/10.15252/ Medicine; February 9, 2019.
emmm.201505889. Print 2016 Nov. 35. Antibody biopolymer conjugates: a novel scientific
24. Khanani AM, et al. Simultaneous inhibition of VEGF approach and platform for extended-durability reti-
and Ang-2 with faricimab in neovascular AMD: nal medicines – first results from a phase 1b Proof of
STAIRWAY phase 2 results. Presented at the 2018 Concept Study of KSI-301, an anti-VEGF ABC” pre-
American Academy of Ophthalmology (AAO) Annual sented by Pravin U. Dugel, MD at American Society
Meeting; 2018 Oct 26; Chicago, United States. of Retina Specialists 2019 on July 27, 2019.
25. Brown DM. Phase 1 study of combination therapy 36. Peterson WM, Yang M. GB-102 for Wet AMD: a
with nesvacumab and aflibercept for neovascular novel injectable formulation that safely delivers active
AMD and diabetic macular edema. Presented at: levels of sunitinib to the retina and RPE/choroid for
American Academy of Ophthalmology annual meet- over four months. Poster presented at; ARVO; 2016
ing; Oct. 14–18, 2016; Chicago, IL. May 1–5; Seattle, WA, USA.
26. Regeneron Provides Update on EYLEA® (afliber- 37. Sharma, Priya, Jeffrey S Heier. The Phase 3 clini-
cept) Injection and Nesvacumab (Ang2 Antibody) cal trial of conbercept for exudative AMD. Retinal
Combination Program [Internet]. Regeneron Physician, PentaVision, Inc., 1 Apr. 2019.
Pharmaceuticals Inc. Regeneron Pharmaceuticals 38. Liu K, et al. Conbercept for treatment of neovascular
Inc.; 2017 [cited 2019 Mar 1]. Available from https:// age-related macular degeneration: results of the ran-
newsroom.regeneron.com/news-releases/news- domized phase 3 PHOENIX study. Am J Ophthalmol.
release-details/regeneron-provides-update-eylear- 2019;000(197):156–67.
aflibercept-injection-and 39. Opthea presents phase 1/2a data of OPT-302 in wet
27. OHR Pharmaceutical provides update on ongoing AMD at the Retina Society 2018 Annual Meeting.
squalamine clinical trial in wet-AMD [press release]. GlobeNewswire News Room, “GlobeNewswire”, 14
New York, NY: OHR Pharmaceutical; April 10, 2017. Sept. 2018,
Available from http://www.ohrpharmaceutical.com/ 40. h t t p s : / / w w w . o p t h e a . c o m / w p - c o n t e n t /
media-center/press-releases/detail/482/ohr-pharma- uploads/2019/08/OPT.-2019_08_07_ASX_Ph2b-
ceutical-provides-update-on-ongoing-squalamine. wAMD-Results_Final.pdf
Aamir A. Aziz, B.S. is a clinical research coordinator

under Arshad Khanani, M.D., M.A. at Sierra Eye Ryan N. Constantine, M.D., Ph.D. completed his M.D.,
Associates in Reno, NV. His degree is in Biochemistry Ph.D. and intern year at the University of Utah. He then
and Molecular Biology from the University of Nevada, completed his ophthalmology residency at the highly
Reno, for which he completed a research thesis focused regarded Duke Eye Center in Durham, North Carolina. After
on muscle fatigue regulated by potassium channels in finishing his vitreoretinal fellowship at Sierra Eye Associates
peripheral Schwann cells. Aamir is currently pursuing a in Reno, Nevada, Dr. Constantine has accepted a position as
career in medicine as a retina surgeon. a retina surgeon with the Orion Eye Center in Bend, Oregon.
Ibrahim Khanani is an incoming freshman at Duke

University studying computer science. His interests lie in
artificial intelligence and applications of machine Arshad M. Khanani, M.D., M.A. is a fellowship-trained
learning. vitreoretinal specialist and is certified by the American Board
of Ophthalmology. As an undergraduate, Dr. Khanani was
honored twice with the Howard Hughes Medical Institute
Research Award. During his medical training, he received
several research awards and designed multiple prospective
clinical trials, leading to publications in major ophthalmol-
ogy journals. Due to his strong interest in clinical research,
Dr. Khanani founded the clinical research section at Sierra
Eye Associates. He has been a principal investigator for over
50 clinical trials and has been a top enroller in the country for
multiple Phase 1–3 trials. He also serves as a member of
clinical trial steering committees and scientific advisory
boards for multiple companies. His articles have been pub-
lished in top ophthalmology journals. Dr. Khanani has also
presented his work at major ophthalmology meetings world-
Fawwaz A. Siddiqui, B.S. currently works as a Clinical wide and has been invited multiple times as a guest speaker
Research Coordinator for Dr. Arshad Khanani at Sierra nationally and internationally. Dr. Khanani has received
Eye Associates. Born and raised in Reno, Nevada, Fawwaz numerous awards of distinction including the Patients’
graduated with a Bachelor’s in Biochemistry and Choice Award, the Compassionate Doctor Recognition
Molecular Biology from the University of Nevada-Reno Award, and the Top 10 Doctor—State Award. He has been
in Spring 2019. Fawwaz is pursuing a career in medicine named in Marquis Who’s Who in the World and has received
and research. the Albert Nelson Marquis Lifetime Achievement Award. Dr.
Khanani has also received the Honor Award and the Senior
Honor Award from the American Society of Retina
Specialists for his contributions. He has also been consis-
tently named one of America’s Top Ophthalmologists and
has also been included in The Leading Physicians of the
World publication. In 2019, he received the Nevada Business
Magazine “Healthcare Heroes—Physician of the Year”
award for his continued dedication to the field of
ophthalmology.

Choroidal Neovascularization

Uploaded by

Copyright:

Available Formats

Choroidal Neovascularization

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Choroidal Neovascularization

Uploaded by

Copyright:

Available Formats

Choroidal

ISBN 978-981-15-2212-3 ISBN 978-981-15-2213-0 (eBook)

© Springer Nature Singapore Pte Ltd. 2020

research is also directed to block the complement factors, and lampalizumab

Pittsburgh, PA Jay Chhablani

1 Choroidal Neovascular Membrane: Historical Perspectives ���������� 1

Part I Basic Science

2 Pathogenesis of Choroidal Neovascularization �������������������������������� 7

Part II Clinical Entities

7 Neovascular AMD: Clinical Features and Imaging������������������������ 73

12 Dystrophy-Related Choroidal Neovascularization ���������������������� 139

Part III Clinical Trials

20 Clinical Trials Related to Choroidal Neovascularization

Part IV Therapy and Rehabilitation

24 Anti-Vascular Endothelial Growth Molecules������������������������������ 331

Jay Chhablani is a Vitreo-Retina Specialist at

1.1 Introduction 1.2 Historical Perspectives

A. Agarwal (*) · K. Bazgain

© Springer Nature Singapore Pte Ltd. 2020 1

1.6 Summary 9. Junius P, Kuhnt H. Die scheibenfo¨rmige Entargung

Krinjeela Bazgain is currently pursuing her M.Ch. in

© Springer Nature Singapore Pte Ltd. 2020 7

inhibits ischemia-induced retinopathy, VEGF- brane, and by releasing cytokines, inflammatory

References sion characteristics in fellow eyes of patients with uni-

© Springer Nature Singapore Pte Ltd. 2020 15

CNV are usually related to breaks in RPE or 3.2.1.1 Presumed Ocular

ity of the cases are diagnosed clinically and CNV

Fig. 3.10 Sarcoid

3.3.2 Myopic Neovascular Choroidal neovascularization in high myopia

3.3.4 Polypoidal Choroidal Histopathology has shown that angioid streaks

Fig. 3.12 Choroidal

cian in the National University of Córdoba, Argentina

Evangelina Esposito , M.D., ChM., is an ophthalmolo-

Julio A. Urrets-Zavalia, M.D., Ph.D., is a vitreo-­retinal

© Springer Nature Singapore Pte Ltd. 2020 37

n­ecessary to recruit monocyte and CX3CR1 major scavenger of oxidative stress. In 3 of 30

PAS Oil Red O vWf ED1

subretinal hemorrhage and exudate which are References

Tapas C. Nag and Sneha Gupta

5.1 Introduction The initial course of the disease, called dry

© Springer Nature Singapore Pte Ltd. 2020 47

5.1.1 Choroidal Neovascularization retinal pigment epithelium (subretinal pigment

Fig. 5.1 Light

Fig. 5.2 Light

The unusual build-up of drusen in the macula is a

Fig. 5.4 A capillary of

Fig. 5.6 Electron

breaks in BM in age-related macular degenera- also antiangiogenic molecules like pigment

Dr. TC Nag is a professor in the Department of Anatomy,

© Springer Nature Singapore Pte Ltd. 2020 57

Table 6.1 Susceptibility genes for age-related macular

Table 6.2 Genetic associations to polypoidal choroidal vasculopathy

Since CNV size determines scotoma size in 6.7 Genes Associated

Table 6.9 Susceptibility genes for myopia 1

Many myopia susceptibility genes were expressed 6.10 Genes Associated

Table 6.10 Susceptibility genes for myopia 2

Kenji Yamashiro received MD from Kyoto University

7.1 Introduction patients manifest the neovascular form of the

© Springer Nature Singapore Pte Ltd. 2020 73

phenotype of neovascularization, which origi-

Pittsburgh, PA Jay Chhablani

1 Choroidal Neovascular Membrane: Historical Perspectives �� 1

Part I Basic Science

2 Pathogenesis of Choroidal Neovascularization �� 7

Part II Clinical Entities

7 Neovascular AMD: Clinical Features and Imaging�� 73

12 Dystrophy-Related Choroidal Neovascularization �� 139

Part III Clinical Trials

20 Clinical Trials Related to Choroidal Neovascularization

Part IV Therapy and Rehabilitation

24 Anti-Vascular Endothelial Growth Molecules�� 331

1.1 Introduction 1.2 Historical Perspectives

1.6 Summary 9. Junius P, Kuhnt H. Die scheibenfo¨rmige Entargung

CNV are usually related to breaks in RPE or 3.2.1.1 Presumed Ocular

3.3.2 Myopic Neovascular Choroidal neovascularization in high myopia

3.3.4 Polypoidal Choroidal Histopathology has shown that angioid streaks

Julio A. Urrets-Zavalia, M.D., Ph.D., is a vitreo-retinal

necessary to recruit monocyte and CX3CR1 major scavenger of oxidative stress. In 3 of 30

5.1 Introduction The initial course of the disease, called dry

5.1.1 Choroidal Neovascularization retinal pigment epithelium (subretinal pigment

Since CNV size determines scotoma size in 6.7 Genes Associated

Many myopia susceptibility genes were expressed 6.10 Genes Associated

7.1 Introduction patients manifest the neovascular form of the

Neovascular AMD presents clinically by typical 7.2.1 Pigment Epithelial

7.3.1 Fluorescein Angiography 7.3.1.1 Classic CNV

7.3.3.1 Type 1 Neovascular Membrane by OCT-A. Grossniklaus and Green hypothesized

to high reflectivity that corresponds to the BVN/ 7.5 Optical Coherence

8.1 he Story of Anti-VEGF

9.3 Fluorescein Angiography 9.4 Optical Coherence

9.7 Treatment removal of CNV: visual increase in 39%, stable

9.7.5 rognostic Factors Associated

10.1 Introduction logical changes, focally. The choroidal thickness

of a detectable core vessel may represent a pro- 10.2.4.3 FFA

10.2.4.4 ICGA 10.2.5.1 Anti-VEGF therapy

11.1 Introduction is yet to be elucidated, and there are mainly two

seen, often associated with granulomatous 11.3.5 Optical Coherence

Fundus autofluorescence (FAF) may also be 11.3.6 Optical Coherence

There is evidence to suggest that inflammatory 11.5.5 Immunosuppressive

12.1 Introduction related to this topic derives from case series or

13.1 Angioid Streaks Secondary choroidal neovascularization

13.2.2 Ocular Imaging Findings show a higher prevalence of BM undulations,

13.2.2.2 eau d’Orange and Reticular

13.3.2 Bevacizumab as a second-line treatment option. To date, this

13.4 Summary • Angioid streaks have a complex pathogenesis,

14.1 Introduction tion [5]. Because it affects younger patients in

14.5.4 Bevacizumab resolution of pigment epithelial detachment

Similar findings in a retrospective compara- 14.6 Key Learning Points

15.1 Introduction vasculature in origin, but as newer imaging

15.3 Diagnostic Imaging 15.4 ssociated Systemic Medical

15.5 Subretinal to that of retinal angiomatous proliferation lesions

16.1 Introduction event can happen either concomitantly or after a

improvement in 5 eyes with peripapillary CNVM 16.12 Conclusions