Strahlenther Onkol

DOI 10.1007/s00066-015-0837-z

O r i g i n a l A rt i c l e

Long-term side effects of radiotherapy for pediatric localized

neuroblastoma
Results from clinical trials NB90 and NB94

Anne Ducassou · Marion Gambart · Caroline Munzer · Laetitia Padovani · Christian Carrie ·
Daphne Haas-Kogan · Valérie Bernier-Chastagner · Charlotte Demoor · Line Claude · Sylvie Helfre ·
Stéphanie Bolle · Julie Leseur · Aymeri Huchet · Hervé Rubie · Dominique Valteau-Couanet ·
Gudrun Schleiermacher · Carole Coze · Anne-Sophie Defachelles · Aurélien Marabelle · Stéphane Ducassou ·
Christine Devalck · Virginie Gandemer · Martine Munzer · Anne Laprie · On behalf of the Neuroblastoma study
group and radiotherapy group of the French Society of Children with Cancer (SFCE)

Received: 25 November 2014 / Accepted: 25 March 2015

© Springer-Verlag Berlin Heidelberg 2015

Abstract localized NB within two French prospective clinical trials:

Introduction Neuroblastoma (NB) is the most frequent indi- NB90 and NB94.
cation for extracranial pediatric radiotherapy. As long-term Patients and methods From 1990–2000, 610 children were
survival of high-risk localized NB has greatly improved, enrolled. Among these, 35 were treated with induction che-
we reviewed treatment-related late toxicities in pediatric motherapy, surgery, and RT. The recommended RT dose
patients who received postoperative radiotherapy (RT) for

A. Ducassou, M.D. () C. Demoor, M.D.

Institut Claudius Regaud, Institut universitaire du cancer Département de radiothérapie, Centre René Gauducheau,
Toulouse Oncopole, 1 avenue Irène Joliot-Curie; IUCT-O, 31059 Nantes, France
Cedex 9 Toulouse, France
e-mail: ducassou.anne@iuct-oncopole.fr S. Helfre, M.D.
Département de radiothérapie, Institut Curie,
A. Ducassou, M.D.  · A. Laprie, M.D., PhD Paris, France
Département de radiothérapie, Institut Claudius Regaud, 31052
Toulouse, France S. Bolle, M.D.
Département de radiothérapie, Institut Gustave Roussy,
M. Gambart, M.D. · C. Munzer, MSc.Res. · H. Rubie, M.D. Villejuif, France
Département d’onco-hématologie pédiatrique, Hôpital des
Enfants, J. Leseur, M.D.
Toulouse, France Département de radiothérapie, Centre Eugène Marquis,
Rennes, France
L. Padovani, M.D., Ph.D.
Département de radiothérapie, Centre Hospitalier et A. Huchet, M.D.
Universitaire, Département de radiothérapie, Centre Hospitalier et
Marseille, France Universitaire,
Bordeaux, France
C. Carrie, M.D. · L. Claude, M.D.
Département de radiothérapie, Centre Léon Bérard, IHOP, D. Valteau-Couanet, M.D., PhD.
Lyon, France Département d’onco-hématologie pédiatrique, Institut Gustave
Roussy,
D. Haas-Kogan, M.D. Villejuif, France
Department of Radiation Oncology, University of California,
San Francisco, USA G. Schleiermacher, M.D., Ph.D.
Département d’onco-hématologie pédiatrique, Institut Curie,
V. Bernier-Chastagner, M.D. Paris, France
Département de radiothérapie, Centre Alexis Vautrin,
Nancy, France

13
2 A. Ducassou et al.

C. Coze, M.D., Ph.D. C. Devalck, M.D.

Département d’hémato-oncologie pédiatrique, Aix-Marseille Département d’onco-hématologie pédiatrique,
Univ et APHM, Hôpital d’Enfants de la Timone, Hôpital des enfants Reine Fabiola,
Marseille, France Bruxelles, Belgique

A.-S. Defachelles, M.D. V. Gandemer, M.D.

Département d’onco-hématologie pédiatrique, Centre Oscar Département d’onco-hématologie pédiatrique, Centre hospitalier,
Lambret, Rennes, France
Lille, France
M. Munzer, M.D.
A. Marabelle, M.D. Département d’onco-hématologie pédiatrique, Centre hospitalier,
Département d’onco-hématologie pédiatrique, IHOP, Reims, France
Lyon, France
A. Laprie, M.D., PhD.
S. Ducassou, M.D. Université Toulouse III Paul Sabatier,
Département d’onco-hématologie pédiatrique, 31000 Toulouse, France
Hôpital des Enfants,
Bordeaux, France A. Laprie, M.D., PhD.
INSERM, UMR825, 24 F-31059 Toulouse, France

was 24 Gy at ≤ 2 years, 34 Gy at > 2 years, ± a 5 Gy boost Patienten und Methoden Von 1990–2000 wurden 610 Kin-
in both age groups. der eingeschlossen. Von diesen wurden 35 mit Chemothera-
Results The 22 patients still alive after 5 years were ana- pie, Chirurgie und RT behandelt. Die empfohlene Bestrah-
lyzed. The median follow-up time was 14 years (range lungsdosis war 24 Gy bei ≤ 2 Jahren, 34 Gy bei > 2 Jahren,
5–21 years). Late effects after therapy occurred in 73 % of ± 5-Gy-Boost in beiden Altersgruppen.
patients (16/22), within the RT field for 50 % (11/22). The Ergebnisse Die 22 Langzeitüberlebenden (≥ 5 Jahre) wur-
most frequent in-field effects were musculoskeletal abnor- den analysiert. Die mediane Nachbeobachtungszeit betrug
malities (n = 7) that occurred only with doses > 31 Gy/1.5 Gy 14 Jahre (Spanne 5–21 Jahre). Spätfolgen nach der Therapie
fraction (p = 0.037). Other effects were endocrine in 3 pa- traten in 73 % der Patienten (16/22) auf, davon 50 % im be-
tients and second malignancies in 2 patients. Four patients strahlten Volumen (11/22). Die häufigsten „In-field“-Toxi-
presented with multiple in-field late effects only with doses zitäten waren Muskel- und Skelettstörungen (n = 7), die nur
> 31 Gy. bei Dosen > 31 Gy/1,5 Gy-Fraktion auftraten (p = 0.037).
Conclusion After a median follow-up of 14 years, late ef- Des Weiteren entwickelten 3 Patienten endokrine Effekte
fects with multimodality treatment were frequent. The most und 2 Patienten Zweitmalignome. Gleichzeitige multiple
frequent effects were musculoskeletal abnormalities and the „In-field“-Spätwirkungen traten bei 4 Patienten nur bei Do-
threshold for their occurrence was 31 Gy. sen > 31 Gy auf.
Schlussfolgerung Mit einer medianen Beobachtungszeit
Keywords Late effects · Radiation therapy · von 14 Jahren waren Spätfolgen nach multimodaler Be-
Neoplasms, second primary · Scoliosis · Musculoskeletal handlung häufig. Am häufigsten waren Muskel-Skelett-An-
abnormalities omalien. Die Schwelle für ihr Auftreten lag bei einer Dosis
von 31 Gy.

Langzeittoxizität nach Strahlentherapie des Schlüsselwörter Spätfolgen · Strahlentherapie ·

lokalisierten Neuroblastoms im Kindesalter Neoplasien, zweite primäre · Skoliose ·
Muskel-Skelett-Anomalien
Ergebnisse der NB90- und NB94-Studien

Zusammenfassung Neuroblastoma (NB) is the most common extracranial solid

Einleitung Das Neuroblastom (NB) ist die häufigste Indi- tumor in childhood [1]. Approximately 50 % of patients
kation für eine extrakranielle pädiatrische Strahlentherapie. present with localized tumors at diagnosis. Non-amplified
Da sich beim lokalisierten Hochrisiko-NB das langfristige MYCN and resectable localized tumors have an excel-
Überleben stark verbessert hat, überprüften wir die behand- lent prognosis, with a 5-year overall survival (OS) rate of
lungsbedingte späte Toxizität bei pädiatrischen Patienten, > 90 %, depending on their histopathological features [2].
die im Rahmen zweier prospektiver klinischer Studien in Due to preoperative chemotherapy, unresectable tumors
Frankreich (NB90 und NB94) eine postoperative Strahlen- have a similar 5-year OS rate of 95 % [3].
therapie (RT) bei lokalisiertem NB erhalten hatten.

13
Long-term side effects of radiotherapy for pediatric localized neuroblastoma 3

However, the survival prognosis for MYCN-amplified to decrease the rate of metastatic relapse and improve OS
(MYCNA) NB is dismal and intensified treatment is now [4].
recommended, including preoperative chemotherapy, sur- The technique was 2D or 3D RT. The clinical target vol-
gery, high-dose chemotherapy (HDC), and radiotherapy ume was the preoperative volume plus a safety margin of
(RT). These treatments can improve OS by 70–80 % [4, 5]. 2 cm (1 cm for a thoracic or cervical tumor), encompass-
RT plays an important role in the local control of NB [6]. ing all areas at risk for microscopic disease as indicated by
RT to the primary is currently delivered in high-risk localized surgical reports and pathological examinations, incorpo-
NB with MYCNA, as well as in most cases of metastatic NB. rating uncertainties of positioning and organ motion, and
Neuroblastoma is, thus, the second most frequent indication including entire adjacent vertebrae. The recommended dose
for RT in pediatric cancer, after brain tumors. Survival is was 24 Gy for children < 2 years old and 34 Gy for those
improving, emphasizing the need for improved knowledge > 2 years. A localized boost of 5 Gy could be delivered to
on the long-term effects of multimodality treatment includ- any site that was deemed to be at particular risk for local
ing RT in this population [7]. Because the median age of failure.
children with NB is < 2 years, these therapies may increase Fractionation was either conventional (1.5–1.8 Gy per
the risk of late adverse effects. The cumulative incidence day, 5 days a week) or hyperfractionated (1 Gy twice a day,
of chronic health conditions is 41.1 % in NB survivors [8]. 5 days a week). Within the dose ranges stated above, deci-
Late effects related to RT are mainly musculoskeletal, neu- sions on final RT dose and fields used for each patient were
rologic, endocrine, and secondary cancers [9]. made on an individual basis and were based on clinical and
From 1990–2000, all children with localized NB were pathological information, such as age at diagnosis, tumor
included in two prospective studies, NB90 and NB94, by location, potential toxicities, lymph node involvement, and
the French Society of Pediatric Oncology (SFOP). Data extent of resection.
were retrospectively analyzed in this population. The aim This study is an unplanned subgroup analysis of the pro-
of our study was to evaluate the late effects of RT in these tocol’s initial design. Medical records of the 35 patients
patients also treated with surgery and chemotherapy. treated with RT were reviewed for information, treatments
received, and survival. Doses delivered to organs at risk
could not be retrieved since the majority of plans were
Patients and methods 2-dimensional. As different doses per fraction were used,
we performed the analysis with both delivered dose and bio-
Treatment logically equivalent dose (BED) with 1.5 Gy/fraction. An
α/β ratio of 3 was selected in order to evaluate the absorbed
Between March 1990 and March 2000, all consecutive dose in bone or soft tissue. The BED with 1.5 Gy/fractions
patients diagnosed with localized NB were registered in the was chosen since it was the most frequent fractionation in
SFOP-Localized Neuroblastoma-90 and -94 studies. A total this cohort and the fractionation now used in the European
of 33 institutions participated in these studies. prospective trials.
The aim of the NB90 study was to evaluate the chemo- Patients were followed-up annually, with a systematic
therapies strategy, i.e., two cycles of carboplatin and VP16, clinical exam and biological exams including blood count,
followed by two cycles of cyclophosphamide, doxorubicin, renal, and hormonal analysis. Follow-up and informa-
and vincristine (CAdO), before surgery for unresectable tion on late effects were collected until September 2012
NB, as described previously [10]. The same drugs were and obtained from computerized clinical notes from each
administered (1–2 cycles) after surgery when there was center. We recorded and graded according to the Common
residual disease and/or lymph node involvement. The NB94 Terminology Criteria for Adverse Events v4.0 (CTCAEv4)
study was a de-escalation trial that included children who all described side effects: musculoskeletal, endocrine, neu-
were < 1 year of age: The children received low-dose cyclo- rologic, vascular, digestive, cardiac, renal, pulmonary, and
phosphamide and vincristine before surgery. There was no second cancers.
chemotherapy after surgery [11]. Musculoskeletal abnormalities (MSAs) were defined as
RT was recommended if there was persistent gross resid- any abnormality, such as spinal curvature (scoliosis, lor-
ual disease at the end of treatment in children aged > 1 year dosis, kyphosis), bony hypoplasia, or muscular hypoplasia
at diagnosis. Patients with MYCNA had a high recurrence on physical examination. In addition, the development of
rate, which was either local or metastatic. Consequently, the malignant neoplasms was included in the analyses.
NB committee recommended additional local RT beginning
in 1992, irrespective of the patient's postoperative status and
age. After 1997, the NB committee also recommended pre-
RT HDC (busulfan: 600 mg/m2 and melphalan: 140 mg/m2)

13
4 A. Ducassou et al.

Statistical analyses 31.1–48.0 Gy). There was no significant difference between

the two groups in location of primary tumor or treatment
Descriptive statistics were reported as absolute frequencies with high dose chemotherapy. Despite the protocol, some
and percentages for qualitative data, whereas medians and children < 2 years old received more than the recommended
ranges were used for continuous variables. Disease progres- 24 ± 5 Gy boost, and some of the older children received
sion, relapse, and death (whatever the cause) were consid- less than the recommended 34 ± 5 Gy boost. This explains
ered events within the event-free analysis. Local control was the absence of difference in age between these two groups.
defined as the absence of relapse in the irradiated primary
tumor site RT field. Local control, event-free survival, and Late effects
overall survival rates were estimated by the Kaplan–Meier
method, and differences between groups were assessed by Late effects for each patient are listed in Table 1 and Figs. 1
the log-rank test. Survival estimates refer to times of 5 and and 2.
15 years after diagnosis, and the related 95 % confidence Overall, 73 % of patients (16/22) presented with long-
intervals (95 % CI) were calculated [12]. All statistical tests term sequelae after treatment, within the RT field in 50 %
were two-sided and a p-value of < 0.05 was considered (11/22) of patients. The most frequent in-field effects were
significant. Analyses were performed using SAS software musculoskeletal abnormalities (n = 7), endocrine effects
(SAS institute Inc., Cary, NC, USA) version 9.1. (n = 3), and second malignancies (n = 2). A total of 29 com-
plications occurred in 16 patients, of which 65.5 % (n = 19)
occurred within the RT fields (Fig. 2). Among these 19 in-
Results field toxicity events, 3 occurred with doses < 31 Gy, and 16
with doses > 31 Gy. Eleven were grade 1 or 2 according to
A total of 610 children were registered in these two stud- CTCAE-v4, whereas 8 were grade 3 or 4. The grade was
ies. Among the 35 irradiated children, 22 were alive 5 years unrelated to the dose. Four patients presented more than one
after diagnosis and were analyzed for radiation-induced late in-field late effect; this represents 29 % (4/14) of patients
effects after a median follow-up time of 14 years (range with a dose > 31 Gy. No more than one late effect occurred
5–21 years). Table 1 describes the 5-year survivors’ detailed below this dose in each child (Fig. 1).
characteristics. Briefly, their median age at treatment was MSAs were the most frequent late effect: they occurred
29 months (range 2–192 months); the primary tumor was in 31 % of patients (7 patients and 10 events) at a median
abdominal in 15 (68 %). International Neuroblastoma Stag- time of 6 years after treatment (range 5–11 years; Fig. 2). All
ing System (INSS) stage 3 was observed in 77 %, and occurred within the RT field and in children having received
MYCNA was found in 45 %. more than a BED of 31 Gy (7/15) compared to none in
patients who received < 31 Gy (0/7; p = 0.038). Considering
Radiotherapy treatment real delivered dose this was also significant with a threshold
of 33 Gy (p = 0.02) under which no MSA occurred. MSAs
The median total given dose to these 22 patients was were scoliosis in 3 patients, kyphosis in two, hyperlordosis
34.5 Gy (range 24–45 Gy). The median fraction-dose per in one, bony hypoplasia in one, and muscular hypoplasia in
day was 1.5 Gy (range 0.8–2 Gy). Hyperfractionated radia- three. According the CTCAEv4, we observed two grade 3
tion (1 Gy twice a day) was delivered to 6 patients. The effects (1 scoliosis and 1 kyphosis), and eight grade 1 or
median BED with 1.5 Gy/fraction was 32.6 Gy (range 21.6– 2 effects. The median height of the patients was 171 cm in
48 Gy). The technique used opposed anteroposterior fields, males (range 149–179 cm) and 160 cm in females (range
mainly using 2D planning techniques. One patient had a 3D 150–166 cm), at adolescent or adult height. They were sim-
plan. Median beam energy was 6 MV (range 1.25–23 MV). ilar with or without MSA. All the MSAs occurred at the
According to the protocol, the entire width of the vertebral site of the primary tumor: eight MSAs occurred after RT in
body was included within the RT field for all children. Mean 6 patients and two occurred before RT, 1 after laminectomy.
and median number of vertebrae within the RT fields was Among the 3 patients with a dumbbell NB, 1 developed
five, and they were mainly thoracolumbar vertebrae in 68 % MSA, 1 after laminectomy and RT, and 1 after RT alone. No
of children. statistical difference was found according to the presence or
In the 22 children who survived longer than 5 years, absence of an intraspinal NB or laminectomy.
we observed two groups according to the dose of RT. The Neither the beam energy, number of vertebrae included in
first group received less than 31 Gy (n = 7, age range 0–4.5 the RT field, type of vertebrae, nor age at treatment affected
years) with a median dose of 24 Gy (range 21.6–30 Gy) and the occurrence of MSAs.
the second group (n = 15, age range 0–16 years) received Among the 5 patients who had late endocrine side effects,
more than 31 Gy with a median dose of 34.5 Gy (range 3 occurred within or at the edge of the RT field. Hypothy-

13
 Table 1 Detailed characteristics of the 22 patients who survived more than 5-yearsTable 1 Detailed characteristics of the 22 patients who survived more than 5-years
Gender Age at INSS MYC-N Dumbbell/ Tumor site High-dose Radiation BED with Vertebrae Follow-up Late effects: Late effects: other In-field
diagnosis amplifica- laminectomy chemo- dose (Gy) 1.5 Gy/fraction in RT (years) MSA effects
(years) tion therapy (α/β = 3) fields (n) (n)
1 M 1.6 3 No No/No Pelvis No 25.6 21.6 3 17.1 Lymphoedema 0
2 F 16.0 3 No No/No Thorax Yes, at 35 31.11 12 15.0 Hepatitis C 0
relapse
3 M 2.1 3 No No/No Abdomen No 24 24 4 4.8 No
4 F 3.8 3 No No/No Pelvis No 36 38.4 1 21.1 Muscular atro- Hypogonadism (gr 3
phy (gr 2) 3), uterine hypo-
plasia (gr 3)
5 F 2.0 3 No No/No Abdomen No 35 31.11 4 21.0 Kyphosis (gr 2
3), Muscular
atrophy (gr 2)
6 F 1.8 3 No No/No Pelvis No 30.6 32.6 6 19.8 Steppage gait, 1
Pilocytic astro-
cytoma, uterine
sarcoma (gr 4)
7 M 0.2 1 Yes No/No Thorax No 24 22.4 5 11.5 Papillary carci- 1
noma (gr 3)
8 F 3.7 3 No Yes/Yes Head and No 35 36.9 8 16.1 Scoliosis (gr Hypoesthesia, 5
neck 3), muscular esophageal
atrophy (gr 2), stenosis (gr 2),
hypothyroidism
(gr 2), respiratory
failure (gr 3)
Long-term side effects of radiotherapy for pediatric localized neuroblastoma

9 F 12.0 3 No Yes/Yes Pelvis No 45 48 1 10.9 Hypoesthesia 0

10 F 3.3 3 Yes No/No Abdomen No 35 31.11 6 11.5 No
11 M 13.7 3 No No/No Abdomen Yes 35 36.9 Unknown 12.7 No
12 M 4.8 1 No No/No Abdomen Yes, at 35 31.11 6 14.6 Hypogonadism 0
relapse
13 M 0.9 1 No No/No Abdomen No 37.4 39 5 16.7 Kyphosis (gr 2) 1
14 M 2.0 3 Yes No/No Abdomen Yes 27 27 7 15.1 Moderate cardiac 1
hypokinesia (gr 1)
15 F 3.0 3 Yes No/No Abdomen Yes 34.5 34.5 3 15.0 Hyperlordosis Occlusive 1
(gr 2) syndrome,
nephrectomy
16 F 1.5 2 Yes No/No Abdomen Yes 33.6 34.35 4 14.0 Scoliosis (gr 1), 2
Bony hypoplasia
(gr 1)
17 M 4.6 3 Yes No/No Abdomen Yes 30 30 8 13.7 GH deficiency 0
18 M 2.8 3 Yes No/No Abdomen Yes 34.5 34.5 5 13.7 No
19 M 1.6 3 No Yes/No Abdomen No 25 25.5 4 13.4 No
20 F 0.8 2 Yes No/No Abdomen Yes 24 24 5 12.8 Hypogonadism 1
(gr 2)
21 F 1.6 3 Yes No/No Abdomen Yes 34.5 34.5 4 12.3 Scoliosis (gr 1) 1
22 M 3.6 3 Yes No/No Abdomen Yes 34.5 34.5 9 11.4 No
BED Biological equivalent dose, gr grade CTCAEv4 for in-field effect, MSA musculoskeletal abnormalities, M male, F female, GH growth hormone, INSS International Neuroblastoma Staging System.
5

13
6 A. Ducassou et al.

Fig. 1 Percentage of patients

surviving 5 years postdiagnosis
presenting with late side-effects,
in-field late effects, second can-
cers, two or more late effects, and
musculoskeletal abnormalities,
based on whether they received
greater than or less than a dose
of 31 Gy

Fig. 2 Number and details of late

effects (29), and their occurrences
within radiotherapy (RT) fields
(16/29) in the 22 patients who
survived > 5 years

roidism occurred after a cervical RT dose of 35 Gy, i.e., uterine hypoplasia, and in an infant after an abdominal RT
36.9 Gy BED. Two cases of hypogonadism occurred: in a dose of 24 Gy in a field that encompassed T12–L4 after
4-year-old girl who had received HDC and 36 Gy (38.4 Gy HDC. The two other endocrine late effects were outside of
BED) to the sacrum for a pelvic tumor and later developed the RT fields: one boy presented with hypogonadism and

13
Long-term side effects of radiotherapy for pediatric localized neuroblastoma 7

was treated with HDC and RT to the abdomen (35 Gy, i.e., is thus much longer than other studies dedicated to RT late
31.11 Gy BED, included T11–L4 and the flank), and one effects which were concerning metastatic NB [14–16]. We
boy developed growth hormone deficiency after treatment compared our results with previous reports in advanced NB
with HDC and thoraco-abdominal RT. with a majority of RT. Laverdiere et al. [17] reported during
All other late effects are detailed in Table 1 and Fig. 2. a 7-year follow-up that at least one late complication in 95 %
The only renal abnormalities described were 2 cases of of 61 advanced stage NB survivors (79 % stage IV), 89 % of
nephrectomy before RT, without biological defect at follow- them having been treated with RT. Comparable rates (93 %)
up; therefore, renal toxicity was not included in the table of long-term sequelae, half of them being severe, were
of late effects. Because 50 % of the long-term surviving observed with a 4-year median follow-up in 16 long-term
patients had undergone HDC, we performed comparative survivors treated for a stage 4 NB, with RT in 8 [18]. Only
analyses on toxicity with or without HDC. There was no dif- 14 and 12 % of MSAs were respectively observed in these
ference in the incidence of late effects with or without HDC. two cohorts, fewer than in our cohort, because of the fewer
rates of RT, the lower delivered dose (mean doses 21 Gy in
Secondary cancers both) and a much shorter follow-up.
In a large cohort of all stages 954 NB, with 48 % of RT,
Three secondary cancers occurred in 2 patients. One patient, RT of the spine was significantly associated with the devel-
treated for thoracic NB with 22.4 Gy BED at 2 years of age opment of severe scoliosis in comparison with siblings, the
presented with a thyroid papillary carcinoma 10 years after 20-year incidence of MSA being 7.8 % [8].
treatment. The 2D RT fields had included vertebrae T2–T6. Le Cohen et al. [19] also showed an additive effect of
The other patient, treated for pelvic NB with 32.6 Gy BED total body irradiation (TBI) in 51 high-risk NB survivors
at 1 year of age, developed pilocytic astrocytoma 11 years with a 6-year follow-up. Eighty percent had received 12 Gy
after treatment, and then a fatal metastatic uterine sarcoma TBI and 88 % a local RT (10–18 Gy); all had undergone
18 years after treatment. None had received HDC. HDC. With a median follow-up of 6 years, a statistical
decrease in linear growth with TBI was observed. MSAs
(scoliosis) were described in only 6 % of patients but were
Discussion underdiagnosed according to the authors. For instance, TBI
is not used anymore in the European high-risk neuroblas-
The data presented herein pertain to patients treated in two toma SIOPEN clinical trial.
large national prospective clinical studies that included all Complications are also frequent after treatment exclusive
patients with localized NB (n = 610) spanning 10 years. To of RT, for advanced stage NB. In a recent chemotherapy
our knowledge this is the longest follow-up published after clinical trial (ENSG5 trial), at least one effect was observed
radiotherapy for localized NB. Moreover, the methodolo- in 75–80 % of patients, with 15–20 % of severe events, dur-
gies reported (chemotherapy, HDC, and AP/PA RT) are still ing a 12.9-year follow-up. Nevertheless, the effects were
the most common therapies for high-risk localized NB. different, with more renal (14 %), and neurological morbid-
As long-term survival is consistently increasing in these ity (hearing loss in 50 %) due to HDC, but lower rates of
patients, improved understanding of the long-term effects hypothyroidism (2 %) and MSAs (2 %) [20].
of multimodality treatment, including moderate to higher MSAs, the most frequent late effect in our study, occurred
doses of RT, is of tremendous importance. in 50 % of patients who had received doses > 31 Gy, whereas
Late effects occurred in 73 % of the patients in our cohort, no MSAs were found in those who had received < 31 Gy:
within RT fields in 50 % of patients. All patients who experi- this difference was statistically significant. Laminectomy
enced more than 1 late effect had received more than 31 Gy. and RT doses are well-known risk factors for developing
The most frequent effect was MSAs, occurring in 31 % of MSAs, though there has been no consensus on the dose that
patients. causes this effect. It has been shown that asymmetric irra-
Long-term toxicity is very frequent in children treated diation of the spine and RT dose > 30 Gy are risk factors
for a NB with a multimodality treatment including RT. This for causing spinal deformity [21]. Doses between 17.5 and
was described by Oeffinger et al. [13], showing a high fre- 39 Gy have been associated with a 50–60 % incidence of
quency of multiple late effects (38 % with 2 or more late scoliosis, whereas the incidence was < 10 % if RT dose was
events), unaffected by RT dose, in 10,000 survivors of pedi- below 17.5 Gy. Laminectomy and RT dose may play a role
atric tumors with a 17.5-year follow-up, including NB. The in causing spinal defects. Paulino et al. [22] reported that the
question of dose in NB was not addressed specifically. If association of laminectomy and RT was the most frequent
several papers have described late toxicity after NB treat- risk factor for scoliosis. We did not find associations with
ment, our series is the first, to our knowledge, focusing on laminectomy, or of any other factors (age, number of verte-
very late effects after RT for localized NB. Our follow-up brae, HDC) in the present study.

13
8 A. Ducassou et al.

The endocrine effects, mainly hypothyroidism, are fre- 50 % of these occurred within the RT field. Late MSAs were
quent in NB and are often due to the association of HDC the most frequent events, occurring only after doses > 31 Gy.
and RT [8, 18, 19]. Hypothyroidism is also frequent in other We suggest that the data described in this article can aid
pediatric cancers, since patients treated for Hodgkin disease communication with the patients’ parents and offer addi-
(HD) are 5.5 times more likely to experience hypothyroid- tional information in the discussion of the optimal dose of
ism after 30 Gy [23]. Our results are in concordance with RT in future prospective trials. Our results will require fur-
these previous reports, since the only patient experiencing ther confirmation from larger, ongoing prospective clinical
hypothyroidism in our series was treated for a head and trials.
neck tumor with a dose of 36.9 BED.
Two patients (5.7 %) developed secondary cancers, a rate Acknowledgments We thank Dr. Ursula Nestle for German editing
and Newmed Publishing for English editing assistance.
similar to previously reported rates of 2–6 % [8, 24]. The
secondary cancers that occurred in the RT fields in our cohort
(thyroid neoplasm and sarcoma) are common histological Compliance with ethical guidelines
radiation-induced neoplasms [24, 25]. The delivered doses
were 22.4 and 32.6 Gy. These secondary cancers occurred Conflict of interest On behalf of all authors, the corresponding author
states that there are no conflicts of interest.
in very young children 10–20 years after RT, which is con-
cordant with previous studies [26].
The accompanying manuscript does not include studies
Our data are meaningful as the patients described in this
on humans or animals.
study received the same treatment as the one most frequently
recommended nowadays, i.e., opposed anterior–posterior/
posterior–anterior fields. This technique allows robust and References
complete coverage of the vertebrae and provides sharp dose
cut-off points at the vertebral body edges [27, 28]. With the 1. Howlader N, Noone AM, Krapcho M et al (2013) SEER Cancer
evolution of radiotherapy and in specific cases, intensity- Statistics Review, 1975–2010. National Cancer Institute, Bethesda
modulated radiation therapy with linear accelerators, tomo- 2. De Bernardi BM, Rubie H et al (2008) Treatment of localised re-
sectable neuroblastoma. Results of the LNESG1 study by the SIOP
therapy or proton therapy can be used to reduce damage to Europe Neuroblastoma Group. Br J Cancer 99:1027–1033
organs at risk, while still homogeneously including the ver- 3. Rubie H, Michon J, Plantaz D et al (1998) Unresectable localized
tebrae adjacent to the tumor site [29–33]. neuroblastoma: improved survival after primary chemotherapy in-
Although chemotherapy induces many late effects, we cluding carboplatin-etoposide. Neuroblastoma Study Group of the
Société Française d’Oncologie Pédiatrique (SFOP). Br J Cancer
thought it was important to specifically study the effects 77:2310–2317
of local radiation therapy. RT is delivered systematically 4. Laprie A, Michon J, Hartmann O et al (2004) High-dose chemo-
to the primary tumor after conventional chemotherapy, therapy followed by locoregional irradiation improves the out-
surgery, and HDC in cases of stage IV disease or high-risk come of patients with international neuroblastoma staging system
Stage II and III neuroblastoma with MYCN amplification. Cancer
localized NB in the ongoing European high-risk protocol 101:1089
of the SIOPEN, trial HR-NBL 1.5 (NCT00030719), in 5. Bagatell R, Beck-Popovic M, London WB et al (2009) Signifi-
which all French patients are included. The dose to the pri- cance of MYCN amplification in international neuroblastoma
mary tumor is 21 Gy in 14 fractions of 1.5 Gy, regardless staging system stage 1 and 2 neuroblastoma: a report from the
International Neuroblastoma Risk Group database. J Clin Oncol
of extent of resection. Although randomized trials explor- 27:365–370
ing RT-specific questions in high-risk NB are lacking, the 6. Castleberry RP, Kun LE, Shuster JJ et al (1991) Radiotherapy im-
Children’s Oncology Group (COG) suggested a dose effect proves the outlook for patients older than 1 year with Pediatric
in their CCG-3891 trial (NCT00026312) as the addition of Oncology Group stage C neuroblastoma. J Clin Oncol 9:789–795
7. Matthay K, Villablanca J, Seeger R et al (1999) Treatment of high-
10 Gy localized RT was improving outcome after 10 Gy risk neuroblastoma with intensive chemotherapy, radiotherapy,
TBI [34]. However, the question of whether an increased autologous bone marrow transplantation, and 13-cis-retinoic acid.
dose should be given for a residual gross disease remains Children’s Cancer Group. N Engl J Med 341:1165–1173
debated [35]. In their following high-risk protocol, the COG 8. Laverdière C, Liu Q, Yasui Y et al (2009) Long-term outcomes in
survivors of neuroblastoma: a report from the Childhood Cancer
(NCT01175356) is evaluating dose escalation in patients Survivor Study. J Natl Cancer Inst 101:1131–11340
with gross disease who receive a dose of 36 Gy compared 9. Paulino AC, Mayr N, Simon JH et al (2002) Locoregional control
to those with a microscopic residual disease who receive in infants with neuroblastoma: role of radiation therapy and late
21.6 Gy; results are pending. toxicity. Int J Radiat Oncol Biol Phys 52:1025–1031
10. Rubie H, Hartmann O, Michon J et al (1997) N-Myc gene ampli-
In these series of children treated with RT during a fication is a major prognostic factor in localized neuroblastoma:
10-year period for intermediate or high-risk localized NB results of the French NB 90 study. Neuroblastoma Study Group
in two consecutive clinical trials, with a median follow-up of the Société Francaise d’Oncologie Pédiatrique. J Clin Oncol
of 14 years, 73 % of patients presented with late effects, and 15:1171–1182

13
Long-term side effects of radiotherapy for pediatric localized neuroblastoma 9

11. Rubie H, Coze C, Plantaz D et al (2003) Localised and unresect- 25. Rubino C, Adjadj E, Guérin S et al (2003) Long-term risk of sec-
able neuroblastoma in infants: excellent outcome with low-dose ond malignant neoplasms after neuroblastoma in childhood: role
primary chemotherapy. Br J Cancer 89:1605–1609 of treatment. Int J Cancer 107:791–796
12. Prentice RL, Kalbfleisch JD, Peterson AV Jr et al (1978) The anal- 26. Flandin I, Hartmann O, Michon J et al (2006) Impact of TBI on
ysis of failure times in the presence of competing risks. Biometrics late effects in children treated by megatherapy for stage IV neuro-
34:541–554 blastoma. A study of the French Society of Pediatric oncology. Int
13. Oeffinger KC, Mertens AC, Sklar CA et al (2006) Chronic health J Radiat Oncol 64:1424–1431
conditions in adult survivors of childhood cancer. N Engl J Med 27. Beneyton V, Niederst C, Vigneron C et al (2012) Comparison of
355:1572–1582 the dosimetries of 3-dimensions radiotherapy (3D-RT) with linear
14. Gatcombe HG, Marcus RB Jr, Katzenstein HM et al (2009) Excel- accelerator and intensity modulated radiotherapy (IMRT) with he-
lent local control from radiation therapy for high-risk neuroblas- lical tomotherapy in children irradiated for neuroblastoma. BMC
toma. Int J Radiat Oncol Biol Phys 74:1549–1554 Med Phys 12:2
15. Bradfield SM, Douglas JG, Hawkins DS et al (2004) Fractionated 28. Paulino AC, Ferenci MS, Chiang K-Y et al (2006) Comparison of
low-dose radiotherapy after myeloablative stem cell transplanta- conventional to intensity modulated radiation therapy for abdomi-
tion for local control in patients with high-risk neuroblastoma. nal neuroblastoma. Pediatr Blood Cancer 46:739–744
Cancer 100:1268–1275 29. Paulino AC (2007) IMRT in abdominal neuroblastoma: response.
16. Browne M, Kletzel M, Cohn SL et al (2006) Excellent local tumor Pediatr Blood Cancer 48:715
control regardless of extent of surgical resection after treatment 30. Gains JE, Stacey C, Rosenberg I et al (2013) Intensity-modulated
on the Chicago Pilot II protocol for neuroblastoma. J Pediatr Surg arc therapy to improve radiation dose delivery in the treatment of
41:271–276 abdominal neuroblastoma. Future Oncol 9:439–449
17. Laverdière C, Cheung N-KV, Kushner BH et al (2005) Long-term 31. Shaffer R, Vollans E, Vellani R et al (2011) A radiotherapy plan-
complications in survivors of advanced stage neuroblastoma. Pe- ning study of RapidArc, intensity modulated radiotherapy, three-
diatr Blood Cancer 45:324–332 dimensional conformal radiotherapy, and parallel opposed beams
18. Perwein T, Lackner H, Sovinz P et al (2011) Survival and late ef- in the treatment of pediatric retroperitoneal tumors. Pediatr Blood
fects in children with stage 4 neuroblastoma. Pediatr Blood Cancer Cancer 56:16–23
57:629–635 32. Hill-Kayser C, Tochner Z, Both S et al (2013) Proton versus
19. Cohen LE, Gordon JH, Popovsky EY et al (2014) Late effects in photon radiation therapy for patients with high-risk neuroblas-
children treated with intensive multimodal therapy for high-risk toma: the need for a customized approach. Pediatr Blood Cancer
neuroblastoma: high incidence of endocrine and growth problems. 60:1606–1611
Bone Marrow Transplant 49:502–508 33. Hillbrand M, Georg D, Gadner H et al (2008) Abdominal cancer
20. Moreno L, Vaidya SJ, Pinkerton CR et al (2013) Long-term fol- during early childhood: a dosimetric comparison of proton beams
low-up of children with high-risk neuroblastoma: the ENSG5 trial to standard and advanced photon radiotherapy. Radiother Oncol
experience. Pediatr Blood Cancer 60:1135–1140 89:141–149
21. Mayfield JK, Riseborough EJ, Jaffe N et al (1981) Spinal de- 34. Haas-Kogan DA, Swift PS, Selch M et al (2003) Impact of radio-
formity in children treated for neuroblastoma. J Bone Joint Surg therapy for high-risk neuroblastoma: a Children’s Cancer Group
63:183–193 study. Int J Radiat Oncol Biol Phys 56:28–39
22. Paulino AC, Fowler BZ (2005) Risk factors for scoliosis in chil- 35. Simon T, Hero B, Bongartz R et al (2006) Intensified external-beam
dren with neuroblastoma. Int J Radiat Oncol Biol Phys 61:865–869 radiation therapy improves the outcome of stage 4 neuroblastoma
23. Sklar C, Whitton J, Mertens A et al (2000) Abnormalities of the in children > 1 year with residual local disease. Strahlenther Onkol
thyroid in survivors of Hodgkin’s disease: data from the Childhood 182:389–394
Cancer Survivor Study. J Clin Endocrinol Metab 85:3227–3232
24. Meadows AT, Friedman DL, Neglia JP et al (2009) Second neo-
plasms in survivors of childhood cancer: findings from the Child-
hood Cancer Survivor Study cohort. J Clin Oncol 27:2356–2362

13