Articles

Long-acting intramuscular cabotegravir and rilpivirine in

adults with HIV-1 infection (LATTE-2): 96-week results of
a randomised, open-label, phase 2b, non-inferiority trial
David A Margolis, Juan Gonzalez-Garcia, Hans-Jürgen Stellbrink, Joseph J Eron, Yazdan Yazdanpanah, Daniel Podzamczer, Thomas Lutz,
Jonathan B Angel, Gary J Richmond, Bonaventura Clotet, Felix Gutierrez, Louis Sloan*, Marty St Clair, Miranda Murray, Susan L Ford, Joseph Mrus,
Parul Patel, Herta Crauwels, Sandy K Griffith, Kenneth C Sutton, David Dorey, Kimberly Y Smith, Peter E Williams, William R Spreen

Summary
Background Cabotegravir and rilpivirine are antiretroviral drugs in development as long-acting injectable formulations. Lancet 2017; 390: 1499–510
The LATTE-2 study evaluated long-acting cabotegravir plus rilpivirine for maintenance of HIV-1 viral suppression Published Online
through 96 weeks. July 24, 2017
http://dx.doi.org/10.1016/
S0140-6736(17)31917-7
Methods In this randomised, phase 2b, open-label study, treatment-naive adults infected with HIV-1 initially received
See Comment page 1468
oral cabotegravir 30 mg plus abacavir–lamivudine 600–300 mg once daily. The objective of this study was to select an
ViiV Healthcare, Research
intramuscular dosing regimen based on a comparison of the antiviral activity, tolerability, and safety of the Triangle Park, NC, USA
two intramuscular dosing regimens relative to oral cabotegravir plus abacavir–lamivudine. After a 20-week induction (D A Margolis MD, M St Clair BS,
period on oral cabotegravir plus abacavir–lamivudine, patients with viral suppression (plasma HIV-1 RNA <50 copies J Mrus MD, Parul Patel PharmD,
per mL) were randomly assigned (2:2:1) to intramuscular long-acting cabotegravir plus rilpivirine at 4-week intervals S K Griffith Pharm D,
K C Sutton MA, K Y Smith MD,
(long-acting cabotegravir 400 mg plus rilpivirine 600 mg; two 2 mL injections) or 8-week intervals (long-acting W R Spreen Pharm D); Hospital
cabotegravir 600 mg plus rilpivirine 900 mg; two 3 mL injections) or continued oral cabotegravir plus abacavir– Universitario La Paz/IdiPAZ,
lamivudine. Randomisation was computer-generated with stratification by HIV-1 RNA (<50 copies per mL, yes or no) Madrid, Spain
during the first 12 weeks of the induction period. The primary endpoints were the proportion of patients with viral (J Gonzalez-Garcia MD); ICH
Hamburg, Hamburg, Germany
suppression at week 32 (as defined by the US Food and Drug Administration snapshot algorithm), protocol-defined (Prof H-J Stellbrink MD);
virological failures, and safety events through 96 weeks. All randomly assigned patients who received at least one dose University of North Carolina,
of study drug during the maintenance period were included in the primary efficacy and safety analyses. The primary Chapel Hill, NC, USA
(Prof J J Eron MD); Hôpital
analysis used a Bayesian approach to evaluate the hypothesis that the proportion with viral suppression for each long-
Bichat Claude Bernard, Paris,
acting regimen is not worse than the oral regimen proportion by more than 10% (denoted comparable) according to France (Prof Y Yazdanpanah MD);
a prespecified decision rule (ie, posterior probability for comparability >90%). Difference in proportions and associated Hospital Universitari de
95% CIs were supportive to the primary analysis. The trial is registered at ClinicalTrials.gov, number NCT02120352. Bellvitge, L’Hospitalet,
Barcelona, Spain
(D Podzamczer PhD);
Findings Among 309 enrolled patients, 286 were randomly assigned to the maintenance period (115 to each of the Infektiologikum, Frankfurt,
4-week and 8-week groups and 56 to the oral treatment group). This study is currently ongoing. At 32 weeks following Germany (T Lutz MD); The
randomisation, both long-acting regimens met primary criteria for comparability in viral suppression relative to the oral Ottawa Hospital, Ottawa, ON,
Canada (J B Angel MD); Broward
comparator group. Viral suppression was maintained at 32 weeks in 51 (91%) of 56 patients in the oral treatment group,
General Medical Center, Fort
108 (94%) of 115 patients in the 4-week group (difference 2·8% [95% CI −5·8 to 11·5] vs oral treatment), and 109 (95%) Lauderdale, FL, USA
of 115 patients in the 8-week group (difference 3·7% [−4·8 to 12·2] vs oral treatment). At week 96, viral suppression was (G J Richmond MD); Hospital
maintained in 47 (84%) of 56 patients receiving oral treatment, 100 (87%) of 115 patients in the 4-week group, and Germans Trias i Pujol, UAB,
UVIC-UCC, Badalona, Catalonia,
108 (94%) of 115 patients in the 8-week group. Three patients (1%) experienced protocol-defined virological failure (two
Spain (B Clotet MD); Hospital
in the 8-week group; one in the oral treatment group). Injection-site reactions were mild (3648 [84%] of 4360 injections) General de Elche & Universidad
or moderate (673 [15%] of 4360 injections) in intensity and rarely resulted in discontinuation (two [<1%] of 230 patients); Miguel Hernández, Alicante,
injection-site pain was reported most frequently. Serious adverse events during maintenance were reported in 22 (10%) Spain (Prof F Gutierrez MD);
North Texas Infectious Disease
of 230 patients in the intramuscular groups (4-week and 8-week groups) and seven (13%) of 56 patients in the oral Consultants, Dallas, TX, USA
treatment group; none were drug related. (Prof L Sloan MD); ViiV
Healthcare, London, UK
Interpretation The two-drug combination of all-injectable, long-acting cabotegravir plus rilpivirine every 4 weeks or (M Murray PhD);
GlaxoSmithKline, Research
every 8 weeks was as effective as daily three-drug oral therapy at maintaining HIV-1 viral suppression through Triangle Park, NC, USA
96 weeks and was well accepted and tolerated. (S L Ford PharmD); Janssen
Research and Development,
Funding ViiV Healthcare and Janssen R&D. Beerse, Belgium
(H Crauwels PhD,
P E Williams PhD); and
Introduction treatment efficacy for patients with HIV, enhancing GlaxoSmithKline, Mississauga,
An estimated 36·7 million individuals were living with patient survival and quality of life.2,3 However, adherence ON, Canada (D Dorey MMath)
HIV worldwide at the end of 2015.1 Advances in highly to medication remains an important challenge; poor *Prof Louis Sloan died in
June, 2017
active antiretroviral therapies (ARTs) have improved compliance can result in treatment failure and the

www.thelancet.com Vol 390 September 23, 2017 1499

 Articles

Correspondence to:
Dr David A Margolis, Research in context
ViiV Healthcare, 5 Moore Drive,
Research Triangle Park, Evidence before this study treatment of HIV-1 infection. An option to treat HIV-1 without
NC 27709-3398, USA To establish the background for this study, we searched PubMed the use of daily oral medications represents a paradigm shift in
david.a.margolis@ publications on the topics of antiretroviral therapy and the HIV-1 treatment landscape. The principal finding of the study
viivhealthcare.com
treatment adherence; long-acting injectable therapies; and the is that among patients who were suppressed on an oral
safety, efficacy, and pharmacokinetics of cabotegravir cabotegravir-based therapy, switching to a long-acting
(GSK1265744) and rilpivirine using the keywords “antiretroviral combination of cabotegravir and rilpivirine maintained
therapy”, “treatment adherence”, “long-acting injectable virological suppression in 90% of patients overall through
therapies”, “cabotegravir”, “GSK1265744”, “rilpivirine”, and week 96 following 4-week or 8-week injectable administration at
“TMC-2782”. We also located package inserts and government rates comparable to remaining on daily oral cabotegravir-based
documents using internet search engines. All searches were therapy. The complete week 96 dataset included here provides
updated as of March 7, 2017. A review of this literature shows an important evidence for both the durability of virological response
ongoing challenge in HIV therapy wherein suboptimal adherence and acceptability of intramuscular injections for chronic use with
to daily oral medication can lead to treatment failure or the dual antiretroviral therapy in patients infected with HIV-1.
emergence of viral resistance. To date, no long-acting injectable
Implications of all the available evidence
regimen is available to patients with HIV. Cabotegravir is an
To date, the class of integrase strand transfer inhibitors has
integrase strand transfer inhibitor with clinically demonstrated
shown a high level of virological efficacy in clinical studies, which
activity against the HIV-1 virus and a physiochemistry and
has translated to global widespread successful use. The ability to
pharmacokinetic profile amenable to its formulation and use as a
employ one of these agents, in partnership with one other
long-acting agent. Rilpivirine is a non-nucleoside reverse
agent, as an effective long-acting injectable agent has the
transcriptase inhibitor approved as an oral agent for the
potential to address the challenges of adherence to daily
treatment of HIV-1 infection in combination with other
medication faced by people living with HIV. The daily
antiretrovirals; its physiochemistry is also appropriate for
psychological burden of being discovered as HIV positive can be
formulation as a long-acting agent. We did a randomised,
eased by less frequent or clinic-based medication dosing and
open-label, parallel group, phase 2b study (LATTE-2) in patients
might be a preferred option for some patients. Following the
with HIV-1 viral suppression on oral medication to evaluate the
advent and proliferation of single-tablet regimens, which
efficacy, safety, and tolerability of cabotegravir plus rilpivirine
themselves constituted a leap forward in dosing convenience,
given as long-acting injections every 4 or 8 weeks, compared
long-acting injectables such as the cabotegravir plus rilpivirine
with daily oral cabotegravir taken with abacavir and lamivudine.
regimen might represent the next revolution in HIV therapy by
Added value of this study providing an option that circumvents the burden of chronic
The LATTE-2 study is the first to investigate the efficacy and daily dosing.
safety of a long-acting injectable antiretroviral therapy for the

emergence of drug-resistant mutations.4 Long-acting two nucleoside reverse transcriptase inhibitors (NRTIs),
injectable ART might provide some patients with a providing proof of principle for a two-drug maintenance
convenient approach to manage HIV infection that regimen using an INSTI and NNRTI.13
avoids daily oral dosing and the need to keep, store, and Long-acting injectable nanosuspension formulations of
transport medications as they undertake their activities cabotegravir and rilpivirine are in clinical development.12,14
of daily living.5 Phase 1 clinical studies investigating long-acting
Cabotegravir (GSK1265744) is an analogue of the cabotegravir and rilpivirine have shown prolonged
integrase strand transfer inhibitor (INSTI) dolutegravir exposures at least 30 days following gluteal intramuscular
that exhibits subnanomolar potency and antiviral activity injections, enabling dosing at once-monthly or longer
against a broad range of HIV-1 strains.6 Oral administration intervals.15,16 Combined administration of long-acting
of cabotegravir once daily has exhibited acceptable safety cabotegravir plus rilpivirine produced no clinically
and tolerability profiles, a long half-life (40 h), and few significant pharmacokinetic interactions, supporting
drug–drug interactions.7–9 Rilpivirine (TMC278) is a non- investigation of these agents as the first-ever long-acting
nucleoside reverse transcriptase inhibitor (NNRTI) that is combination ART regimen.16 Here, we report the efficacy
approved as a 25 mg once-daily oral medication for HIV-1 and safety of long-acting cabotegravir plus rilpivirine,
treatment.10–12 In the phase 2b LATTE trial (ClinicalTrials. given as intramuscular injections every 4 weeks or every
gov identifier, NCT01641809), a two-drug regimen of once- 8 weeks, compared with that of oral cabotegravir plus
daily oral formulations of cabotegravir and rilpivirine abacavir–lamivudine, as maintenance therapy through
demonstrated durable viral suppression in patients whose 96 weeks for individuals who had achieved successful
viral load was previously suppressed to less than 50 HIV-1 HIV-1 viral suppression with oral cabotegravir plus
RNA copies per mL by treatment with cabotegravir and abacavir–lamivudine.

1500 www.thelancet.com Vol 390 September 23, 2017

 Articles

Methods with blocks shared across sites, was used to conceal the
Study design and participants allocation schedule and prevent selection bias.
LATTE-2 is an ongoing phase 2b, randomised, An open-label design was used because a double-blind,
multicentre, open-label, non-inferiority, parallel-group double-dummy design would have resulted in an
trial, consisting of a 20-week induction period, 96-week increased pill burden in all patients, a requirement for
maintenance period, extension period, and long-term sham injections, elevated risk of oral ART non-adherence
follow-up period. The study was done at 50 sites in for oral comparator group patients receiving sham
the USA, Canada, Spain, France, and Germany. injections, limitations to patient-reported preference data
Patients who were HIV-1 positive, were aged 18 years or comparing injectable and oral ART, as well as
older, and had no more than 10 days of previous ART considerable trial design complexities.
treatment, with screening HIV-1 RNA of at least
1000 copies per mL and CD4+ T-cell counts of at least Procedures
200 cells per mm³, were eligible for inclusion. Key Planned analyses were done after all patients had
exclusion criteria included the presence of any major completed weeks 32, 48, and 96 of the maintenance
antiretroviral resistance-associated mutation, pregnancy, period (or discontinued earlier).
moderate or severe hepatic impairment, clinically Adverse events were graded according to the Division of
relevant hepatitis, hepatitis B infection, laboratory values AIDS Table for Grading the Severity of Adverse and
of clinical concern, creatinine clearance less than Pediatric Events (2009).17 Serious adverse events were
50 mL/min, and a need for chronic anticoagulants. defined as any untoward medical occurrence that resulted
Eligible patients received the induction period regimen in death, was life-threatening, required hospitalisation,
of oral cabotegravir 30 mg plus abacavir–lamivudine resulted in disability or incapacity, was a congenital
600 mg–300 mg once daily for 20 weeks. Rilpivirine anomaly or birth defect, or met predefined liver injury
25 mg once daily was added 4 weeks before randomisation criteria. Liver stopping criteria were met when alanine
(week −4 [week 16 of the induction period]) and continued aminotransferase values met or exceeded the upper limit
until the first injection visit (day 1). Patients who tolerated of normal (ULN) by eight times, five times for 14 days, or
the induction period regimen and achieved plasma three times with bilirubin at least two times the ULN (if
HIV-1 RNA less than 50 copies per mL at week −4 were >35% direct bilirubin, bilirubin fractionation is required).
eligible to enter the maintenance period at day 1 and were Pharmacokinetic samples for cabotegravir and rilpivirine
randomly assigned to receive intramuscular injections were collected at day 1 and at weeks 1, 4, 8, 12, 16, 20, 24,
every 4 weeks (long-acting cabotegravir 400 mg plus 25, 28, 32, 36, 40, 41, 44, and 48.
rilpivirine 600 mg; two 2 mL injections) or every 8 weeks Treatment satisfaction was measured using the HIV
(long-acting cabotegravir 600 mg plus rilpivirine 900 mg; Treatment Satisfaction Questionnaire (HIVTSQ). The
two 3 mL injections), with a provision for a 14-day dosing HIVTSQ status version (HIVTSQ[s]) was completed by
window, or to continue receiving oral cabotegravir 30 mg patients at weeks –16 and –4 of the induction period and
plus abacavir–lamivudine once daily for 96 weeks. Long- at day 1 (long-acting predose) and weeks 8, 32, 48, and 96
acting injectable formulations contained 200 mg per mL of the maintenance period or at withdrawal.
of cabotegravir and 300 mg per mL of rilpivirine for
administration as two separate intramuscular injections Outcomes
into the gluteus medius muscle at each dosing visit. Both The primary endpoints were the proportion of patients in
4-week and 8-week dosing regimens had an initial loading the maintenance-exposed population (which consisted of
dose of cabotegravir 800 mg (two 2 mL injections). randomly assigned patients who received at least one
The study was done in accordance with the principles dose of study drug during the maintenance period) with
of the Declaration of Helsinki. Written informed consent HIV-1 RNA less than 50 copies per mL at maintenance
was obtained from all participants, and the protocol was week 32 (using the US Food and Drug Administration
approved by the institutional review board of each study [FDA] snapshot algorithm), the proportion of patients
site. The authors vouch for the accuracy and completeness with protocol-defined virological failures, and incidence
of the data and the analyses and for the fidelity of the and severity of adverse events and laboratory
study to the protocol. abnormalities.
Secondary endpoints evaluated the proportion of
Randomisation and masking patients with plasma HIV-1 RNA less than 200 copies per
A computer-generated allocation sequence created by mL and less than 50 copies per mL; incidence of
the validated software, RandALL (version 2.10; treatment-emergent viral resistance; absolute values and
GlaxoSmithKline, Research Triangle Park, NC, USA) was change from baseline in plasma HIV-1 RNA; absolute
used to randomly assign patients at day 1, with values and changes from baseline in CD4+ cell counts;
stratification by HIV-1 RNA (<50 copies per mL, yes or incidence of disease progression; incidence and severity
no) before week −8 (ie, during the first 12 weeks of of adverse events and laboratory abnormalities over time;
induction period treatment). Central randomisation, absolute values and changes in laboratory parameters

www.thelancet.com Vol 390 September 23, 2017 1501

 Articles

through the week 96 analysis; evaluation of plasma dis­continued from the study; those who had received
pharmacokinetic parameters and steady-state deter­ one or more injections entered a 52-week long-term
minations for cabotegravir and rilpivirine, as well as follow-up period.
pharmacokinetic–pharmacodynamic relationships with
safety and antiviral activity; and treatment satisfaction Statistical analysis
and medication adherence of participants using patient- The primary analysis of the maintenance-exposed
reported outcome questionnaires. population used a Bayesian approach to evaluate the
Protocol-defined virological failure following random­ hypothesis that the proportion of patients with HIV-1
isation was defined as having two consecutive plasma RNA less than 50 copies per mL (FDA snapshot
HIV-1 RNA measurements of at least 200 copies algorithm) at week 32 (and repeated at week 48) for each
per mL. Patients who met the definition of protocol- intramuscular long-acting regimen is not worse than the
defined virological failure before receiving oral regimen proportion by more than 10% (denoted
any cabotegravir plus rilpivirine injections were comparable). This Bayesian approach provided an

386 patients screened

76 failed screening*
65 did not meet inclusion
or exclusion criteria
10 withdrew consent
2 investigator discretion
1 lost to follow-up

309 patients entered induction phase†

21 discontinued
3 adverse event
5 lack of efficacy
2 protocol deviation
3 met stopping criteria
2 lost to follow-up
1 investigator discretion
5 withdrew consent

286 patients entered maintenance phase‡

115 patients received ≥1 dose of 115 patients received ≥1 dose of 56 patients received ≥1 dose of
intramuscular cabotegravir LA intramuscular cabotegravir LA oral cabotegravir plus
plus rilpivirine LA every 4 weeks plus rilpivirine LA every 8 weeks abacavir–lamivudine

11 withdrew 4 withdrew 6 withdrew

7 adverse event 1 adverse event 1 adverse event
2 protocol deviation 1 lack of efficacy 1 lack of efficacy
1 met stopping criteria 1 investigator discretion§ 1 met stopping criteria
1 withdrew consent 1 withdrew consent 1 lost to follow-up
2 withdrew consent

104 ongoing at time of analysis 111 ongoing at time of analysis 50 ongoing at time of analysis
(week 48) (week 48) (week 48)

3 withdrew 1 withdrew consent 3 withdrew consent

1 adverse event
2 withdrew consent

101 ongoing at time of analysis 110 ongoing at time of analysis 47 ongoing at time of analysis
(week 96) (week 96) (week 96)

Figure 1: Trial profile

LA=long-acting. *Multiple reasons for screening failure were reported for two patients. †310 patients were enrolled, but one patient withdrew consent after baseline
visit procedures were done and before study treatment was initiated. ‡Two patients completed the induction period (and had day 1 assessments) but did not enter
the maintenance period and were not randomly assigned (because of investigator discretion and lack of efficacy). §Patient experienced suspected protocol-defined
virological failure at the time of withdrawal, which was subsequently confirmed.

1502 www.thelancet.com Vol 390 September 23, 2017

 Articles

estimate of the (posterior) probability that the hypothesis maintenance-exposed population, whereas data for the
is true given the observed data and pre-trial information.18 induction period and study population characteristics are
A Bayesian posterior probability of at least 90% was summarised using the intention-to-treat exposed
prespecified as the decision rule for claiming population (patients who received at least one study dose
comparability for each comparison. To incorporate during the induction period [week −20 to day 1]).
previous information for the oral group response rate Change from baseline (last value collected up to and
based on data from the LATTE study,13 a beta (23, 2) prior including the date of first induction period treatment at
distribution was assumed to reflect the belief that the week –20) in CD4+ cell count, plasma HIV-1 RNA, and
oral group response rate was between 78% and 99% with patient-reported outcome endpoints are summarised
95% confidence, and a non-informative prior distribution using observed data with no imputation for missing data.
was assumed for the intramuscular response rate. Laboratory abnormalities are presented as maintenance
Sample sizes of 45 patients in the oral cabotegravir plus period treatment emergent, which refers to graded toxic
abacavir–lamivudine group and 90 patients each in the effects that developed or increased in intensity while on
intramuscular groups were chosen to ensure a high treatment in the maintenance period relative to the last
probability that a two-drug long-acting regimen with recorded toxic effect up to and including the date of the
poor response relative to oral cabotegravir plus abacavir– first dose of the maintenance period treatment.
lamivudine once daily would be identified. With the Antiretroviral plasma concentrations at the above-
chosen number of patients per treatment group, and mentioned timepoints following long-acting intra­
assuming true response rates of 82% for long-acting muscular cabotegravir plus rilpivirine administration
intramuscular cabotegravir plus rilpivirine versus 92% were analysed using validated LC-MS/MS methods and
for oral cabotegravir plus abacavir–lamivudine, there was are summarised over time using evaluable data that met
a low probability of falsely concluding that long-acting sample collection window criteria, excluding samples
intramuscular cabotegravir plus rilpivirine was affected by dosing errors (wrong dose) or oral bridging.
comparable with oral cabotegravir plus abacavir– Sampling windows for intramuscular dosing were set
lamivudine (simulated probability=0·064). Each long- relative to the previous injection as follows: 0·5 h for 2 h
acting intramuscular regimen was evaluated against the post-dose samples; 1 day for 1-week post-injection visits;
oral cabotegravir plus abacavir–lamivudine regimen in 2 days for predose samples in the 4-week group, weeks 4
the primary analysis; comparability in antiviral response
rates between the two intramuscular regimens was Intramuscular Intramuscular Oral cabotegravir Total (n=286)
assessed as a prespecified key secondary comparison. cabotegravir LA cabotegravir LA plus abacavir–
Normal approximation 95% CIs for the difference in the plus rilpivirine LA plus rilpivirine LA lamivudine
every 4 weeks every 8 weeks (n=56)
proportions (each long-acting intramuscular regimen vs (n=115) (n=115)
oral regimen) are provided as supportive secondary
analyses (prespecified but without adjustment for Age (years; range) 36 (19–62) 35 (20–64) 35 (19–57) 35 (19–64)

multiple testing). These efficacy analyses used Sex

the maintenance-exposed population. Per-protocol Male 109 (95%) 107 (93%) 46 (82%) 262 (92%)
sensitivity analyses excluding participants with Female 6 (5%) 8 (7%) 10 (18%) 24 (8%)
prespecified protocol deviations were not done as fewer Ethnic origin
than 5% of participants had such deviations (threshold White 94 (82%) 93 (81%) 39 (70%) 226 (79%)
for conducting analysis specified in advance in the African American 12 (10%) 17 (15%) 15 (27%) 44 (15%)
analysis plan). or African heritage
In the snapshot analysis, participants with a last HIV-1 Other 9 (8%) 5 (4%) 2 (4%) 16 (6%)
RNA result less than 50 copies per mL in the analysis Baseline HIV-1 RNA
timepoint window were classified as responders.
Log10 copies per mL 4·46 4·42 4·29 4·39
Participants without HIV-1 RNA data at the visit of (4·00–4·97) (4·05–4·80) (4·01–4·74) (4·03–4·83)
interest (due to missing data or discontinuation of study ≥100 000 copies per mL 28 (24%) 16 (14%) 7 (12%) 51 (18%)
drug before analysis visit window) and patients who
Baseline CD4+ cell count 499 449 518 489
switched ART after week 4 were classified as non- (cells per mm3) (359–624) (343–618) (417–630) (359–624)
responders. For week 32, if a participant had no data Hepatitis C co-infection 5 (4%) 3 (3%) 2 (4%) 10 (3%)
within a window of –2 to +2 weeks, then an expanded
NRTI during induction
window of –6 to +6 weeks was used. For week 48, if a
Abacavir–lamivudine 107 (93%) 107 (93%) 53 (95%) 267 (93%)
participant had no data within a window of −2 to +6 weeks,
then an expanded window of –6 to +6 weeks was used. Tenofovir–emtricitabine 8 (7%) 8 (7%) 3 (5%) 19 (7%)
For week 96, a window of –6 to +6 weeks was used. Data are median (IQR) or n (%) unless stated otherwise. LA=long-acting. NRTI=nucleoside reverse transcriptase
Data for the maintenance and induction plus inhibitor.
maintenance periods combined, respectively, are
Table 1: Baseline demographics and disease characteristics (maintenance-exposed population)
summarised by randomised group using the

www.thelancet.com Vol 390 September 23, 2017 1503

 Articles

Induction period Maintenance period

100
Proportion of patients with virological suppression (%)

80

Outcomes (n, %) at 96 weeks Intramuscular Intramuscular Oral cabotegravir

cabotegravir LA cabotegravir LA plus abacavir–lamivudine
60 plus rilpivirine LA plus rilpivirine LA (n=56)
every 4 weeks every 8 weeks
(n=115) (n=115)

Virological response 100 (87%) 108 (94%) 47 (84%)

40 Virological non-response 0 5 (4%) 1 (2%)
Data in window not below threshold 0 2 (2%) 0
Discontinued for lack of efficacy 0 1 (<1%) 1 (2%)
Discontinued for other reason while below threshold 0 2 (2%) 0
20 No virological data 15 (13%) 2 (2%) 8 (14%)
Discontinued due to adverse event or death 9 (8%) 1 (<1%) 2 (4%)
Discontinued for other reasons 5 (4%) 1 (<1%) 6 (11%)
Missing data during window but on study 1 (<1%) 0 0
0
as 20
ee )
W –16

W 12

W –8
4
y1

k4

W 8

W 2

W 6

W 20

W 4

W 8

W 32

W 36

W 0

W 44
8

6
W line

k1

k7
k4

k8
k–

k2

k6
k1

k2

k4

k5

k8

k9
k
k–
(b ek –

Da
k

ee

k
ee

k
ee
ee
k

ee

ee
ee
ee

ee

ee

ee

ee

ee
ee

ee
ee

ee
e

ee

ee

ee
ee

W
e

W
W

W
W

Study visit
Oral cabotegravir plus abacavir–lamivudine induction (maintenance-exposed population)
Intramuscular cabotegravir LA plus rilpivirine LA every 8 weeks (n=115)
Intramuscular cabotegravir LA plus rilpivirine LA every 4 weeks (n=115)
Oral cabotegravir plus abacavir–lamivudine (n=56)

Figure 2: Proportion of patients with HIV-1 RNA concentration less than 50 copies per mL (FDA snapshot algorithm) by visit in the maintenance-exposed
population and snapshot outcomes at week 96
Error bars show 95% CIs, derived using the normal approximation. FDA=US Food and Drug Administration. LA=long-acting.

and 8 cabotegravir samples in the 8-week group, and Results

midcycle concentrations 4 weeks post-injection in the The first patient was screened in April 28, 2014, and the
8-week group; and 4 days for predose samples in the last patient’s week 96 visit occurred in Nov 10, 2016. Of
8-week group (except for weeks 4 and 8 cabotegravir 386 patients screened, 309 were enrolled in the study
samples, as described earlier). Sparse pharmacokinetic (figure 1). 282 (91%) patients were male, average age
sampling obtained after week 48 and through week 96 was 36·6 years (SD 10·4). 72 (23%) patients had a baseline
not assayed as a matter of routine and therefore is not CD4+ cell count of no more than 350 cells per mm³, and
reported. Plasma concentrations at week 48 are 60 (19%) patients had a baseline HIV-1 RNA of at least
summarised using geometric means and associated 100 000 copies per mL. Baseline characteristics were
95% CIs, whereas plasma concentration–time profiles are balanced among the three dosing groups.
summarised using arithmetic mean (SD). Relationships 288 patients (93%) completed the 20-week induction
between cabotegravir and rilpivirine trough concentrations period (intention-to-treat exposed). 21 patients (7%)
at week 48 (or last available value before week 48 if no discontinued treatment during the induction period
week 48 data available) and week 48 virological failure (five withdrew consent, five for lack of efficacy, three for
were explored graphically in a post-hoc analysis. adverse events, three met predefined liver chemistry
stopping criteria, two for protocol deviation, two were
Role of the funding source lost to follow-up, and one at investigator discretion).
This study was funded by ViiV Healthcare and Janssen 286 patients qualified for and entered the maintenance
R&D. The funders participated in the study design, data period. Patients were randomly assigned (2:2:1) to receive
gathering, analysis, and interpretation. All listed authors intramuscular injections of long-acting cabotegravir plus
meet the criteria for authorship set forth by the rilpivirine every 4 weeks (n=115; 4-week group) or every
International Committee of Medical Journal Editors. All 8 weeks (n=115; 8-week group) or continue the oral
authors had full access to the data and are responsible for cabotegravir plus abacavir–lamivudine regimen (n=56;
the veracity and completeness of the reported data. The oral treatment group; table 1). In the maintenance period,
corresponding author had final responsibility for the 14 (12%) patients in the 4-week group, five (4%) in the
decision to submit the manuscript for publication. 8-week group, and nine (16%) in the oral treatment group

1504 www.thelancet.com Vol 390 September 23, 2017

 Articles

withdrew from the study (figure 1). The most common at week 48 while not suppressed (and subsequently
reasons for withdrawal during the maintenance period confirmed as a protocol-defined virological failure). One
were adverse events (ten patients [3%]) and withdrawn of the two patients with HIV-1 RNA less than 50 copies See Online for appendix
consent (ten patients [3%]). Eight of the ten patients who
withdrew during the maintenance period because of
adverse events were in the 4-week group. Intramuscular Intramuscular Oral cabotegravir
cabotegravir LA plus cabotegravir LA plus abacavir–
During the 20-week induction period, oral treatment
rilpivirine LA every plus rilpivirine LA lamivudine
induced viral suppression (HIV-1 RNA <50 copies 4 weeks (n=115) every 8 weeks (n=56)
per mL) in 282 (91%) patients by day 1 (week 20 of (n=115)
induction period), with 279 (90%) patients achieving less Grade Grade Grade Grade Grade Grade
than 50 copies per mL of HIV-1 RNA within the first 1–4† 3–4‡ 1–4† 3–4‡ 1–4† 3–4‡
8 weeks of treatment. One patient met the criteria for Total adverse events*
protocol-defined virological failure (two consecutive Any event 115 (100%) 21 (18%) 115 (100%) 24 (21%) 54 (96%) 7 (13%)
plasma HIV-1 RNA measurements ≥200 copies per mL), Injection-site pain 112 (97%) 6 (5%) 110 (96%) 8 (7%) 0 0
as a result of poor medication compliance, without Nasopharyngitis 39 (34%) 0 35 (30%) 0 22 (39%) 0
experiencing treatment-emergent resistance. Injection-site nodule 36 (31%) 1 (<1%) 29 (25%) 1 (<1%) 0 0
Following randomisation, 108 (94%) of 115 patients Injection-site swelling 34 (30%) 0 29 (25%) 1 (<1%) 0 0
achieved the primary efficacy endpoint (plasma HIV-1 Diarrhoea 32 (28%) 0 27 (23%) 0 11 (20%) 0
RNA <50 copies per mL; FDA snapshot algorithm) at Injection-site pruritus 33 (29%) 0 25 (22%) 0 0 0
week 32 in the 4-week group, 109 (95%) of 115 in the Headache 27 (23%) 0 29 (25%) 1 (<1%) 14 (25%) 1 (2%)
8-week group, and 51 (91%) of 56 in the oral treatment Injection-site induration 25 (22%) 0 29 (25%) 1 (<1%) 0 0
group (figure 2; appendix p 1). Both long-acting regimens Injection-site warmth 21 (18%) 0 23 (20%) 1 (<1%) 0 0
met prespecified efficacy criteria for demonstrating Upper respiratory tract 13 (11%) 0 23 (20%) 0 7 (13%) 0
comparability relative to the oral comparator group infection
(posterior probability for comparability >90%; appendix Injection-site bruising 14 (12%) 0 20 (17%) 0 0 0
p 2). Treatment differences at week 32 for each group Nausea 18 (16%) 0 16 (14%) 0 9 (16%) 0
receiving long-acting injections compared with daily oral Injection-site erythema 19 (17%) 0 13 (11%) 1 (<1%) 0 0
treatment were 2·8% (95% CI −5·8 to 11·5) for the 4-week Pyrexia 16 (14%) 0 16 (14%) 0 3 (5%) 0
group and 3·7% (−4·8 to 12·2) for the 8-week group. At Gastroenteritis 15 (13%) 0 14 (12%) 0 6 (11%) 1 (2%)
week 48, comparability between regimens was confirmed Fatigue 14 (12%) 0 14 (12%) 0 4 (7%) 0
(appendix p 2), with virological suppression achieved in Syphilis 11 (10%) 0 17 (15%) 0 6 (11%) 0
105 (91%) of 115 patients in the 4-week group, 106 (92%) Back pain 13 (11%) 0 15 (13%) 0 10 (18%) 0
of 115 in the 8-week group, and 50 (89%) of 56 in the oral Insomnia 13 (11%) 0 12 (10%) 0 4 (7%) 0
treatment group (figure 2; appendix p 1). Through Bronchitis 12 (10%) 1 (<1%) 12 (10%) 0 6 (11%) 0
96 weeks of maintenance treatment, 100 (87%) of
Cough 13 (11%) 0 11 (10%) 0 7 (13%) 0
115 patients in the 4-week group, 108 (94%) of 115 in the
Influenza 16 (14%) 0 6 (5%) 0 2 (4%) 0
8-week group, and 47 (84%) of 56 in the oral treatment
Arthralgia 10 (9%) 0 12 (10%) 0 4 (7%) 0
group maintained virological suppression (figure 2;
Anogenital warts 11 (10%) 0 9 (8%) 0 2 (4%) 0
appendix p 1). Virological non-response, as defined by the
Pharyngitis 8 (7%) 0 12 (10%) 0 5 (9%) 0
FDA snapshot algorithm, at week 96 occurred in
Respiratory tract infection 11 (10%) 0 6 (5%) 0 6 (11%) 0
six patients (five in the 8-week group, one in the oral
Asthenia 10 (9%) 0 7 (6%) 0 9 (16%) 0
treatment group; figure 2). Through 96 weeks, no patients
Treatment-related adverse events*
in the 4-week group failed for virological reasons. In
Any event 113 (98%) 10 (9%) 110 (96%) 10 (9%) 21 (38%) 1 (2%)
comparison, at week 48, ten patients met virological non-
Injection-site pain 112 (97%) 6 (5%) 109 (95%) 8 (7%) 0 0
response criteria (one in the 4-week group, eight in the
8-week group, one in the oral treatment group). For the Injection-site nodule 35 (30%) 1 (<1%) 29 (25%) 1 (<1%) 0 0

8-week group, four patients with virological non-response Injection-site swelling 34 (30%) 0 29 (25%) 1 (<1%) 0 0

at week 48 (HIV-1 RNA >50 copies per mL) were Injection-site pruritus 33 (29%) 0 24 (21%) 0 0 0
resuppressed with HIV-1 RNA less than 50 copies per mL Injection-site induration 25 (22%) 0 28 (24%) 1 (<1%) 0 0
at week 96 without a change in therapy. Of the five Injection-site warmth 21 (18%) 0 22 (19%) 1 (<1%) 0 0
patients in the 8-week group with virological non- Injection-site bruising 14 (12%) 0 19 (17%) 0 0 0
response at week 96, two had HIV-1 RNA of at least Injection-site erythema 19 (17%) 0 12 (10%) 1 (<1%) 0 0
50 copies per mL at week 96 (one of whom had HIV-1 Nausea 12 (10%) 0 8 (7%) 0 5 (9%) 0
RNA ≥50 copies per mL at week 48), one discontinued Headache 7 (6%) 0 6 (5%) 0 4 (7%) 0
due to protocol-defined virological failure at week 4, Pyrexia 7 (6%) 0 5 (4%) 0 0 0
one withdrew consent due to intolerability of injections at (Table 2 continues on next page)
week 8, and one withdrew due to investigator discretion

www.thelancet.com Vol 390 September 23, 2017 1505

 Articles

period by a median of 226 cells per mm³ (IQR 145–393) in

Intramuscular Intramuscular Oral cabotegravir
cabotegravir LA plus cabotegravir LA plus abacavir– the 4-week group (n=100), 239 cells per mm³ (111–359) in
rilpivirine LA every plus rilpivirine LA lamivudine the 8-week group (n=109), and 317 cells per mm³ (214–505)
4 weeks (n=115) every 8 weeks (n=56) in the oral treatment group (n=47).
(n=115)
During the induction and maintenance period, in the
Grade Grade Grade Grade Grade Grade maintenance-exposed population, total adverse events of
1–4† 3–4‡ 1–4† 3–4‡ 1–4† 3–4‡
any grade and attribution occurred in 115 (100%) patients
(Continued from previous page) in the 4-week group, 115 (100%) in the 8-week group, and
Injection-site 6 (5%) 0 3 (3%) 0 0 0 54 (96%) in the oral treatment group (table 2). Injection-
discolouration
site pain, the most common injection-site reaction, was
Dyspepsia 6 (5%) 0 1 (<1%) 0 1 (2%) 0
the most frequently reported adverse event in the
Asthenia 3 (3%) 0 2 (2%) 0 3 (5%) 0
intramuscular groups (112 [97%] patients in the 4-week
Data are n (%). LA=long-acting. *Includes all post-baseline induction period and maintenance period adverse events, group, 110 [96%] patients in the 8-week group). Most
as well as long-term follow-up period adverse events for patients withdrawing from intramuscular dosing that injection-site reactions were mild (grade 1; 3648 [84%] of
occurred within 35 or 63 days (4-week group or 8-week group) of the last maintenance period intramuscular injection
until up to and including the start date of the long-term follow-up period on oral highly active antiretroviral treatment.
4360 injections) or moderate (grade 2; 673 [15%] of 4360
†At least 10% in any treatment group for total adverse events and at least 5% in any treatment group for injections) in intensity, with median symptom duration
treatment-related adverse events. ‡Includes only events listed in the grade 1–4 column; other grade 3–4 events that of 3 days (appendix p 3). The most commonly reported
did not meet the 5% or 10% cutoff for the grade 1–4 column are not shown.
adverse events other than an injection-site reaction were
Table 2: Summary of total adverse events and treatment-related adverse events through week 96 in the nasopharyngitis (39 patients [34%] in the 4-week group,
safety maintenance population 35 [30%] in the 8-week group, and 22 [39%] in the oral
treatment group), diarrhoea (32 [28%] in the 4-week
group, 27 [23%] in the 8-week group, and 11 [20%] in the
per mL at week 96 remained on the study and had HIV-1 oral treatment group), and headache (27 [23%] in the
concentrations less than 50 copies per mL at the next 4-week group, 29 [25%] in the 8-week group, 14 [25%] in
scheduled visit. The patient in the 4-week group with the oral treatment group). Serious adverse events
virological non-response at week 48 (HIV-1 RNA, occurred in 13 (11%) patients in each of the intramuscular
59 copies per mL) remained on the study beyond week 48 treatment groups and nine (16%) patients in the oral
and had subsequent viral resuppression with HIV-1 RNA treatment group, only one of which was drug related
less than 50 copies per mL at week 96. The virological (migraine, which occurred in the initial oral induction
non-responder in the oral treatment group discontinued period of the study). During the maintenance period,
as a result of protocol-defined virological failure at week 8. serious adverse events occurred in 11 (10%) patients in
Three patients (two in the 8-week group, [week 4 and each of the intramuscular groups compared with
week 48], one in the oral treatment group [week 8]) met seven patients (13%) in the oral treatment group.
the criteria for protocol-defined virological failure However, none was considered to be related to study
through week 96. Viral genotyping analysis for the treatment. 11 patients (4%) developed an adverse event
patient in the oral treatment group had no treatment- during the maintenance period, which led to withdrawal:
emergent resistance mutations in the genes encoding eight patients (7%) in the 4-week group, two (2%) in the
viral reverse transcriptase, protease, or integrase. Of the 8-week group, and one (2%) in the oral treatment group.
two patients in the 8-week group, a mixture emerged for Two patients (both in the 8-week group) had injection-
one at integrase codon 269 (R269R/G), which did not site reactions leading to withdrawal within 8 weeks of
decrease cabotegravir susceptibility. The second patient initiating dosing. Two deaths occurred during the study:
harboured virus with treatment-emergent reverse one resulted from a motor vehicle accident that occurred
transcriptase mutations K103N, E138G, and K238T, with in the initial oral induction period of the study, and the
phenotypic resistance to efavirenz, rilpivirine, and second occurred after an epileptic seizure in a patient in
nevirapine, and an integrase mutation Q148R, with the 4-week intramuscular treatment group who had
phenotypic resistance to raltegravir, elvitegravir, and received 48 weeks of cabotegravir and 32 weeks of
cabotegravir, while remaining sensitive to dolutegravir. rilpivirine treatment. The seizure event was not
Most patients in all groups maintained HIV-1 RNA less considered likely to be related to cabotegravir or
than 200 copies per mL (100 [87%] in the 4-week group, rilpivirine on the basis of the late time to onset of
110 [96%] in the 8-week group, and 47 [84%] in the oral symptoms and both direct and circumstantial evidence
treatment group). The mean change from baseline in of recreational drug use proximal to the event.
plasma HIV-1 RNA concentration was –2·89 log10 copies Progression of HIV disease was uncommon. One
per mL (SD 0·713) for patients in the 4-week group, patient in the 4-week group experienced disease
–2·77 log10 copies per mL (0·602) in the 8-week group, progression from Centers for Disease Control and
and –2·77 log10 copies per mL (0·582) in the oral treatment Prevention (CDC) Class A to death (epileptic seizure).
group. At week 96 of the maintenance period, CD4+ cell Four patients in the 8-week group experienced disease
counts increased from the beginning of the induction progression from CDC Class A to Class C (one had

1506 www.thelancet.com Vol 390 September 23, 2017

 Articles

A B
100 Intramuscular cabotegravir LA plus rilpivirine LA every 4 weeks 1000 Intramuscular cabotegravir LA plus rilpivirine LA every 4 weeks
Intramuscular cabotegravir LA plus rilpivirine LA every 8 weeks Intramuscular cabotegravir LA plus rilpivirine LA every 8 weeks
PA-IC90 PA-IC90
10 mg orally Cτ 25 mg orally Cτ
30 mg orally Cτ
Mean plasma cabotegravir concentration (SD; µg/mL)

Mean plasma rilpivirine concentration (SD; ng/mL)

10

100

0·1 10
01 4 8 12 16 20 24 28 32 36 40 44 48 01 4 8 12 16 20 24 28 32 36 40 44 48
Week number Week number

Figure 3: Arithmetic mean (SD) plasma concentration–time profiles following every 4 weeks and every 8 weeks administration of (A) cabotegravir LA and (B) rilpivirine LA through week 48
Cτ=concentration at the end of dosing interval. LA=long-acting. PA-IC90=protein-adjusted 90% inhibitory concentration.

Kaposi’s sarcoma, one had pneumonia, and two had (86·6–103·4) for the 4-week group and 64·48 ng/mL
herpes simplex). No patient experienced disease (60·0–69·3) for the 8-week group, which were
progression to CDC Class C or death in the oral eight times and five times greater than in-vitro PA-IC90 of
treatment group. 12 ng/mL against wild-type HIV-1. Accumulation of
Grade 3 or more severe maintenance period treatment- rilpivirine was observed through 24–48 weeks of dosing,
emergent laboratory abnormalities occurred in 32 patients with the lowest rilpivirine trough concentrations
(28%) in the 4-week group, 21 (18%) in the 8-week group, observed after initial intramuscular injections in both
and 12 (21%) in the oral treatment group. Grade 3 or more groups (figure 3). No relationship was observed between
severe maintenance period treatment-emergent ALT cabotegravir concentrations and virological non-
elevations occurred in four (3%) patients in each of the response at week 48. In a post-hoc analysis of seven of
4-week and 8-week groups, and in three (5%) patients the nine patients in the 8-week group with virological
receiving oral treatment, largely attributable to acute non-response at week 48 (FDA snapshot algorithm),
hepatitis C infections. Among patients meeting pre- rilpivirine trough concentrations at week 48 (or last
defined liver stopping criteria, possible drug-induced available trough concentration before week 48) were in
liver injury occurred in two patients (both treated with the lowest 25th quartile among all rilpivirine samples,
oral cabotegravir plus abacavir–lamivudine; one during whereas cabotegravir trough concentrations were
the induction period [before randomisation], and one distributed throughout the quartile ranges. Of the
during the maintenance period). In both cases, liver two patients in the 8-week group with protocol-defined
chemistry abnormalities resolved following treatment virological failure during the maintenance period
discontinuation, and the patients remained clinically through week 48, one patient had non-quantifiable
asymptomatic. Mean change from baseline was evaluated rilpivirine concentration 4 weeks post injection without
across all laboratory parameters, and no clinically treatment-emergent resistance; there was no clear
significant differences were observed through 96 weeks. correlation between cabotegravir or rilpivirine
At week 48, cabotegravir geometric mean trough concentrations at the time of viral rebound with the
concentrations (C0; 95% CI) were 2·58 μg/mL (2·4–2·8) other patient.
for the 4-week group, 1·46 μg/mL (1·3–1·6) for the High levels of treatment satisfaction were observed
8-week group, and 4·47 μg/mL (3·9–5·2) for the oral across all treatment groups (figure 4). At week 96, patients
treatment group, which were 16 times, nine times, and reported very high levels of satisfaction across all three
27 times greater than the in-vitro protein-adjusted groups (4-week, 8-week, and oral treatment, via the
90% inhibitory concentration (PA-IC90) of 0·166 μg/mL HIVTSQ[s]), with 246 (97%) of 254 patients selecting a
against wild-type HIV-1. For rilpivirine, week 48 score of 5 or 6 on a 6-point satisfaction scale. A similar
geometric mean C0 (95% CI) values were 94·64 ng/mL percentage of patients in each of the intramuscular groups

www.thelancet.com Vol 390 September 23, 2017 1507

 Articles

A How satisfied are you with your current treatment? How satisfied would you be to continue with your present form of
treatment?
<1% <1% 2%
100 <1% <1% 4% <1%
<1% 2%
<1% 11%
13% 8%
14%
80 20%

29%

60 33%
Response (%)

40

20

79% 83% 67% 85% 88% 55%

0
Every 4 weeks Every 8 weeks Oral cabotegravir plus Every 4 weeks Every 8 weeks Oral cabotegravir plus
(n=103) (n=109) abacavir–lamivudine (n=103) (n=109) abacavir–lamivudine
(n=49) (n=49)
Intramuscular cabotegravir LA plus rilpivirine LA Intramuscular cabotegravir LA plus rilpivirine LA
6 5 4 3 2 1 0
Very satisfied Very dissatisfied

B How satisfied are you with your current treatment? How satisfied would you be to continue with your present form of
treatment?
100 2% 2%
3% <1% 1% <1%
2% 4%
4% 11% 10% 7%
14%
80 21% 15% 9%

60
Response (%)

35%
40

20

76% 85% 76% 88% 89% 43%

0
Every 4 weeks Every 8 weeks Oral cabotegravir plus Every 4 weeks Every 8 weeks Oral cabotegravir plus
(n=100) (n=108) abacavir–lamivudine (n=100) (n=108) abacavir–lamivudine
(n=46) (n=46)
Intramuscular cabotegravir LA plus rilpivirine LA Intramuscular cabotegravir LA plus rilpivirine LA
6 5 4 3 2 1 0
Very satisfied Very dissatisfied

Figure 4: Summary of patient-reported outcomes at (A) week 48 (maintenance treatment) and (B) week 96
The data are based on the observed case dataset of patients who completed questionnaires at week 48 and week 96 (HIV Treatment Satisfaction Questionnaire,
status version). LA=long-acting.

(≥99%; 99 of 100 in the 4-week group and 107 of 108 in the regimens in patients with HIV-1 infection and contributes
8-week group) reported they would be highly satisfied to to the growing number of studies evaluating simplification
continue their current long-acting regimen, while a lower to two-drug therapy. Both the long-acting injectable
percentage would elect to continue on oral dosing (78%; 4-week and 8-week regimens maintained virological
36 of 46 patients in the oral treatment group). Patients suppression at rates comparable to oral daily three-drug
who discontinued for any reason before week 96 did not ART, with two protocol-defined virological failures
complete the questionnaire at this timepoint, introducing occurring among the 230 patients who received long-
a small degree of selection bias in these results. acting therapy during the 96-week maintenance period.
The long-acting cabotegravir plus rilpivirine regimens
Discussion were generally well tolerated, with no drug-related serious
LATTE-2 is the first study to analyse the efficacy and adverse events and few adverse event-related withdrawals.
safety of fully injectable two-drug long-acting ART Although injection-site reactions were common, they

1508 www.thelancet.com Vol 390 September 23, 2017

 Articles

were transient in nature, mild or moderate in severity, option for patients burdened by life-long daily oral
and did not appear to compromise high levels of patient- medication compliance.
reported satisfaction. The long-term acceptability of This study has some limitations. Although the LATTE-2
administering chronic intra­ muscular injections to study population included patients from five countries,
patients was also shown in LATTE-2, with very few participants were predominantly male and were
withdrawals resulting from injection-site reactions, restricted at entry to CD4+ cell count of at least 200 cells
two patients (<1%) through 96 weeks. per mm³, which does not accurately represent the global
Treatment with the 4-week and 8-week regimens HIV-infected population.1 Therefore, the efficacy,
maintained virological suppression in 87% (n=100) and safety, and pharmacokinetic outcomes of long-acting
94% (n=108) of patients, respectively, compared with cabotegravir plus rilpivirine in different subpopulations
84% (n=47) of patients treated with oral cabotegravir plus infected with HIV-1 needs further evaluation.
abacavir–lamivudine. These proportions are consistent Additionally, LATTE-2 was an open-label study, and
with multiple other studies examining efficacy of cabotegravir was included in both the intramuscular and
switching oral regimens in patients with viral oral comparator group, possibly influencing adverse
suppression.19–21 The LATTE-2 study was unique compared event reporting and limiting adverse event comparisons
with previous switch studies in its evaluation of short- relative to other, standard-of-care ART. Lastly, treatment
term viral suppression, before dosing simplification. Few satisfaction data from clinical trial participants, who
virological non-responders, as determined by the entered the study with an 80% chance of receiving long-
stringent FDA snapshot algorithm, were observed with acting treatment might not be generalisable to a more
either long-acting regimen in the LATTE-2 study, with a diverse population of patients with HIV-1.
higher rate of non-response observed in the 8-week group Results from the LATTE-2 study show high rates of
(five patients [4%]) than in the 4-week group (none). Only efficacy and an acceptable safety profile for long-acting
two patients in the 8-week group, and none in the 4-week cabotegravir plus rilpivirine as injectable two-drug
group, met the criteria for protocol-defined virological maintenance therapy in virologically suppressed patients
failure. One participant had emergence of well described with HIV. These results support the further evaluation of
NNRTI and INSTI mutations that conferred reduced monthly long-acting cabotegravir plus rilpivirine as the
susceptibility. No NNRTI mutations were observed in the first all-injectable ART regimen.
second participant, who had undetectable rilpivirine Contributors
concentrations, and a mixture emerged at an integrase DAM was the project physician and assisted with study design, medical
codon not associated with INSTI resistance22 with no monitoring of the study, data collection and interpretation, critical review
and discussion of the manuscript and final study report, and had final
change in cabotegravir susceptibility. The lack of responsibility for the decision to submit for publication. JG-G, H-JS, JJE,
virological non-responders in the 4-week regimen has led YY, DP, TL, JBA, GJR, BC, FG, and LS were study investigators and
to the selection of an optimised 4-week dosing regimen in participated in the conduct of the study, recruitment and follow-up of
the ongoing phase 3 clinical programme, while the patients, data collection, interpretation, and drafting and review of the
manuscript. DD served as the study statistician; participated in study
week 96 data provide supportive evidence for the long- design, acquisition, analysis, and interpretation of data; and provided
term durable response of both the 4-week and 8-week critical review of the study report. SLF, PP, and HC served as clinical
dosing options, supporting further investigation of both pharmacologists; participated in study design, acquisition, analysis, and
dosing intervals. interpretation of data; and reviewed the manuscript. MSC, MM, JM,
SKG, KCS, PEW, KYS, and WRS participated in study design, analysis,
One of the challenges of injectable long-acting agents and interpretation of the data and reviewed the manuscript. All authors
is the potential for serious systemic adverse events provided input to the report and approved the final version.
without the possibility to curtail exposure to the agent. Declaration of interests
The LATTE-2 study implemented a period of cabotegravir DAM, MSC, MM, JM, SKG, KCS, KYS, PP, and WRS are employed by
and rilpivirine oral dosing, before long-acting ViiV Healthcare and are shareholders in GlaxoSmithKline. JG-G has
cabotegravir or rilpivirine dosing, as a strategy to identify received grant funding and consultancy honoraria from AbbVie,
Bristol-Myers Squibb, Gilead Sciences, and Merck, and grant funding
any early-onset acute safety issues before giving the long- from Janssen and ViiV Healthcare. H-JS has received grant funding and
acting injectable formulations. No drug hypersensitivity consultancy honoraria from Gilead Sciences, Janssen, and Merck, and
reactions were observed in the LATTE-2 study. An oral consultancy honoraria from AbbVie and Bristol-Myers Squibb. JJE has
received grant funding and consultancy honoraria from AbbVie,
lead-in strategy has been implemented in phase 3 HIV
Bristol-Myers Squibb, Gilead Sciences, Janssen, and ViiV Healthcare,
treatment studies, to determine if there is a need for and consultancy honoraria from Merck. YY has received travel grants,
continued use of the oral lead-in. honoraria for presentations at workshops, and consultancy honoraria
The high satisfaction reported by patients in the from AbbVie, Bristol-Myers Squibb, Gilead Sciences, Merck, Johnson
& Johnson, and ViiV Healthcare. DP has received grant funding from
LATTE-2 study suggests that long-acting regimens might
Gilead Sciences and ViiV Healthcare, and consultancy honoraria from
provide a preferred alternative to oral daily therapy for Janssen and Merck. TL has received grant funding and consultancy
patients infected with HIV.23,24 The acceptability and honoraria from AbbVie, Bristol-Myers Squibb, Gilead Sciences, Janssen,
tolerability of injectable dosing options will be an Merck, and ViiV Healthcare, and grant funding from Roche. JBA has
received grant funding and consultancy honoraria from Bristol-Myers
important component of long-term treatment success,
Squibb, Gilead Sciences, Merck, and ViiV Healthcare. BC has received
and a high degree of treatment satisfaction will avail this

www.thelancet.com Vol 390 September 23, 2017 1509

 Articles

grant funding and consultancy honoraria from Gilead Sciences, Janssen, 10 Edurant (package insert). Titusville, NJ: Janssen Therapeutics, 2015.
and Merck; grant funding from ViiV Healthcare; and consultancy 11 Odefsey (package insert). Foster City, CA: Gilead Sciences, Inc, 2016.
honoraria from Abbott. FG has received consultancy honoraria and 12 Williams PE, Crauwels HM, Basstanie ED. Formulation and
non-financial support from Bristol-Myers Squibb, Gilead Sciences, and pharmacology of long-acting rilpivirine. Curr Opin HIV AIDS 2015;
Janssen, and consultancy honoraria from ViiV Healthcare. LS has 10: 233–38.
received grant funding and consultancy honoraria from ViiV Healthcare. 13 Margolis DA, Brinson CC, Smith GH, et al, for the LAI116482 Study
MSC has received consultancy honoraria from ViiV Healthcare. SLF is Team. Cabotegravir plus rilpivirine, once a day, after induction with
currently a full-time employee of GlaxoSmithKline and is a shareholder cabotegravir plus nucleoside reverse transcriptase inhibitors in
in GlaxoSmithKline. HC is a full-time employee of Janssen and antiretroviral-naive adults with HIV-1 infection (LATTE):
stockholder of Johnson & Johnson. DD is employed by and is a a randomised, phase 2b, dose-ranging trial. Lancet Infect Dis 2015;
shareholder in GlaxoSmithKline. PEW is a full-time employee of 15: 1145–55.
Janssen and stockholder of AstraZeneca, GlaxoSmithKline, and Johnson 14 Trezza C, Ford SL, Spreen W, Pan R, Piscitelli S. Formulation and
& Johnson. GJR declares no competing interests. pharmacology of long-acting cabotegravir. Curr Opin HIV AIDS
2015; 10: 239–45.
Acknowledgments 15 Spreen W, Ford SL, Chen S, et al. GSK1265744 pharmacokinetics in
We thank everyone who has contributed to the success of this study— plasma and tissue after single-dose long-acting injectable
including all study participants and their families; the LATTE-2 clinical administration in healthy subjects. J Acquir Immune Defic Syndr
investigators and their staff in Canada, France, Germany, Spain, and the 2014; 67: 481–86.
USA; and the ViiV Healthcare, GlaxoSmithKline, PAREXEL, and 16 Spreen W, Williams P, Margolis D, et al. Pharmacokinetics, safety,
Janssen study team members. The study team wishes to express sincere and tolerability with repeat doses of GSK1265744 and rilpivirine
condolences to the family, friends, and patients of Prof Louis Sloan, (TMC278) long-acting nanosuspensions in healthy adults.
coauthor and long-term HIV researcher who passed away during the J Acquir Immune Defic Syndr 2014; 67: 487–92.
conduct of LATTE-2. Writing and editorial assistance was provided 17 DAIDS Regulatory Support Center. Division of AIDS table for
under the direction of the authors by MedThink SciCom with support grading the severity of adult and pediatric adverse events.
https://rsc.tech-res.com/docs/default-source/safety/daids_ae_
from Jeffrey Stumpf and Diane Neer and was funded by ViiV
grading_table_v2_nov2014.pdf?sfvrsn=8 (accessed June 1, 2017).
Healthcare.
18 Bolstad WM, Curran, JM. Introduction to Bayesian statistics,
References 3rd edn. Hoboken, NJ: John Wiley & Sons, 2017.
1 WHO. Progress report 2016: prevent HIV, test and treat all. 19 Palella FJJ, Fisher M, Tebas P, et al. Simplification to
Geneva: World Health Organization, 2016. http://www.who.int/hiv/ rilpivirine/emtricitabine/tenofovir disoproxil fumarate from
pub/progressreports/2016-progress-report/en/ (accessed ritonavir-boosted protease inhibitor antiretroviral therapy in a
June 20, 2017). randomized trial of HIV-1 RNA-suppressed participants. AIDS 2014;
2 Walensky RP, Paltiel AD, Losina E, et al. The survival benefits of 28: 335–44.
AIDS treatment in the United States. J Infect Dis 2006; 194: 11–19. 20 Arribas JR, Pialoux G, Gathe J, et al. Simplification to coformulated
3 Blanc FX, Sok T, Laureillard D, et al, for the CAMELIA (ANRS elvitegravir, cobicistat, emtricitabine, and tenofovir versus
1295-CIPRA KH001) Study Team. Earlier versus later start of continuation of ritonavir-boosted protease inhibitor with
antiretroviral therapy in HIV-infected adults with tuberculosis. emtricitabine and tenofovir in adults with virologically suppressed
N Engl J Med 2011; 365: 1471–81. HIV (STRATEGY-PI): 48 week results of a randomised, open-label,
4 Glass T, Cavassini M. Asking about adherence: from flipping the phase 3b, non-inferiority trial. Lancet Infect Dis 2014; 14: 581–89.
coin to strong evidence. Swiss Med Wkly 2014; 144: w14016. 21 Trottier B, Lake JE, Logue K, et al. Dolutegravir/abacavir/
5 Margolis DA, Boffito M. Long-acting antiviral agents for HIV lamivudine versus current ART in virally suppressed patients
treatment. Curr Opin HIV AIDS 2015; 10: 246–52. (STRIIVING): a 48-week, randomized, non-inferiority, open-label,
6 Karmon SL, Mohri H, Spreen W, Markowitz M. GSK1265744 phase IIIb study. Antivir Ther 2017; published online April 12.
demonstrates robust in vitro activity against various clades of HIV-1. DOI:10.3851/IMP3166.
J Acquir Immune Defic Syndr 2015; 68: e39–e41. 22 Wensing AM, Calvez V, Günthard HF, et al. 2015 Update of the
7 Spreen W, Min S, Ford SL, et al. Pharmacokinetics, safety, and drug resistance mutations in HIV-1. Top Antivir Med 2015;
monotherapy antiviral activity of GSK1265744, an HIV integrase 23: 132–41.
strand transfer inhibitor. HIV Clin Trials 2013; 14: 192–203. 23 Mobula L, Barnhart M, Malati C, et al. Long-acting, injectable
8 Ford SL, Gould E, Chen S, et al. Lack of pharmacokinetic antiretroviral therapy for the management of HIV infection:
interaction between rilpivirine and integrase inhibitors dolutegravir an update on a potential game-changer. J AIDS Clin Res 2015; 6: 1–5.
and GSK1265744. Antimicrob Agents Chemother 2013; 57: 5472–77. 24 Dolgin E. Long-acting HIV drugs advanced to overcome adherence
9 Spreen WR, Margolis DA, Pottage JCJ. Long-acting injectable challenge. Nat Med 2014; 20: 323–24.
antiretrovirals for HIV treatment and prevention.
Curr Opin HIV AIDS 2013; 8: 565–71.

1510 www.thelancet.com Vol 390 September 23, 2017