Safety, antiviral activity, and pharmacokinetics of long-acting injectable cabotegravir–rilpivirine in virologically suppressed adolescents living with HIV-1 (IMPAACT 2017/MOCHA): 48-week results of a multinational, phase 1/2, single-arm study

Summary

Background

Combined intramuscular long-acting cabotegravir–rilpivirine is the first long-acting combination antiretroviral therapy regimen approved for adults with HIV. The International Maternal Pediatric Adolescent AIDS Clinical Trials Network (IMPAACT 2017)–More Options for Children and Adolescents (MOCHA) study assessed safety, acceptability, tolerability, antiviral activity, and pharmacokinetics of these drugs in adolescents with HIV-1. Here we present secondary outcome data up to week 48.

Methods

In this phase 1/2, open-label, non-comparative, dose-finding trial conducted at 18 sites across Botswana, South Africa, Thailand, Uganda, and the USA, virologically suppressed (HIV-1 RNA <50 copies per mL) adolescents (aged 12–17 years; weight ≥35 kg) with HIV-1 switched prestudy antiretrovirals to 4 weeks of daily oral 30 mg cabotegravir and 25 mg rilpivirine, followed by 600 mg cabotegravir and 900 mg rilpivirine long-acting intramuscular (3 mL each) injections in the contralateral gluteus medius at week 4 and week 8, and then every 8 weeks. Secondary outcomes assessed here were grade 3 or worse adverse event, virological failure (including HIV-1 RNA ≥50 copies per mL and ≥200 copies per mL per the US Food and Drug Administration snapshot algorithm), and pharmacokinetic measures including cabotegravir and rilpivirine predose concentration assessment. This trial is registered with ClinicalTrials.gov, NCT03497676.

Findings

Between July 26, 2021, and Aug 27, 2022, 44 of 55 participants who had participated in cohort 1 and 100 of 115 screened study-naive participants were enrolled into cohort 2 of the study. 74 (51%) of 144 participants were female and 70 (49%) male, median age was 15 years (range 12–17), BMI 19·5 kg/m² (16·0–34·3), weight 48·5 kg (35·2–100·9). 132 (92%) had vertical HIV acquisition. Of 144 enrolled participants, 142 received at least one injection, 140 completed week 48, and 140 received the expected seven injections through to week 48. Of 142 participants with at least one injection, 48 (34%) experienced injection-site reaction, mostly grade 1 resolving within 7 days. 43 (38%) of 140 participants experienced drug-related adverse events; two experienced grade 3 or worse adverse events (one abscess with pain 3 days after injection; one abscess 6 weeks after injection). The most common drug-related non-injection-site reaction adverse events were headache (three), rash (three), and nausea (two). After week 48, one participant experienced a grade 4 adverse event (anaphylaxis per site; post-injection reaction per Clinical Management Committee) that resolved but led to study drug discontinuation. No virological failures occurred through to week 48. Median week 48 observed predose concentrations were 2·77 μg/mL for cabotegravir (IQR 1·99–3·55) and 67·9 ng/mL for rilpivirine (52·8–82·4), approximating those in adults and exceeding the protein-adjusted IC₉₀ of 0·166 μg/mL and 12 ng/mL, respectively.

Interpretation

Week 48 data from the first virologically suppressed adolescents with HIV-1 who switched from daily oral to intramuscular cabotegravir and intramuscular rilpivirine every 8 weeks showed favourable safety, pharmacokinetics, and virological suppression informing both real-world use, where approved, and ongoing regulatory submissions, where approval is pending.

Introduction

Despite massive improvements in antiretroviral drugs and regimens, including reduced pill burden, fewer side-effects, and more affordable costs, many patients worldwide still struggle to adhere to daily oral antiretroviral therapies (ARTs). As part of providing differentiated service delivery, it is beneficial to have several HIV prevention and treatment options from which patients and their health-care providers can choose. The availability of a long-acting all-injectable combination of cabotegravir and rilpivirine for treating adolescents and adults with demonstrated virological suppression has been a major milestone for clinically managing HIV.^1–3

In addition to numerous clinical trials that have shown the safety and efficacy of long-acting cabotegravir–rilpivirine in adults with HIV, a growing body of real-world experience using this regimen across various countries and clinical settings shows favourable results.^4–7 Associated injection-site reactions have been characterised with best practices for injection administration outlined.⁸ For adolescents, clinical trial data for this long-acting all-injectable regimen come from the phase 1/2 International Maternal Pediatric Adolescent AIDS Clinical Trials (IMPAACT) Network 2017–More Options For Children And Adolescents (MOCHA) study.^3,9,10 Interim week 48 results in subcohorts of 31 young South Africans (aged 12–17 years), who were virologically suppressed or with detectable virus and recent adherence challenges and achieved virological suppression on oral antiretrovirals and then switched to intramuscular cabotegravir–rilpivirine in a single-center AFINAty study, were presented at the 2025 Conference on Retroviruses and Opportunistic Infections.¹¹ Real-world data for this regimen in adolescents is limited to two single-centre studies.^12,13 Rousseau and colleagues, describe 19 people age 13–25 years with detectable HIV of 76–390 621 copies per mL (median 1760) in the 3 months leading up to the off-label use of intramuscular cabotegravir–rilpivirine and share that all remained undetectable 6–12 months (mean 7 months [SD 5]) post switch.¹² Of note, seven of the 19 had an undetectable HIV viral load at the time of initiation of injectable cabotegravir or injectable rilpivirine. William and colleagues¹³ described 25 young people (median age 19 years, range 14–23) who transitioned from oral ART to intramuscular cabotegravir–intramuscular rilpivirine;¹³ 19 of 25 were virologically suppressed when last tested before initiating intramuscular cabotegravir–rilpivirine. Over a median duration of 11·8 months (range 0·8–31·3), 100% maintained engagement in care and viral suppression.

IMPAACT 2017 examined the safety and pharmacokinetics of oral cabotegravir and intramuscular cabotegravir and rilpivirine in virologically suppressed adolescents (aged 12–17 years) with HIV-1 in a sequential two-cohort study design. Cohort 1 provided the initial safety and pharmacokinetic data that supported dosing in cohort 2, where adolescents stopped taking oral ART and received intramuscular cabotegravir–rilpivirine every 8 weeks after an initial 4–6 weeks of daily lead-in with the oral formulation of these medications.⁹ IMPAACT 2017 cohort 2, week 24 primary outcome data have now been published and show safety, acceptable drug exposures, and efficacy of this all-injectable regimen.³

Here we describe the outcomes related to cohort 2 secondary objectives to assess the safety, antiviral activity, and repeat-dose pharmacokinetics of long-acting intramuscular cabotegravir–rilpivirine in adolescents with HIV who were virologically suppressed up to week 48.

Methods

Study design and participants

IMPAACT 2017/MOCHA was a phase 1/2, multicentre, open-label, non-comparative, dose-finding study to evaluate the safety, pharmacokinetics, acceptability, tolerability, and antiviral activity of oral cabotegravir intramuscular–rilpivirine in virologically suppressed adolescents (aged 12–17 years; weighing ≥35 kg) with HIV-1.

The previously published study design included two cohorts, and each cohort included two steps.^3,9 Briefly, in cohort 1, participants continued their oral ART regimen and added oral cabotegravir or rilpivirine for 4–6 weeks (step 1), followed by intramuscular long-acting cabotegravir or intramuscular long-acting rilpivirine (step 2). In cohort 2, participants discontinued their ongoing suppressive oral ART regimen and started 4–6 weeks of oral lead-in with daily cabotegravir (30 mg) plus rilpivirine (25 mg; step 3), then switched to intramuscular cabotegravir (600 mg [3 mL]) plus intramuscular rilpivirine (900 mg [3 mL]), each administered in contralateral gluteus medius muscles 4 weeks apart and then intramuscular cabotegravir–rilpivirine every 8 weeks (step 4; appendix p 2). Participants who permanently discontinued injectable study intervention continued on study for an additional 48-week long-term safety and washout pharmacokinetic follow-up after their last injection, while receiving an alternative oral ART regimen.

The full IMPAACT 2017/MOCHA protocol with inclusion (including virologic suppression) and exclusion criteria, study design, and procedures is available on the IMPAACT Network website.

The study protocol, informed consent, and other information that required preapproval were reviewed and approved by applicable ethics committees or institutional review boards.

This trial is registered with ClinicalTrials.gov, NCT03497676.

Procedures

Safety evaluations and HIV-1 viral load measurements were performed at weeks 2, 4, 5, 8, 16, 24 (primary endpoints), 32, 40, and 48 (secondary endpoints). The Division of AIDS (DAIDS) Table for Grading the Severity of Adult and Pediatric Adverse Events was used for grading adverse events throughout the study.¹⁴ Injection-site reactions were defined in the protocol as pain out of proportion to what would be expected when a person receives an intramuscular injection and other local findings, such as tenderness, erythema, redness, induration, swelling, or pruritis.

Table: Summary of adverse events through to week 48 in the evaluable and all-treated analysis datasets and injection site reactions in participants with at least one injection.

	Cohort 1-rollover group	Study-naive group	Total
Evaluable
Any adverse event	39/41 (95·1%, 83·5–99·4)	76/97 (78·4%, 68·8–86·1)	115/138 (83·3%, 76·1–89·1)
Grade 3 or higher adverse event*	5 (12·2%, 4·1–26·2)	14 (14·4%, 8·1–23·0)	19 (13·8%, 8·5–21·7)
Grade 3 or higher drug-related adverse event†	0 (0·0–8·6)	2 (2·1%, 0·3–7·3)	2 (1·4%, 0·2–5·1)
Serious adverse event‡	0 (0·0–8·6)	1 (1·0%, <0·1–5·6)	1 (0·7%, <0·1–4·0)
Serious drug-related adverse event	0 (0·0–8·6)	0 (0·0–3·7)	0 (0·0–2·6)
Premature permanent discontinuation of treatment due to drug-related adverse event	0 (0·0–8·6)	0 (0·0–3·7)	0 (0·0–2·6)
Death due to drug-related adverse event	0 (0·0–8·6)	0 (0·0–3·7)	0 (0·0–2·6)
All treated
Any adverse event	42/44 (95·4%, 84·5–99·4)	76/100 (76·0%, 66·4–84·0)	118/144 (81·9%, 74·7–87·9)
Grade 3 or higher adverse event*	6 (13·6%, 5·2–27·4)	14 (14·0%, 7·9–22·4)	20 (13·9%, 8·7–20·6)
Grade 3 or higher drug-related adverse event†	0 (0·0–8·0)	2 (2·0%, 0·2–7·0)	2 (1·4%, 0·2–4·9)
Serious adverse event‡	0 (0·0–8·0)	1 (1·0%, <0·1–5·5)	1 (0·7%, <0·1–3·8)
Serious drug-related adverse event	0 (0·0–8·0)	0 (0·0–3·6)	0 (0·0–2·5)
Premature permanent discontinuation of treatment due to drug-related adverse event	0 (0·0–8·0)	0 (0·0–3·6)	0 (0·0–2·5)
Death due to drug-related adverse event	0 (0·0–8·0)	0 (0·0–3·6)	0 (0·0–2·5)
At least one injection
Any injection-site reaction§	11/43 (25·6%, 13·2–40·3)	37/99 (37·4%, 27·6–47·2)	48/142 (33·8%, 25·7–41·7)
Abscess	0 (0·0–8·0)	2 (2·0%, 0·2–7·0)	2 (1·4%, 0·2–4·9)
Bruising	1 (2·3%, 0·1–12·0)	0 (0·0–3·6)	1 (0·7%, <0·1–3·8)
Induration	0 (0·0–8·0)	1 (1·0%, <0·1–5·5)	1 (0·7%, <0·1–3·8)
Joint pain	1 (2·3%, 0·1–12·0)	0 (0·0–3·6)	1 (0·7%, <0·1–3·8)
Nodule	1 (2·3%, 0·1–12·0)	4 (4·0%, 1·1–9·9)	5 (3·5%, 1·1–7·9)
Pain	11 (25·6%, 13·2–40·3)	37 (37·4%, 27·6–47·2)	48 (33·8%, 25·7–41·7)
Pruritus	0 (0·0–8·0)	2 (2·0%, 0·2–7·0)	2 (2·0%, 0·2–7·0)
Swelling	2 (4·7%, 0·6–15·5)	3 (3·0%, 0·6–8·5)	5 (3·5%, 1·1–7·9)

Data are n/N (%, 95% CI) or n (%, 95% CI) unless stated otherwise.

^*Grade: 1=mild, 2=moderate, 3=severe, 4=potentially life-threatening, 5=death.

^†Drug-relatedness of adverse events was determined by the site.

^‡Serious adverse events included only International Council of Harmonization criteria-defined serious adverse events and malignancies.

^§An injection-site reaction was defined as an adverse event that results in pain out of proportion of what would be expected when a person gets an intramuscular injection, tenderness, erythema, redness, induration or swelling, or pruritis, regardless of when it occurs after administration of an injection.

The pharmacokinetic sampling strategy is detailed in the protocol. A single predose sample was obtained at weeks 8, 16, 24, 32, 40, and 48.

Outcomes

Secondary safety outcomes assessed through week 48 are described in this manuscript and included grade 3 or worse adverse event, grade 3 or worse adverse event assessed as related to the study products, serious adverse event meeting International Council of Harmonization criteria assessed as related to the study products, permanent discontinuation of study product due to an adverse event assessed as related to the study products, and death due to an adverse event assessed as related to study products by the site investigator. Primary outcomes assessed at week 24 were previously published and supported recent approval by the European Commission.^3,15

Secondary virological response (antiviral activity) outcomes assessed at week 48 included HIV-1 RNA of 50 copies per mL or more and 200 copies per mL or more, per the US Food and Drug Administration (FDA) snapshot algorithm.¹⁶

Pharmacokinetic outcome measures included cabotegravir and rilpivirine predose concentration profiles at week 48.

Outcomes evaluating the tolerability and acceptability of long-acting cabotegravir–rilpivirine through week 48 are reported elsewhere.¹⁷

Statistical analysis

The evaluable analysis set included participants treated exclusively with the dose (intramuscular cabotegravir and intramuscular rilpivirine every 8 weeks) that was confirmed in cohort 1 and who completed all scheduled treatments up to week 48, or experienced any of the following events: death attributable to the study products, study product-related grade 3 or worse adverse events (excluding injection-site reactions), or permanent discontinuation of treatment due to study product-related adverse event (regardless of grade). Study-naive participants (ie, those who did not participate in cohort 1) were treated as the primary subgroup of cohort 2 evaluable participants. The all-treated analysis set included all participants who received at least one dose of any study product, including cohort 1-rollovers (ie, those who participated in cohort 1).

All safety data were considered from the first participant exposure to cohort 2 study treatment through to week 48. Grade 3 or worse adverse events after week 48 but before the week 48 analysis cutoff date of Nov 15, 2023, are also reported. Clinical data summarised in this manuscript after week 48 might evolve and be subject to change after all follow-up is complete. Safety outcome measures were summarised with proportions and exact (Clopper–Pearson) 95% CIs and are presented by group (cohort 1-rollover and study-naive) and total for the week 48 evaluable and all-treated analysis sets.

For the analysis of the secondary outcome measures of virologic response at week 48, occurrence of virologic failure per the FDA snapshot algorithm at thresholds of 50 copies per mL or more, and 200 HIV-1 RNA copies per mL or more, was summarised using proportions and exact (Clopper–Pearson) 95% CIs and is presented by group (cohort 1-rollover and study-naive) and total for the week 48 evaluable and the all-treated analysis sets.

All statistical analyses were performed with SAS (version 9.4).

Repeat-dose pharmacokinetics of intramuscular cabotegravir and intramuscular rilpivirine every 8 weeks in adolescents with HIV who were virologically suppressed through to week 48 were analysed as a secondary study objective. Cabotegravir and rilpivirine concentrations were quantified via liquid chromatographic–tandem mass spectrometry using assays validated per FDA bioanalytical recommendations. Lower limits of quantification were 0·025 μg/mL for cabotegravir and 1 ng/mL for rilpivirine.^18,19 Preanalytic pharmacokinetic blood samples were protected from light.

Pharmacokinetic concentrations and accumulation ratios with the long-acting formulations were summarised using descriptive statistics. Post-hoc analyses evaluated trough concentrations at week 48 by BMI and sex, adjusting for weight, with linear-regression models. Assessment for steady state was done based on 90% CIs for the geometric mean ratios of the week 48 versus week 40 pharmacokinetic concentrations. A geometric mean ratio (week 48 to week 40) 90% CI with an upper bound of 1·10 or more was considered indicative of ongoing accumulation of the drug in the plasma and failure to achieve steady-state conditions by week 48.

Role of the funding source

Representatives of funders participated in the study design, data interpretation, and manuscript writing.

Results

18 centres in five countries (Botswana, South Africa, Thailand, Uganda, and the USA) enrolled 144 participants (44 cohort 1-rollover and 100 study-naive) in cohort 2 between July 26, 2021, and Aug 27, 2022 (figure 1). Baseline data have been previously reported.³

Figure 1: Cohort 2 trial profile.

ART=antiretroviral treatment. *Inclusion or exclusion criteria deviation: was not on stable combination ART before study entry, discovered after study entry. †As of the database freeze date of Nov 15, 2023.

Of 144 participants, 142 received at least one injection, 140 completed week 48, and 140 received the expected seven injections through to week 48. The median (minimum–maximum) total exposure duration (ie, first oral dose through either treatment discontinuation, last day of oral treatment, last injection plus 70 days, or the time of database freeze on Nov 15, 2023) was 90 weeks (8–108) for cohort 1-rollover participants and 73 weeks (2–78) for study-naive participants. Through to week 48, all injection administrations fell within the allowable 7-day window except at week 32, when one of 141 (1%) participants had an injection administration at 8 days or more after the target. This participant (a cohort 1 rollover) switched to oral bridging for 20 days before the week 32 injection due to a planned trip out of the country.

Six participants were excluded from the evaluable analysis population: one rollover participant withdrew from the study due to non-compliance to study visits; one study-naive participant did not meet eligibility criteria and exited the study after 2 weeks; two rollover participants erroneously were underdosed at week 16; one study-naive participant became pregnant and discontinued use of the study drug regimen before week 24; and one study-naive participant was lost to follow-up before week 48.

Through to week 48 in the evaluable analysis set, 76 (78·4%, 95% CI 68·8–86·1) of 97 study-naive and 39 (95·1%, 83·5–99·4) of 41 cohort 1-rollover participants experienced at least one adverse event; 14 (14·4%, 8·1–23·0) of 97 study-naive and five (12·2%, 4·1–26·2) of 41 cohort 1-rollovers experienced a grade 3 or worse adverse event (table). Two (2·1%, 0·3–7·3) study-naive participants experienced a grade 3 or worse drug-related adverse event. One participant experienced a grade 3 injection site abscess with associated grade 3 injection site pain at the cabotegravir injection site 3 days after the injection. Both adverse events were related (common underlying cause) but reported separately, per DAIDS grading criteria. The adverse events resolved within 2 weeks and study treatment was not discontinued. The other participant experienced a grade 3 injection site abscess related to rilpivirine about 6 weeks after the injection. This adverse event resolved within 3 days without any intervention, and study treatment was not discontinued. Per the participant’s self-report, the overall clinical picture for this adverse event was atypical for an abscess––discharge without pain that spontaneously resolved and left a scar by the time the participant was examined by site staff 1 week later. The adverse event was documented as an abscess and adjudicated as a grade 3 event, per the DAIDS grading criteria.

One (1%, 0–6) study-naive participant experienced a serious adverse event; this participant was reported to have concurrently grade 3 malaria and typhoid fever (both events were not related to either study drug and resolved).

All evaluable study-naive and rollover participants completed each visit, and none had an intercurrent event (ie, an event that could affect the existence or interpretation of the outcome measure), a serious adverse event, prematurely discontinued treatment, or death due to a drug-related adverse event through to week 48.

By week 48, 53 (37%) of 144 participants in the all-treated analysis set experienced at least one drug-related adverse event. The most common drug-related adverse events were injection-site pain (47 participants), injection-site nodule (five participants), injection-site swelling (five participants), and headache (three participants; appendix pp 3–4).

Since the previously published week 24 data (data cutoff date: June 7, 2023), four additional participants had at least one grade 3 or 4 adverse event, and one participant who had a grade 3 serious adverse event at week 24 (malaria) had an additional grade 3 serious adverse event (typhoid fever; all data through to Nov 15, 2023 [database freeze], all-treated analysis; appendix p 5). The grade 3 or 4 adverse events reported in the additional participants, since the week 24 data cutoff date were grade 3 haemoglobin decreased, grade 3 platelet count decreased, grade 4 blood creatine phosphokinase increased (all not related), and grade 4 anaphylactic reaction (related). The event including generalised maculopapular rash and chest tightness reported after the week 72 injection of cabotegravir and rilpivirine as a study-drug-related anaphylactic reaction (grade 4) by the site investigator and noted as a serious adverse event led to discontinuation of study intervention (not reflected in the database at the time of database freeze on Nov 15, 2023). Per the protocol, the IMPAACT 2017 clinical management committee independently assessed the event as study-drug-related, not consistent with an anaphylactic event, and most consistent with a post-injection reaction (rapid onset and resolution of symptoms without any medications; not meeting anaphylaxis criteria).²⁰ Post-dose cabotegravir (46·7 μg/mL) and rilpivirine (326 ng/mL) concentrations collected 21·87 h and 21·85 h post dose, respectively, as part of suspected post-injection reaction follow-up, were significantly higher than predose trough ranges of 2·2–4·9 μg/mL and 37–90 ng/mL, respectively, noted in this participant. Given study drug discontinuation, there was no further predose trough sample collection.

By week 48, 142 participants with at least one injection had a total of 1696 injections and reported 196 injection-site reactions: 177 (90%) 196 were grade 1 and 16 (8% were grade 2; 174 (89%) resolved within 7 days (figure 2, appendix p 6). Of 148 injection-site reactions among the study-naive participants, three (2%) were grade 3 and eight (5%) took more than 14 days to resolve. The number of participants with any injection-site reaction was highest at the first injection. Although there was a slight uptick at week 24, there was a downward trend with subsequent injections. At week 48, 13 (9%) of 140 participants reported injection site reaction related to rilpivirine compared with four (3%) related to cabotegravir. No participant withdrew from the study due to injection-site reactions through to week 48.

Figure 2: Injection site reactions at week 48.

Injection site reactions over 48 weeks were stratified by study injection type—ie, intramuscular cabotegravir, intramuscular rilpivirine, or overall.

There was one reported pregnancy with estimated conception occurring during week 5. This participant’s last injection was at study week 8, and the participant had a livebirth of a healthy female infant at 39 weeks (as reported previously).³

All evaluable participants (n=138) were categorised as a virological success at both HIV-1 RNA less than 200 copies per mL and less than 50 copies per mL (100%; 95% CI 97–100), per FDA snapshot algorithm (appendix p 7). Ten participants had viral blips (HIV viral load >50 copies per mL; range 53–457 copies per mL) at some point after study entry while on study treatment. None had two consecutive HIV viral load measurements greater than 200 copies per mL. Six participants had HIV viral loads greater than 50 copies per mL at study entry (range 60–168 053 copies per mL). One of these participants inadvertently stopped taking their prestudy oral ART for 3 weeks before study entry and subsequently was taken off study (inclusion criteria violation). All other participants were categorised as having virological success on study drugs.

After oral administration of 30 mg cabotegravir and 25 mg rilpivirine once daily through to week 4, followed by two injections of intramuscular cabotegravir (600 mg) and rilpivirine (900 mg) 4 weeks apart, and then every 8 weeks, the median week 48 observed predose concentrations for cabotegravir (2·77 μg/mL, IQR 1·99–3·55) and rilpivirine (67·9 ng/mL, 52·8–82·4) approximated those in adults and were well above the respective protein-adjusted IC₉₀ values of 0·166 μg/mL for cabotegravir and 12 ng/mL for rilpivirine (figure 3). At week 48, cabotegravir (geometric mean ratio [week 48 to week 40] 1·022, 90% CI 0·970–1·076) appeared to be at steady state, but rilpivirine (1·130, 1·089–1·173) was not. Of the participants in cohort 2, 138 evaluable provided cabotegravir and rilpivirine data. Of the six participants whose pharmacokinetic data were excluded in part or whole, two permanently discontinued study treatment before receiving intramuscular injections; two received incorrect doses (underdosed) at week 16 but remained virologically suppressed; one became pregnant and entered long-term safety follow-up before the week 24 visit, and a week 48 sample was not available for one. Two participants, both at week 40, had markedly elevated rilpivirine plasma concentrations post-injection (1332 ng/mL at 2·07 h and 2173 ng/mL at 2·08 h after rilpivirine injection, respectively) but normal pre-injection concentrations (96·4 ng/mL and 61·4 ng/mL, respectively) with adverse events consistent with a post-injection reaction (chest discomfort, dizziness, headache, or difficulty breathing and distressed, tachypneic or hyperhidrosis, respectively). For the two participants with markedly elevated post-injection rilpivirine plasma concentrations, only the latter post-injection values were removed from the pharmacokinetic analysis. The subsequent trough concentrations collected before the next dose of rilpivirine were 61 ng/mL and 43 ng/mL, respectively. Although these values were approximately 30–35% lower than the previous predose trough concentrations, they were well within the target range.

Figure 3: Cabotegravir and rilpivirine trough concentrations in IMPAACT 2017/MOCHA participants compared with those of adults in the ATLAS-2M²¹ study.

Cabotegravir (A) and rilpivirine (B) concentration troughs in IMPAACT 2017/MOCHA participants over 48 weeks. Median troughs (solid black lines) and the 5th percentile and 95th percentile (dashed black and green lines) were compared with the median troughs of adults from the ATLAS-2M study (blue lines) and protein-adjusted IC₉₀s (red lines). ATLAS-2M study=the antiretroviral therapy as long acting suppression–2M study. IMPAACT 2017/MOCHA study=International Maternal Pediatric Adolescent AIDS Clinical Trials Network 2017–More Options For Children And Adolescents study.

There were no BMI-associated differences in cabotegravir or rilpivirine trough concentrations evident at week 48 (p=0·86). Females had higher trough cabotegravir (p<0·0001) and rilpivirine concentrations (p=0·024) compared with males at week 48, even after adjusting for weight (appendix p 9).

Discussion

Week 48 analysis of adolescents living with HIV who were virologically suppressed on oral ART and switched to intramuscular cabotegravir–intramuscular rilpivirine every 8 weeks showed that study retention was high, there were no unexpected safety signals, virological suppression was maintained, and pharmacokinetics were comparable to those in adults receiving this treatment regimen.

The most common drug-related adverse event associated with the relatively large 3-mL volume of each cabotegravir and rilpivirine injection in contralateral glutei was injection-site pain, and injection-site reactions were noted in about a third of the study participants.⁸ Of note, proactively planning for the nuances of pain reporting in paediatrics versus adults, the study guidance for injection-site reaction (including pain) ascertainment specified only those signs and symptoms that were out of proportion to what could reasonably be expected after an intramuscular injection. This approach differs from studies in adults, where any pain at the injection site is included as an injection-site reaction. Despite the reported injection site pain, all but one (who received oral bridging for one injection missed due to planned travel) participants received their injections within the study-recommended window, and none discontinued study because of injection site pain or reaction. Although still noted more with the rilpivirine than cabotegravir injection, the majority (91%) of the participants at week 48 had no injection-site reaction noted; information that likely would be helpful to stress with adolescents, both for those considering switching to this injectable regimen and during pre-injection counselling. Consistent with studies in adults, the number of participants with any injection-site reaction was highest at the first injection. However, the slight uptick in injection-site reactions noted at week 24 has not been reported before, and we do not have a hypothesis, actionable interventions to suggest, or anything to look out for in future studies. Further details and discussion of acceptability and tolerability of this all-injectable regimen are described in Lowenthal and colleagues.¹⁷

The only drug-related adverse event that led to study drug discontinuation was based on site investigator assessment of an anaphylactic reaction; however, the clinical management committee deemed this event to be more consistent with a post-injection reaction. In this case, the post-dose cabotegravir, much more so than post-dose rilpivirine, concentrations (research-only tests not tested in real-time and not reportable for clinical decision making) were significantly elevated, consistent with the assessment of post-injection reaction. The two other self-limiting post-injection reactions noted at week 40 were mild (grade 1 and grade 2) and not assessed as anaphylaxis, and participants continued to receive the study drug. Clinically differentiating the immediate, typically short-lived and self-limiting, post-injection reaction from anaphylaxis can be sometimes difficult but is very important, and health-care provider familiarity with the post-injection reaction is necessary. Unlike if anaphylaxis occurs, if a post-injection reaction is suspected cabotegravir and rilpivirine use is typically continued without any clinical concerns. In prior clinical trials, serious post-injection reactions were reported within minutes after the injection of rilpivirine, including symptoms such as dyspnoea, agitation, bronchospasm, abdominal cramping, rash, urticaria, dizziness, flushing, sweating, oral numbness, changes in blood pressure and pain (eg, back and chest pain). These events were reported in less than 1% of participants and began to resolve within a few minutes after the injection. These events might have been associated with inadvertent (partial) intravenous administration and can be associated with the excipients in the long-acting formulations rather than the antiretrovirals themselves.²⁰ There is guidance to redosing after a suspected post-injection reaction and the lower, but still well above the target, trough concentrations collected before the next scheduled dose after two of the rilpivirine-associated post-injection reactions were consistent with this recommendation.

By week 48, viral load data from adolescents enrolled in the IMPAACT 2017/MOCHA study remained overall consistent with what has been reported in clinical trials and real-world data in adults, showing efficacy (no confirmed virological failures) with occasional viral load blips in a small subset of participants without obvious timing or patterns or associations with virological failure.^22–25 The observed cabotegravir and rilpivirine pharmacokinetic profiles met the key exposure targets based on adult data for cabotegravir (oral and intramuscular administration) and rilpivirine (intramuscular). Although cabotegravir appeared to be at steady state by week 48, the time for rilpivirine to reach steady state remains to be determined for this cohort of adolescents. Of note, after 1 year of monthly or every 2 months injections, approximately 80% of the rilpivirine pharmacokinetic steady-state exposure is reached.²⁶ The mean apparent half-life of intramuscularly administered rilpivirine in adults is absorption-rate limited and was estimated to be 13–28 weeks. The observed sex-based difference in cabotegravir pharmacokinetics, with cabotegravir troughs higher in females than males at week 48, is similar to what has been previously noted in adults and, overall, is not clinically significant or actionable.²⁷

One of the limitations of this study, in which all study participants received a daily oral cabotegravir and rilpivirine lead-in for 4–6 weeks, is the inability to characterise the safety and pharmacokinetics of a direct-to-injection approach, where the oral lead-in is omitted. Additionally, because this study required virological suppression before enrolment, it does not speak to the use of this regimen for the much-desired potential use in those who are struggling with adherence to daily oral ART. However, the data support an all-injectable regimen option for the large subset of adolescents with HIV who are virologically suppressed on daily oral medications.

This Article describes the longest experience to date from the first study of adolescents, the majority of whom had perinatally acquired HIV infection and had taken daily oral ART most of their lives, to receive the first all-injectable HIV treatment regimen. These data inform clinical practice and regulatory submissions. Longer-term follow-up of study participants through to week 96 is awaited and will further inform the use of this all-injectable HIV-1 treatment option in adolescents living with HIV.

Research in context.

Evidence before this study

Intramuscular cabotegravir–rilpivirine is the first approved long-acting combination antiretroviral treatment (ART) regimen for people with HIV-1. We searched PubMed periodically between Jan 5, 2017, and April 1, 2025, using keywords “long acting cabotegravir”, “long acting rilpivirine”, “treatment”, AND “adolescents”. Other than the International Maternal Pediatric Adolescent AIDS Clinical Trials Network (IMPAACT 2017)–More Options for Children and Adolescents (MOCHA) study, no clinical trials of this long-acting all-injectable regimen for treating HIV in this age population have been published. The IMPAACT 2017/MOCHA study is the first clinical trial in adolescents living with HIV-1 to evaluate this long-acting all-injectable ART regimen. IMPAACT 2017/MOCHA protocol development was initiated in January, 2017, and initial pharmacokinetics, safety, and acceptability and tolerability results from cohort 1 of this study, where virologically suppressed adolescents living with HIV received either long-acting cabotegravir or rilpivirine in addition to their background daily oral ART, were published in April, 2024. In cohort 2 of IMPAACT 2017/MOCHA, adolescents living with HIV switched from daily oral ART to once every 8 weeks intramuscular cabotegravir–rilpivirine. Cohort 2 week 24 results were published in March, 2025, and showed that this regimen has favourable safety, pharmacokinetics, and antiviral activity in virologically suppressed adolescents (aged 12–17 years) weighing 35 kg or more, which is comparable to the outcomes seen in adults.

Added value of this study

The IMPAACT 2017/MOCHA cohort 2 week 48 results constitute the longest and largest described multinational experience to date of this first all-injectable HIV treatment regimen in a diverse group of virologically suppressed (HIV-1 RNA <50 copies per mL) adolescents (aged 12–17 years) weighing 35 kg or more who switched to this regimen. These data inform the growing clinical use of this all-injectable combination ART in the USA, Canada, and the EU, where the regimen has been approved in adolescents, and supports future submissions seeking regulatory approval of this regimen for adolescents living with HIV in other countries.

Implications of all the available evidence

Week 48 multinational data from virologically suppressed adolescents with HIV-1 who switched from daily oral to intramuscular cabotegravir–rilpivirine every 8 weeks showed favourable safety and pharmacokinetic profiles and durable efficacy. These data help inform regulatory submission in countries where the regimen is not yet approved and the growing clinical use in the USA, Canada, and most recently in the EU, where the regimen is approved.