Switch to injectable cabotegravir–rilpivirine given every 8 weeks in adolescents living with HIV with virological suppression in sub-Saharan Africa (LATA): a randomised, open-label, multicentre, 96-week non-inferiority trial

BACKGROUND: Adolescents living with HIV have poorer treatment outcomes than other age groups and might benefit from long-acting-injectable (LAI) antiretroviral therapy (ART). We evaluated efficacy and safety of LAI cabotegravir-rilpivirine compared with daily oral ART in adolescents living with HIV with virological suppression in Africa. METHODS: LATA, a randomised, open-label, multicentre, non-inferiority trial, recruited adolescents with HIV aged from 12 years to younger than 20 years from five clinics in Kenya, South Africa, Uganda, and Zimbabwe. Participants had to have virological suppression (HIV-1 viral load <50 copies per mL) for more than 12 months and no previous treatment failure. Participants were randomly assigned (1:1) to receive either LAI cabotegravir-rilpivirine (LAI group) or daily oral dolutegravir-lamivudine-tenofovir disoproxil fumarate (TLD; control group). Randomisation was stratified by clinical centre and mode of infection (vertical vs horizontal or other). Participants, clinical staff, and researchers were unmasked to treatment allocation, whereas laboratory staff measuring viral loads were masked. Participants in the LAI group were assigned to receive cabotegravir (600 mg)-rilpivirine (900 mg) intramuscular injections at weeks 4, 8, 16, and every 8 weeks thereafter. Participants in the control group were assigned to receive the oral fixed-dose combination, dolutegravir (50 mg)-lamivudine (300 mg)-tenofovir disoproxil fumarate (300 mg), every day. Viral loads were measured every 24 weeks. The primary outcome was the proportion of participants with two consecutive (confirmed) viral load measurements equal to or higher than 50 copies per mL by week 96, estimated using adjusted Kaplan-Meier methods, by intention to treat. The non-inferiority margin and confidence level depended on the control event rate (8·9% margin, 99% CI, for 6% event rate). If non-inferiority was demonstrated, superiority was assessed at the α=5% level. This trial is completed and registered with ClinicalTrials.gov, NCT05154747. FINDINGS: Between June 22, 2023, and April 15, 2024, 476 of 560 screened adolescents with HIV were enrolled in LATA and randomly assigned to either the LAI group (n=235) or the control group (n=241). Of the 476 randomised participants, 466 (98%) had acquired HIV vertically, and 200 (42%) were recruited from Uganda, 128 (27%) from Zimbabwe, 77 (16%) from Kenya, and 71 (15%) from South Africa. Median age was 16·5 years (IQR 14·8-18·1), 256 (54%) participants were female, 220 (46%) were male, 474 (>99%) were Black, and median duration of previous ART was 11·7 years (IQR 8·5-14·1). Median follow-up was 120 weeks (IQR 104-128). Two participants in the LAI group had confirmed viral load of 50 copies per mL or higher by week 96 (Kaplan-Meier estimated proportion 0·9%, 95% CI 0·0 to 2·2) versus 15 participants (6·4%, 3·7 to 9·8) in the control group (estimated difference -5·5%, 99% CI -10·3 to -1·5). LAI met the non-inferiority criterion and demonstrated superiority over control (p=0·0013). By the end of follow-up, seven participants permanently discontinued LAI (one confirmed viral load ≥200 copies per mL, two LAI-related adverse events [serious hypersensitivity reaction; non-serious drug eruption], one incident tuberculosis, two planning pregnancy, and one participant choice). There were 14 serious adverse events (in 14 participants) in the LAI group, including one homicide, compared with 18 (in 12 participants) in the control group, including one death from metastatic osteosarcoma (hazard ratio [HR] 1·16, 95% CI 0·54 to 2·51; p=0·71). The only treatment-related serious adverse event was hypersensitivity reaction. There were 31 adverse events of grade 3 or higher (in 27 participants) in the LAI group and 34 (in 26 participants) in the control group (HR 1·07, 0·62 to 1·83; 0·81). These events included four injection-site reactions (in the LAI group); otherwise, adverse event profiles were similar between trial groups. INTERPRETATION: In adolescents with HIV with virological suppression, LAI cabotegravir-rilpivirine was superior to TLD for maintaining virological suppression, with no new safety concerns observed. These findings support LAI cabotegravir-rilpivirine as a maintenance treatment option for this population.