Cabotegravir-Rilpivirine Lowers Regimen Failure in LATITUDE Trial

Key Takeaways

• Monthly injections were associated with lower regimen failure than standard oral ART during the randomized phase.
• Participants first received adherence support, conditional incentives, and oral ART before only those with viral suppression entered randomization.
• Overall adverse-event incidence was similar, injection-site reactions were common with injections, and resistance-associated mutations were reported after confirmed virologic failure in both groups.

Monthly cabotegravir-rilpivirine was associated with less regimen failure than standard oral ART at week 48, 22.8% versus 41.2%, in adults with HIV and adherence barriers who entered the randomized phase after achieving viral suppression during step 1. Step 2 follow-up had a median of 48 weeks, and the comparison was limited to participants who reached suppression during that structured run-in period.

The ACTG A5359 LATITUDE trial was an open-label, phase 3, multicenter randomized study conducted at 33 U.S. sites. Eligible participants were adults aged 18 years or older with HIV, prior ART exposure, no hepatitis B coinfection, and evidence of nonadherence. Nonadherence meant a poor virologic response in patients with an active prescription for ART or loss to clinical follow-up with ART nonadherence, each lasting at least 6 consecutive months within 18 months before entry. A total of 453 participants entered step 1, which provided up to 24 weeks of adherence support, conditional economic incentives, and standard oral ART. After suppression was achieved, 306 participants were randomized in step 2.

Participants with suppression were randomized 1:1 to continue standard care or switch to monthly cabotegravir-rilpivirine, with or without an oral lead-in. Among 152 participants assigned to injections, 87 used an oral lead-in lasting about 4 weeks before starting treatment. The injectable regimen used loading doses of cabotegravir 600 mg and rilpivirine 900 mg, followed by maintenance injections every 21 to 35 days after the second and third injections were given within 24 to 32 days after the previous injection. After loading, 94% of injections were given on time, and four participants received oral bridging therapy because injection delays were anticipated. Regimen failure, the primary endpoint for week 48, meant confirmed virologic failure with two consecutive HIV-1 RNA measurements above 200 copies per milliliter or permanent treatment discontinuation during step 2.

In the randomized phase, regimen failure occurred in 29 participants (19%) receiving monthly cabotegravir-rilpivirine and 55 participants (36%) receiving standard care. The between-group difference was -18.4 percentage points, with a 98.4% confidence interval of -32.4 to -4.3 and P=0.002. Virologic failure occurred in 6 versus 34 participants, and treatment-related failure occurred in 9 versus 34 participants. Randomization stopped early after a prespecified interim review on the basis of the efficacy results for virologic failure and treatment-related failure, and subgroup and protocol-version analyses showed no substantial variability. These findings came from the step-2 comparison after suppression had been achieved during the run-in phase.

By week 48, overall adverse-event incidence was similar, with cumulative incidence of 43.5% in the injectable group and 42.4% with standard care. Serious adverse events occurred in 19.3% versus 14.5%, and adverse-event discontinuations affected 4 participants receiving injections and 1 receiving standard care. Injection-site reactions affected 84 of 141 participants on cabotegravir-rilpivirine, most often as pain, swelling, tenderness, or nodules; 3 were grade 3, and 2 led to discontinuation. Resistance-associated mutations were reported in 2 participants with confirmed virologic failure in each group, and INSTI mutations developed in 2 of 5 versus 2 of 22 with available data. The trial remained open-label, and step 1 included support and incentives before randomization.