Treatment outcomes of integrase inhibitors, boosted protease inhibitors and nonnucleoside reverse transcriptase inhibitors in antiretroviral-naïve persons starting treatment

Objectives: Although outcomes of antiretroviral therapy (ART) have been evaluated in randomized controlled trials, experiences from subpopulations defined by age, CD4 count or viral load (VL) in heterogeneous real-world settings are limited. Methods: The study design was an international multicohort collaboration. Logistic regression was used to compare virological and immunological outcomes at 12 ± 3 months after starting ART with an integrase strand transfer inhibitor (INSTI), contemporary nonnucleoside reverse transcriptase inhibitor (NNRTI) or boosted protease inhibitor (PI/b) with two nucleos(t)ides after 1 January 2012. The composite treatment outcome (cTO) defined success as VL ' 200 HIV-1 RNA copies/mL with no regimen change and no AIDS/death events. Immunological success was defined as a CD4 count ' 750 cells/μL or a 33% increase where the baseline CD4 count was ≥ 500 cells/μL. Poisson regression compared clinical failures (AIDS/death ≥ 14 days after starting ART). Interactions between ART class and age, CD4 count, and VL were determined for each endpoint. Results: Of 5198 ART-naïve persons in the International Cohort Consortium of Infectious Diseases (RESPOND), 45.4% started INSTIs, 26.0% PI/b and 28.7% NNRTIs; 880 (17.4%) were aged ' 50 years, 2539 (49.4%) had CD4 counts ' 350 cells/μL and 1891 (36.8%) had VL ' 100 000 copies/mL. Differences in virological and immunological success and clinical failure among ART classes were similar across age groups (≤ 40, 40–50 and ' 50 years), CD4 count categories (≤ 350 vs. ' 350 cells/μL) and VL categories at ART initiation (≤ 100 000 vs. ' 100 000 copies/mL), with all investigated interactions being nonsignificant (P ' 0.05). Conclusions: Differences among ART classes in virological, immunological and clinical outcomes in ART-naïve participants were consistent irrespective of age, immune suppression or VL at ART initiation. While confounding by indication cannot be excluded, this provides reassuring evidence that such subpopulations will equally benefit from contemporary ART.