Abstract
We evaluated treatment duration and viral suppression (VS) outcomes with integrase strand transfer inhibitor (INSTI)-based regimens versus other contemporary regimens among adults in routine HIV care. Eligible participants were seen during January 1, 2007 to June 30, 2018 at nine U.S. HIV clinics, initiated antiretroviral therapy (ART) (baseline date), and had ≥2 clinic visits thereafter. We assessed the probability of remaining on a regimen and achieving HIV RNA <200 copies/mL on initial INSTI versus non-INSTI ART by Kaplan-Meier analyses and their correlates by Cox regression. Among 1,005 patients, 335 (33.3%) were prescribed an INSTI-containing regimen and 670 (66.7%) a non-INSTI regimen, which may have included non-nucleoside reverse transcriptase inhibitors, protease inhibitors, and other agents. In both groups, most patients were male, nonwhite, and aged <50 years. Comparing the INSTI with non-INSTI group, the median baseline log(10) HIV viral load (VL; copies/mL) was 4.6 versus 4.5, and the median CD4(+) cell count (cells/mm(3)) was 352 versus 314. In Kaplan-Meier analysis, the estimated probabilities of remaining on initial regimens at 2 and 4 years were 58% and 40% for INSTI and 51% and 33% for non-INSTI group, respectively (log-rank test p = .003). In multivariable models, treatment with an INSTI (vs. non-INSTI) ART was negatively associated with a regimen switch [hazard ratio (HR) 0.67, 95% confidence interval (CI) 0.56-0.81, p < .001] and was positively associated with achieving VS (HR 1.52; CI 1.29-1.79, p < .001), both irrespective of baseline VL levels. Initial INSTI-based regimens were associated with longer treatment durations and better VS than non-INSTI regimens. Results support INSTI regimens as the initial therapy in U.S. treatment guidelines.