Homeostatic model assessment for insulin resistance index trajectories in HIV-infected patients treated with different first-line antiretroviral regimens

Objective To describe the trajectories of the homeostatic model assessment for insulin resistance (HOMA-IR) index in a cohort of HIV-1 infected patients during their first-line antiretroviral (ART) regimen. Methods Retrospective analysis of naive patients who started ART from 2007 at the Infectious Diseases Unit of the San Raffaele Hospital, Milan. We included patients treated with two nucleoside reverse transcriptase inhibitors (NRTIs, tenofovir, abacavir, lamivudine or emtricitabine), and one anchor drug (ritonavir-boosted protease inhibitor [PI/r], non-NRTI [NNRTI], or integrase strand transfer inhibitor [InSTI]), and with HOMA-IR assessed both before and after the start of ART. Univariate and multivariate mixed linear models estimated HOMA-IR changes during ART. Results Among 618 patients included in the study, 218 received InSTI-, 210 PI/r-, and 190 NNRTI-based regimens. Median follow-up was 27.4 (16.3-41.2) months. Adjusted mean change in HOMA-IR index was significantly higher (P = .041) in patients treated with InSTI-based regimens [0.160 (95% CI: 0.003-0.321) units per year] compared with NNRTI-based regimens [-0.005 (95% CI: -0.184-0.074) units per year]; no difference was observed between patients treated with NNRTI- and PI/r-based regimens or between INSTI-based and PI/r-based regimens. Conclusion InSTI-based first-line ARTs were independently associated with greater increases in HOMA-IR index.