Background: A low CD4/CD8 ratio during antiretroviral therapy (ART) identifies people with heightened immunosenescence and increased risk of mortality. We aimed to assess the effects of integrase strand transfer inhibitor (INSTI)-based, protease inhibitor-based, or non-nucleoside reverse transcriptase inhibitor (NNRTI)-based first-line ART on long-term CD4/CD8 ratio recovery.
Methods: This prospective cohort study included 13 026 individuals with HIV registered in the Spanish HIV Research Network (CoRIS) cohort recruited from 45 Spanish hospitals. We included HIV-positive people who started triple ART (two nucleoside reverse transcriptase inhibitors [NRTI] with a NNRTI, protease inhibitor, or INSTI) and had HIV RNA suppression within 48 weeks. We used piecewise linear mixed models adjusted for potential confounders to compare longitudinal changes in the CD4/CD8 ratio between people receiving three different types of ART. We used Cox proportional-hazard models to compare the times to CD4/CD8 normalisation between the treatment groups, using cutoff ratios of 0·4, 1·0, and 1·5.
Findings: 6804 individuals contributing 37 149 persons-years and 37 680 observations were analysed; median follow-up was 49 months (IQR 22-89). INSTI-based ART was associated with greater CD4/CD8 gain (change per year compared with INSTI was coefficient -0·07 [95% CI -0·08 to -0·06] for NNRTI and was -0·08 [-0·09 to -0·08] for protease inhibitors). Differences were observed from the first year of therapy and were driven by changes in both CD4 and CD8 cell counts. Subanalyses at different time periods suggested that these differences were driven by changes during the first year of ART without significant differences in the adjusted CD4/CD8 ratio trajectories after the second year of ART (change per year compared with INSTI was coefficient -0·03 [95% CI -0·05 to -0·13] for NNRTI and was -0·06 [95% CI -0·08 to -0·04] for protease inhibitors). Although no differences in the time until CD4/CD8 normalisation at a cutoff ratio of no less than 0·4 were reported between any of the groups, compared with the INSTI group, both the NNRTI and protease inhibitor groups showed lower rates of normalisation at cutoff ratios of 1·0 or more (adjusted hazard ratio 0·80 [95% CI 0·72-0·89] for the NNRTI group and 0·79 [0·69-0·89] for the protease inhibitor group), and 1·5 or more (0·79 [0·65-0·95] for the NNRTI group and 0·78 [0·64-0·97] for the protease inhibitor group). No differences were found between the different integrases in the time until CD4/CD8 normalisation. Subanalyses adjusted for the backbone NRTIs and allowing observations after virological failure yielded similar results.
Interpretation: This study provides new evidence that reinforces the positioning of INSTI-based therapies as a first choice and underlines the importance of analysing the effects of therapeutic interventions on biomarkers linked with morbidity and mortality beyond the plasma HIV RNA and the CD4 cell counts.
Funding: Spanish AIDS Research Network (Instituto de Salud Carlos III), European Development Regional Fund "A way to achieve Europe".
Copyright © 2020 Elsevier Ltd. All rights reserved.