Immune outcomes with dolutegravir‐based dual therapy versus bictegravir‐based triple therapy: A multinomial propensity score analysis

OBJECTIVES: To compare immune outcomes after switching to dolutegravir/lamivudine (DTG/3TC), dolutegravir/rilpivirine (DTG/RPV) and bictegravir/emtricitabine/tenofovir alafenamide (B/F/TAF). METHODS: Three Spanish cohorts of adults with baseline HIV-RNA <50 copies/mL were harmonized. Multinomial propensity scores were used for 13 baseline variables. Stabilized inverse probability weights were winsorized at the 1st/99th percentiles. The primary outcome was baseline-adjusted CD4+ T-cell count at week 48. Bootstrap confidence intervals and Holm's correction were used. RESULTS: Of 3057 eligible participants, 1794 had complete covariates and 1438 had week-48 CD4+ T-cell counts. B/F/TAF versus DTG/3TC and DTG/RPV differences were -7.4 (95% CI -45.1 to 30.4) and +5.6 (-24.6 to 35.7) cells/μL (both Holm p = 1.000). Additional baseline CD4+ T-cell percentage adjustment changed the former to -39.8 (-78.5 to -1.2). CD4+ T-cell percentage differences were + 1.66 and + 1.44 percentage points (both Holm p ≤ 0.002). Original secondary CD8+ T-cell count differences were - 92.9 and - 96.5 cells/μL (both Holm p = 0.007); the ratio difference versus DTG/3TC was +0.165 (Holm p = 0.013). CONCLUSIONS: In these three Spanish cohorts, the primary propensity score-adjusted analysis did not detect a difference between regimens in the week-48 CD4+ T-cell count. Original secondary analyses associated B/F/TAF with lower CD8+ T-cell counts than both dual regimens and a higher CD4+/CD8+ T-cell ratio than that of DTG/3TC. The CD8+ T-cell count and ratio contrasts versus DTG/3TC attenuated after additional baseline percentage adjustment. These associations were sensitive to baseline immune adjustment, and their clinical relevance warrants further investigation.