Background. In individuals with viral load (VL) suppression on a boosted protease inhibitor (PI) regimen, a switch to raltegravir (RAL) can be an option in case of comorbidities, but the SWITCHMRK trials challenged this strategy. Here, among individuals with VL suppression on a boosted PI, we compared outcomes between those who continued on the same regimen and those who switched to RAL.
Methods. In this cohort study from the French Hospital Database on HIV, each individual who switched to RAL was matched with up to 3 individuals who continued PI, were being followed up during the calendar period of the switch, and had the same duration of VL suppression (both ±6 months). The primary endpoint was a composite endpoint of hospitalization, or AIDS event or death, and secondary endpoints the immunovirologic responses. To control for measured confounders, the inverse probability treatment weighting (IPTW) method was applied to estimate hazards ratios between the 2 groups.
Results. We matched 282 RAL switchers with 838 nonswitchers. Although several variables differed significantly between the groups, including a higher prevalence of comorbidities in the RAL group, the IPTW method yielded standardized differences <10% for all variables. After IPTW, there was no difference in the risk of hospitalization or AIDS event or death between the 2 groups (13.6% and 10.5%, respectively; hazard ratio, 1.16 [95% confidence interval, .74–1.83]) and no difference in the likelihood of virologic failure or CD4 cell gain.
Conclusions. In individuals with controlled VL on a boosted PI regimen who switched to RAL, none of the endpoints differed with nonswitchers after IPTW.