The US HIV guidelines include within-class switches to simpler and less frequent regimens as a valid reason for changing a regimen. With its low risk of cross-resistance and once daily dosing, clinicians may consider substituting dolutegravir (DTG) for raltegravir (RAL) in an effort to improve adherence and durability. Clinical trials have evaluated INSTI intra-class switching, however, real-world assessments of stable RAL to DTG switching is currently lacking. This study tested the hypothesis that patients stably suppressed on RAL and switched to DTG would not significantly differ in risk of virologic failure from patients who continued on RAL.
Using the OPERA longitudinal database, individuals who initiated RAL after their first prospectively-collected visit were identified. Additionally, they were required to have RAL as their first INSTI regimen and achieve stable suppression (2 consecutive VL<75 copies/mL at least 90 days but not more than 365 days apart). Patients on RAL were followed from the time they first achieved stable suppression with a subset of these being followed after a switch to DTG while still stable. The primary outcome was time to virologic failure (VL>200 copies/mL). DTG-switch to RAL-continuation was compared by estimating hazard ratios (HR) with propensity score adjusted Cox proportional hazards models.
Out of 64,759 HIV+ individuals in OPERA, RAL had been prescribed to 9,677 patients. Of 2,755 eligible stable-suppressed patients taking RAL, 229 (8%) switched to DTG, after a median (IQR) of 1,202 (693, 1763) days on RAL. Those continuing RAL and switching to DTG were followed for a median (IQR) of 616 (275,1164) and 337 (145, 564) days, respectively. No difference in risk of virologic failure was observed for patients switching to DTG compared to those continuing on RAL (weighted HR: 0.74, 95% CI: 0.38, 1.42). In sensitivity analyses, findings were robust to alternative definitions of stable-suppression.
Within-class switching from RAL to DTG was found to be equally successful at maintaining stable viral suppression as compared to continuing on RAL in both crude and adjusted models as well as in sensitivity analyses.