Maraviroc, as a switch option, in HIV-1-infected individuals with stable, well-controlled HIV replication and R5-tropic virus on their first Nucleoside/Nucleotide reverse transcriptase inhibitor plus ritonavir-boosted protease inhibitor regimen: Week 48 Results of the randomized, multicenter March study

Background. Alternative combination antiretroviral therapies in virologically suppressed human immunodeficiency virus (HIV)-infected patients experiencing side effects and/or at ongoing risk of important comorbidities from current therapy are needed. Maraviroc (MVC), a chemokine receptor 5 antagonist, is a potential alternative component of therapy in those with R5-tropic virus. Methods. The Maraviroc Switch Study is a randomized, multicenter, 96-week, open-label switch study in HIV type 1-infected adults with R5-tropic virus, virologically suppressed on a ritonavir-boosted protease inhibitor (PI/r) plus double nucleoside/nucleotide reverse transcriptase inhibitor (2 N(t)RTI) backbone. Participants were randomized 1:2:2 to current combination antiretroviral therapy (control), or replacing the protease inhibitor (MVC + 2 N(t)RTI arm) or the nucleoside reverse transcriptase inhibitor backbone (MVC + PI/r arm) with twice-daily MVC. The primary endpoint was the difference (switch minus control) in proportion with plasma viral load (VL) <200 copies/mL at 48 weeks. The switch arms were judged noninferior if the lower limit of the 95% confidence interval (CI) for the difference in the primary endpoint was < -12% in the intention-to-treat (ITT) population. Results. The ITT population comprised 395 participants (control, n = 82; MVC + 2 N(t)RTI, n = 156; MVC + PI/r, n = 157). Baseline characteristics were well matched. At week 48, noninferior rates of virological suppression were observed in those switching away from a PI/r (93.6% [95% CI, -9.0% to 2.2%] and 91.7% [95% CI, -9.6% to 3.8%] with VL <200 and <50 copies/mL, respectively) compared to the control arm (97.6% and 95.1% with VL <200 and <50 copies/mL, respectively). In contrast, MVC + PI/r did not meet noninferiority bounds and was significantly inferior (84.1% [95% CI, -19.8% to -5.8%] and 77.7% [95% CI, -24.9% to -8.4%] with VL <200 and <50 copies/mL, respectively) to the control arm in the ITT analysis. Conclusions