First prospective comparison of genotypic versus phenotypic tropism assays in predicting virologic responses to maraviroc in a phase 3 study.

Maraviroc (MVC, a CCR5 antagonist) is only fully active against CCR5 tropic [R5] HIV-1, and tropism testing is required prior to initiating treatment. The MODERN study prospectively compared genotypic (GTT) and phenotypic (Trofile®) tropism testing with treatment-naive HIV-1-infected participants randomized 1:1 to either GTT or Trofile tropism assessments. Participants with R5 virus were randomized 1:1 to receive darunavir/ritonavir (DRV/r) with either MVC or tenofovir/emtricitabine. Screening samples were also retrospectively tested using the alternative assay. Positive predictive values (PPVs) for each assay were estimated using both the observed MVC+DRV/r response rate (HIV-1 RNA <50 copies/mL at Week 48) and model-based response estimates. The observed MVC+DRV/r response rate was 146/181 (80.7%) for GTT versus 160/215 (74.4%) for Trofile, with a stratification adjusted difference of 6.6% (95% CI, -1.5% to 14.7%) in favor of GTT. The model-based PPV estimates (±standard error) were 80.5% (±2.38) and 78.0% (±2.35) for GTT and Trofile, respectively (difference, 2.5%; 95% CI, -2.0% to 7.0%). Most participants had R5 results using both assays (285/396; 72%) and, of those, 79.3% (226/285) had HIV-1 RNA <50 copies/mL at Week 48. Both the genotypic and phenotypic tropism assays evaluated can effectively predict treatment response to MVC.