Comparative evaluation of seven resistance interpretation algorithms and their derived genotypic inhibitory quotients for the prediction of 48 week virological response to darunavir-based salvage regimens

Background: the darunavir genotypic inhibitory quotient (gIQ) has been suggested as one of the predictors of virological response to darunavir-containing salvage regimens. Nevertheless, which resistance algorithm should be used to optimize the calculation of gIQ is still debated. The aim of our study was to compare seven different free-access resistance algorithms and their derived gIQs as predictors of 48 week virological response to darunavir-based salvage therapy in the clinical setting. Methods: patients placed on two nucleoside reverse transcriptase inhibitors + 600/100 mg of darunavir/ritonavir twice daily  ±  enfuvirtide were prospectively evaluated. Virological response was assessed at 48 weeks. Darunavir resistance interpretation was performed according to seven different algorithms, of which two were weighted algorithms. Analysis of other factors potentially associated with virological response at 48 weeks was performed. Results: fifty-six treatment-experienced patients were included. Overall, 35 patients (62.5%) had a virological response at 48 weeks. Receiver operator characteristic curve analysis showed that De Meyer's weighted score (WS) and its derived gIQ (gIQ WS) were the most accurate parameters defining virological response, and related cut-offs showed the best sensitivity/specificity pattern. In univariate logistic regression analysis, baseline log viral load (P = 0.028), optimized background score ≥ 2 (P = 0.048), WS >5 (P = 0.001) and WS gIQ ≥ 600 (P < 0.0001) were independently associated with virological response. In multivariate analysis, only baseline log viral load (P = 0.008) and WS gIQ ≥ 600 (P < 0.0001) remained in the model. Conclusions: in our study, although different resistance interpretation algorithms and derived gIQs were associated with virological response, gIQ WS was the most accurate predictive model for achieving a successful virological response.