Abstract
Second-line therapy randomized trials with lopinavir/ritonavir question the value of resistance testing to guide nucleoside reverse transcriptase inhibitor (NRTI) selection. In this study, we investigated the association between baseline drug resistance and treatment outcome after 104 weeks of second-line therapy with NRTIs and either darunavir/ritonavir or lopinavir/ritonavir in West-central Africa. We did an observational analysis of data from 387 individuals in a randomized, open-label 2LADY trial in Burkina Faso, Cameroon, and Senegal. We modeled the association between RTI drug resistance mutations (DRMs) and virological failure (VF) (viral load [VL] \textless50 copies/mL) at week 104 using logistic regressions. Covariates included baseline VL and CD4+ count, demographic, and adherence data. Overall, 193 (49.9%), 150 (38.8%), and 44 (11.4%) individuals had, respectively, low/none (genotypic susceptibility score [GSS] \textless1), intermediate (GSS = 1), and high predicted NRTI activity (GSS \textgreater1) in their prescribed second-line regimen. The average number of DRMs by drug class, the proportion of individuals by GSS category, and the duration of first-line therapy were not associated with VF (p \textgreater .05). High VL at switch was the only consistent prognostic factor for VF after multivariate adjustment (p \textless .01). Suboptimal adherence, high predicted RTI activity, or low NRTI mutations were associated with VF (p \textless .05) when using higher end points for VF or in the intention-to-treat analysis. In conclusion, the use of RTIs with predicted reduced activity does not impair second-line protease inhibitor-based therapy. Therefore, HIV care in resource-limited settings should prioritize strategies to improve adherence and targeted VL testing over drug resistance testing for selecting NRTIs during a protease-based second-line switch.