The impact of drug resistance mutations in plasma virus of patients failing on protease inhibitor-containing HAART regimens on subsequent virological response to the next HAART regimen: results of CPCRA 046 (GART)
ABSTRACT: Objective: To determine the impact of drug resistance mutations in plasma virus of patients failing on a protease inhibitor-containing HAART regimen on the short-term virological response to the next round of HAART therapy.
Methods: Baseline genotypes were obtained for plasma virus from 152 patients entering a multicenter, randomized trial of the clinical utility of GART who had HIV RNA levels (bDNA) available at 4 and 8 weeks following change to a new HAART regimen. The viral load (VL) response (log10) was the change in log10 plasma HIV RNA from baseline to the average of the week 4 and 8 values. in this analysis, the GART and no GART groups were combined and only baseline mutations with 5% frequency were evaluated: T69D, M184V and T215Y/F in RT; and D30N, M46I/L, V82A/F/T, 184V andL90M in HIV protease. A multiple regression analysis was performed of the VL response to the next HAART regimen associated with the presence/absence of each mutation, adjusted for baseline CD4 and HIV RNA, with/without the inclusion of prior drug history.
Results: In multivariate models of RT mutations, the T69D mutation had a significant adverse impact on virologic response (log10 VL +0.55, P=0.02), which persisted even after adjustment for NRTI exposure history (P=0.04). The T215Y/F and M184V mutations, which were present in most patient samples, did not have a significant impact. In multivariate models of PI mutations, the D30N PI-resistance mutation had a positive effect (log10 VL -0.41, P=0.04) whereas the L90M mutation had an adverse impact (log10 VL +0.31, P=0.04) on virological responses. The M46I/L, V82A/F/T, and I84V mutations had no significant impact on virologic responses. The positive benefit of D30N persisted after adjusting for PI exposure history (log10 VL -0.47, P=0.06).
Conclusions: The T69D RT mutation and the L90M PI mutation had a significant adverse impact on response to the next HAART regimen whereas the D30N mutation had a beneficial effect compared to all of the other PI-mutations evaluated. Specific drug resistance mutations in plasma virus of patients failing on HAART regimens were associated with the short-term virological responses to subsequent rounds of HAART.