926    

Background:  Incomplete HIV suppression due to insufficient or erratic drug exposure is associated with evolution of genotypic resistance mutations. We examined pharmacokinetic data in all children receiving combination therapy with ritonavir/indinavir (RTV/IDV) and evolution of resistance mutations in the subset of patients who discontinued (d/c) RTV/IDV.

Methods: P1013 is a prospective, randomized trial of IDV with either boosting or therapeutic doses of RTV (125 mg/m² or 300 mg/m² every 12 hours, respectively) and >1 NRTI. Children 2 to 18 years of age failing >16 weeks of combination therapy with HIV RNA >10,000 copies/mL and <3 primary protease gene mutations were eligible.  Genotypes at study screen and RTV/IDV d/c were scored using the Stanford Algorithm. Intensive pharmacokinetic studies were done in each subject at week 4 of RTV/IDV.  Predose IDV concentrations (ICpre) at weeks 4, 8, 12, 16, and 24 and then every 12 weeks were used to assess adherence. Nonadherence was defined as an ICpre more than 50% below the lowest C12 hours (threshold) from the 4-week pharmacokinetic study.

Results:  Of 31 patients enrolled, 10 sustained HIV RNA decrease >1.0 log10</log> from baseline (median time on RTV/IDV = 71 weeks), despite 7 having 1 or 2 primary protease mutations at screening. Of 10 with virologic decrease, 7/7 with evaluable ICpre data were adherent at 94% of visits. Of 21 who d/c RTV/IDV, 6 were nonevaluable (4 for RTV/IDV refusal within 5 weeks and 2 with incomplete genotype data). Of the 15 evaluable subjects, 14 had virologic failure and 1 refused RTV/IDV at week 69 (median time to d/c = 32 weeks [20 to 94 weeks]) with a median of 50% of visits with ICpre below the adherence threshold. 12/15 had = 1 undetectable ICpre in the visit(s) preceding virologic failure. Median time to RTV/IDV d/c in 6 subjects with new primary resistance mutations was 47 weeks and in 9 subjects without such mutations 25 weeks (p = 0.02). Among the 15 subjects who d/c’d RTV/IDV, there was a significant increase in Stanford score for all PI from screening (median score = 4, susceptible) to d/c (median score = 10 to 44; potentially low to intermediate genotypic resistance) (p = 0.03), but not for NRTI or NNRTI. 

Conclusions: Pharmacokinetic measures of nonadherence preceded virologic failure in 80% of subjects discontinuing RTV/IDV.  In 40% (6/15) evolution of significant genotypic mutations may have contributed to virologic failure. In addition, subjects failing this combination may also have increased resistance to other PI as well as to RTV/IDV.

Keywords: adherence; genotypic mutations; pharmacokinetics