Trends in darunavir resistance-associated mutations and phenotypic resistance in commercially tested United States clinical samples between 2006 and 2012.

Abstract

HIV-1 samples submitted by clinicians from the United States for routine drug susceptibility testing (PhenoSense GT) were evaluated for genotypic and phenotypic resistance to darunavir and other protease inhibitors (PIs). Among these samples (Monogram Biosciences database January 2006-June 2012; N=78,843), isolates harboring zero IAS-USA darunavir resistance-associated mutations (RAMs) increased from 77.7% in 2006 to 92.8% through the first half of 2012 (H1 2012; upward trend, p=0.0008); a downward trend seen for samples with three or more darunavir RAMs (7.5% in 2006 and 2.6% in H1 2012; p=0.002). Among samples with any PI resistance (N=15,932), samples harboring zero darunavir RAMs gradually increased (39.9% in 2006 vs. 55.0% in H1 2012; upward trend, p=0.005), but three or more darunavir RAMs did not change over time (21.7% in 2006 and 19.2% in H1 2012; p=0.27). During this period, the frequency of the 11 individual darunavir RAMs (IAS-USA 2011 list) decreased among all samples. The frequency of each darunavir RAM in PI-resistant samples decreased or remained relatively stable. The prevalence of samples with phenotypic resistance to darunavir (partial-to-full) decreased over time in all samples (8.2% in 2006 vs. 2.3% in H1 2012), as did resistance to other PIs (p<0.006 for all PIs). Phenotypic resistance to darunavir and other PIs also decreased in PI-resistant samples (darunavir: 23.9% in 2006 vs. 17.1% in H1 2012; p<0.013 for all PIs). Since approval of darunavir in 2006, there was a significant decrease in prevalence of samples with genotypic and phenotypic resistance to darunavir in commercially tested HIV-1 isolates. Furthermore, the prevalence of phenotypic resistance to darunavir was lower than all other PIs.