Low abundance drug resistance variants in transmitted HIV drug resistance surveillance specimens identified using tagged pooled pyrosequencing

Abstract

HIV drug resistance (DR) testing using Sanger sequencing (SS) is limited by the inability of the method to identify low abundance drug resistance variants. The application of tagged pooled pyrosequencing (TPP) for HIV DR surveillance is described and the results compared with SS. HIV(+) serum specimens were genotyped using both SS and TPP. Surveillance drug resistance mutations were identified using SS and TPP consensus reads at multiple mixed base identification thresholds (MBITs). Drug resistance patterns were highly concordant between SS and TPP when the MBIT was set at 20%. DR mutations were detected in 7.1% of the subjects, with 1.6% of individuals harboring resistance to NRTI, 3.3% NNRTI and 2.7% PI. Analyzing the TPP reads for each subject confirmed that drug resistance mutations with frequencies <20% were inconsistently detected by SS. Conversely, low abundance drug resistant variants were easily identified using TPP with mixed base identification threshold set at low value. In conclusion, at considerable savings when compared to commercial assays, TPP produces HIV DR profiles that are concordant with those from SS, furthermore, these same data can be used to identify low abundance drug resistant variants.