Innovative and affordable HIV-1 drug resistance testing for resource limited settings.

Abstract

Background: HIV drug resistance (HIVDR) remains a major threat to achieving sustainable viral suppression on antiretroviral treatment (ART). Most countries including those in resource limited settings (RLS) have adopted use of dolutegravir (DTG), a more potent integrase strand transfer inhibitor, leading to an increase in the demand for integrase resistance testing. Current HIVDR testing methods in RLS focus on genotyping the HIV protease (PR) and reverse transcriptase (RT) genes, separate from the integrase (IN) gene. However, amplification of PR and RT separate from IN is expensive and increases the workload for HIVDR genotyping. Therefore, affordable and labour efficient methods that genotype all relevant HIV-1 genes (i.e., the PR, RT and IN genes) are required to guide clinical decisions, especially in RLS where cost is a major limiting factor. Thus, this study aimed to design an affordable in-house HIVDR genotyping method suitable for use in RLS. Methods: Remnant plasma samples were obtained from a CAPRISA 103 study and viral RNA was extracted from 500μl of plasma. We validated the assay using remnant plasma samples from an external quality assessment (EQA) programme. Complimentary DNA synthesis and first-round PCR were performed followed by second-round nested PCR which was designed to amplify an ~2.9kb HIV-1 pol region (PR, RT and IN genes) using 1% gel electrophoresis. Successful second-round nested PCR products were purified using ExoSAP-IT Express PCR Product Cleanup reagent. Sanger sequencing was performed and quality of the sequences were manually edited using Geneious Prime software. HIVDR mutations were assessed using the Stanford HIV drug resistance database. HIVDR mutations using the designed method were compared to previous results obtained on the same samples. Sequence quality was also evaluated using phylogenetic analysis in Geneious software with maximum likelihood tree reconstruction using a generalized time reversible model with proportion of invariable sites and gamma distribution (GTR + I + G), and with 100 bootstrap replicates. Method cost-estimates were done by comparing costs and turn-around time to current genotyping methods. Results: Of 115 plasma samples obtained, 19 samples were not processed due to inadequate plasma volume. Of the 96 processed, we obtained sequence data for 78 (81%). Of those, 75 (96%) had at least one HIVDR mutation in the PR and RT

genes, with no major-IN mutations observed. Only one sample had an E157Q INSTI-accessory mutation. When compared to previous genotypes, only 2/79 (3%) had different phenotypic predictions that affected the choice of subsequent regimens. Of 7 EQA samples, 4 were HIV-1C, 2 were HIV-1D, and 1 was HIV-1A. Genotypic resistance data generated using the IDR method showed 100% concordance with EQA panel results. The overall cost per sample was estimated at ~US$43, with a turn-around time of ~15 hours. Conclusion: We successfully designed an in-house HIVDR method suitable for genotyping HIV-1C PR, RT and IN genes, at an affordable cost of US$64 and shorter turn-around time reduced from ~21 hours to ~15 hours, compared to currently available methods. This HIVDR genotyping method accommodates changes in ART regimens and will help to guide HIV-1 treatment decisions in RLS.