Comparative genotypic study of viral RNA and proviral DNA in HIV-1 patients

Abstract

Highly Active Antiretroviral Therapy (HAART) has been used for HIV-1 treatment for more than a decade. Prior identification of drug resistance associated mutations and viral tropism provides valuable information for physicians in optimizing the best HAART regimens for each patient. Plasma RNA has long been used as the major laboratory marker as it reflects the status of circulating virus. However, it might not be as informative in patients with low viral loads. Proviral DNA was therefore suggested as an additional maker for this purpose.

This project aimed to compare the concordance of resistance interpretation and tropism prediction between plasma RNA and proviral DNA. Peripheral blood samples from 45 HIV-1 patients were collected by Integrated Treatment Centre. Drug resistance mutations (DRMs) of Protease Inhibitors (PIs), Nucleoside Analogue Reverse transcriptase Inhibitors (NRTIs), Non- Nucleoside Analogue Reverse transcriptase Inhibitors (NNRTIs) and Integrase Inhibitors (INIs), and viral tropism were identified with bi-directional Sanger sequencing and interpreted with Stanford HIVdB and Geno2Phenocoreceptor, respectively.

The DRMs identified from plasma RNA and proviral DNA had high concordances among PIs (83.3%), NRTIs (100%), NNRTIs (83.3%) and INIs (91.7%). Major PI mutations were more frequently detected by proviral DNA. In contrast, more NNRTI and accessory INI mutations were detected in plasma RNA. The concordance was lower in HIV-1 tropism test (73.7%). In general, proviral DNA identified more non-R5 variants than plasma RNA. Ten discordant cases were only found in subtype B and CRF01_AE and were mainly contributed by cases (6 out of 10) with R5 tropic variants detected in plasma RNA and non-R5 tropic variants detected in proviral DNA (R5/non-R5). False positive rate (FPR) values obtained were compared and found high concordance that correlation coefficient r equals to 0.83.

In conclusion, proviral DNA and plasma RNA were highly concordant in DRMs interpretation. Since mutations exclusively detected in each genotype, proviral DNA was suggested to be used as a supplementary source of DRMs detection. Frequent non-R5 variant detection in proviral DNA could discourage usage of CCR5 antagonists. Before clinical relevance of proviral DNA tropism is validated, plasma RNA tropism identification was preferred.