Protease and gag diversity and drug resistance mutations among treatment-naive Mexican people living with HIV

Abstract

Introduction: In Mexico, HIV genotyping is performed in people living with HIV (PLWH) failing their first-line antiretroviral (ARV) regimen; it is not routinely done for all treatment-naive PLWH before ARV initiation. The first nationally representative survey published in 2016 reported that the prevalence of pretreatment drug mutations in treatment-naive Mexican PLWH was 15.5% to any antiretroviral drug and 10.6% to non-nucleoside reverse transcriptase inhibitors (NNRTIs) using conventional Sanger sequencing. Most reports in Mexico focus on HIV pol gene and nucleoside and non-nucleoside reverse transcriptase inhibitor (NRTI and NNRTI) drug resistance mutations (DRMs) prevalence, using Sanger sequencing, next-generation sequencing (NGS) or both. To our knowledge, NGS has not be used to detect pretreatment drug resistance mutations (DRMs) in the HIV protease (PR) gene and its substrate the Gag polyprotein.

Methods: Treatment-naive adult Mexican PLWH were recruited between 2016 and 2019. HIV Gag and protease sequences were obtained by NGS and DRMs were identified using the WHO surveillance drug resistance mutation (SDRM) list.

Results: One hundred PLWH attending a public national reference hospital were included. The median age was 28 years-old, and most were male. The median HIV viral load was 4.99 [4.39-5.40] log copies/mL and median CD4 cell count was 150 [68.0-355.78] cells/mm³. As expected, most sequences clustered with HIV-1 subtype B (97.9%). Major PI resistance mutations were detected: 8 (8.3%) of 96 patients at a detection threshold of 1% and 3 (3.1%) at a detection threshold of 20%. A total of 1184 mutations in Gag were detected, of which 51 have been associated with resistance to PI, most of them were detected at a threshold of 20%. Follow-up clinical data was available for 79 PLWH at 6 months post-ART initiation, seven PLWH failed their first ART regimen; however no major PI mutations were identified in these individuals at baseline.

Conclusions: The frequency of DRM in the HIV protease was 7.3% at a detection threshold of 1% and 3.1% at a detection threshold of 20%. NGS-based HIV drug resistance genotyping provide improved detection of DRMs. Viral load was used to monitor ARV response and treatment failure was 8.9%.

Keywords: Antiretroviral therapy; Gag; HIV drug resistance mutations; HIV genotyping; Human immunodeficiency virus; Next generation sequencing; Protease.

Fig. 1

Phylogenetic tree analysis of HIV gag and protease from ART-naive Mexican individuals. Neighbor joining trees were constructed from the aligned sequences of a protease and b Gag. HIV risk factors are denoted by MSM (men who have sex with men), MSW (men who have sex with women), and MBI (men bisexual). The corresponding subgroup for each reference strain is shown next to the accession number

Fig. 2

HIV Protease drug resistance mutation frequency and levels of PI resistance. a Frequency of minor and major PI resistance mutations identified with NGS at different detection sensitivity thresholds in patients without experience of antiretroviral therapy (n = 96). b Levels of resistance to protease inhibitors calculated with the Stanford HIVdb program using the 20% threshold. ATV/r ritonavir-boosted atazanavir. LPV/r = ritonavir-boosted lopinavir. FPV/r ritonavir-boosted fosamprenavir. IDV/r = ritonavir-boosted indinavir. NFV nelfinavir. SQV/r ritonavir-boosted saquinavir

Fig. 3

HIV Gag drug resistance-associated mutation frequency at different detection sensitivity thresholds (n = 96). Gag cleavage site mutations are indicated with an asterisk