Prevalence of gag mutations associated with in vitro resistance to capsid inhibitor GS-CA1 in HIV-1 antiretroviral-naive patients

Sir, HIV capsid (CA) plays an essential role in multiple stages of the viral replication cycle, but it remains an unexplored target for ART. Previous in vitro studies have described a novel class of exquisitely potent and metabolically stable HIV capsid inhibitors (CAIs).1 Recently, preclinical studies have revealed a highly potent CAI, GS-CA1, with an EC50 of 140 pM, which is active against all major HIV-1 group M clades.2,In vitro resistance selection experiments were performed with GS-CA1 and identified five amino acid substitutions in CA protein that independently confer resistance to GS-CA1: L56I, M66I, Q67H, N74D and A105E, all located at the GS-CA1 binding site in CA.2 The aim of this study was to assess the prevalence of CA nucleotide substitutions previously described as being associated with in vitro resistance to the HIV-1 CAI GS-CA1 in antiretroviral-naive patients using ultra-deep sequencing technology at baseline and at time of virological failure (VF) in patients receiving a first-line PI-based regimen. In an observational single-centre cohort, we assessed all antiretroviral-naive patients initiating a first-line regimen including two NRTIs and one PI (i.e. darunavir or atazanavir boosted with ritonavir) between January 2012 and March 2015. VF was defined as two consecutive plasma viral load (VL) values >50 copies/mL following virological suppression or as the absence of reaching VL <50 copies/mL 6 months after antiretroviral initiation. All the patients gave their written informed consent to have their medical chart recorded in the electronic medical record system Nadis® (www.dataids.org; Fedialis Medica, Marly Le Roi, France; CNIL number: 1171457; 24 May 2006), designed for the healthcare medical follow-up of HIV-infected patients. Ultra-deep sequencing of the CA region was assessed at baseline in 137 patients and at time of VF in 14 patients experiencing VF. HIV-1 RNA was extracted from 400 μL of plasma using the EZI Virus Mini Kit v. 2.0® (Qiagen, Hilden, Germany). First, a 1926 bp fragment of the gag gene was amplified from viral RNA using the PCR SuperScript™ III One-Step RT-PCR System with Platinum® Taq High Fidelity (Thermo Fisher, Waltham, MA, USA), according to the manufacturer’s instructions and using the following primers from Gall et al.:3 Pan-HIV-1_1F (5′-AGCCYGGGAGCTCTCTG) and Pan-HIV-1_1R (5′-CCTCCAATTCCYCCTATCATTTT). A 1819 bp fragment was then amplified by a second-round of PCR using the primers: GAGPCRNEST_F (5′-TAAAGCTTGCCTTGAGTGC) and GAGPCRNEST_R (5′-TACTGTATCATCTGCTCCTG). Ultra-deep sequencing was performed using Illumina MiSeq technology (Illumina, San Diego, CA, USA) according to the manufacturer’s instructions. CA sequences were analysed using Geneious software v. 9.0.4 (https://www.geneious.com).4 Only mutations above 2% were retained. CA ultra-deep sequencing was successful at antiretroviral initiation in 132 samples of the 137 tested (96.4%). The distribution of HIV-1 subtypes was as follows: 42 (31.8%) subtype B, 48 (36.4%) CRF02_AG and 42 (31.8%) other ‘non-B’ subtypes. None of the five mutations identified during the in vitro selection experiments was detected among the 132 sequences, either in majority or minority proportion. Regarding the 14 samples obtained at time of VF, no selection of any of these five mutations was observed. The present study, performed on a large dataset of CA sequences obtained from antiretroviral-naive patients and representing a variety of HIV-1 group M subtypes, showed the absence in pre-therapeutic samples of the five GS-CA1 resistance-associated mutations recently described in vitro. The newest antiretroviral classes still under development, such as attachment or maturation inhibitors, can be impacted by the presence of pre-existing mutations. Thus, gag cleavage site mutations decreased susceptibility to the first generation of maturation inhibitors.5,6 Similarly, in vitro phenotypic susceptibility of attachment inhibitors can be impaired by the presence of pre-existing mutations in the CD4 binding site.7,8 However, regarding the CAI drug class, particularly GS-CA1, our findings suggest no need for pre-therapeutic genotypic resistance testing before using the first compound of this new antiretroviral drug class, which is currently at the preclinical development stage. The research leading to these results has received funding from the French National Agency for Research on AIDS and Viral Hepatitis (ANRS) D. D. has received honoraria and travel grants from ViiV Healthcare, Janssen, Gilead Sciences and MSD for participation in advisory boards, educational programmes and international conferences. C. C. has received honoraria and travel grants from ViiV Healthcare, Janssen, Gilead Sciences and MSD for participation in advisory boards, educational programmes and international conferences. All other authors: none to declare.