Data for: Learning from full characterization of HIV proviruses in people receiving Long-Acting Cabotegravir/Rilpivirine with history of replication on the antiretroviral classes

To better understand factors associated with virological response, we characterized retrospectively HIV proviruses of seven People-With-HIV who received long-acting cabotegravir/rilpivirine (CAB/RPV-LA) selected according the following criteria: virological control achieved despite a history of viral replication on one or both corresponding antiretroviral classes (n=6); virological failure (VF) post CAB/RPV-LA initiation (n=1). The last available blood samples before the initiation of CAB/RPV-LA were analyzed retrospectively. Near full-length HIV-DNA genome haplotypes were inferred from Nanopore sequencing using in-vivo Genome-Diversity-Analyzer (iGDA) to search for archived Drug Resistance Mutations (DRMs) and to evaluate the frequency and the intactness of proviruses harboring DRMs. Archived DRMs including G-to-A mutations were found in samples from three subjects who maintained virological control. Genomes harboring DRMs were majorly in minority variants (<20%) and were defective in all cases except for one participant. In this participant, intact genomes with H221Y mutation on reverse transcriptase were detected representing 11 copies/106 PBMCs. The other mutations observed in the participants of the study resulted most likely from hypermutations. The subject with VF presented archived mutations, all associated with defects. Other factors could explain this VF. Our findings highlight the difficulty in interpreting the clinical significance of DRMs when detected in proviral DNA and the need to filter out hypermutated sequences. Detected DRMs could be harbored by defective archived genomes unlikely to contribute to treatment failure.

Methods
A near full-length HIV genome nanopore single genome sequencing (NFG) was performed on HIV DNA after limiting dilution to search for archived DRMs, link them with each other, and evaluate the intactness of proviruses harboring these DRMs. A de-hosting step was performed using the VIRiONT pipeline. A custom-designed bioinformatic pipeline adapted from the in-vivo Genome Diversity Analyzer (iGDA) algorithm was developed to recover representative haplotypes using the following initial criteria: Total reads >30, Min length >1,000 bp, Min freq >0.01 (1%), Mean depth >100 reads.