Background
The relevance of drug resistance mutation (DRM) detection in HIV-1 proviral sequencing for the prediction of virological success on antiretroviral therapy (ART) remains unclear. In this retrospective study, we examined archived INSTI DRMs in people with HIV (PWH) virologically suppressed on INSTI-based regimens.
Methods
PWH with at least one HIV-1 integrase proviral sequencing were screened for inclusion. HIV-1 DNA integrase was sequenced using the Sentosa SQ HIV Genotyping Assay. Genotypic resistance was analyzed using the Stanford HIV Drug Resistance algorithm. Inclusion criteria were (i) at least one DRM conferring full-class INSTI resistance (T66K, E92Q, G118R, E138A/K/T, G140A/C/S, Q148H/K/R, N155H or R263K) on proviral sequencing and (ii) virological suppression on an INSTI-based regimen at the time of sequencing. Mutational load was calculated as the product of DRM frequency and the amount of total HIV-1 DNA. Proviral defectiveness was presumed based on the detection of stop codons or APOBEC-induced G-to-A hypermutation (using the Los Alamos HIV Database’s Hypermut 2 algorithm).
Results
From 883 screened PWH since March 2022, 26 were included for analysis (sex assigned at birth: 69% male; median age: 53.6 [IQR 40–61.8]; subtype B: 58%). Identified DRMs were G140S (n=8), R263K (n=8), E138K (n=7), N155H (n=4), G118R (n=2) and E138A (n=1). Individuals were successfully treated with ART containing DTG (n=17), BIC (n=10), CAB (n=1), EVG (n=1) or RAL (n=1) as shown in figure 1A. DRMs were found in presumably defective proviruses in 18 individuals (69%). A history of virological failure on an INSTI-based regimen was reported for 10 individuals. DRMs previously detected in plasma RNA (n=2) were harbored in presumably intact proviruses. Median total HIV-1 DNA was 2,36 log10 copies/106 leukocytes (IQR 2,24–2,65, n=17). Mutational loads are represented in figure 1B. No virological failure on an INSTI-based regimen was reported after proviral DRM detection (median follow-up: 202 days [IQR 105–366]).
Conclusions
In virologically suppressed PWH, we found a low frequency of archived INSTI DRMs. They were not accountable for any virological failure on INSTI-based regimens during follow-up, even though some of the mutational loads were notably high. This finding is consistent with the detection of DRMs mostly in presumably defective proviruses. Together, our results suggest that archived INSTI DRMs should be cautiously interpreted and do not preclude virological success on INSTI-based regimens.