Background:  We evaluated the prevalence and effect of transmitted low-level HIV-1 drug resistance detected by the oligonucleotide ligation assay (OLA, sensitivity = 5%), but not by consensus sequencing in subjects infected with HIV-1 after 1996.

Methods:  We performed consensus sequencing and OLA on the first available plasma and peripheral blood mononuclear cell (PBMC) specimens from subjects in an observational cohort at the University of Washington Primary Infection Clinic. OLA assessed mutations in reverse transcriptase (K65R, K70R, L74V, M184V, T215F/Y, K103N, Y181C, and G190A) and protease (D30N, I50V, V82S/A/T, I84V, N88D, and L90M). We compared the number of subjects with resistance detected by each method by McNemar exact tests. The Stanford University HIV Drug Resistance Database was used to predict the effect of mutations on ARV activity. Survival analyses evaluated the time to HIV-1 RNA <50 copies/mL among a subset of subjects treated with ARV and were adjusted for HIV-1 RNA levels at start of ARV.

Results:  Specimens from 100 subjects were obtained a median of 30 days after the estimated date of HIV-1 infection. Consensus sequencing detected resistance in 6 subjects, and OLA detected low-level mutations in 28 additional subjects. Compared to consensus sequencing, OLA detected more subjects with mutations in plasma (p <0.0001) and PBMC (p = 0.001). Of 94 subjects, 83 (88%) without mutations detected by consensus sequencing began ARV a median of 45 (IQR 24 to 107) days after infection. Median time to HIV-1 RNA <50 copies/mL was 109 (IQR 63 to 145) days for 55 subjects without mutations, 105 (IQR 57 to 163) days for 14 subjects with low-level mutations treated with <3 active ARV, and 82 (IQR 57 to 105) days for 10 subjects with low-level mutations treated with ≥3 active ARV (p = 0.1). Compared to subjects without mutations, adjusted hazard ratios for time to HIV-1 RNA <50 copies/mL were 1.1 (95%CI 0.5 to 2.3, p = 0.8) for subjects with low-level mutations treated with <3 active ARV and 2.3 (95%CI 1.1 to 4.6, p = 0.02) for subjects with low-level mutations treated with ≥3 active ARV. Only 4 subjects experienced virologic failure during follow-up.

Conclusions:  Consensus sequencing underestimated the prevalence of transmitted HIV-1 drug resistance in ARV-naïve subjects with primary HIV-1 infection. In this pilot project, we found no association between low-level mutations and delayed time to HIV-1 RNA <50 copies/mL, possibly due to the small sample size. With advances in ARV potency, it is also possible that low-level resistance mutations are only clinically relevant at critical codons.