HIV drug resistance detected during acute infection may be a result of transmission from an infected partner, selection by pre-exposure prophylaxis (PrEP), or error prone replication. Here we describe the long-term dynamics of resistance detected during early acute HIV infection.
The Partners PrEP Study was a randomized trial in which serodiscordant partners were assigned to emtricitibine co-formulated with tenofovir (FTC/TDF) or TDF alone compared to placebo, and monitored for HIV infection monthly. In a previous study, baseline samples from 137 HIV serocovnverters were tested for drug resistance using 454 ultra deep sequencing at the time seroconversion was first detected. Here, we conducted resistance testing by 454 on plasma samples at 6, 12 and 24 months following seroconversion from individuals with resistance mutations detected during acute infection at frequencies ≥1% that had a Stanford score ≥30 and cause resistance to non-nucleoside or nucleoside reverse-transcriptase inhibitors (NNRTIs or NRTIs), including both PrEP and non-PrEP antiretrovirals.
There were 35/137 (26%) individuals with resistance detected at frequencies ≥1% at the time seroconversion was first identified: 30 had resistance to non-PrEP antiretrovirals (not TDF or FTC) likely due to transmitted resistance or error-prone replication and 11 individuals had PrEP-related mutations (K65R, K70E and M184IV); of whom 6 had both PrEP-related and non-PrEP mutations. Of the 35, 31 individuals had resistance results available through 12-24 months. Resistance faded to frequencies <1% in 16/31 (52%) individuals by 6 months after seroconversion and in 21/31 (68%) by 24 months. PrEP-selected mutations did not persist. In 5 individuals, resistance persisted at low frequencies between 1-10% of the viral population and in 3 individuals, resistance mutations K103N (n=2) or Y181C (n=1) persisted at frequencies ≥ 99% throughout follow-up.
Among individuals who acquired HIV in a PrEP trial, non-PrEP (non TDF or FTC) mutations accounted for the majority of resistance detected during acute infection. Persistence of high frequency resistance was limited to 2 NNRTI mutations that were most likely transmitted. Most of the resistance present during acute infection faded below detection by 24 months regardless of whether it was selected or transmitted. These data suggest that in the absence of antiretroviral treatment, onward transmission of resistance is most likely to occur early after HIV infection.