Background:
Nirmatrelvir-ritonavir (N-R) and remdesivir (RDV) are SARS-CoV-2 antivirals that reduce the risk of hospitalization and progression to severe COVID-19. N-R and RDV resistance have been described previously, but the frequency and risk factors for emergent drug resistance remains unclear.
Methods:
We enrolled non-hospitalized participants with acute SARS-CoV-2 infection into the POSITIVES study, a prospective observational cohort, where we collected anterior nasal swabs thrice weekly during the first two weeks after diagnosis. From these samples, we performed deep sequencing of [i]nsp5[/i] among N-R treated (n=53) and untreated (n=42) participants and [i]nsp12[/i] among RDV treated (n=14) participants. We compared the incidence of emergent N-R resistance between N-R treated and untreated participants and evaluated its association with post-treatment virologic rebound, while also characterizing emergent RDV resistance.
Results:
Compared with untreated individuals, those treated with antivirals were older, more immunosuppressed, and had received more COVID-19 vaccinations, reflecting guidelines for N-R and RDV use. Emergent N-R mutations expected to at least confer moderate resistance (≥2.5-fold reduced susceptibility to N-R in vitro) were detected more often in those who received N-R than those who did not (5/53 [9%] vs 0/42 [0%], p=0.06). However, these mutations (E166V, H172Y, Q189K, P252L) were detected at low frequencies, with all but one mutation present in <25% of the viral population. Additionally, for those with detectable viral loads at follow-up timepoints after cessation of treatment, all (4/4) of the emergent resistance mutations subsequently reverted to wild type. Viral rebound occurred in 28% (15/53) of participants receiving N-R, but there was no difference in resistance emergence in those who experienced virologic rebound compared to those who did not (2/15 [13%] vs 3/38 [8%], p=0.6). Emergent RDV resistance mutations were detected in two immunosuppressed participants (14%). However, similar to the N-R mutations, all four mutations (V166L, N198S, V792I, M794I) were low-frequency and reverted to wild type at successive timepoints.
Conclusions:
Mutations that confer resistance emerge with N-R and RDV treatment, but they are transient, present at minor frequencies, and are not associated with virologic rebound. These data suggest that the risk of widespread dissemination of significant drug resistance to N-R and RDV remains low.