Ensitrelvir treatment-emergent amino acid substitutions in SARS-CoV-2 3CLpro detected in the SCORPIO-SR phase 3 trial.

Abstract

The impact of treatment-emergent amino acid substitutions (TEAASs) in SARS-CoV-2 3C-like protease (3CL^pro) on clinical and virologic outcomes was evaluated in patients with mild-to-moderate COVID-19 who received ensitrelvir 125 mg in the SCORPIO-SR trial. Individuals were randomised to ensitrelvir or matched placebo once daily for 5 days (first dose <72 h after disease onset). 3CL^pro-TEAASs were identified by sequencing nsp5 encoding 3CL^pro from pre- and post-treatment nasopharyngeal swabs. Time to resolution of a composite of five characteristic COVID-19 symptoms (TTR) was compared between patients with and without most common 3CL^pro-TEAASs in the ensitrelvir arm. The ensitrelvir and placebo intention-to-treat populations comprised 345 and 341 patients, respectively. 3CL^pro-TEAASs were detected in 19/204 (9.3%) ensitrelvir-treated and 3/137 (2.2%) placebo-treated patients with paired sequence data. The most common 3CL^pro-TEAASs in the ensitrelvir arm were M49L (n=12), M49I (n=3) and S144A (n=2). In the placebo arm, all 3CL^pro-TEAASs occurred in ≤1 patient. Median (95% confidence interval) TTR was comparable between patients with and without those TEAASs (158.8 h [112.1-281.9] vs 189.7 h [151.4-234.4]). Mean viral RNA levels declined more slowly in patients with M49L/I or S144A versus those without. Reductions in viral titre were unaffected by these TEAASs. The characteristics of recombinant SARS-CoV-2 with 3CL^pro mutations were explored in vitro. Recombinant viruses with some 3CL^pro mutations had reduced susceptibility to ensitrelvir in vitro, with limited effects on viral and competitive fitness. Continued surveillance is warranted to monitor the spread of viruses with 3CL^pro mutations.