Oral Nirmatrelvir for High-Risk, Nonhospitalized Adults with Covid-19

Abstract

Background: Nirmatrelvir is an orally administered severe acute respiratory syndrome coronavirus 2 main protease (M^pro) inhibitor with potent pan-human-coronavirus activity in vitro.

Methods: We conducted a phase 2-3 double-blind, randomized, controlled trial in which symptomatic, unvaccinated, nonhospitalized adults at high risk for progression to severe coronavirus disease 2019 (Covid-19) were assigned in a 1:1 ratio to receive either 300 mg of nirmatrelvir plus 100 mg of ritonavir (a pharmacokinetic enhancer) or placebo every 12 hours for 5 days. Covid-19-related hospitalization or death from any cause through day 28, viral load, and safety were evaluated.

Results: A total of 2246 patients underwent randomization; 1120 patients received nirmatrelvir plus ritonavir (nirmatrelvir group) and 1126 received placebo (placebo group). In the planned interim analysis of patients treated within 3 days after symptom onset (modified intention-to treat population, comprising 774 of the 1361 patients in the full analysis population), the incidence of Covid-19-related hospitalization or death by day 28 was lower in the nirmatrelvir group than in the placebo group by 6.32 percentage points (95% confidence interval [CI], -9.04 to -3.59; P<0.001; relative risk reduction, 89.1%); the incidence was 0.77% (3 of 389 patients) in the nirmatrelvir group, with 0 deaths, as compared with 7.01% (27 of 385 patients) in the placebo group, with 7 deaths. Efficacy was maintained in the final analysis involving the 1379 patients in the modified intention-to-treat population, with a difference of -5.81 percentage points (95% CI, -7.78 to -3.84; P<0.001; relative risk reduction, 88.9%). All 13 deaths occurred in the placebo group. The viral load was lower with nirmatrelvir plus ritonavir than with placebo at day 5 of treatment, with an adjusted mean difference of -0.868 log₁₀ copies per milliliter when treatment was initiated within 3 days after the onset of symptoms. The incidence of adverse events that emerged during the treatment period was similar in the two groups (any adverse event, 22.6% with nirmatrelvir plus ritonavir vs. 23.9% with placebo; serious adverse events, 1.6% vs. 6.6%; and adverse events leading to discontinuation of the drugs or placebo, 2.1% vs. 4.2%). Dysgeusia (5.6% vs. 0.3%) and diarrhea (3.1% vs. 1.6%) occurred more frequently with nirmatrelvir plus ritonavir than with placebo.

Conclusions: Treatment of symptomatic Covid-19 with nirmatrelvir plus ritonavir resulted in a risk of progression to severe Covid-19 that was 89% lower than the risk with placebo, without evident safety concerns. (Supported by Pfizer; ClinicalTrials.gov number, NCT04960202.).

Figure 1. Randomization, Treatment Assignments, and Follow-up.

Patients were recruited through December 9, 2021, from the United States (105 sites), Bulgaria (30 sites), South Africa (28 sites), Brazil (26 sites), India (19 sites), Mexico (18 sites), Ukraine (17 sites), Turkey (16 sites), Japan and Spain (10 sites each), Russia (9 sites), Argentina and Colombia (8 sites each), Poland and South Korea (7 sites each), Hungary (6 sites), Taiwan (5 sites), Malaysia and Czech Republic (4 sites each), and Thailand and Puerto Rico (3 sites each).

Figure 2. Efficacy of Nirmatrelvir plus Ritonavir (NMV-r) in Preventing Covid-19–Related Hospitalization or Death from Any Cause through Day 28.

Panel A shows efficacy results among patients who were treated within 3 days and within 5 days after symptom onset and who did not receive or were not expected to receive Covid-19 therapeutic monoclonal antibodies at randomization. The average time at risk for an event was computed as the time to the first event or as the time to the last day of participation or day 28, whichever was earlier. The average study follow-up was computed as the time to the last day of participation or day 28, whichever was earlier. Panel B shows the cumulative percentage of patients with Covid-19–related hospitalization or death from any cause through day 28 among patients treated within 5 days after symptom onset. The cumulative percentage was estimated for each treatment group with use of the Kaplan–Meier method. The inset shows the same data on an expanded y axis. Panel C shows subgroup analysis of the differences of the proportions of patients treated within 5 days after symptom onset who had Covid-19–related hospitalization or death from any cause through day 28, estimated for each treatment group with use of the Kaplan–Meier method. P values are based on normal approximation of the data. Study populations are described in Table S2.

Figure 3. Change from Baseline in Log₁₀-Transformed Viral Load over Time (Modified Intention-to-Treat Population).

Panel A shows the adjusted mean change in viral load from baseline among all the patients who received at least one dose of the drug or placebo, had at least one visit between day 1 and day 28, did not receive or were not expected at baseline to receive Covid-19 therapeutic monoclonal antibody treatment, and were treated within 3 days after the onset of Covid-19 (modified intention-to-treat population). Panel B shows findings for the subgroup of patients whose baseline SARS-CoV-2 serology status was negative, and Panel C shows findings for the subgroup of patients whose baseline SARS-CoV-2 serology status was positive. Panel D shows findings among patients whose baseline viral load was more than 10⁴ copies per milliliter, and Panel E shows findings among patients whose baseline viral load was more than 10⁷ copies per milliliter. Patients were excluded from the analysis if the viral load was not detected or if data on baseline viral load were missing. Results obtained with unvalidated swabs were also excluded. Results were obtained with the use of a mixed-effects repeated-measures analysis of covariance model. Treatment, visit, and visit-by-treatment interactions were fixed effects in the analysis. Geographic region, baseline SARS-CoV-2 serology status, baseline viral load, and nasopharyngeal sample site were covariates, and participant was a random effect.