Clinical effectiveness, safety, and viral mutagenicity of oral favipiravir for COVID-19: results from a community-based, open-label, randomized Phase III trial

Abstract

Early community treatment of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) infection may reduce severe coronavirus disease (COVID-19) incidence. We evaluated clinical effectiveness, safety, and SARS-CoV-2 mutagenicity of favipiravir, an oral viral RNA polymerase inhibitor. We performed an open-label, community-based, randomized Phase III trial, recruiting non-hospitalized adults with mild COVID-19 (WHO ordinal severity score [OSS] ≤ 3). Positive cases were invited to web-based self-screening within 24 h using public health data. Exclusion criteria included symptoms for >7 days, pregnancy/breastfeeding, severe renal/liver disease, gout, and licensed antiviral eligibility. Participants were randomized 1:1 to 10 days favipiravir (Day 1: 3,600 mg; days 2-10: 1,600 mg) or no additional treatment. The primary endpoint was worst recorded OSS up to and including Day 15 (intention-to-treat). The target recruitment was 302. Secondary endpoints included adverse event (AE) rate to Day 60, time-to-viral clearance (TTVC), time-to-symptom resolution (TTSR), and SARS-CoV-2 sequencing variant rate (≥5% frequency) at Day 15 (registration ISRCTN: 31062548; EudraCT: 2020-001904-41). A total of 68,788 adults were invited, and 302 (0.4%) were subsequently randomized between December 2020 and July 2022 (favipiravir [n = 152]: standard care [n = 150]). Mean (SD) age was 47.2 (13.2), and 230/302 (76%) were vaccinated. Severe outcomes were infrequent, with no intensive care unit admissions/deaths. There was no difference in the primary endpoint: odds ratio 1.18 (95% confidence interval [CI] 0.63-2.20), TTSR (HR 1.03 [95% CI 0.81-1.31]), or TTVC (HR 1.13 [95% CI 0.65-1.97]). Favipiravir was well tolerated with few AEs but was associated with increased variant frequency, including C-to-U mutations. Community administration of favipiravir for mild COVID-19 was not associated with clinical benefits or safety concerns but was associated with SARS-CoV-2 mutagenicity.CLINICAL TRIALSThis study is registered with ISRCTN as 31062548 and with EU-CTR as 2020-001904-41.

Keywords: COVID-19; antiviral agents; randomized controlled trial.

Fig 1

CONSORT diagram summarizing flow through the study from pre-screening to screening, randomization, treatment allocation, and follow-up. The ITT population table details the proportion of cases in each study arm with missing OSS scores at indicated time points.

Fig 2

OR in pre-specified sub-group analyses regarding vaccination status, number of comorbidities, age (≤50 vs >50) and SARS-CoV-2 genotype (Alpha vs Delta vs Omicron). The red line refers to an OR = 1.

Fig 3

Kaplan-Meier plots summarizing (A) time-to-symptom resolution (TTSR) and (B) time-to-viral clearance (TTVC). A total of 266 out of 302 (88.1%) participants achieved symptom resolution (OSS ≤ 1) by Day 60; there was no difference in TTSR between favipiravir treated cases and controls. A total of 111 out of 302 (36.8%) cases had ≥1 follow-up PCR test and could be included in the TTVC analysis. There was no difference in TTVC between arms.

Fig 4

Differences in the SARS-CoV-2 variant frequency (>5%) relative to the initial consensus sequence in Day 15 nasopharyngeal swab samples collected in favipiravir-treated cases versus controls. Significant differences are highlighted by *P < 0.05 and **P < 0.01. (A) Increase in total variant numbers in favipiravir-treated cases. (B) Data for each of the 12 mutational channels (A-to-C, A-to-G, A-to-T, etc.) revealing increases in C-to-U, consistent with the known action of favipiravir, in addition to increases in G-to-U and U-to-C.