ABSTRACT
Background: Metformin has in vitro activity against SARS-CoV-2. In a published phase 3, quadruple-blinded, placebo-controlled randomized trial of outpatient COVID-19 therapy, metformin resulted in a 42% reduction in ER visits/hospitalizations/deaths by day 14, 58% reduction in hospitalizations/ death by day 28, and 42% reduction in Long Covid through 10 months. This analysis presents the results of viral load sampling performed during that clinical trial. Method(s): Covid-Out trial (NCT04510194) enrolled adults aged 30 to 85 within 3 days of a documented SARS-CoV-2 infection and < 7 days after symptom onset. The trial randomized 1323 participants to metformin (1000mg/day days 2-5;1500mg/day days 6 to 14), ivermectin, fluvoxamine, and/or exact-matching placebo in a 2x3 factorial trial design. Nasal swabs for viral load were an optional component, self-collected from the anterior nares on day 1, 5, and 10. Viral loads were measured via RT-qPCR using N1 and N2 targets in the SARSCoV- 2 nucleocapsid protein, with relative Ct values converted to absolute copy number via calibration to droplet digital PCR. A linear Tobit regression model was used to assess change over time while accounting for left censoring due to the viral load limit of detection. Results were adjusted for other treatment allocations within the factorial design, vaccination status, and baseline viral load. Repeated measures were accounted for using clustered standard errors within participants. Result(s): Samples were available from n = 945, 871, and 775 participants on days 1, 5, and 10, respectively. The mean change from baseline to followup was -0.64 log10 copies/mL (95%CI, -1.16 to -0.13) for metformin versus placebo, which equates to a 4.4-fold greater decrease. The mean change in SARS-CoV-2 from baseline to day 5 was -0.48 log10 copies/mL, and was -0.81 log10 copies/mL from baseline to day 10. The anti-viral effect increased with increased metformin dosing days 6-14. The antiviral effect was larger in those unvaccinated (mean -0.95 log copies/mL) than vaccinated (mean -0.39 log copies/mL). There was no change in viral load vs. placebo for ivermectin or fluvoxamine. Conclusion(s): Metformin lowered SARS-CoV-2 viral load in this quadrupleblinded, randomized clinical trial. The temporal relationship to dose titration suggests a dose-dependent effect. The magnitude of antiviral effect was similar to nirmatrelvir at day 5, greater than nirmatrelvir at day 10. Metformin is safe, widely available, and has few contraindications.