TL;DR
In a large-scale, quadruple-blinded Phase 3 trial, researchers found that a two-week course of the common diabetes drug metformin reduced the risk of developing Long Covid by 42%. In contrast, ivermectin and fluvoxamine showed no significant effect on preventing the condition over a 10-month follow-up period.
What Problem Does This Paper Address?
While vaccines and antivirals have reduced the severity of acute COVID-19, "Long Covid" (Post-Acute Sequelae of SARS-CoV-2) remains a massive public health challenge, affecting millions with symptoms that can prevent work and daily activities.
There is a critical gap in the clinical toolkit: we lack proven, widely available, and inexpensive outpatient treatments that can be administered during the acute phase of infection to prevent the long-term development of the disease. The authors of the COVID-OUT trial sought to determine if three existing oral medications—metformin, ivermectin, and fluvoxamine—could fill this gap.
Methods
The researchers conducted a decentralized, multi-site, phase 3 randomized controlled trial (NCT04510194). Key features of the methodology include:
- Factorial Design: A 2x3 factorial assignment was used to efficiently test three drugs simultaneously while sharing placebo controls.
- Quadruple-Blinded: Participants, care providers, investigators, and outcome assessors were all blinded to treatment allocation.
- Dosing Regimen:
- Metformin: Titrated over 6 days to 1,500mg/day for 14 days total.
- Ivermectin: 430mcg/kg/day for 3 days.
- Fluvoxamine: 50mg twice daily for 14 days.
- Endpoint: The primary outcome of the original trial was acute severe Covid. This paper focuses on a pre-specified secondary outcome: a medical-provider diagnosis of Long Covid reported by the participant through day 300.
Datasets
The "dataset" for this study consisted of 1,125 adults in the US, aged 30 to 85.
- Inclusion Criteria: Participants were required to have a BMI ≥ 25 kg/m² (overweight or obesity), fewer than 7 days of symptoms, and a documented infection within 3 days of enrollment.
- Demographics: The cohort was 56% female (7% of whom were pregnant) and 82.8% white, with a median BMI of 29.8 kg/m².
- Vaccination: 55% of the cohort had received a primary vaccination series before enrollment.
Key Insights and Ideas
- Metformin’s "Off-Label" Potential: While primarily a glucose-lowering drug, metformin has known anti-inflammatory properties. More interestingly, in silico and in vitro models suggest it can inhibit SARS-CoV-2 protein translation by suppressing the mTor pathway.
- Early Intervention is Key: The trial enrolled patients very early in their infection (within 3 days of a positive test). This suggests that preventing the "inflammatory cascade" or reducing viral replication early on is crucial for preventing long-term sequelae.
- Consistency Across Variants: The benefit of metformin was consistent across different viral variants (Alpha, Delta, and Omicron) and regardless of whether the participant was vaccinated.
- Failure of Other Candidates: Despite significant public interest, neither ivermectin nor fluvoxamine showed any preventative benefit for Long Covid in this high-quality trial setting.
Key Results
- Metformin Success: Only 6.3% of participants in the metformin group received a Long Covid diagnosis, compared to 10.6% in the blinded control group.
- Hazard Ratio: The Hazard Ratio (HR) for metformin was 0.58 (95% CI 0.38 to 0.88), representing a 42% relative reduction in incidence.
- Ivermectin/Fluvoxamine Results:
- Ivermectin HR: 0.99 (no effect).
- Fluvoxamine HR: 1.36 (statistically insignificant, but trended toward higher incidence).
- P-Value: The p-value for metformin was 0.009, which remains significant even after Bonferroni correction for multiple testing.
Limitations and Future Work
- Cohort Specificity: The trial specifically targeted adults with overweight or obesity (BMI ≥ 25) and those over age 30. It is unclear if these results generalize to younger, lower-BMI populations.
- Secondary Outcome: Long Covid was a secondary endpoint, not the primary one. While the results are statistically robust, primary endpoints are generally considered the "gold standard" for clinical proof.
- Diagnosis Method: The study relied on participant-reported medical-provider diagnoses. While this reflects real-world clinical practice, it is subject to the evolving definition of Long Covid.
- Future Directions: The authors suggest future research should investigate optimal dosing, whether extended-release metformin is equally effective, and if there is synergy when metformin is combined with other antivirals like Paxlovid (nirmatrelvir).
Summary updated