ABSTRACT
BACKGROUND: Ion channel inhibition may offer protection against coronavirus disease 2019 (COVID-19). Inflammation and reduced platelet count occur during COVID-19 but precise quantification of risk thresholds is unclear. The Recov ery-SIRIO study aimed to assess clinical effects of amiodarone and verapamil and to relate patient phenotypes to outcomes. METHODS: RECOVERY-SIRIO is a multicenter open-label 1:1:1 investigator-initiated randomized trial with blinded event adjudication. A sample of 804 symptomatic hospitalized nonintensive-care COVID-19 patients, follow-up for 28 days was initially planned. RESULTS: The trial was stopped when a total of 215 patients had been randomized to amiodarone (n = 71), verapamil (n = 72) or standard care alone (n = 72). At 15 days, the hazard ratio (hazard ratio [HR], 95% confidence interval [CI]) for clinical improvement was 0.77 (0.52-1.14) with amiodarone and 0.97 (0.81-1.17) with verapamil as compared to usual care. Clinically relevant associations were found between mortality or lack of clinical improvement and higher peak C-reactive protein (CRP) levels or nadir platelet count at 7, 10 and 15 days. Mortality rate increased by 73% every 5 mg/dL increment in peak CRP (HR 1.73, 95% CI 1.27-2.37) and was two-fold higher for every decrement of 100 units in nadir platelet count (HR 2.19, 95% CI 1.37-3.51). By cluster analysis, thresholds of 5 mg/dL for peak CRP and 187 × 103/mcL for nadir platelet count identified the phenogroup at greatest risk of dying. CONCLUSIONS: In this randomized trial, neither amiodarone nor verapamil were found to significantly accelerate short-term clinical improvement. Peak CRP and nadir platelet counts were associated with increased mortality both in isolation and by cluster analysis.