Effect of megadose vitamin C in patients with sepsis: the CEMVIS randomized clinical trial

Previous randomized trials of vitamin C in sepsis have reported conflicting findings, but whether the effect of vitamin C varies according to septic shock status remains unclear. We aimed to evaluate the effect of megadose vitamin C in sepsis and explore whether treatment effect differed according to septic shock status. CEMVIS was a multicenter, randomized, single-blind, placebo-controlled trial conducted in four tertiary hospitals in China. Adults meeting Sepsis-3 criteria were randomly assigned (1:1) to receive intravenous vitamin C (12 g every 12 h for 4 days) or placebo using stratified randomization according to septic shock status. Patients were analyzed according to the modified intention-to-treat principle. The primary outcome was 28-day mortality. Between Feb 8, 2022, and Dec 25, 2024, 114 patients in the vitamin C group and 117 in the placebo group were included in the modified intention-to-treat population. 28-day mortality did not differ significantly between groups (35/114 [30.7%] vs 36/116 [31.0%]; odds ratio 0.99, 95% CI 0.56–1.74; p = 0.97). No significant differences were observed in the duration of mechanical ventilation, vasopressor therapy, and renal replacement therapy, or in changes in SOFA scores over 96 hours. Prespecified subgroup analyses showed significant interaction according to septic shock status (P for interaction = 0.026), with the treatment effects on 28-day mortality numerically favoring vitamin C among patients without septic shock (15.4% vs 34%) and numerically favoring placebo among those with septic shock (38.7% vs 29%). Hypernatremia occurred more frequently in the vitamin C group (40/114 [35.1%] vs 24/117 [20.5%]; p = 0.02). Exploratory restricted cubic spline analyses showed that increasing lactate levels were associated with a progressive shift in treatment effect from potential benefit toward potential harm. Post hoc Bayesian analyses further provided probabilistic characterization of the potential treatment effect according to septic shock status. The results of this trial do not support the routine use of megadose vitamin C to reduce 28-day mortality in patients with sepsis. Treatment effects may vary according to septic shock status. This trial was registered with ClinicalTrials.gov on 3 January 2022 (identifier: NCT05194189).