Population Pharmacokinetics of Intravenous Vitamin C in Adults with Sepsis: A Sub-study of the LOVIT Trial.

Abstract

Background and objectives: High-dose intravenous vitamin C was associated with increased harm in adults with sepsis requiring vasopressors in the large international LOVIT trial. This population pharmacokinetic sub-study of LOVIT aimed to characterise the pharmacokinetics of vitamin C and explore associations with patient characteristics and outcomes.

Methods: Patients who enrolled in the LOVIT trial were randomized to receive 50 mg/kg vitamin C every 6 hours for 96 hours or placebo. Blood samples were collected on study Days 1, 3, and 7 for plasma vitamin C quantification. Population pharmacokinetic modeling was performed, with relevant patient and treatment-related factors tested as covariates. Internal validation was conducted using bootstrap resampling (n = 1000). Univariable and multivariable analyses explored associations between 28-day mortality and patient characteristics, as well as the relationship between vitamin C exposure and severity of illness.

Results: A total of 464 samples from 161 patients who were randomized to vitamin C were analysed. The median age was 64 (interquartile range [IQR] 56-72) years, weight 82 (IQR, 70-94) kg, and APACHE II score 24 (IQR, 18-29); 65.2% were male. A one-compartment model best described the data, with clearance (CL) significantly influenced by CKD-EPI eGFR and APACHE II score. Population mean estimates for clearance and volume of distribution were 3.37 L/h and 45.06 L, respectively. The model was internally validated and demonstrated to be robust and stable. Although greater vitamin C exposure was observed among non-survivors versus survivors (median 1900.5 [IQR 1191.8-2548.2] vs 1017.8 [IQR 630-1836.7] mg·h/L; p < 0.001), only age (odds ratio [OR] 1.063; 95% confidence interval [CI] 1.024-1.104, p = 0.001) and APACHE II score (OR 1.079; 95% CI 1.005-1.158, p = 0.036) were independently associated with 28-day mortality. Severity of illness was also predictive of vitamin C exposure.

Conclusions: Vitamin C pharmacokinetics are highly variable between individuals. Some of this variability can be explained by changes in glomerular filtration and APACHE II score as a measure of illness severity. Although non-survivors had greater vitamin C exposure than survivors, exposure was not independently associated with mortality.