Why Do Valproic Acid Plasma Concentrations Vary from Person to Person?

This multinational pooled study provides the answer.

Introduction

01.Materials and methods

O

02.Results

O

A total of 191 patients were included in the study, 127 of whom were male (66.5%). The median age of the patients was 48 years, and the median weight was 67 kg. Of these patients, 92 were from Pakistan (median age: 54 years; median weight: 71 kg), and 99 were from South Korea (median age: 44 years; median weight: 60 kg). Of the Pakistani patients, 64 were male (69.6%) and 28 were female (30.4%), with a median single dose of 1,000 mg. Of the South Korean patients, 63 were male (63.6%) and 36 were female (36.4%), with a median single dose of 1,000 mg.

The optimal model was a one-compartment model combined with first-order elimination. In the pooled population, the clearance (CL) of valproic acid was 0.931 L/h, with 43.4% interindividual variability, and the volume of distribution (Vd) was 16.6 L, with 22.3% interindividual variability.

Covariate analysis revealed that race and body weight were significant predictors of CL. When these were included, the OFV decreased by 50.4 units. Body weight also had a significant effect on Vd, reducing the OFV by a further 32.8 units. Specifically, the median CL in Pakistani patients (1.21 L/h, range 0.55–2.96 L/h) was significantly higher than in Korean patients (0.97 L/h, range 0.39–3.38 L/h), indicating a significant difference in valproic acid clearance capacity between the two ethnic groups. However, Vd values were relatively similar between the two groups: 17.6 L (range 10.4–27.5 L) for Pakistani patients and 15.2 L (range 8.36–25.1 L) for Korean patients (see Figure 1). This may be related to the comparability of body weight and age between the two groups.

Figure 1. Box-plots showing CL (a) and Vd (b) of valproic acid among Pakistani and South Korean patients

Model evaluation showed that the observed concentrations were in good agreement with the predictions for both the population and individuals, and that the condition-weighted residuals were randomly distributed around the zero line. More than 95% of the values fell within the range of -2 to 2 (see Figure 2). The results of 1,000 bootstrap analyses were highly consistent with the final model parameters, confirming the model's stability and robustness. 

Figure 2. Goodness-of-fit plots of final model showing observed concentrations vs. individual predictions (a), observed concentrations vs. population predictions (b), CWRES vs. population predictions (c) and CWRES vs. time after dose (d).

03.Discussion and conclusions

O

This study is the first to use a combined therapeutic drug monitoring (TDM) dataset from patients in Pakistan and South Korea to reveal significant inter-ethnic differences in the pharmacokinetics of valproic acid through population pharmacokinetic analysis. The median CL in Pakistani patients (1.21 L/h) was approximately 24% higher than in South Korean patients (0.97 L/h). This difference may be attributed to genetic polymorphisms in hepatic metabolic enzymes that vary across different ethnic groups. Valproic acid is primarily metabolised by enzymes such as CYP2C9 and CYP2C19; genetic polymorphisms in these enzymes exhibit significant ethnic group variation, leading to marked differences in drug clearance.

Body weight has also been confirmed to be an important covariate affecting CL and Vd, consistent with numerous previous studies ^{[5, 6, 7]}. The impact of body weight on CL reflects the relationship between hepatic metabolic capacity and body surface area, whereas its influence on Vd relates to the drug's distribution volume in tissues. While there was a significant difference in CL between the two groups, their Vd values were relatively similar, potentially due to comparable body weight and age.

This study is innovative in that it is the first to conduct a pooled analysis of TDM data from patients in Pakistan and South Korea. This analysis reveals the impact of ethnic differences on the pharmacokinetics of valproic acid. However, the study has certain limitations. Due to the limited availability of covariates in the South Korean patient data, some (such as concomitant medications and liver function markers) could not be included in the analysis.

In summary, the population pharmacokinetic model of valproic acid developed in this study confirms that ethnicity and body weight are important covariates affecting clearance. This provides clinicians with a basis for adjusting dosages according to a patient’s ethnicity and body weight, thereby facilitating more precise, individualized dosing.