Background: A population pharmacokinetic model for phenytoin in Asian pediatric patients was developed to determine the influence of concurrent medications, patient demographics, and blood biochemistry on the pharmacokinetic profile of phenytoin.
Methods: Retrospective clinical data were obtained from 66 patients (age, 1-16 years) for the determination of pharmacokinetic parameters of phenytoin using WinNonmix. Data from 49 patients (74.2%) were allocated in the "index" group, and the other 17 patients (25.8%) in the "validation" group. Models were compared by final log likelihood, mean error as a measure of bias, and root-mean-squared error as a measure of precision.
Results: The Michaelis-Menten constant (km) and volume of distribution (V) were fixed at 9.08 mg/L and 1.23 L/kg, respectively. The saturated elimination rate (V × Vmax) of phenytoin was then found to be 0.525 mg/kg per hour (352.9 4 mg/d for a 28.0 kg individual). Patients' body surface area (in square meter) and catalytic activities of liver enzymes aspartate aminotransferase (U/L) and alkaline phosphatase (U/L) appeared to have significant correlation with Vmax, whereas coadministrating drugs with phenytoin did not yield any significant effect. The final model for the saturated elimination rate was (Equation is included in full-text article.) In validation of the final model, the mean error was found to be -0.805 (95% confidence interval, -3.67 to 2.06), and the root-mean-squared error was 7.92 (95% confidence interval, 3.41-12.43).
Conclusions: The obtained results indicated the need to consider patients' body surface area and the catalytic activities of liver enzymes aspartate aminotransferase and alkaline phosphatase when dosing phenytoin. Based on the population pharmacokinetic parameters, a nomogram was subsequently developed for dose individualization of phenytoin in Asian pediatric patients.