Population Modelling of the Absolute Bioavailability and Pharmacokinetics of Phenobarbitone in Infants with Seizures
A. Lanner (1), X. Xiaonian (1), T. Donovan (2), B. Charles (1)
(1) Australian Centre for Paediatric Pharmacokinetics, School of Pharmacy & Mater Health Services, The University of Queensland, Brisbane, Australia; (2) Grantley Stable Neonatal Unit, Royal Women’s Hospital & District Health Services, Brisbane, Australia
Objectives: Seizures occur in 1-2 infants per 1000 term live births and constitute a medical emergency with significant risk of mortality and morbidity, such as impeded neuronal development. Phenobarbitone is the long-standing treatment of first choice, but it has adverse effects including inhibition of brain growth and neuronal toxicity. Little is known of the pharmacokinetics of phenobarbitone in these patients, especially the oral bioavailability. In this study we determined the pharmacokinetics of oral and intravenous phenobarbitone in order to provide guidelines for seizure management in combination with plasma concentration monitoring data.
Methods: A population pharmacokinetic analysis was developed using NONMEM with the G77 compiler and first order conditional estimation. Phenobarbitone concentration data were obtained retrospectively from medical records. Using a one-compartment model, estimates of clearance, volume of distribution, and fraction of absorbed dose were obtained. Interindividual and residual variabilities were estimated using an exponential and additive error model, respectively. The absorption rate constant was fixed to 2 h-1 because of lack of data in the absorptive phase. Several patient characteristics were screened for influence on the typical pharmacokinetic parameter values.
Results: From 1-9 (median 2) plasma samples per infant were obtained from 113 infants of mean (range) age of 13 (1-108) d, gestational age 36.8 (23-42) wk, and current weight of 3.0 (0.59-5.8) kg. The mean plasma concentration was 29.5 (1-93) mg/L. The typical values (interindividual variability) were 12.2 mL h-1 (38%) for clearance and 1.9 L for volume (33.9%) at median weight of 3.27 kg. A linear median-centred weight model on both parameters significantly improved the fit to the data. The systemic oral bioavailability was 0.61 (33.6%) and was unaffected by any patient characteristic. The elimination half-life (based on typical values) was 107.9 h. The residual variability was 20%.
Conclusion: Typical clearance, volume of distribution and half-life were similar to adult values, but the bioavailability was somewhat lower. The half-life was longer than reported in older children. Infants older than 14 d had lower drug levels indicating a possible need for upward dose adjustments. There was considerable interindividual variability in all parameters which justifies continued plasma phenobarbitone concentration monitoring in these patients.