Phenytoin dosage predictions in paediatric patients

Area of Science:

• Pharmacokinetics
• Pediatric Pharmacology
• Drug Metabolism

Background:

• Phenytoin dosing in pediatric patients presents challenges due to growth-related changes and saturable metabolism.
• Accurate dosage adjustments are crucial for therapeutic efficacy and safety in children.

Purpose of the Study:

• To evaluate two pharmacokinetic methods for adjusting phenytoin dosage in pediatric patients based on a single dosing-rate/steady-state concentration pair.
• To compare the predictability, bias, and precision of a Bayesian forecaster versus a fixed Vmax method.

Main Methods:

• Retrospective analysis of 34 pediatric patients with 48 predictions.
• Utilized a Bayesian forecaster and a fixed Vmax method with different a priori parameter estimates.
• Assessed methods based on absolute predictability, bias (mean error), and precision (root mean squared error).

Main Results:

• No significant differences in predictability, bias, or precision were found among the five evaluated methods.
• The Bayesian algorithm demonstrated greater robustness, providing predictions in all cases, unlike fixed Vmax methods.
• All methods showed a tendency for overprediction of dosage, with poorer results for steady-state concentration predictions.

Conclusions:

• Both Bayesian and fixed Vmax methods have limitations in phenytoin dosage adjustment for pediatric patients.
• The Bayesian method offers more consistent predictions across various scenarios.
• Close patient monitoring remains critical regardless of the chosen dosage adjustment method.