Modeling and Simulation to Predict the Pharmacokinetics of Intravenous Cefazolin in Children – A Strategy to Reduce the Number of Pediatric Subjects in Pharmacokinetic Studies

Purpose: In contrast to pharmacokinetic (PK) studies in healthy adults, PK studies in the pediatric population typically involve patients. For ethical reasons, the number of pediatric patients studied should be limited.  Modeling and simulation techniques are useful in the design of pediatric PK studies.

Methods: To predict PK exposure in children undergoing antibiotic perioperative prophylaxis with cefazolin, an allometric-based population PK model for cefazolin was created using NONMEM based upon data from 24 adults in a Phase 1 study. For the simulations, two cohorts of 3,000 children each (10-12 years and 13-17 years) were generated; each child was randomly assigned a representative weight based on their age using CDC growth charts. Using the allometric-based population PK model, cefazolin concentration-time profiles were generated and PK exposures were estimated for each simulated child. Power analyses were conducted using simulation results, consistent with a recent FDA publication (Wang Y, et al. J Clin Pharmacol. Epub 2011 Dec 12).

Results: A two-compartment model with volumes and clearances scaled allometrically best described plasma cefazolin PK data with <15% interindividual variability in PK. Model-based simulations indicated that single dose regimens of 1g for children and adolescents ≤50kg and 2g for those >50kg would provide cefazolin exposures within the range observed in adults receiving a single dose of 2g. Thirty minute infusions may be warranted to obtain peak cefazolin concentrations observed in adults. A sample size of seven subjects should provide sufficient power to determine cefazolin PK parameters if a clinical study should be requested in children aged 10-12 years.

Conclusion: Modeling and simulation is integral to the design of pediatric PK studies and in this exercise was used to identify appropriate cefazolin doses for perioperative prophylaxis in children and to estimate the required sample size for a potential study in children aged 10-12 years.