Objective: The aim of our work was to define the kinetic profile of carbamazepine (CBZ), in order to improve on dosing schedules through a Bayesian approach.
Method: Carbamazepine dose/steady-state trough concentrations data pairs and associated information were collected retrospectively on a population of adult epileptic patients.
Results: Fifty patients (index population) with two or more available concentrations (total of 174 determinations) met our inclusion criteria. Patients were taking CBZ (200-1800 mg/day) in mono- or polytherapy regimens. The analysis assumed a one-compartmental model with first-order absorption and elimination. Due to the data source (only trough concentrations were measured as part of hospital routine), the volume of distribution was fixed at 1.19 l/kg. The final estimates for CL were: 0.075 +/- 0.027 (mono- and polytherapy), 0.069 +/- 0.020 (monotherapy), and 0.106 +/- 0.037 l/h/kg (polytherapy). In order to validate these results, we assessed their predictive capacity using 18 new patients (validation population), submitted to the same inclusion criteria and using Prediction-Error analysis. The results suggested a different CL value for our population compared to earlier published clearance values. The results also pointed to an increased metabolic rate associated with polytherapy. The prediction capacity of the optimization method derived from a Portuguese population made in an a priori evaluation indicated a low error (-0.04 microg/ml), close to the theoretical zero value.
Conclusion: Our results provide specific data on CBZ disposition in a Portuguese population and given the wide variability in the literature values, our data may help improve dosing of CBZ in Portuguese patients.