Influence of coadministered antiepileptic drugs on serum phenobarbital concentrations in epileptic patients: quantitative analysis based on a suitable transforming factor.

This study investigated most suitable transforming factor related to the daily Phenobarbital dose (D) providing a steady-state serum concentration (Ct) and analyzed the influences of concomitant antiepileptic drugs on Ct quantitatively. Data obtained by routine therapeutic drug monitoring from a total of 326 epileptic patients treated with multiple oral administrations of phenobarbital (PB) as a powder, were used for the analysis. A total of 156 patients were administered PB alone, and 92, 57, and 21 patients were coadministered one, two, and three different antiepileptic drugs, respectively. Valproic acid (VPA), carbamazepine (CBZ), phenytoin (PHT), zonisamide (ZNS), clonazepam, and ethosuximide were coadministered with PB. For administration of PB alone, four types of transforming factor corresponding to clearance, i.e., total body weight, total body water volume, body surface area and extracellular water volume (VECW) were proposed. With VECW as a transforming factor, the level/dose (L/D) ratio (:Ct/(D/VECW)) was independent of the patient's age and gender. Ct was dependent on only one variable regarding D/VECW and expressed as Ct=0.989 x (D/VECW). The coadministration of one drug caused a difference in the gradient of the regression line of PB alone, and the influence of each drug was detected by dividing each mean L/D ratio of PB plus one other drug by that of PB alone. VPA, CBZ, and PHT significantly increased (p<0.01) the L/D ratio to 1.48, 1.35, and 1.23 of the value for PB alone, respectively. With coadministration of multiple drugs, the L/D ratio rose significantly (p<0.05) as the number (< or =2) of drugs coadministered increased regardless of the type, and also increased with the concomitant use of 3 drugs compared with 2 drugs. For a more detailed analysis, we defined the parameter eta(i) (i=1, 2, ..., 6) and an alteration ratio Ri, representing the influence of each antiepileptic drug on the L/D ratio of PB alone. A model based on the assumption that each coadministered drug competitively inhibited PB-metabolizing enzyme, was adopted. The analysis clarified that VPA, CBZ, and PHT significantly increased (p<0.05) the L/D ratio of PB to 1.466, 1.177, and 1.186, respectively. In the case of the addition or discontinuance of concomitant treatment with antiepileptic drugs in the same patient, the estimated L/D ratios were calculated using the value of each Ri and compared with the measured value. The mean of prediction error was calculated as 23.1%. Our results appear valid and Ri should be available for clinical use.