I. Bondareva, K. Bondareva
The Research Institute of Physical - Chemical Medicine, Moscow, Russia
Objectives: Phenytoin (PHN) is widely used in the treatment of epilepsy for a long time. The pharmacokinetic (PK) characteristics of PHN increase the risk for toxicity: saturable concentration-dependent metabolism, relatively narrow therapeutic index, wide interindividual PK variability, clinically significant drug-drug interactions. The objective of the study is to develop a nonlinear model of PHN pharmacokinetics and to estimate its parameters from TDM data of adult epileptic patients on chronic PHN – monotherapy.
Methods: PHN monitoring data were routinely collected in the Laboratory of Pharmacokinetics of Moscow Medical University. PHN concentrations were measured by high performance liquid chromatography. The assay error pattern was used as: SD=0.1+0.011C+0.003C*C (where SD is standard deviation of the assay at measured PHN concentration C). The population PK analysis was performed using the NPEM program (USC*PACK software) based on a one-compartment model with the first-order absorption and Michaelis-Menten elimination kinetics. This study included 42 patients (28.7±10.7 years) for whom at least two pairs of measured serum levels (peak – trough strategy) related to different PHN dosages were available (182 PHN serum levels totally, PHN dose 274.4±141.7 mg/d).
Results: Assuming 100% bioavailability of orally-administered PHN, estimated median population PK parameter values for the rate constant of absorption (Kabs = 1.72 1/h, CV = 65.8%), the apparent volume of distribution (Vd = 0.6 L/kg, CV = 54.2%), the maximum rate of metabolism (Vmax = 0.35 mg/kg/h, CV = 36.8%) and the Michaelis-Menten constant (Km = 7.5 mg/L, CV = 40.2%) are in good agreement with those reported in the literature.
Conclusions: The study demonstrated wide interindividual variability in PHN pharmacokinetics and the need for TDM and individualizing of PHN dosage regimens. All feedback methods improve the predictability of steady-state (SS) PHN serum concentration (Css) in comparison with predictions based on the population parameter values. However, a reliable Css value for PHN often can be obtained only after about 2 – 3 weeks on an unchanged dosage regimen. Therefore, Bayesian approach for PHN concentration prediction based on minimum sampling SS or non-SS TDM measurements appear to be preferable. Bayesian feedback adaptive control and the proposed population model can improve PHN dosage adjustment and can identify how close a patient is to the more saturated part of the PK curve.
Reference: PAGE 21 () Abstr 2456 [www.page-meeting.org/?abstract=2456]
Poster: CNS