Reanalysis of carbamazepine and carbamazepine-epoxide pharmacokinetics after multiple dosing of extended release formulation.

Purpose: To model and re-evaluate the pharmacokinetics of carbamazepine (CBZ) and CBZ-10, 11-epoxide (CBZ-E) after 5 day b.i.d. dosing with either Carbatrol (extended-release beads) or Tegretol-XR (an osmotic pump tablet, an Oros tablet) using compartmental method.

Methods: Plasma concentration time profile data from 15 normal healthy adults received, in a randomized crossover fashion, Carbatrol (2 x 200 mg capsules), b.i.d. for 5 days and Tegretol-XR (400 mg), b.i.d. for 5 days were available for analysis from previous study. The compartmental kinetic parameters of CBZ and CBZ-E were simultaneously fitted by assuming: i) one compartment open model with zero order absorption with lag time, and first order elimination for CBZ and ii) one compartment open model with Michaelis-Menten formation with a sigmoidity factor, and first order elimination for CBZ-E. Time to 50% of CBZ plateau concentrations (TC50) was estimated and statistically compared between the two products.

Results: There was a good agreement between simulated and observed plasma concentrations. For CBZ, the fitted parameters were: the first order elimination rate constant (K(10)) 0.024 and 0.022 hr(-1), t(1/2) 27.3 and 30.3 hr, volume of central compartment (V(1) ) 1.119 and 1.160 L/kg, for Carbatrol and Tegretol-XR, respectively. For CBZ-E, the fitted parameters were: the first order elimination rate constant (K (30) ) 0.128 and 0.157 hr (-1), t (1/2) 6.1 and 5.1 hr, volume of central compartment (V (3) ) 0.728 and 0.644 L/kg, V (max) 0.085 and 0.076 mg/hr/kg, K (m) 28.639 and 33.138 mg/mL, for Carbatrol and Tegretol-XR, respectively. The fitted pharmacokinetic parameters of CBZ and CBZ-E were generally consistent with published values from previous studies. A minimal rise in CBZ-E concentrations was observed during the first 12 hours, the finding of which has not been reported before. Consequently, the CBZ-E plasma profiles appear as sigmoid curves and have a different shape compared to those of the CBZ profiles. The inclusion of the sigmoidity factor allowed flexibility in the fitting and optimized the simulation results. When compared to published literature of single dose data, the investigation of CBZ and CBZ-E pharmacokinetics from this study suggested that autoinduction might occur by the fifth day of dosing and might partly contribute to the sigmoidal shape of CBZ-E profiles.

Conclusion: The fitted model well described the plasma profiles of both CBZ and CBZ-E. Carbatrol and Tegretol-XR were similar in their pharmacokinetics based on compartmental analysis.