Objectives: Carbamazepine (CBZ) is metabolized by the cytochrome P450 into carbamazepine-10,11-epoxide (CBZepox), a metabolite with similar pharmacological activity to the parent drug. Recently it has been indicated that most current immunoassays for the determination of CBZ are unable to quantitatively measure its active epoxide. An evaluation of the Cobas Integra immunoassay for the determination of CBZ was carried out, and the results are compared with those obtained for CBZ+CBZepox using high-performance liquid chromatography (HPLC) in patients on monotherapy and polytherapy.
Design and methods: Steady-state serum trough concentrations of CBZ were determined in 119 epileptic patients using the Cobas Integra immunoassay and HPLC. In 91 cases CBZ was administered in monotherapy and in 28 cases in polytherapy with other anticonvulsant drugs.
Results: The study of within- and between-run imprecisiom for the Cobas Integra immunoassay led to clinically acceptable coefficients of variation. A high correlation was found between the concentrations of CBZ obtained using the immunoassay and HPLC (r = 0.981, p < 0.001). In both the group of patients on monotherapy and those on polytherapy, the levels of CBZepox were greater than the clinically acceptable error for CBZ; consequently, there is a clinically significant difference between the total of CBZ+CBZepox concentrations (HPLC) and the concentrations of CBZ (immunoassay). In the group of patients under monotherapy, a high correlation coefficient was obtained between the levels of CBZ and CBZ+CBZepox (r = 0.975, p < 0.001) with an standard error of the estimate similar to the clinically acceptable value.
Conclusions: For the patients on monotherapy, it is possible to make a clinically valid estimation of CBZ+CBZepox from the concentration of CBZ obtained by means the immunoassay. In patients on polytherapy, the analytical determination of CBZepox could be of interest in cases where CBZ+CBZepox would be higher than the critical level of CBZ.