Gefitinib-phenytoin interaction is not correlated with the 14C-erythromycin breath test in healthy male volunteers

center dot The response to gefitinib is variable and could be explained partly by the interindividual variability in gefitinib exposure. center dot Gefitinib is mainly metabolized by CYP3A4 and CYP2D6, and is to a lesser extent a P-glycoprotein (P-gp) substrate. center dot Patients with cancer are at high risk of drug-drug interactions, such as with phenytoin, a potent CYP3A4 inducer. WHAT THIS STUDY ADDS center dot The reduced systemic exposure of gefitinib with multiple-dose phenytoin treatment. center dot An effect attributed to the possible induction of intestinal CYP3A4 because of the lack of correlation between changes in gefitinib disposition and the erythromycin breath test and the lack of association between allelic variant in the ABCB1 gene and baseline and induced oral gefitinib clearance. center dot The CYP2D6 extensive metabolizer seems to be less sensitive to the interaction with phenytoin, as the magnitude of induction of gefitinib clearance was greater in CYP2D6 poor metabolizers than in CYP2D6 extensive metabolizers. AIMS We aimed to describe the pharmacokinetic interaction between phenytoin, a potent CYP3A4 and P-glycoprotein (P-gp) (ABCB1) inducer, and gefitinib, a CYP3A4, CYP2D6 and P-gp substrate. METHODS An open-label, randomized, two-phase crossover study was conducted. Eighteen healthy male volunteers (nine homozygous CC and nine homozygous TT as determined by their ABCB1 C3435T polymorphism in exon 26) received a single oral dose of 250 mg gefitinib alone or after 5 days treatment with phenytoin (5 mg kg(-1) daily). Gefitinib plasma concentrations were determined by high-performance liquid chromatography. Hepatic CYP3A4 activity was evaluated by the C-14-erythromycin breath test (ERMBT) and the ABCB1 and CYP2D6 genetic polymorphisms were determined by the TaqMan allelic discrimination assay and long polymerase chain reaction, respectively. RESULTS Following treatment with phenytoin, mean gefitinib C-max and AUC(0-infinity) decreased by 26 +/- 44% [95% confidence interval (CI) for the difference 5-48%, P = 0.005] and 47 +/- 26% (95% CI for the difference 34-60%, P = 0.001), respectively, and apparent oral clearance increased by 126 +/- 93% (95% CI for the difference 80-172%, P = 0.004). Concomitantly, phenytoin increased the mean ERMBT by 91 +/- 44% (95% CI 75-105%, P < 0.001) from baseline, but the extent of liver CYP3A4 induction was not correlated to the extent of interaction. Furthermore, this interaction was independent of ABCB1 genetic polymorphism. The CYP2D6 genotype was slightly but significantly related to gefitinib clearance (P = 0.04) during the control phase. CONCLUSIONS The significant interaction between gefitinib and phenytoin was not correlated with the erythromycin breath test and was independent of ABCB1 polymorphism, but may involve presystemic CYP3A-mediated intestinal first-pass.