Background and objectives: Exposure to gefitinib (IRESSA, ZD1839), an epidermal growth factor receptor-tyrosine kinase inhibitor, is highly variable between subjects. In an attempt to explain this variability, three pharmacokinetic studies were carried out in healthy volunteers to investigate the relationship between exposure to gefitinib and cytochrome P450 (CYP) 3A phenotype (study 1), CYP3A5 genotype (study 2) and CYP2D6 genotype (study 3).
Methods: In study 1 all 15 healthy volunteers received single oral doses of midazolam (7.5 mg), as a CYP3A probe, and gefitinib (500 mg), separated by an appropriate washout period. Plasma concentrations of midazolam and gefitinib were measured. In study 2, 73 healthy volunteers with previously defined single-dose gefitinib pharmacokinetic profiles were genotyped for CYP3A5. In study 3 a single oral dose of gefitinib (250 mg) was administered to poor and extensive CYP2D6 metabolisers (n = 15 in each group). Plasma concentrations of gefitinib and its major metabolite, M523595, were measured. Plasma concentrations of gefitinib, M523595 and midazolam were measured using high-performance liquid chromatography with tandem mass spectrometric detection, and appropriate pharmacokinetic parameters were determined by non-compartmental methods. Genetic analysis of CYP3A5 (study 2) and CYP2D6 (study 3) alleles was carried out using standard methodology.
Results: In study 1 there was some indication of a correlation between the area under the plasma concentration-time curve from time zero to infinity (AUCinfinity) values of midazolam and gefitinib, although this did not reach statistical significance (p = 0.062, regression analysis). In study 2 eight of 73 volunteers (11%) were identified as CYP3A5 expressers. No apparent relationship was observed between the occurrence of the CYP3A5 expresser genotype and gefitinib plasma clearance or terminal elimination halflife. In study 3 M523595 was not detected in any plasma samples collected from poor CYP2D6 metabolisers. Gefitinib geometric mean AUCinfinity and peak plasma drug concentration were higher in poor CYP2D6 metabolisers compared with extensive metabolisers (AUCinfinity 3060 vs 1430 ng . h/mL, p < 0.05, ANOVA), although the range of values was wide with considerable overlap between the groups. Gefitinib was well tolerated in both groups.
Conclusions: Individual differences in CYP3A expression do not explain all the interindividual variability in gefitinib exposure. There is no apparent relationship between CYP3A5 genotype and gefitinib clearance. The lack of measurable levels of M523595 in poor CYP2D6 metabolisers confirms that production of this metabolite is mediated by CYP2D6. Although higher exposure to gefitinib occurs in individuals who are poor CYP2D6 metabolisers, genotyping prior to initiation of therapy and dosage adjustment are not warranted.