Abstract PD05-09: Prospective Assessment of CYP2D6 by Genotyping, Phenotyping and Measurement of Tamoxifen, 4-Hydroxy-Tamoxifen and Endoxifen in Breast Cancer Patients Treated with Tamoxifen

Abstract

Abstract Background: CYP2D6 is genetically highly polymorphic and several studies support that patients classified as poor-or intermediate-metabolizers achieve less or no benefit from tamoxifen treatment probably because they have lower levels of endoxifen. Genotyping is currently the major method used in studies and in clinical practice. However the ability of genotyping to predict plasma levels of endoxifen is uncertain for an individual patient. We assessed prospectively CYP2D6 activity by genotyping, phenotyping and the measurement of tamoxifen and its metabolites. Methods: Patients were genotyped for CYP2D6 (*1, *3, *4, *5, *6 and *XN) polymorphism. The CYP2D6 phenotype was determined by the dextromethorphan test. Plasma was collected at 2 time points after at least 4 months of treatment with tamoxifen 20 mg daily: tamoxifen, 4-hydroxytamoxifen, N-demethyltamoxifen and endoxifen were measured with high performance liquid chromatography coupled to triple stage tandem mass spectrometry. Linear regression analyses were performed on log transformed concentrations of tamoxifen, its metabolites, and the ratios 4- hydroxytamoxifen/tamoxifen, endoxifen/N-demethyltamoxifen versus the different genotype groups (UM, EM, IM, PM) and the dextromethorphan/dextrorphan ratio. Results: The data of 26 patients are currently available. Geometric mean plasma concentrations (coefficient of variation %) of tamoxifen, N-desmethyltamoxifen, 4-OH-tamoxifen and endoxifen were 377 nmol/L (39%), 482 nmol/L (36%), 5.9 nmol/L (52%) and 60.7 nmol/L (94%), respectively. Genetic variation in CYP2D6 was significantly correlated with endoxifen, 4-hydroxytamoxifen and the ratios of 4-hydroxytamoxifen/tamoxifen and endoxifen/N-demethyltamoxifen: determination coefficients (R-squared) of 44% (P=0.0002), 30% (P=0.0038), 57% (P= <0.0001) and 47% (P=0.0001), respectively. Phenotypes defined by the dextromethorphan/dextrorphan ratio were significantly correlated with endoxifen and the ratios of 4-hydroxytamoxifen/tamoxifen and endoxifen/N-demethyltamoxifen: determination coefficients of 47% (P=0.0002), 39% (P=0.0012) and 59% (P= <0.0001), respectively. Conclusions: Our data confirm a significant correlation between CYP2D6 activity defined by genotyping or by phenotyping and plasma levels of endoxifen. However, the important interindividual variability in the concentrations of the metabolites of tamoxifen remains largely unexplained, suggesting that CYP2D6 genotyping and phenotyping are not sufficient to predict plasma levels of the active metabolites of tamoxifen. These preliminary results are consistent with a potential superiority of monitoring the active metabolites themselves rather than genetic or phenotypic surrogates. The study is ongoing and more data will be presented. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr PD05-09.