Abstract 5033: Optimized CYP2D6 phenotype assignment for plasma endoxifen prediction in breast cancer patients treated with tamoxifen

Abstract

Tamoxifen (Tam), a standard therapy for estrogen receptor (ER)-positive breast cancer, is bioactivated to its active metabolite endoxifen by the cytochrome P450 enzyme CYP2D6. CYP2D6 polymorphisms constituting extensive (EM), intermediate (IM) and poor (PM) metabolizer phenotypes are main determinants of variable endoxifen plasma concentrations and supposed to influence therapeutic efficacy. Genotype-guided prediction of TAM metabolizer status has long been debated to be useful for clinical dose adjustment in patients with impaired metabolism. As this requires accurate genotype-phenotype assignments, we tested several CYP2D6 phenotype classifications including the codeine metabolism-based scoring implemented by the ‘Clinical Pharmacogenetics Implementation Consortium’ (Crews et al. 2014, CPT) for their predictive value of active metabolite concentrations. We applied linear modelling of square-root transformed plasma endoxifen or log-transformed metabolic ratio (MR) endoxifen/desmethyl-TAM depending on CYP2D6 phenotype classification in a cohort of 936 pre- and postmenopausal TAM treated breast cancer patients of European, Middle Eastern Arabic, and Asian descent (Muerdter et al CPT 2011; Saladores et al. TPJ 2015). CYP2D6 diplotypes were defined by PM (*3, *4, *5, *6, *7), IM (*9, *10, *41), EM (*1, *2, *35) or UM (duplicated EM) alleles. CYP2D6 phenotypes were defined by various diplotype groupings and an adapted activity score regarding the *10 allele. Models were adjusted for CYP2C9 and CYP3A5 genotypes, and robust coefficients of determination (R²) with 95% confidence intervals from 10.000 bootstrap replicates were calculated using R package lmrob. Our analyses indicate, that the commonly used codeine-based CYP2D6 phenotype assignment poorly predicts CYP2D6 mediated endoxifen formation. This is evident from a low predictability (<20%) for both metabolite endpoints particularly in Asians, most likely due to a misclassification of abundant IM/IM (*10) diplotypes as EM. Importantly, the active metabolite-to-precursor MR appears to be the endpoint that is more closely linked to CYP2D6 enzyme activity, as it explains up to 68-82% of MR variability. In contrast, absolute endoxifen concentrations were only modestly predictable (38-57%) by CYP2D6 diplotype or other phenotype groupings. Moreover, a reduced activity for the *10 compared to other IM alleles appears to improve endoxifen predictabilty by phenotype. Personalized TAM treatment decisions based on plasma endoxifen prediction should rely on TAM specific CYP2D6 genotype-guided phenotype assignments that quantitatively capture the allele-dosage effects and that include endoxifen-to-precursor MR as the endpoint most directly linked to CYP2D6.

Citation Format: Werner Schroth, Stefan Winter, Michel Eichelbaum, Thomas Muerdter, Matthias Schwab, Hiltrud Brauch. Optimized CYP2D6 phenotype assignment for plasma endoxifen prediction in breast cancer patients treated with tamoxifen [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 5033. doi:10.1158/1538-7445.AM2017-5033