COBRA Abstract, 2003

Consortium on Breast Cancer Pharmacogenomics

Goals

As is the case in the study of genetic pathogenesis of disease, variations in multiple genes may be responsible for important inter-individual variation in response to medicines. Our center was therefore organized to study multiple genetic influences on the multiple effects of one important drug: tamoxifen as a model for other drugs, where multiple genetic variations may alter diverse pharmacologic responses. Tamoxifen was chosen because of its importance in the endocrine treatment of women with breast cancer, because it is a sufficiently mature therapy that we understand that it has multiple effects (i.e. effects on the cardiovascular system, coagulation pathways, uterus and bone as well as on breast cancer) and because pharmacogenetic variants may contribute to both pharmacokinetic and pharmacodynamic variability in response to the drug. We expect that the presence of multiple clinical effects of tamoxifen will allow us to measure multiple response phenotypes in women under tamoxifen treatment, affording a rich, controlled phenotypic dataset in a population of women who are highly genetically characterized for candidate genes in the tamoxifen pharmacokinetic and estrogen effect pathways.

Progress

To enrich and focus the list of candidate genes of importance to the pharmacokinetics of tamoxifen, a comprehensive characterization of the kinetics of tamoxifen metabolism to its primary and secondary metabolites was performed and preliminary clinical trials to investigate the pharmacokinetics of tamoxifen and its metabolites were conducted. While these efforts led us to the identification of key candidate genes to be studied in retrospective and prospective clinical trials, they also resulted in the isolation, identification, bulk synthesis and pharmacodynamic characterization of a novel tamoxifen metabolite. We have isolated, purified, and identified this metabolite as 4-hydroxy-N-desmethyl tamoxifen (previously designated as BX), and named it endoxifen: the end product of the N-demethylation and hydroxylation of tamoxifen. We have developed a bulk synthetic method for endoxifen that allowed us sufficient quantities of the drug to conduct comprehensive efficacy and metabolism studies in vitro. Our data demonstrate that endoxifen is as potent as 4HT as an antagonist of estrogen binding to both the α and β forms of the estrogen receptor (ER_α and ER_β), as an antagonist of estrogen effects on estrogen-dependant gene transcription and of growth in estrogen-dependant breast cancer cell lines, (MCF-7, T47D and BT474 cells). Since many studies of the effects of tamoxifen in vitro have used 4HT, a large body of work remains to be conducted to fully characterize the effects of endoxifen, and our ongoing studies aim toward a comprehensive characterization of endoxifen-mediated gene expression. Our data also demonstrate that endoxifen is present at higher steady-state concentrations than 4-HT in the plasma of women treated with the drug, and that its concentration is significantly diminished in women who carry variant alleles of the CYP2D6 gene, or are prescribed CYP2D6 inhibitors.

To test the effects of patterns of the candidate genes on phenotypic variables, we have begun a prospective trial of the effects of tamoxifen in 200 women with breast cancer and have developed a novel statistical method to iteratively estimate the best predictor of individual phenotypic outcomes, combining both available clinical and genetic predictors. This trial measures a large number of phenotypic variables including hot flashes, breast density, bone mineral density, plasma lipid profiles and coagulation factors as well as the pharmacokinetics of tamoxifen and its metabolites using a novel analytical technique that allows us to assess picogram concentrations of metabolites in plasma. This trial has already demonstrated that a pattern of genetic variants in CYP2D6 and sulfotranferase 1A1 best predict the plasma concentration of endoxifen, and that the size of changes in LDL concentration with tamoxifen can be predicted by ER? genotype. At present, 120 patients have been enrolled at our two clinical centers at the University of Michigan and Indiana University. In addition, data that we have generated on the effect of genotype on the ability of estradiol and tamoxifen to bring about platelet aggregation indicate that women who carry a homozygous variant genotype in exon 7 of eNOS may be more vulnerable to platelet aggregation by estrogens than others. These data have significant practical import for the clinical use of tamoxifen and other anti-estrogens.