Effects of CYP2D6 and CYP3A5 polymorphisms on tamoxifen and its metabolites in Thai breast cancer patients

CYP2D6 polymorphisms appear to affect END concentration through an NDMT subpathway and potentially affect 4OHT concentrations through a 4OHT subpathway in CYP3A5-PM group.

One hundred and thirty-four Thai breast cancer patients from the Thai Tamoxifen Project undergoing TAM treatment who met the inclusion/exclusion criteria were recruited. Plasma samples were assessed for the concentrations of TAM and its metabolites using high-performance liquid chromatography. The data are presented as actual values and metabolic ratios (MR). The hypotheses were tested using Kruskal-Wallis or Mann-Whitney U test, including the simple main effects analysis.

This study aimed to determine the effects of CYP2D6 and CYP3A5 polymorphisms on the levels of tamoxifen (TAM) and its metabolites in the plasma of breast cancer patients. The protocol was designed to test the associations between CYP2D6, CYP3A5 genotypes and phenotypes (extensive metabolizer [EM], intermediate metabolizer [IM] and poor metabolizer [PM]) and TAM, N-desmethyl tamoxifen (NDMT), endoxifen (END) and 4-hydroxytamoxifen (4OHT) concentrations. Patients and

The patients had stage 0-IV breast cancer. The mean age and body mass index were 51.6±11.6 years and 24.0±4.3, respectively. Also, 53.0% of them were premenopausal, 10.4% were perimenopausal and 36.6% were postmenopausal, while 23.1% were CYP2D6-EM/CYP3A5-EM and 20.9% carried only CYP2D6 and CYP3A5 incomplete alleles. The median concentrations of TAM, NDMT, END and 4OHT were 374.7 (interquartile range [IQR] 230.2) ng/mL, 1,064.9 (IQR 599.6) ng/mL, 54.5 (IQR 52.5) ng/mL and 5.0 (IQR 3.1) ng/mL, respectively. MR (TAM-NDMT) and MR (NDMT-END) were statistically different (p=0.013 and p=0.014, respectively), while MR (4OHT-END) was not statistically different within the CYP2D6 phenotype (p=0.594). MR (TAM-4OHT) was not statistically different within the CYP2D6 phenotype (p=0.079), but it was potentially different from CYP3A5-PM (p=0.056). None of the MR was statistically different within the CYP3A5 phenotype.