A simulation study to evaluate treatment effect of tamoxifen by CYP2D6 genotypes compared with AIs

Aims: Tamoxifen is a prodrug, and active metabolites including endoxifen are known to exert its therapeutic effect of tamoxifen. Tamoxifen is converted to endoxifen mainly by CYP2D6 enzymes. The CYP2D6 enzyme activity is known to be associated with CYP2D6 genotypes. Although the role of CYP2D6 genotyping in the prediction of patient outcomes of tamoxifen therapy remains to be controversial, some studies reported the association between CYP2D6 genotypes and therapeutic effect of tamoxifen in patients with breast cancer. 3rd generation aromatase inhibitors (AI) such as letrozole and anastrozole has been replacing tamoxifen since mid-2000s because some studies had showed superior results of AI over tamoxifen (disease-free survival (DFS) HRAI/TAM=0.86 in ATAC trial, HRAI/TAM=0.81 in BIG 1-98 trial). However, studies showing therapeutic superiority of AI over tamoxifen, such as ATAC and BIG 1-98 trial, did not take loss of heterozygosity (LOH) into account, which occurs when genotyping using tumor tissue. Therefore, the objective of this study was to evaluate treatment effect of tamoxifen by CYP2D6 genotypes considering to LOH in comparison with aromatase inhibitors.

Methods: Using the reported result of ATAC and BIG1-98 trials (HRAI/TAM=0.86, 0.81 for DFS), HRPM,IM/EM calculated from the meta-analysis, and the CYP2D6 genotype frequencies in Caucasian populations, which were intended to represent the all the possible scenarios of CYP2D6 genotype frequencies in the subjects enrolled in ATAC and BIG1-98 trials, hazard ratio between AI and tamoxifen in patients with CYP2D6 genotypes associated with EM (HRAI/TAM EM), and that in patients with CYP2D6 genotypes associated with IM or PM (HRAI/TAM IM/PM) were estimated separately in each replicate of the simulation. From the simulation study conducted using NONMEM (version 7.3) and R (version 3.12), median and 95% prediction interval (PI) of HRAI/TAM EM and HRAI/TAM IM/PM were estimated and compared between CYP2D6 genotypes.

Results: The median HRAI/TAM IM/PM (95% PI) was 0.43 (0.23-0.79) in the simulation for ATAC trial and was 0.40 (0.22-0.73) for BIG 1-98 trial. However, the median HRAI/TAM EM (95% PI) was 0.97 (0.84-1.11) in the simulation for ATAC trial and was 0.91 (0.77-1.08) for BIG 1-98 trial. These results suggest that the treatment effect of tamoxifen for postoperative breast cancer patients carrying CYP2D6 genotypes representing extensive metabolizer (EM) is comparable to letrozole or anastrozole.

Conclusion: In this simulation study, the treatment effect of tamoxifen in patients with CYP2D6 genotypes representing EM was comparable to the 3rd generation AI of letrozole and anastrozole. Based on these results, tamoxifen could be a good therapeutic alternative to patients experiencing severe side effects from the 3rd generation AI.