Predictive modeling of adverse events of tamoxifen therapy for breast cancer (results of a cohort study)

Abstract

Relevance. Endocrine therapy is the standard treatment for women with ER-positive breast cancer. The clinical response to Tamoxifen is variable. Approximately 30 % of patients with breast cancer will have a recurrence of the disease within 15 years after treatment, despite ongoing endocrine therapy. This article presents the results of a prospective pharmacogenetic cohort study. The study was conducted in 2018–2019. Aim. To analyze adverse drug reactions to Tamoxifen in the adjuvant regimen in breast cancer patients in relation to the carriage of genetic polymorphisms of genes encoding cytochrome P450 enzymes and drug transporter proteins and to build predictive models based on them. A comparative analysis of the relationship between genetic and non-genetic determinants with adverse events on tamoxifen therapy allowed us to build predictive models of their development. Materials and Methods. The study involved 120 women with pre- and postmenopausal breast cancer who underwent genetic testing for CYP and Pg enzyme gene polymorphisms. Entry criteria: a histologically confirmed diagnosis of breast cancer, taking Tamoxifen at the recommended doses, establishing a diagnosis not earlier than 2007, and obtaining informed voluntary consent to participate in the study. Allelic variants were determined using real-time polymerase chain reaction in the Research Institute for Molecular and Personalized Medicine of the Russian Medical Academy of Continuous Professional Education of the Ministry of Healthcare of the Russian Federation. Results. An associative analysis showed their association with the development of adverse drug reactions (ADR) to Tamoxifen, indicating the clinical significance of different genetic polymorphisms of CYP2D6, CYP3A5, CYP2C9 and ABCB1. The complex associative analysis performed using mathematical modeling made it possible to build predictive risk models for the development of such ADR, such as hot flashes, dyspepsia, bone pain, and asthenia. The resulting regression models were statistically significant (p < 0,001) and demonstrated high diagnostic efficiency. This allows them to be implemented in clinical practice. Conclusion. Thus, models that include both genetic and non-genetic determinants of response may further improve the prediction of individual response to tamoxifen