Development of a predictive model for neutropenia risk in Japanese breast cancer patients treated with doxorubicin and cyclophosphamide.

Chemotherapy-induced neutropenia (CIN) and febrile neutropenia (FN) are important considerations for anticancer drug dosing. Recent studies suggest that CIN may also improve survival. This study aimed to assess the impact of CIN on survival outcomes in breast cancer patients and to develop a predictive model for CIN incorporating genetic factors. In 166 breast cancer patients treated with doxorubicin and cyclophosphamide (AC therapy), neutropenia was classified into three severity groups using CTCAE v5.0 criteria. The relationship between severe CIN and survival was analysed. Additionally, 33 genetic polymorphisms linked to pharmacokinetics and susceptibility were examined, and a predictive model for CIN was built using multivariate logistic regression integrating clinical and genetic factors. Among patients receiving AC therapy, the 10-year survival rates were 80.6% for the Grades 0-2 group, 96.2% for the Grades 3-4 group without FN, and 79.0% for the Grades 3-4 group with FN, indicating that patients in the Grades 3-4 group without FN had significantly longer survival. A predictive model for Grade 4 CIN was developed using seven factors: age, haemoglobin levels, serum creatinine levels, and the genetic polymorphisms ABCB1 rs1128503, ERCC1 rs11615, ERCC2 rs238406 and XRCC3 rs709399. The model had an AUC of 0.744 (95% CI, 0.666-0.823), and including genetic polymorphisms improved predictive accuracy by 0.128 (P = 0.004). CIN was linked to improved survival in breast cancer patients. The predictive model including genetic factors provided a more accurate assessment of CIN risk, potentially enabling more personalized treatment approaches.