Risk factors for fluoropyrimidine-induced cardiotoxicity in colorectal cancer: A retrospective cohort study and establishment of a prediction nomogram for 5-FU induced cardiotoxicity.

Background: Fluoropyrimidine is an important component of systemic chemotherapy for colorectal cancer (CRC). Fluoropyrimidine-induced cardiotoxicity (FIC) may result in delay and discontinuation of chemotherapy and, in severe cases, can even be life-threatening. To date, risk factors for FIC have not been well identified. This cohort study aimed to identify the predictors of FIC in CRC patients and develop a risk prediction nomogram model. Methods: Between January 1, 2018 and December 31, 2020, colorectal cancer patients who received 5-fluoropyrimidine(5-Fu)/capecitabine-based chemotherapy in Affiliated Cancer Hospital of Guizhou Medical University were included. FIC was defined as an adverse cardiovascular event related to fluoropyrimidine that occurred during or within four weeks of completing chemotherapy. Risk factors were determined by LASSO algorithm and multivariate logistic regression analysis. Nomogram for predicting 5-Fu-induced cardiotoxicity was established and internally validated. The concordance index (C-index) and calibration curve were used to evaluate the nomogram's discrimination and accuracy. Results: A total of 916 patients were included for analysis, and 200 [21.8%,95% confidence interval (CI):19.12%-24.47%] experienced FIC. LASSO algorithm and multivariate logistic regression analysis determined that chemotherapy ≤3 cycles (OR=4.694, 95%CI=3.184-6.92), age≥ 60 (OR=1.678, 95%CI=1.143-2.464), BMI>22.97 (OR=1.77, 95%CI=1.202-2.606), and simultaneous use of bevacizumab (OR=2.922, 95%CI=1.835-4.653) were significant risk factors, and were included in the prediction model for 5-Fu induced cardiotoxicity. The C-index (95%CI) was 0.751 (0.706-0.795) by internal validation. For patients treated with capecitabine-based regimen, the incidence of FIC increased with the absolute value of neutrophils (OR=5.177, 95%CI=1.684-15.549) and eosinophils (OR=3.377,95% CI=1.237-9.22). Conclusions: Our study identified risk factors for FIC and established a prediction nomogram model based on chemotherapy cycle, age, BMI and use of target therapy for 5-FU induced Cardiotoxicity. The discriminative prediction model can be used for patient counselling and risk-stratification before undergoing chemotherapy in colorectal cancer.