To develop a prediction model for cardiotoxicity induced by anthracyclines combined with chemotherapy

Abstract Purpose: Anthracyclines are commonly used as curative agents in chemotherapy for patients with breast cancer and lymphoma. Although the chemotherapeutic effect of anthracyclines is good, the cardiotoxicity caused by these drugs cannot be ignored. In this study, we included using anthracycline-based drugs in combination with chemotherapy in patients with breast Cancer and lymphoma, study its after Cancer chemotherapy related heart toxicity (Cancer therapy - related cardiac dysfunction, CTRCD) of the independent risk factors. LASSO regression analysis was used to screen predictive factors to construct a clinical prediction model (nomogram), and the identification accuracy and clinical efficacy of nomogram were verified to provide a theoretical basis for predicting CTRCD in such patients. Methods: 1. Eligible patients were randomly assigned to the training and validation cohorts in a 5:1 ratio according to specified inclusion and exclusion criteria. With CTRCD as the primary endpoint, univariate and multivariate logistic regression analysis was performed on the training cohort to determine the independent prognostic factors of CTRCD. 2. Use LASSO regression to select predictors and develop a nomogram. Then, the calibration Curve, Hosmer-Lemeshow (H-L) goodness-of-fit test, Receiver Operating Characteristic (ROC) curve and Decisive Curve Analysis were used to analyze the nomogram. DCA was used to evaluate the performance of the nomogram. Internal validation was performed to check the stability of the model. Results: 1. A total of 324 patients were included in this study. logistic multivariate regression analysis of the training cohort showed that Body Mass Index (BMI) ≥ 25, previous history of diabetes, Brain Natriuretic Peptide (BNP) ≥ 52pg/ml, pathological stage, low density lipoprotein (LDL) ≥ 3.37mmol/L, and cumulative dose of anthracyclines ≥ 550mg/㎡ were statistically significant (P < 0.005). 2. The C-index of the prediction model was 0.872 (95%CI:0.821–0.923) and 0.886 (95%CI:0.799–0.974), and the p value of H-L test was 0.837 and 0.700 in the training and validation cohorts, respectively. Conclusion: 1. BMI ≥ 25, history of diabetes, BNP ≥ 52pg/ml, LDL ≥ 3.37mmol/L, pathological stage, cumulative dose of anthracyclines ≥ 550mg/㎡ are independent risk factors for CTRCD in breast cancer and lymphoma patients receiving anthracyclines combined chemotherapy.2. The prediction model has good discrimination, and can preliminarily predict the risk of cardiotoxicity in patients with lymphoma and breast cancer using anthracycline-based chemotherapy. This model has certain reference value for the prediction of cardiotoxicity in other chemotherapy regimens using anthracyclines.