Abstract C020: Predicting cerebrovascular diseases death in Black cancer patients using ensemble machine learning models

Abstract Objective: Cerebrovascular disease (CVD) is a significant yet underrecognized cause of mortality in cancer patients, particularly among black patients, a population disproportionately affected by both cardiovascular and oncologic disparities. Despite the increasing awareness of CVD as a competing cause of death in cancer patients, current literature have yet to demonstrate the risk factors associated with CVD as a cause of death in black cancer patients. This study aimed to use advanced machine learning algorithms to predict CVD mortality and identify linear and non-linear associations between clinical features and CVD death in this vulnrable and underserved population. Methods: Data from The Surveillance, Epidemiology, and End Results 17 registry was utilized to predict CVD as a cause of death in Black cancer patients from January 1, 2000, to December 31, 2020. To enhance the predictive accuracy and capture of complex relationships between CVD death and clinical variables, we employed an ensemble machine learning approach. Multiple machine learning models were utilized, including Random Forest, Logistic regression, XGBoost, LightGBM, and Gradient Boosting. These models were furtherly trained, tested, and integrated into a voting ensemble model, levaraging their collective predictive strengths. The dataset included features such as age, sex, cancer type, year of diagnosis, and treatment modalities including adjuvant chemotherapy, adjuvant radiotherapy, adjuvant chemoradiotherapy, surgery alone, primary chemotherapy, primary radiotherapy, primary chemoradiotherapy, and no treatment. Data preprocessing involved cleaning, feature selection, and 80-20 stratified train-test splitting. The performance of the model was evaluated using metrics including accuracy, AUC (macro, micro, weighted), precision, recall, and F1 scores. Feature importance was analyzed to identify the most contributing variables in predicting CVD. Results: A total of 292,680 black cancer patients were included, with a mean age of 57.36 ± 13.68 years. The cohort comprised 152,991 males (52.27%) and 139,689 females (47.73%). Among them, 5,874 patients (2.01%) died from cerebrovascular diseases, while 286,806 patients are alive. The voting ensemble model achieved an accuracy of 0.98, with an AUC macro of 0.88, AUC micro of 0.99, and AUC weighted of 0.88. The average precision scores were 0.64 (macro), 0.99 (micro), and 0.98 (weighted). The F1 scores were 0.64 (macro), 0.98 (micro), and 0.97 (weighted). The precision scores were 0.78 (macro), 0.98 (micro), and 0.97 (weighted). Significant predictors of CVD included an earlier year of diagnosis (with patients diagnosed earlier in the study period from 2000 to 2020 having a higher risk of CVD), older age, lack of any cancer treatment, and specific cancer types (notably gastrointestinal and male genital cancers). Conclusions: The findings suggest that, beyond the predictive capabilities of our model, the model provides valuable insights into key clinical and demographic factors contributing to CVD-related death in black patients. Citation Format: Yaman B. Ahmed, Sung-Min Cho. Predicting cerebrovascular diseases death in Black cancer patients using ensemble machine learning models [abstract]. In: Proceedings of the 18th AACR Conference on the Science of Cancer Health Disparities; 2025 Sep 18-21; Baltimore, MD. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2025;34(9 Suppl):Abstract nr C020.