PO56 | EXTERNAL VALIDATION OF THE SAVED SCORE IN A CANADIAN REGIONAL CANCER CENTER COHORT: A REAL-WORLD ANALYSIS

A. Naassan, R. Fatima, M. Naassan, C. Chisholm | Lakeridge Health, R.S. McLaughlin Durham Regional Cancer Centre, Department of Oncology, Oshawa, ON, Canada

Introduction. Venous thromboembolism (VTE) remains a significant complication in multiple myeloma patients receiving immunomodulatory drugs, with incidence rates of 3-26%. The SAVED score was developed to stratify VTE risk using readily available clinical variables. While initial validation demonstrated strong discriminatory ability in large academic centers, real-world performance in community cancer centers has not been characterized. This study aims to validate the SAVED score in a Canadian Regional Cancer Centre cohort.

Methods. We reviewed 182 multiple myeloma patients who began lenalidomide treatment between January 2013 and February 2025 at the Durham Regional Cancer Centre, with follow-up through January 2026. The SAVED score was calculated using age >80 years (+1), Asian ethnicity (-3), prior VTE (+3), surgery within 90 days (+2), and dexamethasone 120-160 mg monthly (+1) or >160 mg monthly (+2). Patients were classified as low risk (SAVED 0-1) or high risk (≥2). The main result was VTE during lenalidomide therapy, which was confirmed by imaging. Statistical analysis included the Fisher exact test, Kaplan-Meier survival analysis with the log-rank test, and ROC curve analysis. Sensitivity analyses excluded patients with prior VTE and tested alternative cutoffs (≥3).

Results. The median age was 79 years. Overall VTE rate was 12.6% (23/182). Low-risk patients (n=139, 76.4%) had a 5.8% VTE rate vs. 34.9% in high-risk patients (n=43, 23.6%). The odds ratio was 8.77 (95% CI 3.39-22.69, p<0.001). The SAVED score showed good discrimination, with an AUC of 0.735 (95% CI 0.611–0.849). Kaplan-Meier analysis showed clear separation (log-rank p<0.001), with median VTE-free survival of 943 days (2.6 years) in high-risk vs 2,703 days (7.4 years) in low-risk patients. Sensitivity analyses confirmed robustness: excluding prior VTE (n=162) yielded OR 8.73 (p<0.001); an alternative cutoff of ≥3 showed OR 6.88 (p<0.001).

Conclusions. The SAVED score successfully stratified VTE risk in our community cancer center cohort, with high-risk patients demonstrating a 9-fold increased odds of VTE and significantly shorter VTE-free survival. These results support SAVED score-guided thromboprophylaxis decisions in community practice settings where most myeloma patients receive care. Future work should focus on evaluating the impact of SAVED score-guided thromboprophylaxis on VTE reduction and the cost-effectiveness of risk-stratified approaches to thromboprophylaxis in the community cancer center setting.