A predictive model for survival outcome of patients with early-stage medullary breast carcinoma: A population-based study.

e13178 Background: Medullary breast carcinoma (MedBC) is a rare subtype of breast cancer, with a great number of MedBC patients diagnosed with triple-negative breast cancer that is featured by high invasiveness and poor prognosis. The study aims to establish a predictive model for overall survival (OS) in patients with early-stage MedBC. Methods: 606,355 patients with breast cancer were searched from the Surveillance, Epidemiology and End Results (SEER) database. 244 patients with MedBC were collected from the National Cancer Center in a real-world setting. Univariate and multivariate Cox analyses were conducted to determine prognosis-related variables. Survival time and model performance were evaluated. Results: A total of 650, 188 eligible patients were collected from the SEER cohorts and the real-world cohort. A predictive model was established for OS in the training cohort, incorporating age, T stage, N stage, and chemotherapy. The overall C-index was 0.76, the AUC was 0.83, 0.76, 0.75 and 0.76 for 1-, 3-, 5- and 10-year OS, respectively. Patients in the low-risk group achieved a significantly better OS than those in the high-risk group (p = 2.8e-10). In the internal validation cohort, the overall C-index was 0.86, the AUC was 0.78 and 0.83 for 3 and 5-year OS. In the external validation cohort, the overall C-index was 0.90, the AUC was 0.85, 0.85, 0.84 for 3-, 5- and 10-year OS. Patients in the low-risk group gained a significantly improved OS, as compared to those in the high-risk group in both validation cohorts. Conclusions: A clinical model, involving age (65 years), T stage (T3/T4), N stage (N2/N3), and chemotherapy (No) was successfully developed for inferior OS in patients with early-stage MedBC, which may provide useful references for survival assessment of the targeted patient population.