A web-based novel model for predicting prognostic value in patients with invasive micropapillary carcinoma in breast cancer: a real-world data retrospective cohort study.

ABSTRACT

Invasive micropapillary carcinoma (IMPC) accounts for less than 2% of all invasive breast cancers and usually associates with poor survival, so we investigated the prognostic factors for IMPC using a large population-based database and designed a web-based novel model. Clinicopathological prognostic factors were evaluated using the Surveillance, Epidemiology, and End Results (SEER) database. Multivariate Cox regression analysis was performed to evaluate the prognostic value of variables on the overall survival. A web-based nomogram was finally constructed to predict the survival probability. The model was validated in an external dataset. A web-based model, combined with age, radiation, clinical stage, and hormone receptor (HR) immunochemistry status four prognostic factors, was constructed. The C-index (0.714, 95% CI 0.683-0.741), calibration curves, and decision curves showed that this model was superior in prediction. By determining the cut-off values, high-risk group and low-risk group were divided. The Kaplan-Meier survival curves showed that these two groups had significantly different survival rates (P < 0.0001). The result of C-index, calibration curves, and Kaplan-Meier survival curves were consistent in the validation cohort. The novel nomogram with four risk factors resulted in accurate prognostic prediction for IMPC.