A nomogram for predicting the risk of male breast cancer for overall survival.

BACKGROUND

Male breast cancer (MBC) is a rare disease, accounting for <1% of all male carcinomas. Lack of prospective data, the current therapy for MBC is based on retrospective analysis or information that is extrapolated from studies of female patients. We constructed a nomogram model for predicting the overall survival (OS) of MBC patients and verify its feasibility using data from China.

METHODS

Constructed a predictive model using 1224 MBC patients from the Surveillance, Epidemiology and End Results (SEER) registry between 2010 and 2015. The performance of the model was externally validated between 2002 to 2021 using 44 MBC patients from the Fujian Medical University Union Hospital. The independent prognostic factors were selected by univariate and multivariate Cox regression analyses. The nomogram was constructed to predict individual survival outcomes for MBC patients. The discriminative power, calibration, and clinical effectiveness of the nomogram were evaluated by the receiver operating characteristic (ROC) curve, and the decision curve analysis (DCA).

RESULTS

A total of 1224 male breast cancer patients were in the training cohort and 44 in the validation cohort. T status (<0.001), age at diagnosis (<0.001), histologic grade (=0.008), M status (<0.001), ER status (=0.001), Her2 status (=0.019), chemotherapy (=0.015) were independently associated with OS. The diagnostic performance of this model was evaluated and validated using ROC curves on the training and validation datasets. In the training cohort, the nomogram-predicted AUC value was 0.786 for 3-year OS and 0.767 for 5-year OS. In the validation cohort, the nomogram-predicted AUC value was 0.893 for 3-year OS and 0.895 for 5-year OS. Decision curve analysis demonstrated that the nomogram was more benefit than the AJCC stage.

CONCLUSIONS

We developed a nomogram that predicts 3-year and 5-year survival in MBC patients. Validation using bootstrap sampling revealed optimal discrimination and calibration, suggesting that the nomogram may have clinical utility. The results remain reproducible in the validation cohort which included Chinese data. The model was superior to the AJCC stage system as shown in the decision curve analysis (DCA).