Analysis of prognostic factors for Merkel cell carcinoma based on SEER database

Objective To identify prognosis factors and construct a nomogram for predicting cancer-specific survival in patients with resected Merkel cell carcinoma. Methods A total of 1 271 patients were extracted from the Surveillance, Epidemiology and End Results database and randomly divided into training cohort(n=891) and validation cohort(n=380). Univariable and multivariable cox analysis were used to determine prognostic factors, and these factors were used to construct a nomogram for predicting cancer-specific survival in patients with resected Merkel cell carcinoma. Model performance was evaluated by Harrell's concordance index(C-index), the area under the time-dependent receiver operating characteristic curve(AUC) and calibration plots. The predictive power of the model was compared with the traditional TNM staging system. Results Cox analysis indicated age, sex, tumor size, N stage, M stage were associated with cancer-specific survival, while race, marital status and radiation therapy did not show a significant effect on outcomes. The above positive factors were employed to build a nomogram and the nomogram was superior to the 8 th TNM staging system. The C-index of the new model was 0.761, while that of the TNM staging system was 0.711. The AUC of the new model was significantly higher than that of the TNM staging system in both cohorts. Also, the nomogram displayed a good calibration. Conclusion The nomogram is superior to the 8 th TNM staging system in predicting cancer-specific survival of patients with Merkel cell carcinoma and may help doctors to evaluate the prognosis of each patient and facilitate personalized treatments.