Abstract: Objective  To analysize pathological character and prognosis factors of Chinese and American MBC patients,and constructing a nomogram for forecasting 3-year,5-year and 8-year overall survival(OS).
Methods  The clinical data of 5 671 MBC patients diagnosed from 2007 to 2012 were extracted from SEER database,which was assigned to modeling set.A retrospective analysis was performed on 101 MBC cases from Xijing hospital in China from 2004 to 2012,which was assigned to validation set.The clinical baseline characteristics of patients in modeling set and validation set were compared,and the independent risk factors for the overall survival(OS) were screened by Cox equal proportional regression model.Based on the results of multiple regression analysis of the modeling set Cox model,the nomogram models for predicting the OS were constructed.the reliability of the predictions obtained from the models were evaluated by C-index,ROC curve and calibration curve.Finally,the clinical benefit and application value of the model were evaluated by clinical decision curve analysis(DCA).
Results  There were no significant difference in the baseline characteristic between the patients in modeling set and validation set except age,surgery,whether first primary tumor，T stage,chemotherapy and radiotherapy.The results of analysis showed that age,surgery,radiation,T stage and M stage were independent risk factors for OS in MBC patients.Based on the above variables,the nomogram models were constructed and evaluated.The C-index of model in modeling set was 0.87,The C-index of model in validation set was 0.71,and the C-index of model for internal validation was 0.81.Moreover,the ROC curve analysis showed that prediction model had a relatively high accuracy and the calibration curve analysis showed that model predicted OS had a good consistency with the actual observed values.Finally,the clinical decision curve analysis showed that model can bring clinical benefit.
Conclusion  A nomogram can accurately predict the prognosis of MBC patients in China and provide scientific basis for clinical diagnosis and treatment.

Key words: mucinous breast cancer, prognosis, nomogram, prediction model