Abstract:

Objective  To validate the ANZ model，a prognostic model of metastatic breast cancer among the patients of our breast cancer tumor center. Methods  The data of 104 patients starting first-line treatment for metastatic breast cancer from Jan. 2002 to Dec. 2012 were collected as a validation dataset for ANZ model. Results   We validate the ANZ model in 104 metastatic breast cancer patients of our breast tumor center，the C-index is 0.61（0.53，0.70），and the calibration plot suggested that the accuracy of the predicted OS using the nomogram is unsatisfied in the validation cohort. Conclusion  ANZ model showed an inferior prognostic power in the metastatic breast cancer patients of our tumor center. A more accurate prognostic model trained by Chinese metastatic breast cancer patients is needed.

Key words: metastatic breast cancer, overall survival, prognostic model

摘要： 目的 验证晚期乳腺癌患者的生存预测ANZ模型。方法 收集本中心104例辅助治疗后出现复发或转移、拟接受一线治疗的晚期乳腺癌患者的临床和病理学资料，利用ANZ模型预测该部分患者的生存时间，分析该模型对中国晚期乳腺癌患者的生存预测的鉴别能力和准确性。结果ANZ模型对本中心104例晚期乳腺癌患者进行外验证的C-index为0.61（0.53，0.70），calibrationplot提示模型预测生存时间与实际生存时间的拟合度稍差。结论 ANZ模型对本中心辅助治疗后出现复发或转移、拟接受一线治疗的晚期乳腺癌患者的生存预测效能欠佳，可能不是一个适合中国患者的生存预测模型，未来需要建立一个基于中国患者人群、具有更全面建模因素的晚期乳腺癌生存预测模型。

关键词: 晚期乳腺癌, 总生存, 预测模型