预测去势抵抗性前列腺癌患者总生存期Halabi风险列线图的验证研究

目的 在中国上海及周边地区去势抵抗性前列腺癌（CRPC）患者中验证Halabi风险列线图，预测CRPC患者的总生存期。方法 选择2006年8月至2013年12月在复旦大学附属肿瘤医院被确诊为CRPC的228例患者。预测患者生存率通过Halabi风险列线图计算，模型效应用偏倚和校准分析评估。结果 228例患者中死亡150例，中位总生存期为20.8个月（4～48个月）。Halabi风险列线图的准确度为0.61（95％ CI：0.51～0.73）。Halabi列线图预测的中位总生存期为16.4个月（6～31个月）。预测模型校准结果显示，1年和2年肿瘤特异性生存率被低估。结论 Halabi风险列线图未能准确预测中国上海地区CRPC患者的生存率，仍需要更大样本的研究来验证Halabi风险列线图或者建立新的风险列线图来预测CRPC患者的生存情况。

External validation of Halabi nomogram for predicting overall survival of castrate resistant prostate cancer patients

Objective To predict the overall survival of castration resistant prostate cancer(CRPC) patients from Shanghai and the surrounding areas by Halabi nomogram.Methods A total of 228 patients with CRPC in Tumour Hospital Affiliated to Fudan University from August 2006 to December 2013 were enrolled.The overall survival was predicted and calculated by Halabi nomogram.The validity of the model was assessed by discrimination and calibration.Results The median overall survival of the cohort was 20.8 months (4-48 months) followed-up with 150 death events.The predictive accuracy of Halabi nomogram was 0.61 (95%CI 0.51-0.73).Predicted median overall survival was 16.4 months(6-31months).Calibration plot demonstrated that the 1-and 2-year survival rate was remarkably underestimated by the nomogram.Conclusion The present validation study does not confirm the predictive value of Halabi nomogram in a contemporary community series and further studies with a large sample size to validate or develop nomograms for predicting survival in CRPC are necessary.