摘要
Based on the results of TAX 327, a nomogram was developed to predict the overall survival of metastatic castration-resistant prostate cancer （mCRPC） after first-line chemotherapy. The nomogram, however, has not been validated in an independent dataset, especially in a series out of clinical trials. Thus, the objective of the current study was to validate the TAX 327 nomogram in a community setting in China. A total of 146 patients with mCRPC who received first-line chemotherapy （docetaxel or mitoxantrone） were identified. Because clinical trials are limited in China's Mainland, those patients did not receive investigational treatment after the failure of first-line chemotherapy. The predicted overall survival rate was calculated from the TAX 327 nomogram. The validity of the model was assessed with discrimination, calibration and decision curve analysis. The median survival of the cohort was 21 months （docetaxel） and 19 months （mitoxantrone） at last follow-up. The predictive c-index of the TAX 327 nomogram was 0.66 （95% CI： 0.54-0.70）. The calibration plot demonstrated that the 2-year survival rate was underestimated by the nomogram. Decision curve analysis showed a net benefit of the nomogram at a threshold probability greater than 30%. In conclusion, the present validation study did not confirm the predictive value of the TAX 327 nomogram in a contemporary community series of men in China, and further studies with a large sample size to develop or validate nomograms for predicting survival and selecting therapies in advanced prostate cancer are necessary.
Based on the results of TAX 327, a nomogram was developed to predict the overall survival of metastatic castration-resistant prostate cancer （mCRPC） after first-line chemotherapy. The nomogram, however, has not been validated in an independent dataset, especially in a series out of clinical trials. Thus, the objective of the current study was to validate the TAX 327 nomogram in a community setting in China. A total of 146 patients with mCRPC who received first-line chemotherapy （docetaxel or mitoxantrone） were identified. Because clinical trials are limited in China's Mainland, those patients did not receive investigational treatment after the failure of first-line chemotherapy. The predicted overall survival rate was calculated from the TAX 327 nomogram. The validity of the model was assessed with discrimination, calibration and decision curve analysis. The median survival of the cohort was 21 months （docetaxel） and 19 months （mitoxantrone） at last follow-up. The predictive c-index of the TAX 327 nomogram was 0.66 （95% CI： 0.54-0.70）. The calibration plot demonstrated that the 2-year survival rate was underestimated by the nomogram. Decision curve analysis showed a net benefit of the nomogram at a threshold probability greater than 30%. In conclusion, the present validation study did not confirm the predictive value of the TAX 327 nomogram in a contemporary community series of men in China, and further studies with a large sample size to develop or validate nomograms for predicting survival and selecting therapies in advanced prostate cancer are necessary.
