External validation of a predictive model of survival after cytoreductive nephrectomy for metastatic renal cell carcinoma.

Abstract

Background: In 2013, a nomogram was developed for pre- and postoperative prediction of the probability of death (PoD) following CN in patients with metastatic renal cell carcinoma (mRCC). The nomogram, generated from a single-institution database, included patients from the cytokine era and has not been externally validated. We reevaluated this model in contemporary patients treated in the VEGF-targeted therapy era in a multi-institution validation cohort.

Design, Setting, and Participants: Multi-institutional European and North American data from patients who underwent CN between 2006 and 2013 were used for external validation of the pre- and postoperative nomogram based on serum albumin and serum lactate dehydrogenase and the postoperative addition of pathologic primary tumour and nodal stage and intraoperative blood transfusion. In addition, patient characteristics and MSKCC risk factors were collected. Using the original calibration indices and quantiles of the distribution of predictions, Kaplan-Meier estimates and calibration plots of observed versus predicted PoD were calculated. For the preoperative model, a decision curve analysis (DCA) was performed and compared to the MSKCC model.

Results and limitations: Of 1108 patients with a median OS of 27 months [95% CI 24.6-29.4], 546 and 482 patients had full data for the validation of the pre-and postoperative models, respectively. The AUC for the pre- and postoperative model was 0.6794 [95% CI 0.6241-0.7350] and 0.7281 [95% CI 0.6773-0.7739], respectively. In the DCA the preoperative model performs well within threshold survival probabilities of 20-50%. Limitations are the retrospective design and the use of OS instead of Cancer Specific Survival used in the internal validation.

Conclusions: In external validation the pre- and postoperative nomograms perform similarly to the AUCs reported for MSKCC and IMDC mRCC risk prediction models, which were not developed to address the decision on performing CN. Although performance of the preoperative nomogram was lower than in the internal validation, it retains ability to predict early death after CN.