Improving the definition of high-risk patients for tumor recurrence from clear-cell renal cell carcinoma – The U-CISS classification

629 Background: Thirty percent of patients with localized clear cell renal cell carcinoma (ccRCC) will ultimately develop recurrence (local or metastatic) after nephrectomy. Current risk stratification systems still misclassify patients. A better stratification is needed to select high-risk patients who may benefit from adjuvant therapy. We sought to improve the existing UISS by the addition of genetic information in a new UCLA cytogenetic integrated staging system (U-CISS). Methods: A total of 240 patients from UCLA with localized ccRCC and cytogenetic analysis on tumor specimen were included in the study. In a continuation of our previous research, cytogenetic (combined loss 3p-14q) and a pathological high-risk feature, microvascular invasion (MVI) were implemented in the UISS. Association with recurrence free survival (RFS) was analyzed in a uni- and multivariable fashion; prognostic accuracy was tested with the c-index. All tumors that had either MVI, combined loss 3p/14q or both in the UISS low-risk group were placed into the U-CISS intermediate group. Tumors with one or both risk factors in the UISS intermediate group were placed into the new U-CISS high-risk group. Results: Fifty patients developed tumor recurrence. On multivariate analysis, combined loss 3p-14q, and MVI were independent prognostic factors. The U-CISS placed significantly better prognosticated RFS in the high-risk group (7/50 (14%) with UISS vs. 23/50 (46%) with U-CISS) and thus, was more accurate in prognosticating RFS. The c-index for recurrence prognostication was improved in the U-CISS (0.70 vs. 0.65 for the UISS). Furthermore, the U-CISS was a better prognostication tool when the intermediate and high-risk group were combined (prognostication of 74% with U-CISS vs. 68% UISS). Conclusions: The use of U-CISS, which integrates genomic alterations with clinical and pathologic features, allowed a re-allocation of patients to create a better stratification of recurrence risk. This new definition of high-risk of recurrence could significantly improve selection of patients who are in greatest need of closer surveillance and/or adjuvant treatment. C.L. and N.K. contributed equally to first authorship.