Recurrence risk groups after nephrectomy for renal cell carcinoma

Abstract

IntroductionThere is no consensus on the follow-up protocol after nephrectomy for renal cell carcinoma (RCC), and the identification of recurrence risk groups (RRG) is required. ObjectiveEstablish recurrence risk groups (RRG). Material and methodsA retrospective analysis of 696 patients with renal cancer submitted to surgery between 1990 and 2010; 568 (81.6%) patients treated with radical nephrectomy and 128 (18.4%) treated with partial nephrectomy.Pathological variables were classified as: 1st-level variables (1LPV): pTpN stage and Fuhrman grade (FG); and 2nd level pathological variables (2LPV): sarcomatoid differentiation (SD), tumor necrosis (TN), microvascular invasion (MVI) and positive surgical margins (PSM).Univariate and multivariate analysis have been performed using Cox regression to determine 1LPV related to recurrence. Based on 1LPV, we classified patients into three RRG: Low (LRG) <25%; Intermediate (IRG) 26–50% and High (HRG) >50%.We performed univariate and multivariate analysis with the 2LPVs for each RRG. With these data, patients were reclassified as RRG +.ROC curves were used for comparison of RRG and RRG+. ResultsThe median follow-up was 105 months (range 63–148). There were 177 (25.4%) patients with recurrence: 111 (15.9%) distant, 34 (4.9%) local and 32 (4.6%) distant and local.In the multivariable analysis, Fuhrman grade HR = 2,75; p = 0,0001 and pTpN stage HR = 2,19;p = 0,0001 behaved as independent predictive variables of recurrence.Patients were grouped as RRG AUC = 0,76; p = 0,0001:- LRG (pT1pNx-0 G1-4; pT2pNx-0 G 1-2 ): 456 (65.5%) patients.- IRG (pT2pNx-0 G 3-4 ; pT 3-4 pNx-0 G 1-2 ): 110 (15.8%) patients.- HRG (pT 3-4 pNx-0 G 3-4 ; pT1-4pN+): 130 (18.6%) patients.After multivariate analysis with 2LPV, RRG were reclassified RRG+ AUC = 0,84, p = 0,0001:-LRG+ (LRG without TN, SD and/or PSM(+))-IRG+ (IRG; LRG with TN)-HRG+ (HRG; LRG with SD and/or PSM(+); IRG with TN and/or SD) ConclusionThe inclusion of 2LPV to the classification according to VP1N improves the discriminating capacity of RRG classification.