Application of TNM, 2002 version, in localized renal cell carcinoma: is it able to predict different cancer-specific survival probability?

OBJECTIVES: To verify whether the latest version of the TNM staging system (2002) could predict different cancer-specific survival in patients with localized renal cell carcinoma (RCC; Stage T1-T2N0M0). METHODS: According to the 2002 TNM staging system, we reassigned the pathologic stage of 702 patients who had undergone surgical treatment for RCC from 1976 to 2000. We selected 491 patients with localized RCC (pT1-T2N0M0). In 334 patients (68.0%), we had performed radical nephrectomy; in 121 (24.6%), elective nephron-sparing surgery; and in 36 (7.3%), imperative nephron-sparing surgery. Cancer-specific survival was estimated according to the Kaplan-Meier method. The log-rank test and Cox's proportional hazard model was used for univariate and multivariate analysis, respectively. RESULTS: Of the 491 tumors, 249 (50.7%) were classified as pT1a, 155 (31.6%) as pT1b, and 87 (17.7%) as pT2. The median follow-up was 75 months. The 5-year and 10-year cancer-specific survival probabilities were, respectively, 97.4% and 95.6% in the pT1a patients, 92.5% and 89.8% in the pT1b patients, and 89.3% and 78.5% in the pT2 patients. The survival curve comparison was statistically significant both between pT1a and pT1b (log-rank test, P = 0.01) and between pT1a and pT2 (log-rank test, P = 0.0007). No statistically significant difference was observed between the pT1b and pT2-specific survival probabilities (log-rank test, P = 0.42). CONCLUSIONS: The 2002 TNM staging system does not seem able to predict different cancer-specific survival between pT1b and pT2 RCC. These data highlight the need to define an optimal breakpoint to stratify patients with localized RCC.