OBJECTIVE To examine the cancer-specific mortality (CSM) of patients with T4N0-2M0 renal cell carcinoma (RCC) treated with either nephrectomy (RN) or no surgery (NS). PATIENTS AND METHODS Of 43 143 patients with RCC identified in the Surveillance, Epidemiology and End Results database, 310 had tumours involving adjacent organs with no evidence of distant metastases (T4NanyM0) and had RN (246, 79.4%) or NS (64, 20.6%). Kaplan-Meier analyses, Cox regression and competing-risks regression models were used to compare the effect of RN vs NS on CSS. RESULTS In patients with T4N0 disease the median survival benefit associated with RN vs NS was 42 months (48 vs 6 months, P <0.001). Conversely, the median survival in patients T4N1-2 was no different between RN and NS (9.3 vs 9.1 months, P = 0.9). Multivariable analyses in T4N0 cases indicated a substantial survival disadvantage for patients having NS vs RN (hazard ratio 4.8, P <0.001). Conversely, in patients with N1-2 stages, the CSS was virtually the same for NS and RN (hazard ratio 0.9, P = 0.9). Competing-risks regression models confirmed the benefit of RC in patients with T4N0 and the lack of benefit in those with T4N1-2 disease, after controlling for other-cause mortality. CONCLUSION Our data suggest a survival benefit in patients with T4N0 RCC treated with RC. By contrast, RN seems to have no effect on survival in patients with evidence of nodal metastases.
OBJECTIVE To examine the cancer-specific mortality (CSM) of patients with T4N0-2M0 renal cell carcinoma (RCC) treated with either nephrectomy (RN) or no surgery (NS). PATIENTS AND METHODS Of 43 143 patients with RCC identified in the Surveillance, Epidemiology and End Results database, 310 had tumours involving adjacent organs with no evidence of distant metastases (T4NanyM0) and had RN (246, 79.4%) or NS (64, 20.6%). Kaplan-Meier analyses, Cox regression and competing-risks regression models were used to compare the effect of RN vs NS on CSS. RESULTS In patients with T4N0 disease the median survival benefit associated with RN vs NS was 42 months (48 vs 6 months, P <0.001). Conversely, the median survival in patients T4N1-2 was no different between RN and NS (9.3 vs 9.1 months, P = 0.9). Multivariable analyses in T4N0 cases indicated a substantial survival disadvantage for patients having NS vs RN (hazard ratio 4.8, P <0.001). Conversely, in patients with N1-2 stages, the CSS was virtually the same for NS and RN (hazard ratio 0.9, P = 0.9). Competing-risks regression models confirmed the benefit of RC in patients with T4N0 and the lack of benefit in those with T4N1-2 disease, after controlling for other-cause mortality. CONCLUSION Our data suggest a survival benefit in patients with T4N0 RCC treated with RC. By contrast, RN seems to have no effect on survival in patients with evidence of nodal metastases.