A critical assessment of the prognostic value of clear cell, papillary and chromophobe histological subtypes in renal cell carcinoma: a population-based study.
OBJECTIVE To assess the magnitude of the effect of histological subtype (HS, the three most common being clear cell, papillary and chromophobe) on cause-specific mortality (CSM) from renal cell carcinoma (RCC). PATIENTS AND METHODS Univariable and multivariable Cox regression models included data from 11 618 patients treated with nephrectomy between 1988 and 2004 in nine Surveillance Epidemiology and End Results registries. We tested whether HS represents an independent predictor of CSM, and whether HS adds to the ability of other variables to predict CSM. The covariates comprised age, year of surgery, T stage, nodal status, M stage and Fuhrman grade. RESULTS In a multivariable model predicting CSM, HS was an independent predictor (P = 0.03), but failed to improve the accuracy of the model (+0.1% gain when HS was included in the model). CONCLUSION Although we confirmed that HS is an independent predictor for CSM, there was no gain in accuracy when HS was added to standard predictors of CSM. From a practical perspective, this implies that patients with clear cell, papillary and chromophobe HS share similar natural histories after nephrectomy, provided that other cancer characteristics are accounted for. From a statistical perspective, in multivariable models of CSM, the clear cell, papillary and chromophobe HS might be included as a single entity.
OBJECTIVE To assess the magnitude of the effect of histological subtype (HS, the three most common being clear cell, papillary and chromophobe) on cause-specific mortality (CSM) from renal cell carcinoma (RCC). PATIENTS AND METHODS Univariable and multivariable Cox regression models included data from 11 618 patients treated with nephrectomy between 1988 and 2004 in nine Surveillance Epidemiology and End Results registries. We tested whether HS represents an independent predictor of CSM, and whether HS adds to the ability of other variables to predict CSM. The covariates comprised age, year of surgery, T stage, nodal status, M stage and Fuhrman grade. RESULTS In a multivariable model predicting CSM, HS was an independent predictor (P = 0.03), but failed to improve the accuracy of the model (+0.1% gain when HS was included in the model). CONCLUSION Although we confirmed that HS is an independent predictor for CSM, there was no gain in accuracy when HS was added to standard predictors of CSM. From a practical perspective, this implies that patients with clear cell, papillary and chromophobe HS share similar natural histories after nephrectomy, provided that other cancer characteristics are accounted for. From a statistical perspective, in multivariable models of CSM, the clear cell, papillary and chromophobe HS might be included as a single entity.