Nomograms incorporating serum C-reactive protein effectively predict mortality before and after surgical treatment of renal cell carcinoma.

OBJECTIVES: To incorporate C-reactive protein into nomograms estimating survival in patients with renal cell carcinoma.
METHODS: Patients undergoing surgery for renal cell carcinoma from 2005-2012 were studied retrospectively. Multivariable Cox proportional hazards regression and competing risks regression models including stage, grade, C-reactive protein levels and presence of metastatic disease were constructed. Outcomes analyzed include overall mortality overall mortality and renal cell carcinoma-specific mortality.
RESULTS: The cohort included 516 patients with a mean follow up of 1.7 years (SD 1.4 years). One- and 3-year renal cell carcinoma-specific mortality was 8.8% and 13.5%, respectively. Four nomograms were generated using overall mortality and renal cell carcinoma-specific mortality as end-points, two each for pre- and postoperative counseling. The factor with the largest effect on all nomograms was preoperative C-reactive protein. Based on the internal validation with bootstrapping, the concordance indices for renal cell carcinoma-specific mortality in the preoperative nomogram, postoperative nomogram, and the Mayo Clinic stage, size, grade and necrosis score were 0.889, 0.893, and 0.832, respectively (P = 0.005 and 0.002 comparing with stage, size, grade and necrosis scores for preoperative or postoperative nomograms). For overall mortality, the preoperative nomogram, postoperative nomogram, and stage, size, grade and necrosis score showed concordance indices of 0.866, 0.897, and 0.828, respectively (P = 0.123 and 0.008 compared with stage, size, grade and necrosis score for preoperative or postoperative nomograms).
CONCLUSIONS: We have generated nomograms incorporating serum C-reactive protein levels that effectively predict overall mortality and renal cell carcinoma specific mortality. Our findings warrant external validation.