Abstract

Renal cancer surgery can adversely impact long-term function and survival. We evaluated predictors of chronic kidney disease 5 years and nonrenal cancer mortality 10 years after renal cancer surgery. We analyzed the records of 4,283 patients who underwent renal cancer surgery from 1997 to 2008. Radical and partial nephrectomy were performed in 46% and 54% of patients, respectively. Cumulative probability ordinal modeling was used to predict chronic kidney disease status 5 years after surgery and multivariable logistic regression was used to predict nonrenal cancer mortality at 10 years. Relevant patient, tumor and functional covariates were incorporated, including the preoperative glomerular filtration rate (A), the new baseline glomerular filtration rate after surgery (B) and the glomerular filtration rate loss related to surgery (C), that is C = A - B. In contrast, partial or radical nephrectomy was not used in the models due to concerns about strong selection bias associated with the choice of procedure. Multivariable modeling established the preoperative glomerular filtration rate and the glomerular filtration rate loss related to surgery as the most important predictors of the development of chronic kidney disease (Spearman ρ = 0.78). Age, gender and race had secondary roles. Significant predictors of 10-year nonrenal cancer mortality were the preoperative glomerular filtration rate, the new baseline glomerular filtration rate, age, diabetes and heart disease (all p <0.05). Multivariable modeling established age and thepreoperative glomerular filtration rate as the most important predictors of10-year nonrenal cancer mortality (c-index 0.71) while the glomerular filtration rate loss related to surgery only changed absolute mortality estimates 1% to 3%. Glomerular filtration rate loss related to renal cancer surgery, whether due to partial or radical nephrectomy, influences the risk of chronic kidney disease but it may have less impact on survival. In contrast, age and the preoperative glomerular filtration rate, which reflects general health status, are more robust predictors of nonrenal cancer mortality, at least in patients with good preoperative function or mild chronic kidney disease.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call