New Baseline Renal Function after Radical or Partial Nephrectomy: A Simple and Accurate Predictive Model

Abstract

Purpose: Preoperative estimation of new baseline glomerular filtration rate after partial nephrectomy or radical nephrectomy for renal cell carcinoma has important clinical implications. However, current predictive models are either complex or lack external validity. We aimed to develop and validate a simple equation to estimate postoperative new baseline glomerular filtration rate.

Materials and methods: For development and internal validation of the equation, a cohort of 7,860 patients with renal cell carcinoma undergoing partial nephrectomy/radical nephrectomy (2005-2015) at the Veterans Affairs National Health System was analyzed. Based on preliminary analysis of 94,327 first-year postoperative glomerular filtration rate measurements, new baseline glomerular filtration rate was defined as the final glomerular filtration rate within 3 to 12 months after surgery. Multivariable linear regression analyses were applied to develop the equation using two-thirds of the renal cell carcinoma Veterans Administration cohort. The simplest model with the highest coefficient of determination (R²) was selected and tested. This model was then internally validated in the remaining third of the renal cell carcinoma Veterans Administration cohort. Correlation/bias/accuracy/precision of equation were examined. For external validation, a similar cohort of 3,012 patients with renal cell carcinoma from an outside tertiary care center (renal cell carcinoma-Cleveland Clinic) was independently analyzed.

Results: New baseline glomerular filtration rate (in ml/minute/1.73 m²) can be estimated with the following simplified equation: new baseline glomerular filtration rate = 35 + preoperative glomerular filtration rate (× 0.65) - 18 (if radical nephrectomy) - age (× 0.25) + 3 (if tumor size >7 cm) - 2 (if diabetes). Correlation/bias/accuracy/precision were 0.82/0.00/83/-7.5-8.4 and 0.82/-0.52/82/-8.6-8.0 in the internal/external validation cohorts, respectively. Additionally, the area under the curve (95% confidence interval) to discriminate postoperative new baseline glomerular filtration rate ≥45 ml/minute/1.73 m² from receiver operating characteristic analyses were 0.90 (0.88, 0.91) and 0.90 (0.89, 0.91) in the internal/external validation cohorts, respectively.

Conclusions: Our study provides a validated equation to accurately predict postoperative new baseline glomerular filtration rate in patients being considered for radical nephrectomy or partial nephrectomy that can be easily implemented in daily clinical practice.

Keywords: glomerular filtration rate; kidney neoplasms; nephrectomy.