CUA 2017: Creation of a Prediction Tool for Renal Function After Partial and Radical Nephrectomy: Personalizing Decision-Making for Renal Cancer Surgery

Toronto, Ontario ( This study aimed to create a preoperative prediction tool for renal function outcomes at various time points following partial nephrectomy (PN) and radical nephrectomy (RN) in order to help guide the choice of surgical approach.

The Mayo Clinic Nephrectomy Registry was queried for patients who underwent PN or RN for a renal tumour between 1997 and 2013. Exclusions were nodal or distant metastases, venous tumour thrombus on imaging, and preoperative estimated glomerular filtration rate (eGFR) <15 mL/min. Parsimonious linear regression models predicting eGFR were also created for PN and RN using backward selection of candidate preoperative predictors, and eGFR predictions at one year were presented. Adjusted R2, a value ranging from 0−1 that represents the proportion of total variation in eGFR explained by the model, was used to quantify predictive ability.

The analytic cohort included 1525 and 935 patients undergoing PN and RN, respectively. Mean (standard deviation [SD]) preoperative eGFR and tumour size were 72 (20) mL/min and 3.4 (1.9) cm, respectively, for patients undergoing PN, and 65 (18) mL/min and 7.1 (3.8) cm, respectively, for patients undergoing RN. The model for PN included age, presence of a solitary kidney, smoking status, performance status, body mass index (BMI), preoperative eGFR, tumour size, and open vs. lap surgical approach (R2=0.78), while the model for RN included age, diabetes, BMI, preoperative eGFR, tumour size, and surgical approach (R2=0.68). As an example using these models, a 68-year-old, non-smoking, non-diabetic, Eastern Cooperative Oncology Group (ECOG) 0, binephric patient with a BMI of 20kg/m2, a preoperative eGFR of 100 mL/min, and a 6.5 cm renal mass will have a predicted eGFR of 85 mL/min following open PN and 63 following laparoscopic RN at one year.

In summary, the authors created a prediction tool for renal function following RN and PN. If validated in other larger cohorts, this tool may be useful during patient counselling by providing personalized predicted renal function outcomes. 

Presented By: Bimal Bhindi, MD, Mayo Clinic, Rochester, MN, United States

Written By: Hanan Goldberg, MD, Urologic Oncology Fellow (SUO), University of Toronto, Princess Margaret Cancer Centre   Twitter: @GoldbergHanan at the 72nd Canadian Urological Association Annual Meeting - June 24 - 27, 2017 - Toronto, Ontario, Canada