External Validation of a Postoperative Renal Function Predictive Model Following Partial Nephrectomy

Area of Science:

• Nephrology
• Urology
• Medical Informatics

Background:

• Accurate estimation of kidney function post-partial nephrectomy (PN) is crucial for patient management.
• A predictive model for post-operative estimated glomerular filtration rate (eGFR) was developed by the Veterans Affairs National Health System.
• The model achieved a coefficient of determination (R ²) of 0.66 and predicted GFR within 30% for 82% of patients (P30).

Purpose of the Study:

• To externally validate the predictive model's performance in a distinct patient cohort.
• To assess the model's accuracy in predicting 1-year post-operative eGFR in patients from a renal hypothermia trial.

Main Methods:

• The model was validated using data from a randomized clinical trial involving 175 patients.
• Glomerular filtration rate (GFR) was measured using ^99mTc-DTPA plasma clearance pre- and post-PN.
• Model performance was evaluated using R², calibration slope, and precision metrics (P10, P20, P30).

Main Results:

• The model demonstrated strong performance with an R² of 0.67 in the trial cohort.
• Precision metrics showed 41.7% within 10% (P10), 73.7% within 20% (P20), and 89.1% within 30% (P30) of predicted GFR.
• A calibration slope of 1.04 indicated good calibration across various baseline renal function levels.

Conclusions:

• The predictive model for post-operative GFR performed well in patients from the renal hypothermia trial.
• The model, utilizing ^99mTc-DTPA plasma clearance measurements, can support informed decision-making for patients undergoing partial nephrectomy.
• The findings suggest the model's utility in counseling patients about expected kidney function outcomes after PN.