Estimated glomerular filtration rate from the renal hypothermia trial: clinical implications

Objective: To assess if estimated glomerular filtration rate (eGFR) can replace measured GFR (mGFR) in partial nephrectomy (PN) trials, using data from a randomised clinical trial.

Patients and methods: We conducted a post hoc analysis of the renal hypothermia trial. Patients underwent mGFR with diethylenetriaminepentaacetic acid (DTPA) plasma clearance preoperatively and 1 year after PN. The eGFR was calculated using the 2009 Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) creatinine equations incorporating age and sex, with and without race: 2009 eGFRcr(ASR) and 2009 eGFRcr(AS), and the 2021 equation that only incorporates age and sex: 2021 eGFRcr(AS). Performance was evaluated by determining the median bias, precision (interquartile range [IQR] of median bias), and accuracy (percentage of eGFR within 30% of mGFR).

Results: Overall, 183 patients were included. Pre- and postoperative median bias and precision were similar between the 2009 eGFRcr(ASR) (-0.2 mL/min/1.73 m² , 95% confidence interval [CI] -2.2 to 1.7, IQR 18.8; and -2.9, 95% CI -5.1 to -1.5, IQR 15, respectively) and 2009 eGFRcr(AS) (-0.3 mL/min/1.73 m² , 95% CI -2.4 to 1.5, IQR 18.8; and -3.0, 95% CI -5.7 to -1.7, IQR 15.0, respectively). Bias and precision were worse for the 2021 eGFRcr(AS) (-8.8 mL/min/1.73 m² , 95% CI -10.9 to -6.3, IQR 24.7; and -12.0, 95% CI -15.8 to -8.9, IQR 23.5, respectively). Similarly, pre- and postoperative accuracy was >90% for the 2009 eGFRcr(ASR) and 2009 eGFRcr(AS) equations_. Accuracy was 78.6% preoperatively and 66.5% postoperatively for 2021 eGFRcr(AS).

Conclusion: The 2009 eGFRcr(AS) can accurately estimate GFR in PN trials and could be used instead of mGFR to reduce cost and patient burden.