eGFR Inaccuracy in the Assessment of Living Kidney Transplant Donor Eligibility

Accurate glomerular filtration rate (GFR) estimation is crucial for evaluating living kidney donors, especially when measured GFR (mGFR) is unavailable. This study compares the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI), full age spectrum (FAS), and European Kidney Function Consortium (EKFC) creatinine- and/or cystatin C-based estimated GFR (eGFR) equations against iohexol mGFR to determine the optimal equation for donor eligibility assessment in a US population.

1210 kidney donor candidates were retrospectively analyzed, comparing eGFR equations based on creatinine, cystatin C, or both against iohexol mGFR. Accuracy metrics (P10, P30, mean bias) and subgroup analyses for Black donors and age-based bias were evaluated. The classification performance of eGFR equations in donor eligibility was examined.

The 2021 CKD-EPIcr [AS (age, sex)] demonstrated the best accuracy across the overall cohort (P10 48.2%, P30 93.2%, mean bias -4.8 mL/min/1.73 m²). The 2021 CKD-EPIcr-cys (AS) excelled in Black donors (P10 60.3%, P30 95.4%, mean bias -3.6 mL/min/1.73 m²). Cystatin C-based equations showed higher negative bias, with the largest underestimation observed in older donors. All equations demonstrated <35% positive predictive value (PPV) for rejecting ineligible donors (<60 mL/min/1.73 m²) but >85% PPV in determining acceptable donors (≥90 mL/min/1.73 m²). Overall, 2021 CKD-EPIcr-cys (AS) was the most reliable for identifying acceptable donors (F1 score 83.9 at ≥90 mL/min/1.73 m²). Using age/sex-specific thresholds improved performance of all equations in donor eligibility classification compared to absolute thresholds.

No eGFR equation reliably rejected ineligible donors. 2021 CKD-EPIcr-cys (AS) exhibited the best overall performance for donor eligibility assessment.