Abstract Background and hypothesis Glomerular filtration rate (GFR) estimation is crucial in renal transplantation. While creatinine-based estimating equations (eGFRcrea) are widely used, their performance is suboptimal due to non-GFR determinants. Recent KDIGO guidelines recommend greater use of cystatin C-based equations, either alone (eGFRcys) or combined with creatinine (eGFRcrea+cys). This study compared the performance of CKD-EPI (2009/2021) and EKFC equations in a large cohort of kidney transplant recipients, also analyzing discordance between eGFRcrea and eGFRcys. Methods This multicenter retrospective study included 1,348 stable kidney transplant recipients from France and the Netherlands. Measured GFR (mGFR) served as the reference. Creatinine and cystatin C were measured using standardized assays. Performance was assessed using bias, P20, P30, and imprecision. Discordance was defined as 20% difference between eGFRcrea and eGFRcys. Results EKFC equations outperformed CKD-EPI counterparts for all biomarkers. EKFCcrea+cys had the highest P30 (87.6%) and was the only unbiased combined equation. Discordant eGFR results were common between eGFRcrea and eGFRcys (42% with EKFC, 50% with CKD-EPI). Combined equations showed better performance in discordant groups, particularly with EKFC. CKD-EPI equations performed poorly in patients with eGFRcrea lower than eGFRcys. Conclusion EKFC equations—especially EKFCcrea+cys—provide superior accuracy in KTR. Discordance between eGFRcrea and eGFRcys is frequent, reinforcing the value of combined equations. Future studies should validate these findings across diverse populations and assess the prognostic impact of discordant GFR estimates.
Delanaye et al. (Wed,) studied this question.