Discordance between creatinine and cystatin C-based estimation of glomerular filtration rate (eGFR) in solid organ transplant recipients

Abstract

BackgroundEstimation of glomerular filtration rate is a critical component of assessing kidney function post-solid organ transplantation and in monitoring risk of acute injury. Our objective was to evaluate estimated glomerular filtration rate (eGFR) as derived from creatinine (eGFRcr), cystatin C (eGFRcys), and both (eGFRcr-cys) in a cohort of transplant recipients. MethodsA total of 47 unique post-solid organ transplant patients receiving tacrolimus were included. Residual specimens were assayed for creatinine (Jaffe and enzymatic), cystatin C, and tacrolimus. eGFR was estimated using the 2021 CKD-EPI formulae. Results were compared by Deming regression and bias was assessed using non-parametric cumulative distribution plots. Percent agreement in chronic kidney disease (CKD) by stage was evaluated across equations. ResultseGFRcys relative to eGFRcr estimates demonstrated a median bias of –22 mL/min/1.73 m2 and an overall 21.3 % [95 % CI: 12.1, 35.0] agreement in CKD staging. eGFRcr-cys demonstrated a median bias of −14 mL/min/1.73 m2 and overall agreement of 34.0 % [95 % CI: 22.3, 48.4] relative to eGFRcr (enzymatic). Discordance increased proportionally with eGFR and did not differ by creatinine assay (Jaffe or enzymatic). ConclusionCystatin C incorporation leads to markedly negative biases between eGFR estimates in patients with solid organ transplant, implying lack of applicability of eGFRcys or GFRcr-cys in the transplant setting. This highlights the dependency of patient characteristics on equation performance and the need to consider confounding factors in interpretation and utilization of cystatin C based equations.