Background: Renal dysfunction is common prior to liver transplantation (LTx); however, the best way to determine the glomerular filtration rate (GFR) pre-LTx is unclear. Our purpose was to evaluate the performance of two commonly used serum creatinine- (Scr) based GFR estimating equations relative to a measured radionuclide GFR (rGFR). We also studied if pre-LTx GFR was associated with incident chronic kidney disease (CKD) and death post-LTx. Methods: We studied 426 consecutive LTx from 1990-2014 (56 ±11 years, 67% were male, 80% were Caucasian, 29% had hepatitis C, 19% alcoholic liver disease, 11% NASH, 11% hepatoma, and MELD-Na score was 22.2±8.5). We excluded combined organ transplants, re-transplants, and those without rGFR within 1 year prior to LTx. We analyzed the performance (bias and precision) of the CKD-Epi and the MDRD equations vs. rGFR using Bland-Altman plots. We examined the association of each GFR method with a composite outcome of CKD stage 4 (GFR <30 ml/min/1.73 m²), initiation of chronic dialysis, or patient death. Results: The pre-LTx rGFR was 92 ml/min/1.73 m²(IQR 69-116). The correlation coefficient of rGFR with CKD-Epi and MDRD was 0.61 and 0.58, respectively. The MDRD equation showed less bias than CKD-Epi (-4.7 vs. -11.1 ml/min), whereas CKD-Epi was more precise (28.1 vs. 32.9 ml/min). There were 183 patients that reached the composite endpoint over a median follow up of 6 years. All GFR measures were significantly associated with the outcome in univariate analysis. However, after adjusting for potential confounders (age, race, BMI, etiology of end-stage liver disease, pre-existing hypertension and diabetes) in a backward selection approach, only the CKD-Epi equation remained significantly associated with the composite outcome.

« Back to 2015 American Transplant Congress