Should the CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration) equation be used to evaluate renal function in the hepatic failure patient awaiting liver transplantation?

Purpose: Estimation of renal clearance in patients with hepatic dysfunction is difficult, at best, to determine by mathematical equation.  Currently several methods are utilized to estimate CrCl which include Cockcroft & Gault(CC&G) equation, modification of diet in renal disease (MDRD-6&-4) equations, the Nix model, and a newly described CKD-EPI(Chronic Kidney Disease Epidemiology Collaboration) equation.  The aim of this study is to compare the CKD-EPI and other methods of renal clearance estimation to actual, collected 24 hour creatinine clearance in hepatic failure patients.

Methods: This IRB approved study collected 24 hour CrCl from patients awaiting liver transplant.  The actual 24 hr CrCl was then compared to the CC&G, MDRD-6/-4, Nix model, and the CKD-EPI formula.  Strength of correlation was then determined by the Pearson R method with good correlation being set at 0.75.

Results: A total of 120 patients were evaluated with a mean age of 53 years, 59.2% male, and 13% African-American.  Best correlation was seen with the CC&G and Nix’s formula (r=0.680 vs. r=0.683) with the MDRD-6/-4 and CKD-EPI a lower correlation (r=0.563, 0.592, 0.575).  The mean calculated GFR from collection (99 ml/min) was very different from calculated means[88 ml/min(CC&G), 82.9 ml/min(CPK-EPI), 83.5 ml/min(MDRD-4), 65.3 ml/min(MDRD-6) and 86.3 ml/min(Nix’s)]. For patients with MELD>15, the Nix’s formula better correlated CrCl than other methods(r=0.841,0.772 and 0.625).  Based on type of liver disease, the MDRD-6 had the best correlation for immune mediated diseases(r=0.962).  For patients with CrCl<60ml/min, correlation was low, but similar(r=0.528, 0.507, 0.521, 0.531, & 0.528)with similar results forCrCl >60ml/min (r=0.557, 0.40, 0.447, 0.418, & 0.539).

Conclusion: The newer equations did no better than the CC&G in estimating CrCl for patients awaiting transplant. Due to lower than expected Pearson r with the new renal validated formulas, addition work should be undertaken to estimate clearance in hepatic failure patients.