BACKGROUND AND AIM: Evaluation of renal function, that is, glomerular filtration rate (GFR), has become very important, but conventional mathematical formulae for GFR assessment are inaccurate in patients with cirrhosis. The aim of the present study was to compare serum creatinine (sCr)-based and serum cystatin C (cysC)-based estimated GFR (eGFR) formulae with Chromium-ethylenediaminetetraacetic acid GFR ( Chr-GFR) in patients with stable decompensated cirrhosis.
METHODS: In 129 Caucasian patients with decompensated cirrhosis, we assessed sCr-based GFRs [Modification of Diet in Renal Disease and chronic kidney disease-epidemiology (CKD-EPI)-sCr formulae], cysC-based GFRs [Hoek, Larsson, and CKD-EPI-cysC equations], and the mathematical formulae, which combined both sCr and cysC [i.e. CKD-EPI-sCr-cysC and the specific for cirrhotics formula recently proposed by Mindikoglu et al. (Mindikoglu-eGFR)]. Multivariate linear regression analysis was used for GFR predictors in our cohort.
RESULTS: The correlations between Chr-GFR and all mathematical formulae were good (Spearman r  > 0.68, P < 0.001). Modification of Diet in Renal Disease and CKD-EPI-sCr had lower bias (6.6 and -4.8, respectively), compared with the other eGFRs, while Mindikoglu-eGFR and CKD-EPI-sCr-cysC formulae had greater precision (17.1 and 17.3, respectively), compared with the other eGFRs. CKD-EPI-sCr and Mindikoglu-eGFR had higher accuracy (39% and 41%, respectively), compared with the other eGFRs. The factors independently associated with the Chr-GFR were age, cysC, and sCr, and the new derived formula had lower bias (0.89) and similar precision (17.2) and accuracy (41%) with Mindikoglu-eGFR formula.
CONCLUSION: The specific mathematical formulae derived from patients with cirrhosis seem to provide superior assessment of renal function, compared with the conventional used sCr-based and cysC-based formulae.