Estimating Glomerular Filtration Rate in Cirrhosis Using Creatinine-Based and Cystatin C-Based Equations: Systematic Review and Meta-Analysis

Abstract

Accurate estimation of kidney function in cirrhosis is crucial for prognosis and decisions regarding dual-organ transplantation. We performed a systematic review/meta-analysis to assess the performance of creatinine-based and cystatin C (CysC)-based eGFR equations compared with measured GFR (mGFR) in patients with cirrhosis. A total of 25 studies (n = 4565, 52.0 years, 37.0% women) comprising 18 equations met the inclusion criteria. In all GFR equations, the creatinine-based equations overestimated GFR (standardized mean difference, SMD, 0.51; 95% confidence interval [CI], 0.31-0.71) and CysC-based equations underestimated GFR (SMD, -0.3; 95% CI, -0.60 to -0.02). Equations based on both creatinine and CysC were the least biased (SMD, -0.14; 95% CI, -0.46 to 0.18). Chronic kidney disease-Epi-serum creatinine-CysC (CESC) was the least biased but had low precision and underestimated GFR by -3.6 mL/minute/1.73 m² (95% CI, -17.4 to 10.3). All equations significantly overestimated GFR (+21.7 mL/minute/1.73 m² ; 95% CI, 17.7-25.7) at GFR <60 mL/minute/1.73 m² ; of these, chronic kidney disease-Epi-CysC (10.3 mL/minute/1.73 m² ; 95% CI, 2.1-18.4) and GFR Assessment in Liver Disease (12.6 mL/minute/1.73 m² ; 95% CI, 7.2-18.0) were the least biased followed by Royal Free Hospital (15 mL/minute/1.73 m² ; 95% CI, 5.5-24.6) and Modification of Diet in Renal Disease 6 (15.7 mL/minute/1.73 m² ; 95% CI, 10.6-20.8); however, there was an overlap in the precision of estimates, and the studies were limited. In ascites, overestimation of GFR was common (+8.3 mL/minute/1.73 m² ; 95% CI, -3.1 to 19.7). However, overestimation of GFR by 10 to 20 mL/minute/1.73m² is common in patients with cirrhosis with most equations in ascites and/or kidney dysfunction. A tailored approach is required especially for decisions regarding dual-organ transplantation.