Introduction: Serum creatinine (Creat) is not a reliable marker of renal function in patients (pts) with cirrhosis mostly because it is highly influenced by several extrarenal factors. In contrast, recent studies showed that cystatin-C (CysC), a renal biomarker produced by all nucleated cells, is a more sensitive marker of decreased GFR than Creat and a good predictor of hepatorenal syndrome. Postoperative acute kidney injury (AKI) increases morbidity and mortality of liver transplantation (OLT). The goal of this study was to evaluate preoperative Cystatin C levels as a predictor of post-OLT AKI in adults with cirrhosis.

Methods: Prospective study including consecutive pts transplanted in two centers from 2019 to 2021. CysC (immunoturbidimetric assay, normal 0.4-0.99 mg/dl) and Creat (mg/dL) were determined upon admission for the transplant and within 12 hours of surgery. Cut-off values for renal dysfunction were > 1.1 for Creat and >1.4 for CysC (selected based on previous reports). Post-OLT renal dysfunction was defined by AKIN criteria assessed 48 hours after surgery. Preoperative, intraoperative and donor variables were analyzed by univariate and multivariate analysis.

Results: The study included 131 pts aged 52 ± 12.5 years (61% males). HCV was the most frequent etiology of cirrhosis (24%) and 35% of pts had hepatocellular carcinoma. Pre-OLT Creat was 0.85 ± 0.42 mg/dL, Cys 1.46 ± 0.46 mg/dL and MELD-Na 20 ± 8. Increased serum Creat was observed in 15/131 pts (11.4%). In contrast, 66/131 pts (50.4%) had CysC values >1.4 mg/dL (p <0.001). Among the 66 pts with elevated CysC, 53 (80.3%) had normal Creat. No deaths occurred before the primary endpoint of the study. The incidence of postoperative AKI was 43.5% (57/131). Among patients with elevated CysC, 39/66 (59.1%) develop AKI compared to 18/65 (27.7%) with normal CysC (p < 0.001). Independent predictors of post-OLT AKI on multivariate analysis were CysC >1.4 (OR 3.47 95% CI 1.35-9.52, p= 0.012), BMI>30 (OR 3.08 95%CI 1.2-8.43, p=0.024), red blood cell transfusions ≥ 5 units (OR 5.93 95%CI 1.65-24.8, p= 0.011) and preoperative hipoalbuminemia < 2.5 gr/dl (OR 3.96 95%CI 1.39-12.1, p= 0.012).

Conclusions: A large proportion of OLT candidates with cirrhosis and normal Creat have elevated levels of CysC that appears to be a better marker of renal dysfunction in this patient population. Preoperative CysC, but not Creat, was a potent predictor of post-OLT AKI. Early preoperative identification by CysC of pts at risk of post-OLT AKI may play an important role at the time of selecting the immunossupressive regimen, using either induction or nephroprotective agents.