Low C-reactive Protein and Urea Distinguish Primary Nonfunction From Early Allograft Dysfunction Within 48 Hours of Liver Transplantation

Abstract

Primary nonfunction (PNF) is a life-threatening complication of liver transplantation (LT), but in the early postoperative period, it can be difficult to differentiate from early allograft dysfunction (EAD). The aim of this study was to determine if serum biomarkers can distinguish PNF from EAD in the initial 48 h following LT.

Materials and methods: A retrospective study of adult patients that underwent LT between January 2010 and April 2020 was performed. Clinical parameters, absolute values and trends of C-reactive protein (CRP), blood urea, creatinine, liver function tests, platelets, and international normalized ratio in the initial 48 h after LT were compared between the EAD and PNF groups.

Results: There were 1937 eligible LTs, with PNF and EAD occurring in 38 (2%) and 503 (26%) patients, respectively. A low serum CRP and urea were associated with PNF. CRP was able to differentiate between the PNF and EAD on postoperative day (POD)1 (20 versus 43 mg/L; P < 0.001) and POD2 (24 versus 77; P < 0.001). The area under the receiver operating characteristic curve (AUROC) of POD2 CRP was 0.770 (95% confidence interval [CI] 0.645-0.895). The urea value on POD2 (5.05 versus 9.0 mmol/L; P = 0.002) and trend of POD2:1 ratio (0.71 versus 1.32 mmol/L; P < 0.001) were significantly different between the groups. The AUROC of the change in urea from POD1 to 2 was 0.765 (95% CI 0.645-0.885). Aspartate transaminase was significantly different between the groups, with an AUROC of 0.884 (95% CI 0.753-1.00) on POD2.

Discussion: The biochemical profile immediately following LT can distinguish PNF from EAD; CRP, urea, and aspartate transaminase are more effective than ALT and bilirubin in distinguishing PNF from EAD in the initial postoperative 48 h. Clinicians should consider the values of these markers when making treatment decisions.

FIGURE 1.

Schematic representation of CRP and urea production pathways in the hepatocyte. A, Diagram demonstrating the process of CRP and urea production within the hepatocyte. B, Urea cycle showing metabolites and enzymes (italics) involved in sequential steps. Diagram based on that provided by Bigot et al. CRP, C-reactive protein; IL-6, interleukin-6.

FIGURE 2.

Patients included and excluded in the study. CNS, central nervous system; HAT, hepatic artery thrombosis; MHN, massive hemorrhagic necrosis ; PV, portal vein; VOO, venous outlet obstruction.

FIGURE 3.

ROC curves demonstrating the ability of different biomarkers to predict PNF, grouped according to biomarker. A, ROC curves for predicting PNF from commonly measured biochemical markers. ALT, alanine aminotransferase; AST, aspartate aminotransferase; CRP, C-reactive protein; INR, international normalized ratio; PNF, primary nonfunction; ROC, receiver operating characteristic. B, ROC curves separated by postoperative day for the commonly measured biochemical markers.

FIGURE 4.

Plots demonstrating the relationship between biomarker levels and PNF rate. Binary logistic regression model graphs for the markers (absolute values only) that had an area under the receiver operating characteristic curve of ≥0.75. The solid line on the graph demonstrates the PNF incidence (y-axis) that can be expected at the given value demonstrated on the x-axis. AST, aspartate transaminase; CRP, C-reactive protein; INR, international normalized ratio; PNF, primary nonfunction.

FIGURE 5.

Timing of graft failure due to PNF and its relationship to existing scoring systems. Bar graph demonstrating the time point of either retransplant or death due to PNF following liver transplantation. The dashed lines demonstrate published early graft function prediction models. EASE, Early Allograft Failure Simplified Estimation; L-GrAFT, Liver Graft Assessment Following Transplantation; MEAF, Model for Early Allograft Failure; PNF, primary nonfunction.