Bile Biochemistry Following Liver Reperfusion in the Recipient and Its Association With Cholangiopathy

Abstract

Cholangiocytes secrete bicarbonate and absorb glucose, producing bile with alkaline pH and low glucose content. These functions of cholangiocytes have been suggested as a marker of bile duct viability during normothermic ex situ liver perfusion, and they are now monitored routinely after reperfusion in our center. In this study, we reviewed the composition of bile immediately after reperfusion in liver transplant recipients to determine normal posttransplant parameters and the predictive value of bile biochemistry for the later development of cholangiopathy. After reperfusion of the liver graft, a cannula was placed in the bile duct to collect bile over a median 44-minute time period. The bile produced was analyzed using a point-of-care blood gas analyzer (Cobas b221, Roche Diagnostics, Indianapolis, IN). A total of 100 liver transplants (35 from donation after circulatory death and 65 from donation after brain death) were studied. Median bile pH was 7.82 (interquartile range [IQR], 7.67-7.98); median bile glucose was 2.1 (1.4-3.7) mmol/L; median blood-bile-blood pH difference was 0.50 (0.37-0.62); and median blood-bile glucose difference was 7.1 (5.6-9.1) mmol/L. There were 12 recipients who developed cholangiopathy over a median follow-up of 15 months (IQR, 11-20 months). Bile sodium (142 versus 147 mmol/L; P = 0.02) and blood-bile glucose concentration differences (5.2 versus 7.6 mmol/L; P = 0.001) were significantly lower and were associated with ischemic cholangiopathy. In conclusion, bile biochemistry may provide useful insights into cholangiocyte function and, hence, bile duct viability. Our results suggest bile glucose is the most sensitive predictor of cholangiopathy.

Fig. 1

A graph illustrating the relationship of glucose concentration in the bile and blood after reperfusion of the liver in the recipients. The straight line indicates where bile and blood glucose are the same. We would expect all readings to lie below the line. Patients with cholangiopathy (indicated by the circles) had less difference between biliary and blood glucose, but there were patients without cholangiopathy whose values were on or near the line of equivalence. The graph also illustrates the saturation of the glucose transporter, where the bile glucose starts at a higher blood glucose concentration.

Fig. 2

The relationship between bile and blood pH following reperfusion of the liver in the recipients. The straight line represents the equivalence between bile and blood pH. With one exception, bile pH was higher than that of blood. Circles represent the patients with proven cholangiopathy.

Fig. 3

An ROC curve of the blood‐bile glucose difference, bile sodium, and bile pH as a predictor for the development of cholangiopathy. A blood‐bile glucose difference of <6.5 mmol/L has 83% sensitivity and 62% specificity for predicting cholangiopathy (AUROC, 0.80; P = 0.001). No such cutoff value can be made from the bile sodium and bile pH. The AUROC (95% CI)s were 0.80 (0.68‐0.91; P = 0.001) for the blood‐bile glucose difference; 0.71 (0.53‐0.89; P = 0.02) for bile sodium; and 0.64 (0.48‐0.79; P = 0.12) for bile pH.