PANCREATIC FLUSH INJURY IN COMBINED PANCREAS-LIVER RECOVERY

The function of pancreatic grafts harvested from six pancreas-liver (PL)donors was compared to that of nine pancreas-alone (PA) donors. All donors had comparable physiological parameters. Pancreas and liver were flushed in situ with 4C Collins solution and the portal vein was vented immediately. The pancreaticoduodenal grafts were reflushed and stored in Collins solution (three PA and two PL), silica gel-filtered plasma (six PA and two PL), or University of Wisconsin solution (two PL). Later they were revascularized by the iliac vessels and drained into the bladder. All pancreas recipients were insulin-dependent. Serum glucose, amylase, lipase, blood urea nitrogen (BUN), creatinine, protein, albumin, and urine amylase, bicarbonate and pH were monitored for 12 days. Data were analyzed using the Mann-Whitney U-test and Fischer's exact test. The PL grafts received a significantly higher aortic flush volume (5983 cc vs 1622 cc, P = 0.001) than those in the other group. Recipients of PL grafts had higher serum amylase (335 vs 250 IU/l) and lipase (1048 vs 424 IU/l), significantly lower levels of urine bicarbonate (11.2 vs 27.1 mEq/l, P = 0.01), pH (6.8 vs 7.3, P = 0.04), and amylase (9202 vs 19 981 IU/l, P = 0.05) than those in the PA group. Moreover, they required longer periods of gastric suction (4.8 +/- 1.7 vs 2.7 +/- 3.1 days, P = 0.04), and despite central venous nutrition (4/6 vs 1/9 patients, P = 0.02) had significantly lower serum protein (6.0 +/- 0.7 vs 5.2 +/- 0.4 g/dl, P = 0.02) and albumin (2.7 +/- 0.3 vs 3.3 +/- 0.4 g/dl, P = 0.01) when compared to the other group and to the pre-transplant values (P = 0.03). These data suggest that high volume aortic flush induces pancreatic graft injury. Suggestions for improvement are discussed.