Steven M. Strasberg

Sinusoidal lining cell damage : the critical injury in cold preservation of liver allografts in the rat

We have previously defined viability limits in a rat transplantation model. All liver allografts stored in a simple preservation solution (NaCl 0.9%, CaCl2 2 mM) at 4 degrees C for 4 hr or at 37 degrees C for 1 hr were viable upon transplantation, but all those stored at 4 degrees C for 8 hr or at 37 degrees C for 2 hr were nonviable. Only cold-preserved, nonviable livers showed increased vascular resistance, platelet trapping and an initially low, but then high, rise in aspartate transaminase (AST) upon reperfusion, all suggesting injury to the microcirculation, with secondary injury to the hepatocyte. In the present study, we investigated the morphological changes that occur in livers stored for the defined critical times, using light and electron microscopy after perfusion-fixation. Accurate and reproducible identification of specimens as belonging to viable or nonviable and warm- or cold-preserved could be made in this way. Preservation in the cold first resulted in reversible changes consisting of cellular swelling, alterations of intracellular organelles, and partial denudation of the sinusoidal lining (cold-preserved viable group). Later, under conditions of nonviable cold preservation, detachment of cell bodies of sinusoidal lining cells with nuclear changes and almost complete denudation of the sinusoidal lining was observed. Endothelial cells of larger vessels were only injured mildly. In contrast, under conditions of warm preservation, changes involving mitochondria and later nuclei were found in hepatocytes, and blebbing was more extensive. Endothelial cells were spared relatively. We also examined livers stored in isotonic citrate solution at 4 degrees C for 8 hr and 16 hr, the critical times determined for this solution in another model of rat liver transplantation. The findings were very similar to storage in saline with respect to the changes in the sinusoidal lining cells after cold preservation for the two critical times. The results provide convincing evidence of a qualitative difference between warm and cold preservation injury, with relatively selective damage to hepatocytes or sinusoidal lining cells, respectively. Endothelial damage represents the primary event, resulting in the loss of organ viability following hypothermic storage. Thus morphology may serve as a useful viability marker after preservation.

Treatment of primary liver graft nonfunction with prostaglandin E1

Primary nonfunction following orthotopic liver transplantation is characterized by rapidly rising serum transaminases, minimal bile production, and severe coagulopathy, which can progress to hypoglycemia, hepatic encephalopathy, and acute renal failure. Untreated it has a mortality of over 80% and to date the only treatment has been retransplantation. As a result of the beneficial effect of Prostaglandin E1 infusion in patients with fulminant hepatic failure, this trial was conducted to determine whether PGE1 would be of value in primary nonfunction. We have encountered 16 cases of primary nonfunction in 94 liver transplants, an incidence of 17%. Initially in the program, there were 6 occurrences of nonfunction that did not receive PGE1; 3 underwent retransplantation (2 survivors), 2 died awaiting another liver, and in one recovery of hepatocellular function occurred with supportive care but the patient died of cytomegalovirus infection. Ten patients received PGE1 within 4-34 hr of transplantation. Within 12 hr of treatment, 8 patients responded with a significant fall in the AST (129 U/hr) whereas, in the untreated group, the AST continued to rise (267 +/- 102 U/hr) at the same rate as prediagnosis (337 +/- 95 U/hr). At the conclusion of the infusion (4-7 days) in the 8 responders, there were significant decreases in AST (4386 +/- 546 U/L to 102 +/- 21 U/L), prothrombin time (22 +/- 2 to 12 +/- .4 sec) and partial thromboplastin time (45 +/- 3-29 +/- 4 sec), and significant increases in coagulation factor V (26 +/- 8 to 95 +/- 12%) and factor VII (10 +/- 5 to 61 +/- 4%). No serious side effects occurred, although 2 patients developed diarrhea, and abdominal cramps. Two patients treated with PGE1 were retransplanted at 10-36 hr and were considered nonresponders. Graft survival was 80% in the PGE1-treated group and 17% in the untreated group (P less than 0.05) and patient survival was 90% and 33%, respectively. This study suggests a potential benefit of PGE1 in the treatment of primary nonfunction.

MARKERS OF ALLOGRAFT VIABILITY IN THE RAT RELATIONSHIP BETWEEN TRANSPLANTATION VIABILITY AND LIVER FUNCTION IN THE ISOLATED PERFUSED LIVER

The relationship between transplant viability and liver function has been examined. Wistar rat livers were preserved at 4°C for increasing intervals and then transplanted into Wistar rat recipients. Two critical times were identified, the longest preservation period with 100% transplantation success (4 hr) and the shortest preservation period with 100% transplant failure (8 hr). The comparable critical times were also identified in livers preserved at 37°C (1 hr and 2 hr). Liver functions were studied by the isolated perfused liver technique in other rat livers stored at 4°C or 37°C for the critical times. Two liver function tests, AST and LDH concentration in perfusate, discriminated between viable and nonviable livers across as well as within preservation groups. AST gave the best separation between viable and nonviable livers. Some functions such as ALT concentration in perfusate separated viable from non viable allografts only within preservation groups. Other liver functions were more sensitive to preservation temperature than allograft viability. Oxygen consumption after cold preservation for either critical time was about twice control levels. Urea production was far below control levels in warm-preserved livers but almost normal in cold-preserved livers. Our results indicate that AST release into perfusate can be used as a screening technique to optimize preservation methods, reserving transplantation for confirming the most promising results.

Effect of mucous glycoprotein on nucleation time of human bile

The purpose of this study was to determine whether mucous glycoprotein is the nucleating factor responsible for the rapid in vitro nucleation time of gallbladder bile from persons with cholesterol gallstones. Ultracentrifugation and ultrafiltration of abnormal bile removed all detectable mucous glycoprotein, yet bile that had been filtered exhibited as rapid a nucleation time as unfiltered bile. When abnormal bile was heated to 95 degrees C for 60 min, nucleation time was significantly prolonged. Rapid nucleation time could be restored to heated abnormal bile by addition of small volumes of unheated bile. Purified human mucous glycoprotein accelerated nucleation time of human bile, but mucous glycoprotein from control patients was as effective as that from gallstone patients. There was a direct relationship between mucous glycoprotein concentration and effect on nucleation time. Mucous glycoprotein may be important in the early stages of stone formation, but it is probably not the agent responsible for the sharp discrimination between control bile and gallbladder bile from patients with cholesterol stones found in the in vitro nucleation time test. The markedly prolonged nucleation time of heated abnormal bile is preliminary evidence that the nucleating factor may be a heat-labile protein other than mucous glycoprotein.

Biliary Proteins and the Nucleation Defect in Cholesterol Cholelithiasis

A study was performed to determine whether differences in gallbladder proteins might be present in patients with rapidly nucleating bile. Gallbladder and hepatic bile protein concentrations were measured using a fluorometric assay. The method was validated by an independent technique, i.e., hydrolysis and amino acid analysis. Persons with cholesterol gallstones had significantly higher gallbladder bile protein concentrations than patients without gallbladder disease or patients with pigment stones. The protein concentration correlated with the in vitro nucleation time in the cholesterol stone group. Gallbladder bile proteins were also purified by chromatography and gradient ultracentrifugation. Proteins from patients with cholesterol gallstones accelerated the nucleation time of control bile, whereas protein from controls had little effect. Hepatic bile protein concentrations were similar in persons with and without cholesterol gallstones. The gallbladder-to-hepatic bile ratios of a variety of solutes were examined. The ratio for protein in the cholesterol gallstone group can be explained straightforwardly by water reabsorption in the gallbladder, whereas the very low ratio in patients without cholesterol gallstones suggests that their gallbladders reduce protein mass by a process such as protein absorption or degradation during water absorption in the gallbladder.

The simultaneous determination of oxidized and reduced glutathiones in liver tissue by ion pairing reverse phase high performance liquid chromatography with a coulometric electrochemical detector

The oxidized (GSSG) and reduced (GSH) forms of glutathione were quantified simultaneously in rat liver tissues by ion-pairing reverse phase high pressure liquid chromatography coupled to a coulometric electrochemical detector (HPLC-EC). Other biological thiols namely cysteine, cystathionine, homocysteine and methionine were shown to be well resolved from the glutathiones. Standard curves for GSH and GSSG were linear over the range of glutathione concentrations found in biological tissues with a correlation coefficient (r) greater than 0.999. Rat liver tissue content of GSH (16.96 +/- 4.29, n = 5) and GSSG (0.467 +/- 0.188 n = 5) found in this study are similar to results reported by other investigators. This method is also applicable to determine glutathione levels in rat bile samples. The advantages and disadvantages of employing coulometric over amperometric, as well as precautions in establishing HPLC-EC for the detection of oxidized and reduced glutathiones is discussed.

Adenine nucleotide tissue concentrations and liver allograft viability after cold preservation and warm ischemia

The relation between adenine nucleotide liver concentrations and the viability of liver allografts after cold preservation and warm ischemia was studied. A rat model was used with storage times defined in terms of allograft viability. Livers were excised and stored for 4 hr at 4°C or 1 hr at 37°C (viable if transplanted) or for 8 hr at 4°C or 2 hr at 37°C (not viable if transplanted) in a solution containing 0.9% NaCl and 2 mM CaCl2. Adenine nucleotide, malondialdehyde, and glutathione concentrations were measured in liver biopsies at the end of the storage periods and in control livers. During cold preservation, ATP concentrations decline, but degradation is largely halted at AMP, and this is independent of the length of storage or viability of the allograft. Graft failure is not due to lack of availability of intramitochondrial substrate (AMP) for rephosphorylation to

Prediction of the outcome of transplantation in man by platelet adherence in donor liver allografts. Evidence of the importance of prepreservation injury.

We examined platelet adhesion in thirty human donor livers to determine if the degree of platelet adhesion could predict outcome of transplantation. Wedge liver biopsies were taken at the start of the donor operation (biopsy 1) and 1 hr after reperfusion in the recipient (biopsy 2). Biopsies were stained with a monoclonal antibody against platelet glycoprotein Ib and graded for platelet adhesion. Hematoxylin and eosin-stained sections were examined for polymorphonuclear leukocyte adhesion and necrosis. Platelet adhesion was much more frequent and extensive than expected in biopsy 1. Nine of 30 biopsies showed moderate or high-grade platelet adhesion. Thus in this study endothelial cell damage was present in about one-third of donors before the donor operation. The injury was not detectable by routine microscopic or clinical examination or biochemical tests. The degree of platelet adhesion in biopsy 1 predicted development of PMN adhesion and necrosis in biopsy 2 and postoperative transaminase concentrations and prothrombin times in recipients. During preservation and implantation some livers converted from low to either moderate or high grades of platelet adhesion. The grade of platelet adhesion in biopsy 2 predicted postoperative outcome as measured by transaminase and PT levels. Patients whose platelet grade converted to a higher level during preservation and implantation did not do as well as patients who remained at a low adhesion grade. These findings strongly suggest that the degree of platelet adhesion is an important determinant in assessing outcome and may provide a means of measuring the status of liver allografts prior to transplantation.

Nucleation of cholesterol from vesicles isolated from bile of patients with and without cholesterol gallstones

Bile was obtained from patients with and without cholesterol gallstones at surgery. Biliary vesicles were separated from micelles by gel filtration. The cholesterol/phospholipid ratio in vesicles was much higher than in micelles. Cholesterol crystals nucleated from vesicular fractions, but nucleation from the micellar fractions was slow or did not occur at all. Cholesterol nucleated from vesicles obtained from bile of control patients as rapidly (2.4 days +/- 0.7) as from patients with stones (2.4 days +/- 0.9) and there was no difference in the vesicular cholesterol/phospholipid ratio. The effect of alteration of the bile salt environment was studied by changing the concentration of sodium cholate in the eluting buffer. At low concentrations (5 mM) only vesicles were eluted from the column. These vesicles had a relatively low cholesterol/phospholipid ratio and cholesterol nucleated slowly from these vesicles. At higher concentrations the proportion of micelles increased. The proportion of vesicles decreased progressively but their cholesterol/phospholipid ratio increased and the nucleation time fell. These studies demonstrate that cholesterol nucleates from vesicles in the absence of micelles, that control vesicles are not protected by tightly bound antinucleating substances and that exposure of vesicles to micelles strips relatively more phospholipid than cholesterol from the vesicular fraction, resulting in vesicles with higher cholesterol/phospholipid ratios and shorter nucleation times.

Correlation of donor nutritional status with sinusoidal lining cell viability and liver function in the rat

We have demonstrated that the sinusoidal lining cell injury sustained by rat liver allografts during hypothermic storage is a critical determinant of graft viability. The present study was designed to examine the effect of donor nutritional status on hepatic microcirculation and graft function. Rat livers from four nutritional groups (group I, fasted; group II, fed; group III, intraperitoneal glucose; and group IV, fed plus intraperitoneal glucose) were excised and stored for 24 hr in Marshalls isotonic citrate solution. Then the livers were perfused under anoxic conditions with trypan blue. The percentage of nonviable SLC in each group was 26.7±8.1, 24.9±7.9, 17.6±6.9, and 5.9±1.9 in groups I, II, III, and IV respectively; i.e., there was a significant improvement in SLC viability with nutritional repletion in group IV. Electron microscopy was performed on livers from groups I and IV following 30-hr preservation in University of Wisconsin solution and after 16-hr preservation in Marshalls isotonic citrate solution. Biopsies were taken at the end of storage and after 1 hr of reperfusion at 37°C. At the end of preservation group IV livers contained glycogen and had much more normal liver ultrastructure than group I livers. After reperfusion there was partial recovery of normal SLC morphology in both groups and depletion of glycogen in group IV. Liver function was studied on the isolated perfused rat liver system at 37°C following 30-hr storage in UW solution. Transaminase release into the perfusate was significantly lower in nutritionally repleted livers than in livers from fasted animals. A significant reduction in perfusate platelet count occurred only in livers from fasted animals. The results show that nutritional repletion can reduce the injury of cold preservation to both hepatocytes and endothelial cells and improve liver function in the postpreservation period.