The aim of this study was to develop a porcine model of post-operative liver failure (POLF) that could accurately reproduce all the neurological and metabolic parameters of the corresponding clinical syndrome that may develop after extensive liver resections.

In our model, we induced POLF by combining extended left hepatectomy and ischemia of the small liver remnant of 150 min duration. Subsequently, the remnant liver parenchyma was reperfused and the animals were closely monitored for 24 h.

Twelve Landrace pigs (weight 25-30 kg) were randomly assigned in two groups; eight of them constituted the experimental group, in which POLF was induced (POLF group, n = 8), whereas the rest of them (n = 4) were included in the control group (sham laparotomy without establishment of POLF). RESULTS (MEANS ± SD): All POLF animals gradually developed neurological and biochemical signs of liver failure including, among many other parameters, elevated intracranial pressure (24.00 ± 4.69 versus 10.17 ± 0.75, P = 0.004) and ammonia levels (633.00 ± 252.21 versus 51.50 ± 9.49, P = 0.004) compared with controls. Histopathologic evaluation of the liver at the end of the experiment demonstrated diffuse coagulative necrosis and severe architectural distortion of the hepatic parenchyma in all POLF animals.

Our surgical technique creates a reproducible porcine model of POLF which can be used to study the pathophysiology and possible therapeutic interventions in this serious complication of extensive hepatectomies.

Although the use of non-heart-beating donors (NHBD) is the oldest type of organ transplantation, the results were and still are disappointing. To consider using a liver from NHBD, it is of importance to assess the graft viability. Our aim was to assess the role of reduced liver glutathione (rGSHL) as a potential predictive marker of liver function before transplantation. Autotransplanted livers were subjected to 0, 60, and 90 minutes of ischemia in 20 pigs. We analyzed systemic cardiocirculatory parameters, bowel ischemia by endotoxin, endotoxin-neutralizing capacity, oxidative stress, hepatic perfusion parameters, liver enzymes, local bowel ischemia, and liver oxidative stress (rGSHL and oxidized glutathione in the liver). Autotransplantation was comparable to donor explantation/recipient transplantation with respect to systemic and hepatic parameters. Liver ischemia for 0, 60, and 90 minutes resulted in survival in 100% (NHBD-0), 71% (NHBD-60), and 57% (NHBD-90) of animals. Of all parameters, only hepatic microperfusion, pHi of the sigmoid colon, and bowel ischemia by endotoxin in the NHBD-90 group showed significant changes compared to NHBD-60 and control animals. Although systemic endotoxin-neutralizing capacity and total glutathione in erythrocytes levels were mainly influenced by cold perfusion, hepatic oxidative stress increased with ischemia time. The cut-off value of 11.5 ng/mmol of rGSHL could distinguish survivors from nonsurvivors, independent of the ischemia time. In conclusion, rGSHL has the potential of becoming an important viability marker, as it could predict survival in autotransplantation NHBD model regardless of the ischemia time. Further investigation to declare reasons for differing rGSHL levels within the liver is required.

Ischemia/reperfusion (I/R) injury is frequently encountered during hepatic surgery. We recently showed that I/R accelerates the outgrowth of pre-established colorectal micrometastases. The aim of this study was to assess the influence of ischemia time, gender, age, and liver steatosis on the accelerated outgrowth of colorectal metastases following I/R.

Five days after tumor cell inoculation, mice were subjected to 20, 30 or 45 min of left lobar I/R. To assess the influence of age, gender, and liver steatosis on I/R-accelerated tumor growth, we compared old with young mice, male with female mice, and mice with healthy livers with mice with steatotic livers. Endpoints were extent of tissue necrosis and tumor growth.

With increasing ischemia times, tissue necrosis and I/R-accelerated tumor growth increased, with a significant stimulatory effect at 30 and 45 min of ischemia. I/R-stimulated outgrowth of micrometastases was further increased by 33% in aged mice and by 42% in steatotic livers and was associated with increased tissue necrosis. In female mice tissue necrosis had decreased by 47% and tumor growth was reduced in both control and clamped liver lobes. The stimulatory effect of I/R on metastasis outgrowth was similar in male and female mice.

I/R-accelerated outgrowth of colorectal micrometastases largely depends on the duration of the ischemic period, with a safe upper limit of 20 min in mice. The stimulatory effects of I/R on tumor growth are exacerbated in aged mice and in steatotic livers.

The number of liver transplant recipients is increasing but donor organ shortages have become more severe. The effect of milrinone, a type 3 phosphodiesterase inhibitor (PDEI), on non-heart-beating donor grafts was evaluated using an orthotopic liver transplantation model in rats.

Type 3 PDEI or normal saline (control group) was given intravenously to the donor animals for 60 min continuously (50 microg kg-1 min-1 ) before 60 min of warm ischaemia followed by cold preservation and subsequent transplantation. Survival, serum chemistry, bile output, histopathological findings and tissue cyclic 3',5'-adenosine monophosphate (cAMP) concentrations were then compared.

Five of seven animals in the PDEI group were alive at 7 days, compared with only one of seven rats in the control group (P < 0.01). Serum levels of alanine aminotransferase 2 and 6 h after reperfusion, and hyaluronic acid levels 6 h after reperfusion, were significantly lower in the PDEI group than in the control group. Bile output from the transplanted graft was significantly greater in the PDEI group than in controls 2 h after reperfusion (P < 0.01). The mean necrotic area 6 h after reperfusion was also reduced in the PDEI-treated grafts (P < 0.01). cAMP levels in liver tissue at the end of both warm and cold ischaemia, and 2 and 6 h after reperfusion, were significantly higher in the PDEI group compared with those in the control group.

Type 3 PDEI attenuated the graft injury caused by warm and cold ischaemia and subsequent reperfusion injury via an increase in intracellular cAMP levels. This treatment may be a novel pharmacological intervention for safe and efficient usage of liver grafts from non-heart-beating donors.

Intermittent hepatic pedicle clamping is thought to cause less hepatic reperfusion injury compared with continuous clamping. The mechanisms underlying this difference are unknown. We examined the relationship between intermittent ischemia/reperfusion and the production of free radicals using electron spin resonance spectrometry.

Alpha-(4-pyridyl 1-oxide)-N-tert-butylnitrone was administered to rats as a spin trap agent. Continuous clamping (15, 30, or 60 minutes) or intermittent clamping (four cycles of 15-minute ischemia and 5 or 15 minutes of reperfusion) of hepatic pedicle was carried out. After reperfusion, free radical production in the liver was measured by an electron spin resonance spectrometer, and the level of hepatic injury was evaluated by measuring liver enzyme.

Longer periods of ischemia increased free radical production after reperfusion. There was no significant increase in free radical production or liver enzymes when the duration of ischemia was 15 minutes. Free radical production and liver damage were significantly less severe in intermittent pedicle clamping than in continuous clamping for 60 minutes, especially when the duration of the reperfusion between four cycles of ischemia was 15 minutes.

These results indicate that intermittent pedicle clamping lessens free radical production when compared with continuous clamping, although many free radicals are produced.

Warm ischemia is known to induce substantial damage to the liver parenchyma. With respect to clinical liver transplantation, the tolerance of the liver to warm ischemia and the preservation of these organs have not been studied in detail. In isolated reperfused pig livers we proceeded according to the following concept: Livers were subjected to 1 or 3 h of warm ischemia. Subsequently, these organs were preserved by either normothermic perfusion or cold storage (histidine-tryptophan-alpha-ketoglutarate, HTK) for 3 h each. After storage, liver function was assessed in a reperfusion circuit for another 3 h. Parameters under evaluation were bile flow, perfusion flow, oxygen consumption, enzyme release into the perfusate (creatine kinase, glutamic oxaloacetic transaminase (GOT), lactic dehydrogenase, and glutamic pyruvic transaminase), and histomorphology. Damage to the liver was lowest after warm ischemia of 1 h. The results after cold storage were superior to those after normothermic perfusion (GOT: 3.2 +/- 0.3 and 2.6 +/- 0.2 U/g liver; cumulative bile production: 14.7 +/- 2.1 and 9.4 +/- 1 ml, respectively; P < 0.05). In contrast, we found substantial damage at the end of reperfusion in livers undergoing 3 h of warm ischemia under both preservation techniques with severe hepatocellular pyknoses and essentially altered nonparenchymal cells. The results suggest that pig livers undergoing 1 h of warm ischemia and cold storage for 3 h with HTK solution may lead to functioning after transplantation.

To investigate the tolerance to warm ischemia of liver grafts from non-heart-beating donors, porcine orthotopic liver transplantation was performed using grafts obtained at various periods after cardiac arrest. Graft viability was investigated in relation to changes in hepatic adenine nucleotide metabolism. In donors, livers were divided into four groups according to warm ischemic time after cardiac arrest (group 1: 0 min, n=3; group 2: 30 min, n=3; group 3: 60 min, n=5; group 4: 90 min, n=4). Thereafter, the livers were flushed and preserved for 4 hr using 4 degrees C Euro-Collins solution. After surgery, all of the recipients in groups 1, 2, and 3 survived more than 4 days, except for one pig in group 3 that died of bleeding from an arterial catheter on day 2. By contrast, all of the recipients in group 4 died within 12 hr. The serum glutamic oxaloacetic transaminase concentration at 4 hr after reperfusion of the graft was significantly higher in group 4 (mean+/-SE, 2563+/-556 IU/L) than in groups 1, 2, and 3 (298+/-29 IU/L, 1226+/-222 IU/L, and 1181+/-174 IU/L, respectively). The adenylate energy charge of the liver graft recovered at 1 hr after reperfusion of the graft to 0.852+/-0.013, 0.845+/-0.003, and 0.842+/-0.003 in groups 1, 2, and 3, respectively. The recovery was significantly suppressed in group 4 (0.796+/-0.011). The hepatic adenosine triphosphate concentration also was significantly lower in group 4 compared with the other groups. The present study suggests that liver allografts can be used from non-heart-beating donors subjected to warm ischemia for less than 60 min. Postoperative survival is associated with prompt recovery of the adenylate energy charge of the liver graft.

Hepatic resection with total vascular isolation has been reported to reduce hemorrhage. Addition of supraceliac aortic clamping putatively avoids hemodynamic instability, but may increase morbidity.

This technique was used in 99 major liver resections utilizing scalpel division and suture hemostasis.

Livers were normal in 86 patients, cirrhotic with no portal hypertension in 5, and cirrhotic with portal hypertension in 8. There was 1 death in 91 patients with no portal hypertension due to hepatic failure or bleeding esophageal varices. There were 59 hemihepatectomies and 40 segmentectomies. Median operating time was 145 and 110 minutes, respectively, and mean transfused blood was 4 and 0 units, respectively, with minimal morbidity.

Use of total hepatic vascular isolation with routine supraceliac aortic clamping is a safe and expedient method of hepatic resection that limits blood loss and maintains hemodynamic stability, but does not increase morbidity. However, the presence of portal hypertension precludes safe resection.

We attempted to investigate the process of ischemia-induced disturbances in the rat liver, employing the electrical bio-impedance technique. The electrical bio-impedance was measured continuously over 6 h by the 4-electrode method, at various incubation temperatures, in six liver samples extracted from male Wistar rats. The electrical properties of biological tissues can be expressed in terms of three parameters: extracellular resistance (Re), intracellular resistance (Ri) and cell membrane capacitance (Cm). These three parameters were calculated from the measured values of the electrical impedance by the curve-fitting technique, using a computer program. The Re value increased rapidly after the rat livers were extracted, and then decreased slowly. The Re value reached a peak after about 13 min at 36 degrees C, and then decreased slowly, becoming constant after 3 h. There was a negative correlation between the Tmax of Re (the time when Re reached a maximum) and the incubation temperature (R = -0.973, P < 0.001). The Ri value decreased once in the early stage after extraction, followed by almost no change and then an increase after 4 h at 36 degrees C. The Cm showed a similar pattern of change to the Re value, and a negative correlation was also found between the Tmax of Cm and the incubation temperature (R = -0.969, P < 0.001). The increase in the Re and Cm values, and the decrease in the Ri value for quite long periods after the blood flow has stopped, suggest an increase in the resistance of extracellular fluid due to a decrease in its volume, an increase in cell membrane capacitance due to cell swelling, and a decrease in cellular fluid resistance due to an increase in its volume. The time when the Cm value decreases rapidly after an initial gradual decrease after the peak corresponds well with the time when the Ri value begins to increase, from which it is estimated that cell lysis proceeds and that the flow of extracellular fluid into the cell begins at this time. The findings of this study suggest the possibility of estimating the changes in liver tissue or the tissue structure due to ischemia.

Initial poor function and primary nonfunction are important problems in clinical transplantation. The incidence of primary nonfunction is about 6% and that of initial poor function is about 15%. Grafts with initial poor function have a higher graft failure rate in the first 3 mo after transplantation. Severe steatosis and cold preservation in University of Wisconsin solution for over 30 hr will alone cause primary nonfunction. However, primary nonfunction is probably most often caused by the presence of multiple relative risk factors. The major donor-relative risk factors are moderate steatosis, cold preservation over 12 hr and donor age over 50 yr, whereas retransplantation, high (United Network of Organ Sharing class 4) medical status and kidney failure are recipient relative risk factors. The most important perioperative risk factor is warm ischemia time. Rates of primary nonfunction and initial poor function might be reduced by avoidance of combinations of risk factors. Several tests have been developed to predict primary nonfunction and initial poor function, but none is yet clinically efficient.

To evaluate the role of thromboxane A2 (TXA2) in ischemic liver injury, the serum changes in thromboxane B2 (TXB2) and 6-keto-prostaglandin F1 alpha (6-K-PGF1 alpha) following warm ischemia of the total canine liver were examined, and the protective effect of a TXA2 synthetase inhibitor was assessed. Total liver ischemia was performed for 60 min on two groups of dogs: a control group, in which ischemia alone was performed, and an OKY-046 group, which received a TXA2 synthetase inhibitor. A temporary active portacaval shunt was used to eliminate the effects of splanchnic venous stasis during clamping of the hepatic pedicle. Postoperative changes in liver function, assessed by the transaminase enzyme levels, and in prostaglandins were recorded and the histologic liver findings of both groups 1 week after ischemia were compared. The levels of 6-K-PGF1 alpha increased after reperfusion in both groups, while those of TXB2 increased in the control group but maintained low levels in the OKY-046 group. Liver function was better and histologic changes less marked in the OKY-046 group than in the control group, suggesting the important role of TXA2 in ischemic liver injury and the usefulness of a TXA2 synthetase inhibitor for protecting the liver against ischemic injury.

The tolerance of the liver to ischaemia during intermittent clamping of the hepatic pedicle was compared with that during a continuous Pringle manoeuvre. Three groups of rats undergoing total durations of clamping of 60, 90 and 120 min were studied. A temporary peroperative portacaval shunt was used to exclude the effects of splanchnic venous stasis and allow independent study of the effects of hepatic ischaemia. In each group, three methods of portal clamping were evaluated: a continuous Pringle manoeuvre (n = 10), a 30-min intermittent clamping (n = 10) and a 15-min intermittent clamping (n = 10). The clamp release time between the periods of liver ischaemia was 5 min. Survival at day 7 and postoperative changes in liver function (transaminase enzymes, bilirubin, bromsulphthalein elimination, liver adenosine 5'-triphosphate levels) were recorded. Intermittent clamping of the pedicle was tolerated significantly better than continuous clamping. This method optimizes the ability of the liver to tolerate extended periods of ischaemia. For a given duration of ischaemia, no additional improvement could be produced by shortening the intermittent clamping period from 30 to 15 min. These data suggest that, when the Pringle manoeuvre is used, it should be applied intermittently rather than continuously.

The influence of glycogen content, temperature, and Euro Collins (EC) solution on membrane potential (Vm) and intracellular sodium activity (aNai) were measured in cells of superfused porcine liver slices by means of double-barrelled ion-sensitive microelectrodes. Vm was -26.1mV in fasted pigs and -20.6mV after glucose feeding, when measured in HEPES-buffered solution (P less than 0.0001). aNai was not measurably affected by glucose feeding. During superfusion with Tyrode solution, lowering the temperature from 35.5 degrees C to 15.5 degrees C led to a fast Vm decrease of roughly 2mV followed by an increase of 1-3mV. At the same time, aNai increased from 12.8 to 18.2mM within 10 min. Superfusion with EC solution for 10 min caused comparable changes in fed and fasted pigs. Vm depolarized at either temperature by about 16mV. At 35.5 degrees C the initial aNai of 17.5mM was roughly halved, whereas at 15.5 degrees C it decreased from 21.0 to 14.3mM. The results suggest that the nutritional state markedly affects the electric properties of liver. However, the effect on membrane potential of high-potassium organ-protective solutions seems to be distinctly more pronounced. Moreover, cellular Na+ activity decreases in consequence of an extracellular Na+ reduction with protective solutions, which might be balanced to some extent by a simultaneous temperature decrease.

Over a 10-year period from 1980, 46 patients with liver injuries were referred after primary admission to other hospitals. Previous surgery had been performed in 40 cases and the predominant reasons for referral were uncontrollable bleeding and postoperative sepsis with biliary leakage. Of 30 such cases, 19 were treated by liver resection; all were of a limited nature and no major hepatectomies were performed. Only two deaths occurred in these patients and this low mortality rate supports our conservative approach to liver trauma. Other reasons for referral were late biliary stenosis, complicated penetrating injury and intrahepatic haematoma. Three cases were referred with postoperative hepatic failure; two responded to resection of infected necrotic tissue and liver transplantation was attempted in the third. Injuries to liver segments 6 and 7 were those most frequently referred for assistance with bleeding, and all patients were safely transferred after intra-abdominal packing. This injury is particularly suitable for resection by segmentectomy rather than a formal hepatectomy, which has been associated with a high mortality rate in trauma cases. Further patients with intractable injuries might be salvaged by liver transplantation.

Platelet activating factor (PAF) is a newly discovered inflammatory chemical mediator, which was reported to play a pivotal role in various types of shock. There is also a great possibility that PAF plays an important role in the shock caused by hepatic inflow occlusion. In the present study, the effect of CV6209, a PAF antagonist, on the shock caused by the occlusion was investigated. Intravenous 3 micrograms/kg of PAF caused hypotension in Wistar rats (n=6), and pretreatment with intravenous 3 mg/kg of CV6209 significantly (p less than 0.01) prevented the hypotension (n=6). Forty-five minutes of hepatic inflow occlusion caused hypotension in rats during the occlusion period, and the hypotension continued even after restoration of blood flow in control group (pretreated with saline i.v. only, n=5). In contrast, this hypotension was significantly (p less than 0.01) reversed in PAF antagonist group (pretreated with 3 mg/kg of CV6209 i.v., n=5). In sham-operated rats (n=6), arterial pressure remained unchanged and not hypotensive during the monitoring period. The survival rate of rats 90 minutes after declamp was 30% in control group (n=20), and that was significantly (p less than 0.05) improved to be 65% in PAF antagonist group (n=20). In conclusion, PAF plays an important role in the shock and death caused by temporary hepatic inflow occlusion, and a PAF antagonist could be a therapeutic drug against temporary hepatic inflow occlusion.

The effect of 60 min of normothermic liver ischemia on cellular levels of adenine nucleotides (ATP, ADP, and AMP), energy charge (EC), and bile excretion was studied. Following ischemia the concentration of ATP was reduced to 12% of preischemic and control values within the first 10 min and remained low during the remaining ischemic period. EC values were also low. During 120 min of reperfusion, ATP increased to 34% of the ATP level found in the control group. EC values increased immediately to reach values not significantly different from those of control animals. Bile flow was nonexistent during ischemia and increased during reperfusion. The increase paralleled those of ATP and EC. Bile flow seems to reflect the degree of liver ischemia and may be used as a functional parameter.