Non-heart-beating donor porcine livers: The adverse effect of cooling

Authors


Abstract

Normothermic preservation has been shown to be advantageous in an experimental model of preservation of non-heart-beating donor (NHBD) livers, which have undergone significant warm ischemic injury. The logistics of clinical organ retrieval might dictate a period of cold preservation prior to warm perfusion. We have investigated the effects of a brief period of cold preservation on NHBD livers prior to normothermic preservation. Porcine livers were subjected to 60 minutes of warm ischaemia and then assigned to following groups: Group W (n = 5), normothermic preservation for 24 hours; and Group C (n = 6), cold preservation in University of Wisconsin solution for 1 hour followed by normothermic preservation for 23 hours (total preservation time, 24 hours). Synthetic function (bile production and factor V production) and cellular damage were compared on the ex vivo circuit during preservation. There was no significant difference in the synthetic function of the livers (bile production and factor V production). Markers of hepatocellular damage (alanine aminotransferase and aspartate aminotransferase release), sinusoidal endothelial cell dysfunction (hyaluronic acid), and Kupffer cell injury (β-galactosidase) were significantly higher in Group C. The histology of the livers at the end of perfusion was similar. In conclusion, a brief-period cold preservation prior to normothermic perfusion maintains the synthetic function and metabolic activity but results in significant hepatocellular damage, sinusoidal endothelial cell dysfunction, and Kupffer cell injury. Transplant studies are required to establish whether livers treated in this way are viable for transplantation. (Liver Transpl 2005;11:35–38.)

Liver transplantation is an established treatment for end stage liver failure. The donor organ shortage is one of the principal causes of increasing waiting lists and death of patients on the waiting list. Non-heart-beating donor (NHBD) livers are a potential means to expand the donor pool. Controlled NHBD (Maastricht Category 3, death anticipated) livers can be used for transplantation.1 Uncontrolled NHBD livers (Maastricht Categories 1 and 2, death not anticipated) are not generally utilized for liver transplantation due to a high rate of primary nonfunction.2 This contrasts with the situation in renal transplantation in which there is increasing use of organs from uncontrolled NHBD with good results.3

A combination of prolonged warm ischemic insult with conventional cold preservation leads to a poor outcome in uncontrolled NHBD liver transplantation. Warm ex vivo perfusion has been used to resuscitate livers that have suffered 60 minutes of warm ischemia in a porcine transplant model4 and porcine kidneys that have suffered 120 minutes of warm ischemia.5 Warm perfusion has the added advantage of allowing viability assessment of the organs while on the circuit before transplantation6. The logistics of clinical organ retrieval might dictate a period of cold preservation prior to warm perfusion. We recently demonstrated that normothermic perfusion failed to resuscitate porcine livers after 60 minutes of warm iscaemia and 4 hours of cold preservation of NHBD livers.7 We next wanted to test the impact of a brief period of cold preservation on NHBD livers, which were subsequently preserved by normothermic perfusion.

Abbreviations

NHBD, Non-heart-beating donor; HA, hyaluronic acid.

Materials and Methods

White Landrace pigs (40 kg) were used. All animals were treated in accordance with the animal protection act 1986 of the United Kingdom. Following mobilization of the liver, heparin was given, the pigs were exsanguinated and cardiac arrest induced. During a period of 60 minutes of circulatory arrest, cannulas were placed into the vena cava, portal vein, and hepatic artery.

Experimental Design

Porcine livers were assigned to following groups: Group W (n = 5), normothermic preservation for 24 hours; and Group C (n = 6), cold preservation in University of Wisconsin solution for 1 hour, followed by normothermic preservation for 23 hours (total preservation time, 24 hours).

Synthetic liver function and cellular damage were compared between the 2 groups on the ex vivo normothermic circuit during the period of warm perfusion. Perfusion was performed according to the methodology previously described.8 Briefly, this uses cardiopulmonary bypass equipment to deliver warm, autologous, oxygenated blood to the liver through both the hepatic artery and the portal vein at normal body temperature (39°C for a pig) and physiological pressures and flow parameters.

Parameters Assessed on the Circuit

Synthetic liver function was measured by bile and factor V production. The bile duct was cannulated and the hourly bile production measured. For the factor V assay, dilutions of standard and test plasma were mixed with substrate plasma (deficient in factor V). A modified prothrombin time was then performed, and the ability of the test and standard plasmas to correct the prolonged prothrombin time of the substrate plasma was compared. Factor V levels were measured at 0, 4, 8, 12, and 20 hours of perfusion.

For biochemistry, heparinized blood was immediately centrifuged. Plasma was stored at −70°C until analyzed. Aspartate aminotransferase and alanine aminotransferase were measured as markers of hepatocellular injury at 0, 1, 2, 4, 6, 8, 12, 16, and 20 hours of perfusion on plasma using an automated analyzer (Abbott Aeroset, Maidenhead, UK). For evaluating sinusoidal endothelial cell injury, hyaluronic acid (HA) in the perfusate was measured at 0, 1, 2, 4, 6, 8, 12, and 20 hours of perfusion by a sandwich enzyme linked immunosorbent assay technique (Corgenix [UK] Ltd., Peterborough, UK). β- galactosidase, a lysosomal enzyme mainly originating from Kupffer cells and a marker of Kupffer cell activation,9 was measured as previously described by a microtiter plate fluorometric method.10

At the end of perfusion, each liver was sectioned and multiple random samples were obtained and fixed in formalin. At least 5 sections were examined by 2 blinded observers. The sections were scored using a semiobjective scale as previously published.7 The sections were scored for necrosis, architectural destruction, apoptosis, sinusoidal congestion, sinusoidal dilatation, and hepatocellular vacuolization. Statistical analysis was performed using SPSS software (SPSS Inc., Chicago, IL). Results are expressed as mean ± SEM. The Mann-Whitney U test was performed to compare the data at each time point. P < .05 was considered significant.

Results

Function of the Livers

Bile Production

Both groups had similar bile production. Group W had a mean bile production of 10 ± 4 mL/hour at 2 hours and remained steady above 9 mL/hour throughout the rest of perfusion. Group C produced bile at a mean of 5 ± 2 mL/hour at 2 hours, which increased to a mean of over 7 mL/hour at 4 hours and remained at this level thereafter for the rest of perfusion. This difference was not statistically significant (Fig. 1A).

Figure 1.

Function of the livers. (A) Bile production. Each point represents mean ± SEM for the experiments. Differences not significant (P = NS). (B) Factor V levels. Expressed as percentage of starting value. Each point represents mean ± SEM for the experiments. Differences not significant (P = NS).

Factor V

The factor V levels in the perfusate before starting the perfusion were 337 ± 13U/dL in Group W and 383 ± 95U/dL in Group C. After 20 hours of perfusion, the factor V levels were 307 ± 16U/dL in Group W versus 291 ± 66U/dL in Group C, and this difference was not significant (Fig. 1B).

Cellular Injury

Hepatocellular Injury

There was a greater release of enzymes into the perfusate from Group C compared to Group W. The mean alanine aminotransferase levels by 20 hours of perfusion were 68 ± 19 IU/L in Group W and 166 ± 36 IU/L in Group C. The rise in aspartate aminotransferase mirrored that of alanine aminotransferase, with mean levels at 20 hours of perfusion of 671 ± 408 IU/L in Group W and 1826 ± 303 IU/L in Group C. The differences in the levels were significant (Fig. 2A and 2B).

Figure 2.

Hepatocellular injury. (A) Alanine aminotransferase levels in the perfusate. Each point represents mean ± SEM for the experiments. X represents statistically significant differences (P < .05). Normal range, 31–58 U/L (B) Aspartate aminotransferase levels in the perfusate. Each point represents mean ± SEM for the experiments. X represents statistically significant differences (P < .05). Normal range 14–43 U/L.

Sinusoidal Endothelial Cells

HA was rapidly cleared from the perfusate, with levels falling to a mean of 10 ± .4 ng/mL in Group W and 19 ± 4 ng/mL in Group C after 1 hour of perfusion. In Group W the levels remained low, with a small rise to 44 ± 30 ng/mL by the end of perfusion. In contrast, in Group C there was a large increase in HA levels after 4 hours, reaching 2483 ± 1436 ng/mL by the end of perfusion. In Group C the rise in HA was not uniform in any of the perfusions. The levels increased in 3 of the 6 livers but remained low in 3 of the 6 livers. The difference in the levels was significant at 2, 6, and 8 hours (Fig. 3A).

Figure 3.

Sinusoidal endothelial cell injury. (A) Logarithmic scale of HA levels in the perfusate. Each point represents mean ± SEM for the experiments. X represents statistically significant differences (P < .05). (B) β-galactosidase levels in the perfusate. Each point represents mean ± SEM for the experiments. X represents statistically significant differences (P < .05).

β-Galactosidase

β-galactosidase peaked at 83 ± 13 U/mL at 1 hour in Group W and progressively declined through the rest of the perfusion. In contrast, in Group C it peaked at 1061 ± 93U/mL at 4 hours and then progressively declined. These differences were significant (Fig. 3B).

Histology

On histological analysis, both Group W and Group C livers showed preserved tissue architecture with minimal necrosis. The mean score was 4.6 ± 1.6 for Group W and 6.8 ± 1.7 for Group C and this difference was not statistically significant (P = .3).

Discussion

The resuscitation of NHBD organs is an issue of increasing relevance with the rapid development of the use of these organs. We have recently shown that a combination of 60 minutes of warm ischemia and 4 hours of cold preservation renders pig livers nonviable.7 Other investigators have shown that a period of warm perfusion can restore cellular energy stores and resuscitate organs after warm ischemia. This has been shown in porcine livers and kidneys.4, 5

In an experimental setting, it is possible to use warm perfusion with minimal exposure of the organ to cooling. Clinically, the logistics of warm preservation are such that a brief period of cold preservation immediately after retrieval would greatly simplify the procedure. We have shown that 4 hours of cold preservation is excessive and are hence investigating whether a shorter period of 1 hour is compatible with organ viability.

The addition of 4 hours of cold preservation to 1 hour warm ischemia results in severe cellular injury and loss of function.4, 7 We have now shown that 1 hour of cold preservation leads to a significant graft injury (hepatocellular injury, Kupffer cell activation, and sinusoidal endothelial cell dysfunction), but it is less severe than that seen after 4 hours of cold preservation, and this is not associated with major graft dysfunction as evident by steady bile production and factor V production throughout the perfusion period.

It has been previously demonstrated that cold ischemia causes greater damage to sinusoidal endothelial cell compared to hepatocytes, whereas warm ischemia damages the hepatocytes more than the sinusoidal endothelial cell.11 This study provides further evidence that warm ischemia alone results in minimal hepatocellular injury and Kupffer cell activation with no significant injury to sinusoidal endothelial cells. In Group C the injury to the livers on the circuit was not uniform. Three of the 6 livers had an elevated HA, whereas 3 of the 6 livers had low HA by the end of the perfusion, despite the same uniform methodology used for all the perfusions. These 3 livers with elevated HA also showed greater hemorrhage and necrosis on histology. The liver architecture was preserved in both groups.

Transplant experiments need to be performed to demonstrate viability of these livers. For various reasons, the ex vivo experiments cannot be directly applied to in vivo model. In vivo, “no reflow” following microvascular thrombosis contributes to graft injury.12 The anticoagulation used for ex vivo perfusion decreases the risk of microvascular thrombosis. In addition, being an isolated ex vivo circuit there is no opportunity to recruit more leukocytes and platelets. In the interest of producing a consistent, reproducible preclinical model, we used donor anticoagulation. Clearly, in the clinical NHBD situation, this may not be possible. Heparin can be given in controlled NHBD in countries such as the United States where legislation allows but this will definitely not be relevant in uncontrolled NHBD. One option in nonheparinized donors is to use streptokinase as pre-flush,13 but its expense prohibited us from using it in this experimental model.

In conclusion, we have demonstrated that although the introduction of a brief period of cold preservation causes significant cellular injury, the function of ischemically damaged porcine liver is largely preserved.

Ancillary