SEARCH

SEARCH BY CITATION

Keywords:

  • Electron leakage;
  • HOPE;
  • hypothermic oxygenated perfusion;
  • liver transplantation;
  • marginal liver;
  • mitochondria;
  • normothermic oxygenated perfusion;
  • reactive oxygen species;
  • sinusoidal endothelial cell

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Scientific Methods to Test Liver Perfusion
  5. Machine Perfusion Systems
  6. Effects of Machine Perfusion on Marginal Grafts
  7. Ideal Machine for Clinical Use
  8. Future Perspectives
  9. References

Due to the critical shortage of deceased donor grafts, clinicians are continually expanding the criteria for an acceptable liver donor to meet the waiting list demands. However, the reduced ischemic tolerance of those extended criteria grafts jeopardizes organ viability during cold storage. Machine perfusion has been developed to limit ischemic liver damage but despite its proven biochemical benefit, machine liver perfusion is not yet considered clinically due to its low practicability. In this review, we summarize our understanding of the role of machine perfusion in marginal liver preservation. The goal is to highlight advantages or disadvantages of current perfusion techniques and to explain the underlying mechanisms. We provide evidence for the need of a liver perfusion performance shortly before implantation, and point out promising designs.


Abbreviations: 
OLT

orthotopic liver transplantation

DCD

donation after cardiac death

IPL

isolated perfused liver

ICAM-1

intercellular adhesion molecule 1

vWF

von Willebrand factor

ROS

reactive oxygen species

HOPE

hypothermic oxygenated perfusion

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Scientific Methods to Test Liver Perfusion
  5. Machine Perfusion Systems
  6. Effects of Machine Perfusion on Marginal Grafts
  7. Ideal Machine for Clinical Use
  8. Future Perspectives
  9. References

The persisting gap between the number of patients in need of a liver transplant and the number of available grafts has led to an enhanced interest in using ‘extended criteria’, previously called marginal, donor livers for orthotopic liver transplantation (OLT). A number of factors have been identified as risk factor for poorer outcome including hepatic steatosis (1), donation after cardiac death (DCD) (2), also called nonheart-beating donor, and prolonged cold ischemia (e.g. >12 h) (3). The combination of these factors in a donor puts a recipient at higher risk of postoperative complications. The incentive to use these organs is, however, high; for example including DCD donors may increase the pool of available organs by up to 20% (4).

Obviously, new strategies must be developed for the safe use of many of those grafts. A number of new and promising data regarding machine liver perfusion has emerged over the past decade with currently some attempts to bring this technology to patients. In this review, we will therefore focus on machine perfusion, first looking at perfusion system using normothermic perfusates and next those using cold perfusate. Finally, we attempt to present the ‘ideal’ perfusion machine for clinical use considering effectiveness and practical issues in providing suitable organs for OLT. For more detailed information on the historical development and technical aspects of machine perfusion, we would refer to other reviews (5–8).

Scientific Methods to Test Liver Perfusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Scientific Methods to Test Liver Perfusion
  5. Machine Perfusion Systems
  6. Effects of Machine Perfusion on Marginal Grafts
  7. Ideal Machine for Clinical Use
  8. Future Perspectives
  9. References

No specific general tests are available that precisely determine total hepatic injury during or after OLT. One of the models most often used in experimental surgery to analyze hepatic preservation/reperfusion injury is the isolated perfused liver (IPL). This approach offers the possibility to continuously detect changes in perfusate biochemistry or liver tissue during perfusion without the influence of extrahepatic factors (9). The application of blocking or stimulating substances gives the opportunity to test specific reactions. While injury of hepatocytes and endothelial cells can be analyzed during normothermic oxygenated perfusion, clear prognostic markers of bile duct viability are currently not available.

In addition, assessment of viability of the IPL appears limited and strongly depends on the perfusion conditions used. It should be recognized that an IPL is considered as a dying organ with a life span of approximately 4–6 h under usual conditions (9). Therefore, it appears mandatory to validate IPL systems in control experiments (nonischemic, nonpreserved livers) for physiological function and minor cell injury.

In this review, we focus on results based on transplant models. We also consider data on the IPL if control experiments are provided that demonstrate no relevant increase in cytosolic enzyme release after several hours of reperfusion as well as sufficient bile flow.

Machine Perfusion Systems

  1. Top of page
  2. Abstract
  3. Introduction
  4. Scientific Methods to Test Liver Perfusion
  5. Machine Perfusion Systems
  6. Effects of Machine Perfusion on Marginal Grafts
  7. Ideal Machine for Clinical Use
  8. Future Perspectives
  9. References

Normothermic oxygenated perfusion

Normothermic liver perfusion already has been studied in the early developmental stage of liver preservation. The major challenge, that could not be overcome, was the need for a carrier providing sufficient oxygen to the liver without vascular damage and infection. Nonetheless, some investigators have recently shown the feasibility of normothermic machine liver perfusion using autologous blood and adequate nutrients for periods of up to 72 h (10).

Perfusion route, perfusion pressure and perfusate oxygenation:  Normothermic perfusion was tested in pig livers through dual perfusion via portal vein and hepatic artery (10–13). Single portal perfusion was usually preferred in rat livers (14–16). In both models perfusion pressure was adjusted to physiological values (7–10 mmHg for the portal vein, 90–100 mmHg for the hepatic artery) (10–13). Perfusate oxygenation ranged between 14 and 20 kPa (pO2) in sanguineous systems (10,11) and 60–70 kPA (pO2) in asanguineous perfusion (14).

Perfusate composition:  In principle, the various components of normal blood provide an ideal perfusate for warm perfusion analogous to cardiopulmonary bypass. In addition, multiple metabolites for energy support must be included to balance the high catabolism activity of the liver under normothermic conditions. The group of Peter Friend in Oxford designed a perfusion machine for pig livers using whole blood supplemented with heparin, taurocholate, prostacyclin and many nutrients (10–12). Peter Neuhaus’ group in Berlin added a dialysis unit to remove the putative toxins accumulating during machine perfusion (13).

Effect of normothermic oxygenated perfusion tested in transplant models:  Unpreserved rat livers were normothermically perfused for 6 h and successfully transplanted thereafter (16). Normothermic oxygenated perfusion of pig liver grafts after 60 min of warm ischemia resulted in survival of the recipients in contrast to the untreated group (13). Recently, a Spanish study showed salvage of human DCD livers applying normothemic liver perfusion directly after warm ischemia (17); deceased donors (8 min of cardiac arrest) were transported under continuous reanimation followed by rapid cannulation of the femoral vessels and normothermic oxygenated in situ perfusion with autologous heparinized blood for 3 h. Five of 10 recipients survived after transplantation (17). However, this concept required start of in situ perfusion within 120 min after cardiac arrest before organ retrieval.

No transplant study is currently available investigating normothermic oxygenated perfusion after combined warm ischemia and cold storage.

Mechanism of protection:  The central concept behind normothermic perfusion is to maintain normal function of the liver during the whole period of preservation and enable immediate graft function. Two studies designed an ex vivo pig model to mimick the real clinical situation of organ allocation for OLT; a warm ischemic period of 60 min (cardiac arrest) was followed by cold ischemia for organ transport (1 h or 4 h). Afterwards, the grafts were normothermically oxygenated and perfused with assessment of viability. However, normothermic oxygenated sanguineous perfusion after this combination of warm and cold ischemia failed to rescue the grafts (18,19).

One reasonable explanation for this shortcoming could be the presence of leucocytes and platelets in the perfusate, which may multiply reperfusion injury during normothermic oxygenated perfusion after warm and cold ischemia.

Advantages of normothermic oxygenated perfusion:  One significant advantage of normothermic oxygenated perfusion relating to the physiological state of organ perfusion is the ability to assess liver viability during preservation by analysis of bile flow, urea production, oxygen uptake and release of several cytosolic enzymes (10–13). Second, in theory, liver grafts could be preserved for several days without risk of organ damage including a normal endothelial cell lining (10).

Disadvantages of normothermic oxygenated perfusion: Normothermic perfusion with a sanguineous perfusate seems to protect only in the absence of exposure to cold storage (18,19). Currently, multiorgan harvesting is associated with cooling followed by cold preservation. Thus, to avoid significant graft damage, warm perfusion should begin immediately before or after harvesting of a donor liver, that is in absence of any cooling. This approach probably will not be accepted by those involved with harvest of other organs.

Hypothermic oxygenated perfusion

Hypothermic oxygenated perfusion (HOPE) is based on the concept that oxidative energy production via mitochondrial electron transport sustains at low temperatures. The value of hypothermic machine perfusion becomes apparent when taking into consideration that mitochondria are central players in mediating damage in the ischemic cell (6). However, mitochondrial and other cell processes are unlikely to function normally in the hypothermic state. The dysregulation of the system secondary to the altered activity of different enzymes and RNA transcription may also result in harm.

Perfusion route and application:  Several different perfusion routes have been tested for cold machine perfusion. Seventy-two hour continuous hypothermic perfusion of pig livers through the portal vein alone resulted in viable livers (20). Twenty-four hour retrograde perfusion via the hepatic veins of rat livers was similar to portal vein perfusion while perfusion via the hepatic artery alone was less beneficial (21). All of those studies used isolated perfusion models to test reperfusion injury.

Perfusion pressure and flow:  A critical target of injury of cold preservation, with or without perfusion system, is the sinusoidal endothelial cell (24). The perfusion conditions must be carefully adjusted with most studies selecting a low portal vein pressure of 3–4 mmHg and low arterial pressure of 20–30 mmHg in case of dual perfusion. Portal flow rates are reported within a wide range between 0.14 and 3.5 mL/min/g liver (20–23,25–28). Both, constant flow or constant pressure systems have been used for HOPE (22,27,28). Prolonged application of constant flow during hypothermia injures sinusoidal endothelial cells due to increased vascular resistance and shear stress (29,30). One study correlated perfusion pressure and parenchymal flow (31). Cold portal perfusion with constant pressure rates at 25% of normothermic conditions resulted in complete perfusion of the liver without induction of endothelial injury (31).

Perfusate oxygenation:  Oxygenated perfusion enables the graft to restore tissue energy charge (32,33). Even brief periods of temporary HOPE during or after long-term storage of rat livers significantly improved recovery of cellular energy charge and glycogen content (34). Magnetic resonance studies in pig livers showed that ATP signals were absent in grafts preserved by cold storage, but within minutes of HOPE, hepatic ATP signals were detectable (34). Reported perfusate oxygenation in the cold ranged between 9 and 95 kPA (20–23,25–31).

Few studies are available comparing oxygenated versus complete deoxygenated cold liver perfusion (35–37). All suggest that continuous supply of oxygen is necessary for successful hypothermic perfusion preservation.

Perfusate composition:  The choice of the colloid has remained a focus of research. The Amsterdam group recently proposed a switch to polyethylene glycol as the major colloid for liver machine perfusion (38). A perfusate with physiologically low potassium concentration avoids an increased vascular resistance of livers during cold perfusion (29). Some ingredients of the original University of Wisconsin and histidine-tryptophan-ketoglutarat solutions were identified to play a former unknown role (Table 1). For example, lactobionate chelates calcium down to the millimolar level (39) and acts as potent cryptic inhibitor of metalloproteinases (MMP), which are released by endothelial cells during preservation (40).

Table 1. Machine perfusion solutions for hypothermic oxygenated perfusion (HOPE)
ComponentBelzer MPS (UW gluconate) (22)HTK (47)Polysol (50)Celsior (21)IGL-1Modified UW (28,33,43,44)
Colloid
 Hydroxyethylstarch5% (50 g/L)     
 PEG  2 g% 0.03 mmol/L 
Electrolytes
 Na/K ratio100/2515/10120/25100/15120/2525/120
 Ca0.5 mmol/L0.015 mmol/L0.27 mmol/L0.25 mmol/L 2 mmol/L
Buffer
 HEPES10 mmol/L 20 mmol/L   
 KH2PO415 mmol/L 6 mmol/L 25 mmol/L25 mmol/L
 MgSO4    5 mmol/L5 mmol/L
 NaHCO325 mmol/L     
 Histidine 198 mmol/L 30 mmol/L  
Antioxidants
 Allopurinol1 mmol/L 1.2 mmol/L 1 mmol/L1 mmol/L
 Glutathione3 mmol/L 3 mmol/L3 mmol/L3 mmol/L3 mmol/L
 Adenosine  5 mmol/L 5 mmol/L5 mmol/L
 Adenine5 mmol/L 5 mmol/L   
 Ribose5 mmol/L     
 Deferoxamine 10 mmol/L
Impermeants 
 Raffinose  2.7 mmol/L 30 mmol/L30 mmol/L
 Gluconate115 mmol/L 95 mmol/L   
 Lactobionate 80 mmol/L100 mmol/L100 mmol/L
 Trehalose  5.3 mmol/L   
 Mannitol 30 mmol/L 60 mmol/L  
Energy substrates
 Glucose10 mmol/L 11 mmol/L   
Amino acids
 Tryptophan 2 mmol/L    
 Ketoglutarate/glutamic acid 1 mmol/L 20 mmol/L  
 Amino acids, vitamins  11 mmol/L   
Additives
 Bactrim2 mL/L     
 Gentamycine     80 mg/L
 Insulin40 units    40 units
 Dexamethasone16 mg/L    8 mg/L
 Vitamins  0.15 mmol/L   
Osmolarity310 mosm/L310 mosm/L  290 mosm/L295 mosm/L
pH7.47.47.4 7.47.3

Effects of hypothermic oxygenated perfusion tested in transplant models: Six transplant studies (25–28,41,42) were done concerning HOPE. Guarrera et al. compared 5-h cold storage with 5-h HOPE through the portal vein and hepatic artery in pig livers. Posttransplant results of liver function were comparable between the groups, with all animals surviving for 5 days (25). Pienaar et al. reported successful continuous HOPE of canine livers for 72 h through the portal vein only. Seven of eight dogs survived following OLT (26). In a rat liver transplant study, Lee et al. showed the benefit of HOPE; 5-h cold machine perfusion after 30 min of warm ischemia led to survival of five from six recipients (27).

The data on improved outcome with the use of prolonged HOPE is still scarce, as only three studies indicate successful liver transplantation following HOPE for more than 24 h (26,41,42).

In contrast to those investigators, who applied their perfusion systems through the whole preservation period, a new concept of HOPE was presented recently (28,36). This approach considers practicability by limiting the perfusion period to an endischemic perfusion for only 1 h. A recent study confirmed the benefit of short-termed HOPE in a rat liver transplant model through single portal vein perfusion (28).

Mechanism of protection:  Hypothermic machine perfusion acts through several pathways. Mitochondrial respiration is reactivated as hypothermic machine perfusion oxidizes mitochondrial electron complexes prior to rewarming and reperfusion (43). Upon reperfusion, mitochondrial electron overload may be prevented with decreased production of reactive oxygen species (ROS), evident in rat livers by a decrease in lipid peroxidation and preserved cellular glutathione (44).

Other studies suggest that injury also occurs to the sinusoidal endoplasmatic reticulum during cold and warm ischemia (45,46). ATP deficiency can promote Ca2+ escape from the endoplasmatic reticulum into cytosol (46), triggering increased calpain activity, actin disassembly and release of MMP (40). This, in turn, may lead to the expression of von Willebrand factor (vWF) (40) and intercellular adhesion molecule 1 (ICAM-1) (47) on the sinusoidal surface. HOPE with UW (28) or HTK solution (47) was shown to reverse not only ATP depletion but oppose calpain and MMP activity possibly because of the specific actions of lactobionate and histidine (Figure 1). This observation correlated with a remarkable change in sinusoidal endothelial appearance after short-term cold machine perfusion in cold stored DCD donor rat livers (28,33). Livers exposed to the machine perfusion system disclosed intact microvilli along the endothelial surface in contrast to untreated livers (Figure 2).

image

Figure 1. HOPE possibly acts through several mechanisms. First, cellular energy charge increases by oxidative phosphorylation (44). Second, a rise of cellular calcium and activation of metalloproteinases can be counteracted by lactobionate or histidine (39,40). Third, overreduction of respiratory chain is decreased (43). It results in a change in mitochondrial redox state and less activation of sinusoidal endothelial cells which contribute synergistically to less release of ROS and less adhesion of platelets and leukocytes.

Download figure to PowerPoint

image

Figure 2. Electron microscopy of rat livers exposed to 45 min of warm ischemia followed by 5 h of cold storage and 3 h of normothermic reperfusion on the isolated perfused liver system (28). Livers treated by 1 h HOPE before reperfusion showed intact sinusoidal endothelial surface (arrows) in contrast to the untreated group (arrows).

Download figure to PowerPoint

Advantages of hypothermic oxygenated perfusion:  The strength of the cold oxygenated machine perfusion approach is its capacity to restore hepatocytes and endothelial cells under nonphysiological conditions. In contrast to sanguineous normothermic oxygenated perfusion, hypothermic machine perfusion preconditions the liver cells prior to the introduction of blood cells (Figure 1).

This accumulating data suggests that hypothermic oxygenated machine perfusion currently appears as an attractive and practical strategy to rescue cold stored marginal livers.

Disadvantages of hypothermic oxygenated perfusion:  During cold perfusion there is a time dependent increase in vascular resistance which bears the risk of damage to the sinusoidal endothelium and endosplasmatic reticulum, particularly when cold perfusion extends beyond 18 h (29,30,45).

The second potential risk of HOPE is the exposure to continuous oxygenation at low temperatures. Indications for the role of ROS under oxygenated low temperature conditions have come from studies on isolated cell systems (48). Whole liver perfusion, however, resulted in minor oxidative stress proved by different investigators (28,33,47,49) which included glutathione, deferoxamine or superoxide dismutase in their perfusate. Based on this, the risk of potential dangerous cytokine or ROS release appears low despite exposure to oxygen in the cold if sufficient concentrations of antioxidants and iron chelators are available in the perfusate.

Effects of Machine Perfusion on Marginal Grafts

  1. Top of page
  2. Abstract
  3. Introduction
  4. Scientific Methods to Test Liver Perfusion
  5. Machine Perfusion Systems
  6. Effects of Machine Perfusion on Marginal Grafts
  7. Ideal Machine for Clinical Use
  8. Future Perspectives
  9. References

Several investigators have confirmed advantages of HOPE in marginal livers. A recent study showed improved function of macrosteatotic rat livers by HOPE (50), tested on the IPL. Cold stored livers were energetically recharged by HOPE and showed improved viability upon ex vivo reperfusion (33). Cold stored nonheart beating rat livers were successfully rescued by short-term HOPE as demonstrated ex vivo and after transplantation (28). Despite exposure of oxygen, no signs of reperfuson injury occurred during cold machine perfusion (28,33).

In contrast, introduction of blood cells and oxygen to the ischemic liver at normothermic conditions (normothermic oxygenated perfusion) provokes reperfusion injury dependent on the type and the duration of the previous ischemia. Accordingly, normothermic oxygenated perfusion failed to protect preserved DCD liver grafts (18,19). No studies are available concerning normothermic oxygenated perfusion of steatotic livers or after prolonged cold storage.

The effect of HOPE or normothermic oxygenated perfusion on bile duct viability is unknown. As several studies on DCD livers demonstrated that ischemic cholangiopathy is the main cause of long-term graft failure, it will be of great importance to investigate bile duct viability after cold or warm machine perfusion of DCD livers. Currently, no transplant study is available on this issue.

Ideal Machine for Clinical Use

  1. Top of page
  2. Abstract
  3. Introduction
  4. Scientific Methods to Test Liver Perfusion
  5. Machine Perfusion Systems
  6. Effects of Machine Perfusion on Marginal Grafts
  7. Ideal Machine for Clinical Use
  8. Future Perspectives
  9. References

Most of the published machine perfusion techniques do not consider practicability because it is generally recommended to apply perfusion immediately after organ harvest and during the whole preservation period. From a clinical point of view, perfusion during organ transport would be unavoidable which bears the risk of perfusion failure due to logistic reasons. In addition, back table preparation after perfusion would lead to secondary ischemia. In contrast, we suggest a short-term machine perfusion performed after arrival of the harvested donor organ at the center. A period of approximately 1–2 h usually accumulates after completion of back table preparation which allows interventions on the preserved and prepared graft during recipient hepatectomy without delay of the transplant procedure. Either hypothermic oxygenated low-pressure perfusion or normothermic asanguineous oxygenated perfusion are conceivable. The latter technique requires, however, full nutritional support and dual perfusion through portal vein and hepatic artery.

It remains unclear how long machine perfusion should be performed for maximal benefit. Short-term oxygenated perfusion potentially lowers the risk of possible release of ROS and reduces endothelial shear stress. On the other hand, prolonged machine perfusion reloads energy storages more effectively.

Future Perspectives

  1. Top of page
  2. Abstract
  3. Introduction
  4. Scientific Methods to Test Liver Perfusion
  5. Machine Perfusion Systems
  6. Effects of Machine Perfusion on Marginal Grafts
  7. Ideal Machine for Clinical Use
  8. Future Perspectives
  9. References

Cold storage causes injury to the liver graft regardless of the preservation strategy, only the degree of injury varies. Good quality grafts tolerate well preservation periods of up to 12 h while marginal grafts probably should be implanted earlier. Methods which can discriminate livers exhibiting delayed graft function under reperfusion are still lacking. It remains difficult, therefore, to decide which graft will benefit from a rescue strategy by machine perfusion. Grafts from DCD donors (warm ischemia > 30 min), steatotic grafts (>30% macrosteatosis) and grafts with prolonged cold storage (> 16 h) are currently not considered by most surgeons for transplantation because of the high risk of graft failure (1–3) and qualify therefore for an optimizing strategy. Based on practicability, we suggest that those marginal livers could be best improved by a short-term cold oxygenated perfusion after back table preparation (28,51). No transportable devices are necessary with such an endischemic approach. More studies are needed in large animals and on discarded human livers to bring this strategy to humans.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Scientific Methods to Test Liver Perfusion
  5. Machine Perfusion Systems
  6. Effects of Machine Perfusion on Marginal Grafts
  7. Ideal Machine for Clinical Use
  8. Future Perspectives
  9. References