Metabolomic Analysis, Perfusate Composition, and Pseudo-physiology of the Isolated Liver During Ex Situ Normothermic Machine Perfusion

The liver fulfills complex functions of metabolism, synthesis, storage, and redistribution of nutrients. During the ex situ phase of transplantation, such functions are suppressed by lowering the temperature of the graft to 0 °C to 12 °C. On the contrary, they may be activated in the graft during the ex situ phase using normothermic machine perfusion (NMP). Considering the temperature range at which it is performed (35 °C–38 °C), NMP ideally recreates in situ physiological conditions for the liver in the ex situ setting in order to maintain liver homeostasis if not initiate processes of regeneration and repair. Nonetheless, most liver NMP being performed clinically is done on devices and under protocols designed to last for only a few hours and providing far from physiological conditions. Aside from issues related to hepatic artery perfusion (continuous as opposed to pulsatile) and inclusion of short-duration membrane oxygenators, another common shortcoming of liver NMP for extended perfusion preservation is that it supplies little more than heparinized red blood cells diluted in colloid and supplemented continuously with a limited array of nutrients and electrolytes, insulin, antimicrobials, and a single bile acid to increase bile flow.1 Although normothermic liver perfusion performed in this manner may keep a graft viable over a relatively brief period, it creates a pseudo-physiological situation, in which some liver functions may be maintained but others are notably altered. As a reflection of this fact, a study performed by Zhiyong and colleagues2 and published in this issue of Transplantation describes metabolomic analyses performed on livers undergoing ischemia-free liver transplantation (IFLT), comparing metabolic and other changes arising between the in situ and ex situ settings. IFLT involves connecting a liver to an ex situ NMP device while still in the donor and maintaining the connections and ongoing hepatic arterial and portal blood flow during recovery, transportation, and transplantation into the recipient. Although challenging, IFLT offers a unique opportunity to evaluate changes in liver function arising during ex situ NMP relative to in situ perfusion of the same liver in its native state, without the confounding effects of either warm or cold ischemia. In the study by Zhiyong et al, stable arterial and venous flows, blood gas parameters, transaminase release, and lactate clearance were observed during ex situ liver NMP lasting 6 h on average. However, metabolic and synthetic activities measured in samples collected after 2 h of ex situ NMP were significantly reduced relative to those measured in situ before livers were disconnected from their donors. Ex situ livers suffered urea cycle dysfunction, and progressively rising urea levels were observed throughout NMP. Authors also observed substantial increases in glucose levels in the NMP perfusate during the first hour, an observation that they were not able to justify but might be explained—at least in part—by glucose included in bags used to collect and transport blood to the machine. Normal liver function in vivo depends on continuous interaction with and inputs from the rest of the body, many of which are known and some of which remain to be discovered: Removal of metabolic wastes not cleared by the liver, such as urea produced from ammonia breakdown, is managed by the kidney. Pancreatic hormones reaching the liver include not only insulin but also glucagon, which aids in hepatic glucose control by stimulating the liver to convert stores of glycogen and circulating amino acids into glucose (glycogenolysis and gluconeogenesis, respectively). Nutrients absorbed from the gut or released from other parts of the body (eg, adipose tissue and muscle) do not flow into the liver continuously but change according to the circadian variations and stressed versus nonstressed states. Bile acids absorbed in the terminal ileum and returned to the liver via enterohepatic circulation include not a singular molecule but a diversity of primary and secondary steroid acids; they play a critical role in digestion of lipids and bind their receptors to aid in maintenance of liver metabolic homeostasis. Gut microbe-derived metabolites represent another important component of the gut-liver axis, supporting hepatic homeostasis and metabolic function. Moving beyond normal liver function to regeneration, platelets and leukocytes have been implicated in initiation of intrahepatic processes of repair.3,4 An experimental study comparing ex situ liver NMP performed with diluted red blood cells versus whole blood observed a trend toward better outcomes with the latter.5 There is also an important role for bone marrow-derived progenitor cells in the regeneration of certain liver cell populations (eg, endothelial cells), and their absence has been seen to cause significant impairment in hepatic regeneration.6 Table 1 summarizes common perfusate components used during ex situ liver NMP, as well as aspects that might be modified and improved and interesting concepts for future research. One ex situ liver NMP device (Liver4Life, Wyss Zurich, Switzerland) has incorporated some of these items, such as dialysis, glucose control via infusion of insulin and glucagon, discontinuous (ie, circadian) infusion of amino acids and lipids, and inclusion of platelets.8 Using this protocol, whole and partial liver grafts from both pigs and humans have been preserved ex situ for periods of up to 7 to 10 d.9 Recently, the liver of a 29-year-old donor with ongoing sepsis due to multiple drug-resistant organisms and a segment 1 lesion was treated with antimicrobials and analyzed for 3 d on the device, after which point it was successfully transplanted into a human recipient.10 TABLE 1. - Common ex situ liver NMP protocol components, aspects needing improvement, and areas for further research Commonly administered Commonly missing but important Oxygenated red blood cells or hemoglobin-based oxygen carrier Dialysis Plasma or other colloids Glucagon Electrolytes Lipids Bicarbonate buffer (priming solution) Circadian variation in infusion of nutrients Continuous infusion of nutrients (amino acids, glucose, multivitamins, trace elements) Single bile acid a Anticoagulant Broad-spectrum antimicrobials Vasoactive substances Methylprednisolone May be administered in excess Potentially important (need further research) Glucose Range of bile acids Insulin Platelets Oxygen b Leukocytes Bicarbonate c Gut microbe-derived metabolites Bone marrow-derived progenitor cells aSodium taurocholate or ursodeoxycholic acid.bHyperoxia during ex situ liver NMP has been seen to have negative effects that include on-device vasoplegia and greater reperfusion injury among transplanted recipients.7cAlthough bicarbonate is commonly added to adjust the pH of the NMP priming solution, there might be a protective role for mild acidosis during initial liver reperfusion on the device and raising of the temperature to normothermic range.NMP, normothermic machine perfusion. Both standard ex situ NMP and IFLT represent an exciting frontier in the field of liver transplantation. Currently, they are means to transplant more livers with greater confidence, improved logistics, and less ischemia. Their true potential, however, lies in the opportunity to study, treat, and modify livers over the course of days to weeks. In spite of rising clinical application, there is ongoing need for research on in situ liver physiology and the physiological adequacy of ex situ conditions in order to reach the ultimate goal using NMP and IFLT to recreate human conditions for the liver outside the human body.