Prolonged Normothermic Machine Perfusion: Buying More Time for Liver Graft Assessment and Repair

Static cold storage (SCS) has been the gold standard of organ preservation for decades. Although metabolism is significantly decreased during SCS, there is still anaerobic activity in the organ, leading to an array of adverse intracellular changes.¹ Considerable drawbacks of SCS are the limited storage time and challenges in assessing quality and viability of the organ. Ex situ normothermic machine perfusion (NMP) represents a platform that does enable graft assessment before transplantation. During NMP, the organ is kept at body temperature while provided with oxygen and nutrients, creating an optimal environment in which the organ can maintain aerobic metabolic activity. NMP thus provides the opportunity to perform functional assessments aiding in the decision of whether it is safe to transplant the liver. This approach could help to increase organ utilization, especially when using extended criteria donor (ECD) organs. A recent European randomized trial in liver transplantation performed by Nasralla et al² demonstrated safety and potential benefits of NMP over SCS, providing evidence of ameliorated consequences of ischemia-reperfusion injury. Median duration of NMP in this trial was 9 h (with a maximum of 23 h), using a perfusion device certified for ex situ liver graft preservation up to 24 h. Extending the time of ex situ graft preservation beyond 24 h may, at least in theory, create opportunities for liver repair and regeneration. Prolonged NMP (beyond 24 h), however, requires a more sophisticated machine than the ones that are currently commercially available and should, for example, include a dialysis unit for the removal of waste products.³ Recently, 2 groups have described such an advanced perfusion device that can be used for prolonged NMP for up to a week or even longer, showing preserved viability. However, liver transplants have not been performed.^4,5

The Zürich group reported in a recent issue of Nature Biotechnology on the first clinical transplantation of a donor liver after 3 d of ex situ NMP.⁶ The liver was derived from a 29-y-old patient, who underwent multivisceral transplantation because of an invasive abdominal desmoid fibromatosis associated with chronic intraabdominal abscesses and sepsis (University of Zurich, personal communication). The patient’s liver was offered for domino transplantation but subsequently rejected by all Swiss centers because of an unclassified tumor in segment 1, requiring further pathological analysis. Sepsis originating from chronic intra-abdominal abscesses represented an additional reason not to accept the liver. After hepatectomy, the liver first underwent hypothermic (6–8 °C) oxygenated perfusion for 2 h and 40 min, during which the liver was prepared for NMP. The portal vein (PV), hepatic artery (HA), vena cava inferior, and common bile duct were cannulated, enabling perfusion in a closed system. NMP was performed using the Wyss perfusion machine, a sophisticated perfusion device that was extensively described previously.⁵ This machine provides oxygenated blood through the HA and PV using physiological pressures. HA hemodynamics were controlled by automated administration of vasodilators and vasoconstrictors. The liver was placed on a silicon mat, mimicking diaphragm movement to prevent pressure injuries. An integrated dialysis unit maintained a physiological electrolyte balance and facilitated the removal of waste products, including urea. The machine was primed with a leukocyte-free red blood cell–based perfusion fluid. Flow rates in the HA were 0.2 to 0.5 L/min, with a mean arterial pressure of 65 mm Hg. PV flows ranged from 0.8 to 0.9 L/min with a pressure between 10 and 12 mm Hg. Timing of transplantation allowed for histopathological diagnostic workup, performed by day 4, after passing a viability assessment that included physiological response to the vasoactive drugs, quality and quantity of bile production, low markers of injury, and inflammation in the perfusate, in addition to acceptable histology. The histological results of the tumor in segment 1 revealed a benign perivascular epithelioma, and broad-spectrum antimicrobial agents were continuously given to treat potential bacterial contamination. During perfusion, lactate levels in the perfusate dropped quickly; there was continuous bile production, and perfusate aspartate transaminase and alanine transaminase concentrations stayed relatively low, as did interleukin 6, indicating indeed an absence of pressure necrosis of the liver and an absent inflammation. After 68 h of NMP, the liver graft was successfully transplanted without signs of postreperfusion syndrome. Peak serum aspartate transaminase and alanine transaminase concentrations in the recipient remained low at 309 and 138 U/L, respectively. The international normalized ratio returned to normal levels by day 4 after transplantation. During NMP, there was very low to no cellular proliferative activity, as shown by the low expression of MIB1/Ki67 and PH3 immunostaining in a liver biopsy. Interestingly, by day 10 after transplantation, a very strong proliferative activity was observed in a protocol biopsy. This observation raises the intriguing question of whether repair on the pump was not necessary or not possible. Another interesting observation was a weight loss of 25% at the end of the NMP, with an equal reduction in hepatocyte volume.

It should be noted that this liver graft was derived from a living donor and thus not intrinsically injured by the consequences of brain death or ischemia-reperfusion injury, nor did it need to be repaired before the domino transplantation. This certainly raises some questions on how results reflect repair of a deceased donor liver. Others may argue that prolonged hypothermic oxygenated machine perfusion may also provide safe extension of ex situ graft preservation time⁷ (Figure 1). From a hepatobiliary injury point of view, many will argue that this domino liver could have been transplanted without NMP. Moreover, the histology of the tumor in segment 1 could also have been obtained before the transplant/donation procedure. Nevertheless, most of us would not have considered this liver primarily for transplantation, and therefore, this graft seemed suited to show the feasibility of prolonged perfusion. The question of whether graft repair and regeneration of a preinjured ECD liver are possible during ex situ machine perfusion, however, remains to be answered.

FIGURE 1.: An overview of the different liver preservation methods and the current reported maximum preservation duration. Traditional SCS is safe for up to 12 h. Hypothermic oxygenated perfusion (dual or single; [D]HOPE) is typically performed after SCS for 2 h.⁸ Prolonged preservation with DHOPE has been reported for up to 20 h.⁹ “Regular” NMP is possible for up to 24 h.² Prolonged NMP enables graft preservation for up to 1 wk and maybe even longer.^4-6 (D)HOPE, (dual) hypothermic oxygenated machine perfusion; IRI, ischemia-reperfusion injury; NMP, normothermic machine perfusion; SCS, static cold storage.

In summary, Clavien et al are the first to show that transplantation of a human liver after 68 h of ex situ NMP can be safe and successful, stimulating further research into long-term liver graft preservation. Prolonged NMP allows us to take the next step in organ preservation and assessment. Moving forward, the ultimate goal will be to enable ex situ recovery and regeneration of preinjured ECD liver grafts to expand the donor pool and decrease waitlist mortality.