Salvaging the fatty liver for transplant: is short duration NMP enough?

Obesity is an epidemic in the US. As of 2020, the Center for Disease Control (CDC) notes that 40% of the ~258 million US adults suffer from obesity. This represents just more than a 100 million people suffering from obesity. In addition, about 23 million people suffer from severe obesity with a body mass index >40 kg/m².¹ Obesity is the primary driver of the increased prevalence of fatty liver disease. Patients with obesity are also at an increased risk for other comorbidities of the metabolic syndrome, putting them at higher risk of death from causes other than a fatty liver, especially cardiovascular disease and stroke. The end result of this epidemic is that we are identifying a greater proportion of organ donors with varying degrees of liver steatosis.

Transplantation of steatotic livers is associated with an increased degree of ischemia-reperfusion injury (IRI) and release of inflammatory cytokines from the graft. The consequences of this can range from severe reperfusion syndromes with immediate vasoplegia and circulatory collapse to distant organ dysfunction with acute kidney injury, liver allograft dysfunction, and primary nonfunction (PNF). Early allograft dysfunction (EAD) in these grafts is frequently accompanied by multiorgan dysfunction resulting in increased ICU length of stay and hospital resource utilization.^2,3 Ito et al² reviewed the degree of IRI in more than 500 patients with postreperfusion biopsy. They demonstrated that patients with moderate to severe IRI had a higher incidence of developing EAD. In addition, on multivariate analysis risk factors for EAD were cold ischemia time and macrovesicular (MaS) steatosis ≥20%. Furthermore, there was a cumulative effect if both risk factors were present in the same graft.

This study by Patrono et al⁴ therefore comes at an opportune time. In this study, the authors analyze the efficacy and safety of normothermic machine perfusion (NMP) in the setting of MaS ≥30%. To this end, they identified 12 grafts that had biopsy-proven MaS ≥30% and underwent NMP for a minimum of 4 hours, after a period of static cold storage. The primary criterion for livers to be considered for transplantation was lactate level ≤4 mmol/L at 2 hours of NMP plus at least 3 of the following criteria: stable pH ≥7.3 without the need for repeated bicarbonate supplementation, evidence of glucose metabolism; bile production ≥2 mL/h; macroscopically homogeneous perfusion; hepatic artery flow ≥150 mL/min; and portal vein flow ≥500 mL/min. These are similar to previously published studies from the UK.⁵

A total of 71% of these steatotic grafts were utilized following NMP. However, there was a 14% PNF rate resulting in a successful transplant rate of 57%. Not taking into account PNF livers, the EAD was ~30% and after a median follow-up of 12 months, patient and graft survival rates were 100% and 80%, respectively. The median time before discard was 4 hours 44 minutes. Of the 2 livers that experienced PNF, one met all viability criteria for NMP but experienced a rebound increase in lactate after hour 3. The second liver met all but glucose utilization viability criteria. It too was noted to have a rebound lactate increase at hour 5. In addition, while meeting the flow criteria, PNF grafts had a 38% decrease in hepatic artery flow and a more modest 16% decrease in portal vein flow at hour 2 of NMP compared with functioning livers. The median AST/ALT was also higher in the perfusate in PNF livers. There was no difference in pH levels in the PNF and functioning grafts.

These data highlight the challenges of currently suggested viability criteria for high-risk grafts. Importantly, one of the grafts that was used and functioned well did not attain the lactate level suggested by the criteria. Other criteria for viability testing have also been challenged. Nasralla et al⁶ showed that lack of bile production on NMP was not associated with posttransplant liver function. Mergental et al⁵ evaluated marginal discarded grafts on NMP and demonstrated that of the grafts that were transplanted 14% did not meet the pH criteria, 18% did not meet bile production criteria, and 10% did not meet the glucose metabolism criteria. Despite this, there were no patients with PNF. However, 18% had nonanastomotic biliary strictures. Viability criteria defined by Birmingham, Cambridge, and Gronigen groups when used for marginal grafts show a varied utilization rate—71% utilization with Birmingham versus 0%–4% utilization of organs with other criteria—emphasizing the need for standardization and better markers of viability in marginal grafts.⁷

Another important finding of this study was the 43% effective discard/PNF rate. Similar NMP discard rates of 44%–81% in steatotic grafts with or without additional high-risk features were seen in other studies.^5,6 Therefore, to fully increase utilization of steatotic grafts, additional strategies are needed.

Strategies to make steatotic grafts safe for transplant can broadly be divided into those related to reducing the initial ischemic injury and those related to recovery, rehabilitation, and organ optimization in the ex situ environment. Reducing cold ischemia times by placing livers on NMP at the donor center may help in reducing the cold-induced injury. However, utilizing ischemia-free recovery and reperfusion techniques with NMP will likely yield the greatest benefit by preventing the first cold ischemic hit completely. In the current study, Patrono and colleagues also identified that short-duration NMP may not be adequate to fully evaluate parameters like lactate clearance. Our group has recently presented its novel findings of the utilization of extended NMP, up to 7 days, in rehabilitating marginal grafts.⁸ One of the important results was that the normalization of lactate can take up to 14 hours in marginal grafts. While not all grafts will require such long durations of NMP, this length of NMP can also allow for strategies related to organ optimization.

Organ optimization can be performed by reducing IRI and potentially defatting the liver. Gene silencing using RNA interference in an NMP environment may be effective in reducing apoptosis, and IRI mediated by pathways for TNF-α, NF-κB.⁹ Other techniques include delivering human liver stem-like cell–derived extracellular vesicles during NMP. This has been shown in an animal model to reduce hepatocellular damage.⁹ Liver defatting has been described in animal models using varied cocktails with up to a 50% reduction in lipid content.⁹ Defatting in human livers on NMP has been attempted with varying success. Banan et al¹⁰ attempted defatting in 2 livers with a 10% reduction in macrosteatosis in an 80% MaS liver and no change in a 30% MaS liver. Boteon and colleagues demonstrated a 40% reduction in macrosteatosis in 5 human livers. This is promising, but 2 of the 5 livers in the study group had only 0–5% steatosis.⁹

In conclusion, in the midst of the obesity epidemic, development of benchmarks for viability of steatotic grafts on NMP and strategies for increasing utilization of NMP-treated steatotic donor grafts is critical. The ideal approach would likely incorporate a reduction in the first cold ischemic hit followed by longer duration NMP, which is needed to better evaluate these grafts. Furthermore, extended NMP platforms may be best suited for organ optimization strategies ranging from interventions to reduce IRI or to initiate defatting of these grafts.