Liver transplantation (LT) provides the only definitive cure for end-stage liver disease, though access is limited due to a shortage of donor organs. The growing use of ex situ machine perfusion technology for dynamic liver preservation prior to LT has significantly expanded the use of marginal and extended-criteria grafts in recent years (1, 2). However, sizeable knowledge gaps remain with respect to liver physiology during normothermic machine perfusion (NMP) that limit its potential use as a platform for rehabilitation of untransplantable or severely injured grafts. Recent work from our collaborative group has delineated differences in liver physiology during NMP relative to hepatocellular function as defined by the balance between ischemia-reperfusion injury (IRI) and recovery of cellular homeostasis via autophagy (3). Therefore, leveraging the equilibrium between graft injury and recovery with adjunct therapeutic interventions during NMP may present a reliable avenue for rehabilitation of graft function and transplantation of livers with inadequate function.

One approach to “tipping the scale” between IRI and autophagy may be to target cell death pathways that represent the consequence of irrecoverable injury (4, 5). Apoptosis, pyroptosis, and necroptosis represent different forms of tightly controlled stress-induced regulated cell death mechanisms in the liver (6). Under normal physiologic conditions, apoptosis allows for controlled removal of injured liver cells followed by cellular proliferation of surrounding healthy hepatocytes. Autophagy is another homeostatic control mechanism in which damaged organelles and proteins are degraded, thus protecting cells from undergoing apoptosis. Crosstalk between apoptosis and autophagy can dictate a cell’s fate (5). However, under abnormal conditions, apoptosis can be pathologic if the equilibrium between turnover and proliferation is disrupted, resulting in chronic inflammation, fibrosis, and even cancer (7). Dysregulated apoptosis has been described in the setting of liver disease and transplant for decades and numerous studies have shown that therapeutic inhibition of apoptosis results in mitigation of IRI (8–13).

Given that our previous work demonstrated the association of hepatocellular function during NMP with the balance between liver injury and autophagy, we hypothesized that discarded human livers with inadequate hepatocellular function during NMP would be characterized by an insufficient autophagic response with resultant overwhelming apoptosis. We subsequently tested whether adjunct delivery of an irreversible pan-caspase inhibitor, emricasan, during NMP would improve hepatocellular function of discarded human livers by shifting the equilibrium between cell death and autophagy.

Pathologic apoptosis, represented by an imbalance between regulated cell death and cell renewal, is an unregulated process that can be sustained and injurious (7). Release of intra-cellular contents, including DAMPs and cytokines, activate immune cells and propagate the cycle of cell death and inflammation. In this study of discarded livers subjected to NMP, livers with IHF were characterized by abundant apoptotic cell death, robust and sustained activation of innate immunity, and decreased activation of ER stress response pathways (autophagy) when compared to livers with AHF. Addition of an irreversible pan-caspase inhibitor, emricasan, to the perfusate composition resulted in significantly lower levels of apoptotic cell death and mitigation of innate immune responses during NMP. Comprehensive analysis of transcriptomic data indicate addition of emricasan resulted in suppression of gene expression related to innate immunity such as cytokine-mediated signaling and leukocyte migration, as well as decreased concentrations of DAMPs and pro-inflammatory cytokines in the circulating perfusate. Transcriptional changes were seen at the level of both upstream regulators and downstream target gene expression.

Early applications of emricasan in rodent LT models demonstrated improved survival when drug was given at the time of liver procurement (10–12). In a subsequent randomized clinical trial, Baskin-Bey et al. demonstrated significantly lower AST and ALT levels in LT recipients when emricasan was added to the preservation flush solution compared to a placebo group, though the trial was not powered to detect differences in graft survival or EAD (13). Despite these promising results, routine clinical application of emricasan never reached wider use likely owing to the complexity of incorporating drug therapies into the clinical workflow at the time of organ procurement. The advent of perfusion technologies as a platform for organ rehabilitation is likely to provide a solution to these obstacles as organs can be dynamically preserved, monitored, and treated ex situ prior to implant (20). This study is a step forward in this direction, demonstrating that treatment of discarded DCD livers with emricasan during NMP after a period of cold and warm ischemia mitigated injury driven by innate immune and pro-inflammatory responses during ex situ preservation.

The physiologic benefits of emricasan-adjunct NMP are likely multifactorial. Studies of individual caspase enzymes in the last decade have revealed varied functionality, with some acting as initiators of cell death (caspase 8,9,10) and others as executioners involved in end-stage proteolysis (caspase 3,6,7). A third class named inflammatory caspases (caspase 1,4,5) play a significant role in inflammasome activation and IL-1β release, often characterized as pyroptosis (6, 19). Inhibition of inflammatory caspases by emricasan may be one reason EM-treated livers had uniformly depressed IL-1β perfusate levels and markedly suppressed transcriptional enrichment of innate immune responses. Additional studies are needed to determine whether equivalent results can be obtained with specific inflammatory caspase inhibitors versus the pan-caspase inhibitor used here. A narrower therapeutic target may also minimize off-target consequences. For these reasons, we cannot definitively conclude that inhibition of apoptosis in of itself was the mechanism by which ex situ innate immune responses were mitigated, or whether inhibition of pyroptosis had a significant beneficial effect. However, necroptosis does not appear to have a major role during liver NMP, nor does it appear to be unintentionally activated by pan-caspase inhibition.

Despite these limitations, prevention of unregulated apoptosis (and pyroptosis) and thus release of DAMPs by injured hepatocytes likely minimizes activation of resident immune cells and further cytokine release. This appears to disrupt the vicious cycle wherein injured hepatocytes undergo unregulated apoptosis and activate a cascade of cytokine-driven innate immune responses that in turn drive further cell death. Using a rat liver NMP model, Scheuermann et al. likewise demonstrated higher cytokine and DAMP concentrations with resultant downstream signaling consequences, including apoptosis. Less inflammatory signaling was seen when rat livers were perfused at lower temperatures (21). Furthermore, it is of interest that the cytokine signature seen in IHF and some AHF livers was redolent of the pattern seen in sera of liver transplant recipients who experienced IRI in a study by Sosa et al. (18). In contrast, the suppressed proinflammatory cytokine signature seen in EM livers more closely resembled non-IRI livers. However, histopathology did not demonstrate a significant difference in inflammatory cell infiltrate, necrosis, or reperfusion injury between individual groups. This may be a limitation of sample size or the scoring system, which was extrapolated from post-LT reperfusion biopsies, though it warrants more cautious interpretation of these results.

We further investigated ER stress responses to delineate potential crosstalk between apoptosis and autophagy. Prior studies have shown when cellular stress overwhelms autophagy, cells are directed toward apoptosis (5). We therefore hypothesized that inhibiting apoptosis would increase flux through autophagy in ischemic-injured hepatocytes, thereby promoting cell recovery and improving hepatocellular function. In fact, we found a dampened autophagic response. AHF livers demonstrated significantly more activation of autophagy, corroborating our findings from a recent smaller study (3). On the other hand, EM livers demonstrated comparatively moderate autophagic flux, with an appropriate decrease in the ER stress marker phospho-eiF2α, a modest increase in LC3B-II:I ratios, and a cyclic reactivation of autophagy inhibition by 6 hours of NMP. It is possible that the dampened inflammatory response due to emricasan treatment does not require such a strong cellular stress response. A more likely consideration is that caspase inhibition was too far downstream and did not achieve the intended effect on autophagy. Targeting proteins more upstream at the nexus between autophagy and apoptosis may be a better therapeutic option (22).

With respect to rehabilitation of liver hepatocellular function, it is not possible to draw definitive conclusions given the variability between EM livers. Isolated examination of EM-3 and EM-4 livers (and to a lesser extent EM-5) appear to demonstrate better than expected lactate clearance, based on the initial slow clearance followed by a more rapid drop. We were unable to evaluate post-transplant function of the research livers as no perfusion device had received regulatory approval in the United States at the time of this study. Alternatively, whole blood reperfusion studies could have been incorporated into the emricasan-treated livers but would have been limited by our use of retrospectively reanalyzed samples for the control groups. However, the feasibility of a clinical trial is no longer a barrier with recent FDA approval of two NMP devices.

Another limitation worth discussion is the wide variability seen in human liver studies. The relatively small group size in this study limited our ability to make more definitive conclusions. Despite this, we were able to demonstrate statistically significant differences in both transcriptional and protein analyses with clinically important implications for real-world practice. Future studies with a larger study population may be better suited for anticipated clinical trials given the resource-intensive nature of preclinical discarded human research. Another consideration was the apparent non-superiority in improving cholangiocellular function, as emricasan did not appear to improve IRI in cholangiocytes. This is a known limitation of end-ischemic NMP in DCD livers (23) and a precursor period of hypothermic oxygenated perfusion with or without controlled rewarming is likely required to prevent detrimental IRI in cholangiocytes (2, 24). The observed trend toward higher bile production in emricasan livers likely reflects the preserved hepatocellular function of bile secretion but not the cholangiocellular function of bile detoxification and alkalinization (25). In contrast, one factor that favors the wider application of emricasan is its excellent safety and tolerance profile, having been demonstrated in clinical trials for both liver transplant and non-alcoholic steatohepatitis (13, 26). Finally, the endpoints (cleaved cytokeratin 18, caspase-3/7 activity) used to determine therapeutic efficacy (26, 27) demonstrated a significant measurable difference within a clinically relevant perfusion timeframe, making incorporation of adjunct therapy a viable option during routine NMP. For these reasons, translation of these preclinical findings into a pilot clinical trial appears feasible. An initial series of standard criteria donor livers undergoing emricasan-adjunct NMP followed by LT would allow drug safety and adverse event monitoring. This could be followed by a larger randomized clinical trial with attention paid to early allograft dysfunction and graft survival as endpoints. In conclusion, therapeutic pan-caspase inhibition during liver NMP mitigates innate immune and pro-inflammatory responses in the ex situ setting and may be a valuable tool for improving hepatocellular function prior to transplantation.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/ Supplementary Material .

The studies involving human participants were reviewed and approved by Massachusetts General Hospital Institutional Review Board and Lifespan Institutional Review Board. The patients/participants provided their written informed consent to participate in this study.

SR: conceptualization, investigation, methodology, formal analysis, funding acquisition, visualization, writing – original draft, and writing – review and editing. JSant: investigation, data curation, software, methodology, formal analysis, and visualization. AO: investigation, data curation, software, methodology, formal analysis, and visualization. MH: investigation, methodology, formal analysis, and visualization. SB: investigation and formal analysis.

TC: investigation and formal analysis. RV: investigation and formal analysis. IR: investigation, methodology, and formal analysis. AS: investigation, methodology, and formal analysis. JM: resources, supervision, funding acquisition, and writing – review and editing. PG: resources, supervision, funding acquisition, and writing – review and editing. KU: resources, supervision, funding acquisition, project administration, and writing – review and editing. JSand: resources, methodology, formal analysis, funding acquisition, project administration, supervision, and writing – review and editing. HY: resources, methodology, formal analysis, funding acquisition, project administration, supervision, and writing – review and editing. All authors contributed to the article and approved the submitted version.

AO and JSand are supported by National Institute of Environmental Health Sciences (T32ES007272). SR, HY, and RV are supported by the Massachusetts General Hospital Executive Committee on Research. This research was funded by the National Institutes of Diabetes and Digestive and Kidney Diseases (R01DK096075, R01DK107875, R01DK114506), the National Science Foundation (EEC 1941543, ATP-Bio), and the Rhode Island Hospital/Brown University Department of Pediatrics.

We would like to thank the donors, their families, and the organ procurement organization staff for making this work possible.

SR, HY, and KU are inventors on pending patents relevant to this study and have a provisional patent application relevant to this study. KU has a financial interest in Sylvatica, a company focused on developing organ preservation technology. KU’s interests are managed by the Massachusetts General Hospital and Mass General Brigham in accordance with their conflict of interest policies.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.