Metabolomics Based Prediction Model for Liver Graft Viability During Normothermic Regional Perfusion in Donation After Circulatory Death

With the increasing shortage of available donor organs, liver transplantation from donation after circulatory death (DCD) has been established in many countries to offer a therapeutic strategy for patients with end-stage liver disease beyond standard donation after brain death (DBD). In France, a controlled DCD (cDCD) program based on the use of normothermic regional perfusion (NRP) during organ procurement has been successfully implemented since 2015. This program has since shown excellent post-transplant outcomes with one-year patient and graft survival rates >90%. However, owing to very strict donor and graft selection criteria nearly one third of all potential liver grafts were not transplanted because of presumed poor quality. Furthermore, survival in cDCD liver transplantation which did not adhere to the current selection criteria was significantly lower (68% vs 94%) compared to the highly selected population. Thus, to further expand selection criteria and reduce discard rates without compromising outcomes, there is an urgent need for novel objective methods to assess graft quality prior to transplantation.

Graft quality assessment prior to transplantation has been a key challenge in liver transplantation for decades and still today the decision to accept or decline a graft before transplantation relies for the most part on ''gut feeling" of the procurement or transplant team. Currently, selection criteria for cDCD in France are based on donor data, liver biopsy results and hepatocyte injury markers AST and ALT. While donor data are only indirect indicators of graft quality, liver graft biopsy are highly pathologist dependant and do not provide a dynamic assessment. Hepatocyte injury markers during normothermic perfusion have been shown to only have limited value in predicting post-transplant graft function. In addition, several recent studies have suggested a superior predictive value of specific biomarkers in the liver through metabolomics. There is growing evidence that dynamic graft preservation such as NRP may even allow a direct analysis of key injury metabolites within solid organ grafts.

During cDCD procurement, liver grafts are exposed to donor warm ischemia which in combination with static cold storage causes ischemia-reperfusion injury (IRI) of the graft during implantation in the recipient. IRI is directly correlated to detrimental post-transplant complications such as primary non function and ischemic cholangiopathy or even recipient death.

Studies in animal models have revealed that mitochondria play a major role in IR injury and several mitochondrial signature metabolites for example succinate have been identified in various solid organ grafts such as livers, hearts and kidneys. Based on these results, our team and others have recently identified the release of a small auto fluorescent molecular compound of the mitochondrial respiratory chain, Flavin Mononucleotide (FMN), during the early phase of hepatic I/R injury. Furthermore, FMN has been shown to serve as a surrogate marker for impaired cellular energy production of the liver graft prior to transplantation and enable accurate prediction of post-transplant liver graft viability. Given the natural fluorescence of FMN, a real time quantification method has been established allowing to rapidly assess viability of human liver grafts during the procurement process. Of note, a team from the United Kingdom has validated this real-time FMN quantification during NRP of cDCD liver grafts. Interestingly, since mitochondrial injury is a universal signature of graft injury during the transplant process, FMN has the potential to serve as viability marker in other solid organs such as hearts, lungs and kidneys.

In addition, metabolomics approach allow the identification and quantification of many metabolites in a biological system through the use of high-throughput analytical technologies. It can give valuable information about what has, is and will occured in a specific tissu. In Liver transplantation, efforts have focused on studying early event of IR which are known to influence long term outcomes. It is however a complex pathway involving numerous metabolites. Metabolomic is therefore gaining interst in field of LT because it can enlight underlying metabolic process occurring during organ preservation and can lead the way toward a better evaluation of graft quality.

Thus, this study aims at developing a robust bio-clinical prediction model of liver graft viability during NRP using in depth metabolomics of perfusate and tissue in combination with available donor and graft characteristics