Dysregulated coagulation in normothermic machine perfusion of gene-edited pig liver ex vivo: molecular mechanisms, genetic engineering strategies and implications for clinical translation

Liver xenotransplantation has advanced more slowly than kidney and heart xenotransplantation, and this delay cannot be attributed to immunology alone. When human blood comes into contact with porcine liver tissue, hemostasis is the first system to fail. In a recent decedent study using cross-circulation through an extracorporeal pig liver, 80% to 90% of circulating and transfused platelets were cleared within an hour, while the xenograft maintained its integrity, produced bile, cleared ammonia, maintained pH and supported the international normalized ratio, although values measured in mixed plasma require cautious interpretation. Normothermic machine perfusion is what makes this route practical, since the graft is perfused outside the body: exposure is limited to hours or days, the graft is discarded afterwards, and the recipient avoids lifelong immunosuppression. A phase 1 trial in acute-on-chronic liver failure has now been cleared, which moves coagulation dysregulation from a preclinical question to a clinical one. Here, coagulation dysregulation is no longer a preclinical problem, but rather a problem in early-phase xenoperfusion. In this review, we summarize what is currently known about the reasons behind coagulation dysfunction in this specific liver xenoperfusion setting. We first describe mechanisms that are related to the species barrier: von Willebrand factor binding human glycoprotein Ib without shear, asialoglycoprotein receptor 1-mediated platelet phagocytosis in the sinusoid, mismatched thrombomodulin-protein C axis, and continuous production of porcine coagulation proteins by the xenograft and their entry into human plasma. Then, we attempt to determine what the perfusion circuit adds to that. The genetic engineering strategies developed to address each mechanism are discussed. The gene-edited constructs used in humans to date target complement regulation, the protein C axis, purinergic control, and macrophage recognition, and none of them target von Willebrand factor or ASGR1. We conclude the review with a suggestion for a minimum dataset for coagulation assessment, as the data presented in different studies are extremely difficult to compare with one another.