Profound Coagulopathy Associated With Pig-to-Primate Xenotransplants: How Many Transgenes Will Be Required to Overcome This New Barrier?

Coagulation and Thrombotic Disorders Associated with Pig Organ and Hematopoietic Cell Transplantation in Nonhuman Primates. Transplantation 2000; 70:1323. Bühler L, Basker M, Alwayn IPJ, Goepfert C, Kitamura H, Kawai T, Gojo S, Kozlowski T, Ierino FL, Awwad M, Sachs DH, Sackstein R, Robson SC, and Cooper DKC The progress toward clinical application of pig-to-human organ xenotransplantation is rather like a good horror story. Just as the hero looks like winning, you turn the page and there is a new and unexpected horror facing him or her. So it is with our progress toward clinical xenotransplantation. Hyperacute rejection has been defeated by transgenic expression of supranormal amounts of complement regulators, to the clamor of the commentators and the NASDAQ. Meanwhile, in the scientific salt mines, we are still working out the details, such as whether there is any advantage in expressing more than one regulator, and even whether the regulators need be of human origin, as pig regulators may be sufficient provided they are in excess (1). However, back to the story, on the next page of our book we find that Bühler et al. have summarized their large experience and presented us with incontrovertible evidence of a major new barrier to our clinical dream. Pig-to-primate renal and hemopoietic cell transplantation provoke profound and lethal disturbances of coagulation. Several articles from the Harvard groups (referenced in their article) culminating in the article in this issue, and one from our group (2), have highlighted the features of this barrier. Essentially, pig-to-primate renal transplants either without any therapy or with various components of the tolerance induction regimen (TIR) are complicated by disseminated intravascular coagulation (DIC). This is severe to the point of lethality and intractable until the graft is removed. Furthermore, primates treated with pig blood stem cells or bone marrow, with or without TIR, develop thrombotic thrombocytopenic purpura (TTP) or thrombotic microangiopathy. It is not clear yet whether this TTP is an independent entity; a full description will be published separately. These studies raise many questions. First, why has this observation not been made before? Second, is the phenomenon restricted to renal and hemopoietic cell xenografts? Third, what is/are the mechanism(s)? Fourth, how might it be prevented or treated? Clearly the answers must be interrelated. The coagulopathy is neither restricted to baboon recipients, as Bühler et al. have observed it in cynomologous monkey recipients, nor is it restricted to non-transgenic pig donors, because Cowan et al. (2) used triple-transgenic (CD55, CD59 and H-transferase) donors. It is not a consequence of TIR as it does not occur when TIR is given without the xenograft, and can occur in xenografts performed without any immunosuppressive therapy. So far its occurrence has been reported only in renal and hemopoietic xenografts, but it is not universally observed. It has not been reported by the Cambridge (Imutran) and Nextran groups in either renal or heart xenotransplantation, although coagulation parameters have not been reported in their studies, presumably because of lack of any overt clinical or hematological problems; however, there seemed to be an increase in thrombotic graft anastomotic problems perioperatively. The occurrence of coagulopathy does not appear to correlate with xenoantibody as it has been observed in both xenoantibody-depleted and untreated recipients, and continued plasmapheresis in particular is a feature of the Nextran regimen but not a usual component of the Imutran regimen. A component of both companies’ treatment regimens, which is not a component of TIR, is cyclophosphamide, and it is possible that the use of cyclophosphamide has an unexpected and unexplained benefit, possibly by a direct action on endothelium or on platelet function. Bühler et al. discuss the molecular incompatibilities in regulation of hemostasis between pig and man, and by implication between pig and primate, and the topic has been reviewed by Lawson and Platt (3) and more recently by Simon Robson (4), who has been the major contributor to this area and is also a co-author of this report. Because quiescent endothelium normally maintains an anticoagulant profile to the adjacent blood, any failure of endothelial regulators of coagulation caused by molecular incompatibilities could be expected to have serious procoagulant consequences. There are three defined molecular incompatibilities. First, porcine von Willebrand factor (vWF) has supranormal capacity to bind human platelet GP1b, which occurs in the absence of shear stress, unlike the interaction between human vWF and GP1b. It is possible that this incompatibility, probably together with platelet binding by Fc receptors to endothelial-bound IgG anti-αGal xenoantibody, contributes to the immediate fall in circulating platelets seen in Bühler’s and our studies. Second, porcine tissue factor pathway inhibitor (TFPI) is relatively ineffective at binding human factor Xa. Consequently there is a failure to inhibit Xa and inefficient generation of the TFPI/Xa complex which is needed to inhibit the VIIa/tissue factor complex. Third, porcine thrombomodulin (TM) binds human thrombin poorly. Normally the protein C anticoagulant pathway is markedly accelerated on endothelium by TM; reduced TM-thrombin binding results in inefficient generation of activated protein C (aPC) which is an inactivator of factors Va and VIII. In the xenogeneic setting the resulting defect is magnified because aPC and TFPI (both of which are defective) are normally synergistic. Together these incompatibilities result in the conversion of large amounts of prothrombin to thrombin, which in turn activates clotting and further enhances it by activating endothelial cells (EC) and by causing platelet aggregation. In addition, the rejection process activates EC with resulting loss of the natural anticoagulant molecules heparan sulfate, TM and CD39. The latter, also known as vascular ATPDase, normally degrades ATP and ADP to AMP with ultimate generation of adenosine. ADP is a crucial agonist of autocrine/paracrine activation of platelets and of EC, while adenosine has potent anti-platelet aggregatory and anti-inflammatory activity. If we accept the proposition that the ultimate success of clinical xenotransplantation will depend on genetic modification of the donor to avoid the need for multiple hazardous therapeutic assaults on the recipient, we must face the issue of how many genetic modifications are technically possible. Our experience suggests that triple transgenics are readily generated (2), thus at least 6 transgenes could be assembled by crossbreeding. However, we may also need to allow for crossbreeding of the transgenic modifications onto an α1,3 galactosyltransferase homozygous knockout background. It is a long story, and we are still reading Chapter 2.