Therapeutic approaches to counteract immediate blood-mediated inflammatory reaction in islet transplantation.

I t is not too common to witness the definition of a key, novel biological phenomenon that profoundly influences both our thinking and our approach to an experimental clinical problem. This has been indeed the case, in the field of islet transplantation, with the description of the immediate bloodmediated inflammatory reaction (IBMIR) by Bennet and Korsgren (1–3), because its impact on transplantation success appears now of cornerstone importance. We may say that it was a brilliant intuition, and that the authors of the original observation approached the problem with an elegantly designed model system that allowed them to explore the phenomenon by the use of novel technology. In particular, the use of an ex vivo monitoring system to assess coagulation and complement activation in a closed loop system led to the conclusion that islets, when they come in contact with blood, act as potent activators of multiple pathways that encompass coagulation, platelet sequestration, polymorphonuclear cell recruitment and complement activation. Activation of these events led to insulin dumping and islet cell death shortly after the initial encounter of the islets with blood (1–3). At least some of the molecular bases of these events have been elucidated, and one of the key players in the triggering of this complex set of cascades has been recognized in the expression of tissue factor (TF) on the islet surface (4, 5). TF activates coagulation, and the subsequent thrombin generation leads to the triggering and sequestration of platelets. Platelets bind to islet surface probably through collagen. IBMIR occurs in the clinical setting, and it is likely to lead to early/immediate loss of a significant portion of the transplanted tissue in clinical trials of islet allotransplantation (6). IBMIR is also likely to amplify the subsequent specific immune alloresponse, because the activation of coagulation, complement, and polymorphonuclear cells have recognized roles in making antigen recognition more efficient. Now that the phenomenon is quite well described and convincingly shown to occur in the clinical setting, the fundamental question is: how do we prevent it? Prevention of IBMIR could clearly result in clinical transplantation success with lower islet masses, and could also lead to a blunted immune response, perhaps requiring less potent immunosuppressive regimens to be administered. There are two fundamental approaches that could be conceived to counteract the effects of IBMIR. The first is based on systemic treatment of the recipient, to prevent coagulation and complement activation at the transplantation site; the second is based on manipulations of the transplanted tissue to minimize its intrinsic characteristics that trigger IBMIR. The advantage of the former is its quite immediate clinical applicability, the primary disadvantage being that a systemic treatment may have generalized side effects (namely bleeding). In a mirrorlike fashion, manipulation of the tissue prior to implant would have the advantage of its localized effect, but the potential disadvantages linked to the quite cumbersome technological approach needed to achieve the goal. Gene therapy approaches, as an example, have intrinsic limitations of safety and efficacy, and there is concern about the immunogenicity of viral-encoded products. In current clinical trials, islets are infused in the patients’ portal system in a solution that contains heparin, and therefore we may claim that, even before the formal demonstration of IBMIR occurrence, our approach was intuitively aimed at the prevention of coagulation in the portal venous system. The effectiveness of heparin, on the other hand, especially in preventing complement activation is less that optimal, and there have been recent proposals that alternative, more efficient compounds should be considered, such as low molecular weight dextran sulfate and melagatran (7, 8). The paper by Dwyer et al. published in this issue of Transplantation (9) convincingly presents evidence on the effectiveness of CD39 expression on islets in the prevention of IBMIR. CD39 is an ectonucleotidase that degrades the platelet agonist adenosine triphosphate (ATP), therefore holding promise as an IBMIR antagonist. The authors show that transgenic mice that express CD39 on islets have normal glucose metabolism, and that their islets, when exposed to human blood, fail to induce coagulation in roughly half of the experiments performed, with an average coagulation time in all experiments that is twice that of the controls. The observation of a normal blood glucose control in the mice that over express human CD39 is key to suggest the viability of this interventional strategy, because ATP levels are of paramount importance in regulating stimulation-secretion coupling in islet beta cells (10), and any strategy aimed at preventing IBMIR has to be void of unwanted effects on beta cell responsiveness to glucose. The donor-recipient combination explored in this paper is xenogeneic (mouse islets and human blood), but it is easy to conceive that comparable efficacy of CD39 over expression should exist in an allogeneic combination. Although overexpression of CD39 on human islets would pose the challenges mentioned earlier (gene therapy approaches), this modification of the tissue prior to transplantation could be easily envisioned as highly beneficial and technically feasible in a context of xenogeneic transplantation where porcine islets are used for human recipients. Generation of transgenic This work was supported by National Institutes of Health grants 5R01 DK25802 and 5U42 RR016603, NIDDK grant 1UO1DK70460, the JDRF, and the DRIF. The Diabetes Research Institute, University of Miami, Miami, FL. Address correspondence to: Camillo Ricordi, M.D., Diabetes Research Institute, University of Miami Miller School of Medicine, 1450 NW 10th Avenue, Miami, FL 33136. E-mail: [email protected]. Received 4 April 2006. Accepted 7 April 2006. Copyright © 2006 by Lippincott Williams & Wilkins ISSN 0041-1337/06/8203-312 DOI: 10.1097/01.tp.0000228879.22020.0f