Avoiding Damage to Transplanted Human Islets During Implantation Is Important

The article by Korsgren et al. in this issue of Transplantation is a welcome attempt to inject a little more thought in place of activity in the field of clinical islet transplantation. In essence I agree with the points made by Korsgren, so readers who are short on time and looking for a take-home message can stop here. However, I will discuss some interesting aspects of the arguments and observations that Korsgren makes. There can be few who would deny that the last 4 years have seen remarkable change in the field of islet transplantation, with insulin independence and long-term islet allograft function documented in more than a dozen centers worldwide. However, the supply of donor human pancreas remains a problem, islets being in competition with vascularized pancreas transplantation and the scale of operation needed to process pancreases to deliver sufficient islets to routinely produce normoglycemia is daunting, since islets from 2 or 3 donors are usually required. In the United Kingdom these factors have so far been insurmountable and repeatedly successful islet transplantation has not become reality at any United Kingdom centre. Korsgren argues that for the continued success of clinical islet transplantation it is essential that the donor/recipient ratio be reliably reduced to 1:1. I would not demur from that conclusion. Korsgren then uses a combination of local data and historic studies to define the quantity and distribution of islet tissue in the normal human pancreas. This estimate is then compared with the reported islet yields and outcomes for human islet allotransplantation to support an overall conclusion that experienced islet isolation centers achieve yields from individual pancreases which are routinely better than 50% of the islet tissue content and may on occasion approach >90%. Again I find the argument mostly plausible, although I think one should always remember that there is no standardized reference for comparison of islet yields between Centers, the measurements of islet volume being based on the cube of the measurement actually taken and therefore exquisitely sensitive to error, and thus the likelihood of over-estimation bias must be high. I would broadly agree that there is relatively small margin for further improvement in islet isolation efficiency from the “ideal” donor pancreas, although I would estimate the potential for gain to be slightly larger than Korsgren, maybe 10%–20%, which would still be a useful increase. However, there are some donor pancreases where the islet yield is very poor, and all Centers processing human pancreas for islets still experience this variability. The margin for improvement in these cases is much greater, and special treatment of such pancreases would possibly greatly improve the yield of islets regularly obtained, if a way of identifying them predigestion can be found. Korsgren next argues that there is little evidence that adult islet tissue, once implanted and functional, is able to expand or improve its function by anything more than a minor degree. I would agree with that conclusion, adding to it our own unpublished data on primate islet autotransplantation to the intraportal site, where some juvenile animals initially gained excellent glycemia control after total pancreatectomy and islet autotransplantation, but subsequently became glucose intolerant at 1–3 years posttransplantation, associated with a doubling or tripling in body weight. The crux of Korsgren’s argument is that a high proportion of the allografted human islet tissue transplanted is being lost due to involvement in a damaging rapid clotting process that was initially described by Bennett from Korsgrens group in Uppsala (1) and termed the instant blood mediated inflammatory reaction or IBMIR by them. My colleagues in Oxford have confirmed the existence of the phenomenon (2) and found a surprising severe extent of damage, which is in keeping with the scenario Korsgren puts forward. Indeed, I now consider that this is the likely reason for the poor results of our 1990s Oxford series of nine single donor islet allografts performed in diabetic patients who had already undergone successful kidney transplantation for treatment of diabetes-induced renal failure. In some of these patients the amount of islet tissue transplanted approaches that which would be expected to reverse diabetes yet over half the patients showed no detectable C peptide production (unpublished data on file). This finding was previously inexplicable but as no anticoagulant was used during the islet infusion I now think it likely that the islets were largely lost to damage by IBMIR. It is interesting to note that the Edmonton group routinely used heparin infusion intraportally with the islets, which was added because of a previous disastrous intraportal thrombosis in a patient rather than any specific knowledge of IBMIR. Heparin has since been shown to reduce IBMIR (1, 2), although not to completely abrogate it, and would explain the success of the Edmonton protocol using multiple donors, whilst leaving the margin for improvement suggested by Korsgren. Clearly, knowledge of the etiology of IBMIR would help enormously in designing avoidance strategies. However, progress in this field is limited by the fact that so far the demonstrable full blown phenomenon of IBMIR is restricted to fresh allogeneic human blood contacting freshly prepared human islets, and although some damage can be demonstrated in vitro, the major effect is seen only in vivo. Similar phenomena can be demonstrated in nonhuman primate models and possibly in models using pigs, but it is not found in the easily available laboratory species such as the mouse and rat recipient. Indeed the use of a blood clot as a method to aggregate islets for transplantation to sites such as the kidney capsule was originally developed by our group as a reliable technique for transplanting islet tissue which allowed quantitative measurement of islet survival. The IBMIR phenomenon is associated with abnormal rapid clotting, necrosis of the islet tissue, leukocyte infiltration (including polymorphs and monocytes), and complement activation in the absence of antibody binding (although the Oxford group found low level IGG IGM binding in the absence of alloantibody). Why the major difference between species remains unexplained. A reasonable assumption was that the clotting process itself was likely to be the initiator of IBMIR-induced damage via complement activation, and certainly the effects of heparin and complement inhibitory proteins in experimental models are in keeping with this. The Uppsala group identified the initiating cause of the clotting cascade as the expression of tissue factor (TF) (3). But before this explanation is accepted beyond question it should be remembered that certain meaningful control groups have not been performed. For example, it has not been possible to investigate IBMIR in an autogenous human transplantation system because anticoagulation by any method makes the subsequent demonstration of IBMIR unreliable. When cadaveric donor blood is taken at the same time as pancreas donation it has to be anticoagulated during the 4–8 hr minimum taken to prepare the islets, making subsequent in vitro tests for IBMR unreliable. Of course in vivo autogenous islet transplantation is not possible using human islets from cadaveric donors, but islet autotransplantation after total pancreatectomy has been undertaken routinely in some centers and long-term function described (4). Since these patients are already undergoing potentially life-threatening procedures it has not been ethically appropriate to add any further intervention such as biopsy that would allow assessment of IBMIR, and so only the outcome data are available. Nevertheless, the improved function seen in these patients who achieved normoglycemia with remarkably few islets transplanted is suggests that IBMIR may be less severe or even not affect autogenous islets. In keeping with this conclusion is the unique experience of multiorgan transplantation with islet transplantation undertaken in Pittsburgh (5), where remarkably good function was seen with relatively few islets. Lastly, my own studies of islet autotransplantation in the totally pancreatectomized nonhuman primate found normoglycemia to be regularly obtainable after intraportal infusion of approximately 2000 IEQ per kilogram (6, 7), again suggesting IBMIR might not operate in autogenous fresh islet transplantation. This raises the intriguing question of allo-MHC specificity being involved in the mechanism of IBMIR (but not alloantibody as this was excluded) and would suggest IBMR was not just caused by tissue factor expression. Finally, there is an intriguing similarity between the pathology of IBMIR and the described rapid destruction of xenogeneic islets in primate recipients (8) and the role of the IBMIR in the xenograft rejection response need examination. More work is needed to fully define the cause of IBMIR, the effect of agents such as compliment inhibitors and maneuvers such as tissue culture, to allow development of effective strategies for prevention.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.