The current status of pancreatic islet transplantation in Britain

Diabetes implies raised glucose levels but the hallmark of type 1 diabetes is not so much the height of the glucose levels as their instability. It is because of this increased blood glucose variance that attempts to normalise HbA1c result in a substantially increased risk of hypoglycaemia. Because glucose levels cannot be completely normalised, complications cannot be totally prevented.1 In order to achieve near-normal glucose levels, a closed-loop system is necessary whereby the amount of insulin in the circulation is modulated from minute to minute in order to keep glucose levels within a tight and acceptable range. Transplantation of a new population of pancreatic beta-cells can achieve this. However, transplantation necessitates immunosuppression; and immunosuppression carries with it not only the risk of side effects from the drugs themselves but also an increased risk of infections and of cancers. Skin cancers are particularly increased so that, for example, the long-term post-transplant incidence of skin cancer is 41%.2 For this reason, the hope for the past 25 years has been that an artificial pancreas, in which a glucose sensor drives an insulin pump, would supplant beta-cell transplants. However, clinical trials with an intravenous sensor and intraperitoneal (and hence intraportal) insulin delivery were abandoned four years ago because of sensor problems and the system now being developed uses subcutaneous sensing and subcutaneous insulin delivery. Such a system necessarily has a much slower response time3 so that it will be slow to increase insulin levels in response to hyperglycaemia and, perhaps even more importantly, slow to reduce insulin levels when glucose levels fall. Moreover, whereas physiologically both neural influences and the secretion of incretins modulate beta-cell algorithms so that more insulin is secreted at a given glucose level at mealtimes, regulatory authorities have been reluctant to allow ‘meal announcement’ whereby the patient ‘tells' the pump that they are about to eat. For the foreseeable future therefore, the only way to achieve near-normal glucose control is with a transplant. Whole pancreas transplantation began in the mid-1980s following the introduction of ciclosporin and more than 20 000 have been done in the USA alone. Levels of HbA1c are restored to normal, hypoglycaemia is abolished, patients can eat whatever they wish and no longer have to do finger-pricks to monitor their glucose levels. Unsurprisingly, quality of life improves dramatically. Retinopathy has been shown to regress in the short term4 though the longer-term outcome is more controversial.5 The histological changes of diabetic nephropathy are reversed,6 though clinically this may be counterbalanced by the nephrotoxicity of the immunosuppressive drugs, and neuropathic symptoms improve.7 The 10-year survival of patients in end-stage renal failure with a successful pancreas transplant as well as a renal transplant is 75–80% compared to 20–25% in patients whose pancreas transplants fail.8, 9 Thus, in this group of patients, pancreas transplantation is as life-saving as liver transplantation but it takes longer for its benefit to become apparent. The problem with whole pancreas transplantation is the post-operative morbidity. The exocrine pancreas causes a high incidence of post-operative complications, including in particular pancreatitis, graft thrombosis, fistulae and abscesses. Relaparotomy rates between one in seven and one in three are reported.10 Many pancreases are lost within the first month and patients can even die, so that a solitary pancreas transplant may not bring the benefits seen with a simultaneous pancreas-kidney transplant.10, 11 Since the patient with type 1 diabetes does not need the whole pancreas but only the insulin-secreting beta-cells, the idea of separating and transplanting just the beta-cell containing islets is very attractive. Although this was done experimentally in 1972, clinical trials did not begin until Ricordi developed a method for isolating the islets in quantity in the late 1980s. The complications seen with a whole organ transplant do not of course occur, but since the purified islet preparation is injected into the portal vein one has instead the potential risk of portal tributary thrombosis. More commonly, since the patients are heparinised to prevent this, one sees haematomas and haemorrhages from the per-hepatic needle track. To avoid this, some centres report excellent results with injection of the islets into a mesenteric vein under direct vision via a mini-laparotomy12 though one loses the sexiness of a transplant without surgery. Early clinical results were poor with a one-year insulin independence rate of 8–10% (compared to ∼80% with whole pancreas transplants).13 As a result, only a few hundred were done. However, in at least 50% of patients, C-peptide became positive, implying that beta-cells did survive but not in sufficient numbers to normalise glucose levels. Sadly, no data were gathered at the time on the effect on HbA1c or other indices of diabetic control and insulin independence was the only outcome measure. In 2000, out of the blue, James Shapiro in Edmonton reported 100% insulin independence at one year in seven patients.14 He had shown in animals that the combination of steroids with a calcineurin inhibitor, as used in standard immunosuppressive protocols, was toxic to islets. Instead of just modifying the protocol accordingly, he changed it completely. Induction was with an interleukin-2 receptor antagonist instead of anti-thymocyte globulin (ATG). Maintenance immunosuppression was still with tacrolimus but in low doses and steroids were replaced by sirolimus. No azathioprine or mycophenolate mofetil (MMF) was used. As if that were not enough, he also transplanted non-uraemic patients, targeted diabetic instability and, perhaps crucially, was prepared to take advantage of the relative safety of the procedure to use two and even three transplants per patient. Eight years on where are we? Some 50 centres around the world have discovered for themselves just how difficult it is to match the Edmonton results.15 With larger numbers of patients transplanted, the one-year insulin independence rate has fallen to 80%, which is comparable to the results of whole organ transplantation. Sadly, however, whereas whole pancreas transplants are lost only slowly thereafter, insulin independence after islet transplantation falls quite steeply to ∼50% at three years and little more than 10–15% at five years.16 However, patients who have to restart insulin still have C-peptide secretion and have levels of HbA1c only just above the normal range. In other words, they behave rather like newly diagnosed patients in the honeymoon period. Crucially, since they no longer have hypos their quality of life improves dramatically. Given that they were selected on the basis of their diabetic instability, frequent hypoglycaemia and hypoglycaemic unawareness in the first place, those of us in the field would argue that this represents a successful outcome. Although one cannot assume that long-term effects will be the same as with whole pancreas transplants, the Milan group has shown beneficial effects of islet transplantation on multiple physiological parameters, as well as improved quality of life.17 Seven-year patient survival is 90% following a kidney and a successful islet transplant compared to only 51% if the islet graft fails. They speculate that the benefit may be due to restoration of circulating C-peptide. If one typically has to process at least two pancreases to get a transplantable preparation, and patients need two transplants each, clinical islet transplantation currently has to be regarded as an extravagant procedure. Nevertheless, the National Commissioning Group (NCG), which has agreed to fund the programme in the UK, calculates that it will be worth it if it obviates the cost of acute emergency treatment of recurrent, severe and often asymptomatic hypoglycaemia. We might add that we have had two patients die of asymptomatic hypoglycaemia on our islet transplant waiting list, which at least reassures us that we are selecting the right patients. Islet transplantation is also indicated in insulin-dependent patients who are already immunosuppressed, most of whom will have had renal transplants, since the risk:benefit ratio is acceptably low. In the UK, three centres have successfully performed islet transplants: King’ s and the Royal Free in London, and Oxford. The NCG funded programme is such that the two London centres collaborate as one, and it also includes three peripheral centres (Bristol, Manchester and Newcastle) which will transplant islets isolated in London or Oxford. At present, there is no experience in Britain of transplanting islets isolated in another centre but other groups, notably Geneva and Uppsala, have been operating hub and spoke systems successfully for many years.18 What of the future? In the first place, we need a source of beta-cells in quantity. Even if the surgeons could totally abolish the post-operative morbidity associated with whole pancreas transplantation, the 600 multi-organ donors per year in Britain would solve the problem for only a small proportion of the 200 000 patients with type 1 diabetes. For islet transplantation the arithmetic is currently even worse but in the future, if beta-cells can be produced in the lab, it will not be whole pancreases which are transplanted. Stem cells get a lot of publicity but are still many years away and the problem is that, if you were to transplant a patient with type 1 diabetes with a population of beta-cells identical to those they were born with, they would destroy them for the same reason they destroyed their native beta-cells. Type 1 diabetes is an auto-immune disease. However, the idea of genetically modifying stem-cell derived beta-cells may not be so far-fetched and animals genetically predisposed to develop auto-immune diabetes do not do so if the glutamic acid decarboxylase (GAD) gene is silenced.19 Xenografts from genetically modified animals represent a possible alternative. Secondly, we need to understand the cause of long-term graft failure following islet transplantation. Islet transplantation is very unusual in that one is using immunosuppression not only to prevent graft rejection but also to prevent disease recurrence. The contribution of auto-immunity to graft failure is unknown but pancreatectomised patients given an islet allograft do better than those with type 1 diabetes.13 Furthermore, it has recently been shown that patients with T cells activated against GAD and IA-2 at the time of their transplant never come off insulin (compared to a 90% insulin independence rate if activated T cells are not present).20 Interestingly, antibody levels are not predictive. An alternative explanation is suggested by the fact that pancreatectomised patients given islet autografts (who need no immunosuppression) do better than those given allografts.13 Our current drug protocols are diabetogenic in animal models21 and the very high portal vein drug levels following an oral dose are often well into the toxic range.22 We may quite simply be poisoning the islets we transplant. We need either to use different drugs or to change the route of drug delivery, or to transplant the islets into a different site. An alternative transplantation site might have other potential advantages. We have recently shown that even in a syngeneic model 45–60% of islets fail to engraft successfully following intraportal transplantation.23 Islets are integrated ‘cities' of several thousand cells with their own blood and nerve supply. They have a very high oxygen requirement and the hypoxic environment of the portal vein is probably the last place they should be transplanted. Shapiro’ s 2000 landmark paper is not the end of the story but it may hopefully be the start of the last chapter. Different centres are studying different immunosuppressive protocols (including some which are more effective against auto-immunity) as well as other ways of improving clinical outcomes. Minneapolis, using multiple modifications of the Edmonton protocol, recently reported a 50% five-year insulin independence rate following a single transplant.24 Edmonton is now reporting a 100% insulin independence rate in a small series using Campath based immunosuppression.25 But this is where we came in … The NCG is being far-sighted in agreeing to finance a procedure for which there remain so many unanswered questions. However, having now the platform on which to build the necessary research puts Britain in a strong position to join other international centres in studies aimed at making single donor islet transplants the norm and improving the long-term results so that far more patients can benefit from this important new clinical procedure.

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.