Islet Yield Remains a Problem in Islet Autotransplantation FREE

Chronic pancreatitis is a severe, debilitating condition that causes severe upper abdominal pain that often is not adequately controlled by opiate analgesia, commonly resulting in narcotic dependency and abuse. Although most patients can be treated conservatively, in a few the pain is unremitting and intractable, thereby destroying their quality of life. In this cohort, it has been recognized for many years that surgery often is the best method of controlling or even removing the symptom of pain.¹ The type of surgery depends on the macroscopic appearance of the pancreatic duct, with drainage procedures used for dilated or obstructed ducts and near-total or subtotal pancreatectomy used for small duct disease. Long-term follow-up has shown good results in terms of pain relief using the rationale outlined in the previous sentences.² However, one of the inevitable sequelae of total pancreatectomy is that the patient is rendered diabetic and insulin dependent; this is a serious complication in any patient group but especially in those with a high incidence of alcohol and other drug abuse.¹

For the best part of a century (since 1902), before the discovery of insulin, attempts were made to treat diabetes with transplantation of pancreatic tissue.^3- 4 Successful prevention of diabetes in pancreatectomized dogs using autotransplantation of unpurified, collagenase-digested pancreatic tissue into the splenic parenchyma was described in the late 1970s.^5- 6 Given these findings in canine and rodent models,^7- 8 results from human islet autotransplantation (IAT) have been less successful than anticipated. Islet autotransplantation in humans has produced variable results, and, despite euglycemia immediately after surgery, there have been many reports of graft failure in the subsequent weeks and months.^9- 11 The principal factor in determining whether IAT is successful in a specific case is islet yield; there seems to be a direct correlation between the number of islets transplanted and the duration of insulin independence. The minimum number of islets for IAT likely to produce insulin independence in patients who undergo near or total pancreatectomy is 200 000 (3000 islets/kg of body weight),^11- 12 which corresponds to the number of islets required to achieve insulin independence after IAT in large animal models.^13- 15 Therefore, it follows that any technical improvements in IAT that increase functional islet yield should have a direct effect on the rate and duration of insulin independence after IAT.

The goal of this review was to identify the various factors affecting the number and percentage of possible islets extracted from the pancreas during removal and preparation of the gland for IAT, with particular reference to the role each plays in determining the final islet yield.

Chronic pancreatitis results in fibrosis and calcification of the gland, which can have a progressively adverse effect on the endocrine and exocrine functions of the pancreas. It has been shown that the proportion of patients with chronic pancreatitis requiring exogenous insulin therapy increases markedly in a short time. A study² of patients with chronic pancreatitis who had undergone drainage procedures but not pancreatic resections revealed an increase from 24% to 60% in the proportion of patients requiring insulin therapy during 5-year follow-up. The likelihood of successful IAT is remote in patients who have impaired glucose tolerance or even diabetes mellitus before surgery.

Some studies^16- 18 have shown a significant reduction in islet yields from cadaveric donors who underwent a long period of hospitalization and malnutrition. As suggested by other authors,¹⁹ it would be reasonable to assume that similar mechanisms are likely to be in effect in the patient with malnourished chronic pancreatitis, thereby reducing the anticipated islet mass available for IAT.

It has long been established that reducing warm ischemia time (WIT) improves the outcome and long-term function of solid organ transplants.^20- 21 In a porcine model, the detrimental effect of WIT has also been shown to affect the mass and quality of the islets acquired by pancreatic digestion and islet isolation.²² The reduction in WIT is usually achieved in cadaveric donors by en bloc infusion of cold organ perfusion solutions (University of Wisconsin solution or Euro-Collins solution) through the femoral vessels or, more commonly, through the aorta and portal vein.²³ Regarding kidney transplantation using a living donor, WIT is kept to a minimum by maintaining the relatively simple and consistent vascular supply to the organ as it is dissected free, with ligation of the vein and artery as a final act before removal of the organ. However, preservation of the vascular supply to the pancreas as it is resected is more difficult given the multiple arterial supply of the gland. This is further complicated in the fibrotic and inflamed pancreas of the patient with chronic pancreatitis by the destruction of the tissue planes and the desire to preserve as much of the surrounding anatomic structure as possible. These 2 considerations can significantly add to the duration of warm ischemia to which the gland is subjected.

Digestion and islet isolation of the human pancreas have been attempted using many techniques, most commonly that described by Ricordi et al,²⁴ which requires intraductal distention of the pancreas with a collagenase-containing solution before atraumatic digestion within an isolated digestion chamber. For optimal islet isolation, it is important that the whole of the gland is distended with collagenase before beginning the digestion process. Theoretically, there is a risk of incomplete and uneven pancreas distention if the pancreatic capsule is breached at the time of resection, again leading to reduced mass of the islet isolate.

In addition to the effect of WIT on the pancreas, duration of cold ischemia and gland preservation has been shown to affect the returns from islet isolation. It has been demonstrated in animal models that as little as 3 hours of cold ischemic preservation (in University of Wisconsin solution, Euro-Collins solution, or silica gel–filtered plasma) can have a detrimental effect on the number of islets isolated and on the rates of insulin independence achieved in transplanted cases.^25- 26 The deleterious effect of cold ischemia has also been demonstrated to be cumulative, with further significant reductions in islet return with increasing storage intervals. Work on human cadaveric pancreata has produced similar findings.²⁷ Although vascular pancreatic grafts can be successfully preserved for up to 72 hours in University of Wisconsin solution,²⁸ there is still considerable uncertainty and debate regarding which of the myriad preservation solutions is actually the best for storage of pancreata before islet isolation.^29- 31 However, regardless of which solution is used, there is a decline in islet yield and successful islet transplantation with increasing preservation time. This reduction in isolated islet mass and demonstrable viability is such that some authors³² argue that it is an inefficient use of time and resources to process pancreata with cold ischemic times greater than 16 hours for IAT.

Most studies of pancreas storage and subsequent islet isolation use glands that have been prepared by intravascular flushing either as an isolated process or as part of a multiple-organ retrieval program. Evidence indicates that intraductal distention at the time of harvest improves the eventual islet yield. This ductal distention can be accomplished by either simple preservation solution²⁷ or a collagenase-containing solution.³³

Fibrosis of the pancreas, which is an inevitable consequence of chronic pancreatitis, also affects the islet isolation process. The intraductal distention of the pancreas might be uneven because of the scarring of the pancreatic parenchyma, which in turn can lead to incomplete and uneven digestion of the pancreas, with the consequent reduction in final islet mass.^34- 35

All the methods described for islet isolation contain essentially the same fundamental steps^24,36- 37: (1) the pancreas is excised from the donor; (2) the gland is digested using a collagenase, with or without mechanical mincing; and (3) the islets are washed by centrifugation. This method produces a pellet of unpurified islets and acinar cells. The principal area for variation in pancreas processing to this point (for a given method) is in the effectiveness of the collagenase in digesting the gland. The collagenase most commonly used is a natural product from the bacteria Clostridium histolyticum and varies markedly in its level of activity,³⁸ thereby affecting the mass of isolated islets potentially available from any given pancreas preparation. In recent years, there has been increasing interest in and use of the enzyme preparation Liberase. This preparation has been shown to produce superior results compared with various collagenases in large animal models and human studies. The findings indicate that not only does the number of islets isolated improve but also their in vitro function is superior compared with islets isolated using crude collagenase preparations.^39- 41 In 1995, Gill et al⁴² demonstrated that Liberase produced no adverse effects at concentrations 10 orders of magnitude greater than would be anticipated in islet isolates, which indicates that there is no apparent safety bar to the continuing use of Liberase in the field of human islet isolation.⁴²

Purification of the islet/acinar pellet is another area of the isolation procedure that is technically demanding and has an enormous impact on the final islet mass. The purifying process is performed using a density gradient (either continuous or discontinuous Ficoll) to separate the contents of the postcollagenase pellet by centrifugation. The end result should produce a layer of islets either at a density boundary (discontinuous Ficoll) or within a small density spectrum (continuous Ficoll). The aforementioned variability in the extent of pancreatic digestion will considerably affect the size and densities of the fragments and therefore the degree of success that will be achieved by density separation. These difficulties in isolation of islets are reflected in the wide variation in islet mass (islet equivalents per gram of pancreas) and purity of islet isolates achieved interinstitutionally and intrainstitutionally.

In animal models, successful IAT has been achieved using a number of transplantation sites: intraportal, intrasplenic, subrenal capsule, and free intraperitoneal dispersal.^7,43- 46 However, in clinical practice it has again been difficult to repeat these levels of success. Human IAT is now focused on the intraportal and intrasplenic sites because these have consistently shown the best results in terms of insulin independence after IAT. There is also some indication that the islet mass required for successful IAT is reduced when the intraportal route is used compared with the intrasplenic route. However, intraportal IAT has a number of significant complications directly due to the IAT procedure: increased intraportal pressure, hepatic venous thrombosis, hepatic infarction, and disseminated intravascular coagulopathy.^47- 49 The increase in portal pressure is thought to be a direct result of the injection of often large volumes (>50 mL) of islet isolate into the portal vein. On the other hand, thrombotic complications are considered by some authors⁴⁹ to be secondary to the action of the pancreatic thromboplastins released by digestion of the exocrine component of the pancreas within the islet isolate.

The complications of intraportal IAT have been reduced by decreasing the volume of fluid in which the islets are suspended, which is possible owing to the increased islet purity of the isolate and therefore smaller volume produced by the gradient Ficoll centrifugation. However, in the process of improving the purity of the islet isolate it is inevitable that there will be some loss in the number of islets in the final preparation. Indeed, the final mass of islets seems to be inversely related to the degree of purity of the same isolate.³⁸ In IAT, this is of particular importance because the initial number of "potentially" available islets found in the fibrosed pancreas of the patient with chronic pancreatitis might well be close to the minimum number of islets required to achieve insulin independence after IAT. The balance between purity and volume of isolate must therefore be carefully considered.

The future of IAT depends on increasing the yield of islets available at the end of the process of excising, digesting, and purifying the pancreas. The conservation of each islet is especially important in IAT because of the limited, and sometimes just sufficient, numbers of islets potentially available from the chronically inflamed and fibrosed gland. The current methods of digestion, although based on the "automated" Ricordi technique, still require a human subjective decision to determine the end point at which the digestion process should be stopped and the islets harvested. There is further potential for islet loss in purification of the isolate, although this has been improved with the use of COBE-style (a centrifuge that allows large volumes to be processed in sterile conditions) cell separators. This is because the basis for gradient purification, the successful cleaving of exocrine cells from the islets to produce the density difference, depends on the function of the enzyme preparation used to digest the pancreas. As mentioned earlier, crude collagenase is inherently inconsistent in its performance. The introduction and increasing use of Liberase as the digestion agent could help remove some of the variability from this step that was previously seen with collagenase use.

The duration of pancreatic ischemia, both warm and cold, must also be considered when attempting to optimize islet returns. There is overwhelming evidence that reducing the interval between removal of the gland and the islet isolation process improves the number and function of the islets collected. Although making up only approximately 3% of the mass of the pancreas, the islets of Langerhans receive 10% of its blood supply. Therefore, minimizing WIT is also likely to improve the islet yield because these highly vascularized and metabolically active units will be among the first cells in the pancreas to be affected by ischemia. This is again especially crucial in IAT because pancreatectomy in these patients is a more technically challenging and time-consuming procedure than in a disease-free gland.

If total pancreatectomy and IAT is to become a more widely used treatment for "end-stage" chronic pancreatitis, all methods of improving postoperative graft function should be explored. It is generally accepted that the mass of transplanted islet cells is the most important indicator of successful outcome after IAT. Of the many areas affecting the final islet yield, several have seen recent improvements and modifications: minimization of cold ischemic time and use of Liberase digestion agents and large-scale COBE-type cell separators. Because of the nature of the disease, little can be done to alter the condition of the gland at the time of resection and processing. However, one area that warrants further investigation is the reduction in WIT during the difficult and often prolonged resection.

Corresponding author and reprints: Guy J. Maddern, FRACS, University of Adelaide, Department of Surgery, The Queen Elizabeth Hospital, Woodville Road, Woodville, SA 5011 Australia (e-mail: guy.maddern@adelaide.edu.au).