SPECIAL FOCUS y 10-year anniversary issue
Review
Expert Review of Clinical Immunology Downloaded from informahealthcare.com by Nyu Medical Center on 01/12/15 For personal use only.
Islet transplantation for Type 1 diabetes: where are we now? Expert Rev. Clin. Immunol. 11(1), 59–68 (2015)
Aaron J Ahearn‡, Justin R Parekh‡ and Andrew M Posselt* Department of Surgery, Transplant Division, University of California, Room M896, 505 Parnassus Avenue, San Francisco, CA, 94143, USA *Author for correspondence: Tel.: +1 415 353 1473 Fax: +1 415 353 8709 [email protected] ‡
Authors contributed equally
Pancreatic islet transplantation is a minimally invasive procedure that can restore normoglycemia and insulin independence in Type 1 diabetics without the surgical complications associated with vascularized pancreas transplantation. The advances made in this field over the past decade have dramatically improved patient outcomes, and the procedure is now transitioning from an experimental treatment to a clinical reality. Nonetheless, a number of important issues continue to hamper the success of islet transplantation and must be addressed before there is widespread clinical acceptance. These include the relative inefficiency of the islet isolation process, the progressive loss of islet function over time and the need for multiple donors to achieve insulin independence. Here, we discuss the current status of islet transplantation and examine its future as a treatment for Type 1 diabetes. KEYWORDS: glycemic control • hypoglycemic unawareness • immunosuppression • insulin independence • islet transplantation • transplant site • Type 1 diabetes
The first reports of islet transplantation in animals in 1972 generated tremendous hope for future human application and the first reports of successful human islet transplantation in 1989 by both Warnock and Scharp only bolstered that enthusiasm [1–3]. However, it has taken more than 20 years to make islet transplantation a viable treatment for Type 1 diabetes. As recently as 1999, the International Islet Registry data reported that 1-year insulin independence rates after islet transplantation were less than 10%, and rates of primary nonfunction were as high as 70–80% [4]. Expectations changed dramatically when the results of the landmark paper by Shapiro et al. reported seven consecutive patients who became insulin independent after islet transplantation and maintained independence for more than 1 year [5]. Since Shapiro’s report in 2000, islet transplantation has continued to grow and make improvements. According to the Collaborative Islet Transplant Registry, which has accumulated data on more than 80% of islet transplants performed in North American, European and Australian centers, there have been more than 800 clinical islet transplants performed to date. The number of islet transplants peaked at more than 100 per year in 2002 but then steadily fell
informahealthcare.com
10.1586/1744666X.2015.978291
until 2008 when improvements in collagenase products and the ongoing Clinical Islet Transplantation (CIT) consortium trials led to an increase in activity, reaching approximately 75 islet transplants performed in 2010 [6]. Ongoing support from the US Congress has also helped to maintain this recent resurgence in islet transplantation. Currently, a portion of government funding for diabetes research is allocated towards islet transplant research, and there is a mandate that the Centers for Medicare and Medicaid cover costs of Medicare patients enrolled in islet transplant trials; support that has been crucial to advancing the field [7]. Ongoing research efforts largely fall under the CIT – a consortium of academic centers (nine in the US, one in Canada and one in Sweden) that initiated a series of pilot studies in 2004 to improve engraftment and long-term insulin independence [8]. The CIT has also initiated two Phase III registration trials to demonstrate that islet transplantation can lead to improved glycemic control and prevent hypoglycemic events in Type 1 diabetic patients with severe hypoglycemia and glycemic lability (CIT-07) and in Type 1 diabetic patients who had received a kidney transplant (CIT-06). Both studies have completed enrollment and
2015 Informa UK Ltd
ISSN 1744-666X
59
Review
Ahearn, Parekh & Posselt
pancreas. Under carefully controlled clinical good manufacturing practice conditions, the pancreatic tissue must be Islet transplant donor Transport digested using a mix of collagenase and container protease enzymes in such a fashion that Digestion Sampling Digested the acinar cells are dispersed without port islets damaging the islets. Islets are then isoDigestion lated based on processes derived from the chamber Temp. monitor work of Ricordi et al. and further modified by Lakey et al., who described the controlled delivery of enzyme solutions Shaker into the pancreatic duct to facilitate Heater digestion of acinar tissue [10–12]. (FIGURE 1) The reproducible digestion of pancreatic COBE 2991 Immunosuppressive exocrine tissue to allow for subsequent and engraftment Dilution islet isolation is one of the most imporPump Recirculation therapy solution tant steps in islet manufacture. Success Purified has historically been limited by batch-toIslet transplant islets batch variability of the collagenase recipient Exocrine Purified islets enzyme formulations and narrow activity tissue in culture parameters outside which islets are Immediate Islet isolation degraded along with acinar tissue [13]. Islet transplant, Density, gradient transplantation Success of islet preparation is generally culture, or centrifugation long distance measured by quantification of islet equivIslet in portal vein transportation alents (IEQ). IEQs are determined by staining islet cell preparations with dithiFigure 1. Human pancreatic islet manufacturing and transplantation. Islets are zone to discriminate islets from exocrine isolated from a deceased donor pancreas using a multi-step process and infused directly tissue. Next islets are measured by light into the portal vein of the recipient. The islets lodge in the small portal venules and microscopy to estimate their size within eventually begin to make insulin in response to blood glucose levels. 50 mm. The IEQ is then determined by Reprinted with permission from [81] The Biochemical Society (2006). multiplying the number of islets by a cormost participants have reached the primary endpoint [9]. These rection factor to convert these diameters [14]. Although some registration trials are extremely important for the field since the studies have questioned the consistency of this method [15], results will be used to obtain biologic licensure by the US FDA other measurements have been proposed (Han-Hung Huang that, in turn, will help move islet transplantation from being a Acta Diabetol 2013 50). IEQs remain the de facto system of research protocol to a clinically acceptable procedure that is reim- standardization for islet cell preparations and for standardizabursable by third-party insurance. tion across clinical trials. This standard has allowed the field to Overall, in the last decade and a half, the field of islet trans- evolve its islet preparations to maximize clinical results. plantation has made significant strides in achieving insulin indeIn 1995, a highly purified formulation of collagenase (Liberpendence as well as reducing hypoglycemia events and some of aseTM HI) became available and was found to be very effective the secondary complications of diabetes. These clinical improve- in dispersing acinar tissue without injuring the islets [16]. This ments have been the direct result of maximizing the isolation, product quickly became the preferred blend for clinical islet engraftment and preservation of islet cell mass. Nonetheless, a isolation by many islet transplant centers [5,13,17]. Unfortunately, number of important issues continue to hamper the success of the enzyme was manufactured using neural tissue and other the procedure and must be addressed before there is widespread specified risk materials from cattle, and its use in clinical isolaclinical acceptance. These include the relative inefficiency of the tions was discontinued in 2007 due to concerns about potential islet isolation process, the loss of more than 60% of the islet mass transmission of bovine spongiform encephalopathy. The immediately after intraportal infusion as a result of the instant enzyme formulations that subsequently became available blood-mediated inflammatory response (IBMIR), the progressive included the SERVATM blend (a mixture of a controlled cliniloss of islet function over time, and the inability to accurately cal good manufacturing practice grade collagenase and neutral monitor islet allografts for rejection. protease; SERVA Electrophoresis GMBH, Heidelberg, Germany) and other collagenase blends made without bovine prodImprovements in pancreatic islet isolation technology ucts or with recombinant technology [18,19]. The in vitro To obtain islet preparations that are acceptable for clinical use, biochemical activities of these products appear to be similar to the islets must be separated for the acinar tissue of the donor LiberaseTM HI, but the parameters for optimal use are different
Expert Review of Clinical Immunology Downloaded from informahealthcare.com by Nyu Medical Center on 01/12/15 For personal use only.
Pancreas donor
60
Expert Rev. Clin. Immunol. 11(1), (2015)
Expert Review of Clinical Immunology Downloaded from informahealthcare.com by Nyu Medical Center on 01/12/15 For personal use only.
Islet transplantation for Type 1 diabetes
Review
and many experienced centers had inconGMP facility sistent results using these products for clinical islet isolation. The CIT consortium addressed this serious problem by launching a concerted effort to establish the optimal parameters for its use and islet quality and yields are now consistently good. Once digested, the islets must be isolated from the exocrine tissue using either a Biocoll-based gradient or other similar gradient system [20]. Recently, the COBE2991 cell apheresis centrifuge system has become the preferred method for separating the islets from the pancreatic digest (FIGURE 1). The purity of the islets obtained after this preparation is central to the success of the procedure because impure preparations do not funcFigure 2. Steps in the islet manufacturing process. Top row: (A) islets are isolated tion as well as purified preparations, and in a good manufacturing practice facility using stringent protocols. (B) Enzymatic digesinfusion of larger tissue volumes assocition in a temperature controlled chamber is the first step in the isolation process. (C) ated with impure preparations can cause Islets are separated from exocrine tissue using gradient centrifugation. Bottom row: appearance of the islet product at various stages of the manufacturing process. (A) elevated portal pressures and increase the Donor pancreas at the beginning of digestion. (B) Islets (stained red) with contaminating risk of portal vein thrombosis. The develexocrine tissue (C) The purified islet product. (D) Settled islet product in culture medium. opment of standardized density gradients The islets comprise the light colored pellet at the bottom of the tube. has dramatically improved the efficacy of Reprinted with permission from [82] American Diabetes Association (2013). this step, and there is relatively little islet loss. Once purified, islets can be cultured for 24–72 h after isolation to improve purity and allow com- Islet delivery & transplant site pletion of viability assays [21]. This in vitro culture also serves as Once the islets are ready for transplantation, they are infused directly into the recipient’s portal venous system. (FIGURE 2) The viability assays as some islets will not survive this brief culture. In addition to the purification and isolation protocols, intraportal site is the most commonly used site for human islet advances in donor selection and pancreas procurement play a transplantation, mainly because this site has been extensively crucial role in the overall outcomes. Most centers with expe- studied in animal models of islet transplantation and the infusion rience in the field now report that islet yields are highest process is relatively straightforward and well tolerated. During when younger donors (2.5 mg/dl or need for transfusion or surgery) after percuta9000–10,000 IEQ/kg and thus patients who receive signifi- neous islet transplantation was 9%, though this was subsequently cantly less than this amount are prepared for preemptive reduced by avoiding aspirin therapy before transplant and additional transplants that are performed as early as 2 months improved ablation of the catheter tract with Gelfoam and metal after the first infusion [22]. This approach is thought to mini- coils [23,24]. Other alternative approaches include a minimize graft attrition due to high metabolic activity and thus laparotomy in the operating room and to infuse the islets directly into a portal vein tributary [25] and recanulization of the umbilical improve long-term function. informahealthcare.com
61
Expert Review of Clinical Immunology Downloaded from informahealthcare.com by Nyu Medical Center on 01/12/15 For personal use only.
Review
Ahearn, Parekh & Posselt
vein [2]. The risk of bleeding is significantly less, but the procedure requires general anesthesia and abdominal surgery, both of which can have complications. With either approach, the need for systemic anticoagulation after the transplant places these patients at higher risk of bleeding and they must be monitored accordingly. Hypotension unrelated to bleeding can occur during the islet infusion and is thought to result from the inflammatory response caused by the intravascular inoculum. It is generally self-limited and responds to fluid resuscitation and transient vasopressor administration [26]. The incidence of portal vein thrombosis has decreased as a result of improved purity of the islet inoculum, and is estimated to occur in about 3% of patients [27]. The thrombosis is usually partial and resolves with long-term anticoagulation, but there is a report of complete portal vein thrombosis in a combined islet–liver transplant recipient who required emergent retransplantation of the liver [28]. Hepatic steatosis is seen in as many as 20% of intraportal islet recipients and is thought to result from increased insulin levels in the periislet parenchyma. The clinical relevance of this finding is not known at present [29–32]*add 3 more refs. In addition to these procedural issues, there are a number of physiologic reasons that make the liver a suboptimal site for islet transplantation. First, oxygen content of portal blood is very low (15–20 mmHg), and this relative hypoxia can compromise the viability and survival of the infused islets [33,34]. This hypoxic state is further exacerbated by the local stagnation of portal blood that develops in the small portal venous tributaries after islet transplantation and persists for several days. Second, intravenous infusion of the islets exposes them to a severe non-specific inflammatory response called the IBMIR that rapidly destroys much of the islet inoculum [35,36]. IBMIR develops when tissue factor expressed on the islet surface is exposed to blood, triggering platelet aggregation, complement activation and leukocyte infiltration, all of which contribute to accelerated islet loss. Third, the location of the islets directly within the portal system exposes them to high levels of immunosuppressive medications being absorbed from the gut, and many of these, such as tacrolimus and sirolimus have direct beta cell toxicity [37]. Together, these factors are thought to be responsible for a loss of 50–75% of the islet inoculum in the first several days after transplantation, and are an important reason why most patients require islets from multiple donors to achieve independence [35,36]. Several approaches have been tried to minimize early destruction of intraportally placed islets, including strategies to promote anticoagulation such as systemic therapy with heparin or low-molecular-weight dextran and coating of the islet surface with heparin or urokinase [38–40]. In addition, a number of agents have been used to inhibit the inflammatory response: most promising are fusion proteins that inhibit the action of the inflammatory cytokines IL-1b and TNF-a [41–43]. One such agent, the soluble TNF-a fusion protein etanercept is currently considered standard of care in most protocols [44]. Although these approaches have improved engraftment and early islet survival, the overall clinical outcomes are only modestly better, with most patients requiring 62
two or more islet infusions to achieve independence and functional graft loss occurring in 50–60% of patients within 2–3 years after transplantation [45]. Although several anatomical sites other than the liver have been tried in animal models with varying success (renal subcapsule, testicle, thymus, anterior chamber of the eye, peritoneum and subcutaneous space), most are impractical or hazardous to use in the clinical setting or require islet numbers significantly greater than those obtainable during routine clinical isolations [46,47]. Initial attempts to transplant islet cells failed to achieve insulin independence [48]. However, more recent studies have demonstrated success of transplantation of islets in mice and humans [49,50]. In particular, Maffi et al. have successfully transplanted autologous pancreatic islet cells in four patients who underwent total pancreatectomy with detectable C-peptide levels at 944 days. Another site with potential clinical utility is the gastric or intestinal submucosal space. This site is attractive because it is well vascularized, accessible endoscopically and does not expose the islets to IBMIR. Encouraging results have been reported in rodent and large animal (pig) models, with adequate graft function and evidence of improved glucose control compared with the intraportal route [51,52]. Recently, Wzola et al. described the first successful case of gastric submucosal islet autotransplantation in a patient with chronic pancreatitis. This patient was not a candidate for intraportal infusion due to preexisting partial portal vein thrombosis. The patient underwent a pancreatectomy and approximately 250,000 IEQ (3700 IEQ/kg) were isolated from the pancreas. These were injected with endoscopic guidance into 18 gastric submucosal sites at the time of surgery. There were no complications related to the procedure. Remarkably, the patient’s pre-pancreatectomy and 30-day post-pancreatectomy fasting and glucose-stimulated C-peptide levels were similar, and the patient did not require any insulin or oral hypoglycemic agents after the procedure [53]. Additional studies are needed to determine whether this site has broader clinical applicability. Advances in immunosuppressive strategies
Although significant progress has been made in optimizing islet isolation, most of the improvements in clinical outcomes are the direct result of refinements in the treatment of the autoand alloimmune responses to the islet grafts. The observation that islet engraftment and duration of insulin independence are significantly better in patients undergoing islet autotransplantation versus allotransplantation suggests that immunologic and immunosuppressive drug-related effects play an important role in mediating graft loss [54] For example, some of early islet loss observed with protocols using non-depleting induction agents was likely due to incomplete suppression of alloreactivity at the time of transplant [55]. In addition to this early loss, long-term function of the remaining islets was likely compromised by the direct toxicity of some maintenance immunosuppressive agents. Calcineurin inhibitors (CNIs), prednisone [5,56,57] and even mycophenolic acid have all been shown to have a toxic effect on islet cells [5,56–58]. Expert Rev. Clin. Immunol. 11(1), (2015)
Expert Review of Clinical Immunology Downloaded from informahealthcare.com by Nyu Medical Center on 01/12/15 For personal use only.
Islet transplantation for Type 1 diabetes
To address these issues, new immunosuppressive protocols were developed that included potent T-cell depleting drugs and minimizing the use of agents toxic to islets. Hering et al. was the first to show that T cell depletion (initially with OKT3 and subsequently with polyclonal anti-lymphocyte antibody thymoglobulin) at the time of transplant, along with ‘islet-sparing’ maintenance immunosuppression based on sirolimus and low-dose tacrolimus substantially improved engraftment and long-term independence rates [55]. Alemtuzumab, an anti-CD52, T cell depleting antibody, has also shown promise as an induction agent, further strengthening the argument that T cell depletion at the time of islet infusion may improve long-term outcomes [59]. Simultaneous with this development came the routine use of etanercept. As discussed above, the rationale behind this agent is that TNF-a has significant in vitro toxicity to islets, and administration of TNF-a inhibitors improves islet engraftment in both murine and primate models. Thus, in 2005, Hering et al. published the results of eight patients treated with a combination of thymoglobulin and etanercept for induction. This trial showed that this therapy was safe and demonstrated 100% insulin independence after a single islet transplant. Subsequent comparisons confirm that T cell depleting regimens in conjunction with etanercept outperform both thymoglobulin only and daclizumab induction in terms of both 3 and 5 year rates of insulin independence (FIGURE 3) [60]. In addition to increasing the level of immunosuppression at the time of islet administration, development of protocols that minimize exposure to prednisone and CNIs is crucial to improving long-term islet survival. Both CNIs and steroids have been shown to have direct islet toxicity and a number of studies have shown that steroid withdrawal and minimization of CNI exposure with the addition of an m-TOR inhibitor such as sirolimus improves overall outcomes [5,37,56,61]. The use of m-TOR inhibitors is limited by the development of proteinuria in a number of patients. Several centers have taken this approach one step further by developing immunosuppressive protocols consisting of induction with thymoglobulin followed by maintenance immunosuppression that avoids CNIs and steroids, and instead relies on novel monoclonal antibodies against specific T cell antigens. The antibodies include efalizumab (Raptiva) that is directed against the anti-leukocyte functional antigen-1 and affects T cell migration and activation, and belatacept (LEA29Y, Nulojix) that inhibits the costimulatory signal needed to activate T cells [62,63]. These biologic agents were selected for their potent immunosuppressive properties, the absence of beta cell and renal toxicity, and lack of many of the side effects observed with other currently employed drugs. Results from these relatively small studies are promising, with high rates of insulin independence after single-donor islet transplantation and insulin independence rates exceeding 50% at 5 years. These results are comparable to those seen after pancreas transplantation in non-uremic Type 1 diabetic patients [64]. Though efalizumab was very well tolerated in the islet patients, it was linked to the development of progressive multifocal informahealthcare.com
Review
Tacrolimus (low dose) Sirolimus Etanercept Anti-thymocyte globulin
–2
–1
Donor islet isolation
0
+1
+2
+7
+10
+28
Days after transplant Islet transplant
Islet culture
Figure 3. Immunosuppressive protocol associated with improved islet engraftment, better long-term islet function and increased incidence of insulin independence after single transplants. T cell depleting induction immunosuppression and etanercept are given during the peritransplant period. Sirolimus and low-dose tacrolimus are used for maintenance therapy.
leukoencephalopathy in a small number of patients receiving the drug as treatment for psoriasis and was voluntarily withdrawn from the market by its manufacturer. The drug was subsequently discontinued in all islet patients, but surprisingly, most retained graft function despite relatively low levels of immunosuppression that consisted primarily of sirolimus and mycophenolate [65]. Encouraging results were also observed with belatacept, and a number of centers continue to use it as a maintenance agent for islet as well as kidney transplantation [64]. Indications for islet transplantation
While allogeneic islet transplantation is a well-tolerated procedure and can provide excellent glucose control, it requires long-term immunosuppression that carries significant risks. Consequently, islet transplantation alone is currently limited to adult patients who have difficulties in controlling diabetes despite optimization of medical therapy. Specifically, potential recipients must be Type 1 diabetics with no detectable C-peptide (
Review
Expert Review of Clinical Immunology Downloaded from informahealthcare.com by Nyu Medical Center on 01/12/15 For personal use only.
Islet transplantation for Type 1 diabetes: where are we now? Expert Rev. Clin. Immunol. 11(1), 59–68 (2015)
Aaron J Ahearn‡, Justin R Parekh‡ and Andrew M Posselt* Department of Surgery, Transplant Division, University of California, Room M896, 505 Parnassus Avenue, San Francisco, CA, 94143, USA *Author for correspondence: Tel.: +1 415 353 1473 Fax: +1 415 353 8709 [email protected] ‡
Authors contributed equally
Pancreatic islet transplantation is a minimally invasive procedure that can restore normoglycemia and insulin independence in Type 1 diabetics without the surgical complications associated with vascularized pancreas transplantation. The advances made in this field over the past decade have dramatically improved patient outcomes, and the procedure is now transitioning from an experimental treatment to a clinical reality. Nonetheless, a number of important issues continue to hamper the success of islet transplantation and must be addressed before there is widespread clinical acceptance. These include the relative inefficiency of the islet isolation process, the progressive loss of islet function over time and the need for multiple donors to achieve insulin independence. Here, we discuss the current status of islet transplantation and examine its future as a treatment for Type 1 diabetes. KEYWORDS: glycemic control • hypoglycemic unawareness • immunosuppression • insulin independence • islet transplantation • transplant site • Type 1 diabetes
The first reports of islet transplantation in animals in 1972 generated tremendous hope for future human application and the first reports of successful human islet transplantation in 1989 by both Warnock and Scharp only bolstered that enthusiasm [1–3]. However, it has taken more than 20 years to make islet transplantation a viable treatment for Type 1 diabetes. As recently as 1999, the International Islet Registry data reported that 1-year insulin independence rates after islet transplantation were less than 10%, and rates of primary nonfunction were as high as 70–80% [4]. Expectations changed dramatically when the results of the landmark paper by Shapiro et al. reported seven consecutive patients who became insulin independent after islet transplantation and maintained independence for more than 1 year [5]. Since Shapiro’s report in 2000, islet transplantation has continued to grow and make improvements. According to the Collaborative Islet Transplant Registry, which has accumulated data on more than 80% of islet transplants performed in North American, European and Australian centers, there have been more than 800 clinical islet transplants performed to date. The number of islet transplants peaked at more than 100 per year in 2002 but then steadily fell
informahealthcare.com
10.1586/1744666X.2015.978291
until 2008 when improvements in collagenase products and the ongoing Clinical Islet Transplantation (CIT) consortium trials led to an increase in activity, reaching approximately 75 islet transplants performed in 2010 [6]. Ongoing support from the US Congress has also helped to maintain this recent resurgence in islet transplantation. Currently, a portion of government funding for diabetes research is allocated towards islet transplant research, and there is a mandate that the Centers for Medicare and Medicaid cover costs of Medicare patients enrolled in islet transplant trials; support that has been crucial to advancing the field [7]. Ongoing research efforts largely fall under the CIT – a consortium of academic centers (nine in the US, one in Canada and one in Sweden) that initiated a series of pilot studies in 2004 to improve engraftment and long-term insulin independence [8]. The CIT has also initiated two Phase III registration trials to demonstrate that islet transplantation can lead to improved glycemic control and prevent hypoglycemic events in Type 1 diabetic patients with severe hypoglycemia and glycemic lability (CIT-07) and in Type 1 diabetic patients who had received a kidney transplant (CIT-06). Both studies have completed enrollment and
2015 Informa UK Ltd
ISSN 1744-666X
59
Review
Ahearn, Parekh & Posselt
pancreas. Under carefully controlled clinical good manufacturing practice conditions, the pancreatic tissue must be Islet transplant donor Transport digested using a mix of collagenase and container protease enzymes in such a fashion that Digestion Sampling Digested the acinar cells are dispersed without port islets damaging the islets. Islets are then isoDigestion lated based on processes derived from the chamber Temp. monitor work of Ricordi et al. and further modified by Lakey et al., who described the controlled delivery of enzyme solutions Shaker into the pancreatic duct to facilitate Heater digestion of acinar tissue [10–12]. (FIGURE 1) The reproducible digestion of pancreatic COBE 2991 Immunosuppressive exocrine tissue to allow for subsequent and engraftment Dilution islet isolation is one of the most imporPump Recirculation therapy solution tant steps in islet manufacture. Success Purified has historically been limited by batch-toIslet transplant islets batch variability of the collagenase recipient Exocrine Purified islets enzyme formulations and narrow activity tissue in culture parameters outside which islets are Immediate Islet isolation degraded along with acinar tissue [13]. Islet transplant, Density, gradient transplantation Success of islet preparation is generally culture, or centrifugation long distance measured by quantification of islet equivIslet in portal vein transportation alents (IEQ). IEQs are determined by staining islet cell preparations with dithiFigure 1. Human pancreatic islet manufacturing and transplantation. Islets are zone to discriminate islets from exocrine isolated from a deceased donor pancreas using a multi-step process and infused directly tissue. Next islets are measured by light into the portal vein of the recipient. The islets lodge in the small portal venules and microscopy to estimate their size within eventually begin to make insulin in response to blood glucose levels. 50 mm. The IEQ is then determined by Reprinted with permission from [81] The Biochemical Society (2006). multiplying the number of islets by a cormost participants have reached the primary endpoint [9]. These rection factor to convert these diameters [14]. Although some registration trials are extremely important for the field since the studies have questioned the consistency of this method [15], results will be used to obtain biologic licensure by the US FDA other measurements have been proposed (Han-Hung Huang that, in turn, will help move islet transplantation from being a Acta Diabetol 2013 50). IEQs remain the de facto system of research protocol to a clinically acceptable procedure that is reim- standardization for islet cell preparations and for standardizabursable by third-party insurance. tion across clinical trials. This standard has allowed the field to Overall, in the last decade and a half, the field of islet trans- evolve its islet preparations to maximize clinical results. plantation has made significant strides in achieving insulin indeIn 1995, a highly purified formulation of collagenase (Liberpendence as well as reducing hypoglycemia events and some of aseTM HI) became available and was found to be very effective the secondary complications of diabetes. These clinical improve- in dispersing acinar tissue without injuring the islets [16]. This ments have been the direct result of maximizing the isolation, product quickly became the preferred blend for clinical islet engraftment and preservation of islet cell mass. Nonetheless, a isolation by many islet transplant centers [5,13,17]. Unfortunately, number of important issues continue to hamper the success of the enzyme was manufactured using neural tissue and other the procedure and must be addressed before there is widespread specified risk materials from cattle, and its use in clinical isolaclinical acceptance. These include the relative inefficiency of the tions was discontinued in 2007 due to concerns about potential islet isolation process, the loss of more than 60% of the islet mass transmission of bovine spongiform encephalopathy. The immediately after intraportal infusion as a result of the instant enzyme formulations that subsequently became available blood-mediated inflammatory response (IBMIR), the progressive included the SERVATM blend (a mixture of a controlled cliniloss of islet function over time, and the inability to accurately cal good manufacturing practice grade collagenase and neutral monitor islet allografts for rejection. protease; SERVA Electrophoresis GMBH, Heidelberg, Germany) and other collagenase blends made without bovine prodImprovements in pancreatic islet isolation technology ucts or with recombinant technology [18,19]. The in vitro To obtain islet preparations that are acceptable for clinical use, biochemical activities of these products appear to be similar to the islets must be separated for the acinar tissue of the donor LiberaseTM HI, but the parameters for optimal use are different
Expert Review of Clinical Immunology Downloaded from informahealthcare.com by Nyu Medical Center on 01/12/15 For personal use only.
Pancreas donor
60
Expert Rev. Clin. Immunol. 11(1), (2015)
Expert Review of Clinical Immunology Downloaded from informahealthcare.com by Nyu Medical Center on 01/12/15 For personal use only.
Islet transplantation for Type 1 diabetes
Review
and many experienced centers had inconGMP facility sistent results using these products for clinical islet isolation. The CIT consortium addressed this serious problem by launching a concerted effort to establish the optimal parameters for its use and islet quality and yields are now consistently good. Once digested, the islets must be isolated from the exocrine tissue using either a Biocoll-based gradient or other similar gradient system [20]. Recently, the COBE2991 cell apheresis centrifuge system has become the preferred method for separating the islets from the pancreatic digest (FIGURE 1). The purity of the islets obtained after this preparation is central to the success of the procedure because impure preparations do not funcFigure 2. Steps in the islet manufacturing process. Top row: (A) islets are isolated tion as well as purified preparations, and in a good manufacturing practice facility using stringent protocols. (B) Enzymatic digesinfusion of larger tissue volumes assocition in a temperature controlled chamber is the first step in the isolation process. (C) ated with impure preparations can cause Islets are separated from exocrine tissue using gradient centrifugation. Bottom row: appearance of the islet product at various stages of the manufacturing process. (A) elevated portal pressures and increase the Donor pancreas at the beginning of digestion. (B) Islets (stained red) with contaminating risk of portal vein thrombosis. The develexocrine tissue (C) The purified islet product. (D) Settled islet product in culture medium. opment of standardized density gradients The islets comprise the light colored pellet at the bottom of the tube. has dramatically improved the efficacy of Reprinted with permission from [82] American Diabetes Association (2013). this step, and there is relatively little islet loss. Once purified, islets can be cultured for 24–72 h after isolation to improve purity and allow com- Islet delivery & transplant site pletion of viability assays [21]. This in vitro culture also serves as Once the islets are ready for transplantation, they are infused directly into the recipient’s portal venous system. (FIGURE 2) The viability assays as some islets will not survive this brief culture. In addition to the purification and isolation protocols, intraportal site is the most commonly used site for human islet advances in donor selection and pancreas procurement play a transplantation, mainly because this site has been extensively crucial role in the overall outcomes. Most centers with expe- studied in animal models of islet transplantation and the infusion rience in the field now report that islet yields are highest process is relatively straightforward and well tolerated. During when younger donors (2.5 mg/dl or need for transfusion or surgery) after percuta9000–10,000 IEQ/kg and thus patients who receive signifi- neous islet transplantation was 9%, though this was subsequently cantly less than this amount are prepared for preemptive reduced by avoiding aspirin therapy before transplant and additional transplants that are performed as early as 2 months improved ablation of the catheter tract with Gelfoam and metal after the first infusion [22]. This approach is thought to mini- coils [23,24]. Other alternative approaches include a minimize graft attrition due to high metabolic activity and thus laparotomy in the operating room and to infuse the islets directly into a portal vein tributary [25] and recanulization of the umbilical improve long-term function. informahealthcare.com
61
Expert Review of Clinical Immunology Downloaded from informahealthcare.com by Nyu Medical Center on 01/12/15 For personal use only.
Review
Ahearn, Parekh & Posselt
vein [2]. The risk of bleeding is significantly less, but the procedure requires general anesthesia and abdominal surgery, both of which can have complications. With either approach, the need for systemic anticoagulation after the transplant places these patients at higher risk of bleeding and they must be monitored accordingly. Hypotension unrelated to bleeding can occur during the islet infusion and is thought to result from the inflammatory response caused by the intravascular inoculum. It is generally self-limited and responds to fluid resuscitation and transient vasopressor administration [26]. The incidence of portal vein thrombosis has decreased as a result of improved purity of the islet inoculum, and is estimated to occur in about 3% of patients [27]. The thrombosis is usually partial and resolves with long-term anticoagulation, but there is a report of complete portal vein thrombosis in a combined islet–liver transplant recipient who required emergent retransplantation of the liver [28]. Hepatic steatosis is seen in as many as 20% of intraportal islet recipients and is thought to result from increased insulin levels in the periislet parenchyma. The clinical relevance of this finding is not known at present [29–32]*add 3 more refs. In addition to these procedural issues, there are a number of physiologic reasons that make the liver a suboptimal site for islet transplantation. First, oxygen content of portal blood is very low (15–20 mmHg), and this relative hypoxia can compromise the viability and survival of the infused islets [33,34]. This hypoxic state is further exacerbated by the local stagnation of portal blood that develops in the small portal venous tributaries after islet transplantation and persists for several days. Second, intravenous infusion of the islets exposes them to a severe non-specific inflammatory response called the IBMIR that rapidly destroys much of the islet inoculum [35,36]. IBMIR develops when tissue factor expressed on the islet surface is exposed to blood, triggering platelet aggregation, complement activation and leukocyte infiltration, all of which contribute to accelerated islet loss. Third, the location of the islets directly within the portal system exposes them to high levels of immunosuppressive medications being absorbed from the gut, and many of these, such as tacrolimus and sirolimus have direct beta cell toxicity [37]. Together, these factors are thought to be responsible for a loss of 50–75% of the islet inoculum in the first several days after transplantation, and are an important reason why most patients require islets from multiple donors to achieve independence [35,36]. Several approaches have been tried to minimize early destruction of intraportally placed islets, including strategies to promote anticoagulation such as systemic therapy with heparin or low-molecular-weight dextran and coating of the islet surface with heparin or urokinase [38–40]. In addition, a number of agents have been used to inhibit the inflammatory response: most promising are fusion proteins that inhibit the action of the inflammatory cytokines IL-1b and TNF-a [41–43]. One such agent, the soluble TNF-a fusion protein etanercept is currently considered standard of care in most protocols [44]. Although these approaches have improved engraftment and early islet survival, the overall clinical outcomes are only modestly better, with most patients requiring 62
two or more islet infusions to achieve independence and functional graft loss occurring in 50–60% of patients within 2–3 years after transplantation [45]. Although several anatomical sites other than the liver have been tried in animal models with varying success (renal subcapsule, testicle, thymus, anterior chamber of the eye, peritoneum and subcutaneous space), most are impractical or hazardous to use in the clinical setting or require islet numbers significantly greater than those obtainable during routine clinical isolations [46,47]. Initial attempts to transplant islet cells failed to achieve insulin independence [48]. However, more recent studies have demonstrated success of transplantation of islets in mice and humans [49,50]. In particular, Maffi et al. have successfully transplanted autologous pancreatic islet cells in four patients who underwent total pancreatectomy with detectable C-peptide levels at 944 days. Another site with potential clinical utility is the gastric or intestinal submucosal space. This site is attractive because it is well vascularized, accessible endoscopically and does not expose the islets to IBMIR. Encouraging results have been reported in rodent and large animal (pig) models, with adequate graft function and evidence of improved glucose control compared with the intraportal route [51,52]. Recently, Wzola et al. described the first successful case of gastric submucosal islet autotransplantation in a patient with chronic pancreatitis. This patient was not a candidate for intraportal infusion due to preexisting partial portal vein thrombosis. The patient underwent a pancreatectomy and approximately 250,000 IEQ (3700 IEQ/kg) were isolated from the pancreas. These were injected with endoscopic guidance into 18 gastric submucosal sites at the time of surgery. There were no complications related to the procedure. Remarkably, the patient’s pre-pancreatectomy and 30-day post-pancreatectomy fasting and glucose-stimulated C-peptide levels were similar, and the patient did not require any insulin or oral hypoglycemic agents after the procedure [53]. Additional studies are needed to determine whether this site has broader clinical applicability. Advances in immunosuppressive strategies
Although significant progress has been made in optimizing islet isolation, most of the improvements in clinical outcomes are the direct result of refinements in the treatment of the autoand alloimmune responses to the islet grafts. The observation that islet engraftment and duration of insulin independence are significantly better in patients undergoing islet autotransplantation versus allotransplantation suggests that immunologic and immunosuppressive drug-related effects play an important role in mediating graft loss [54] For example, some of early islet loss observed with protocols using non-depleting induction agents was likely due to incomplete suppression of alloreactivity at the time of transplant [55]. In addition to this early loss, long-term function of the remaining islets was likely compromised by the direct toxicity of some maintenance immunosuppressive agents. Calcineurin inhibitors (CNIs), prednisone [5,56,57] and even mycophenolic acid have all been shown to have a toxic effect on islet cells [5,56–58]. Expert Rev. Clin. Immunol. 11(1), (2015)
Expert Review of Clinical Immunology Downloaded from informahealthcare.com by Nyu Medical Center on 01/12/15 For personal use only.
Islet transplantation for Type 1 diabetes
To address these issues, new immunosuppressive protocols were developed that included potent T-cell depleting drugs and minimizing the use of agents toxic to islets. Hering et al. was the first to show that T cell depletion (initially with OKT3 and subsequently with polyclonal anti-lymphocyte antibody thymoglobulin) at the time of transplant, along with ‘islet-sparing’ maintenance immunosuppression based on sirolimus and low-dose tacrolimus substantially improved engraftment and long-term independence rates [55]. Alemtuzumab, an anti-CD52, T cell depleting antibody, has also shown promise as an induction agent, further strengthening the argument that T cell depletion at the time of islet infusion may improve long-term outcomes [59]. Simultaneous with this development came the routine use of etanercept. As discussed above, the rationale behind this agent is that TNF-a has significant in vitro toxicity to islets, and administration of TNF-a inhibitors improves islet engraftment in both murine and primate models. Thus, in 2005, Hering et al. published the results of eight patients treated with a combination of thymoglobulin and etanercept for induction. This trial showed that this therapy was safe and demonstrated 100% insulin independence after a single islet transplant. Subsequent comparisons confirm that T cell depleting regimens in conjunction with etanercept outperform both thymoglobulin only and daclizumab induction in terms of both 3 and 5 year rates of insulin independence (FIGURE 3) [60]. In addition to increasing the level of immunosuppression at the time of islet administration, development of protocols that minimize exposure to prednisone and CNIs is crucial to improving long-term islet survival. Both CNIs and steroids have been shown to have direct islet toxicity and a number of studies have shown that steroid withdrawal and minimization of CNI exposure with the addition of an m-TOR inhibitor such as sirolimus improves overall outcomes [5,37,56,61]. The use of m-TOR inhibitors is limited by the development of proteinuria in a number of patients. Several centers have taken this approach one step further by developing immunosuppressive protocols consisting of induction with thymoglobulin followed by maintenance immunosuppression that avoids CNIs and steroids, and instead relies on novel monoclonal antibodies against specific T cell antigens. The antibodies include efalizumab (Raptiva) that is directed against the anti-leukocyte functional antigen-1 and affects T cell migration and activation, and belatacept (LEA29Y, Nulojix) that inhibits the costimulatory signal needed to activate T cells [62,63]. These biologic agents were selected for their potent immunosuppressive properties, the absence of beta cell and renal toxicity, and lack of many of the side effects observed with other currently employed drugs. Results from these relatively small studies are promising, with high rates of insulin independence after single-donor islet transplantation and insulin independence rates exceeding 50% at 5 years. These results are comparable to those seen after pancreas transplantation in non-uremic Type 1 diabetic patients [64]. Though efalizumab was very well tolerated in the islet patients, it was linked to the development of progressive multifocal informahealthcare.com
Review
Tacrolimus (low dose) Sirolimus Etanercept Anti-thymocyte globulin
–2
–1
Donor islet isolation
0
+1
+2
+7
+10
+28
Days after transplant Islet transplant
Islet culture
Figure 3. Immunosuppressive protocol associated with improved islet engraftment, better long-term islet function and increased incidence of insulin independence after single transplants. T cell depleting induction immunosuppression and etanercept are given during the peritransplant period. Sirolimus and low-dose tacrolimus are used for maintenance therapy.
leukoencephalopathy in a small number of patients receiving the drug as treatment for psoriasis and was voluntarily withdrawn from the market by its manufacturer. The drug was subsequently discontinued in all islet patients, but surprisingly, most retained graft function despite relatively low levels of immunosuppression that consisted primarily of sirolimus and mycophenolate [65]. Encouraging results were also observed with belatacept, and a number of centers continue to use it as a maintenance agent for islet as well as kidney transplantation [64]. Indications for islet transplantation
While allogeneic islet transplantation is a well-tolerated procedure and can provide excellent glucose control, it requires long-term immunosuppression that carries significant risks. Consequently, islet transplantation alone is currently limited to adult patients who have difficulties in controlling diabetes despite optimization of medical therapy. Specifically, potential recipients must be Type 1 diabetics with no detectable C-peptide (
