Gene-Altered Islets for Transplant: Giant Leap or Small Step?

On July 20, 1969, Neil Armstrong uttered the now immortal words “That’s one small step for man, one giant leap for mankind” on the occasion of man first setting foot on the lunar surface. For diabetes researchers, finding a “cure” is quite analogous to sending a man to the moon. In pursuit of that laudable goal, much excitement has been generated on several fronts: islet transplantation (from allogeneic or xenogeneic sources), stem cells, gene therapy, closed loop insulin pumps, and immunotherapy to induce tolerance, to name a few. Diabetes research is not unique in that many paths appear so bright at first only to turn either most difficult, or worse, prove to be blind. Predictions surrounding the therapeutic use of monoclonal antibodies, antiangiogenesis agents for cancer therapy, and hormone replacement therapy for a wide panoply of postmenopausal ills are but a few recent examples of most-promising therapies that have fallen quite short (at least to date) of initial, overly enthusiastic hopes. In this issue of Endocrinology, Lopez-Talavera et al. (1) report their studies on the marriage of two promising technologies, gene therapy and islet transplantation. Using a rat islet allotransplantation model, their data (discussed in greater detail below) suggest that the sum may be greater than the constituent parts. Before analyzing the manuscript reported herein, some discussion is warranted of both the promise and problems associated with these two technologies: islet transplantation for diabetes and gene therapy. Although transplantation-based therapies for type 1 diabetes mellitus (T1DM) (whole pancreas or isolated islets) have matured, with often quite dramatic leaps forward over the past 20 yr, significant issues remain (2). Considering the more mature pancreas transplant procedure first, many centers experienced in the technique now report that more than 85% of their transplant recipients achieve normoglycemic insulin independence 1 yr after surgery (3, 4). Despite this remarkable success rate, several features conspire to limit the pancreas transplantation field for the vast majority with diabetes. Most important is the severe discrepancy between those in need (in the United States alone, an estimated 1 million with T1DM, and at least 17 million others with type 2 diabetes) and the number of cadaveric pancreata available for transplantation. Organ procurement organizations have stringent criteria for organ donation such that in the United States only about 12,000 brain-dead individuals are considered suitable as organ donors each year, and families consent to organ donation in only about half of those cases (thus, about 6000 organ donors are identified each year). Of those 6000 human pancreata, only about 1500 are considered suitable for solid organ transplantation, and for many reasons. One reason is that most pancreatic cells make digestive enzymes (islet cells constitute less than 5% of the pancreatic cell mass) and because ischemia is damaging to those cells, most transplant surgeons are reluctant to transplant an organ with a cold ischemia time of more than 12 h. Furthermore, unless the transplant surgeon is confident of the donor pancreas integrity, he/she is reluctant to sew that organ into a recipient for fear of potentially serious complications. Thus, surgeons will often decline pancreata from older donors (typically defined as a donor older than age 50 or 55), or pancreata from donors brain dead from major trauma (for fear of unapparent pancreatic damage), or from obese donors in whom the pancreas is often encased in fat, making visual inspection of the gland difficult. Another issue that continues to limit pancreas transplantation is the intrinsic toxicity associated with the immunosuppressive agents required to prevent organ allograft rejection post transplantation. Immunosuppression is known to increase a patient’s risk for certain infections and malignancies due to the agents’ very nature, one designed (using present technology) to indiscriminately temper immune responses (5, 6). In addition, however, the immunosuppressive agents, like all medicines, have their own toxicity profiles with a cadre of problems (depending upon the specific agent used) like impairment of renal function, bone marrow suppression, hyperlipidemia, hypertension, tremor, mouth ulcers, diarrhea, etc. Indeed, one recent report documented that for other-than-kidney allograft recipients, the incidence of new onset renal insufficiency (presumably from the immunosuppressive agents) was 7–21%, depending upon the organ transplanted and other clinical variables (7). Last but not least, although the patient with a successful pancreas transplant typically reports a huge and positive impact on quality of life, whether the procedure influences important end points like end-stage renal disease, neuropathy, or even mortality continues to be debated (8, 9). In a medical marketplace ever increasingly dictated by cost effectiveness, the great expense associated with pancreas transplantation, immunosuppression, and other required posttransplantation follow-up, especially in view of the uncertainty surrounding impact on important hard end points like end-organ complications and survival, pancreas transplantation’s rather limited role in diabetes care is understandable. For many of the reasons cited above, however, investigators interested in developing new therapies for those with brittle diabetes have long dreamed about the ability to isolate islets from cadaveric pancreata so that just those cell clusters could be transplanted. The technique does, by its very nature, have several advantages when compared with pancreas transplantation. Important intrinsic advantages include: 1) As currently practiced, islets are infused into the portal vein via a catheter placed percutaneously such that the operating room and associated surgical complications are eliminated or at least markedly reduced. 2) One transplants only the needed islets, and not the other pancreatic cellular mass, not needed by the patient with diabetes. Paradoxically, it is cells other than the islet cells that cause most of the problems in pancreas transplant recipients. 3) Many of the pancreata not suitable for whole organ transplantation can be used for islet isolation. Despite these obvious and real advantages, islet transplantation languished for years, while remarkably persistent and steadfast investigators worked to improve islet isolation techniques and clinical protocols supporting insulin independence for islet recipients. Technical improvements in islet isolation worked out by Ricordi and others (10–13) are now sufficiently robust, although still imperfect, that in experienced hands islets suitable for transplantation can now be isolated from about 50% or more of donor organs. And in what is now regarded as a landmark advance, the team at the University of Alberta in Edmonton, Canada developed a protocol built chiefly upon two improvements (transplanting a sufficient islet mass—usually requiring islets from two or more donor organs, and a steroid-sparing immunosuppressive regimen) and reported in 2000 that they had rendered seven consecutive patients insulin independent after an islet transplant (14). Since that report, others, including an investigative team at the National Institutes of Health, have been able to largely reproduce Edmonton’s findings (15). As is nearly always the case with a new therapy, however, with that more extensive experience has come knowledge of limitations. For instance, the islet transplant procedure itself has been associated with potentially serious complications in nearly 20% of recipients (chiefly partial portal vein thrombosis or intraabdominal bleeding after the portal vein cannulation) (16). Other, at this point largely theoretical, concerns have been raised about elevations in portal vascular resistance measured after islet infusion (17), and the effect of the intrahepatic islets on the surrounding liver parenchymal cells both in animal models (18), and in the clinic (19). Furthermore, whereas the glycemia control currently achieved after islet transplantation is much improved for patients with so-called brittle T1DM, the blood glucose control is not normal for most (20, 21) nor commensurate with that typically achieved after solid organ transplantation. Furthermore, it must be stated that islet transplantation experience remains limited so that durability of the islet function over time remains unknown with any degree of certainty. Most significant, the problems discussed above for pancreas transplantation regarding the limitations associated with immunosuppressive therapy, costs associated with transplantation and required follow-up, and pancreas donor supply all conspire to limit islet transplantation’s wide clinical applicability as well (22). With regard to the latter point in particular, the donor-potential recipient disparity is particularly acute in that only half of the islet isolation procedures yield transplant quality islets, and that recipients typically require two or more islet preparations. Thus, any islet transplant recipient would usually require that at least one and often up to four (or more) cadaveric pancreata be invested toward the desired clinical goal. Gene therapy has also been the focus of tremendous enthusiasm within both the lay press and the scientific literature due to its virtually unlimited potential for diseases including cancer, inborn errors of metabolism, certain immunodeficiency disorders, hemoglobinopathies, and other genetically defined illnesses like cystic fibrosis. Limiting the field, however, are still yet imperfect vectors to safely and efficiently transfer genes of interest into the diseased cell (23, 24). The most commonly employed vectors for cellular research studies, the adenoviral vectors, transduce cells with appropriate cell surface receptors at high rates, but the gene transfer is transient (i.e. the transduced gene does not integrate into the host cell genome and will eventually be lost) and when used in vivo, the adenoviral vectors induce an aggressive immune response. Retroviral-based vectors integrate permanently into the genome, allowing passage of the integrated transgene to all subsequent progeny, yet these vectors transduce only dividing cells and are less efficient. Regardless the gene vector type, clinical experience has taught that they can be associated with serious toxicity. In a well-publicized early trial, one patient died of fulminant hepatic failure and other sequela shortly after receiving an adenoviral gene therapy agent (25, 26). More recently, and after the first successful clinical application of gene therapy in the treatment of X-linked severe combined immunodeficiency using a retroviral based gene vector, two of the 11 treated patients developed a leukemia felt most likely related to the chromosomal location site of gene integration (23, 27). Regulatory agencies and responsible investigators now struggle with how to proceed with this exciting field while protecting patient safety and recognizing the limits of our knowledge. The manuscript by Lopez-Talavera et al. (1) carefully and clearly makes several important observations that may represent an important step forward for both islet transplantation and for gene therapy. By way of introduction, this research team has already published results indicating that: 1) Islets from transgenic mice engineered to overexpress hepatocyte growth factor (HGF) under rat insulin promoter control more potently restore euglycemia when transplanted into diabetic recipients than do islets from control mice (28, 29), and 2) Adenoviral vector-induced HGF expression in islets then transplanted into immune incompetent mice with streptozotocin-induced diabetes (30), when compared with appropriate control islets, also more effectively restores euglycemia. The team now extends those observations in several important ways including that they: 1) adapted the rodent islet allograft model to rats to facilitate the intraportal islet delivery of allogeneic islets, thus more closely mimicking the approach used for clinical islet transplantation, and 2) employed an immunosuppressive regimen postislet transplant that closely mimics that used clinically. Among the important observations described are that: 1) The FK506 and rapamycin-based immunosuppressive regimen of the Edmonton protocol, at least in this rat model, while effectively blocking allograft rejection also induces insulin resistance (predominantly a rapamycin-induced effect) and decreased islet insulin secretion (predominantly a FK506 effect) such that the two agents conspire to cause diabetes in the rat. 2) A purposefully limited islet allograft mass (in this case 10 islet equivalents per gram rat body weight) creates a model system with the requisite sensitivity to observe improvements in islet function. 3) Islets overexpressing HGF under the control of an adenoviral vector more effectively, and with considerable longevity, function across the allogeneic barrier in diabetic rat recipients treated with FK506 and rapamycin. The model and the manuscript have some limitations. For instance, although their treated rats attain circulating FK506 and rapamycin levels roughly equivalent to levels achieved in the clinic, the agents are delivered to the rats by injection (FK im and rapamycin sc) as opposed to by mouth as is done in the clinical setting, and this may have influenced their findings. One would guess, however, that were the immunosuppressive agents administered by mouth, effects on islet function would be further magnified because portal vein concentrations of the immunosuppressive agents would likely be even greater than the concentrations achieved after parenteral administration. Furthermore, one must presume that rats are far more susceptible than humans to the diabetogenic effects of FK506 and rapamycin because many patients are now treated with those drugs without the hyperglycemia they describe. None of these limitations, however, significantly detract from the authors’ main messages addressed to the scientific community attempting to develop islet transplantation for widespread clinical application. Indeed, certain aspects of their approach are potentially advantageous for the gene therapy field. For instance, whereas adenoviral vectors are attractive due to their high efficiency, and small risk for malignant transformation (because they do not integrate into the genome), they are less attractive because the transduced gene expression is only transient, the vectors stimulate an immune response, and due to the potential toxicity discussed above. Using the approach reported by Lopez-Talavera et al. (1), however, adenoviral-mediated transient gene expression appears to be sufficient to promote islet function long term, the immunosuppression required after the allogeneic transplant would be expected to minimize the antiadenoviral vector immune response, and by exposing the islets to the vector ex vivo, one markedly diminishes the amount of vector introduced in vivo, and thereby also likely diminishes the risk for the recipient. The next and a difficult step will be to plan and perform those studies required to assess safety and efficacy before moving to clinical testing. For instance, testing adenoviral-mediated HGF expression in islets from a large animal model, ideally nonhuman primates, will allow more relevant conclusions regarding safety and efficacy. Lopez-Talavera et al. (1) are not the first to recognize the potential utility of gene therapy-altered islets for clinical use. Numerous strategies designed to subvert the antiislet immune response (31, 32), render islets more resistant to apoptosis (33–35), improve their function via improving vascularization, and/or promote islet proliferation/differentiation are notable examples (36–38). Furthermore, HGF is but one of several factors with experimental evidence supporting a potential role in either promoting islet function, survival, or proliferation with other notable examples including placental lactogen (39), PTHrP (40), glucagon-like peptide-1 or agents that increase glucagon-like peptide-1 receptor signaling (41–45), gastrin (46), vascular endothelial growth factor (47, 48), islet neogenesis-associated protein (49), fibroblast growth factor (38), and still others (50). I began this editorial paraphrasing Armstrong’s words, “giant leap or small step?” Although only time will answer that question, we can say with confidence that whether it is a leap or a small step, the movement is forward. Modern care for individuals with diabetes has improved on many fronts such that disease prognosis is now measurably and quite significantly (both statistically, and clinically) better than it was just 20 yr ago (51). The approach now reported suggests that gene therapy-altered islets may more efficiently restore euglycemia to patients with T1DM and thus help overcome the major limitation of the limited islet supply. For the gene therapy field, using the adenoviral vector to transduce cells ex vivo and to thereby avoid systemic toxicity may provide additional support for this form of therapy. Whether or not the approach reported by Lopez-Talavera et al. (1) represents an important building block toward the “moon shots” of a diabetes cure and/or gene therapy, it is unquestionably an advance and one worthy of aggressive additional investigation. Hepatocyte growth factor; type 1 diabetes mellitus. This work was supported by NIH-NIAMS Grant AR-42238 (to L.J.G.).