Editorials POINT-COUNTERPOINT (SEE P. 1563)

Counterpoint: Clinical Islet Transplantation: Not Ready for Prime Time

-Cell replacement therapy’s great lying cause of the apparent excess mortality from the Edmonton group have exposed ␤ promise is that it can safely post– transplantation is not known several problems (17). and effectively restore - but is likely caused by the chronic immuno- independent euglycemia to indi- suppression required to prevent allograft re- Insufficient safety viduals with diabetes. Although difficult jection and the nephrotoxicity associated Immunosuppression-related risks. All and expensive, modern insulin-based with that therapy. Impaired kidney func- islet transplant recipients must be immu- treatment for has lead to tion is, after all, a known major risk factor nosuppressed to prevent rejection, remarkable improvements in disease for cardiovascular disease and excess mor- and such therapy is associated with signif- prognosis (1–3). Based on recent popula- tality in diabetes (7–10). The accompanying icant toxicity. Calcineurin phosphatase tion-based epidemiological studies, we editorial by Mineo et al. (11) cites a study by inhibitors remain the mainstay of many have estimated that an individual diag- Gruessner et al. (12) that reports an ϳ13% antirejection regimens, and the agents are nosed with type 1 diabetes today faces an mortality in subjects awaiting PTA. Our well known to progressively impair renal excess mortality over the next 20 years of analysis using the same national data function. Indeed, depending on the organ ϳ2% or ϳ0.1% per year (4). Even indi- showed that the 4-year mortality for such transplanted, Ojo et al. (18) calculated viduals with long-standing type 1 diabe- patients was much less (ϳ7.9%) (5). One that 7–21% will develop renal failure, and tes sufficiently problematic to be listed for plausible explanation for the difference be- early studies testing cyclosporine in indi- a pancreas transplant have an annual tween the two studies, both of which used viduals with recent-onset type 1 diabetes mortality Ͻ2.0% per year (5). The per- the same United Network for Organ Shar- were also marred by drug-induced de- ceived weaknesses associated with inten- ing (UNOS) data, is that our analysis ex- clines in renal function (19). Moreover, sive insulin therapy are its cost estimated cluded patients with impaired renal and (the antirejec- in 1996 to be 116,000 USD over a lifetime function (serum creatinine Ͼ2.0 mg/dl) tion agents originally used in the Edmon- (6) and its inconvenience requiring me- while the report from Gruessner et al. did ton Protocol) are known to impair insulin ticulous attention to diet, exercise, fre- not, and renal insufficiency is known to in- sensitivity (20) and islet vascularization quent daily blood glucose measurements, crease subject mortality. We suspect that (21), and in vitro studies suggest direct and multiple daily injections. In addition, our observation suggesting that solitary ␤-cell toxicity (22). Tacrolimus is also as- insulin therapy carries with it an in- whole may actu- sociatedwithhypertensionandhypercho- creased risk of serious hypoglycemia (1). ally increase mortality risk has resulted in a lesterolemia such that islet transplant We attempt to address whether current consistent decline in such procedures (PTA recipients more often require antihyper- ␤-cell replacement therapies overcome plus PAK) performed in the U.S. in the past tensive and cholesterol-lowering medica- the shortcomings associated with medical 5 years (Fig. 1). tions following the transplantation management. Certainly, ␤-cell replace- (14,17). Non–antigen-specific immuno- ment therapy should not increase the sub- suppression is also known to increase an jects’ risk above that associated with Islet transplantation individual’s risk for malignancy. Of 325 standard clinical care. Isolated islet transplantation has been islet allograft recipients entered into the For patients with type 1 diabetes and proposed as an alternative for these pa- Clinical Islet Transplant Registry (CITR) end-stage kidney disease, simultaneous tients because it has at least two potential as of April 2008, 14 (4%) developed var- pancreas-kidney (SPK) transplantation has advantages. First, by transplanting only ious neoplasms and, of these, four were appropriately achieved standard of care sta- the islets required for insulin secretion, classified as possibly or definitely related tus. SPK recipients can reasonably expect one avoids many surgical and postopera- to immunosuppression (14). Most impor- improved and sustained insulin-indepen- tive complications associated with whole tant, seven deaths have been reported dent metabolic control, and the surgery pancreas transplantation. Second, at least among the 325 CITR islet allograft recip- demonstrably improves survival rates com- theoretically, islets can be isolated from ients (one because of viral meningitis pared with those medically treated (5). organs otherwise deemed unsuitable for while on immunosuppression, five from However, the options are much less clear for whole . strokes [within 2–3 years postinfusion], subjects with long-standing diabetes and Unfortunately, subsequent experi- and three from unknown causes). preserved kidney function defined as a se- ence has considerably dampened the tre- Some have argued that islet trans- rum creatinine Ͻ2.0 mg/dl. (And with mendous enthusiasm that accompanied plantation represents an ideal clinical modern therapies, an ever-decreasing mi- the “Edmonton Protocol” in 2000 (13) model to test new immunosuppressive nority will develop renal failure.) For such that reported that seven out of seven iso- regimens. We question that thinking. patients, pancreas transplant alone (PTA) or lated islet transplant recipients main- Many factors can lead to islet allograft fail- pancreas-after-kidney (PAK) transplanta- tained insulin-independent euglycemia ure including islet graft quality, alloim- tion, despite reasonable insulin-indepen- for at least 1 year. More extensive experi- mune rejection, recurrent anti–␤-cell dence rates, does not improve survival and ence from several centers worldwide autoimmunity, and glucose or lipotoxic- may even increase mortality (5). The under- (14–16) and a more prolonged follow-up ity associated with the intrahepatic loca-

1570 DIABETES CARE, VOLUME 32, NUMBER 8, AUGUST 2009 Khan and Harlan

sites, or islet encapsulation may improve islet survival, such potential advances have not been validated in clinical trials. A recent small series of five consecutive sub- jects given a supplemental islet infusion, then combination therapy with exenatide and etanercept, reported sustained insu- lin independence for up to 18 months (31). These results are encouraging; how- ever, the numbers are small and the fol- low-up is short. The most encouraging results cited in the editorial by Mineo et al. such as from a study by Vantyghem et al. (32) raise sev- eral questions. First, Vantyghem et al. Figure 1—UNOS data showing that the number of SPK transplantation procedures performed (32) infused more islets than most groups every year in the U.S. has remained stable. The number of solitary pancreas transplants (PTA and (on average, islets from 2.7 donors/ PAK) performed has declined since the published report by Venstrom et al. (5) in 2003 showing recipient). The most straightforward increased mortality in this group. explanation for their high insulin inde- pendence rate may be simply that better early ␤-cell function (conferred by trans- tion following their intraportal infusion pendence is infrequently attained. Only planting more islets) predicts better late (rev. in 23). And yet, clinical investigators ϳ10% of the recipients from the Edmon- islet function. Even so, islet transplanta- lack assays for reliably predicting islet ton trial were still insulin independent 5 tion is limited by its expense and the in- quality or to adequately monitor anti-islet years following their transplant, while sufficient islet supply. If the protocol immune responses. We are unable to di- only 14% of the islet allograft recipients employed by Vantyghem et al. was widely agnose acute rejection (unlike kidney al- from the Immune Tolerance Network adopted (as discussed more completely in lograft rejection) until the islet grafts have (ITN) trial remained insulin independent the “Cost and availability” section), the failed, and rescue therapies do not exist. by 2 years following their procedure average cost for organs alone would in- With so many factors possibly contribut- (15,24,25). Twenty-three percent of the crease by ϳ50,000 USD per patient and ing to graft failure, and considering our islet allograft recipients reported to the the available organs will be spread more inability to precisely measure the various CITR were insulin independent at 2 years; thinly among potential recipients. Sec- contributing factors, we argue it is diffi- however, most had multiple islet infu- ond, the editorial by Mineo et al. states cult or impossible to critically assess a sions (14). These results stand in stark that Vantyghem et al. achieved 5-year in- new immunosuppressive regimen in the contrast to those obtained in solid pan- sulin independence rates of ϳ57%. How- islet transplant recipient. Further, we ar- creas allograft recipients in whom the in- ever, Vantyghem et al. only estimated gue that immunosuppressive agents sulin intendance rate at 5 years is nearly 5-year graft survival from limited data; of shown to be effective and safe for one threefold higher; thus, the glycemic con- the 14 subjects in the study, only 7 had indication generally transfer well to an- trol is better and more sustained and re- reached the 3-year postislet transplant other. Our point is that newer immuno- quires only one donor pancreas (26,27), time and only 3 reached the 4-year post- modulatory strategies can be more whereas most islet allograft recipients re- transplant time period. We argue that effectively tested in other transplant set- quire islets from two or more donors. Fur- great care should be exercised when pro- tings (in particular, kidney transplanta- thermore, in studies such as the ITN trial, jecting results based on such limited data. tion); should safe and effective strategies insulin independence was defined rather Third, the islet transplant recipients in be identified, then testing that approach loosely (i.e., “FBS not to exceed 140 mg/dl this study lost creatinine clearance at a Ϫ Ϫ in islet transplantation makes great sense. more than three times/week and 2-h post- rate of 1.8 ml min 1 year 1 per 1.73 Procedure-related complications. Al- prandial BG [blood glucose] not to exceed m2. This rate of renal function loss is less though isolated islet transplantation has 180 mg/dl more than four times/week”). than those reported by other groups using fewer acute complications than whole or- This begs the question of why glycemic the same immunosuppressive agents and gan transplantation, it is not free of seri- thresholds for islet transplant recipients levels. In fact, therapeutic use of cal- ous complications. Procedure-related should be any different than for the gen- cineurin phosphatase inhibitors is invari- events include hemorrhage, infections, eral population. A key challenge facing ably associated with decreasing renal anemia, portal vein thrombosis, cholecys- islet transplantation is this inexorable de- function for the group receiving that titis, and lymphopenia (0.06–0.36 cline in islet function that is presumably treatment. (Though within a treatment events/person-year) (14). We acknowl- secondary to auto- and alloimmune at- group, some individuals seem more resis- edge that the rate of these acute compli- tack, with perhaps other factors at play tant to the nephrotoxicity than others.) Is cations has declined in recent years. such as immediate islet loss postinfusion there any reason to suspect that individ- and poor revascularization postengraft- uals with type 1 diabetes who receive an Insufficient efficacy ment (28,29). Although developments islet transplant are somehow resistant to Following the initial enthusiasm regard- such as anticoagulation to minimize the calcineurin inhibitor–induced nephro- ing the Edmonton protocol, it has become immediate islet loss that occurs during toxicity? Indeed, even Vantyghem et al. quite clear that long-term insulin inde- the infusion (30), alternate implantation report that of the 14 subjects in their

DIABETES CARE, VOLUME 32, NUMBER 8, AUGUST 2009 1571 Clinical islet transplantation study, 5 developed new microalbumin- lowing their transplant. There was no sta- nors should, for example, renal replace- uria following islet transplantation. tistical difference in rate of decline in GFR ment therapy be required in the future. before or after transplant; however, sub- Islet transplantation improves jects had a steeper decline in GFR while Cost and availability Ϫ metabolic control, but at what cost? on medical therapy (0.45 ml min 1 Clinical islet transplantation is an expen- Ϫ Most islet transplant recipients retain month 1) than was predicted for the gen- sive proposition. First-year expenses for a Ϫ some graft function, i.e., continued circu- eral population (0.083 ml min 1 typical procedure are estimated to be Ϫ lating C-peptide, for several years follow- month 1). Interestingly, the rate of GFR Ͼ150,000 USD (41): ϳ20,000 USD for ing infusion (14,15,17,25). This partial decline in this study’s medically treated hospital and follow-up visit charges, islet graft function serves to facilitate the population was greater than rates previ- 30,000 USD for a year’s immunosuppres- individual’s ability to achieve desired gly- ously reported for patients with long- sive agents, and 100,000 USD for isolat- cemic control targets and to decrease the standing diabetes. Conversely, the ing islets because the charge for each frequency of severe hypoglycemic. Re- transplant recipients in this study experi- pancreas is about 25,000 USD, each re- ports also suggest an improved quality of enced a much slower GFR decline (0.12 cipient requires, on average, islets from Ϫ Ϫ life (33,34). However, if this is the only ml min 1 month 1) than had previ- two donors, and the islet isolation effort clinical gain from a complicated, expen- ously been described in other large clini- succeeds only about half the time. (Thus, sive procedure with significant risks, is it cal ␤-cell transplantation trials. The four pancreata are typically required for justified? results are intriguing but confusing given each recipient.) The lifetime cost is likely that (except for the use of antithymocyte to be much higher because the recipients Islet transplantation and globulin during induction) the antirejec- would require routine medical visits asso- complications tion therapy was not substantially different ciated with immunosuppressive agent Renal function. Impaired renal function from previous clinical islet transplantation complications, laboratory monitoring of is a major determinant of cardiovascular trials. levels, and disease and mortality in patients with di- Retinopathy. Thomson and colleagues medical/surgical care for complications. abetes (7–10). Any new alternative ther- (37) report that islet allograft recipients And given that most recipients continue apy for diabetes must reduce, and appeared to have less progressive retinop- to take insulin, these costs are added to certainly not hasten, the renal disease pro- athy. Although this finding is promising the costs associated with routine diabetes gression. Unfortunately, with some ex- and consistent with results following care. Aside from costs, broadly applying ceptions, most published clinical islet pancreas transplantation (38,39), we islet transplantation is otherwise limited (15,35) and whole pancreas transplanta- point out the study’s small size and state by the supply of brain-dead donor organs tion studies have shown a progressive de- that glycemic control following islet (currently only ϳ8,000 per year in the cline in renal function in patients transplant is not as persistent or physi- U.S.), inefficient isolation techniques, following the procedure (18,36). In the 3 ological as that following whole organ and the need for specialized centers to years following islet allograft transplanta- transplantation. perform the procedure. Indeed, Mineo et tion, the participants in the Edmonton Macrovascular complications. Islet al. argue for potential benefit using pan- trial had declining glomerular filtration transplantation has not been shown to in- creata from UNOS for islet transplanta- Ϫ rate (GFR) (on average, 0.39 ml min 1 fluence cardiovascular disease given that tion; i.e., of the Ͼ8,000 donors available Ϫ month 1) and many had new or progres- the islet recipient number remains rela- each year, only ϳ25% were used for sive microalbuminuria (35). In the ITN tively small and the follow-up is short. whole organ transplantation, leaving an trial, the GFR also declined at a rate of However, some of the immunosuppres- additional 6,000 organs that could be Ϫ Ϫ 0.45 ml min 1 month 1 (15). In com- sive agents commonly used to maintain used as a source for clinical islet trans- parison, the GFR declined at 0.114 ml graft function in islet recipients are plantation trials. And yet, given that organ Ϫ Ϫ min 1 month 1 in the Epidemiology of known to raise blood pressure and cause procurement for clinical purposes man- Diabetes Interventions and Complica- lipid abnormalities—two major risk fac- dates that each organ procured pay an tions (EDIC) medically treated subjects tors for cardiovascular events. Further, equal share of all costs required (ϳ25,000 Ϫ Ϫ and at 0.183 ml min 1 month 1 in solitary pancreas transplantation, with USD for each organ procured) and if all another study of subjects with long- much higher rates of insulin indepen- 6,000 organs were procured for islet iso- standing type 1 diabetes (36). dence, has not been shown to improve lation efforts, the annual cost for those The principal criticism of the pub- mortality, and some data suggest it may organs alone would be 6,000 times lished studies has been the lack of appro- actually worsen survival (5). $25,000 USD or 150,000,000 USD. priate control groups to compare rates of HLA sensitization. Most recipients re- Clearly, such costs are difficult to justify various end-organ complications. Re- quire islets from multiple donors to during these fiscally constrained times. cently, Thompson and colleagues (37) achieve insulin independence. Islets are published their findings comparing islet allocated based only on ABO compatibil- Conclusion transplantation with intensive medical ity, so recipients risk sensitization to mul- Type 1 diabetes is a life-changing diagno- therapy on the progression of type 1 dia- tiple HLA antigens. Indeed, Shapiro and sis that requires the patient’s nearly con- betes–associated complications. The colleagues (40) recently reported that a stant attention. Although insulin therapy crossover design study included 42 sub- third of islet transplant recipients develop is cumbersome, for most it can deliver jects and was prospective but was non- anti–donor-specific HLA and good metabolic control with resulting randomized and unblinded; patients that most develop antibodies against both protection from end-organ damage. Ad- enrolled in an islet transplant protocol class I and class II antigens. This group of mittedly, multiple daily insulin injections were followed prior to and (for some) fol- individuals will have fewer suitable do- are difficult for many, and over time, ap-

1572 DIABETES CARE, VOLUME 32, NUMBER 8, AUGUST 2009 Khan and Harlan propriate glycemic control can be increas- No potential conflicts of interest relevant to C. Steady progress and current challenges ingly difficult with more frequent and this article were reported. in clinical islet transplantation. Diabetes severe hypoglycemia. ␤-Cell replacement Care 2009;32:1563–1569 12. Gruessner RW, Sutherland DE, Gruessner could help overcome these difficulties. ●●●●●●●●●●●●●●●●●●●●●●● Even so, islet transplantation does not (for AC. Mortality assessment for pancreas References transplants. Am J Transplant 2004;4: more than a few years) negate the need for 2018–2026 insulin and the associated inconveniences 1. Diabetes Control and Complications Trial Research Group. The effect of intensive 13. Shapiro AM, Lakey JR, Ryan EA, Korbutt and has not convincingly been shown to treatment of diabetes on the development GS, Toth E, Warnock GL, Kneteman NM, favorably change the end-organ compli- and progression of long-term complica- Rajotte RV. Islet transplantation in seven cation rate, at least with regard to the tions in insulin-dependent diabetes mel- patients with type 1 diabetes mellitus kidney, which is the most significant de- litus. N Engl J Med 1993;329:977–986 using a glucocorticoid-free immunosup- terminant of patient survival. Moreover, 2. Nathan DM, Cleary PA, Backlund JY, Ge- pressive regimen. N Engl J Med 2000; the transplantation procedure and subse- nuth SM, Lachin JM, Orchard TJ, Raskin 343:230–238 quent immunosuppression have signifi- P, Zinman B. Intensive diabetes treatment 14. Alejandro R, Barton FB, Hering BJ, Wease cant associated risks. Most concerning is and cardiovascular disease in patients S. 2008 update from the Collaborative Is- with type 1 diabetes. N Engl J Med 2005; let Transplant Registry. Transplantation the worsening renal function observed in 2008;86:1783–1788 islet recipients on immunosuppression, 353:2643–2653 3. Pambianco G, Costacou T, Ellis D, Becker 15. Shapiro AM, Ricordi C, Hering BJ, Auch- but recipients also face increased risk of DJ, Klein R, Orchard TJ. The 30-year nat- incloss H, Lindblad R, Robertson RP, Sec- infections, neoplasms, and deleterious ural history of type 1 diabetes complica- chi A, Brendel MD, Berney T, Brennan metabolic and cardiovascular effects. Fur- tions: the Pittsburgh Epidemiology of DC, Cagliero E, Alejandro R, Ryan EA, ther, islet transplantation is an expensive Diabetes Complications Study experi- DiMercurio B, Morel P, Polonsky KS, Re- therapy that cannot be widely applied be- ence. Diabetes 2006;55:1463–1469 ems JA, Bretzel RG, Bertuzzi F, Froud T, cause of limited supply of organs and 4. Khan M, Harlan DM. Transplant-based Kandaswamy R, Sutherland DE, Eisen- poor islet survival. Thus, there is little treatments for the patient with long-stand- barth G, Segal M, Preiksaitis J, Korbutt clinical benefit to be gained from clinical ing type 1 diabetes. In Diabetes: Translating GS, Barton FB, Viviano L, Seyfert-Margo- islet transplantation, while there are nu- Research into Practice. Greenbaum CJ, Har- lis V, Bluestone J, Lakey JR. International trial of the Edmonton protocol for islet merous serious concerns regarding its rison LC, Eds. New York, Informa Health- care USA, 2008, p. 193–208 transplantation. N Engl J Med 2006;355: safety and efficacy. The risks associated 5. Venstrom JM, McBride MA, Rother KI, 1318–1330 with islet transplantation lead us to con- Hirshberg B, Orchard TJ, Harlan DM. 16. Hirshberg B, Rother KI, Digon BJ, Lee J, clude that it should only rarely, if ever, be Pancreas transplantation decreases sur- Gaglia JL, Hines K, Read EJ, Chang R, performed to improve metabolic control vival for patients with diabetes and pre- Wood BJ, Harlan DM. Benefits and risks or to reduce hypoglycemia risk. Rather, served kidney function. JAMA 2003;290: of solitary islet transplantation for type 1 we suggest that improved insulin delivery 2817–2823 diabetes using steroid-sparing immuno- systems can achieve the same goal with far 6. Stern Z, Levy R. Analysis of direct cost of suppression: the National Institutes of less cost and minimal risk to nearly all standard compared with intensive insulin Health experience. Diabetes Care 2003; patients, and for those failing best efforts treatment of insulin-dependent diabetes 26:3288–3295 17. Ryan EA, Paty BW, Senior PA, Bigam D, using insulin, whole pancreas transplan- mellitus and cost of complications. Acta Diabetol 1996;33:48–52 Alfadhli E, Kneteman NM, Lakey JR, Sha- tation remains the best “last resort” 7. Go AS, Chertow GM, Fan D, McCulloch piro AM. Five-year follow-up after clinical therapy. We continue to enthusiastically CE, Hsu CY. Chronic kidney disease and islet transplantation. Diabetes 2005;54: support ongoing research to overcome the risks of death, cardiovascular events, 2060–2069 the problems that at present shroud islet and hospitalization. N Engl J Med 2004; 18. Ojo AO, Held PJ, Port FK, Wolfe RA, transplantation’s great promise. 351:1296–1305 Leichtman AB, Young EW, Arndorfer J, 8. Klausen K, Borch-Johnsen K, Feldt-Ras- Christensen L, Merion RM. Chronic renal MAHFUZUL H. KHAN, MD mussen B, Jensen G, Clausen P, Scharling failure after transplantation of a nonrenal DAVID M. HARLAN, MD H, Appleyard M, Jensen JS. Very low lev- organ. N Engl J Med 2003;349:931–940 els of microalbuminuria are associated 19. Canadian-European Randomized Con- From the Diabetes Branch, National Institute of Di- with increased risk of coronary heart dis- trol Trial Group. Cyclosporin-induced re- abetes and Digestive and Kidney Diseases, Na- ease and death independently of renal mission of IDDM after early intervention: tional Institutes of Health, Bethesda, Maryland. function, hypertension, and diabetes. Cir- association of 1 yr of cyclosporin treat- Corresponding author: David M. Harlan, davidmh@ culation 2004;110:32–35 ment with enhanced insulin secretion. Di- mail.nih.gov. 9. Gerstein HC, Mann JF, Yi Q, Zinman B, abetes 1988;37:1574–1582 DOI: 10.2337/dc09-0611 Dinneen SF, Hoogwerf B, Halle JP, Young 20. Johnston O, Rose CL, Webster AC, Gill JS. © 2009 by the American Diabetes Association. J, Rashkow A, Joyce C, Nawaz S, Yusuf S. Sirolimus is associated with new-onset di- Readers may use this article as long as the work is properly cited, the use is educational and not for Albuminuria and risk of cardiovascular abetes in kidney transplant recipients. profit, and the work is not altered. See http:// events, death, and heart failure in diabetic J Am Soc Nephrol 2008;19:1411–1418 creativecommons.org/licenses/by-nc-nd/3.0/ for and nondiabetic individuals. JAMA 2001; 21. Cantaluppi V, Biancone L, Romanazzi details. 286:421–426 GM, Figliolini F, Beltramo S, Ninniri MS, 10. Valmadrid CT, Klein R, Moss SE, Klein Galimi F, Romagnoli R, Franchello A, Sal- BE. The risk of cardiovascular disease izzoni M, Perin PC, Ricordi C, Segoloni Acknowledgments— This research was sup- mortality associated with microalbumin- GP, Camussi G. Antiangiogenic and im- ported by the Intramural Research Program of uria and gross proteinuria in persons with munomodulatory effects of rapamycin on the National Institute of Diabetes and Diges- older-onset diabetes mellitus. Arch Intern islet endothelium: relevance for islet tive and Kidney Diseases/National Institutes of Med 2000;160:1093–1100 transplantation. Am J Transplant 2006;6: Health. 11. Mineo D, Pileggi A, Alejandro R, Ricordi 2601–2611

DIABETES CARE, VOLUME 32, NUMBER 8, AUGUST 2009 1573 Clinical islet transplantation

22. Nir T, Melton DA, Dor Y. Recovery from engraftment is critical for successful clin- FC, Sutherland DE, Steffes MW. Effects of diabetes in mice by beta cell regeneration. ical islet transplantation. Diabetologia pancreas transplantation on glomerular J Clin Invest 2007;117:2553–2561 2008;51:227–232 structure in insulin-dependent diabetic 23. Rother KI, Harlan DM. Challenges facing 30. Ichii H, Ricordi C. Current status of islet patients with their own kidneys. Lancet islet transplantation for the treatment of cell transplantation. J Hepatobiliary Pan- 1993;342:1193–1196 type 1 diabetes mellitus. J Clin Invest creat Surg 2009;16:101–112 37. Warnock GL, Thompson DM, Meloche 2004;114:877–883 31. Faradji RN, Tharavanij T, Messinger S, RM, Shapiro RJ, Ao Z, Keown P, Johnson 24. Ryan EA, Lakey JR, Rajotte RV, Korbutt Froud T, Pileggi A, Monroy K, Mineo D, JD, Verchere CB, Partovi N, Begg IS, Fung GS, Kin T, Imes S, Rabinovitch A, Elliott Baidal DA, Cure P, Ponte G, Mendez AJ, M, Kozak SE, Tong SO, Alghofaili KM, JF, Bigam D, Kneteman NM, Warnock Selvaggi G, Ricordi C, Alejandro R. Harris C. A multi-year analysis of islet GL, Larsen I, Shapiro AM. Clinical out- Long-term insulin independence and transplantation compared with intensive comes and insulin secretion after islet improvement in insulin secretion after medical therapy on progression of com- transplantation with the Edmonton pro- supplemental islet infusion under ex- plications in type 1 diabetes. Transplan- tocol. Diabetes 2001;50:710–719 enatide and etanercept. Transplantation tation 2008;86:1762–1766 25. Ryan EA, Lakey JR, Paty BW, Imes S, Kor- 2008;86:1658–1665 38. Chow VC, Pai RP, Chapman JR, butt GS, Kneteman NM, Bigam D, Rajotte 32. Vantyghem M-C, Kerr-Conte J, Arnals- O’Connell PJ, Allen RD, Mitchell P, RV, Shapiro AM. Successful islet trans- teen L, Sergent G, Defrance F, Gmyr V, Nankivell BJ. Diabetic retinopathy after plantation: continued insulin reserve pro- Declerck N, Raverdy V, Vandewalle B, combined kidney-pancreas transplanta- vides long-term glycemic control. Pigny P, Noel C, Pattou F. Primary graft tion. Clin Transplant 1999;13:356–362 Diabetes 2002;51:2148–2157 function, metabolic control, and graft sur- 39. Pearce IA, Ilango B, Sells RA, Wong D. 26. Sutherland DE, Gruessner RW, Dunn DL, vival after islet transplantation. Diabetes Stabilisation of diabetic retinopathy fol- Matas AJ, Humar A, Kandaswamy R, Care 2009;32:1473–1478 Mauer SM, Kennedy WR, Goetz FC, Rob- 33. Poggioli R, Faradji RN, Ponte G, Betan- lowing simultaneous pancreas and kidney ertson RP, Gruessner AC, Najarian JS. court A, Messinger S, Baidal DA, Froud T, transplant. Br J Ophthalmol 2000;84: Lessons learned from more than 1,000 Ricordi C, Alejandro R. Quality of life af- 736–740 pancreas transplants at a single institu- ter islet transplantation. Am J Transplant 40. Campbell PM, Senior PA, Salam A, La- tion. Ann Surg 2001;233:463–501 2006;6:371–378 branche K, Bigam DL, Kneteman NM, 27. Sutherland DE, Gruessner AC. Long-term 34. Barshes NR, Vanatta JM, Mote A, Lee TC, Imes S, Halpin A, Ryan EA, Shapiro AM. results after pancreas transplantation. Schock AP, Balkrishnan R, Brunicardi FC, High risk of sensitization after failed islet Transplant Proc 2007;39:2323–2325 Goss JA. Health-related quality of life after transplantation. Am J Transplant 2007;7: 28. Korsgren O, Nilsson B, Berne C, Felldin pancreatic islet transplantation: a longitu- 2311–2317 M, Foss A, Kallen R, Lundgren T, Salmela dinal study. Transplantation 2005;79: 41. Guignard AP, Oberholzer J, Benhamou K, Tibell A, Tufveson G. Current status of 1727–1730 PY, Touzet S, Bucher P, Penfornis A, Bayle clinical islet transplantation. Transplanta- 35. Senior PA, Zeman M, Paty BW, Ryan EA, F, Kessler L, Thivolet C, Badet L, Morel P, tion 2005;79:1289–1293 Shapiro AM. Changes in renal function Colin C. Cost analysis of human islet 29. Korsgren O, Lundgren T, Felldin M, Foss after clinical islet transplantation: four- transplantation for the treatment of type 1 A, Isaksson B, Permert J, Persson NH, year observational study. Am J Transplant diabetes in the Swiss-French Consortium Rafael E, Ryden M, Salmela K, Tibell A, 2007;7:91–98 GRAGIL. Diabetes Care 2004;27:895– Tufveson G, Nilsson B. Optimising islet 36. Fioretto P, Mauer SM, Bilous RW, Goetz 900

1574 DIABETES CARE, VOLUME 32, NUMBER 8, AUGUST 2009