Recent studies demonstrate that improved islet transplant outcomes could be observed with enhanced islet isolation, glucocorticoid-free immunosuppression, and provision of an adequate islet mass of more than 10,000 islet equivalents (IE)/kg body weight. These improvements have resulted in benefits to type 1 diabetic subjects, including long-term c-peptide secretion, improved glycemic control, and reduced hypoglycemia episodes (1–3). Despite these advances, graft dysfunction leading to a return to insulin therapy has been observed as the duration of follow-up has increased and more recent studies demonstrate that patients who received islet transplants still developed deteriorated renal function (4, 5). Other studies have examined the effect of islet transplantation on diabetes complications (6). Conclusions from these studies are limited by the need for a control group of subjects who received medical therapy targeted to meet current standards.
We hypothesized that health outcomes for type 1 diabetic subjects could be improved by islet cell transplantation (ICT) compared with medical therapy. To better define the effects of islet transplantation on long-term complications, we examined the natural history of metabolic control, nephropathy, retinopathy, and neuropathy in a single-center, prospective, crossover cohort study. Results were compared with outcomes after intensive medical therapy targeted to all current standards. This report details results after a mean of 3 years follow-up in each arm.
MATERIALS AND METHODS
Eligible subjects were 20 to 65 years of age with more than 5 years diabetes duration, were c-peptide negative, and had evidence of retinopathy and mild nephropathy (urine albumin-to-creatinine ratio >2.0 mg/mL and glomerular filtration rate (GFR) >70 mL/min). Exclusion criteria included ischemic heart disease, previous transplant, recurrent infections, and malignancy (except basal or squamous skin cancer). Patients were not excluded on the basis of insulin dose or weight. Fifty subjects were enrolled from January 2002 to January 2005. Eight subjects withdrew, none of whom received ICT, and are excluded from analysis. All subjects gave written informed consent and the study was approved by the Institutional Review Board of the University of British Columbia.
All subjects began intensive medical therapy upon enrollment, consisting of intensive glucose management, angiotensin blockade, and control of lipids and blood pressure to recommended levels (7–9). Insulin adjustment followed established methods (10) with individual glucose targets titrated, trading optimal HbA1c with minimal hypoglycemia.
The medical patient visits were conducted every 3 months and they were contacted through telephone by clinic nurses every 2 weeks.
Islet Isolation and Transplantation
Islets were isolated from pancreata of adult heart-beating cadaver organ donors in the Ike Barber Human Islet Transplantation Laboratory of Vancouver General Hospital (11). Briefly, collagenase was perfused through the pancreatic duct, digestion was completed in a chamber, and purification was by continuous density gradient purification using Ficoll hypaque. Additional islets were retrieved by repurifying impure tissue fractions. Transplantation occurred under local anesthesia in the interventional radiology suite, where a catheter was advanced percutaneously to the main portal vein with fluoroscopic guidance and islets were infused. Repeat infusions from different donors were completed as necessary until more than 12,000 IE/kg were provided. Immunosuppression consisted of antithymocyte globulin (ATG) induction followed by introduction of sirolimus (n=2) or mycophenolate mofetil (n=29) and tacrolimus. For subsequent infusions of islets, induction was with the interleukin 2 receptor blocker Basiliximab.
Assessment of Renal Function
A primary endpoint was the rate of change in GFR, as measured by the blood clearance of 99mTc-diethylenetriaminepentaacetate every 6 months in the medical and post-ICT groups (12). GFR values were analyzed using simple linear regression for all measurements for each subject to calculate the rate of change. To be included in the analysis, patients had at least two GFR measurements at least 12 months apart. Individual patient results were combined to determine the mean intrapatient rate of change in GFR expressed as the slope of the regression line.
Assessment for Retinopathy
Ophthalmologic assessment was performed by a retina specialist repeatedly (5.5±1.8 visit/patient) who was aware of the treatment category. Annual seven-field stereo fundus photographs were performed by a certified photographer (13). Severity of retinopathy was assessed using the international scale, which has three levels of nonproliferative diabetic retinopathy (NPDR) ranging from mild to severe and one level of proliferative diabetic retinopathy (PDR) (14). Grading was performed to compare levels of severity at entry into medical therapy, just before the initial islet transplant and at the last follow-up visit. Predefined endpoints for each eye were (1) NPDR advancing by more than one level severity (2); progression of PDR to a severity level that qualified for laser photocoagulation or (3) development of macular edema with visual acuity less than 20/40, which qualified for laser treatment.
Neuropathy was assessed by nerve conduction studies of seven peripheral nerves performed annually. Measurements were completed for sensory (median, ulnar, and sural) and motor (median, ulnar, peroneal, and tibial) in the same laboratory under standard conditions (15).
Data were separately analyzed for the entire cohort of 42 subjects who received medical therapy (including 11 subjects who received only medical treatment) and 31 subjects who received ICT. For all parameters, this interim report includes data up to July 15, 2008. For nephropathy component of the investigation, the study was powered to detect 25% of difference in the rate of change of renal function over 5 years between medical therapy and ICT. Recent data suggest a rate of decline in GFR of 1 mL/min per year in the general population, 1 to 2 mL/min per year in patients with microalbuminuria, and 4 to 5 mL/min per year for overt proteinuria (16). For the purposes of this study, we estimated a rate of decline in GFR of approximately 2.5 mL/min per year.
For retinopathy, primary endpoint was progression in an individual eye defined as meeting any of the endpoints. An eye was only counted as progressing once regardless of the number of laser sessions given. Two eyes that were blind and five that had a vitrectomy before entry to the study (total, 6) were excluded since they were not at risk for the study endpoints. Continuous variables were presented as mean±standard deviation if normally distributed and as median if not.
For neuropathy, raw nerve conduction velocity (NCV) data were compiled for all nerves and by different groups of nerves. The NCV (m/sec) was combined to produce a single overall mean value. Measurements over time were compared. Only patients with at least 12 months of follow-up were included.
HbA1c results were presented as medians for each quarter since the data within each quarter were not normally distributed. All other results were expressed as mean±standard deviation. Comparisons were made using t tests. For retinopathy data analysis was with chi-square for proportional variables. All group analyses were performed by intention to treat.
Study Subjects and Follow-up
Table 1 demonstrates the baseline characteristics of the groups. Baseline characteristics were similar except that the HbA1c level immediately preceding islet cell transplant (7.0±0.7) was lower than that at entry into intensive medical therapy (8.1±1.2). Weights (kg) for the medical group were 73.2±13.9 at entry and 72.5±13.1 at last visit. Corresponding values for ICT were 70.2±10.4 before first ICT and 69.6±11.9 at the last.
Duration of follow-up is shown in Figure 1. The duration of follow-up was similar for medical (34±18 month) and ICT (38.4±18 month) groups.
Islet Transplantation and Metabolic Outcomes
Islets suitable for transplantation were isolated from 70 pancreases yielding 494,000±204,000 IE per islet infusion. Thirty-one patients received 70 infusions, distributed per patient as follows: 1 infusion in 7; 2 infusions in 12; 3 infusions in 9; and 4 infusions in 3. Final IEs per kilogram implanted was 12,998±6069. The goal was to achieve and maintain insulin independence as long as possible. If patients were initially insulin independent, then returned to insulin therapy, they were eligible to receive a supplemental infusion. Among the seven recipients of a single infusion, one has remained insulin independent after 12,000 IE/kg, one is on hold because of severe CMV infection after starting immunosuppression, and five await a second infusion to achieve targets of more than 12,000 IE/kg. Therefore, 25 of 31 had completed transplant. Three of these 25 no longer receive immunosuppression; two were stopped because of side effects, and one because of progressive skin cancer. At the time of this report, 16 of 25 patients who have completed their transplant (64%) remain insulin independent. Insulin independence has been observed for more than 5 years (n=1), more than 4 years (n=2), and more than 3 years (n=1). Twenty patients have received the GLP-1 agonist exenatide at a dose of 10 μg subcutaneously twice daily (17). Mean fasting c-peptide (pmol/mL) in insulin-independent patients was 448±184 versus 241±141 in those who have resumed some insulin (P<0.01). Seven patients who achieved initial insulin independence for greater than 1 month but later required insulin despite maintained immunosuppression were treated with supplemental donor islets (20,256±2641 IE/kg) and exenatide. Six of these seven achieved insulin independence at 18.0±16.0 day postinfusion and c-peptide increased significantly from 163±61 to 540±266 at 1 month postinfusion and was sustained at 543±192 after 6 months with HbA1c at 6.5%±0.5%. Patients with partial islet graft function who have resumed insulin currently take 33% to 75% of their pretransplantation dose. Figure 2 demonstrates median quarterly HbA1C for medical versus ICT subjects. Measurements for subjects were combined for each quarter of follow-up and the median value plotted as shown. At all time periods studied, HbA1C (%) was lower for ICT and pooling all numbers to calculate total glycemic exposure during the study period, HbA1C was 7.4 for medical versus 6.6 for ICT (P<0.01).
Next, we evaluated change of renal function. GFR declined for medical (0.45±0.7 mL/min/month) and ICT (0.12±0.7 mL/min/month). Differences between these groups were not significant (P=0.1). The rate of change (slope of GFR decline) in ICT did not differ significantly from 0, nor from that expected in the general population (1 mL/min/yr); in contrast the decline for the medical group was greater than 0 and faster than expected for general population. Subgroup analyses were performed to determine the change observed after initial medical therapy (just before transplant) and for those who had not yet received ICT and remained on medical therapy. Corresponding values of decline for these groups were 0.49±0.8 and 0.33±0.4, respectively (P vs. ICT=NS).
Next, we evaluated effects on progressive retinopathy. The eyes with mild NPDR did not progress. In others, more progression occurred on medical therapy (10 of 82 eyes) than after ICT (0 of 51 eyes, P<0.01). During medical therapy alone, one subject had severe NPDR in one eye that progressed to PDR necessitating laser therapy. Another subject in the medical therapy group awaiting ICT progressed from severe NPDR to PDR in one eye and required laser therapy. Considering only ICT subjects, progression was observed in 6 of 51 eyes on medical treatment compared with none after transplant (P<0.02).
Finally, we studied nerve conduction. Table 2 summarizes NCV in medical versus ICT. No significant deterioration was observed from baseline in either group.
Landmark studies of the Edmonton Protocol have stimulated keen interest in the role of ICT to control the devastating complications of type 1 diabetes. Initial studies demonstrated feasibility of this therapy to achieve insulin independence, to reduce metabolic liability, and to correct severe hypoglycemia (1–3). In contrast with published reports that focus on brittle diabetes, the present study addresses the important question of effects of islet transplants alone on secondary complications of diabetes by comparing outcomes with controls receiving best medical therapy set to current standards of practice (7–9). Our results show improved metabolic parameters after ICT compared with medical therapy throughout all study periods in follow-up: HbA1C was improved with ICT and the difference seems to grow over the study period. This is accompanied by consistent evidence of c-peptide secretion by the transplanted islets. Together, these data reconfirm that islet transplants can reduce insulin dependence and improve metabolic control in many patients. Nevertheless, the grafts seem to be subject to partial loss of function, necessitating a return to insulin therapy in 38% of the recipients. This suggests that antiapoptotic strategies to induce and maintain islet cell function continue to be urgently required. Recent studies by our group and others suggest that incretin-mediated support for transplanted islets may help to sustain more durable long-term function (17, 18). Although this may confound interpretation of data compared with the medical group, there is no indication to give exenatide to c-peptide negative patients with no islets and no literature to suggest that exenatide has any effect on retinopathy, neuropathy, or nephropathy. It is difficult to single out this drug compared with any other aspect of the treatment given to transplanted but not medically treated patients. Our conclusion is that transplanted patients have less progression than medical patients and we can only speculate on what aspect the transplant is responsible for the difference.
The current data confirm and extend our earlier observation 1.5 yr ago that renal function does not deteriorate after ICT compared with best medical therapy (19). The current analysis shows that renal function is not statistically different in the two treatment arms. Of significance, the slope of decline in GFR after medical therapy was significantly more than 0 and greater than expected for the general population, whereas the decline of 0.12±0.7 mL/min per month after ICT was not. The results suggest that additional follow-up might be required to assess differential effects of ICT on diabetic nephropathy. These data seem more promising than the international trial (5), which demonstrated a decline in estimated GFR of 0.45 mL/min per month in 36 subjects followed up for an average of 41 months. Although the current data are reassuring, it must be interpreted with caution since maintenance immunosuppression currently depends on calcineurin inhibitors that carry risks of nephrotoxicity.
This study shows that ICT is more effective than intensive medical therapy in slowing progression of advanced diabetic retinopathy. This result agrees with recent literature on pancreas transplantation in which advanced retinopathy was stabilized (20, 21). In contrast with the studies for pancreas transplantation, this study details a concurrent medical control group. Improved results after ICT may be attributable to better glycemic control (as evidenced by lower HbA1C levels) or the presence of c-peptide, which has been demonstrated to decrease fluorescein leakage across the blood-retinal barrier (22). Finally, immunosuppressive drugs used in the ICT group could potentially arrest progression of diabetic microangiopathy though we have been unable to identify literature reporting such an effect.
No differences were observed in the change of NCV in any of the subgroups analyzed during medical versus ICT therapy. An expected change in NCV for diabetes is a decline of 0.5 m/sec per year (23). Although the current data have not demonstrated a benefit of ICT, the lack of deterioration is encouraging and, at least in the ICT group, is consistent with the observation that c-peptide conserves sensory nerve function (23). Conserved function in the current study may be attributable to mild neuropathy at study entry.
The present study demonstrates some limitations of islet transplantation observed in previous studies. Multiple islet infusions are required to induce and sustain insulin independence. This observation stresses the need to improve the engraftment and long-term survival of islet allografts. Infusions of islets from multiple donors have been implicated in sensitization to donor antigens. The current study has detected panel-reactive antibodies in 2 of 31 recipients (data not shown). The risk of this observation must be balanced with benefits of improved metabolism, conserved renal function, and visual acuity. The immunosuppressive regimen was associated with intolerable side effects (CMV infection, fatigue, and skin cancer) in three subjects, which necessitated cessation. Specific immunosuppression protocols require optimization to improve long-term tolerance of islet transplants.
Major limitations of the current study include the fact that subjects were not randomly allocated, nor were determinations of endpoints masked. The study design did not include obtaining fundus photographs prior to laser treatment. The indication for laser given in the operative report for the patients was in all cases an approved indication for laser by the American Academy of Ophthalmology and there have been no cases in which laser has not been given for an approved indication. In future studies, this could be addressed by examination of fundus photographs by investigators who are unaware of the treatment category. The study is planned to continue until there is 5-year follow-up. We believe this interim analysis is important because of potential significance of observations that an early decline in GFR is not observed and progression of vision-threatening retinopathy is less likely to occur after ICT.
In conclusion, results of this study suggest that human subjects have improved glycemic control after ICT compared with controls receiving intense medical therapy. Furthermore, some subjects have less progression of diabetic retinopathy. These data may serve as justification for a randomized, controlled trial to investigate the role of ICT to prevent complications of type 1 diabetes mellitus.
The authors are grateful to Maria Lopes Chau and Crystal Robertson for assisting with manuscript preparation.
1. Shapiro AMJ, Lakey JRT, Ryan EA, et al. Islet transplantation in seven patients with type 1 diabetes mellitus using a glucocorticoid-free immunosuppressive regimen. N Engl J Med
2000; 343: 230.
2. Hering BJ, Kandaswamy R, Ansite JD, et al. Single donor, marginal dose islet transplantation in patients with type 1 diabetes. JAMA
2005; 293: 830.
3. Froud T, Ricordi C, Baidal DA, et al. Islet transplantation in type 1 diabetes mellitus using cultured islets in steroid-free immunosuppression: Miami experience. Am J Transplant
2005; 5: 2037.
4. Ryan EA, Paty BW, Senior PA, et al. Five year follow-up after clinical islet transplantation. Diabetes
2005; 54: 2060.
5. Shapiro AMJ, Ricordi C, Hering BJ, et al. International trial of the Edmonton Protocol for islet transplantation. N Engl J Med
2006: 355; 1318.
6. Fiorina P, Folli F, Bertuzzi F, et al. Long-term beneficial effect of islet transplantation on diabetic macro- micro-angiopathy in type 1 diabetic kidney transplanted patients. Diabetes Care
2003; 26: 1129.
7. The DCCT Research Group. The implementation of treatment protocols in the diabetes control and complications trial. Diabetes Care
8. American Diabetes Association. Standards of medical care for patients with diabetes mellitus. Diabetes Care
2002; 25(suppl 1): S33.
9. Canadian Diabetes Association. 2003 clinical practice guidelines for the prevention and management of diabetes in Canada. Canadian Diabetes Association; 2003.
10. Thompson DM, Kozak SE, Sheps S. Insulin adjustment by a diabetes nurse educator improves glucose control in insulin-requiring diabetic patients: A randomized trial. Can Med Assoc J
1999; 161: 859.
11. Warnock GL, Meloche RM, Thompson D, et al. Improved human pancreatic islet isolation for a prospective cohort study of islet transplantation vs. best medical therapy in type 1 diabetes mellitus. Arch Surg
2005; 140: 735.
12. Perrone RD, Steinman TI, Beck G, et al. Utility of radioisotopic filtration markers in chronic renal insufficiency: Simultaneous comparison of 125I
Tc-DTPA and inulin. Am J Kidney Dis
1990; 16: 224.
13. Grading diabetic retinopathy from stereoscopic colour fundus photographs—An extension of the modified Airlie-House classification. ETDRS report #10. Early Treatment Diabetic Retinopathy Study Research Group. Ophthalmology
1991; 98: 786.
14. Wilkinson CP, Ferris FL, Klein RE, et al. Proposed international clinical diabetic retinopathy and diabetic macular edema disease severity scales. Ophthalmology
2003; 110: 1677.
15. Hansson P, Lindblom U, Lindstrom P. Graded assessment and classification of impaired temperature sensibility in patients with diabetic polyneuropathy. J Neurol Neurosurg Psychiatry
1991; 54: 527.
16. K/DOQI Clinical Practice Guidelines for Chronic Kidney Disease: Evaluation, Classification, and Stratification. Am J Kidney Dis
2002; 39: 51.
17. Ghofaili KA, Fung M, Ao Z, et al. Effect of Exenatide on beta-cell function after islet transplantation in type 1 diabetes mellitus. Transplantation
2007; 15: 24.
18. Froud T, Faradji RN, Pileggi A, et al. The use of Exenatide in islet transplant recipients with chronic allograft dysfunction: Safety, efficacy and metabolic effects. Transplantation
2008; 86: 36.
19. Fung M, Warnock GL, Ao Z, et al. The effect of medical therapy and islet cell transplantation on diabetic nephropathy: An interim report. Transplantation
2007; 84: 17.
20. Pierce IA, Ilango B, Sells RA, et al. Stabilization of diabetic retinopathy following simultaneous pancreas and kidney transplant. Br J Ophthalmol
2000; 84: 736.
21. Giannarelli R, Coppelli A, Sartini M, et al. Effects of pancreas-kidney transplantation on diabetic retinopathy. Transplant Int
2005; 18: 619.
22. Johansson BJ, Kernell A, Sjoberg S, et al. Influence of combined c-peptide and insulin administration on renal function and metabolic control in diabetes type 1. J Clin Endocrinol Metab
1993; 77: 976.
23. Ekberg K, Brismar T, Johansson B-L, et al. C-peptide replacement therapy and sensory nerve function in type 1 diabetic neuropathy. Diabetes Care
2007; 30: 71.