American Diabetes Association – 72nd Scientific Sessions

June 8-12, 2012; Philadelphia, PA; Report – Type 1 Therapies (Cure Related) – Draft

Executive Highlights

There were a few interesting data nuggets and updates on type 1 therapies presented at this year’s ADA, but not too much movement overall. Most notably, Dr. Tihamer Orban (Joslin Diabetes Center, Boston, MA) presented three-year results from a phase 2 study with abatacept in subjects newly diagnosed with type 1 diabetes (150-LB). Following treatment for two years with either abatacept or placebo and observation for an additional year, subjects treated with abatacept had significantly higher mean AUC C-peptide levels than those treated with placebo. We’ll look forward to learning more about abatacept’s (BMS’ Orencia, currently used for rheumatoid arthiris) potential in phase 3 testing, particularly as there was no difference in insulin dose between the study’s two groups at 36 months, despite C-peptide differences. Results from the DELAY Trial meanwhile excitingly suggested that the effect of teplizumab on C-peptide loss in type 1 diabetes is not limited to the new-onset period (85-OR). We wonder about the impact on Macrogenics from these results. Studies with IL-1 targeted therapies were not as successful. In phase 2 studies, the IL-1 receptor antagonist anakinra did not preserve C-peptide secretion in adults with recent onset type 1 diabetes, while the anti-IL-1 monoclonal antibody canakinumab did not have an effect in newly diagnosed subjects ages 6-45. As at last year’s ADA, presenters voiced interest in exploration of combination therapies for treatment of type 1 diabetes – we know there are quite a few hurdles (ranging from preclinical to regulatory) to overcome before such testing commences, but we’ll be eagerly waiting for word on what surely seems to be the next step for type 1 immune therapies. On the islet transplantation front, we found presentations on the CIT-07 islet transplantation protocol (156- OR), benefits of using liraglutide in transplantation (158-OR) and 11-year islet transplantation trends quite interesting – overall, this year’s ADA emphasized that islet transplantation field is certainly not speeding along, we are getting slowly getting smarter on what therapies and procedures might be helpful for improving engraftment, insulin independence, and outcomes.


Table of Contents 


Oral Sessions: Late-Breaking Abstracts


Tihamer Orban, MD (Joslin Diabetes Center, Boston, MA)

Dr. Orban presented results from the three-year follow up of a study that examined the effects of abatacept (CTLA4-Ig, BMS’ Orencia) on preservation of beta-cell function in type 1 diabetes. As a reminder, results from this trial presented at last year’s ADA suggested that abatacept could significantly preserve C-peptide compared to placebo after 24 months (for more details, see our ADA 2011 Full Report at Participants were randomized to 27 intravenous infusions of either placebo or abatacept over two years. Subjects were then not treated but followed for an additional year. The study remained double-blinded throughout. Adjusted mean AUC C-peptide was significantly higher in the abatacept than placebo-treated group at 36 months (p=0.033). Although abatacept treatment seemed to have a major effect at the beginning of treatment, after six months, rates of C-peptide decline became parallel between the study’s two arms. Abatacept treatment delayed decline in beta cell function by 9.5 months compared to placebo. Notably, participants treated with abatacept had significantly lower A1cs during the study. However, there was no difference in insulin dose among abatacept and placebo-treated groups at 24 through 36 months. In light of these findings, Dr. Orban suggested that abatacept might be useful in individuals at high risk of type 1 diabetes or as one component in studies testing combinations of treatment strategies.

  • Dr. Orban presented results from three-year follow up of a phase 2 trial examining the effects of abatacept on preservation of beta-cell function in type 1 diabetes. Results from this trial presented at last year’s ADA suggested that abatacept could significantly preserve C-peptide compared to placebo after 24 months (for more details, see our ADA 2011 Full Report at As a reminder, abatacept (CTLA4-Ig, BMS’s Orencia) is a selective T cell costimulation modulator that impairs the full activation of T cells. It has been approved in the United States for use in rheumatoid arthritis and juvenile idiopathic arthritis. It functions by binding to the CD80/86 complex, thus impairing the signaling between antigen presenting cells and T cells.
  • As a reminder, participants (n=112) in this trial were randomized to receive 27 intravenous infusions of either abatacept or placebo over two years. Infusions were given in an outpatient setting without any intravenous medication. Abatacept was administered at 10 mg/kg, with a maximum dose of 1000 mg. After treatment ended, subjects were followed for an additional year. The study remained double-blinded throughout. The study’s primary outcome measure was mean C-peptide area under the curve after two hours of a four-hour mixed meal tolerance test. Secondary outcomes included rate of change of C-peptide, time to first stimulated C-peptide less than 0.2 nmol/L, A1c levels and insulin doses at regular time points throughout the study, and several safety and mechanistic parameters. Out of 77 patients originally in the abatacept arm, 64 had 36-month data collected (73 had two-year data collected). Of the 35 patients originally in the placebo group, 29 had data collected at 36 months (30 did at 24 months). Furthermore, 69% and 66% of abatacept and placebo-treated patients, respectively, received all 27 study doses.
  • Adjusted mean C-peptide was significantly higher in the abatacept than placebo- treated group at 36 months (p=0.033). The mean AUC C-peptide for the abatacept group was 0.215 nmol/L [0.168, 0.265], while that for the placebo group was 0.135 [0.069, 0.205]. Notably, abatacept seemed to have a major effect only at the beginning of treatment. Although abatacept was given continuously for 24 months, declines in beta cell function became parallel between the study’s two groups after six months. According to Dr. Orban, this trend suggests that there is likely significant T-cell activation that the drug can have an effect on at the time of diagnosis. However, after that, the autoimmune process may be co-stimulation independent.
  • Abatacept treatment delayed C-peptide decline by 9.5 months versus placebo. Participants treated with abatacept had significantly better A1cs during the study. Additionally, significantly more abatacept than placebo-treated subjects had C-peptides above 0.2 nmol/L. However, there was no significant difference in insulin dose among the two groups at 24 through 36 months. Subgroup analyses showed no association between age or C-peptide entry status and three-year C-peptide level.
  • Based on these findings, abatacept might be useful in individuals at high risk of type 1 diabetes or as one component in studies testing combinations of treatment strategies.

Questions and Answers

Q: Do you have any data on hypoglycemia in the follow-up? If patients had more C-peptide, was there a change in hypoglycemia. Also, do you think six months of treatment would be enough in a phase 3 trial?

A: There was no difference in terms of metabolic complications for treated and non-treated patients. The second question is very good. It seems that you may not need a full two years of treatment to achieve an effect. We don’t know how long treatment needs to be though. We will need to try a shorter course and see if there is still an effect.

Q: Did you have CGM data?

A: No.

Q: A1c levels were not static in the trial. They started from around 6% and went up to 7.5 or 8%. There was a continuous increase over the three years. Can you talk about this?

A: I think that when you’re treating young children and teenagers, it’s pretty typical for A1c to gradually increase after two or three year post-diagnosis. It is partially due to the eventual loss of C-peptide. However, there was a difference in A1c between the two groups and A1c levels can have a huge impact on future complications.

Q: Since there were differences in A1c levels, maybe it is reasonable to assume that there were also differences in glucose levels between the two groups. Did you try to relate C- peptide to glucose levels?

A: We didn’t.

Q: Have you looked at any immune parameters from treatment to post treatment?

A: Yes, we did. We will be presenting this data in Vancouver in focus meetings though, so I would rather not say what the results are.


Oral Sessions: Human Immunology and Diabetes


Piotr Tronzkowski, MD, PhD (University of Lodz, Lodz, Poland)

Type 1 diabetes has been hypothesized to develop due to a combination of either low levels of regulatory T-cells (T-regs) or impaired T-reg function, or excessive numbers of autoaggressive cells. Dr. Tronzkowski and his colleagues explored whether correcting the imbalance between low levels of T-regs and high levels of autoreactive cells could stop or delay development of diabetes. Their study enrolled 11 patients five to 18 years old who had been diagnosed with type 1 diabetes within the last two months, had C-peptide levels of at least 0.4 ng/ml, and had at least one anti-islet autantibody. Subjects were infused with up to 20x106/kg autologous CD3+CD4+CD25highCD127- T-regs sorted with a single-use- sample-line FACS sorter and expanded under good manufacturing practice conditions. After administration, subjects’ peripheral blood T-reg levels almost doubled, suggesting a correct dose was used. Up to now, no adverse events have been noted with the therapy (this is the most important outcome since it is a phase 1 study). In terms of clinical outcomes, treated patients had higher mean C- peptide levels than non-treated controls used for comparison (n=122) starting at four months post- transfer, suggesting that therapy prolonged beta cell function. T-reg treated patients also had lower mean insulin doses than controls at two, four, and six months. One treated patient notably no longer needed insulin at four months. However, T-reg levels declined to baseline by 9 -12 months, which might have limited the efficacy of the protocol – Dr. Tronzkowski’s group has now received approval for a protocol that allows for re-injection of T-regs. The group is also exploring further method modifications.

Questions and Answers

Q: Did you do a functional assay?

A: We performed a functional assay, as well as a biological contamination assay to make sure the cell swere pure. The details of that will be published next month in Diabetes Care. All the details will be there.

Q: Did you check for CD127?

A: Yes, there was usually a small population of positive cells, but not more than 15-20%. They were very often negative.


Kevan Herold, MD (Yale University, New Haven, CT)

Successful immune therapy trials have generally enrolled subjects within three months of diagnosis because it was thought that therapies lost efficacy as beta cell mass declined. However, prospective studies have clearly shown that clinically significant insulin secretion is still detectable in patients after one year, suggesting that therapies might have an effect even outside the three-month window. In the “Delay” trial, Dr. Herold and his colleagues examined whether treatment with the anti-CD3 monoclonal antibody teplizumab could have an effect even after the new onset period (defined as later than 3 months post-diagnosis). This randomized, controlled trial tested whether a 14-day course of teplizumab could prevent C-peptide decline after one year in subjects with diabetes of four to 12 months duration. Patients in the study (n=58) were classified into two strata based on diabetes duration – those in stratum 1 were enrolled at four to eight months post-diagnosis and those in stratum 2 were enrolled at nine to 12 months. At twelve months, C-peptide AUC was significantly (21.7%) higher in drug versus placebo-treated patients, suggesting that the effect of teplizumab on C-peptide loss in type 1 diabetes is not limited to the new-onset period. Notably, stratum 1 patients drove this difference. A 37.2% improvement in C-peptide AUC was seen at 12 months in those teplizumab vs. placebo-treated patients with A1cs ≤ 6.5. However, no difference in this metric was observed between drug vs. placebo-treated patients with A1cs > 6.5%, at either six or twelve months. Drug efficacy was greatest in younger (8-14 year old) subjects. According to Dr. Herold, the basis for differences in response in younger subjects and those with lower A1cs is unclear.

  • Successful immune therapy trials have typically enrolled subjects within three months after diagnosis because it was thought that therapies lost efficacy when beta cell mass declined. However, this hypothesis had never been formally tested and it is now clear from prospective studies of insulin secretion that clinically significant insulin secretion is still detectable in most patients after the first year of disease.
  • In the “Delay” trial, Dr. Herold and his colleagues tested whether treatment with the anti-CD3 monoclonal antibody teplizumab could have an effect even after the new onset period (defined as after 3 months post-diagnosis). This randomized, controlled trial tested whether a 14-day course of teplizumab could prevent C-peptide decline after one year. It enrolled subjects (n=58 by the study’s end) with type 1 diabetes of duration four- 12 months, and a stimulated C-peptide of at least 0.2 pmol/ml. Randomization was 1:1 drug to placebo, by disease duration stratum (stratum 1 included patients with diabetes duration of four to eight months at baseline, while stratum 2 subjects had diabetes of duration 9-12 months at baseline). The study’s primary endpoint was difference in C-peptide in response to a mixed-meal test, adjusted for baseline, one year after enrollment. The average age of study participants was 12, which Dr. Herold said was important to note when comparing this study to previous ones. A potential confounding factor was that baseline A1c was different in drug vs. placebo treated patients. At one year, all subjects with detectable insulin were eligible to receive the active drug.
  • C-peptide AUC was significantly (21.7%) higher in drug versus placebo-treated patients after 12 months. Notably, stratum 1 patients drove this difference. Drug-treated subjects used significantly less insulin at twelve months, again driven by those whose diabetes duration was four to eight months. There was no significant effect of treatment on A1c, either in terms of baseline adjusted A1c or change in A1c.
  • Interestingly, a 37.2% improvement in C-peptide AUC was seen at 12 months in teplizumab vs. placebo-treated patients with A1cs ≤ 6.5. However, no difference in this metric was observed in treated vs. placebo patients with A1cs > 6.5%, at either six or twelve months. Notably, drug efficacy was greatest in younger (8-14 year old) subjects. Additionally, treatment was associated with an increase in IL-10 secretion by CD8+ T cells and clinical responders could be distinguished from nonresponders by an increased in absolute number of CD8 central memory cells. Reported grade 3 and 4 adverse events associated with treatment included transient cytopenia, cytokine release, and drug hypersensitivity. In general, adverse events were well tolerated, not frequent, and reversible.

Oral Sessions: Novel Agents for Diabetes Management


Paolo Pozzilli, MD (Barts and the London School of Medicine and Dentistry, London, UK)

Dr. Pozzilli reported results from the DIA-AID1 phase 3 trial for Andromeda’s type 1 diabetes therapy DiaPep277. As a reminder, the study randomized 457 adolescents or adults (16-45 years) recently diagnosed with type 1 diabetes (within three months of screening) to receive either 1.0 mg DiaPep277 or placebo for two years. At baseline, average age was 27 years, time from diagnosis was 2.8 years, and A1c was 7.3% in the DiaPep277 arm and 7.5% in the placebo arm (p=ns). Topline results previously announced demonstrated that DiaPep277 treatment led to: 1) a statistically significant preservation of stimulated C-peptide secretion (via a glucagon-stimulated test) over placebo (23.4% relative change; p=0.037; mITT analysis) – the study’s primary endpoint; 2) a greater percentage of individuals achieving an A1c < 7% at the study’s (56% vs 44%; p=0.035); 3) no significant improvement in fasting C-peptide over placebo; 4) no significant improvement in mixed-meal stimulated C-peptide over placebo; 5) a lower rate of hypoglycemic events in the per protocol population over placebo (p<0.05); and 6) a greater percentage of individuals achieving a composite endpoint ofA1c < 7.0% and an insulin dose < 0.5 IU/kg/day in the completer population over placebo (41% vs 30%, p=0.035). Additional results reported today by Dr. Pozzilli included than an even greater preservation of glucagon- stimulated C-peptide occurred with DiaPep277 in the completer population over placebo (29% relative change; p=0.011) and that there were no significant differences in the occurrence of overall adverse events, drug related adverse events, or severe adverse events between the groups. Interestingly, during Q&A, Dr. Pozzilli hypothesized that the non-significant result for change in mixed-meal stimulated C- peptide may have been due to stimulation of the incretin axis with the mixed-meal test. For more details on the DIA-AID1 study design, topline results, and development program, please see the November 22, 2011 Closer Look at We also note that DiaPep277 was granted orphan drug status by the FDA in May.

Questions and Answers

Q: Was there any change in weight?

A: There was no change in weight.

Q: Have you look at differences in efficacy depending on the number of autoantibodies expressed?

A: We are examining that right now.

Q: Were there any elevations in adverse events at all in the DiaPep277 arm?

A: There was some bruising near the infusion site. However, we found the therapy to be very safe.

Q: Have you had the chance to look at different subgroups yet, like age, C-peptide, duration of diabetes, HLA haplotype, etc.?

A: We are still looking into that. There are certainly a number of interesting subgroups that we can look at that might show differential efficacy.

Q: Can you speculate why you didn’t see anything in the mixed-meal tolerance test?

A: The glucagon stimulation test has been used for many years as the gold standard test for C-peptide response. That is particularly true in adults. The fact that there was no difference in C-peptide with the mixed-meal test leads me to speculate that the mixed-meal test may have stimulated the incretin axis. We are unsure of what the GLP-1 response was in this patient population and how the GLP-1 response impacted beta cells. With the glucagon stimulation test, however, you get a more direct hit on beta cells. So, these tests rely on different mechanisms. The glucagon stimulation test is less affected by age, body weight and a number of other factors compared to the mixed-meal test.

Q: Why did you choose a 16 years and up population?

A: This decision was based upon a previous trial conducted in children that demonstrated DiaPep277 was not effective in preserving glucagon stimulated C-peptide. The idea was just to focus on the adult population, in which DiaPep277 was suggested to be effective in preserving glucagon stimulated C-peptide in a separate trial. Given that the drug is quite safe, my personal view is that we have to reconsider using DiaPep277 in specific subgroups of patients with new onset disease at a younger age. For example, we could look at individuals with low risk HLA haplotypes and a lower number of expressed islet autoantibodies. If numerous autoantibodies are already expressed and an individual possesses strong HLA haplotypes, the drug probably won’t work.

Oral Sessions: Clinical Islet Transplantation


Michael Rickels, MD (University of Pennsylvania School of Medicine, Philadelphia, PA)

Dr. Rickels presented data comparing beta cell secretory capacity in people with type 1 diabetes who had received an islet transplantion with the Clinical Islet Transplantation 07 (CIT07) protocol vs the Edmonton protocol. The two protocols differed in the induction therapy, anticoagulation therapy, and insulin therapy used, but both protocols utilized the same immunosuppression maintenance therapy (see below for details). A glucose-potentiated arginine test was used to assess beta cell secretory capacity at 75 days post-transplant in 11 people that underwent the CIT07 protocol and five individuals who received an islet transplantation with the Edmonton protocol. Insulin response after the glucose- potentiated arginine test was significantly greater (p<0.05) with the CIT07 protocol than the Edmonton protocol. Based on this data, Dr. Rickels concluded that engrafted islet beta cell mass was markedly improved with the CITo7 protocol. Moreover, because fewer islets were transplanted in patients that underwent the CIT07 protocol, he argued that a greater proportion of islets appeared to survive with this protocol compared to the Edmonton protocol. While he could not discern which aspect of the CIT07 protocol conferred the improved results, he suggested that it was likely that each played a role.

  • Although the Edmonton and CIT07 protocols used similar immunosuppression maintenance therapies, the protocols used different immunosuppression induction therapies, anticoagulation therapies, and insulin regimens. Dr. Rickels drew particular attention to the immunosuppression induction therapy used in the CIT07 protocol. Anti- thymocyte globulin therapy (6 mg/kg divided) began at two days pre-transplant and continued for two days post-transplant. Etanercept therapy (50 mg IV f/b 25 mg SQ x 3) began at implantation and continued for 10 days.

Table 1: Edmonton vs CIT07 Protocol


Edmonton Protocol

CIT07 Protocol

Induction Therapy


Anti-Thymocyte Globulin Etanercept

Maintenance Therapy

Tacrolimus (low dose) and Sirolimus

Tacrolimus (low dose) and Sirolimus

Anticoagulation Therapy

Heparin with Islets

Heparin for One Week

Insulin Regimen

For Hyperglycemia

Intensive Insulin Therapy for Two Months

  • The study used glucose-potentiated arginine tests at 75 days post-transplant to measure beta cell secretory capacity. Five individuals with type 1 diabetes received a transplant through the Edmonton protocol (14,313 islet equivalents/kg body weight), and 11 individuals with type 1 diabetes received a transplant though the CIT07 protocol (10,090 islet equivalents/kg body weight). 11 individuals without type 1 diabetes who did not receive an islet transplant were also included in the study as an additional comparison. At 75 days post- transplant, average age was 34 years, A1c was 6.1%, FPG was 100 mg/dl, and BMI was 21 kg/m2 in the Edmonton protocol arm. In the CIT07 protocol arm, average age was 30 years, A1c was 5.6%, FPG was 103 mg/dl, and BMI was 23 kg/m2. In the non-diabetic arm, average A1c was 5.3%, FPG was 81 mg/dl, and BMI was 24 kg/m2. Of note, the average number of islet equivalents per kg body weight received in the CIT07 arm was significantly less than that received in the Edmonton protocol arm (p<0.05).
  • Insulin response after the glucose-potentiated arginine test was markedly greater (p<0.05) with the CIT07 protocol than the Edmonton protocol. While exact numbers were not given, we estimated about a 20 μU/ml increase in acute insulin response based off the chart shown. However, the CITo7 protocol still had a significantly less insulin response (p<0.001) compared to the group without type 1 diabetes. There was no statistically significant difference observed in insulin response following the non-glucose potentiated arginine test between the Edmonton and CIT07 arms. At the end of his presentation, Dr. Rickels noted that seven participants in the CIT07 arm had completed reassessment one year post-transplant, and each exhibited similar or greater beta cell secretory capacity (based on the glucose-potentiated arginine test). Based on these results, Dr. Rickels concluded that maintenance immunosuppression with low-dose tacrolimus and sirolimus (as introduced by the Edmonton protocol) may not be toxic to transplanted islets. While he could not discern which aspect of the CIT07 protocol conferred the improved results, he suggested that it was likely that each played a role.


Peter Senior, MD, PhD (University of Alberta, Edmonton, Canada)

Dr. Senior presented the results from a study assessing combined sitagliptin and pantroprazole therapy on islet regeneration after transplant. Eight individuals (68 kg [149.6 lb], 2 islet transplants, 57 months since last transplant; 6.3% A1c) with early islet graft dysfunction were treated with sitagliptin (100 mg daily) and pantoprazole (80 mg daily – editor’s note – the slide said 80 mg, but the abstract said 40 mg) for six months. Dr. Senior defined early graft dysfunction as A1c levels > 6% and < 7.5%, fasting glucose >7 mmol/L (126 mg/dl), or random glucose >10 mmol/L (180 mg/dl) and total insulin use < 10 units/day. Glycemic control and graft function were assessed at baseline, three months and six months after intervention, and following a three-month washout period. Dr. Senior described the several trends observed over these time intervals. First, while A1c decreased after one month of treatment, A1c levels plateaued through the remaining treatment course and increased when treatment was withdrawn. Second, he found no statistically significant differences in mean fasting and stimulated C- peptide/glucose ratios throughout the study. Third, while CGMS data showed no significant change in sensor detected average glucose levels, he suggested three reassuring findings: 1) the sensor data suggested relatively good glycemic control during the study (6.8 mmol/L [122.4 mg/dl], 7.1 mmol/L [127.8 mg/dl], 6.8 mmol/L [122.4 mg/dl], and 7.0 mmol/L [126 mg/dl] at baseline, three, six, and nine months, respectively); 2) treatment did not increase the proportion of time spent in hypoglycemia; and 3) standard deviation was numerically lower with treatment suggesting no deterioration in control. Fourth, Dr. Senior pointed to five patients who achieved insulin independence after six months of treatment. However, when the drugs were withdrawn, these patients resumed insulin use with increased dose requirements. Finally, Dr. Senior noted that the therapy was safe and well tolerated with no significant side effects (although there were two cases of transient mild gastrointestinal side effects) or weight changes. Dr. Senior concluded that while treatment may improve graft function, improvements were not sustained when treatment was withdrawn. Furthermore, he saw no evidence for beta cell regeneration with sitagliptin and pantoprazole intervention.

Questions and Answers

Q: While there was no sustained regeneration, there was no parallel group you didn’t treat.

A: We chose a very narrow number. There is the potential we’ve slowed down a decline we might have seen otherwise.

Q: The only point changed was people being transiently taken off insulin. What were the criteria for a patient to be taken off insulin?

A: Stopping insulin is subjective criteria. We were fairly strict in terms of blood glucose parameter, pre- meal and post-meal levels, which is why it’s important we’ve taken them off insulin. Indeed A1c margin did better in the first month after treatment. When you remove the DPP-4 inhibitor, a number came back saying they didn’t have as good control, granted that is anecdotal.


Eduardo Peixoto, MD (University of Miami Miller School of Medicine, Miami, FL)

Dr. Peixoto described a prospective study investigating the effect of liraglutide on islet graft function, glycemic control, and gastric emptying in a subset (n=6) of islet transplant recipients. Participants were switched from exenatide (stopped three days before baseline, poorly tolerated due to nausea) to 0.6 mg liraglutide and progressed to 1.2 mg liraglutide. While all patients achieved the intended study dose, two patients had to return to a lower dose – one due to constipation and one due to diarrhea. Liraglutide’s effects were assessed by multiple parameters at baseline and six months. Dr. Peixoto drew attention to an improved stimulation C-peptide to glucose ratio (p=0.03) and an improved GLP-1 response to a mixed meal tolerance test (both a decreased time to peak levels and a higher peak; p <0.01) from baseline. Dr. Peixoto suggested that the reduced time to peak GLP-1 finding might suggest that liraglutide acts to increase endogenous GLP-1 secretion. He noted that A1c, C-peptide, glucagon, and acetaminophen absorption during MMTT did not change significantly. Dr. Peixoto also revealed that while most participants experienced increased appetite, most also lost weight (-2.85 kg (-6.27 lb). He concluded that the higher GLP-1 levels (basal and stimulated) observed with liraglutide treatment could lead to longer islet survival and looked forward to 12- month results, which he indicated would be reported shortly.

  • Dr. Peixoto detailed a prospective study testing the effects of liraglutide intervention in a subset of islet transplant recipients (n=6) with active graft function. Participants switched from exenatide use (stopped at least three days before baseline) to daily liraglutide, starting at 0.6 mg and progressing to 1.2 mg after one week. Insulin, glucose, c-peptide, glucagon, and GLP-1 levels after 300 minute MMTT were assessed as well as body composition, A1c, and glucose finger stick readings at baseline and 6 months after intervention. He noted the six participants represented a group with a markedly prolonged duration of type 1 diabetes (26-54 years).
  • Dr. Peixoto commented that liraglutide was tolerated in the majority of subjects. While all patients initially achieved the study dose of 1.2 mg daily, dose levels were decreased in two patients – one due to constipation and one due to diarrhea. Exenatide had been poorly tolerated in participants, most experiencing nausea with use. No nausea was observed with liraglutide.
  • Compared to baseline, Dr. Peixoto pointed to a significantly increased stimulation index (AUC C-peptide/AUC glucose) and a significantly lower time to peak GLP-1 levels (60 minutes) during a MMTT (p=0.03 and p <0.01 respectively). Mean GLP-1 levels also increased with liraglutide treatment during fasting (1.65 vs 8.65 pM/l at baseline and six months, respectively; P>0.05) and during a MMTT (13.95 vs 42.73 pM/l at baseline and six months, respectively; P<0.01). While an overall increase in GLP-1 levels was expected, the reduction in time to peak GLP-1 levels was most striking to Dr. Peixoto. He commented that this result potentially indicates that liraglutide acts to increase endogenous GLP-1 secretion. Glucose, c-peptide, and glucagon response to MMTT and A1c did not change significantly from baseline (baseline A1c 6%). Additionally, patterns in acetaminophen absorption did not change from baseline.

Questions and Answers

Q: Do you know if this subgroup experienced a benefit from exenatide or was their inability to tolerate the drug such that they had no response to exenatide?

A: They did not all use same dose. So yes, they had benefits, but for us it was a greater benefit with liraglutide because we were able to achieve the target dose.

Q: It is striking that there wasn't a stronger effect of liraglutide. Could that be that they had been on exenatide?

A: Yes, we were discussing that. At baseline, we know they didn’t have exenatide in their blood. But as for the chronic effects yes, you’re right. However, there was no difference in A1c.

Meet the Expert Sessions


George Eisenbarth, MD, PhD (Barbara Davis Center, Aurora, CO)

In this highly interactive session, Dr. Eisenbarth answered audience questions about the pathogenesis and prevention of type 1 diabetes. Discussion during the session touched upon a number of epidemiological and clinical topics, including the percent of beta cells remaining at diagnosis, the hygiene hypothesis, genetic or immunological factors that determine development of complications, and how one explains that diabetes is occurring more often both earlier in life and in the third through sixth decades. The audience also posed interesting questions related to immunomodulation and therapies that could prevent or reverse type 1 diabetes, including BCG vaccination, exogenous insulin administration, anti-CD3 therapy, glucagon antagonism, and vaccines against the trimolecular complex. Most notably, Dr. Eisenbarth predicted during the session that we will be preventing type 1 diabetes within twenty years, either through a combination of drugs or small molecules that target the pocket in which T-cells recognize peptides.

Questions and Answers

Q: What percent of beta cells are remaining when type 1 diabetes is diagnosed?

A: I think about 20%. There are a fair number of beta cells remaining at the onset of type 1 diabetes, and the amount varies between individuals. But I think there are quite a few beta cells worth saving.

Q: How do you explain that diabetes comes earlier and earlier in life now?

A: We don’t know at this point in time. I like the general hypothesis that our environment has changed. It’s clearly an environmental factor. Perhaps kids are not exposed to as many early infections as they used to be.

Q: Can you talk more about the hygiene hypothesis, and especially talk about Bacteroides fragilis and if it is a key player that gets T-regulatory cells altered?

A: There are many components of the hygiene hypothesis. The question is what bacteria are doing in the induction of disease. What is clear in animal studies is that the microbiome affects immune development, often in a pathogenic way. I think it’s going to be a very complicated issue. I will be very surprised if something as simple as a long-term change in the microbiome has changed the risk. We are so ignorant right now. For example we don’t know if the development of autoimmunity has been going up steadily or if it has been going up stepwise. We would have a very different hypothesis if it was stepwise. We can’t just use development of diabetes as the indicator, because the lag time from development of autoantibodies to diabetes can take a decade. A child with autoantibodies at nine months can develop diabetes at nine years of age or 14 years of age. I think that complicates our understanding of why children are developing diabetes earlier. We don’t know for sure if what has changed is the rate or progression from the first antibody or when the first antibody occurs.

Q: Can you talk about the effect of supplementation of vitamin D?

A: There are contradictory studies like the DAISY studies from Scandinavia that have suggested a relationship between vitamin D risk and risk of type 1 diabetes. Vitamin D is a potent immunomodulatory drug. It affects antigen-presenting T-cells in particular. There are groups that are both looking at vitamin D itself and modified forms of vitamin D. Dr. Robyn Lucas’s group, for example, uses vitamin D as a therapeutic on cells in vitro.

Q: Dr. Faustman’s group at Massachusetts General Hospital is testing the hypothesis that BCG could have a role in regenerating beta cells. I don’t know whether you have any opinion about their work.

A: I am not aware of work with BCG regenerating beta cells specifically versus its immunologic effects. A lot of these studies started in the NOD mouse model. But that is a very easy model to affect diabetes in. At the Barbara Davis Center, we did a trial with BCG. It didn’t work. We might not have used the right dose or the right amount, but I am still a skeptic of BCG.

Q: What is the most likely reason we are seeing more people diagnosed in their third, fourth, fifth, and sixth decades of life. What is keeping the autoimmune response dormant?

A: So actually Dr. Andrea Steck recently published data about this from the DAISY study. She was trying to predict the age of onset of type 1 diabetes and why someone can go 20, 30, 40 years with antibodies and then finally present with diabetes. We have had identical triplets that we started publishing on in 1986. Now all three triplets have diabetes. The first triplet was diagnosed as a child, and the last triplet developed it at 57. These are genetically identical triplets that still vary in the spectrum of developing diabetes. Anyways, what Dr. Steck found is that we can predict the approximate age of onset of type 1 diabetes with two variables. The first is the age at which autoantibodies first appear. That is straight forward, because we believe that when autoantibodies first appear, insulitis begins killing beta cells. Second is the level of insulin autoantibodies, not GAD or IA2 or ZT8. The suggestion is that those two factors help explain the age of onset. Slowness of development of diabetes relates to patchiness of killing. What this means is that if you look at the pancreas of someone with type 1 diabetes, it looks like vitiligo. You have an area where there are no beta cells, and a lobule where there are beta cells. Over time, more and more lobules are targeted and beta cells carved out. We don’t know what all the rate controls are, but we know they correlate with levels of insulin autoantibodies. We used to draw it as a straight curve, now we draw it as a squiggly line because levels of autoantibodies vary over time.

Q: What about patients who remain free of chronic complications? Is this genetically determined or is this related to the immunological response?

A: There is a very beautiful study by George King called the Medalist Study from Joslin. It looks at patients who survived type 1 diabetes for more than 50 years to look for potential explanations. The very simple hypothesis is that those who keep more C-peptide have fewer complications.

I would also mention that within type 1 diabetes, there are patients who have other types of diabetes. Most patients with type 1 diabetes have islets that lack beta cells. But there are a significant number of Hispanic and African American patients diagnosed with type 1 diabetes whose islets all have beta cells and who don't have autoantibodies. A type 2 diabetic would also look like this.

Q: As far as I know, there is no good strategy for preventing diabetes based on immunomodulation?

A: We don’t have an immunosuppressive or immunomodulatory therapy for type 1A diabetes. It is very clear that the therapies that have been sued to date delay the loss of beta cells. Anti-CD3, anti-CD-20, and abatacept stop C-peptide loss for six months. Then the beta cell loss line is parallel between when placebo and drug is given. I would say that this type of delay is not adequate even though you can still see an effect 2-3 years after. It is not sufficient risk-benefit wise.

Q: With all the work that is going on, in the next twenty years, do you see a turnaround of something that can reverse or prevent type 1 diabetes?

A: I think yes. I see two parallel pathways. People from the US know what a Hail Mary pass is. If we’re lucky, one of our Hail Mary passes might be a combination of drugs. It will be successful and we will build on that. On the other hand, we need a lot more basic knowledge of what is going on. Right now, we know that antibodies aren’t causing the disease but T-cells are. But we have no really great T-cell assay. When we do trials we have to wait several years to use C-peptide. I would say though that the structures that are primarily causing the disease are getting to be known. When the T cell sees an antigen, the HLA molecule is doing the presenting so the class 2 molecule binds the peptide, and the T-cell recognizes it. What is peculiar is that the T-cell sees the peptide in a specific register. It is a very specific interaction that involves side chains going into specific pockets. Now that we know this structure, we can design small molecules to target these pockets.

Q: What are your thoughts on antagonizing glucagon signaling and “reversing” type 1 diabetes.

A: I think we can reverse it this way only when we have some beta cells left. But for patients with 99.9% of their beta cells gone, that might not be effective.

Q: Do you think it’s possible to produce a vaccine against the T-cell that can interfere with pathogenesis?

A: We are trying to make antibodies against the complex. We can prevent half of diabetes in the NOD mouse. There are a wealth of ways we are targeting that structure.

Q: Assuming that preventing type 1 diabetes is expensive or may have side effects, what are the criteria will you use to determine who to use this for?

A: One advantage is that we can now pretty well predict which patients will get type 1 diabetes, and the assays are getting simpler and simpler. The assays themselves can be done on fingersticks. It can be done relatively easily. The studies of prevention are going to go back from the highest risk groups to the lowest risk groups. As we get smarter, we try to identify the simplest ways to achieve prevention. If a virus is a factor, we may want to have trials of vaccination for certain Coxsackie viruses. It will depend where the science goes. The one other hard part is that the easiest time to prevent type 1 diabetes is before autoantibodies appear. And the first trials of that type of prevention are being done with environmental factors, but we also have oral insulin. In the Pre-POINT study, oral insulin is being given to individuals who have a 50% risk of developing type 1A diabetes. However, the reality is that the best time to have generalized prevention would be before the immune switch even occurs. To do this, you would have to use HLA, family markers, and other genes.

Q: Is there evidence that early use of exogenous insulin can delay the development of diabetes?

A: The DPT Trial examined that, and it did not delay diabetes. With oral insulin, there may be a delay. TrialNet is hopefully going to close their enrollment and conduct a ten-year study. The previous TrialNet study with oral insulin was for oral tolerance. In a subset analysis of the study, it was seen that individuals who came into the study with autoantibodies and were given oral insulin had a seven-year delay in disease onset.

The one thing we can do at the onset of diabetes is intensive insulin therapy. And the DCCT showed that quite a while ago and hopefully we can do the intensive insulin therapy better. TrialNet is using in- hospital insulin at diagnosis and continuous glucose monitoring and intensive insulin therapy to study this.

Q: How do you explain the fact that the program that used the otelixizumab molecule was stopped prematurely?

A: All of the monoclonal antibodies target more generalized molecules like CD3; none of them target specific molecules. In terms of anti-CD3, there were two large phase 3 trials and both had certain flaws. Either they used one-tenth of the dose used in phase 2 trials or they changed their outcome variable. Thus far, it is pretty consistent that C-peptide decline can be delayed.

Q: There have been a few studies in mice regarding alpha 1-antitrypsin, and there is one clinical trial ongoing. Do you have any insight on how it works?

A: There are two studies that I know of. The immune tolerance network is doing a study, and Peter Gottlieb is doing a small study at the Barbara Davis Center. In terms of how it works, the alpha 1- antitrypsin circulates at a pretty high level. The question is when you give it as a drug, how much do you increase that level. Is it sufficient? And in man, we obviously don’t know yet. The first studies are trying to look for a biological signal. And in the animal model, I’m not sure at all how it’s working. I’m told that Dr. Strom has done a key experiment where alpha 1-antitrypsin is given to the NOD mouse. I would say that there are a number of foreign agents that prevent diabetes when given to the NOD mouse though.

Symposium: Beta Cell Death, Cytokine Toxicity, and Islet Antigenicity


Thomas Mandrup-Poulsen, MD, PhD (Steno Diabetes Center, Gentofte, Denmark)

Dr. Mandrup-Poulsen summarized findings from a phase 2 trial of the IL-1 receptor antagonist anakinra in adults with recent onset type 1 diabetes. Subjects in the trial (n=69) were randomized 1:1 to receive either once-daily 100 mg recombinant human IL-1Ra (anakinra) or placebo on top of ADA- specified antidiabetic therapy. The study’s primary endpoint was change in standardized mixed-meal stimulated C-peptide (as measured by the AUC from 0-120 minutes) after 1, 3, 6, and 9 months, while secondary endpoints included peak C-peptide, time to peak C-peptide, insulin requirement, A1c, FPG, and IL-6 and CRP levels at the same timepoints. Notably, the trial’s treatment and placebo groups were imbalanced post-randomization for baseline AUC C-peptide, A1c and totally daily insulin dose. All endpoint analyses were adjusted to account for this. After nine months, there was no difference in loss of AUC C-peptide between the study’s two groups. There were also no differences in any of the study’s secondary endpoints between anakinra and placebo-treated patients. In terms of adverse events, significantly more anakinra than placebo-treated patients experienced skin reactions (p<0.0009). According to Dr. Mandrup-Poulsen, possible interpretations of these negative results are that 1) IL-1 blockade alone may be more efficacious in autoinflammatory type 2 diabetes than in autoimmune type 1 diabetes, 2) IL-1Ra therapy was instituted too late, or 3) higher doses of IL-1 blockade may be needed (the dose used in this trial is FDA-approved for rheumatoid arthritis) for the therapy to be effective in type 1 diabetes. Dr. Mandrup-Poulsen concluded by suggesting that the time may have come to explore combining IL-1 blockade with adaptive immunomodulation; this view, he noted, is supported by recent findings that combining IL-1 blockade with anti-CD3 reverses diabetes in NOD mice better than either therapy alone (Ablamunits et al., Diabetes, 2012).

Questions and Answers

Q: Was there any evidence of an effect on inflammatory markers?

A: As noted from the baseline characteristics slide, serum IL-6 and CRP were very elevated. Since the treatment didn’t show an effect on inflammatory markers, we expect that the lack of effect was not due to that, but to systematic inflammation.

Q: My question is about the age of the patients. Could this also be a factor since the study was in adults and not children?

A: You will now hear a report about anti-IL-1 in a young age group.



Antoinette Moran, MD (University of Minnesota Medical School, Minneapolis, MN)

Dr. Moran presented the results of a randomized, double-blind controlled trial that examined whether canakinumab (an antibody targeting IL-1β) could preserve beta cell function in recent onset type 1 diabetes. Study participants (n=69; ages 6-45) were randomized 2:1 to receive either canakinumab or placebo. At baseline, participants were about 2.5 months from diagnosis. Notably, due to high enthusiasm, enrollment for the trial was accomplished over just five months. Compliance was also excellent in the trial, with 91% of 828 potential injections administered. The study’s primary outcome was adjusted geometric mean AUC C-peptide at one year as measured from the first two hours of an MTT. Secondary outcome measures included geometric mean C-peptide AUC, rate of change of C- peptide, time to first stimulated C-peptide, and A1c at months 3, 6, 9, and 12. At the end of twelve months, canakinumab was not found to preserve C-peptide secretion significantly better than placebo (p=0.88). It similarly had no effect on A1c. Interestingly, patients who had more C-peptide and lower A1cs at baseline responded significantly better to canakinumab treatment than those with lower C- peptide or higher A1cs. No serious adverse events related to the medication were reported. Additionally, rates of neutropenia (which investigators were concerned about because of issues seen in gout trials) were similar in the study’s two arms. Dr. Moran noted that despite the trial’s negative results, she is not ready to give up on canakinumab. Again invoking Ablamunits et al.’s paper in Diabetes, she suggested that anti IL-1 β therapy used as part of a rational combination therapy could still be useful for tackling the type 1 diabetes autoimmune process. She also suggested that cankakinumab might have been used too late in this trial, and might be most effective in pre-type 1 diabetes.

Questions and Answers

Q: There was really high compliance in the trial. Can you put this in the context of using canakinumab in a prevention trial?

A: Once you have a child who has diabetes, there’s a very high level of motivation. These families were quite motivated and really the nuisance of coming into clinic was probably the biggest problem. Injections, etc. were less of a problem.

Q: I wonder if you did any work looking at inflammatory markers.

A: We don’t have that yet. The last patient came in at the end of April, so we were scrambling to assemble the C-peptide data.

Q: Did you expect to see a honeymoon phase? It didn’t look like children had it. Do you think that could have affected study design in some way?

A: I think they were in a honeymoon period when we started the drug. The average insulin dose then was 0.3 units/kg of insulin per day, which is a small dose per children. We caught them in the honeymoon phase, and then both the placebo and canakinumab-treated groups acted the same.

Symposium: ADA Diabetes Care Symposium


Weiqiong Gu, MD (Shanghai Ruijin Hospital, Shanghai, China)

Dr. Gu presented the results of a study that examined differential success of autologous hematopoietic stem cell transplantation (AHSCT) for treatment of recent onset diabetes in subjects with and without a history of ketoacidosis. According to Dr. Gu, the purpose of the protocol was to stop the autoimmune destruction of beta cells with high dose immunosuppressive therapy and subsequently reset the deleterious immune system with a reconstituted one originated from autologous hematopoietic stem cells. 28 patients (11 with and 17 without a history of prior ketoacidosis) with a mean age of 17.6 years and diabetes duration of less than six months were enrolled in the trial. There were no significant differences in age, sex, BMI, duration of diabetes, and baseline A1c levels between the DKA and non- DKA groups, though there was a trend towards decreased GAD levels in the non-DKA versus DKA group. Stem cells for the protocol were mobilized with cyclophosphamide and GCSF and then collected from peripheral blood. Patients were conditioned with cyclophosphamide (200 mg/kg) and rabbit antithymocyte globulin (4.5 mg/kg), infused intravenously with collected stem cells, and followed-up with for 4-42 months. Post-treatment, the overall response rate (defined as the proportion of patients experiencing a reduction of daily insulin dose greater than 50%) was 83.7% ± 7.2%. 58.6% ±9.5% of patients had their insulin doses reduced by more than 75% (a “very good response”), while 53.8% ±9.5% of patients experienced insulin independence after treatment. C-peptide concentrations were significantly greater than pre-treatment in all subjects. Notably, non-DKA patients experienced significantly greater insulin independence post-treatment than those who had had DKA in the past (70.6% vs. 27.3%, p= 0.031). Dr. Gu suggested that this might be explained by diabetic ketoacidosis decreasing the capacity for beta cell recovery. Non-DKA patients additionally had significantly enhanced C-peptide levels, Cmax (peak value during OGTT), and AUCC (area under C-peptide release curve during OGTT) that remained elevated during 24 months of follow-up. Treatment was associated with a number of side effects. All patients had digestive tract symptoms, including nausea, vomiting and diarrhea, most had febrile neutropenia, alopecia, and bone marrow suppression, and 38.5% had blood transfusions. No severe drug toxicity or organ damage was reported.

Questions and Answers

Q: What you showed is a source of tremendous debate. My question for you is do we know that the stem cells themselves are actually working? Could it not be that the effect you’re seeing is directly attributable to the intensive immunosuppressive regimen that you used?

A: We don’t know how the hematopoietic stem cells differentiate in the body but we can make sure that this kind of treatment is a combination therapy combined with immunosuppressive drugs and the hematopoietic stem cells. I think in this therapy, the suppressive drugs such as ATG play an important role. We all know ATG can increase C-peptide, as seen in former clinical trials. Meanwhile, standard hematopoietic stem cells may be differentiated into T-lymphocytes and can secrete some soluble cytokines, which can promote beta cell regeneration.

Q: What was the maximum duration of disease of the patients entered in the protocol?

A: The maximum should be less than six months.

Q: Another group, the one that used the Voltarelli protocol, observed a relapse at sixteen months.

A: In our data, the relapse rate is relatively high, maybe seven or fifteen patients. Really, the net benefit of AHSCT needs to be further studied.


Michael Rickels, MD (University of Pennsylvania, Philadelphia, PA)

Dr. Rickels described trends in safety and efficacy outcomes of 677 recipients who had received either islet alone or islet-after-kidney transplants during three eras: 1999-2002, 2003-2006, and 2007-2010. Data was derived from the Collaborative Islet Transplant Registry (CITR). Measured outcomes included insulin independence, C-peptide levels, fasting glucose, A1c levels and percent A1c change, absence of severe hypoglycemia, vital status, serious adverse events, and GFR. The registry’s data notably shows a spike in enthusiasm for transplantation after the Edmonton Protocol, a drop thereafter, and a spike again in 2007. In terms of baseline characteristics, recipients’ age and diabetes duration have increased from 1999-2010. Rates of severe hypoglycemia prior to transplant have remained steady at 90% over the period, while use of lipid-lowering drugs has increased significantly over the era. Notable trends in islet donation include an increase in donor BMI to the obese range, increased use of islet culture and longer islet culture durations, and an increase in the number of islet equivalents per infused transplant. Use of calcineurin + mycophenolate maintenance immunosuppression has increased in the latter half of this past decade, while use of IL2RA induction immunosuppression has decreased. Insulin independence post-transplant has increased significantly by era (from 27% in 1999-2002, to 37% in 2003-2006 and 44% in 2007-2010; p= 0.02). Factors predictive found to be predictive of insulin independence over the era include total infusions, recipient age, culturing islets for ≥ 6 hours, use of T-cell depletion + a TNF-α inhibitor, and a stimulation index of ≥ 1.5. Islet reinfusion rates significantly decreased between 1999- 2006 and 2007-2010 (p < 0.01). As insulin independence has increased from 1999-2010, graft survival with positive C-peptide has increased accordingly. The percentage of patients with an A1c ≤ 6.5% or which fell by 2% post-transplant increased significantly from 1999-2002 to 2003-2006 (p=0.03). Mortality has not significantly decreased by era, while rates of serious adverse events have (p=0.02). Kidney function has remained relatively stable in all three eras.

Symposium: Beta Cell Replacement Therapies for Severe Hypoglycemia Unawareness


Camillo Ricordi, MD (University of Miami, Miami, FL)

Dr. Ricordi described the Clinical Islet Transplantation (CIT) Consortium’s recent clinical trial efforts and early data from CIT-07, a phase 3 trial examining islet transplantation alone. Dr. Ricordi reminded the audience that the CIT aims to conduct clinical trials that will help bring the first licensed biologically active cell product to market. It has two phase 3 trials in progress – CIT-07 (examining islet transplant alone) and CIT-06 (examining islet after kidney transplant) – as well as five additional phase 2 trials. At this time, the consortium is able to release about 50% of the islets it receives for transplant. It has recently made advances in islet product manufacturing and testing, key enzyme availability, and access to suitable donor pancreases. Subjects enrolled in the CIT trials have experienced substantially reduced insulin use and glycemic lability. After this high-level overview, Dr. Ricordi focused on CIT-07 for the rest of his presentation. CIT-07 is a prospective, single-arm multicenter phase 3 study in which subjects (ages 18-65 with no stimulated C-peptide secretion, marked glycemic liability, and a history of ≥1 severe hypoglycemia episodes) receive up to three separate islet infusions. Subjects are given a regimen of immunosuppressive medicines including antithymocyte globulin, sirolimus, and low-dose tacrolimus, supplemented by etanercept. They are followed for 24 months after their last transplant. Some of the trial’s outcomes (measured at varying timepoints; see for specifics and full outcomes) include proportion of subjects with an A1c of < 7.0% and free of severe hypoglycemic events, proportion of insulin independent subjects (using a more stringent insulin independence definition than other trials), glucose and C-peptide levels as derived from an MMTT, and glycemic liability. Enrollment for the trial was completed in 2011, and all subjects in the trial will reach the study’s primary endpoint by late 2012. Follow-up data is expected by late 2013, after which a BLA will be submitted. Preliminary data for the trial indicates that the procedure has had a very favorable safety profile, with no death or protocol related malignancies reported and very few serious adverse events observed.


Rainer Gruessner, MD (University of Arizona, Tucson, AZ)

Dr. Gruessner provided an overview of pancreas transplant alone (PTA) and progress made in its performance. The goals of PTA, he said, are to avoid severe hypoglycemia and complications. The procedure is indicated for people with brittle type 1 diabetes, nonuremic diabetic patients with severe autonomic neuropathy or microalbuminuria, and those who want to avoid onset or progression of secondary diabetic complications. Most patients who receive PTA have had diabetes for 20+ years and more females than males receive it. The procedure significantly reduces the need for subsequent kidney transplant, and the wait-time for it is markedly lower than for simultaneous kidney-pancreas transplant. According to Dr. Gruessner, PTA can be lifesaving, particularly for patients who are uremic and have to wait a long time for a pancreas-kidney transplant. It also has positive effects on diabetic complications, including nephropathy, retinopathy, and cardiovascular complications. Moreover, at this time, only whole organ pancreas transplantation like PTA can achieve consistent, long-term insulin independence (this is the eventual goal with islet transplantation, but we are not there yet). Nevertheless, there are still very few centers that specialize in PTA, partially because recent improvements in the procedure have been overshadowed by previous high morbidity rates. Per Dr. Gruessner, PTA has become very safe as failures from technical complications have decreased over time. One-year graft survival is above 95%, while 5-year immunological graft loss is much lower than in the past and 5-year unadjusted patient survival is above 90%.

Joint ADA/JDRF Symposium – Antigen-Specific Tolerance in Type 1 Diabetes


Stephen Miller, PhD (Northwestern University Medical School, Chicago, IL)

Dr. Stephen Miller discussed induction of tolerance using IV-injected antigen-modified syngeneic splenocytes and nanoparticles. Interestingly, he noted that if one couples splenocytes with a variety of diabetic antigens (via ECDI covalent linking), only insulinB 9-23specifically protects young NOD mice from development of clinical diabetes, consistent with the hypothesis that insulinB 9-23is the initiating or very early diabetogenic autoepitope in the NOD diabetes model. However, once mice have newly onset diabetes, only induction of tolerance to intact insulin is effective at preventing the transition to overt diabetes, suggesting that the progression from the pre-disease to preclinical disease period is marked by epitope spreading. IV injection of ECDI-fixed allogeneic splenocytes additionally helps induce tolerance to allogenic antigens during islet transplantation. According to Dr. Miller, ECDI-fixed splenocytes help induce tolerance via recapitulation of how tolerance is maintained in the hematopoetic compartment. A phase 1 trial in MS patients has suggested that this strategy may be able to induce antigen-specific regulatory T cells in humans. Antigen-coupled nanoparticles may also be efficient at inducing tolerance, without use of cellular carriers. This strategy is currently being tested in a type 1 diabetes model.

Questions and Answers

Q: Now that you have MS clinical trial data, where do we stand with respect to a type 1 clinical trial?

A: ITN [the Immune Tolerance Network] was at some time trying to get stuff going with regards to this, but I haven’t been associated with the ITN for the last three years. I am waiting to see what happened in the MS trial, as it may give us some indication of the efficacy of this therapy. But I think the future is to get away from cell therapy, because it has a lot of complexity and cost. Apparently, we can use FDA approved PLG beads studded with antigen. We might want to wait and do additional toxicity studies with PLG particles delivered intravenously.


Olli G. Simell, MD, PhD (University of Turku, Turku, Finland)

After briefly reviewing the results for the DIPP nasal insulin prevention trial, Dr. Simell discussed the study design for the INIT II trial, which is similarly exploring whether nasal insulin administration can prevent or delay type 1 diabetes development in at-risk individuals. Dr. Simell expressed excitement for INIT II, highlighting that the study will addresses a number of limitations associated with past nasal insulin prevention studies, including short study durations, small administered insulin doses, and small study populations. According to ( identifier: NCT00336674), INTI II is expected to complete in 2016.

  • In the DIPP nasal insulin prevention trial, the administration of nasal insulin was not demonstrated to prevent or delay the onset of type 1 diabetes in at-risk children. Conducted at three hospitals in Finland, the study randomized 224 infants with HLA-conferred susceptibility and positive expression of two or more type 1 diabetes associated autoantibodies to receive either nasal insulin (1.0 unit/kg; n=115) or placebo (n=109). The mediation duration of intervention was 1.8 years. Type 1 diabetes was diagnosed in 49 children randomized to receive insulin and 47 children randomized to receive placebo. Among the children who continued treatment until diagnosis or study end, 42 children developed type 1 diabetes in the insulin arm and 38 developed type 1 diabetes in the placebo arm. The study was terminated early due to a lack of observed benefit of nasal administration on type 1 diabetes prevention (Nanto-Salonen et al., Lancet 2008).
  • The INIT II trial is a randomized, double-blind, placebo-controlled trial investigating the ability of nasal insulin administration to prevent or delay type 1 diabetes development in at-risk children and young adults (n=120). The study is being conducted at 15 sites in Australia, New Zealand, and Germany. Study participants will be randomized to receive 440 units of nasal insulin or placebo administered daily for the first seven days of the trial, then one day each week for the next 12 months. Participants will be followed for five years or until type 1 diabetes diagnosis. The primary endpoint of the study is type 1 diabetes diagnosis. Secondary endpoints include changes in beta cell, insulin resistance, islet autoantibody levels, and T cell responses to islet autoantigens. Inclusion criteria for the study include having a first degree or second degree relative who was diagnosed with type 1 diabetes before 40 years of age, the presence of autoantibodies to two or more islet antigens, and a normal OGTT and IVGTT. The trial is expected to complete in December 2016 ( identifier: NCT00336674).


Symposium: An Islet Investigation to Advance Research in Diabetes Mellitus


Mark Atkinson, PhD (University of Florida, Gainesville, FL)

Dr. Atkinson reported on the progress made by the Network for Pancreatic Organ Donors with Diabetes (nPOD) in the past year and detailed why the organization’s activities were important to furthering diabetes research. Currently, nPOD supports 79 research projects globally, up from 63 in 2011. The organization recovers approximately 60 pancreases per year from individuals with diverse age, gender, and diabetes backgrounds. The nPOD database makes available patient records and thousands of high quality images of tissue samples that allow visiting clinicians and scientists to make interpretations and develop research questions about the pathology of type 1 diabetes. Furthermore, nPOD has expanded it’s reach in the past year to include the nPod – T (transplant) program, and in the coming year it hopes to launch nPOD –V (viral workgroup). Dr. Atkinson was also very excited to announce their plans to establish a novel reporting mechanism for sharing findings called LabKey. He emphasized that what we can learn from collaborations and studies facilitated by nPOD are crucial to developing a better understanding of the pathogenesis of type 1 diabetes.

  • Designed to “encourage openness” and collaboration, LabKey offers a “facebook- like” portal to discuss findings. Investigators involved in nPOD supported studies will be encouraged to deposit findings in this online portal. Subsequently, they will be able to “friend” others, review their data, and set up collaborations.
  • How do you obtain tissue samples from nPOD? Throughout the presentation, Dr. Atkinson stressed that the services offered by nPOD are free, thanks to the generosity of JDRF. He encouraged researchers to visit to view in more detail what nPOD does, learn about investigators and projects, or even apply to obtain pancreatic tissues. He expressed that the more people who become involved in nPOD, whether through simply joining the discussion or conducting research, the more the program could achieve.

Questions and Answers

Q: How does nPOD receive the pancreases? Is that directly from organ transplant?

A: Recoveries are performed by transplant surgeons, so they follow protocols, but again the pancreas is procured as if it were being used in transplantation.

Q: Is autolysis a concern?

A: It was a concern at the start. It was actually one of the reasons that nPOD wasn’t created sooner. But autolysis was one of the myths we overcame, and we found that if you follow protocols, the pancreas is quite sustainable. We have a strict cut-off window for how long after surgery we can receive them (18 hours), and within the first 100 we received, we only had two cases of autolysis.

Q: Have you had any success in getting pancreases of other forms of diabetes?

A: Great question. We have pathologists and endocrinologists who look extensively at the pancreas and try to determine what odd pathologies of type 1 diabetes mean. Right now we’re not entirely sure, so that is why we group them into “other” category.

Q: In your extensions of nPOD, I did not see nPOD for complications? Is that still a desire?

A: We still have a dream for nPOD- complications. We believe our protocol, our system is ideal for studying this. It’s just that we’re in a position where we need to convince our funding agencies to support us with that vision.

-- by Cati Crawford, Ben Kozak, Kira Maker, Lisa Rotenstein, and Kelly Close