Practical Ways to Achieve Targets in Diabetes Care: Barbara Davis Center Keystone Conference 2011

July 15th – July 17th, 2011; Keystone, CO Full Commentary

Executive Highlights

As always, we were privileged to be part of an audience of so many distinguished physicians, nurses, and diabetes educators at the 2011 Barbara Davis Center Keystone Conference. Finding and communicating ways to improve care for patients with diabetes was the unequivocal theme of this conference, and we are extremely grateful to Dr. Satish Garg (Barbara Davis Center, Denver, CO) for organizing it yet again.

One major theme at Keystone this year was individualizing therapy and selecting patients that can benefit from therapies either new to the market or in late-stage development, such as basal insulin degludec, ultra-rapid acting insulins, novel incretin therapies, PTP-1b receptor antagonists, glucokinase activators, and SGLT-2 inhibitors. For example, Dr. Robert Henry’s (UCSD, San Diego, CA) comprehensive discussion of therapies on the horizon included a review of SGLT-2 inhibitors; notably, when asked about the increased number of cases of bladder and breast cancer in dapagliflozin-treated patients, Dr. Henry described two aspects of the data on breast cancer that made him doubt the risk was attributable to the drug: 1) the cases of breast cancer varied widely in pathologic type; and 2) many of these patients had pre-existing nodules before the trial began. Dr. Henry was slightly more concerned about the risk of bladder cancer given that nearly all cases were of a single pathologic type. While the short duration of the trials introduced uncertainty, he stressed the importance of post-approval evaluations of the association between dapagliflozin and cancer risk.

Surrounding medical devices, we heard both domestic and international perspectives on CGM, pumps, and artificial pancreas developments. Dr. Bruce Bode (Atlanta Diabetes Associates, Atlanta GA) discussed the current role of pump therapy in patients with type 2 diabetes and called for future randomized controlled trials and cost-effectiveness studies in order to better characterize the benefits to patients and society. During one panel session, Dr. Eric Renard (Montpellier University Hospital, Montpellier, France) argued emphatically for the practical benefits of intra-peritoneal insulin delivery (which is still practiced in Europe), asserting that it was a “pity that such technology is not available in the United States.” Dr. Aaron Kowalski (JDRF, New York NY) discussed the challenge of combining diabetes knowledge with mathematical expertise in order to create closed-loop systems, but was quite optimistic and emphasized the efforts of the JDRF-Helmsley Sensor Development Initiative in developing closed-loop technologies. He expressed further frustration with the FDA, saying that regulators are “examining safety exclusively while not fairly examining the therapeutic challenges of controlling diabetes today.” Dr. Roman Hovorka (University of Cambridge, UK) further elaborated on the closed-loop experience, reminding us that physiologic delays in insulin absorption, as well as sensor accuracy, are the most important challenges in developing reliable closed-loop systems.

We were particularly intrigued by the last group of sessions, which were devoted to the topic of prevention and cures for type 1 diabetes. Several topics were addressed, including immunologic manipulation of T regulator cell populations, achieving tolerance to important type 1 autoantibodies, islet cell transplantation, cellular therapy, and beta cell preservation at early onset. Each speaker was careful to avoid spurring false hope, while at the same time providing a clear idea as to productive next steps for investigating potential cure therapies. Dr. George Eisenbarth (Barbara Davis Center, Denver, CO) asserted that we do not have any established way of resurrecting residual beta cell mass, and in the future, we will likely only be able to preserve rather than resurrect beta cells. The progress in islet transplantation was quite intriguing as well, as five-year insulin independence rates have now reached the levels achieved with pancreatic transplantation with new immune tolerance techniques such as T lymphocyte depletion. However, Dr. Ron Gill (University of Colorado, Denver, CO) and others left us with a healthy degree of skepticism, since specific immune tolerance has never been achieved in any autoimmune disease, including type 1 diabetes. By the end of the sessions, we were convinced that the search for a cure is certainly not over, despite the recent setbacks in late-stage therapies for type 1 diabetes.

We were also very pleased to finally meet Dr. Linong Ji of Peking University, the current president of the Chinese Diabetes Society (CDS). He provided us with an important perspective of the disease burden in China as a whole, as well as the experience that Chinese patients have with SMBG. Given that the disease burden is growing quickly alongside rapid rates of economic growth and urbanization, this is will undoubtedly be an important market for diabetes in the years to come.

Table of Contents 

Upcoming News in Diabetes


Matthew Riddle, MD (Oregon Health Sciences University, Portland, OR)

Dr. Matthew Riddle provided us with a useful reminder of the glycemic efficacy of degludec. In the major clinical reports comparing degludec with existing basal insulins, degludec has shown equivalence in glucose reduction endpoints, with some indications of superiority in incidence of hypoglycemia. Zinman et al. divided patients into three arms: degludec once daily, degludec three times weekly, and insulin glargine once daily. A1c reduction, fasting plasma glucose, and SMBG profiles were the same for all three arms. Heise et al. divided a similar population of patients into three arms: a 70/30 mix of degludec plus aspart, a 55/45 mix of degludec plus aspart, and an insulin glargine-only group. In this study, there was a blunting of post-prandial spikes in both arms using a degludec plus aspart co- formulation, but this did not affect overall A1c rates or hypoglycemia. In a study of type 1 patients, Birkeland et al. showed similar reductions in A1c between glargine and degludec, and statistically significant reductions in hypoglycemia, with the greatest benefits demonstrated with nocturnal hypoglycemia. Dr. Riddle ended by describing the major barriers in the use of new basal insulins: the titration of dosage, adherence to prescribed regimens, and the ability to control post-prandial hyperglycemia. Overall, he left us with the impression that new basal insulins like degludec will confer important advantages to particular types of patients. The greatest benefit will be seen by those with type 1 diabetes, long-duration type 2 diabetes, low insulin doses, and consistent activity schedules. According to Dr. Riddle, those with highly variable eating and activity levels will likely struggle on long- acting analogs.


Bruce Bode, MD FACE (Atlanta Diabetes Associates, Atlanta, GA)

Dr. Bode approached the title question of his talk in a clear and systematic way. To begin, he noted there are several theoretical benefits to ultra-rapid acting insulins: 1) less post-prandial glycemic excursion; 2) less hypoglycemia; and 3) less weight gain. All of these benefits relate to the decreased time the insulin spends circulating systemically. The main mechanisms currently being studied to further increase the speed of insulin include faster ways of delivering existing insulins (intradermal infusion, warming the infusion site, and co-formulation with hyaluronidase), insulin analogues with ultra-rapid pharmacokinetics (Novo Nordisk), and alternate delivery routes (inhaled and intra-peritoneal delivery). Dr. Bode maintained that we have come a long way with rapid-acting insulins, but a few years remain before any of the more rapid-acting therapies in development will actually be brought to market. Dr. Bode noted that while each method is promising, they all have safety limitations. On efficacy, he mentioned that none have demonstrated enhanced blood glucose-lowering efficacy (A1c or SMBG) for open-loop therapy except for Mannkind’s Afrezza, which received a significant setback with the FDA’s CRL. Dr. Bode reminded that audience that at this time, none of these novel methods have been tested in a closed-loop system. However, he summarily believed that ultra-rapid acting insulins will be an important therapy for certain patients in the future.


Linong Ji, MD (President, Chinese Diabetes Society, People’s Hospital, Beijing, China)

We were pleased and honored to hear from Dr. Ji at this meeting, who gave us a nice global perspective on diabetes. He informed us that the prevalence of diabetes has grown faster in China and India than in the United States, most likely due to increased rates of economic growth in these emerging economies. He also introduced us to specific types of Chinese herbal medicine that are still used to treat diabetes in China, in addition to Western therapies. Although a Cochrane Systematic Review concluded that Chinese herbal remedies should not be used to treat diabetes, there is compelling evidence that the Xiaoke herb can be effective in treating patients with type 2 diabetes (please see page 298 of our ADA 2011 Full Report at for our coverage of Dr. Ji’s lecture). Specifically, a new randomized controlled trial demonstrated that Xiaoke herbal therapy is non-inferior to glyburide for glycemic reduction, and has a lower incidence of hypoglycemia. We look forward to learning more about diabetes in China in the years ahead, and especially greater clarity on the ability of herbal therapies to treat type 2. In this increasingly constrained economic environment, such therapies, if proven effective, could provide a way to cut costs though at this stage we would be very cautious about acceptance..

Questions and Answers

Q: Could you discuss glargine and its safety during pregnancy?

Dr. Riddle: The bottom line is that there is no difference between NPH and glargine in study outcomes for either mother or baby. It should be noted that the statistical power for these studies is not strong, but at least there was no safety signal.

Comment: My understanding was that glargine was still off-label for treatment in pregnancy because of the higher documented incidence of cancer.

Dr. Riddle: It is not “off-label” for that reason. Currently, it is not recommended for pregnancy because of the lack of data supporting its efficacy and safety in pregnant women specifically. We still do not know the risk/benefit ratio of attempts at strict glucose control in pregnant women, for both mother and baby. The cancer story is another one altogether, and very complicated, as you know there was concern raised by retrospective analyses published in a European journal a couple of years ago. Answering this question is a very thorny scientific problem, and my opinion is that the jury remains out on this.

Dr. Eric Renard, MD, PhD (University of Montpellier, Montpellier, France): I want to give an update on intra-peritoneal insulin delivery in Europe, as this was mentioned in the ultra rapid-acting insulin talk. Around 500 people have been treated by intra-peritoneal delivery using implanted pumps made in the US and extensive research has been shown for a reduced risk of hypoglycemia and sustained glucose control. There are patients who have been on intra-peritoneal insulin for more than 20 years now, and they have A1cs of around 6%. It is a very efficient way of treating diabetes. In addition, it is actually the case that closed-loop studies have been done for intra-peritoneal delivery technology way before other closed-loop systems were tested. Roche is actually developing a new Diaport device. It is a pity that such a technology is not available in the Untied States.

Dr. Satish Garg, MD (University of Colorado Denver, Boulder, CO): A question for Dr. Kendall. Is there a chance that one could use both A1c and fasting glucose?

Dr. Kendall: The current recommendation is if you get an A1c, repeat it. However, the problem is what do you do if you have a FPG of 125 mg/dl and the other test is normal? Do you take your best result or worst result? Well, the clinical intervention is still the same. While on a lab basis you still have some purists, don’t keep looking for your best test result; just use your clinical judgment. Try to repeat that test, and when these data get published, we will have more evidence.

Q: In the vein of lowering costs for insurance companies concerning the number of strips used, how often do you recommend that type 2 patients check their blood sugars?

Dr. Kendall: Fortunately, I had to prepare a talk about this. For SMBG, if you tie it to no therapeutic change then there is no effect. But if you use it to guide therapeutic changes, then there is another 0.3% to 0.4% change in A1c with no added weight gain or risk of hypoglycemia. If you test and don’t use the data, it’s a waste of money.

Dr. Garg: There was a lot of ADA data on degludec in the form of abstracts and posters. Could you provide some highlights of this data? Are there differences between degludec and glargine?

Dr. Riddle: There were a lot of different abstracts. So many, actually, that people are pooling information and doing their own internal meta-analyses. What I took away was that degludec works great, and once- daily injection is easy and does not have any side effects unique to it. Hypoglycemia seems to be slightly less with degludec than with glargine. Twenty-four hour hypoglycemia, however, is not clearly very different. The main advantage seems to be at night. To me, it’s a little less of an impressive difference than that between detemir and glargine. It is not clear whether this will play out in differences in glucose control. We need a pooled analysis of hypoglycemia rates.

Dr. Bode: There were 14 presentations at ADA. I agree with Matt. There’s always a trend toward lower FPG, and this was only significant in five of the studies. There is not a significant trend toward lowering A1c, but there is a trend toward achieving lower nocturnal hypoglycemia (25-40% reduction). Overall hypoglycemia is not different. This is a very predictable insulin, and there is testimony that it’s smoother for patients. Interestingly, however, you do not see less dropout rates in the degludec arms for these studies. It does appear very safe. The drug has zero-order kinetics; you could give it eight hours apart and 40 hours apart and still have less hypoglycemia.

Dr. Riddle: Whenever we get a new product, it does take a little time in order for us to figure out how to use it. The patients who are most likely to benefit are type 1 patients where a pump is not being used. If you do have better overnight average basal, you just need to do prandial dosing a little differently. Daytime hypoglycemia is more due to not making appropriate mealtime insulin decisions rather than basal rate irregularities.

Prevention of Cardiovascular Disease and Diabetes


David M. Kendall, MD (former Chief Scientific and Medical Officer, American Diabetes Association)

According to Dr. Kendall, while the psychological tenets of primacy and recency would lead us to believe that aggressive therapy does not reduce the risk of micro- or macrovascular complications, analysis of intermediate data provides strong evidence that both can be improved with aggressive glycemic control. He reminded the audience that ACCORD, ADVANCE, and the VADT trials all provided controversial evidence that aggressive control does not reduce cardiovascular mortality (and in the case of ACCORD, an increase in all-cause mortality) in some patients. However, he feels that reasons for these outcomes are mistakenly attributed to age and a “lower A1c,” missing important nuances in the data. Meta- analyses suggest that, in the entire population of patients with type 2 diabetes, there is a modest decrease in cardiovascular events and no increase in all-cause mortality with intensive control. Despite these relatively encouraging outcomes with meta-analyses, however, care should be individualized to patients, using the data as an important guide.


Marian Rewers, MD, PhD (Professor and Clinical Director, Barbara Davis Center, Denver, CO)

Dr. Rewers began his presentation with a sobering statement: while the overall prognosis for type 1 diabetes continues to get better over time, cardiovascular outcomes have not improved as quickly as the progress made with microvascular complications. Few type 1 diabetes patients meet even the minimal set of clinical recommendations and remain at high risk for CVD. Many non-modifiable risk factors for CVD in patients with type 1 diabetes are well characterized, and new markers are still being characterized. According to Dr. Rewers, our understanding of risk factors for type 1 patients is still evolving.


Silvio Inzucchi, MD (Yale School of Medicine, New Haven, CT)

Dr. Inzucchi provided some interesting data surrounding the prospects for coronary artery disease (CAD) screening in diabetes, concluding that while there may be some benefits to screening certain patients, routine screen should not be recommended at this time. Dr. Inzucchi reviewed the topline results of the Detection of Ischemia in Asymptomatic Diabetics (DIAD) studies. DIAD-1 discovered that abnormal myocardial perfusion was present in about 22% of asymptomatic patients (silent myocardial ischemia), with markedly abnormal perfusion detected in 6% of patients. Turning to DIAD-2, he knocked down the long-standing belief that myocardial ischemia increases over time in patients with diabetes. In that study, the data demonstrated that 79% of all initially detected perfusion defects had resolved to normal after three years. Finally, in the DIAD-3 analysis, patients with type 2 diabetes, randomized to coronary artery disease screening by nuclear imaging did no better than those patients who were followed with usual care. Dr. Rewers closed by emphasizing some of the unknown factors relating to this topic, questioning whether certain patients (young type 1s, high risk patients identified by novel markers) or certain procedures (coronary CT) would yield different results.

Questions and Answers

Q: Is there any data from a policy standpoint on when you start a statin? For example, at what age?

Dr. Inzucchi: Medical therapy statins are now two cents a day. For diabetics, there is no real consensus on statins, and remember that whenever the ADA has a consensus conference, there is usually no real consensus. [Laughter]

Dr. Rewers: Regarding the fact that all of those patients regressed, I would just come to the conclusion that it’s not a good test and that there were a lot of false positives initially. What do you think of that possibility?

Dr. Inzucchi: My nuclear cardiology colleagues would disagree. This all may represent some microvascular disturbance. It is a good test. Regular use of aspirin, statin therapy, and a low threshold for ACE or ARB-based therapy are pretty good pieces of suggestive evidence.

Dr. Rewers: Still, there is little evidence that atherosclerotic disease can spontaneously resolve like that, so I’m a big skeptic.

Q: In terms of therapies and screening for patients with diabetes, do you have any suggestions for hospital systems that are looking at dashboards so I can deal with primary care doctor frustrations?

Dr. Kendall: A lesson can be derived from the APGAR score. The APGAR score has horrible science behind it, but what they did is not give all-or-none credit. You sort of get scored on a relative scale, and this is where NCQA is starting to get it right. There has to be the ability to do what he said from Silvia’s paper. There is no way the system, despite its desire to be simplistic, can do that. The ADA is all of us and we put this together as an optimized goal for the general population. This can’t simply be applied. You sit down with them and say, “How can we give partial credit or a graded scoring system?” For babies, if you were pink but not breathing, that’s a bummer. Doing all things reasonably well is a good goal, and you get there by doing a collective score rather than individual cut points.

Q: Any there any hospital systems doing it that way?

Dr. Kendall: NCQA is getting there. The data support that being on statin therapy is generally a good thing. Zero, one, two scoring for A1c is sensible. Make it so that you can still get 2 points if you’re under 8% and you’ve listed the reasons. The problem that occurred when the UK implemented this system was that all of a sudden there were good reasons why huge numbers of patients didn’t get below target.

Dr. Rewers: When you looked at the DCCT data comparing A1c to complication risk, you would see that it’s not linear; it is an exponential curve. Therefore, theoretically, it is wrong to assign the same point going from 8% to 7% vs. 10% to 9%. You would have saved a lot more if you just did the former.

Dr. Kendall: Yes, this is why a 100-point scale makes more sense. We’ve worked endless hours on how to get this right. Measurement folks have taken this way too far. Just because they haven’t figured out how to measure it doesn’t mean you should compromise care.

Q: How are you addressing vitamin D screening and treatment for patients?

Dr. Rewers: This is a bit of a trap. The evidence is not completely there to justify treatment at insufficient levels or even sufficient levels. We are getting much less sun than we used to, mainly because of conditioning and long hours of work. This may be a big problem not only for diabetes but for other things. A lot of evidence is suggesting that it has changed.

Dr. Kendall: I have no idea what to do. Every time I go to a meeting I go home and take more vitamin D. [Laughter]

New Treatment Options for Type 2 Diabetes


Matthew Riddle, MD (Oregon Health Sciences University, Portland, OR)

Dr. Riddle provided us with a useful lesson in the physiology of incretin therapy, as well as the nuanced differences between each specific therapy. His goal was to provide some guidance, based on the limited evidence available, as to which patients would benefit most from each type of therapy. He emphasized that reasonable conclusions often cannot be made, since we are “babies” in our understanding of this important pathway. For the clinical audience, he highlighted important differences in the clinical effects between GLP-1 agonists and DPP-4 inhibitors. Importantly, he also covered some of the evidence behind combination therapy using incretins and other oral medications, which are promising. Finally, he outlined much of the evidence behind the adverse effects of GLP-1 agonists, including pancreatitis, and reminded us that the high cost of these medications is a major barrier to patient use.

  • While GLP-1 agonists and DPP-4 inhibitors affect the same biological pathway, there are important clinical differences between them. In summary, GLP-1 agonists potentiate insulin secretion, suppress glucagon, and reduce calorie intake to a greater degree than DPP-4 inhibitors, with each of these effects maintained over time. DPP-4 inhibitors, however, are actually thought to mimic the physiologic balance of endogenous GLP-1 more closely than GLP-1 agonists. Notably, while it is well known that GLP-1 agents slow gastric emptying, this effect is not maintained over time, likely due to GLP-1 receptor sensitization (Nauck et al.).
  • Studies have shown that long-acting GLP-1 agonists have a more pronounced A1c- lowering effect and slightly greater weight reduction than short-acting analogs; however, short acting drugs more readily control post-prandial glucose excursions. This is thought to be due, at least in part, to the blunted effect on gastric emptying over time with long-acting agonists. This effect is seen even within one day in the same patient; Christiansen et al. showed a very nice suppression of postprandial glucose excursion with long-acting agonists, but by dinner, this effect was absent.
  • Studies show promising combinations of GLP-1 analogs and existing oral therapies, indicating that the best of each drug can be achieved simultaneously. Buse et al. compared glargine alone with exenatide BID + glargine, and found favorable A1c reduction to 6.7%, nice weight control, little hypoglycemia, and a favorable reduction in post-prandial blood glucose levels. Similar results have been demonstrated for combination therapy with exenatide plus metformin plus glargine. Dr. Riddle was careful to emphasize that none of these combinations are FDA-approved.
  • Safety and cost considerations are the main barriers to ubiquitous use of GLP-1 drugs. Studies have shown an increased risk of pancreatitis with exenatide or sitagliptin and even an increase in pancreatic and thyroid cancers. These studies are quite controversial, however, since they are based on reports sent by physicians who were probably aware of the public speculation surrounding the cancer risks of incretin drugs. Finally, a 90-day supply of sitagliptin and saxagliptin can be purchased for $620 while Byetta and Victoza are over $800 each. This compare very unfavorably to glimepiride, which is just $42 for a 90-day supply. Patient access to such drugs due to their price, despite their efficacy, is a barrier that cannot be ignored.


Robert Henry, MD (President, Medicine and Science, American Diabetes Association)

Dr. Henry outlined many of the studies investigating potential novel therapies for the treatment of diabetes. His discussion of SGLT-2 inhibitors was particularly relevant, given that the FDA is currently evaluating dapagliflozin. While Dr. Henry did describe the risk of genitourinary and urinary tract infections with SGLT-2 inhibitors, he did not discuss the current controversy over whether dapagliflozin increases the risk for breast and bladder cancers. We note that this issue was discussed extensively in the expert panel and was the primary reason for No-votes among panel members (for more information on the dapagliflozin advisory committee panel, see our full commentary at

  • PTP 1B is a novel therapeutic target important for inhibiting the insulin-signaling pathway; such inhibition potentiates insulin secretion. The compound was discovered about a decade ago and is still being studied in early trials. It is ubiquitously expressed and particularly replete in skeletal muscle. It has been discovered that LDL is also reduced with inhibition. The only phase 2 results of the novel agent involve 26 patients who received 200 mg per week for three months. Compared to placebo, FPG was reduced by 22 mg/dl, while LDL dropped by 11 mg/dl. Frucotasamine also decreased by 24.5 mol. A1c was not used the study duration was three months. Further studies are needed to confirm the therapeutic effect and safety.
  • While glucokinase inhibitors have generated much excitement over the last ten years, these drugs have been withdrawn from human trials, most likely due to severe hypoglycemia and tachyphylaxis over time. The molecule is found in the pancreas and liver and is important for glucose-stimulated insulin release. No test drugs have passed phase 2 trials as of yet. Two are actively recruiting for phase 2 trials, but there is no human data available on compounds in active development.
  • SGLT-2 inhibitors are the furthest along the development process, with dapagliflozin currently being evaluated by the FDA and canagliflozin in phase 2 trials. The mechanism of the drug class is to inhibit glucose reabsorption in the proximal tubule of the kidney. The relative efficacy of these drugs is impressive from both a glucose-lowering and cardiovascular risk standpoint: phase 3 studies for dapagliflozin show a significant reduction in A1c, a decline in systolic blood pressure (4-5 mm Hg), and weight loss (2-4 kg) over 24 weeks of study. However, safety issues are of paramount concern with dapagliflozin. There is a statistically significant increase in GU infections (mainly fungal infections). Dr. Henry did not discuss the apparent increase in the number of dapagliflozin-treated patients with breast and bladder cancers. Interestingly, canagliflozin’s phase 2 data shows the drug does not appear to have the urinary tract infection issues associated with dapagliflozin.


Bruce Bode, MD (MD FACE (Atlanta Diabetes Associates, Atlanta, GA)

Dr. Bode summarized the current role of pump therapy in patients with type 2 diabetes and hypothesized where it is going. According to Dr. Bode, there are currently 25,000 to 50,000 patients with type 2 diabetes on pump therapy. Notably, while the penetration of pump therapy in patients with type 1 diabetes is 33% according to Dr. Bode, penetration in type 2 is only about 1% (according to 2008 data from Medtronic; we suspect penetration has increased since this data was released). Dr. Bode argued in strong favor of pump therapy for patients with type 2 diabetes. First, he asserted that because the natural history of type 2 involves reduced beta cell function over time, pumps do a better job of mimicking normal physiology than MDI. Additionally, considering 89% of type 2 patients on insulin do not inject outside the home, insulin pumps may be quite helpful in these individuals. One 16-week, open label, uncontrolled pilot study (Edelman et al., DT&T 2010: 12(8): 628-33) using an Animas insulin pump plus metformin in 56 patients showed a 1.2% reduction in A1c from a baseline of 8.5%. Although there was no placebo arm, a larger drop in A1c was correlated with a higher baseline. Other studies of pumps in type 2 diabetes showed that fasting plasma glucose trended lower, post-breakfast glucose was lower, and that close to 90% of patients preferred pump therapy over their prior regimen. Dr. Bode concluded by stating that, while the current evidence supports improved glucose levels and an increased quality of life for patients with type 2 diabetes on pump therapy, future randomized-control trials are needed to fully characterize these benefits. In addition, cost-effectiveness studies are needed to support reimbursement of pumps for type 2 diabetes.


Robert H. Eckel, MD (University of Colorado Anschutz, Aurora, CO)

Dr. Eckel described the high incidence of dyslipidemia in patients with type 2 diabetes, simultaneously noting that patients with type 1 diabetes often have more favorable-appearing lipid profiles. He presented much of the evidence behind the benefits of controlling lipids aggressively in patients with both type 1 and type 2 diabetes. He concluded with his own opinions about adequate lipid control: every patient with type 2 diabetes should be on a statin, and those with fasting TG levels >200 mg/dl should be on fibrate therapy. For patients with type 1 diabetes, statins should be used uniformly to lower LDL to <100 mg/dl once adult age has been reached. Otherwise, ATP-III guidelines should be followed.

Questions and Answers

Q: My name is Dr. Rickstaw and I’m a primary care provider. Who might help me understand the chemistry of my diabetic patients as far as potassium goes? I noticed on your slides that the SGLT transporter in the proximal tubule – is that the area where we’re having trouble with patients losing potassium in Type 2 diabetes?

Dr. Henry: They have done studies looking at electrolyte changes with dapagliflozin, and have only shown increases in urine sodium. There is no real observed increased in urine potassium, and no translation of anything into changes of serum electrolytes. Uric acid goes down with dapagliflozin, but I have not seen the data yet for canagliflozin.

Q: I’m just asking generally: why do patients with type 2 diabetes have trouble with their losing potassium in the urine?

Dr. Henry: This is mainly due to type IV renal tubular acidosis caused by increased urine flow volume in the distal tubule. It is mainly the acidosis.

Q: Since the GLP-1 analogs achieve post-prandial blood glucose reduction, have they been studied in patients with impaired glucose tolerance?

Dr. Riddle: There have been studies done in non-diabetics and they do put out nearly 30-50 g of glucose in the urine. There is a trend for weight loss in 14 days and a potential for using these agents as a preventative therapy in individuals with prediabetes. However, this hasn’t been explored at all. You would think that if there are weight control problems, it would make sense to consider treating such patients for multiple targets. This hasn’t been done yet and studies continue ahead.

Q: Dr. Garg: There has been some noise around SGLT-2 possibly causing bladder cancer. Can you comment on that?

Dr. Henry: There have been press releases from Bristol-Meyers Squibb. The dapagliflozin arm had nine cases of breast cancer compared to only one case in the comparator group. For bladder cancer, the difference was also nine versus one. When looking at the specific cases of breast cancer, there were four different types of breast cancer, and four were already known to have nodules before starting therapy. With regard to bladder cancer (mainly males), five out of six were of a single type. Even so, it does seem implausible that these cases would occur so quickly. This will be something that will have to be looked at more closely, however. Please bear in mind that we’re talking about ten total patients out of more than 8000 patients in the study cohorts, including comparators. But this is still important to look at.

Dr. Garg: Is SGLT-2 expressed in the bladder?

Dr. Henry: This was looked at. Unfortunately the breast tissue was not examined, and I don’t know whether there is SGLT-2 in the breast. But it was not known up until the time of the press release. I believe there is SGLT-2 in the bladder.

Dr. Garg: Why can’t we go to an LDL level of less than what you stated? Why not down to 70? Is there any reason not to go down that low?

Dr. Eckel: While there is data to support reducing LDL to <100 in patients with type 1 diabetes, I don’t know that there’s enough evidence to have anything other than an opinion on this. Statins should be indicated to get LDL under 100, but that’s pretty cavalier in its own right, and we probably won’t even make a case before that.

Dr. Garg: How young can we go in treating lipids?

Dr. Eckel: I am only treating adults’ lipids to this level.

Dr. Gottlieb, MD (University of Colorado Health Sciences Center, Denver, CO): Is there a lipid level that’s actually too low? Do clinicians need to be aware of a lower limit of lipids?

Dr. Eckel: At this point there is no evidence that too low of an LDL is an issue.

Q: For patients with type 1 diabetes- weight gain is very frustrating. Do you have experience using GLP-1 agents off-label for type 1 diabetics?

Dr. Riddle: The reason why GLP-1 might be beneficial in patients with type 1 is the potential suppression of glucagon post-prandially, as well as a decrease in calorie consumption. Clearly we’ve seen it in STAR-3, if people eat small meals and require less insulin they might do better. Dr. Garg can speak to this, as he’s conducting these investigations.

Dr. Garg: We should have the information next year. This is a large study. We have measured everything, and the study is double-blinded.

Dr. Eckel: I want to make sure that people understand, because I did make an earlier comment about this – an LDL of 30 is not good, but the risk of getting that low is pretty insignificant.

Dr. Bode: There is some concern for suicide in patients with a lower LDL. While this effect is small in magnitude, it is not trivial.

Q: What do you do for statin-intolerant diabetic patients?

Dr. Eckel: Most of these patients have statin-induced myopathy. There is an algorithm of about eight things to do in the setting of statin-induced myopathy. But there are no good clinical trials comparing one versus two versus three.

Q: What would you recommend for patients who have an LDL of about 100 and who have HDL above that?

Dr. Eckel: Great question. We know that HDL is inversely related to coronary artery disease. Use of the ratio clinically, however, is controversial. HDL function is adversely modified in patients with diabetes, and that is a summary statement of a significant amount of data.

Obesity and Type 2 Diabetes

James O. Hill, PhD (University of Colorado Anschutz, Aurora, CO); Judith Regensteiner, PhD, (University of Colorado Denver School of Medicine, Denver, CO); Daniel Bessesen, MD (University of Colorado School of Medicine, Denver, CO); Diane Karl, MD (Oregon Health and Science University, Portland, OR)

A major theme in this series of talks was that obesity has a strong psychological component. This may explain why so many in vivo-success stories for small molecules with physiologic targets turn into failures when tested clinically in human trials. Even existing oral therapies are mostly modest in their magnitude of effect. The benefits of exercise cannot be overstated and are certainly supported in the most recent DPP trials, even independent of effects on weight loss. Bariatric surgery, which could certainly be considered the “gold standard” therapy for sending diabetes into remission, has a morbidity and side effect profile that is significant and certainly cannot be a universal solution. Most importantly, Dr. Bessesen made a strong case that there is a huge demand for pharmacologic therapy to treat obesity among patients, and that the bias against such drugs lies mainly among clinicians and regulators.

Questions and Answers

Q: In our own practice, and in relationship to obesity, we have seen that we are missing a huge psychological component. How food affects us. I’m not sure we really understand that. Patients don’t understand why they overeat, and I suspect that it has little to do with physiology and much more to do with psychology. What can we do to address that?

Dr. Bessesen: The problem is, a lot of us don’t have training for things like depression. There should be screening for this, in addition to diuretic use, and laxative use, and people should be referred as appropriate. We will all have to develop local strategies.

Dr. Hill: I agree, and that doesn’t mean that the biology is not important, but this illustrates how complicated this issue is.

Q: What about the use of ADD/ADHD medication for weight loss? Are there any studies looking at the safety and efficacy of these medications?

Dr. Bessesen: There are lots of drugs that we could have talked about. Orlistat, phentermine, and other drugs we actually have clinical trials on. ADHD drugs, while potentially causing weight loss effects, don’t have clinical trials or FDA approval for this indication. We do not have data, and do not prescribe for that purpose.

Q: I had a 95th percentile BMI patient and gave her the spiel about diet and exercise, but she says she eats less than her 9 year old brother, who has a BMI in the 30th percentile for his age. Is this girl really eating less based on what you’ve seen clinically?

Dr. Hill: The answer is overwhelmingly likely to be no, she’s not eating less. But I’ve seen that “proving” to her that she is eating a lot may not get you where you want to go.

Dr. Regensteiner: People do delude themselves into not eating. But if you point that out and beat them up for being delusional about what they’re eating, it doesn’t help. They eat in small, tiny bites, maybe a pound of fudge in one hour, but that still counts. Particularly with kids, you have to work with behaviors and customize your therapy.

Q: I am a nurse practitioner and in my last job I was involved in a clinical study of a thermogenic agent released by Lilly. I left that job right as we were midway through the study. Is there an update on that?

Dr. Bessesen: As far as brown adipose tissue thermogenic agents, there aren’t any near the market that I know of.

Comment: I just want to comment about the ADD medications. Back in the 60s and 70s we used synthroid and amphetamines for weight loss, and a lot of people dropped dead from that combination. Appetite suppression decreases about a month or two of use for ADD meds, and I’ve dealt with this. It is actually rare to find kids who actually lose weight on these medications.

Continuous Glucose Monitoring and the Pump


Aaron Kowalski, PhD (Assistant Vice President, Treatment Therapies, JDRF)

Dr. Kowalski discussed the challenge of combining diabetes knowledge with mathematical expertise to create closed-loop systems. He gave us a useful summary of the existing algorithmic approaches that the JDRF has investigated: control-to-range, nocturnal pump control, overnight closed-loop, and even dual hormone approaches with insulin, glucagon, and amylin. Notably, Dr. Kowalski is more of a believer in the inclusion of amylin than glucagon. He emphasized that it is nonsensical that we have pumps today that continue to infuse basal rate insulin despite the fact that they have CGM information that the patients are hypoglycemic. He is a big supporter of glucose shut-off systems like the Medtronic Veo, and stated that the “gaining regulatory approval will be challenging.” Dr. Kowalski informed the audience of JDRF’s hard regulatory work, including regular meetings with key FDA staff and intriguingly, reviewing and approving all submissions from academic sites before they are sent to the FDA. Dr. Kowalski rejected the FDA’s concerns over DKA risk following insulin pump suspension, calling the risk “miniscule” and backed by clinical data. He stated that the FDA was “a bugaboo of his,” and that we have a very conservative FDA, “examining safety exclusively while not fairly examining the therapeutic challenges of controlling diabetes today.” He concluded by reminding us of the JDRF-Helmsley Sensor Development Initiative, which will no doubt make strides in device development for patients with type 1 diabetes.


Eric Renard, MD, PhD (University of Montpellier, Montpellier, France)

Dr. Renard began by describing what he considers to be the numerous benefits of intra-peritoneal delivery of insulin, a technology that is approved and available in Europe. The system is designed by Medtronic Minimed and combines an IV sensor located close to the central vein and an implantable insulin pump, connected by a subcutaneous wire. Insulin effects were quicker (peak effect after 20 minutes), far better than the subcutaneous route. Generally, the PID algorithm (and virtually every closed-loop algorithm) controls blood glucose well overnight, but does not do as well post-prandially. According to Dr. Renard, this is due to the difficulty of mimicking the physiological occurrences after food ingestion (i.e., the closed loop system does not have a cephalic phase). Overall, the Europe experience has shown that mean BG can be improved by closed loop systems, and time within normal range improves significantly with all closed-loop systems. He closed by expressing much faith in the closed loop system, as it could certainly offer improvements over the open-loop system. Regardless of Dr. Renard’s enthusiasm, we are cautious about potential market acceptance of intra-peritoneal delivery. We feel it is unlikely that patients would be willing to use such an invasive system over standard pump therapy, despite its improved abilities to regulate blood glucose. Even if some patients accepted it conceptually, we doubt payors would be positive on the notion.


Roman Hovorka, MD (University of Cambridge, UK)

Dr. Hovorka began by introducing the spectrum of insulin delivery and glucose sensing technology, from open loop to semi-closed loop to fully closed-loop. He gave a detailed description of the low glucose suspend (LGS) system currently available in Europe, the Medtronic Veo. Dr. Hovorka reviewed some of the most recent clinical evidence, concluding that the LGS is safe and well tolerated and may reduce the risk of prolonged hypoglycemia. He also reminded that audience of the FDA’s recent draft guidelines on LGS systems (see our June 23, 2011 Closer Look at, which will require at least a 10% reduction in either severe hypoglycemia, CGM-based event rate (prevention), or hypo AUC reduction (mitigation), and non-inferiority for A1c (a margin of 0.4%). He also described an overnight closed-loop device that was studied in a randomized crossover trial in children and adults. Subjects were randomized to either standard pump therapy or closed loop for 1-2 weeks, and then were switched over to the other therapy for 1-2 weeks. Glycemic variability and time-in-target-range was significantly better for the overnight-closed loop portion of each group’s therapy, although mean overnight glucose was not different. Unfortunately, the technology could not prevent hypoglycemia when exercise and prandial insulin overdosing were not announced to the algorithm. According to Dr. Hovorka, these highlight important limitations in closed-loop systems, especially in patients who may have unpredictable and inconsistent schedules. Favorable results were also seen in pregnant patients with type 1 diabetes. Overall, the technology is progressing rapidly and favorably, although the delay in insulin absorption, as well as sensor accuracy, will continue to be challenges ahead.


Steven Edelman, MD (University of California, San Diego, CA)

Dr. Edelman noted that CGM has been so successful for patients with type 1 diabetes primarily because every day is different for them. In addition, he emphasized that type 1 patients on MDI stand to benefit just as much as CSII patients. For patients with type 2 diabetes, CGM has also proven itself quite effective, especially as a motivational and educational tool. In the type 2 population, he summarized the latest evidence supporting use of CGM. We previously covered the late-breaking abstract by Vigersky, et al., which showed that episodic use of CGM can improve both short-term and long-term glycemic control (see page 133 of our ADA 2011 Full Report at Because patients were taken off CGM after 56 days, and still largely maintained glycemic control, an important take-away was that CGM is a powerful tool for sustained behavior modification. He also reminded us of the 2009 study by Yoo et al. in Diabetes Research and Clinical Practice, which randomized poorly controlled (A1cs of 8- 10%) patients to CGM and SMBG. The study showed a greater reduction in A1c for the CGM group. Dr. Edelman further reviewed a 2008 Murphy, et al. study appearing in BMJ, which showed a more pronounced reduction in A1c and reduced birth weight for patients in the CGM arm. Finally, Dr. Edelman shared a profound story about one of this patients who underwent the “extreme diabetes makeover,” a diabetes education program that he runs through his non-profit, Taking Control of your Diabetes. The patient was a well-educated, successful lawyer who had type 2 diabetes and an A1c of 11.8%. CGM motivated him to achieve better control, and his A1c dropped to below 7%. We were struck by this inspiring story of the power of CGM to motivate patients and hope that data continues to mount for the benefits of CGM in patients with type 2 diabetes.

Questions and Answers

Q: Do you have any data on CGM use in the inpatient setting?

Dr. Hovorka: Bruce Buckingham is doing a study and we think it is going to happen. We think this could be achieved with significant success.

Q: Is there any thought about changing where we sense from and what we pump into? The sites are less than ideal, and we’ve gotten much better through algorithmic approaches. Are we going to be able to take this to the next level? Are we going to have to be able to get past the intrinsic physiologic challenges that come from using subcutaneous delivery?

Dr. Moshe Phillip, MD (Schneider Children's Medical Center of Israel, Petah Tikva, Israel): Why use just one type of site? Do both subcutaneous and intravenous sensing!

Dr. Hovorka: For administration, we might use oral or inhalational methods, but one might not be able to do that on a continuous basis.

Q: But are there novel ways that we can approach even from a sensing standpoint? Patients always tell me that they wished insulin didn’t take so long to work. I don’t mean to imply that we haven’t gone a long way, but can we go further?

Dr. Phillip: We can do better by adding automation. We are aware of what people are with and we know we can do it with new tools. Probably the most likely route to fix the insulin problem is intra-peritoneal. The devices right now, Roche’s Diaport, is approved in Europe. Is it widely applicable for a bunch of patients? That remains to be seen, but next generations of these devices are much smaller, less obtrusive, and not fully implanted. These things will come. I think we can get very good with interstitial fluid sensing.

Dr. Renard: The question is a bit different for delivery, though. Most would like to have a long-term sensor that is stable, that they don’t have to recalibrate and change so often. This is the main way to improve the system, I think.

Q: Do you think glucose oxidase-based sensing is the best thing, or is there something better in the future?

Dr. Renard: It’s a matter of biocompatibility. We need to improve this, have it remain stable, and with fewer calibrations. The main problem is the interface between the body and the system itself. This creates a screen that presents a difficult engineering problem to solve.

Dr. Kowalski: At the JDRF, we are looking at the redundancy of sensors. Closed-loop, I think, is much better, and the mathematics is much more sophisticated than what people can compute. Glucose-oxidase is not the limiting factor.

Q: In practice, I see that patients with stable nighttime values often have good daytime values as well. I didn’t see that in your data, but I would think that if you looked at the second day of overnight control vs. closed-loop control you might see a consistent effect.

Dr. Hovorka: We did follow the subjects after breakfast and we observed improvements. This might be a benefit, but it depends on the breakfast, etc.

Dr. Phillip: You’re absolutely right, it does happen, but we never collected the data. I do think you’re right.

Outcomes of Other Clinical Trials in Diabetes


Linong Ji, MD, (President, Chinese Diabetes Association, Peking University, Beijing, China)

Dr. Ji thoroughly convinced us that the diabetes epidemic in China is pressing, and growing. He again outlined the staggering demographics of the type 2 diabetes burden in China: over 16 years, the prevalence of diabetes tripled from 1.0% to 3.2% (1980-1996), and reached 5.5% by 2000. It is higher among urban residents (11.4%) relative to rural residents (8.2%), most likely due to richer diets in these regions. Dr. Ji argues that, given this disease burden, the self-monitoring statistics of patients is particularly concerning. Only 30.7% of patients with a known diagnosis had their A1c checked in the past 12 months, and only 41.7% of patients self-monitor their glucose through SMBG or urine checks. In patients with type 1 specifically, the mean frequency of checks was 1.92 1.31 times per week. In addition, the self-reported survey showed that 64.1% of patients had an A1c >8%. He concluded by summarizing the major barriers to SMBG in China: 1) low familiarity with the SMBG concept; 2) deficient patient education and self-management; 3) low commitment from care providers, and 4) lack of reimbursement for glucose test strips by the Chinese government.


Steven Edelman, MD, (University of California, San Diego, CA)

Dr. Edelman introduced his talk by acknowledging that the landscape for type 2 diabetes treatments is becoming more and more complex, but that this complexity allows for novel combinations of therapies that are better than single therapies alone. Sitagliptin plus metformin has shown favorable, placebo- adjusted results in multiple studies for A1c reduction, with no significant difference from single agent therapies in hypoglycemia rates. He did discuss the effect of sitagliptin on patients with type 1 diabetes, highlighting upcoming evidence from studies that Dr. Satish Garg is doing with his team. He also summarized data from Ellis, et al., a crossover study in type 1 patients that showed reduced A1c values during both periods for sitagliptin as well as significant reductions in insulin dosages. The overall incremental reduction in A1c with sitagliptin was -0.27% (p = .025). CGM results from this study showed reductions in mean blood glucose and other indices of glycemic control. Finally, he covered colesevelam, a bile acid sequestrant that lowers LDL and also blood glucose with an unknown mechanism of action (mean A1c reduction of 0.5%). Dr. Edelman informed the audience that Dr. Garg also published a paper this year in Diabetes, Obesity, and Metabolism investigating the role of colesevelam in lowering LDL and in glucose control. The drug reduced LDL by 10% through 12 weeks of treatment and achieved a significant A1c reduction over placebo.

Questions and Answers

Q: I am a nurse practitioner at Walter Reed. It seems like the better our drugs become for type 2 diabetes, the more evidence that comes out about adverse events. What are you discussing with patients about the use of Actos and bladder cancer with regards to duration and dose?

Dr. Edelman: That’s a great question. When we discuss side effects with patients, it depends on how you say it and what you say. If you really want to scare people, you can. Pioglitazone is now associated with bladder cancer, CHF, etc. The laundry list is long. What I try to do is put together a few highlights of the numbers and some of the data put out by the FDA that puts it into perspective. If you’re a pro-TZD person you might say things that are more acceptable for the drug. We really have to look at the side effects of these drugs over time. If you look at the list of drugs taken off the market, it’s a long laundry list. We have to look at these GLP-1 agonists over time for renal failure, pancreatitis, etc. I’m sort of becoming more conservative in my age and I have to look at the side effects over time.

Dr. Garg: Drugs showed toxicity even when they went for approval. I personally have never written a prescription for TZDs, but having said that, this is a judgment you need to make based on the literature.

Dr. Edelman: I predict it will be taken off the market.

Dr. Garg: Yes, I agree.

Q: I like the idea of taking blood sugars over a period of time and presenting that info to the patient in order to change behavior. I tried doing something like that. We tried it in a study with three groups: Caucasians with mental illness on an atypical antipsychotic, migrant workers now living in the United States who do not speak English, and Navajo Indians. We asked them to check their blood sugar five times per day, including once at bedtime. We achieved big differences in postprandial levels after dinner and also before bedtime. We asked people to adjust what they ate at dinner and had their provider give some feedback. We got a composite A1c drop of 0.9% for all three groups.

Dr. Polonsky: I think you should try to do a harder study next time. [Laughter]

Dr. Garg: This is a great example. We need to empower the patients to use the information, and it sounds like you did that.

Dr. Polonsky: To follow up further, we’re really trying to consider what blood glucose monitoring can be, and how it can be a motivational tool in addition to CGM.

Dr. Edelman: What really bugs me is the limitation on test strips placed by the VA hospital and insurance companies. We can use these things for behavioral modification and payors don’t realize that they’ll probably save money over time if they just cover it. And this is clearly not amenable to a double-blinded placebo controlled trial.

Dr. Hovorka: Most patients can analyze the records now and make adjustments. I work mostly with kids and adolescents. Testing is not such a big issue for them, but keeping the records in a sophisticated way is the difficult part. There is a disconnect between what we try to accomplish and what they do in real life. How can we improve in this area?

Dr. Polonsky: Never work with teenagers or children again. [Laughter] What we do is this: we try to pose a meaningful question for the teenager that they might be interested in answering. “I wonder why you didn’t do so well in tests.” Target something of interest to them and try to convince them to the point where they say, “Oh, this testing stuff isn’t just a way for my parents and doctor to bug me.” Focus on a meaningful question for them.

Dr. Edelman: Text them! They like texting. [Laughter]

Q: I try to stress the physiological and the physical connection with my patients. But some of the psychotropic drugs increase blood glucose. Would psychiatric drugs be better for these patients if we emphasized the ones that actually decreased the blood glucoses?

Dr. Edelman: What I’ve seen is that it’s really trial and error. Sometimes, patients are affected by one medication and not the other. We don’t really know why, but I do think it’s important to monitor them closely. Certainly these drugs are important for psychological reasons but you have to watch what happens to them. Talk to the psychiatrist and psychologist and try to switch them to an equivalent one that has a different mechanism if you get adverse metabolic effects.

Dr. Hovorka: There isn’t really a known biochemical mechanism. The individual is eating more and gaining weight. The fact is that the most effective drugs had the highest impact on weight gain! We should think about the best atypical antispychotics and make additional efforts to control weight. Try and look at the whole context.

Q: I work as an endocrinologist. What is the mechanism of action for A1c reduction in colesevelam? Is it GLP-1 mediated? And should its effect on triglycerides be taken into account?

Dr. Edelman: You should look at the baseline triglycerides, as this can be made worse. The A1c reduction mechanism is not known. Being at the VA, those labs are easy to get and you can follow them over time with EMRs.

Dr. Garg: We don’t know the reason for the A1c reduction, but we think it’s probably through GLP, and that’s why we did measure the GLP levels in the study. Ideally, we would like to use this drug in combination with one of the DPP-4 inhibitors, and that’s one of the studies we are about to start after looking at the data for this larger study. GLP response was only seen at four weeks and not at 12 weeks. We don’t know the actual reasons, but the study should be done with a much larger size to clarify this.

Q: My question is regarding the SMBG study group. Five percent dropped out, and the effective group that met the protocol was 50%. Did you follow up with other 50% in order to find out what’s common to them?

Dr. Polonsky: This is a relatively poor group of patients and their starting A1c was higher to begin with. We were impressed that people stayed in the study to the degree they did. Two-thirds managed to follow the protocol pretty well. The dropouts tended to be poor and African American, had lower levels of education, and life was harder for them, including making regular medical visits, which had been a challenge.

Can We Really Prevent and/or Cure Diabetes?


Peter Gottlieb, MD (University of Colorado Health Sciences Center, Denver, CO)

Dr. Gottlieb began his talk with a caveat: that we are still “a little bit away” from a cure and that he probably would not have used the title assigned to his talk. However, his overall tone was optimistic, and he gave a number of compelling arguments throughout his presentation. He first outlined the main immunologic mechanisms involved with type 1: a) the balance of autoreactive and regulatory T cells; and b) antigen presentation and innate immunity. Possible approaches addressing the balance of autoreactive and regulatory T cells include cord blood stem cells, GCSF (recruitment of naïve cells with bias towards Treg), dendritic cells (loaded with protective cytokines or antigens), CD4+CD25+Treg cells, and depletion or modulation of T effector cells (e.g., anti-CD3). Regarding antigen presentation and innate immunity, macrophages play an especially important role in the progression of type 1. Overall, the data suggests that IL-1 may play an important role in disease pathogenesis and that innate immune pathways may be possible targets for immune intervention. Therapeutic antibodies against IL-1 are currently in development (Anakinra, Omni AAT, TrialNet Prevention trial). Other approaches include dietary interventions, oral insulin, probiotics with insulin peptide and IL-1o, and insulin targeted toward specific APC populations with or without anti-inflammatory treatment (anti-IL1). Dr. Gottlieb was especially enthusiastic about TrialNet’s efforts — there are 117 North American affiliates working on a novel immune tolerance mechanism involving ingestion of an oral antigen that promotes regulatory lymphocytes and protective cytokines that inhibit beta cell auto-immunity. In trials that have come out of this project, time to diabetes has been significantly greater for treated groups (DPT-1 oral study). The main limitations thus far involve dosing, adjuvant therapy, timing, and route, which may be critical to efficacy despite our lack of knowledge about them. Finally, he ended with a bold statement: multicenter trial networks and combination therapy targeting multiple pathways will help us find a cure for type 1 diabetes in the next decade. We hope so, but until we see better data on the ability to achieve immune tolerance, we remain skeptical.


Massimo Trucco, MD (University of Pittsburgh, Pittsburgh, PA)

In this presentation, Dr. Trucco reviewed an interesting biological attempt to prevent diabetes. He first noted that at the stage of thymic selection, any mechanism (viruses, etc) that allows anti-self T cells to survive triggers autoimmunity. The main mechanism behind cellular therapy involves presenting insulin antigens without co-stimulation mechanisms in order to have dendritic cells become regulatory, as opposed to stimulatory of an auto-reactive immune response. These so-called anti-sense dendritic cells were infused in 14 carefully selected diabetic patients in a phase 1 investigation, and compared with three “control” dendritic cells. After infusion, A1c did not worsen in these patients (although it did not get better), but type 1 autoantibody titers did decline in one patient who was strongly positive. Interestingly, B regulatory cells (B220+ CD11c-) did trend upwards in a few patients, prompting further study and characterization of this mechanism. In a study by Giannoukakis et al. (Diabetes Care, 2011 in press) the data showed that these cells could significantly increase in number after anti-sense dendritic cell administration. Going forward, the aims of phase 2 trials will be to improve the indices of functional beta cell mass in new-onset diabetes (as opposed to established diabetes), including the elimination of autoantibodies and an increase in regulatory T cells and B cells. He concluded by citing a June 28, 2011 article appearing in the Lancet by Bart O. Roep, urging investigators to avoid raising false expectations among patients and physicians, especially given the recent disappointing phase 3 immune intervention trials. We agree.


Ron Gill, PhD (University of Colorado Denver, Denver, CO)

Dr. Gill began by stating that it would be disingenuous to say that we will definitely cure type 1 diabetes, a view that explains his addition of the talk’s subtitle. A common theme throughout this talk, however, was that the rate of progress has been impressive and we should be encouraged. He intentionally focused more on islet cell transplantation than reversal at disease onset or primary disease prevention. Dr. Gill described islet transplantation as a “biological prosthesis, patented by God.” He also described three ways in which transplantation could occur: transplant with chronic immune suppression (current method), transplant with more benign immune suppression (under development), and a durable transplant without immune suppression (effectively a cure). What makes the latter so difficult is the fact that replaced islets have to counter both “transplantation immunity” and “autoimmune injury,” making the issue of immune tolerance paramount to success.

  • Dr. Gill emphasized the fact that prior to the year 2000, less than 10% of islet recipients ever achieved insulin dependence. These patients had received the Edmonton protocol, which included tacrolimus and rapamycin given intermittently for 56 days, followed by daclizumab administration after 90 days (all immune modulators). Post-transplantation, all patients had blood glucose levels that stayed largely within the normal physiologic range, with almost no hypoglycemia! A1c levels dropped from a mean of 6.6% to 4.8%. However, longer-term follow-up after five years saw gradual attrition of islet cell function. While C-peptide secretion was preserved, less than 10% of patients maintained insulin independence.
  • Dr. Gill discussed the Clinical Islet Cell Transplant Consortium, a current multi- center demonstration trial for islet transplant alone and islet-after-kidney transplant. This has produced new approaches such as T lymphocyte depletion strategies (TCD), pioneered by Bernhard Hering of the University of Minnesota. Insulin independence was achieved in 100% of the cohort with this therapy initially, and at 5 years, insulin independence rates were comparable to those seen in pancreatic transplant (ITA 57%, PTA 52%). This progress has taken a much shorter time to achieve than it took to develop the Edmonton Protocol.
  • Dr. Gill ended the talk by broaching the “elephant in the room”: we do not yet have a robust means of inducing immune tolerance in any autoimmune disease, including type 1 diabetes. An important reality check is that we still chronically immune suppress patients who undergo clinical transplantation. Additionally, long-term solid organ allograft survival has only modestly improved over many years. The five-year pancreas transplant survival remained between 40-50% in both 1994 and 2003.


George S. Eisenbarth, MD, PhD (Executive Director, Barbara Davis Center, Denver, CO)

Dr. Eisenbarth began by asserting that we do NOT have any established way of resurrecting residual beta cell mass. Urging cautious optimism, he did feel there was progress on this front. He began by emphasizing a very important distinction between insulin auto-antibodies (IAA) and others: IAA is the only autoantibody whose magnitude is correlated with time-to-onset of type 1 diabetes. He also reminded us of the pathologic distinction characterized by Gianani et al. in Diabetologia 2010: type A patients have atrophic islets throughout the pancreas, whereas type B patients have no pseudoatrophic islets, although the beta cell mass is markedly decreased. With this in mind, he reviewed the potential methods for resurrection: “euglycemia” methods (the DCCT Preservation Protocol and TrialNet’s Metabolic Protocol, both of which investigate whether good control early on delays disease progression), pancreas transplant and immune suppression, and finally, novel immunotherapeutics. He presented data surrounding the DCCT Intensive Therapy Protocol, where beta cell function was more robust and time to retinopathy and nephropathy was delayed with intensive therapy. Dr. Eisenbarth also reminded the audience of the ongoing TrialNet Metabolic Protocol. The study places patients on a three or four day hybrid closed-loop session within seven days of diagnosis, with patients continuing on CGM and manual pump control after discharge. As a reminder, promising interim data was presented at ADA in a talk by Dr. Stuart Weinzimer and a poster by Dr. Bruce Buckingham and colleagues (for more information, see our ADA 2011 Full Report in the July 15, 2011 Closer Look at Finally, Dr. Eisenbarth touched on the prospect of using combination approaches (euglycemia plus novel antigen-specific therapeutics) to achieve success in beta cell preservation. However, he does not feel that beta cell resurrection is likely.

Questions and Answers

Q: Dr. Gill, you showed a slide that said C-peptide productivity was maintained in 80% of the individuals but 90% after five years still needed insulin. Why so?

Dr. Gill: Good question. What’s not shown is how much C-peptide per patient was produced. While there are still residual beta cells, they are not making enough insulin to control their blood sugar. To clarify, they did still have enough to prevent hypoglycemia. At the University of Alberta, they are trying to find out how much C-peptide they make and what the rate of decline is. The challenge that remains, however, is asking islets to remain in an avascular state – by definition those tissues are chronically hypoxic. The question is, can they function? There are a lot of animal models that show achievable function for quite a number of weeks, but this will be a challenge. This becomes a tissue-engineering question. For durable function, this is the kicker.

Q: A child in my family has mild hyperglycemic symptoms and four uncles with type 1 diabetes. They just sent his blood in to TrialNet and his antibodies are negative. What should we make of this? Will he get diabetes?

Dr. Eisenbarth: We don’t measure all four known antibodies right now. He also might have a different form of diabetes that is prevalent in Hispanic or African American populations. He likely does not have type 1A, as teenagers often do not. As we follow children, there are some who have lost their autoantibodies. Ninety percent of patients at onset will have one of the four antibodies. It might be useful for the uncles who are alive to have their GAD antibodies tested.

Q: We currently screen newborns for about 50 or so metabolic diseases. Do you see a time where we’d screen newborns for autoantibodies and follow these kids along intensely?

Dr. Eisenbarth: As I said, the antibodies aren’t present at birth; they come up later. However, eventually what you’re describing will occur through genetic analysis and HLA typing. At present, the highest risk individuals are those with HLA DR 3/4, such that 2% of newborns in Colorado are DR 3/4 positive, and 30-40% of patients with type 1 diabetes are DR 3/4 positive. Note that developing diabetes is a one in five risk only for patients with HLA DR3/4. But eventually, we will probably have some form of surveillance.

Q: I have a question about autoantibody positive patients who might have a bit of polyuria and polydipsia, but not overt diabetes. What should we do with these patients in the clinic?

Dr. Gottlieb: My recommendation at this point would be to certainly screen them for overt diabetes going forward. Currently the anti-CD3 trial might also be available to them. We do think that IGT is probably a step in the progression, and there does seem to be some progression. Periodically screen them and figure out what they’re doing. As we develop therapies that are antigen-specific, we will likely identify patients with IGT as potential candidates for therapy.

Q: Could someone address the issue of a typical teenager in their honeymoon period who is just taking their basals, eating carbs, doesn’t go up that high, and does always require insulin? They often say, “I always correct.” Given the data you presented, are they wearing out their beta cells in their honeymoon by not achieving the best control? Or should I just let them enjoy their honeymoon period while it lasts?

Dr. Gottlieb: Given the metabolic studies we mentioned, including dosing for meals, it is very likely that this could be helpful. In the islet transplant work that we did, the challenge was giving insulin in an appropriate way. If done correctly, it can be very, very helpful. You just have to try to push them to take mealtime insulin.

Q: My question is to Dr. Trucco. Can you describe the rationale for using dendritic cells specifically for new-onset patients in a bit more detail?

Dr. Trucco: These dendritic cells are always producing newer antigens for the T cells. One reason why we didn’t keep the antisense and the peptide that George was describing is that we need to intervene at the onset of disease. We believe that this dendritic cell can pick up anything, and the idea is to flip the balance. It should be a little bit of an advantage over the aggressive/auto-reactive T cells. They block T cell clones. My major concern, however, is that in the results, we will see some effect but it won’t last long. If we do get it to last a year or two, then maybe you can renew the process.

Q: Do you envision using this type of intervention in kids who just became positive for their first antibody?

Dr. Trucco: I would not go that far. When they start to have a couple of autoantibodies present, perhaps that will be a time to intervene.

-- by Adam Brown, Shereef Elnahal, Sanjay Trehan, and Kelly Close