European Association for the Study of Diabetes (EASD) 48th Annual Meeting Preview

48th Annual Meeting September 30, 2012 – October 5 2012; Berlin, Germany – Full Commentary – Draft

Executive Highlights

In this final report, we provide our complete coverage of the 48th Annual Meeting of the EASD, held in Berlin, Germany from September 30 to October 5, 2012. This year 18,127 participants attended EASD 2012, up substantially from 17,462 attendees last year in Lisbon and 17,301 in Stockholm in 2010, and even greater than 17,890 at this year’s ADA in Philadelphia! This year’s six-day long EASD meeting (the longest meeting ever!) featured six tracks showcasing 361 oral presentations and 39 symposia, identical to last year’s 361 oral presentations and 39 symposia, but with a substantial increase in corporate symposia (from 15 in 2011 to 24 in 2012 – wow!), and a slight reduction in posters (from 1,152 in 2011 to 1,120 in 2012). There were 92 exhibitors in 2012, up a sizable amount – 30%! – from 2011.

To help you sort through all of the learnings, we’ve organized our commentary into sections: (1) Incretin Therapies; (2) Insulin and Insulin Therapies; (3) Diabetes Technology; (4) Novel Therapies and Basic Science; (5) Obesity and Prediabetes; (6) Complications and Cancer; (7) Type 1 Diabetes Treatments and Cure-Based Therapies; (8) Regulations and Reimbursement; and (9) Exhibit Hall.

Titles highlighted in yellow are new additions to the full report that were not initially included in our live daily updates from Berlin. Additionally, titles highlighted in blue represent an assortment of one dozen of the most memorable EASD 2012 talks from the perspective of the Close Concerns team. These top talks have also been collected in one document that can be found at It was difficult to choose our very favorites, and this document is just meant to serve as a starting point for EASD learning for those who have limited time to go through the full report. We chose a dozen talks from over 300 different presentations we attended, and we narrowed our favorites down to one on the ADA/EASD position statement, three on devices, two on novel therapies, four on incretins, and two related to insulin. Notable speakers in this list include Dr. Dan Drucker in his masterful Claude Bernard lecture, Dr. Roman Hovorka on the closed loop, and many others.

Below we outline major themes discussed at the meeting, followed by a table of contents for your reference.

  • On the drug front, we were heartened to see advancements on efforts to improve patient experience with incretins through progress in once-weekly GLP-1 therapy, a yearlong- implantable GLP-1 device, a once-weekly DPP-4 inhibitor, a more selective DPP-4 inhibitor, and GLP-1/basal insulin combination therapy. We learned from Dr. Michael Nauck (Diabetes Centre, Bad Lauterberg im Harz, Germany) that the highest dose tested of Novo Nordisk’s once-weekly GLP-1 agonist, semaglutide, performed better than the highest dose of Victoza in terms of A1c reduction and weight loss, but with a worse nausea profile that resulted in the decision to carry forward a lower dose to phase 3. Additionally, we paid close attention to Intarcia’s ITCA-650, an implantable extended-release device for exenatide that would be implanted once at a low titration dose for three months and then replaced every six to 12 months thereafter. Promisingly, in phase 2 studies ITCA-650 provided superior A1c reduction, similar weight loss, and less nausea at 12 weeks compared to the equivalent dose of Byetta (exenatide twice-daily); as we understand it, implantation is an easy process and we hope explantation can become so too. Merck’s once-weekly DPP-4, MK-3102, demonstrated significantimprovements in A1c over placebo at all six doses tested in a phase 2b trial; a once-weekly DPP-4 certainly may offer an adherence benefit as a first-line option for patients who are not tolerant of metformin and we believe would even be appealing to patients (and payors) for those on metformin, since we know that plenty of patients forget to take even easy-to-use daily medicine. We also attended a presentation by Dr. Jonathan Rosenblum (ActivX Biosciences) on preclinical pharmacodynamics data for KRP-104, a more selective DPP-4 inhibitor that is currently in phase 2 trials. Interestingly, Dr. Rosenblum asserted that he would refuse to take sitagliptin because of its impact on the liver – a notable statement given the general consensus about the safety of the DPP-4 class – and that ActivX seeks to make KRP-104 the best-in-class DPP-4 inhibitor based on functional differentiation, something that no other DPP-4 inhibitor has shown. Additionally, we learned that four-year data from the DURATION-1 follow-up study for Bydureon (exenatide once- weekly; the longest-term GLP-1 data we have seen to date) demonstrated a solidly durable A1c reduction and moderate weight-loss durability out to four years – durability was always mentioned even in the early days as an advantage of GLP-1 and this view was certainly reinforced by this data.
  • GLP-1/basal insulin combination therapy continues to be a topic of interest when discussing attractive options for the treatment of type 2 diabetes. Dr. Tina Vilsboll (University of Copenhagen, Copenhagen, Denmark) strongly advocated for this combination at the Lilly corporate symposium, and panelists at a round-table discussion during Sanofi’s corporate symposium agreed with this sentiment – it’s now a foregone conclusion from our view that this combination is widely used and will become ever more so when there are combination applications. Speaking of, we attended an oral presentation on the GetGOAL-L study for Sanofi’s lixisenatide/Lantus (insulin glargine) that demonstrated that the combination reduced placebo- adjusted A1c beyond Lantus alone – although not as big as the placebo-adjusted A1c reduction conferred by a combination of linagliptin and basal insulin as reported in a different oral presentation (0.36% vs. 0.65% placebo-adjusted reduction for lixisenatide/Lantus vs. linagliptin/basal insulin, respectively). We found this result noteworthy even though the lack of head-to-head data makes a true comparison impossible at this point. Still, whatever turns out to be the best formulation, it’s certainly absolutely clear at this point that combining incretins and basal insulin has become quite popular.
  • EASD 2012 included noteworthy discussions on the merits of early insulin use. The topic received focused attention during a debate in which Dr. Jack Leahy (University of Vermont College of Medicine, Burlington, VT) argued that early insulinization provides better glycemic control than the early use of oral agents, and that the weight gain and hypoglycemia associated with insulin can be minimized when given with metformin, or to patients with lower A1cs. In contrast, Dr. Guntram Schernthaner (University of Vienna, Vienna, Austria) argued that early insulinization provides little benefit, as demonstrated by ORIGIN, but increases the risk of hypoglycemia and exacerbates weight gain. Dr. Rury Holman (University of Oxford, Oxford, UK), in support of early insulinization, reasoned that since insulin use is usually inevitable for type 2 diabetes, there is no reason to wait. He maintained that insulin helps patients quickly achieve glycemic control, may prolong beta cell function, and enhances the effectiveness of other oral agents. The topic received further attention during a panel discussion sponsored by Sanofi: the KOLs generally favored the early use of insulin, but commented that patients’ and HCPs’ negative attitudes toward insulin therapy represent a significant treatment barrier. From a patient perspective, we love the idea that all patients would start of getting at least some insulin so that the mystery would be taken out of it and so that going back on insulin (eventually) would beimbued with a fair degree of normalcy and wouldn’t come with all the drama that often accompanies patients going on both basal and mealtime insulin.
  • EASD attendees received an influx of new phase 3 data on insulin degludec. Degludec’s lower risk of hypoglycemia was a recurring theme: a 52-week head-to-head trial of insulin degludec vs. insulin glargine found that while degludec and glargine provided similar antihyperglycemic efficacy, degludec led to lower rates of overall hypoglycemia and of nocturnal hypoglycemia. Novo Nordisk also highlighted the potential for degludec to be dosed flexibly – eight to 40 hours after the previous injection: the FLEX T1 (type 1 diabetes) study and the FLEX T2 study showed that degludec “flexible” dosing led to a significantly lower rate of nocturnal hypoglycemia compared to glargine. Also of interest were data from two phase 3 studies investigating degludec given three times weekly (3TW), which shed light into the ultimate decision to dose degludec once a day: degludec 3TW failed to achieve non-inferiority in reducing A1c compared to glargine and led to greater rates of hypoglycemia. In addition, a trial comparing degludec with sitagliptin found that degludec provided greater A1c and fasting plasma glucose reductions, but more weight gain and hypoglycemia (as would be expected).
  • One major disappointment of EASD was that little new data was presented on other novel insulins or currently available insulin therapies – with a couple of exceptions. A study characterizing Lilly’s investigational PEGylated insulin lispro (LY2605541) found that LY2605541 has ~100-fold greater selectivity for the insulin receptor (IR-A) over IGF-1 compared to insulin lispro, as well as significantly lower IGF-1 binding compared to other insulin analogs. On the heels of their presentations at ADA, Dr. Hertzel Gerstein (McMaster University, Hamilton, Canada) reviewed the results of the ORIGIN trial and Dr. Matthew Riddle (Oregon Health and Science University, Portland, OR) revealed new analyses based on baseline characteristics of patient subgroups. EASD 2012 also included discussion on insulin glargine’s cancer risk. Results from the International Study of Insulin and Cancer (ISICA) showed no difference in breast cancer risk in patients taking glargine vs. other insulins; the study also found that the duration and dose of insulin glargine were not associated with increased breast cancer risk. A meta-analysis of 18 studies, including ORIGIN, found similar results – that the risk of cancer is not elevated with insulin glargine compared to other insulins.
  • Speakers investigated whether improved insulin delivery could confer additional glycemic benefits, with mixed results. Interim results from an ongoing trial of InsuLine’s InsuPad (a patch-like device that provides localized heating after injection) in type 2 diabetes showed a significant 23% reduction in maximal glucose excursions in those using InsuPad vs. traditional insulin injection. This led Dr. Andreas Pfutzner (IKFE, Mainz, Germany) to hypothesize that patients using the InsuPad should require less insulin (~20-30%) for postprandial glucose control in real-world settings. The device was well tolerated in terms of skin irritation, though we note that wearing the pad (and replacing it every day) adds hassle factor . In the same symposium, a study of Generex’s buccal spray insulin, Oral-lyn, had less-than-striking results. Dr. Andreea Soara (University Campus Bio Medico, Rome, Italy) reviewed a six-month intervention study that randomized 22 patients with IGT to Oral-lyn (12 puffs at each meal) or a control group. Oral-lyn treatment improved A1c by 0.4% (baseline: 6.2%) after six months of treatment, but the benefit was lost during the six-month post-trial washout period. We certainly salute the company for such an ambitious trial in people with IGT, though there’s no question this isn’t for the faint of heart, especially considering the economics of treating people with prediabetes. The other major highlight in insulin delivery came in the exhibit hall, where we got a closer look at Debiotech’s new Jewel Patch Pump (see exhibit hall theme below). Developing novel insulin delivery methods is an area desperate for innovation in our view. We hope to seemore in the coming years on MannKind’s Afrezza, BD’s microneedles, and perhaps even the “holy grail,” oral insulin, though we aren’t holding our breath on the last.
  • On the device front, Dexcom’s star-studded G4 Platinum corporate symposium was the highlight in CGM at EASD 2012, though late news on the European launch of Abbott’s Navigator II was also buzzworthy. The excitement for Dexcom’s new CGM (see our report on the post-EASD FDA approval call at was tangible during the company’s very well-received corporate symposium on the second day of the conference. We saw new cuts of G4 Platinum accuracy data from Dr. Thomas Peyser (VP Science and Technology, Dexcom, San Diego, CA), preliminary results from Dr. Bruce Buckingham (Stanford University, Stanford, CA) on a fantastic nocturnal remote monitoring study at a diabetes camp using the G4 Platinum CGM in pediatrics, and a sneak peak at Dexcom’s future pipeline (e.g., predictive algorithms, the accuracy of the G4 AP version) from the eloquent Dr. Jay Skyler (University of Miami Miller School of Medicine, Miami, FL). For the first time that we can recall, Dexcom’s future technology– combining remote monitoring, highly accurate CGM, and potentially predictive alerts – waspositioned as an alternative to low glucose suspend (LGS). This would reinforce competition between Dexcom and Medtronic based not only on CGM device quality, but an approach to insulin therapy and hypoglycemia mitigation. Next-generation CGM news also came in an outstanding presentation by the esteemed Dr. Roman Hovorka (University of Cambridge, UK), where we learned that Abbott “silently” launched its FreeStyle Navigator II two to three weeks before EASD. We’re excited to hear this given the impressive accuracy and AP researcher enthusiasm for the original FreeStyle Navigator and hope to hear more about the device soon. Roche also emphasized its CGM in development (which it hardly ever mentions!), and sales representatives at the exhibit hall said to expect data at ATTD 2013. News on Medtronic’s Enlite was relatively absent this year, though this was likely to be expected since the sensor has been on the European market since April 2011. While an intermittent CGM study in pregnancy had disappointing results, the study without question speaks more to the need for continuous use and next-generation technology rather than an inherent lack of efficacy.
  • While the sessions brought little new news on the artificial pancreas, data on Medtronic’s predictive low glucose management (PLGM) algorithm and a per- protocol analysis from the CAT trial showcased the benefits of closed-loop algorithms on hypoglycemia. Using computer simulations, Dr. Barry Keenan (Medtronic Diabetes, Northridge, CA) showed how the new PLGM algorithm is expected to reduce the number of hypoglycemia events (<70 mg/dl) by 18% and the average duration of hypoglycemia by 50%, a significant improvement over the Veo’s corresponding reductions of 1% and 28%. We cannot wait to see the clinical data on the system, hopefully at ATTD 2013 in Paris. We also saw the per-protocol analysis of the CAT Trial, which tested closed-loop control by two algorithms (Padova/Pavia/UVa’s MPC algorithm vs. Cambridge’s MPC algorithm) against open-loop control. The presentation confirmed the algorithms’ benefits in reducing hypoglycemia that was seen in the intent-to-treat analysis presented at ADA 2012 (see page 86 of our coverage at Last but certainly not least was Dr. Hovorka’s impassioned and highly comprehensive argument in favor of pursuing a “mechanical” solution to type 1 diabetes. In addition to the FreeStyle Navigator II launch news (see above), he provided a valuable update on the three-week overnight closed-loop home studies that the Cambridge team is currently conducting. The glucose traces from the first patients look quite good, and the study has recruited six out of a planned 16 participants.
  • On the novel drugs front, SGLT-2 inhibitors continued to be a hot topic of discussion at EASD. Throughout the conference, speakers highlighted low risk of hypoglycemia, weight loss, and an insulin-independent mechanism of action as benefits of the SGLT-2 inhibitor class. On the safety front, speakers acknowledged the increased incidence of genital and urinary tract infections seen with the drug class, but did not think such undesired side effects would prevent their use (speakers noted that many people with diabetes have previously experienced genital and urinary tract infections, and that the infections are treatable, oftentimes with over-the-counter medications). In addition, several speakers expressed doubt that the cancer imbalances seen in the dapagliflozin development program were due to the drug treatment. In terms of where SGLT-2 inhibitors might fit into the treatment paradigm, Dr. John Wilding (University of Liverpool, Liverpool, United Kingdom) suggested that they could be used after failure of metformin; Dr. Kamlesh Khunti (University of Leicester, Leicester, United Kingdom) expressed similar sentiments, commenting that dapagliflozin would be suitable for a primary care setting. Dr. Jochen Seufert (University of Freiburg, Freiburg, Germany) noted that SGLT-2 inhibitors could theoretically be effective as a treatment for type 1 diabetes, given their insulin- independent mechanism of action.
  • Additionally, on the topic of novel drugs, EASD included interesting discussion on glucagon antagonists, proglucagon-derived molecules, and glucagon fusion peptides. A 12-week study of Lilly’s glucagon antagonist LY2409021 found that the drug provided significantly greater improvements in A1c compared to placebo, with few side effects and little hypoglycemia. Also notable were preclinical data on Zealand Pharma’s GLP-1/gastrin dual agonist ZP3022, which provided similar glycemic control to liraglutide and increased beta cell mass at eight weeks. Several presenters also noted the potential for glucagon-mediating therapies. During the 44th Claude Bernard Lecture, Dr. Daniel Drucker (University of Toronto, Toronto, Canada) gave an extensive review of the therapeutic potential of glucagon, GLP-1, and GLP-2, noting that while glucagon antagonism lowers glucose levels and improves islet cell function, such therapies could potentially have negative side effects due to glucagon’s important roles in liver cell survival and liver lipid homeostasis. In addition, Dr. Benjamin Field (Imperial College London, London, United Kingdom) and Dr. Matthias Tschop (Helmholtz Center, Munich, Germany) explained that glucagon/GLP-1 dual agonists lower blood glucose levels and reduce body weight. In his own presentation on glucagon-based incretin hybrids, Dr. Richard DiMarchi (Indiana University, Bloomington, IN) reviewed preclinical data demonstrating that glucagon/GLP-1 and gastrin/GLP-1 dual agonists reduce hyperglycemia and body weight and improve other metabolic parameters. While the data were not new, his talk reflected the increasing excitement surrounding these novel therapies.
  • Several presentations discussed diabetes complications, and we came away with the impression that even the newest options for treatment are not yet good enough. In diabetic macular edema, anti-VEGF holds promise (as demonstrated by three-year follow-up data on the RESTORE trial for Lucentis), but some speakers (such as Dr. Pascale Massin [University of Paris, Paris, France], Dr. Jost Jonas [University of Heidelberg, Heidelberg, Germany], and Dr. Per-Henrik Groop [Helsinki University Central Hospital, Helsinki, Finland]) were still dissatisfied by its high cost, its invasiveness as an intra-vitreal injection, the need for frequent follow up, and the dearth of long-term experience. We also attended two presentations on the COMBO-DN study of pain management therapies in diabetic neuropathy. Sadly, a combination of duloxetine and pregabalin did not provide better pain relief than high doses of either monotherapy, and we learned from Dr. Solomn Tesfaye (University of Sheffield, Sheffield, UK) that “the best we can hope for is 50% pain reduction in 50% of patients.” We also learned from epidemiologist Dr. Trevor Orchard (University of Pittsburgh, Pittsburgh, PA) that renal function appears to be a more important risk factor for coronary artery disease for people with better controlled A1cs – another reason in our growing list that renal disease deserves more attention than it gets. On diabetic nephropathy, we attended an encouraging talk by Dr. Merlin Thomas (Baker IDI Heart and Diabetes Institute, Melbourne, Australia) who H2-rich water as a particularly promising option in diabetic nephropathy. (Dr. Thomas also conveyed excitement about bardoxolone methyl, though prospects for this drug – and for near-term CKD therapies overall – became much bleaker with the stoppage of bardoxolone’s pivotal trial two weeks later.)
  • Discussions of potential cures for type 1 diabetes were largely focused on two topics: the Protégé trial and the potential role of enteroviruses in causing (and maybe also preventing) type 1 diabetes. In a symposium dedicated to the possible role of enteroviruses and type 1 diabetes, Dr. Matthias von Herrath (La Jolla Institute for Allergy and Immunology, La Jolla, CA) presented research showing that while serious viral infections might cause type 1 diabetes, more mild infections may be preventative. He suggested that a vaccine against serious, diabetes-causing viruses might be effective. In the same symposium, Dr. Heikki Hyöty (University of Tampere, Tampere, Finland) argued that a vaccine could prevent more than 50% of type 1 diabetes cases. Another focus of discussions was the anti-CD3 inhibitor teplizumab’s Protégé study. Dr. Johnny Ludvigsson (Linköping University, Linköping, Sweden) presented data showing that a 14-day teplizumab regimen had no significant impact on C-peptide preservation at one year, but that the impact was significant at two years. Dr. Ludvigsson (and in a later session Dr. Kevan Herold [Yale University, New Haven, CT]) argued that while the Protégé trial failed to meet its primary endpoint, the study should not have halted teplizumab development. In our opinion, even greater focus and time should be placed on good study design at the very beginning of therapeutic investigations to make sure we do not end up wondering “what if,” as we currently are for teplizumab. Other notable presentations on type 1 diabetes treatments included reviews of the phase 1 alpha-1 anti-trypsin RETAIN study and the phase 3 DiaPep277 DIA-AID 1 study, which suggested each of these therapies has beta cell protective effects.
  • The new ADA/EASD Position Statement and the concept of individualizing therapy were often the center of discussion throughout EASD 2012. The highlight was a particularly noteworthy symposium fully devoted to the new position statement on the conference’s concluding day. Co-authors Drs. David Matthews (University of Oxford, Oxford, UK) and Silvio Inzucchi (Yale University, New Haven, CT) provided valuable and enthusiastic reviews of the document, highlighting its rationale, development, and use in clinical practice. Both spoke highly of the position statement’s comprehensive, non-algorithmic approach, which helps address the heterogeneity of treating type 2 diabetes and the limitations of previous algorithms. The enthusiastic mood was somewhat dampened by Dr. Amanda Adler’s (Addenbrooke’s Hospital, Cambridge, UK) sardonic presentation to wrap up the session – she succinctly summed up the thorough position statement (“after metformin, add what you will”), called for future documents to be written under a more rigorous and transparent process, and proposed several possible reasons why the guidelines might or might not make a difference. The audience, in a poll preceding the talk, seemed to share her pessimism, with ~20% believing the guidelines will make a difference and ~80% believing they won’t. We certainly believe training of PCPs on the guidelines would make a major difference. Clinical impact or not, it was clear throughout EASD that companies are fully embracing the concept – Lilly Diabetes had an entire symposium devoted to individualizing treatment options, while BMS/AZ, Sanofi, and Roche all featured talks on the new position statement and/or the value of personalizing therapy. In many ways, the fact that the new statement embraces all therapies is a plus for pharmaceutical companies, who cannow market their products in niches that will fulfill certain patient needs (e.g., weight loss, efficacy, tolerability, etc.). In our view, the jury is still out on whether PCPs can handle it though we believe some policy enhancements could enable this, such as better reimbursement for HCP time, more time for patients with CDEs and doctors overall.
  • Reimbursement challenges appear to be omnipresent in the current financial environment. During IDF Europe’s first-ever pre-conference symposium at EASD, Dr. Lutz Heinemann (Science & Co., Dusseldorf, Germany) noted that in Europe there is a desire to tighten regulatory standards for devices, and he highlighted that devices (especially CGM) are still not well reimbursed in Europe (we learned that there are only 300 patients total on CGM in Germany, compared to at least 30,000 in the US). During the same session, Mr. Keith Tolley (Tolley Health Economics, Derbyshire, United Kingdom) reviewed France’s, Germany’s, and the United Kingdom’s reimbursement policies, commenting that there has been an increasing focus on price in countries that previously adopted free pricing systems (e.g., Germany and the UK); meanwhile, increasing price pressures are likely in countries that are going through particularly challenging especially financial times, such as Greece and Spain.
  • The bustling EASD 2012 Exhibit Hall featured 92 exhibitors, up from 70 the year before in Lisbon. There were a number of new technologies in particular that many, including us, were eager to see. For example, we saw Insulet’s small pod at the Ypsomed booth – it had begun shipping about a week and a half earlier in the EU, and from the crowds at the booth, there was definitely significant interest. As well, we had a fresh look at Debiotech’s updated Jewel patch pump, a device that has flown under the radar since its debut at ADA 2010 – it has never been launched and we hadn’t been sure of its status, but trials will soon commence to support submission to European regulators and the FDA. The newest feature is a pump controller that doubles as a sleek, fully functional cellphone and includes an integrated blood glucose meter . We got a full working demo of the pump and a much better feel for how Debiotech plans to differentiate the Jewel from other pumps (the BGM/cellphone controller, a 500-unit reservoir, on-demand bolusing without the controller, size, cloud-based data capabilities). Other notable device news came from Roche: the launch of the Accu-Chek Combo pump-BGM system in the US and news that the first data on the company’s own CGM will emerge at ATTD 2013. Dexcom was also present in the exhibit hall, showcasing their new G4 Platinum sensor and Studio software that was reviewed in more detail during a packed corporate symposium the day prior. Medtronic had a busy booth throughout the conference as well, drawing attendees seeking more information on the Veo low glucose suspend pump. On the drug side, the biggest areas of focus seemed to be SGLT-2 inhibitors (BI and J&J Janssen) and DPP-4 inhibitors (BMS/AZ, Merck, Novartis, BI/Lilly). Relative to last year’s big-time launch of Bydureon at the Lilly booth, GLP-1 had less fanfare in the 2012 exhibit hall – Victoza had its strong and well advertised presence in Novo Nordisk’s booth, though as might be expected given the BMS/AZ acquisition of Amylin, Lilly devoted just a small corner of its booth to Byetta and Bydureon. The three big insulin manufacturers all took a very international focus in their exhibits, emphasizing Life for a Child (Lilly); the vast unmet international needs in diabetes care (a striking five-column sculpture displaying the 366 million people worldwide that have diabetes, the 50% of those are diagnosed, the 50% of diagnosed patients that receive professional care, the 50% of those that actually reach treatment targets, and the small remaining portion that are free of complications) and the GAPP2 survey that looks into insulin and type 2 diabetes globally (Novo Nordisk); and various diabetes programs in India, China, and Saudi Arabia (Sanofi). There were games galore in this year’s exhibit hall as well: two hypoglycemia simulators, various uses of motion-sensing videogame technology and plenty of Jeopardy! to go around. And as we’ve come to expect at every EASD,there was definitely no shortage of gargantuan espresso machines, artful smoothies, and reusable shopping bags to carry freebies home.


Table of Contents 


1. Incretin Therapies

Oral Presentations: Incretin-Based Therapies


Michael Nauck, MD, PhD (Diabetes Centre Bad Lauterberg, Harz, Germany)

Dr. Michael Nauck presented new phase 2 data for semaglutide, Novo Nordisk’s once-weekly GLP-1 agonist (expected to advance into phase 3 trials in the first half of 2013; for more details on semaglutide development, see our Novo Nordisk 2Q12 report at In this 12-week dose-ranging trial, 411 patients were randomized to receive semaglutide once weekly, liraglutide (Victoza) once daily, or placebo. Semaglutide produced a dose-dependent reduction in A1c from a baseline of 8.1%. The 1.6   mg dose of semaglutide provided superior A1c reduction (1.7% [1.2% placebo-adjusted]) than the highest dose of liraglutide (1.3% [0.8% placebo-adjusted]). A dose-dependent effect of semaglutide on weight   loss was also observed, with patients on the 0.8 mg, 0.8 mg titrated, and 1.6 mg titrated semaglutide doses achieving better weight loss (2.2 kg [4.8 lbs], 2.4 kg [5.3 lbs], and 3.6 kg [7.9 lbs] placebo-adjusted weight loss, respectively) than those taking the highest dose of liraglutide (1.4 kg [3.1 lbs] placebo- adjusted weight loss). Notably, titration of semaglutide reduced nausea while maintaining the same A1c reduction and weight loss. Results from this trial will allow Novo Nordisk to define the dose range for phase 3 trials to optimize the balance between efficacy and side effects; Dr. Nauck stated that the highest dose of semaglutide tested in this trial (1.6 mg titrated) would not be carried forward to phase 3 due to  its side effect profile (31% of patients in the 1.6 mg titrated semaglutide arm withdrew from the study, with most citing GI adverse events as the reason).

  • Dr. Nauck explained that the structure of semaglutide is generally similar to that of liraglutide. Both have a free fatty acid attached via a spacer molecule at Lys-26, but semaglutide has an amino acid substitution at position 8 (Ala  → AiB [alpha-aminoisobutyric acid]) that prevents DPP-4 degradation. Semaglutide has about a 150-hour (6.25-day) half-life).
  • In this 12-week trial, 411 patients were randomized to receive semaglutide once- weekly, liraglutide (Victoza) once-daily, or placebo. Patients on semaglutide received 0.1 mg, 0.2 mg, 0.4, mg 0.8 mg, 0.8 mg titrated or 1.6 mg titrated semaglutide once weekly; patients given liraglutide received 1.2 mg titrated or 1.8 mg titrated liraglutide (n=43-50 per treatment arm). In the 0.8 mg titrated semaglutide group, patients were given 0.4 mg semaglutide for the first week and moved up to 0.8 mg for the second week. In the 1.6 mg titrated semaglutide group, patients were titrated from 0.4 mg the first week to 0.8 mg the second week, and 1.6 mg the third week. Patients had an average baseline A1c of 8.1%, BMI of 30.9 kg/m2, and weight of 87.5 kg   (193 lbs). Patients had an average age of 55 years, were 35% female, and had a very short duration of diabetes (2.6 years). Interestingly, Dr. Nauck implied that the study purposely enrolled people with short disease duration, suggesting to us that Novo Nordisk may position the drug as an early-stage treatment option.
  • Semaglutide produced a dose-dependent reduction in A1c from baseline, and the 1.6 mg dose of semaglutide provided superior A1c reduction (1.7% [1.2% placebo- adjusted] than the highest dose of liraglutide (1.3% [0.8% placebo-adjusted]). See the table below for full details. Additionally, more patients on the higher doses of semaglutide achieved A1c ≤6.5% or 7.0% than patients on liraglutide.
Treatment Dose

Absolute A1c Reduction

Placebo- Adjusted A1c Reduction



≤6.5% A1c



≤7% A1c

Placebo   0.5% - 4%



Once- Weekly

0.1 mg 0.6% 0.1% 13%


0.2 mg 0.9% 0.4% 28% 45%
0.4 mg 1.1% 0.6% 20% 56%
0.8 mg 1.5% 1.0% 50% 73%
0.8 mg 


1.4% 0.9% 45% 69%

1.6 mg 


1.7% 1.2% 63% 81%

Liraglutide Once-Daily

1.2 mg


1.2% 0.7% 34% 59%

1.8 mg


1.3% 0.8% 36%


  • A dose-dependent effect of semaglutide on weight loss was also observed, with patients on the 0.8 mg, 0.8 mg titrated, and 1.6 mg titrated semaglutide doses achieving better weight loss than those taking the highest dose of liraglutide. Semaglutide 1.6 mg titrated produced a 3.6 kg (7.9 lbs) placebo-adjusted weight loss; 0.8 mg titrated a 2.4 kg (5.3 lbs) placebo-adjusted weight loss; and the 0.8 mg untitrated, a 2.2 kg (4.8 lbs) placebo-adjusted weight loss. For comparison, liraglutide 1.8 mg produced a 1.4 kg (3.1 lbs) placebo-adjusted weight loss.
  • Mild nausea was the most common side effect, and was ameliorated by titration (60% of patients on the 0.8 mg untitrated semaglutide experienced nausea [68% of whom described it as mild] vs. 39.5% of patients on the 0.8 mg titrated semaglutide). However, 31% of patients on the highest dose of semaglutide (1.6 mg titrated) withdrew from the study, with most of them citing GI adverse events as the cause. 57.4% of those on the highest dose reported nausea (59% described it as mild, 34% as moderate, and 7% as severe). Dr. Nauck stated that Novo Nordisk would not carry the 1.6 mg dose forward to phase 3 due to the high rate of GI side effects. Serious adverse events (SAEs) were rare, and adverse events (AEs) increased with dose in both semaglutide and liraglutide. As expected, no major hypoglycemic events were reported, and  minor hypoglycemia was uncommon.

Questions and Answers

Q: Did nausea or GI side effects decrease over time?

A: Not that I’m aware of; it’s a good question and maybe we should look whether this was the case.

Q: Did you study the effects of oral contraceptives on the pharmacokinetics of semaglutide?

A: We did not study those effects; it was not the topic studied in this phase 2 trial.

Q: Any data on antibodies?

A: No but the molecule is very similar to liraglutide and is very much like GLP-1 so I wouldn’t expect any antibody formation.

Q: I think this study nicely shows that getting to somewhat higher GLP-1 levels increases efficacy beyond what we see with use of currently available GLP-1 analogs today. Couldn’t one think about different methods of slowly ramping up dose initially to retain the beautiful A1c-lowering effects of the highest doses rather than going forward with low doses?

A: I think this is what should be done for phase 3; there is no hurry, and in principle this and other studies show that ramping up helps with GI side effects and so we’ll keep doing it carefully and slowly – that is  fine for treatment of a chronic disease.

Q: You have reported a significant proportion of withdrawal. When you presented the reduction in A1c, was it based on the ITT population or per-protocol population?

A: That was as assigned with last observation carried forward.



Ronnie Aronson, MD (LMC Diabetes & Endocrinology, Calgary, Canada)

This 24-week study examined the efficacy and safety of once-daily 20 µg lixisenatide in patients with longstanding type 2 diabetes inadequately controlled on basal insulin (with or without metformin). Patients were randomized to receive lixisenatide (n=329) or placebo (n=167) in a 2:1 fashion. Lixisenatide treatment brought about a significant reduction in A1c beyond placebo over the 24-week treatment period (0.74% vs. 0.38% from a baseline of 8.4%; p<0.001), and also led to significant improvements in two-hour postprandial glucose, glucose excursions, seven-point profiles, body weight, and basal insulin dose. Side effects with lixisenatide were mainly gastrointestinal; there was no significant increase in hypoglycemia versus placebo.

  • In this 24-week study, patients with longstanding type 2 diabetes inadequately controlled on basal insulin (±metformin) were randomized to receive lixisenatide 20 µg once daily (n=329) or placebo (n=167). Main inclusion criteria were: 1) type 2 diabetes known for at least one year; 2) the use of basal insulin at a stable dose (≥30 U/day ±20%) for at least three months; and 3) A1c between 7.0-10.0% at screening. Main exclusion criteria were fasting plasma glucose above 13.9 mmol/l (250 mg/dl) and BMI ≤20 kg/m2. No concurrent antidiabetic medications were permitted other than metformin at a stable dose of ≥1.5 g/day for at least three months. Patients were titrated in a two-step process (10 µg at baseline, 15 µg at one week, and 20 µg at the two-week mark). At baseline, patients had an average age of 57- 58, diabetes duration of 11 years, A1c of 8.4%, two-hour PPG of 16.1-16.5 mmol/l (290-297  mg/dl), fasting plasma glucose of 8.1 mmol/l (146 mg/dl), BMI of 31.9-32.6 kg/m2, and duration of insulin use of 1.7-1.8 years. Approximately 80% of patients were on metformin at baseline.
  • Patients receiving lixisenatide treatment experienced significant reductions in A1c, two-hour postprandial glucose, glucose excursions, seven-point profiles, body weight, and basal insulin dose versus placebo. Over the course of 24 weeks, patients on lixisenatide experienced a significantly greater reduction in A1c than patients on placebo (0.74% vs. 0.38%; p<0.001). A higher percentage of patients on lixisenatide achieved A1c targets of <7% and ≤6.5% than placebo – 28.3% of patients on lixisenatide and 12.0% of patients on placebo achieved an A1c of <7% (p<0.0001), while 14.5% of patients on lixisenatide and 3.8% of patients on placebo achieved an A1c ≤6.5% (p=0.0003). The lixisenatide arm also achieved a significant reduction in two-hour postprandial glucose versus placebo (5.5 mmol/l [99 mg/dl] vs. 1.7 mmol/l [31 mg/dl]; p<0.0001) and in glucose excursion (4.1 mmol [73.8 mg/dl] vs. 0.3 mmol/l [5.4 mg/dl]). In addition, patients on lixisenatide experienced a significant reduction in seven-point profile versus placebo (1.49 mmol/l [27 mg/dl] vs. 0.61 mmol/l [11 mg/dl]; p<0.0001). The lixisenatide arm achieved an average reduction in body weight of 1.3 kg (2.9 lbs), compared to a reduction of 0.1 kg (0.2 lbs) with placebo. Patients in the lixisenatide arm reduced their basal insulin dose an average of 5.6 units, versus 1.9 units with placebo (p=0.012).
  • During the study, patients on lixisenatide experienced a higher rate of gastrointestinal adverse events than those on placebo. Most nausea with lixisenatide occurred during the first three weeks of treatment; by the fourth week the frequency of nausea was reduced (please see table below).

Type of adverse event, n (%)

Basal insulin ±MET + lixisenatide  (n=328)

Basal insulin ±MET + placebo (n=167)


86 (26.2)

14 (8.4)


27 (8.2)

1 (0.6)


24 (7.3)

9 (5.4)


17 (5.2)

7 (4.2)

  • Dr. Aronson commented that the incidence of hypoglycemia was comparable between treatment arms. 27.7% of patients receiving lixisenatide experienced symptomatic hypoglycemia, compared to 21.6% of patients receiving placebo; 1.2% of patients receiving lixisenatide experienced severe symptomatic hypoglycemia, whereas no patients on placebo experienced severe symptomatic hypoglycemia.

Questions and Answers

Q: Looking at the meal test and your seven-point profile, lixisenatide seemed to bring  about the greatest reductions in postprandial glucose with the morning meal. Do you think it would be helpful to do the meal test after the evening meal, or do you think that it suggests lixisenatide should be a twice-daily injection?

A: Lixisenatide once daily and BID has been studied, and there was no significant difference between the two, and that’s why it’s being presented as a once-daily medication in all phase 3 studies. There also has been a study looking at dosing with the morning meal and with the evening meal versus placebo, and both had identical benefits. There is a study underway (or about to start) comparing morning meal dosing to  the largest meal of the day dosing, and that will be interesting to investigate in particular.

Q: Do you have any data on antibodies?

A: Across the phase 3 studies, about 20% were antibody positive at the end of the protocol. There were no differences in efficacy related to antibody positivity or negativity.

Q: It’s been suggested that lixisenatide increases heart rate. Do you have the heart rate and blood pressure data?

A: There was no difference in heart rate in this study, or in the phase 3 studies overall. There was a small reduction in systolic blood pressure of 2-3 mm Hg across the phase 3 program.

Q: There was one death in the study of a patient on lixisenatide. Could you comment on that please?

A: The death was related to sudden cardiac death, and not related to the study protocol or medication as far as investigators were aware.

Q: Did you see statistically significant differences in postprandial glucose after lunch and dinner as you saw after breakfast? (I’m assuming the data you showed us was for the breakfast meal.)

A: I can’t recall the exact statistical analysis. I think the standard errors of mean were sufficiently separated that it would be statistically significant, but I can’t say for certain.

Q: In this study, was basal insulin dosed in the morning or during the evening?

A: In this study, it was dosed in the morning.

Q: What about metformin?

A: Metformin was continued as before, most typically BID.



Hannele Yki-Jarvinen, MD, PhD (University of Helsinki, Helsinki, Finland)

This study evaluated the efficacy after 24 weeks and safety after 52 weeks of linagliptin as an add-on therapy to basal insulin, alone or in combination with metformin and/or pioglitazone. At the 24-week mark, patients on linagliptin achieved an average A1c reduction of 0.58%, compared to an average increase of 0.07% with placebo (p<0.0001), from a baseline of 8.3%. Linagliptin treatment was shown to be effective independent of renal function or type of insulin used. There were no significant differences in the rate of adverse events between groups; linagliptin was not associated with an increased risk of hypoglycemia.

  • In this study, patients with type 2 diabetes on basal insulin (alone or in combination with metformin and/or pioglitazone) were randomized to receive linagliptin 5 mg once daily (n=631) or placebo (n=630). Efficacy was assessed at 24 weeks, and safety at 52 weeks. For the first 24 weeks, patients remained on a stable insulin dose; following this period, basal insulin was titrated as appropriate. At baseline, patients were on average 60 years of age, with BMI of 31 kg/m2, A1c of 8.3%; over 80% had diabetes duration of five or more years. Those on linagliptin had average baseline fasting plasma glucose of 8.2 mmol/l (148 mg/dl) and insulin dose of 41.5 U/day, compared to 8.4 mmol/l (151 mg/dl) and 40.1 U/day for placebo. Approximately 75% of patients were on metformin monotherapy, 1% were on pioglitazone  therapy, and 7% on metformin plus pioglitazone at baseline.
  • At the 24-week mark, patients on linagliptin achieved an average A1c reduction of 0.58%, compared to an average increase of 0.07% with placebo (p<0.0001). Linagliptin treatment was shown to be effective independent of renal function or type of insulin used. Patients on linagliptin experienced a slight reduction in body weight (<0.2 kg [<0.4 lbs]) at the 24-week mark, compared to a slight increase (<0.2 kg [<0.4 lbs]) with placebo.

Questions and Answers

Q: Was basal insulin titrated during the study? My impression was that only a small proportion of patients reached an A1c below 7%.

A: The ideal study in a way would be to first titrate basal insulin properly, and then add linagliptin. In the first 24 weeks, we did not allow for any changes in insulin dose.

Q: During the second period of the study, how was insulin titrated?

A: The instructions were to titrate to fasting plasma glucose below 6.0 mmol/l by increasing the basal insulin dose, but I don’t think that this really happened, at least judging from the results.

Q: In the linagliptin group, patients used less insulin, yet the hypoglycemia rate was similar to the placebo group. How do you explain that? One would expect less hypoglycemia in the linagliptin group.

A: The insulin dose was the same – it was 40 units at baseline. During the first 24 weeks of the study, the insulin dose did not change; it was similar between groups. The similar hypoglycemia rates observed is consistent with the idea that linagliptin does not increase the risk of hypoglycemia.

Q: Did you measure serum amylase on a regular basis during the study? Was there any change in the risk of pancreatitis in the two groups?

A: There were no hints of pancreatitis being more common in the linagliptin or placebo group. We didn’t measure amylase.


Leigh MacConnell, PhD (Amylin Pharmaceuticals, San Diego, CA)

Dr. MacConnell presented four-year results of the DURATION-1 follow-up study for exenatide once weekly – the longest-term data we have seen to date. The original DURATION-1 study compared exenatide once weekly (Bydureon) to exenatide twice daily (Byetta) over 30 weeks. After trial completion, all patients were put on Bydureon. Of the original 295 patients in the intent-to-treat population (ITT), 176 (60%) completed the four-year extension. At the end of four years, patients sustained an A1c reduction of 1.7% from a baseline of 8.2% (generally similar to the 1.9% reduction observed at the end of 30-weeks for DURATION-1; notably, Dr. MacConnell stated during Q&A that the A1c reduction in the ITT population using the last observation carried forward was 1.4% and the A1c reduction for patients who did not complete the trial was 1.1%). With regard to weight, patients sustained, on average, a 2.5 kg (5.5 lbs) weight loss from baseline after four years, compared to the ~4 kg (8.8 lbs) weight loss observed after the 30 weeks of DURATION-1. Safety analyses were performed on the ITT population; during four years 20% of patients experienced a serious adverse event (SAE), and three patients died. Dr. MacConnell stated that the causes of death were unrelated to treatment,  and no pattern of SAEs was reported. No major hypoglycemia episodes were reported, and minor hypoglycemia only increased over four years in patients taking a concomitant sulfonylurea (n=116). Improvements in blood pressure, total cholesterol, LDL, and triglycerides were also observed. Overall, we believe this represents strong evidence of Bydureon’s durability in terms of efficacy and safety – four years is much longer than the average patient expects to take a given compound and expect it to work.

  • The four-year DURATION-1 follow-up study examined long-term safety and efficacy of exenatide once weekly (EQW). The original DURATION-1 study compared exenatide once weekly (Bydureon) to exenatide twice daily (Byetta) over 30 weeks. After trial completion, all patients were put on Bydureon. Of the original 295 patients enrolled in the study, 176 (60%) completed the four-year extension. Most withdrawals were due to “withdrawal of consent” (n=62; 21%) with no reason specified; 10% (n=29) withdrew due to an adverse event, and Dr.   MacConnell stated that there was no clear trend in the type of adverse event leading to  withdrawal. Baseline characteristics between the ITT and four-year completer populations were consistent (average age of 54 years, A1c 8.2%, BMI 35 kg/m2, diabetes duration of 7 years, and the majority on a background of metformin [33%] or a sulfonylurea ± metformin or TZD [39%]). Efficacy data presented were for the four-year completer population, and safety data presented were for the intent-to-treat (ITT) population.
  • Patients sustained an average 1.7% A1c reduction from baseline after four years of treatment. FPG at four years was reduced by 2.1 mmol/l (~38 mg/dl) from baseline, compared to the 46 mg/dl FPG drop observed in DURATION-1. A1c and FPG plotted every year over the four years show that most of the rise in A1c and FPG between the end of the trial and the end of the four-year extension occurred from year one to two, with a slight upward trend from year two to three, and stabilization from year three to four. The majority of patients (55%) achieved the ADA recommendation of A1c <7%, and 36% achieved the AACE guideline of A1c <6.5%. At year four, patients also demonstrated improvements in HOMA-B and HOMA-S from baseline (26% and 13% improvement, respectively). Dr. MacConnell acknowledged during Q&A that the completer population is likely biased toward patients achieving better results on Bydureon (she stated during Q&A that the A1c reduction in the ITT population using the last observation carried forward was 1.4%, and the A1c reduction for patients that did not complete the trial was 1.1%).
  • Weight loss was fairly durable; after four years, patients sustained an average 2.5 kg (5.5 lb) weight loss from baseline, compared to the ~4 kg (8.8 lb) weight loss observed in the first 30 weeks of DURATION-1. However, weight tends to increase with age anyway, and without a control group, it is hard to say whether these patients would have gained more weight had they not been taking Bydureon. After four years, 71% of patients had lost weight from baseline, and 61% saw both a weight reduction and A1c reduction.
  • No unexpected safety concerns arose over four years. No pattern of serious adverse events (SAEs) was reported; 20% of patients experienced a SAE, and three patients died. Dr. MacConnell stated that the causes of death were unrelated to treatment. Withdrawal rate due to adverse events was low (8%); 2% withdrew due to GI adverse events. Hypoglycemia was nearly nonexistent in patients who were not using a concomitant sulfonylurea (SFU; n=179) and remained stable over four years in this group. In patients using a concomitant SFU (n=116), incidence of minor hypoglycemia increased slightly during the first year (~20% of patients experienced a minor hypoglycemic episode during the first year). No major hypoglycemic events were reported over four years. The most common side effects reported were nausea and injection site irritation, as expected (15 and 6 per 100 patient years, respectively), but they largely subsided after the first 20 weeks, and Dr. MacConnell indicated they were mild or transient.
  • Cardiometabolic risk factors also improved after four years. Improvements in blood pressure were observed (no significant difference in systolic blood pressure, but a 2.7 mmHg reduction in diastolic blood pressure was recorded); improvements were most marked in those with abnormal baseline blood pressure (systolic and diastolic reductions of 0.7 mmHg and 5.3 mmHg from a baseline of 140 mmHg and 82 mmHg, respectively). Total cholesterol, LDL, and triglycerides also improved. Maximum improvement in all of the aforementioned cardiometabolic risk factors was observed at two years and remained stable thereafter.

Questions and Answers

Q: I am most interested in the time course of what happened to A1c. It appeared that durability did not set in until after year three. Does it take three years to achieve durability, or was it just the selection process and those that stuck with it after three years had a durable effect?

A: Certainly a limitation to this open-ended open-label assessment is that of course it doesn’t have a control arm. I think there is also bias in that patients in the completer population are enjoying their experience with exenatide once weekly.

Q: What additional therapies were added over the four years?

A: The protocol required they maintain their stable doses, but over the long term there was more fluctuation in therapy. 20% of patients either added or increased dosage of medications, but about 20% also dropped a medicine or decreased their dose.

Q: Did blood glucose control deteriorate in the patients that withdrew? What would the analysis look like if they had been included?

A: We have an ITT analysis using last observation carried forward. In that population, the A1c reduction was 1.4% vs. 1.7% in the completer population; we also looked specifically at those patients that did withdraw early and looked at A1c upon withdrawal, and they saw an about 1.1% reduction. So patients terminated early aren’t seeing as robust of an effect as those remaining on therapy, which was not surprising.

Q: Were there any heart rate data?

A: We see an increase in heart rate of about three to four beats per minute. It happens fairly early on and remains stable thereafter so there is no continued increase.

Q: Were patients on concomitant blood pressure or antihypertensive agents?

A: The majority of patients were on lipid-lowering drugs or antihypertensive agents. Most remained stable in their dosage and about 10% of patients increased or added lipid-lowering or antihypertensive agents.


Tadej Battelino, MD, PhD (University Children’s Hospital Ljubljana, Ljubljana, Slovenia)

Dr. Tadej Battelino detailed his group’s dose escalation study of liraglutide in pediatric patients with  type 2 diabetes. He began by noting that though the prevalence of type 2 diabetes in children is increasing in the US, Asia, and Europe, metformin and insulin remain the only two treatments available to these patient populations. His group’s five-week trial assessed the safety, tolerability, and PK/PD profile of liraglutide in 21 pediatric participants (baseline age 14-16 years, A1c 7.8-8.3%, and average BMI 40 kg/m2). The study randomized seven participants to placebo and 14 to ascending doses of liraglutide (starting at 0.3 mg and increasing by 0.3 mg/week until reaching 1.8 mg in the fifth week; dose escalation was based on tolerability and levels of fasting plasma glucose). Adverse events were mild with liraglutide and mild/moderate with placebo, with no observed serious adverse events. Hypoglycemia was uncommon in both study groups and no major hypoglycemic episodes were reported. Participants did not experience any clinically significant shifts in calcitonin levels, and an external hormonal safety board found no evidence of hormonal disruption. The PK profiles for liraglutide in pediatric patients were similar to those reported in adults with type 2 diabetes. Interestingly, liraglutide provided a significant reduction in A1c (-0.86%) compared to a slight A1c increase with placebo (+0.04%), though no statistically significant effects on fasting plasma glucose or body weight were observed.

Questions and Answers

Q: Were you not a little surprised that you didn’t see a difference with respect to body weight? I know it’s only a five-week study and patients were only on liraglutide 1.8 mg for a short period of time, and not all patients reached that dose. But were you surprised?

A: Yes, we were surprised. I think for good reason. I believe a bigger study of a longer duration, especially in an obese population, is needed in this age group because this could be a very important indication.

Q: Can you explain why you saw an important difference in A1c but no difference in fasting plasma glucose?

A: I appreciate that this looks difficult, however, as you know, in early onset type 2 diabetes, it’s mainly post-prandial glucose that contributes to A1c rather than fasting plasma glucose. So if we understand the natural course of type 2 diabetes and take into account the very young age of this population, I don’t think it’s very surprising.

Q: I think to prove safety you need a considerably larger group of patients.

A: There’s no question about it. It’s not proof that liraglutide is very safe. We wanted to start in a small patient group to see if there’s a major problem. There was no major problem, so a larger phase 3 trial is ongoing in Europe.

Q: Did you see any changes in pulse or blood pressure?

A: Nothing major. The truth is, that higher dose was used only for one week, and if you combine the three highest doses, patients were on them for only three weeks. We plan to plot individual graphs, but there is no major issue.

Oral Presentation: Novel Therapies (Incretins)


Robert Henry, MD (University of California, San Diego, California)

Dr. Robert Henry presented new phase 2 data for Intarcia’s extended-release-exenatide device, ITCA 650. Dr. Henry opened by reasoning that presently, poor adherence to GLP-1 therapy is hindering patients’ ability to optimize glucose control. ITCA 650 is a miniature osmotic pump system that delivers exenatide subcutaneously, eliminating the need for self-injection. A phase 2 dose-ranging study found that ITCA 650 provided better A1c reduction than the equivalent total daily dose of exenatide twice daily over 12 weeks; higher doses of ITCA 650 conferred even greater A1c and weight benefits. Additionally, ITCA 650 produced a lower rate and shorter duration of nausea than the equivalent dose of exenatide twice daily. Based on the balance between each dose’s A1c reduction, weight loss, and tolerability, data supported the use of 20 µg/day as the initial dose and increasing to a final dose of 60 µg/day as the regimen for phase 3 trials (expected to begin in 1Q13). We hope that, if brought to market, ITCA 650 could help improve adherence to GLP-1 treatment to improve patient diabetes outcomes in the long run.

  • Dr. Henry opened by reasoning that presently, poor adherence to GLP-1 therapy is hindering patients’ ability to optimize glucose control. All currently approved GLP-1 agonists (Novo Nordisk’s Victoza, BMS/AZ/Amylin’s Bydureon, BMS/AZ/Amylin’s Byetta), and all late-stage candidates for that matter, require self-injection on a regular basis. Dr. Henry stated that adherence to GLP-1 therapy for patients with type 2 diabetes ranges from 30-70%, and that this has relegated this promising class to a third- or fourth-line position for treatment. He believes that reducing the burden of injections will lead to earlier and broader use along with better patient outcomes. ITCA 650 is a miniature osmotic pump system that delivers exenatide subcutaneously, eliminating the barrier of self-injection. The device is about the size of a matchstick and would be implanted subcutaneously. In in vitro tests, it demonstrated the ability to release fluid at a constant rate for 12 months.
  • A phase 2 dose-ranging study found that ITCA 650 provided better A1c reduction than the equivalent total daily dose of exenatide twice daily; higher doses of ITCA 650 conferred even greater A1c and weight benefits. The trial consisted of two stages. In the first stage, 155 patients were randomized to exenatide twice daily (n=53; total of 20 μg/day), ITCA 650 20 μg/day (n=51), or ITCA 650 40 μg/day (n=51) (patients had average baseline A1c’s  of 8.0%, 7.9%, and 8.0%, respectively in the three treatment arms). After 12 weeks, exenatide  twice daily produced a 0.7% A1c reduction, and both ITCA 650 20 μg/day and 40 μg/day  produced a 1.0% A1c reduction. During stage two of the study, patients remained on the same  dose of ITCA 650, or had the dose increased by 40 μg/day, and patients on exenatide were switched to ITCA 40 μg/day or 60 μg/day (overall, n=20 for 20 μg/day, n=42 for 40 μg/day,  n=46 for 60 μg/day, and n=23 for 80 μg/day). Those on 20 or 40 μg/day achieved 0.9% and 1.0% A1c reductions, respectively, and those on 60 or 80 μg/day achieved 1.3% and 1.4% reductions, respectively. A1c lowering was durable out to week 48. Weight loss on the 20 μg/day dose was 0.8 kg (1.8 lbs) at week 24, and for all higher doses was greater than 3 kg (6.6 lbs). Weight loss at for exenatide at week 12 was not disclosed; as we recall, it was around 3-4 kg at about a year.
  • Additionally, ITCA 650 produced a lower rate and shorter duration of nausea than an equivalent dose of exenatide twice daily. The incidence of nausea was consistently lower for ITCA 650 20 μg/day compared to 20 μg/day of exenatide twice daily; the mean duration of nausea on ITCA 650 20 μg/day was 17 days, versus 47.7 days for exenatide injections. Another testament to ITCA 650’s favorable tolerability profile was the 93% completion rate for stage one and 95% completion rate for stage two of the trial; 89% of patients in the exenatide arm   completed stage one of the study, which is also very high. In general, though this is great to see,   we aren’t typically successful forecasting adherence or patient interest from trials since there is certainly an unrepresentative cohort of patients that are “early adopters” – too, some patients participate in trials because they are paid.
  • The 20 to → 60 μg/day dose sequence was chosen as the best regimen to use in phase 3 trials. Since the 80 μg/day did not outperform the 60 μg/day dose and, as expected, exhibited worse tolerability, it was removed from consideration. The 20 μg/day dose was selected as the introductory dose given, its good initial efficacy and favorable tolerability. The 60 μg/day dose exhibited generally similar average A1c reduction to the 40 μg/day dose, but conferred  better A1c reductions for those with higher baseline A1cs, and allowed a larger percentage of patients to achieve targets of ≤7.0% or ≤6.5%. Phase 3 trials investigating the use of ITCA 650 for six-to-12 months will start in 1Q13. We look forward to hearing details on physician comfort with explanting the device; implanting, as we understand it, has been very straightforward.

Questions and Answers

Q: Were antibodies measured during these studies?

A: Levels were about 30%. They produced no effect on glycemic reduction.

Q: Were they neutralizing antibodies?

A: They were not neutralizing.

Q: What was the incidence of pancreatitis?

A: There was no incidence of pancreatitis at all.



Ira Gantz, MD (Merck Sharp & Dohme Corp, Whitehouse Station, NJ)

Dr. Ira Gantz revealed Merck’s phase 2b data for the once-weekly DPP-4 inhibitor MK-3102. As we understand it, this is not actually “Januvia” once weekly though we assume it is something similar to Januvia. The study randomized 685 patients with type 2 diabetes to placebo or MK-3102 0.25 mg, 1 mg, 3 mg, 10 mg, or 25 mg. Over 12 weeks, all five doses of MK-3102 provided significant reductions in A1c (greatest placebo-adjusted change of -0.71%), as well as significant improvements in 2-hour post-meal glucose (greatest placebo-adjusted change of -45 mg/dl) and fasting plasma glucose (placebo-adjusted change of -21.6 mg/dl; full data in table below). No meaningful change in body weight was observed with any dose of MK-3102 relative to baseline. Regarding safety, similar rates of adverse events and hypoglycemia were observed across all six treatment groups, and those taking MK-3102 exhibited no significant changes in heart rate or blood pressure. Based on these encouraging results, Merck recently announced that it is advancing MK-3102 (at the 25 mg dose) into phase 3. While a once-weekly oral should certainly benefit patients from an adherence perspective, we point out that most patients would still likely take other orals at some point that would likely be dosed at least once daily; that said, for newly diagnosed patients, a once-weekly, with its adherence advantages, could serve them very well. From a payer perspective, of course, most will probably have to try metformin first although ultimately, according to Dr. Eric Topol (Scripps, San Diego CA),about 22 percent of type 2 patients do not respond  to metformin. Many patients already take fixed dose combinations of DPP-4 inhibitors and metformin,  so adding MK-3102 would actually be adding one pill and one co-pay. That said, there would likely be adherence advantages for many that have problems remembering to take the fixed dose combination  and again, this prescribed as a very early therapy might work very well. We would also be interested in how it does with people with pre-diabetes.

  • MK-3102 is an orally administered, highly-selective (IC50 = 1.6 nM) DPP-4 inhibitor with a half-life long enough to support once-weekly dosing (t1/2 of ~63 hours for the 25 mg dose). Eliminated mainly through renal excretion, the drug will likely require no dose reduction in patients with mild or moderate renal insufficiency, though like most once-daily DPP- 4 inhibitors, MK-3102 will likely be given at a lower dose in patients with severe renal impairment (as a reminder, linagliptin [Lilly/BI’s Tradjenta] is the only DPP-4 inhibitor that does not require dose adjustment in this patient subpopulation). Dr. Gantz also emphasized that because MK-3102 neither inhibits nor induces cytochrome P450 enzymes, drug-drug interactions are not  anticipated.
  • The double blind, phase 2b study randomized 685 patients with type 2 diabetes to placebo or one of five doses of MK-3102 (0.25 mg, 1 mg, 3 mg, 10mg, or 25 mg) for 12 weeks on a background of metformin. Participants taking oral anti-diabetic drugs (OAD) prior to the trial underwent an eight-week wash-out period while those not on an OAD immediately began the single-blind placebo run-in period before randomization. Baseline characteristics were similar across all six treatment groups, with a mean baseline age of 54-56 years, BMI of 29-30 kg/m2, A1c of 7.9-8.1%, and duration of type 2 diabetes of three to five years. The study completion rate was similarly high between the six groups, ranging from 87% to 98%.
  • All doses of MK-3102 provided statistically significant (p <0.001) reductions in A1c compared to placebo. At all doses above 0.25 mg, A1c continued to decline throughout the 12- week treatment period. Three-point post-meal glucose tests revealed that all doses of MK-3102 provided statistically significant improvements in 2-hour post-meal glucose. A similar trend was observed for change in fasting plasma glucose (data provided below). No meaningful change in body weight was observed with any dose of MK-3102 relative to baseline. Dr. Gantz highlighted that the glycemic efficacy demonstrated by MK-3102 in this study was comparable to those reported for the once-daily DPP-4 inhibitors.


Outcome at Week 12



MK-3102 Doses





0.25 mg


1 mg


3 mg


10 mg


25 mg


Mean Change in A1c (%)














Placebo-adjusted  change














Mean Change in 2-hour PMG (mg/dl)














Placebo-adjusted  change














Mean Change in FPG (mg/dl)














Placebo-adjusted  change













  • All six treatment groups posted similar rates of adverse events (31-44%) and drug- related adverse events (5-8%), and no dose-dependent increase in the incidence of adverse events was observed. Only one serious drug-related adverse event was reported, in the MK-3102 25 mg treatment group. All MK-3102 treatment groups experienced a low incidence of hypoglycemia (0-2%) similar to that observed with placebo, and no episodes of severe hypoglycemia were reported during the study. Furthermore, those taking MK-3102 exhibited no significant changes in heart rate or blood pressure.

Questions and Answers

Q: What is the effect of renal impairment on the level of the drug in plasma?

A: A clinical pharmacology was done in patients with mild, moderate, or severe chronic renal failure and in patients on dialysis. And the effect is in patients with severe chronic renal insufficiency and dialysis.

There’s a modest increase in the AUC. So our plan is that in that specific population, we’ll reduce the dose of MK-3102 that we’re going forward with by 50%. That’s not because there are any adverse events associated with the greater AUC. It’s that we want to provide the same amount of coverage in those patients as in those without renal failure.

Q: What is your estimate regarding how many prescriptions are for monotherapy vs. for combination with metformin, because metformin is normally given twice daily. There are combination products, including those with DPP-4 inhibitors. How does that translate into the adherence advantage in those patients taking the combination?

A: I can’t quote you a specific prescription number, but as you know, not everyone is tolerant of metformin and in some patients, metformin is contraindicated. And as far as adherence, even if you’re taking metformin, those fixed dose combinations aren’t always available. And it might be an attractive option for a patient taking metformin to take MK-3102.


Jonathan Rosenblum, PhD (ActivX Biosciences, La Jolla, CA)

Dr. Jonathan Rosenblum presented results from in vitro and in vivo animal studies on KRP-104, a novel DPP-4 inhibitor, which Dr. Rosenblum stated ActivX is positioning as best-in-class by differentiating its functional selectivity and pharmacodynamic profile, as it has nearly 100% DPP-4 inhibition in humans.. Dr. Rosenblum noted that upon ingestion, KRP-104 is quickly converted into GIS-103, a DPP-4 inhibitor that cannot enter the intracellular space, in contrast to currently available DPP-4 inhibitors. Thus, because GIS-103 is confined within a small space, it has a relatively high potency due to its high maximum concentration. Additionally, since GIS-103 co-localizes with DPP-4 in the extracellular space, less administered drug is required to block DPP-4 action. Dr. Rosenblum thus argued that KRP-104  gives you “more bang for your buck” than its competitors. Enzymatic activity-based profiling assays revealed that GIS-103 inhibits almost no DPP-9 in an intact cell, while vildagliptin (Novartis’ Galvus) inhibits roughly 55% of DPP-9 when added to intact cells at its therapeutic Cmax (maximum concentration that a drug achieves). Similarly, after oral administration of KRP-104 to rats, the ratio of GIS-103 in tissues to plasma was significantly smaller than that of either vildagliptin or sitagliptin (see table below for more details). Though the data appears to indicate that KRP-104 has lower off-target action than currently marketed DPP-4 inhibitors. Dr. Rosenblum did not comment on whether this finding is clinically significant, even though KRP-104 has been studied in clinical trials. Phase 1 and phase 2 data indicate that KRP-104 provides comparable glycemic efficacy to other DPP-4 inhibitors; they have not at this stage shown better efficacy. Thus, most of the drug’s benefits, ActivX has described, seem to involve improvements in side effect profile (though we have not seen any data on differences in adverse event rates between KRP-104 and other DPP-4 inhibitors). If KRP-104 “differentiation” is characterized as primarily safety-related ,we are dubious how much of an impact this form of differentiation will have because safety is a minimal concern for most patients and providers using the therapeutic class. Notably, in a later conversation we had with Dr. Rosenblum he raised concerns about the long-term impact of potentially high levels of sitagliptin (Merck’s Januvia) in tissues, and that these drugs might not be as safe as they are commonly considered to be.

Tissue to Plasma Area Under Curve (AUC0-8) Ratios





















Oral Presentations: Mechanisms of Incretin Action


Carolina Solis-Herrera, PhD (University of Texas Health Science Center, San Antonio, TX )

Dr. Carolina Solis-Herrera’s crossover study investigated the mechanisms underlying the action of sitagliptin and metformin. The study included 16 people with type 2 diabetes (mean age 47 years, BMI 33.5 kg/m2, A1c 8.8%, and duration of diabetes 1.5 years). The participants underwent four treatment periods (six weeks on drug followed by a two-week washout period) during which they took metformin 2000 mg, sitagliptin 100 mg, sitagliptin/metformin combination therapy, or placebo in a randomized order. At the end of each six-week treatment period, the participants received a meal tolerance test (MTT) labeled with 14C-glucose, after which blood samples were taken regularly for six hours. Both sitagliptin and metformin significantly decreased post-meal plasma glucose and total glucose levels,   and sitagliptin/metformin combination therapy provided a larger reduction than either drug alone. The reduction in total glucose was not attributed to post-meal oral glucose levels, but rather to a decrease in endogenous glucose production, which was reduced with metformin and the combination therapy, and  to a lesser extent with sitagliptin. Notably, post-meal glucagon levels decreased by 26% with sitagliptin and 33% with the combination therapy compared to placebo, with no effect observed with metformin. In addition, sitagliptin and the combination therapy resulted in a ~2-fold increase in bioactive GLP-1 levels compared to placebo (metformin provided no change). Taken together, the results indicate that metformin and sitagliptin reduce post-meal glucose levels to the same extent, but through two different mechanisms: metformin suppresses hepatic glucose production while sitagliptin increases GLP-1 levels, thereby decreasing glucagon secretion.

Questions and Answers

Q: You mentioned that you measured bioactive GIP – do you have data on GIP in this experimental set?

A: We did measure GIP, but it increased in all treatment arms so we didn’t include the data.

Q: So there was no change in GIP between the treatment groups?

A: No, GIP increased equally in all the treatment arms – a four-fold increase.

Q: I think you described that sitagliptin alone lowered glucose and had a little bit of a stimulation effect on insulin secretion that wasn’t statistically significant. For the assessment, and based on what sitagliptin does to insulin secretion and the beta cell’s response to rising glucose, wouldn’t it make sense to calculate the ratio of insulin secretion to the rise in glucose? I would assume that with sitagliptin, you would get a significant rise in how beta cells responds to an increase in glucose.

A: We did calculate several indexes that we didn’t show here because we wanted to show most of the data we have. We did measure C-peptide and other parameters, as well as the ratio of insulin secretion to glucose, and we did find that they all improved. We found a slightly greater improvement with the combination therapy.


Oral Presentations: Metformin Action and Benefits


Hitoshi Kuwata (Kansai Electric Power Hospital, Osaka, Japan)

Dr. Hitoshi Kuwata investigated metformin’s effect on GLP-1 and GIP insulin secretion in 29 patients with type 2 diabetes after one-day administration of metformin. Patients had a mean A1c of 7.6%, mean BMI of 27.7 kg/m2 (which Dr. Kuwata indicated was a bit higher than average for Japanese patients with type 2 diabetes), and a mean disease duration of 3.1 years. Patients were given metformin 500 mg at lunch and at dinner on day 1 of the study and at breakfast on day 2 of the study. GLP-1, GIP, insulin secretion, C-peptide, and glucagon secretion were measured at 30 minute intervals after breakfast on day 2. Metformin significantly lowered postprandial glucose area under curve (AUC) by about 200 mM*min and increased total postprandial GLP-1 AUC vs. control (increase of about 1000 pM*min), but had no effect on postprandial insulin secretion, C-peptide, or total GIP AUC. Intact GLP-1 AUC was also elevated to almost twice that of the control with no decrease in DPP-4 AUC, indicating that the increase in intact GLP-1 was due to enhanced GLP-1 secretion and not inhibition of DPP-4 degradation. Dr. Kuwata also stated that metformin improved beta cell function, presumably drawing these conclusions from measures of C-peptide (though this was not explicitly stated or written on the slide), and that the improved beta cell function correlated with enhancement of GLP-1 secretion. Therefore, he concluded that metformin’s selective enhancement of GLP-1 secretion improved beta cell function, though we believe this may be an overstatement given that a correlation cannot prove causation. Furthermore, the improvement in beta cell function was not accompanied by an improvement in postprandial insulin secretion, so the GLP-1 increase may be clinically irrelevant. We would have liked to see measurements taken of other downstream indicators of GLP-1 action (e.g., glucagon secretion).


Symposium: GLP-1 Beyond the Pancreas


Mansoor Husain, MD (University of Toronto, Toronto, Canada)

In his fast-paced presentation, Dr. Mansoor Husain opened with a review of type 2 diabetes and cardiovascular (CV) disease before discussing the CV effects of GLP-1, DPP-4 inhibitors, and GLP-1 agonists in both animal and humans before ending with a brief overview of meta-analyses and CV outcome trials. He noted that animal models indicate that GLP-1 receptors are expressed in cardiac and vascular tissue. Furthermore, GLP-1 agonists have exerted cardiovascular and vasodilatory effects in isolated tissue, suggesting that their CV effects can occur independent of the central nervous system and of insulin. Both the absence or inhibition of DPP-4 improved survival in animal models of myocardial infarction. Dr. Husain then discussed several human studies and noted that GLP-1 agonists have been shown to improve blood pressure, endothelial function, and infarct size. While early results from human studies and meta-analyses appear positive, data from cardiovascular outcome studies are needed to confirm the effects of GLP-1 agonist therapy on cardiovascular disease.

  • Dr. Husain began his talk by showing data from 400 patients in UKPDS, which illustrated that higher A1c levels predict cardiovascular disease (Stratton et al., BMJ 2000). Dr. Husain noted that type 2 diabetes coexists with several risk factors of CVD (such as obesity and hypertension), and that hyperglycemia represents a robust biomarker for CVD. However, previous studies have found it difficult to demonstrate that lowering A1c using anti- diabetic therapies provides a reduction in macrovascular outcomes.
  • Dr. Husain then discussed the cardiovascular (CV) effects of GLP-1 in animal models. He first reminded the audience since only 10% of secreted GLP-1 reaches systemic circulation, researchers have begun to question whether and how GLP-1 can work at very low concentrations. Dr. Husain then reviewed prior studies, which demonstrated that GLP-1 receptors are present in cardiac and vascular tissue, and that GLP-1 and its metabolites protect the heart from ischemia reperfusion injury and increase coronary flow. He highlighted that studies in animals without a functional GLP-1 receptor have indicated that GLP-1’s CV effects are not  entirely dependent on the receptor.
  • Dr. Husain then turned to the CV effects of DPP-4 inhibition in experimental models. He first noted that a genetic mouse model lacking DPP-4 exhibited improved outcomes after experimental myocardial infarction (MI) compared to wild type mice (Sauve et al., Diabetes 2010). Previous studies have also shown that in mouse models of diabetes, the pharmacological inhibition of DPP-4 increases survival over untreated controls following MI.
  • Moving on to GLP-1 agonist therapy, Dr. Husain showed data illustrating a potential positive cardiovascular effect. In a mouse model of MI, pre-treatment with liraglutide reduced cardiac rupture, decreased infarct size, and increased survival through mechanisms independent of weight loss, but apparently dependent on the GLP-1 receptor. Before turning to human studies, Dr. Husain presented unpublished data from his lab showing that GLP-1 agonism appears to provide cardioprotection in a high-fat diet model of obesity and cardiomyopathy, potentially through several pathways.
  • Dr. Husain then reviewed several human studies to illustrate general effects of incretin therapy on cardiovascular-related outcomes. In addition to promoting weight loss, GLP-1 agonists have been shown to improve blood pressure, endothelial function, and lipids. Dr. Husain highlighted that GLP-1 agonists cause a potentially negative impact on heart rate, though he noted that the clinical significance of this effect is currently unclear: while an increase  in heart rate of roughly ten beats per minute has been correlated with cardiovascular morbidity, GLP-1 agonists such as liraglutide only produce an increase of two to three beats per minute. Dr. Husain stated that at the moment, physicians have not observed a rise in cardiovascular outcomes due to this increase in heart rate. On the contrary, studies have shown that exenatide reduces infarct size in patients with acute myocardial infarction (Lonborg et al., Eur Heart J 2011), and that GLP-1 agonists provide beneficial effects on max VO2, heart failure, and quality of life scores in patients with heart failure, though the study cohort was small (Sokos et al., J Cardiac Fail 2006).
  • Dr. Husain ended his talk with a brief overview of some meta-analyses. One such analysis included 40 studies of DPP-4 inhibitors and found a 31% reduction in certain hard cardiovascular endpoints. Another meta-analysis of roughly 6,500 people on GLP-1 agonist therapy found a 50% reduction in major adverse cardiovascular events cardiovascular outcomes relative to placebo (Monami et al., Ex Diabetes Res 2011).



Darleen Sandoval, PhD (University of Cincinnati, Cincinnati, OH)

Dr. Darleen Sandoval reviewed her preclinical findings on GLP-1’s role in regulating long-term energy balance in rodent models. She found that: 1) GLP-1 in the central nervous system (CNS) is necessary for regulating adiposity; 2) while peripheral GLP-1 (the GLP-1 produced in the intestine) acts as a neurotransmitter to control glucose homeostasis, it does not act in this way to regulate energy balance/total body weight; and 3) the GLP-1 receptor is not necessary for the full weight loss and insulin-sensitizing effects of vertical sleeve gastrectomy in rats. If preclinical findings about mechanism of GLP-1 action prove to be true in humans, perhaps an analog that produces less nausea or more weight loss could be developed.

  • GLP-1, aside from its well-known production in L cells and action on the pancreas, is also made in the brain and acts on receptors in the brain. It is produced in the brainstem and has receptors in the brainstem, hypothalamus, and amygdala. Using mouse models, Dr. Sandoval explored whether GLP-1 in the central nervous system (CNS) impacts long- term energy balance. This question was based on past research that had found a correlation between hindbrain pre-proglucagon (PPG; the precursor peptide to GLP-1) expression and fat mass.
  • Dr. Sandoval found that the GLP-1 receptor in the CNS was necessary for regulating adiposity. When Dr. Sandoval’s team selectively administered exendin-9 (a GLP-1 receptor antagonist) in the cerebral ventricle of mice, mice fed on a regular or high fat diet both gained roughly 2% more body weight after five weeks compared to those administered a saline control. Furthermore, selectively knocking out the GLP-1 receptor in the brain and feeding mice a high fat diet for three to four weeks resulted in mice with higher adiposity than wild type mice (~28% vs. ~20% adiposity).
  • She then examined whether there was an interaction between brain and peripheral GLP-1 in regulating energy balance. She stated that it was unlikely that intestinal GLP-1 acts as a hormone on the brain because it is cleared from circulation so quickly by DPP-4 and the liver. But GLP-1 receptors are found on neurons in the portal vein, suggesting GLP-1 might act as a neurotransmitter after being secreted from intestinal L cells. Studies suggested that GLP-1 receptor activity in the portal vein was responsible for regulating glucose homeostasis, but not  food intake. She then hypothesized that intestinal GLP-1 receptors act to stimulate neurons in the intestine as a means of interaction between the intestinal and central nervous systems. In mice that only expressed PPG (the GLP-1 precursor) in the intestine (and not the brain), no reduction  in body weight was observed compared to the whole body PPG knockout, suggesting that  intestinal activity of GLP-1 was not sufficient for regulating body weight.
  • Finally she examined the impact of GLP-1 on bariatric surgery, demonstrating that  in a rat model, the GLP-1 receptor was not necessary for full effects of vertical sleeve gastrectomy. Percent change in body mass after surgery was the same in GLP-1 whole body knockout mice and wild-type mice. Additionally, post-meal insulin response was the same in both wild-type and knockout mice.



Philip Newsome, PhD (University of Birmingham, Birmingham, UK)

In his presentation, Dr. Philip Newsome gave a valuable overview of the prevalence and pathology of non-alcoholic fatty liver disease (NAFLD), as well as a review of preclinical and clinical evidence supporting the use of GLP-1 agonists as a potential therapy. NAFLD is present in roughly 30% of  western populations and is strongly associated with excess weight– 80-90% of people who are morbidly obese suffer from the disease. NAFLD is an independent risk factor for cardiovascular disease, and it is  of particular concern for people with type 2 diabetes, as nearly 9% of patients die from liver disease. After briefly reviewing the pathology of NAFLD, including the “three-hit hypothesis”, Dr. Newsome discussed preclinical data showing that GLP-1 agonists can reduce hepatic lipogenesis, potentially through multiple mechanisms, some independent of the GLP-1 receptor. In addition, results from the LEAD trial of liraglutide (studying liraglutide for weight loss) indicate that the GLP-1 agonist lowers  ALT (alanine aminotransferase; a marker of liver damage) in patients, without increasing the risk of side effects. Looking forward, Dr. Newsome hoped that additional data from ongoing and future clinical trials will clarify the therapeutic potential of GLP-1 agonists in treating NAFLD.

  • Dr. Newsome began his presentation by discussing the prevalence of non-alcoholic fatty liver disease (NAFLD). He noted that FLD is often asymptomatic, with patients having only trivially abnormal liver function tests. However, obesity is a significant driver of the disease; as childhood obesity increases, the growing prevalence of NAFLD in youth and young adults has become a topic of increasing concern. Dr. Newsome reminded the audience that NAFLD is a common disease that affects roughly 30% of western populations along with 80-90% of people who are morbidly obese. He cited a study from Hong Kong that found that over a quarter of the population (27%) has NAFLD (Wong et al., GUT 2011) and noted that 10-15% of the US population has the disease.
  • NAFLD is a predominant cause of negative cardiovascular outcomes, and FLD is an independent risk factor for cardiovascular disease. Because the risk of death increases drastically as the disease progresses, Dr. Newsome recommended that patients with early-stage NAFLD (such as simple steatosis) focus on preventing CV events while those with advanced NAFLD focus on maintaining liver health. NAFLD is a significant concern for people with type 2 diabetes, as nearly 9% of patients die from liver disease; males have a higher risk than females.
  • Dr. Newsome briefly reviewed the pathology of NAFLD, noting that rather than being harmful, the conversion of free fatty acids to triglycerides inside the liver may represent a protective response, while the accumulation of unconverted free fatty acids could be an important driver of the disease (Yamaguchi et al, Hepatol 2007). Previous studies indicate that free fatty acids act in multiple pathways that can mediate liver damage (Parekh et al., Gastroenterology 2007) and that require further study. Dr. Newsome then succinctly described the “three-hit hypothesis,” which posits that NAFLD results from three sequential events: 1) fat accumulation (steatosis), which makes the liver more susceptible to damage; 2) enhanced lipid peroxidation, which accelerates the generation of reactive oxygen species; and 3) inadequate hepatocyte regeneration (Dowman et al., QJM 2010)
  • Dr. Newsome highlighted the lack of effective treatments and believes that GLP-1 agonist therapy may be a favorable option for obese people with an elevated A1c (>7.5%). GLP-1 analogs have been shown to reduce hepatic lipogenesis in preclinical models of NAFLD (notwithstanding the limitations of such models). The mechanism responsible for GLP-1 agonists’ effects is not clear – the beneficial outcomes may result from GLP-1-mediated weight loss, improved insulin sensitivity, or through other pathways. Previous studies show that human hepatocytes express the GLP-1 receptor, suggesting that GLP-1 may act directly on hepatocytes (Gupta et al., Hepatology 2010). Stimulation of GLP-1 appears to activate the insulin-signaling cascade though, as Dr. Newsome noted, GLP-1 agonists may have other effects independent of the GLP-1 receptor. Additional studies show that GLP-1 agonists modulate hepatic lipogenesis by suppressing the expression of lipogenic enzymes and transcription factors in primary hepatocyte cultures, suggesting that GLP-1 has beneficial effects independent of insulin signaling (Ben- Shlomo et al., JHep 2012). Data also suggest that GLP-1 may have anti-inflammatory properties.
  • Dr. Newsome then turned to clinical data, noting that GLP-1 analogs have shown promise and safety in patient cohorts with NAFLD. While a meta-analysis of GLP-1 agonists found no clear effect of the drugs on liver enzyme activity (Vilsboll et al., BMJ 2012),  data from the LEAD trials of liraglutide (which study liraglutide for weight loss) showed a significant reduction in alanine aminotransferase (ALT) with liraglutide. Dr. Newsome also noted that liraglutide treatment led to no increase in side effects in people with abnormal liver function tests, a result he found reassuring. Currently, assessments of NAFLD and NASH (nonalcoholic steatohepatitis, an extreme form of NAFLD) in clinical trials are performed via liver biopsies, though Dr. Newsome is hopeful that non-invasive procedures are imminent.


44th Claude Bernard Lecture


Daniel Drucker, MD (University of Toronto, Toronto, Canada)

The 44th Claude Bernard lecture, given by the inimitable Dr. Daniel Drucker, marked the end of a fantastic EASD 2012. Dr. Drucker comprehensively explored the roles of various proglucagon-derived peptides (including glucagon and GLP-1) based on his work in rodent models. He began by highlighting how wide-ranging the physiological effects of GLP-1 are. He then discussed his findings about, and potential clinical implications for, three peptides released from the cleavage of proglucagon: glucagon, GLP-1, and GLP-2. Glucagon antagonism has demonstrated robust glucose lowering capabilities and islet-enhancing effects in mice, but Dr. Drucker has also found that glucose plays important roles in  liver cell survival and liver lipid homeostasis. Therefore, any glucagon inhibiting effects must balance these positive and negative effects. He then discussed cardioprotective benefits of GLP-1 seen in mice studies and argued that we need more careful analysis to identify its mechanism of action. Finally, Dr. Drucker shared evidence of GLP-2’s role in controlling nutrient absorption and intestinal defense mechanisms. He ended his talk with a tribute to Claude Bernard and a moving dedication of the lecture to his parents.

  • The successful attenuation of glucagon to treat type 2 diabetes depends on understanding how to find the right balance when manipulating this system. While the importance of glucagon in glucose homeostasis is widely known, Dr. Drucker has also found an essential role of glucagon in cell survival and lipid homeostasis in the mouse liver.   Additionally, he has found that glucagon receptor signaling in the mouse liver attenuates pancreatic alpha cell proliferation; glucagon receptor knockout mice exhibit increased alpha cell proliferation and beta cell function. He proposed that if human alpha cell proliferation is similarly repressed in response to hepatic glucagon signaling, that one might use this principle to generate   a treatment for type 1 diabetes by using a glucagon receptor antagonist to promote alpha cell proliferation and then differentiate those extra alpha cells into beta cells. But since glucagon is necessary for hepatocyte survival and hepatic lipid homeostasis, one must be certain to balance  the glucose-lowering and islet-enhancing effects against potential liver harm when determining  the desired level of glucagon inhibition. Indeed, Dr. Drucker cited the glucagon receptor  antagonist MK-0893 as an example – in human trials it produced robust glucose lowering effects but increased levels of hepatic transaminases and lipids (it was subsequently dropped from development).
    • As a side note, Dr. Drucker’s proposed type 1 diabetes therapy detailed above seems like a very creative approach to us, but still of course very early stage at this point – glucagon receptor antagonists are currently in clinical development, but as we understand it, differentiation of alpha cells into beta cells would likely first necessitate de-differentiation into a precursor cell before it could be turned into a beta cell. With the complexity of both dedifferentiation and differentiation, we imagine meticulous coordination would be required.
  • Dr. Drucker then turned his attention to the hot topic of potential cardioprotective effects of GLP-1; he argued that we need more careful analysis to identify the mechanism by which GLP-1 produces such robust cardioprotection in mice (indeed this may help us understand if we should expect to see cardioprotection in humans). We know   that activating the GLP-1 receptor (GLP-1r) increases survival in murine myocardial infarction, improves cardiovascular output, reduces accumulation and migration of macrophages in blood vessels, and has vasodilatory effects. Dr. Drucker stated that we don’t yet understand the relative contributions of direct vs. indirect actions of GLP-1 on the heart and blood vessels, but know there are contributions from both. Some have attributed GLP-1’s cardiovascular effects to a direct mechanism of action on the heart because they have detected GLP-1r on cardiomyocytes.  However, Dr. Drucker challenged this notion, noting that his lab has only found the receptor on arterial cardiomyocytes, and not in ventricular cardiomyocytes, and that other labs have used antibodies that give false positive readings in detecting GLP-1r.
    • Dr. Drucker then analyzed the mechanisms behind GLP-1 secretion from L cells, demonstrating that progesterone activates GLP-1 secretion even without having to enter the cell. This could have exciting therapeutic implications if we could stimulate GLP-1 secretion from the gut lumen without having a drug enter systemic  circulation.
  • Finally, Dr. Drucker shared evidence of GLP-2’s role in controlling nutrient absorption and intestinal defense mechanisms. GLP-2 enhances intestinal blood flow, enhancing the absorption of nutrients. GLP-2 also regulates the mucosal inflammatory response in the intestine and bacterial populations. GLP-2 is being investigated in clinical trials for short bowel syndrome.


Corporate Symposium: Diabetes Care Today: Individualizing Treatment Options (Lilly Diabetes)


Tina Vilsboll, MD (University of Copenhagen, Copenhagen, Denmark)

Dr. Tina Vilsboll argued in favor of combining GLP-1 therapy with insulin. She reviewed their complementary effects: when used together, they produce an additive A1c reduction; GLP-1 offsets the weight gain associated with insulin; GLP-1 can reduce insulin dose; and GLP-1 can reduce  hypoglycemia associated with insulin use. Dr. Vilsboll also believes that GLP-1 agonists’ glucagon- suppressing effects are often under-emphasized. She stated that at least 50% of the benefit derived from GLP-1 agonist therapy is due to improved alpha cell function resulting in a reduction of glucagon secretion, as well as hepatic glucose production. She stated that patients with type 2 diabetes experience fasting hyperglucagonemia and inappropriate glucagon responses resulting in a doubling of post- prandial glucagon secretion compared to healthy people. The glucagon-suppressing effects of GLP-1 agonists stop when patients become hypoglycemic – the glycemic-dependent nature of GLP-1 is, of course, one of its biggest advantages..

  • Dr. Vilsboll highlighted the often under-emphasized effects of GLP-1 agonists on glucagon secretion, stating, “Glucagon is the neglected hormone” of type 2 diabetes. She noted that patients with type 2 diabetes have both fasting hyperglucagonemia and a doubling of the post-prandial glucagon response. GLP-1 suppresses glucagon secretion and most importantly, will not increase patients’ risk of hypoglycemia since the glucagon suppression stops once normoglycemia is reached. DPP-4 inhibitors, she said, also exhibit a glucagon effect.
  • Adding GLP-1 to existing insulin therapy or adding insulin therapy to existing GLP-1 therapy are both effective options. Dr. Vilsboll cited a Novo Nordisk trial that found that adding liraglutide to basal insulin plus metformin resulted in additional A1c reduction, no  increase in hypoglycemia, and only 1 kg (2.2 lbs) of weight gain after one year. Another trial demonstrated that adding exenatide to insulin glargine increased the number of patients meeting treatment targets (from 35% to 60%), produced a 3 kg (6.6 lbs) placebo-adjusted weight loss, did not increase the risk of hypoglycemia, and was insulin sparing. Dr. Vilsboll argued that data from clinical trials support the physiological and pharmacological rationale for combining GLP-1 and insulin therapy due to their complementary modes of action. While insulin is an effective and inevitably necessary treatment in most cases of type 2 diabetes, its efficacy is offset by weight gain (up to 4 kg [8.8 lbs] of weight gain, according to Dr. Vilsboll), and risk of hypoglycemia, which GLP-1 helps to mediate. Dr. Vilsboll also added that, in general, she favors earlier use of combination treatment, stating, “Why should we wait for one treatment to fail before moving on  to another?”



Peter Diem, MD (University of Bern, Bern, Switzerland); Rury Holman, MD (University of Oxford, Oxford, United Kingdom); Tina Vilsboll, MD (University of Copenhagen, Copenhagen,  Denmark)

Q: What are determinants of metformin failure?

Dr. Diem: There are certainly patients for whom metformin doesn’t work. I think often times metformin failure may be that patients do not like the GI side effects and just don’t use it. It’s difficult to define.

Q: What is the place of long-acting GLP-1 agonists?

Dr. Vilsboll: There are a lot of trials that have evaluated different GLP-1 analogs. One head to head comparison trial was LEAD-6 (which compared exenatide twice daily with liraglutide), and in that trial, liraglutide was more efficacious in A1c lowering. Recently the DURATION-6 trial compared liraglutide to exenatide once-weekly. There, to my surprise, the results favored liraglutide. One could say they did not use enough exenatide once-weekly in that trial. From DURATION-1, exenatide once-weekly had beautiful results. From head to head trials, it seems like once-daily is at least as good and maybe a bit better. But  the GLP-1 agonists are a bit different: exenatide has a human backbone and a bit of immunogenicity.

Q: Peter, do sulfonylureas [SFUs] harm beta cells? Should this be taken into account and may it affect success from future therapies?

Dr. Diem: The ADOPT trial clearly shows a rise in A1c following the initial drop with SFUs. Whether the beta cells are really harmed irreversibly, I’m not really sure. If you look at HOMA analysis in UKPDS patients on SFU, it was even higher. Currently we do not have a good way to assess beta cell function once patients are treated with different options. Most analyses I’ve seen are speculations of true function of  beta cells in theses situations.

Dr. Holman: A recent study in Diabetes Care from someone working on a triple therapy in oral agents found that there are no differences. It’s interesting [Editor’s note: we believe he is referring to Dr. Ralph DeFronzo’s long-awaited triple therapy data, which will appear first in journal format in several months, according to Dr. DeFronzo].

Dr. Vilsboll: In mouse islets, if you add SFU in vitro you actually increase islet cell apoptosis, but if you add GLP-1 to human islets in vitro you see a decrease in apoptosis.

Dr. Holman: If I were treating mice, then that would be compelling.

Q: Glucotoxicity improves with any glucose lowering agent, not just GLP-1.

Dr. Vilsboll: Point taken.

Q: What about use of metformin in pregnancy?

Dr. Diem: We don’t use it. We use insulin during pregnancy.

Dr. Vilsboll: But you use it until they become pregnant – it’s not that it’s really dangerous because you use it in women with PCOS and then you stop it when they become pregnant.

Dr. Holman: Though reasonable data exists showing that it’s safe in pregnancy now.

Q: Can you speak to the use of metformin post-liver transplant?

Dr. Diem: I’d look at the reason for the liver transplant. Depending on the reason for liver deterioration, I might not use it at all. I’m not aware of any data that you need to use stricter criteria for kidney function post-liver transplant. As long as liver function is okay, we should use it. In general, I would tend to use lower doses. I very rarely use more than two grams per day.

Dr. Holman: I think we need to use it cautiously, and the guidelines now allow for an eGFR down to 30 with no evidence of increased lactic acidosis.

Q: Does GLP-1 have action on extra-pancreatic glucagon?

Dr. Vilsboll: There are quite a few things we don’t know about glucagon. The intestine might actually secrete some glucagon. Right now we’re conducting a study with biopsies of the GI tract of type 2 diabetes and healthy subjects to see if L cells actually secrete intestinal glucagon. So that’s a hot and interesting topic I’m not capable of addressing right now.

Q: Some people on metformin and DPP-4 inhibitors get hypoglycemia – why does this still happen when both have a low risk of hypoglycemia?

Dr. Diem: The risk is not zero. Perhaps it is also related to the fact that counter regulation in type 2 is not normal.

Dr. Vilsboll: Even placebo can induce hypoglycemia, and it depends on how hypoglycemia is defined. Additionally, other things like alcohol may be involved. But in general, GLP-1 works in a strictly glucose dependent manner and doesn’t push patients into hypoglycemia.

[Note: additional talks from this corporate symposium can be found in the “Insulin and Insulin Therapies” section of this report]


Corporate Symposium: A Comprehensive Therapeutic Approach to Diabetes Management (Sponsored by Sanofi)


Michaela Diamant, MD, PhD (VU Medical Center, Amsterdam, The Netherlands)

Highlighting the differences between short- and long-acting agents, Dr. Diamant compared available and late-stage-investigational GLP-1 receptor agonists with regard to pharmacologic qualities and results of head-to-head trials. Long-acting GLP-1 receptor agonists (liraglutide, exenatide QW, albiglutide QW) have a more potent effect on fasting glucose, but short-acting agonists (exenatide BID, lixisenatide) more dramatically blunt post-meal spikes of glucose and glucagon. The greater postprandial effects of short-acting agents may reflect greater delay of gastric emptying: an effect that seems to be mediated by the vagus nerve (Veedfald et al., EASD 2012). With continuous stimulation of GLP-1, gastric emptying seems to re-acclimate and come back toward normal (Nauck et al., Diabetes 2011). Short-acting agents, by stimulating GLP-1 only intermittently, could lead to delayed gastric emptying that is more sustainable (though short-acting agents also lead to nausea that is more persistent than with long-acting agents). As for body-weight effects, Dr. Diamant suggested that these depend in part on the agonist’s ability to cross the blood-brain barrier. This hypothesis would explain why albiglutide QW (which is large due to binding with albumin) caused less weight loss than liraglutide in the Harmony 7 trial (Pratley et al., ADA 2012).



Julio Rosenstock, MD (Dallas Diabetes and Endocrine Center, Dallas, TX)

After explaining the rationale for using GLP-1 agonists in combination with basal insulin, Dr.  Rosenstock reviewed results from several clinical trials of said combination therapy, focusing on the use of lixisenatide with insulin glargine. He noted that earlier trials demonstrated that GLP-1 agonists could further enhance glycemic control when added on to basal insulin therapy, and vice versa – exenatide improved glycemic control and weight when added to insulin glargine (Buse et al., AIM 2011), and insulin detemir improved glycemic control (without weight gain) when added to liraglutide (Rosenstock et al., Diabetes 2011). Dr. Rosenstock reviewed the results of the GetGoal-L and GetGoal-L-Asia trials, emphasizing that lixisenatide provided robust reductions in two-hour postprandial glucose levels  (earlier in his presentation he pointed out that lixisenatide has a stronger postprandial effect than liraglutide), and conferred incremental A1c reductions in combination with insulin glargine. Finally, he noted that it could be possible to combine longer-acting GLP-1 agonists with basal insulin as well – one recent trial compared once-weekly albiglutide versus three-times-daily insulin lispro as add-ons to insulin glargine (Rosenstock et al., Diabetes 2012). Dr. Rosenstock hoped that people would not waste energy debating whether a GLP-1 agonist or basal insulin should be initiated first; regardless of which  is started first, eventually both will be needed. During the following panel discussion, he stated that as a rule of thumb, those with A1c above 8.5% should start on basal insulin, and those with A1c in the 7-8% range should start on a GLP-1 agonist first.



William Cefalu, MD (Louisiana State University School of Medicine, Baton Rouge, LA), Michaela Diamant, MD, PhD (Diabetes Center, VU University Medical Center, Amsterdam, The Netherlands), Julio Rosenstock, MD (Dallas Diabetes and Endocrine Center, Dallas, TX)

Dr. Cefalu: In your clinical practice, do you use basal insulin in combination with GLP-1 agonists?

Audience: 57% yes; 43% no.

Dr. Cefalu: What would be the preferred order of combination therapy for a patient with an A1c less than 8%?

Audience: 35% basal insulin followed by a GLP-1 agonist; 53% a GLP-1 agonist followed by basal insulin; 12% immediate combination therapy.

Dr. Cefalu: What would be the preferred order of combination therapy for a patient with an A1c greater than 8%?

Audience: 59% basal insulin followed by a GLP-1 agonist; 25% a GLP-1 agonist followed by basal insulin; 16% immediate combination therapy.

Dr. Rosenstock: I agree with this. Whatever you choose is fine. Two things are very critical. We know that insulin works for almost everyone. GLP-1 does not work for everyone – it’s not that effective for 25-30% of patients, and 25-40% of patients don’t tolerate them. Also, GLP-1 agonists are more expensive than  insulin. But, if you want to base it on A1c, I think that if someone is above 8.5%, basal insulin is the way to go. For an A1c between 7-8%, GLP-1 can do very well. For an A1c of 8.0-8.5%, do whatever you want.

Dr. Diamant: At this point, more patients are being started on basal insulin, so the studies you presented with add-on of GLP-1 receptor agonist reflect a practice that will be more common.

Dr. Rosenstock: Plus, as Dr. Gerstein said, we’ve been using insulin for 90 years. We know what to expect. We don’t know what will happen after 10-15 years of GLP-1 receptor agonist use.

Dr. Cefalu: With regard to the mechanism of postprandial control, you discussed effects of time course. Can you comment on the differences in short- and long-acting GLP-1 receptor agonists?

Dr. Diamant: In an infusion study, researchers looked at the effects on gastric emptying, insulin secretion, glucagon secretion, and peripheral glucose disposal. They made some estimates of relative contribution and calculated that gastric emptying was responsible for about 50% and glucagon for another 30%. If you slow down gastric emptying in a way that doesn’t vein off over time, that effects postprandial glucose. This would be for the short-acting GLP-1 receptor agonists: both exenatide BID and lixisenatide.

Dr. Cefalu: You’ve shown a lot on A1c response to GLP-1 receptor agonists. Could you sum up who will respond best and discuss in which situations you would initialize patients to combination therapy with insulin and GLP-1 receptor agonist?

Dr. Diamant: I think that predicting responders is the million-dollar question. The data vary. Some suggest that if you start earlier in the course of disease you have more response, but how do you assess response: A1c response at three months? A combination of body weight and A1c change at three months? Some of my patients have been on two shots of exenatide for eight years, but I don’t know if from this we can get data on what predicts response.

Dr. Rosenstock: In our 12-week study we got a sense of what determines response. In randomized controlled trials we give drug the drug to everyone and then in a randomized comparison to a control group. Maybe the way to go in the future is to try, as we do in clinical practice, choosing cutoffs for continuation. In clinic it’s “the-n-of-one trial” you try it in a patient for 12-24 weeks and assess the response based on some predetermined criteria. If there is a response, continue; if not, don’t continue it. Likewise in the UK, liraglutide has to show a reduction of 1.5% in A1c and 3% body weight over six months. If the patient gets a response the health system pays for liraglutide use; if not, they don’t.

Dr. Cefalu: Would GLP-1 receptor agonists be more effective following basal insulin optimization, once you’ve broken glucotoxicity?

Dr. Diamant: It makes sense, I think. I think most clinicians would like to optimize insulin first and then afterward see what can be done, for instance, to lower postprandial hyperglycemia.

Dr. Rosenstock: Everything works better after you optimize insulin; for example, insulin sensitivity is improved. Indeed, sometimes you hear about GLP-1 deficiency, but it is just dysfunction. Once you optimize insulin you improve native GLP-1 secretion.

Dr. Cefalu: For a patient with an A1c of 9.5%, do you think there is any benefit of starting GLP-1 and insulin therapy at the same time?

Dr. Diamant: We don’t have data on this. It’s definitely an interesting issue, because you could say, well, we are advocating to start combination therapy in these patients because they have such poor control. It needs to be studied.

Dr. Rosenstock: It is an exciting area. I myself am not fond of premix insulin. I know it causes more hypoglycemia and weight gain and so on. But a coformulation of basal insulin plus a GLP-1 could be a very exciting area. As we speak, there are trials ongoing, with degludec and liraglutide, and glargine in combination with lixisenatide.

Dr. Cefalu: Where do we go from here? What do you want to see?

Dr. Rosenstock: I think we need to move further and intervene much earlier. We need better and more consistent organization in studies, especially with regard to insulin optimization. I think that patients should self-titrate, up by one-to-two units per week. We need long-term studies to show durability of effect. We should design long-term studies to find the responders. Not all therapies are good for everyone.

Dr. Diamant: I agree. Head-to-head studies with basal-bolus vs. basal-GLP-1 receptor agonists would be valuable, provided they are designed well with regard to titration and of sufficiently long duration to convince reimbursement authorities. I think they should be two-to-three years long, with combined endpoints and low rates of hypoglycemia, to show cost-effectiveness. I think the design of studies is still a major issue.

[Note: additional talks from this corporate symposium can be found in the “Insulin and Insulin Therapies” and “Type 1 Diabetes Therapies” sections of this report]


Corporate Symposium: Delivering Innovation in Type 2 Diabetes – Tailored Approaches with SGLT-2 and Incretin-Based Therapies (Sponsored by BMS/AZ)


Edoardo Mannucci, MD (Careggi Teaching Hospital, Florence, Italy)

Dr. Edoardo Mannucci began the second session of the symposium, which was focused on  cardiovascular risk and incretins. When the audience was asked, “How much do you think the improvement of glucose control reduces CV risk?” A) no effect; B) marginal decrease in risk; and C) significant decrease in risk. A slight majority of the audience (56%) voted that improving glucose control significantly reduces CV risk, while 37% voted that it marginally decreases risk. Dr. Mannucci said this split voting pattern is not surprising and representative of the literature, which overall seems to point to a moderate improvement (about a 10% drop in major adverse cardiac events [MACE] with a 1% drop in A1c) in cardiovascular risk with better glucose control, but that there have been mixed results  (sometimes even within the same trial). While reviewing the impact different therapies have on cardiovascular risk Dr. Mannucci noted that in the Rosiglitazone Evaluated for Cardiac Outcomes and Regulation of Glycaemia in Diabetes (RECORD) trial (n=4,447) found that there were no significant differences between rosiglitazone and sulfonylureas (SFUs) for MACE or cardiovascular mortality. He rosiglitazone was removed from the market because it increased cardiovascular risk, causing us to question why sulfonylureas have not faced the same fate. On the positive side, he reported that DPP-4 inhibitors reduced MACE by about 30% and all cause mortality by 40% in phase 3 trials.



Petra-Maria Schumm-Draeger, MD, PhD (Clinic Munich Bogenhausen, Munich, Germany)

Dr. Petra-Maria Schumm-Draeger detailed the Saxagliptin Assessment of Vascular Outcomes Recorded in Patients with Diabetes Mellitus-Thrombosis in Myocardial Infarction (SAVOR-TIMI) 53 study being carried out by Bristol Myers Squibb and Astra Zeneca. The multicenter (523 sites), randomized, double- blind, placebo controlled, multinational (25 countries), phase 4 trial is comparing cardiovascular risk and outcomes of saxagliptin vs. a placebo, and has a primary endpoint that is a composite of cardiovascular death, nonfatal myocardial infarction, or nonfatal stroke. The study has completed enrollment (n≈16,500) of people with type 2 diabetes (A1c between 6.5-12.0%) who either have established cardiovascular disease or cardiovascular risk factors, and will run for about five years. Results are expected for mid 2014, which will make this one of the first CV outcomes studies to report. (Canagliflozin’s CANVAS trial will end in April 2013, and lixisenatide’s ELIXA trial in May 2014).



Kamlesh Khunti, MD, PhD (University of Leicester, Leicester, United Kingdom)Paola Fioretto, MD, PhD (University of Padova, Padova, Italy); Samy Hadjadj, MD, PhD (Poitiers University Hospital, Poitiers, France); Andreas Pfeiffer, MD (Charité University Hospital, Berlin, Germany); Edoardo Mannucci, MD (Careggi Teaching Hospital, Florence, Italy); Laurie Baggio, PhD (Samuel Lunenfeld Research Institute, Toronto, Canada); Petra-Maria Schumm-Draeger, MD, PhD (Clinic Munich Bogenhausen, Munich, Germany); Jiten Vora, MD (Royal Liverpool University Hospital, Liverpool, United Kingdom)

Q: If there is a 40% reduction in all-cause mortality for people on DPP-4 inhibitors, shouldn’t we stop ongoing studies comparing DPP-4 inhibitors to placebo?

Dr. Mannucci: The data I presented was showing reduced mortality in the first year; we don’t know what happens after fiver or ten years. So we need a longer study to determine that.

Q: For the SAVOR-TIMI 53 study, do you think having an A1c range from 6.5% to 12% is too large and could result in bias?

Dr. Schumm-Draeger: It is true that this is a large range. I think we have to include this whole spectrum in order to get a full answer on cardiovascular risk. It is a large study so that it will be representative for well and not well patients.

Dr. Mannucci: If this trial had a metabolic endpoint then this range would have been too wide, but what we are looking at here is cardiovascular outcomes. It is correct to have a range as wide a range as possible to have it be applicable to as big a population as possible.

Dr. Khunti: We can’t just go for tight A1c targets for everyone, and this study will give us more information on what patients – and A1cs – should be taking a DPP-4 inhibitor.

Dr. Vora: Looking into the future do you think we will see differences between DPP-4 inhibitors and GLP-1 agonists in terms of cardiovascular risk?

Dr. Baggio: That is a very good question. Based on the preclinical data and the ongoing clinical trials I think there will be differences. One thing that sticks out in my mind is that with some of the GLP-1 agonists there were reports of small increases in heart rate, but to my knowledge DPP-4 inhibitors have not been associated with any increase in heart rate. So I do think we will see some difference. Also GLP-1 agonists works through the GLP-1 receptor but the DPP-4 inhibitors have a number of potential substrates. The problem is they are hard to measure in vivo. It is not clear if they will be physiologically relevant, we just have to wait for the studies.

Q: What do you think of using dapagliflozin in type 1 diabetes patients who may become pregnant?

Dr. Fioretto: I think that dapagliflozin can be used in type 1 patients, particularly patients who are overweight. We have emphasized that the mechanism of action is totally insulin independent. As far as pregnancy is concerned, I do not know the data on that so I cannot address that question.

Q: Could you please elaborate on the clinical importance of the 2-3 kg [4.4-6.6 lbs.] weight loss? If you are continuously losing glucose in your urine why do you not continue to lose weight?

Dr. Hadjadj: Even if losing two to three kg does not do a lot in terms of medical practice it does do a lot in terms of the relationship you have with your patient. Any type 2 diabetes patient is involved in some body weight control program. If we can move their weight in the right direction and give them that motivation then that is probably more important than any clinical benefit.

Dr. Fioretto: I think that losing several kilos is hugely important for motivation. Getting sustained weight loss like this is not easy in people with type 2 diabetes.

Dr. Vora: Do you know the epidemiological benefits of 3 kg [6.6 lbs.] of weight loss?

Dr. Mannucci: Losing 5% of your body weight produces clinically relevant improvements in glycemic control and many other factors. The difference between the weight loss on dapagliflozin and the weight gain on an SFU is five kg and that is more than 5% of the body weight of most of our patients.

Dr. Pfeiffer: Why do they not continue to lose weight? Is it because they eat more? It needs more studying.

Dr. Fioretto: There is not a clear answer to that. I think that the glucosuria is not the only reason for the weight loss. We hypothesize that patients get reset and eat more. But there may also be something going on at the liver level. I think we need to do more studies to find out why they do not continue to lose weight.

Dr. Pfeiffer: Do you think that the ADA/EASD guidelines are realistic? They say we have to put our patient in the drivers seat. So do you think we can really do this?

Dr. Mannucci: I don’t find it unrealistic. What we are really missing right now are studies with good subgroup analysis.

Dr. Vora: What do you think the guidelines are based on?

Dr. Mannucci: For the personalization guides, I would say nothing. For the other parts I would say expert opinion.


Corporate Symposium: Asking The Tough Questions in Type 2 Diabetes Treatment! (Sponsored by Boehringer Ingelheim / Lilly Diabetes)


The intention of this session was to challenge conventional wisdom – particularly in the earlier use of DPP-4s. The key messages appeared to be that DPP-4 inhibitors should be used as second line therapy after metformin instead of sulfonylureas (SU), and that linagliptin (BI/Lilly’s Tradjenta) offered an advantage over other DPP-4s because it can be used in patients with chronic kidney disease, whereas all other DPP-4s are contraindicated in those with renal impairment. The audience seemed to be aligned with the first point – in a later poll, 58% of them said they already prescribed DPP-4 after metformin compared to 28% preferring a SU.

Perspective 1: Optimizing Glycemic Control –Putting the Guidelines and Latest Evidence in Context

Melanie Davies, MD (University of Leicester, UK)

  • Evidence from large-scale outcomes trials generally promotes maintaining a low A1c while minimizing hypoglycemia and weight gain. The UKPDS showed that intensive therapy leads to better outcomes (both micro- and macro-vascular). But the ACCORD trial led to some confusion. Even though the intensive arm successfully reached an A1c of 6.4%, there was increased mortality. In the ADVANCE trial, the intensive group also reached 6.4% A1c, but with no detrimental effects. The difference was hypoglycemia and weight gain – which should ideally be minimized.
  • The ADA/EASD guidelines have been recently changed to allow many options of second line therapy after metformin, including insulin and the incretin based therapies. We should also be taking individual patient characteristics more into account.
  • A large meta-analysis of metformin showed no reduction in all-cause mortality or cardiovascular mortality, so the evidence base for its use as a platform is not clear- cut. There is also no particular mortality benefit with metformin in combination with insulin. On the other hand, DPP-4s showed A1c reduction with little adverse effects. However, we need more evidence to establish DPP-4s as first line therapy.


Perspective 2: The Effect of Patient Characteristics on Treatment Outcomes

Brian M Frier, MD (University of Edinburgh, Scotland);

  • Dr. Brian Frier reiterated that there are many factors that should influence therapy choices for patients. These include: A1c targets, age, duration of diabetes, hypoglycemia, renal function, co-morbidities, psycho-social and cognitive status, and socio-economic status.
  • Very strict control is not appropriate for certain groups. The only major intensive therapy trial that started close to the date of diagnosis is UKPDS – the others start around a decade after diagnosis – so we can’t generalize their results to all patients. It seems that strict control is good for patients in the early stages in diabetes, but could even be harmful in the more advanced stages, depending on your interpretation of ACCORD.
  • Hypoglycemia is probably underestimated as a cause of cardiovascular mortality in type 2 diabetes. Hypoglycemia is also related to fracture risk (from falls) in older women. The risk of hypoglycemia increases greatly with chronic kidney disease (CKD). Metformin, SUs and insulin should be discontinued in CKD.
  • Linagliptin (Tradjenta, BI/Lilly) is the only diabetes drug that is suitable for patients on dialysis, and only pioglitazone (Takeda’s Actos) and linagliptin are suitable for patients with eGFR <30 mL/min/1.73 m2.
  • Intensive therapy is contraindicated in patients with advanced age, limited life expectancy, a less motivated attitude, higher risk of hypoglycemia, longer disease duration, those with more co-morbidities and those with established vascular complications.


Perspective 3: Do Patient Characteristics Influence Choice of DPP-4 Inhibitor?

Bernard Zinman CM, MD, FRCPC, FACP (University of Toronto, Canada)

  • The ideal drug for diabetes has many aspects. These include: safe, efficacious, durable control, well tolerated, low risk of hypoglycemia, weight neutral or weight loss, can be used at all stages of the disease, provides complimentary mode of action with other medication.
  • There are some differences between the five available DPP-4 inhibitors, although the efficacy is quite similar. The most cut-and-dried difference that we are aware of, though, is that the share of renal excretion is different – linagliptin has only 5%, versus 87% for sitagliptin (Merck’s Januvia) and 85% for vildagliptin (Novartis’ Galvus). This means that linagliptin can be used in patients with greater degrees of renal impairment.
  • There are many advantageous aspects of linagliptin in efficacy, safety, and convenience. The long term durability of linagliptin is not firmly established but data suggest that it works out to at least two years. Efficacy seems unaffected by patient age, or duration of diabetes. A meta-analysis of all the linagliptin studies shows a 0.7% reduction in A1c, which is similar to the other DPP-4s. Adverse events are low – there were only two events of pancreatitis in 2,566 patients receiving linagliptin versus zero in the control group.
  • Prof. Zinman stated that “unless there is a financial reason, for me, DPP-4s seem a superior choice than sulfonylureas. I am bold enough to go to a DPP-4 inhibitor after metformin rather than a sulfonylurea.”



Again this session, we emphasized the use of DPP-4s as a replacement for SUs, but also got into an interesting discussion of how prescriptive the guidelines should be. Should physicians be told exactly when to use a DPP-4, or do we trust them to make the correct individualized patient decisions?

Perspective 1: Hypoglycemia and Cardiovascular Risk

Brian M Frier, MD (University of Edinburgh, Scotland)

  • The heart already has several problems in diabetes, so the effects of hypoglycemia are superimposed on top of a weakened organ. Existing problems from diabetes include coronary artery disease (e.g. atherosclerosis), autonomic dysfunction, and a diseased cardiac muscle  (cardiomyopathy).
  • Hypoglycemia is associated with cardiac ischemia. There is a very high release of adrenaline (similar to a major trauma or heart attack) in hypoglycemia, which magnifies the symptomatic response. In hypoglycemia, studies show that the heart has to do a lot of extra work. This is fine if you have a young, healthy heart, but if the heart is diseased there is a bigger risk of ischemia. Hypoglycemia is also associated with negative ECG changes. Studies also show that during hypoglycemia, coronary blood flow is compromised in patients with diabetes compared to healthy patients. The normal autonomic reflex response to stress is blunted by antecedent hypoglycemia.
  • Hypoglycemia is also associated with endothelial dysfunction, blood coagulation abnormalities, and inflammation, as well as the sympatho-adrenal response mentioned above. The effect on the vasculature may persist a lot longer than the hypo episode– which sets the heart up for future problems.
  • Remember that 80% of people with type 2 diabetes die of heart related issues – the ‘smoking gun’ here may well be hypoglycemia. In ACCORD, the cause of excess mortality was not established, but clearly intensive therapy leads to higher hypoglycemia. In VADT severe hypoglycemia was a predictor of death. A retrospective cohort study of patients with type 2 diabetes on oral medications showed that the cohort with higher hypoglycemia had the higher mortality.
  • “Is the recommended target of 6.5% A1c appropriate for all patients? Because of hypoglycemia, we have to be very cautious”.


Perspective 2: DPP-4 Inhibitors vs. Sulfonylureas 

Bernard Zinman CM, MD (University of Toronto, Canada)

  • “Not all therapies are created equal, and some are associated with more weight gain and hypoglycemia.” Comparing linagliptin and sulfonylurea (SU) over time, we see similar A1c lowering out to two years, with possibly a little bit better durability for linagliptin. But the SUs have much worse weight gain and hypoglycemia than linagliptin (and all the other DPP-4s and GLP-1s).
  • There is some suggestive (but not definitive) data showing a lower relative risk for adjudicated cardio-vascular events with linagliptin versus glimepiride, particularly for non fatal stroke. In a prospective small-scale study, linagliptin was associated with reduced cardiovascular risk, but it was again not definitive.
  • However, the ongoing CAROLINA study is designed to conclusively evaluate the cardiovascular safety of linagliptin versus an active comparator (glimepiride). CAROLINA will study 6,000 patients with or without a metformin background over a six to seven year follow up period. The primary endpoint is the time to first occurrence of the primary composite endpoint. There is no placebo group, but this study addresses the clinically relevant question ‘which drug is better as second line therapy?’
  • Prof. Zinman applauded the FDA for insisting on a comprehensive post-approval cardiovascular study for new diabetes drugs. He feels that the evidence suggests that linagliptin, like other DPP-4s, are not associated with increased cardiovascular risk.



Brian M Frier, MD (University of Edinburgh, Scotland); Bernard Zinman CM, MD (University of Toronto, Canada)

Q: Is there any evidence that metformin is toxic?

A: No, we don’t think it is toxic at all. But we need to avoid it in certain circumstances, such as in renal impairment.

Q: What is the risk of hypoglycemia with linagliptin when used in patients with cardiac failure.

A: The risk of hypoglycemia with linagliptin is very low, not zero, but extremely low.

Q: Can we use linagliptin in a patient that already has pancreatitis?

A: I would avoid using DPP-4 and GLP-1 in those patients. Not because we believe they cause it, but we should just be cautious.

Q: Hypoglycemia unawareness is not uncommon in patients on intensive insulin therapy. Do we have electrophysiological studies of the heart in these patients?

A: No, not on this group. It’s less common in type 2 diabetes. I don’t believe that they have any enhanced risk, since the adrenal response is minimized. There is a lot of asymptomatic hypoglycemia going on, and we still need to learn if this can trigger the same symptoms. I suspect that it will and this is an area that needs a lot of further explanation. We can’t show cause and effect in ACCORD unfortunately.

Q: Aren’t the current ADA/EASD guidelines too vague, and give the physician too much room to make errors given the breadth of options they are given by the algorithms?

Dr. Zinman: I couldn’t agree with you more. What PCPs need is not a longer and longer list of drugs. We can be more prescriptive, using the latest evidence. I am critical of the current round of guidelines.

Dr. Ferrannini: Yes, but we can’t tell physicians that we know more about the situation than they do. There just isn’t good enough evidence to support one track versus another given the large numbers of drugs available.

Dr. Davies: The big barrier to improvements is clinical inertia. More and more patients are being treated in primary care. Fewer physicians read the guidelines carefully. We need a clear scenario for when we should use a DPP-4. This would be better than giving every patient every choice.

Dr. Ferrannini: But there is no evidence that this would improve compliance. Guidelines aren’t being followed anyway. The desire is to put the treatment in the hands of the doctors rather than the trialists. Diabetes is a very complex disease, but now we want an algorithm that fits all patients, all over the world, in all circumstances. We thought that this was just too ambitious.

Q: Audience Poll - What is your typical drug choice after metformin?

A: SU 28%, insulin 3%, GLP-1 7%, DPP-4 58% (!), TZD 4%

[Note: additional talks from this corporate symposium can be found in the “Novel Therapies” section of this report]


Corporate Symposium: 25th Anniversary of the Discovery of GLP-1 (Sponsored by the Samuel Lunenfeld Research Institute)


Jens Hull Holst, MD, PhD (University of Copenhagen, Copenhagen, Denmark)

Dr. Jens Holst presented a history of GLP-1 discovery (calling his presentation “a personal account, the way [he] feel that it happened”) starting from the initial search to identify the hormone in the distal small intestine involved in the incretin effect and ending with GLP-1’s various therapeutic benefits  today. He focused specifically on proving how endogenous GLP-1 stimulates insulin and reduces glucagon secretion in healthy subjects. In particular, Dr. Holst identified sustained weight loss over two years and restoration of glucose tolerance in individuals with prediabetes as indicative that the discovery of GLP-1 was not just of physiological importance, but also had great clinical potential – we were keen to hear this influential researcher talk about the potential for pre-diabetes. He then highlighted the importance of GLP-1 in the efficacy of glucose tolerance improvement almost immediately after Roux-en-Y gastric bypass surgery. Finally, Dr. Holst addressed why GLP-1 therapy might be appropriate in type 1 diabetes. Dr. Holst argued that, due to its inhibitory effects on gastric emptying and glucagon secretion, GLP-1 could help patients with type 1 diabetes that have no residual beta cell function. He presented data suggesting that, while GLP-1 did not appear to awaken silent beta cells, its glucagon lowering effects were independent of insulin secretion. Furthermore, after four weeks of treatment with liraglutide in patients with and without beta cell function, patients saw a major reduction in insulin dose and a decrease in hypoglycemia. During Q&A, Dr. Holst said there were other incretins that could be valuable to patients, including gastrin and secretin.

Questions and Answers

Q: I noticed in the study after bariatric surgery that the pattern of what you saw with exendin 9-39 [a GLP-1 receptor antagonist] was consistent on insulin secretion but it appeared to have an even larger effect on glucose concentrations pre-operatively. How is that consistent with over secretion of GLP-1 after the operation, which should give you more substrate for the exendin 9-39 GLP-1 receptor antagonist to work on?

A: This does raise a question of using exendin 9-39 as a tool. I suppose the interference of glucagon function is important here as well, and I didn’t show you that. The glucagon behavior in these people and with exendin after operation, is something we don’t understand right now.

Q: Although those patients don’t have a stomach, I guess we assume there is an effect of GLP-1 on the small intestine. Does that interfere with the absorption process?

A: That’s possible.

Q: For GLP-1 targets such as gastric emptying, GLP-1 is a major player, but there are other players that are GPR receptors. What is your view on those targets in medication and therapy in the future?

A: This is a beautiful field for others to cover in this room. I’ll just throw a bit of salt in the wound and say that one of the studies we did involved looking at the twenty-fold, at least, GLP-1 induction in response to meal intake, and in those meals, there was no fat.

Q: So you told us about the situation around 1987 and 1988 that there were reasons to believe that there are new incretins to be discovered. So how about today? Is there still some room for incretins that haven’t been discovered so far?

A: I know that my own mentor has published a paper where we say a focus has been on GLP-1, but what about gastrin and secretin? What about them? Is there room still? I could add that the double incretin knockout mice suggest that if you take away most GIP and GLP-1, they aren’t terribly sensitive bodies.

Q: What do you think we should make of improvements in insulin resistance? Is that related to gut hormones or just pure caloric restriction?

A: Several groups have studied this, showing caloric restriction of the kind we’re looking at here acutely will improve hepatic insulin sensitivity. Peripheral insulin sensitivity is not changed shortly after bypass; that takes months and is probably related to loss of fat, particularly intramuscular fat and possibly liver fat. I’m quite sure it is not just related to the gut hormone.



Fiona Gribble, PhD (Cambridge Institute for Medical Research, Cambridge, United Kingdom)

Dr. Fiona Gribble detailed findings from an L cell model that she developed to investigate mechanisms   of GLP-1 synthesis and secretion. As a reminder, L cells are located in the intestine and secrete the GLP-1 hormone. Her major finding was that nutrients in the intestinal lumen modulate L cell GLP-1 secretion by depolarization of the cell upon nutrient transport from the lumen into the L cell cytosol.

  • Dr. Gribble reviewed the gut’s role as an endocrine hormone, as well as the synthesis of GLP-1. While endocrine cells account for only ~1% of the gut lining, in total they make up the largest endocrine organ of the body thanks to the gut’s substantial length. GLP-1 is a cleavage product of the proglucagon peptide that, in the pancreas is cleaved to release glucagon, but in the L cells of the intestine is cleaved to release GLP-1 and GLP-2.
  • Dr. Gribble described the transgenic mouse L cell model that her team developed to investigate the mechanisms of GLP-1 secretion. Dr. Gribble’s team engineered a transgenic mouse expressing the GLU-Venus construct in which every cell that expressed proglucagon (i.e., L cells in the intestine and alpha cells in the pancreas) also glowed fluorescent yellow. This allowed for the visualization and identification of L cells in the intestine, which otherwise would not have been possible since L cells represent a small percentage of all gut cells.
  • Dr. Gribble then explained that the mechanism for GLP-1 secretion from L cells involves nutrient-induced depolarization. Glucose from the intestinal lumen enters L cells through sodium-coupled glucose transporters (primarily through SGLT-1) along the gut border. Each glucose molecule that enters is coupled with a Na+ ion. Therefore, glucose stimulation increases the inward flow of positive ions into the cell, depolarizing the cell and allowing for an action potential to signal (mediated by an increase in intracellular Ca2+ concentration) for the release of GLP-1. In SGLT-1 knock-out (KO) mice, GLP-1 is not secreted when the mouse is challenged with glucose by oral gavage. Additionally, Ca2+ concentration within L cells from SGLT-1 KO mice did not increase in response to the addition of glucose. Based on these findings, Dr. Gribble stated that there is a strong body of evidence suggesting that a major way in which L cells sense glucose in the gut lumen is through the activity of SGLT-1. She also presented data demonstrating that amino acids, as well as certain di- and tri-peptides, can similarly trigger GLP- 1 release from L cells via ion-coupled transport that results in L cell depolarization.

Questions and Answers

Q: Is there a possibility that GLP-1 could beget GLP-1? Have you tested this, or do we know if there are GLP-1 receptors directly on L cells or those of any of other hormones that positively feed back onto themselves like GLP-1 does on beta cells?

A: We don’t find GLP-1 receptors on L cells themselves, but if you give GLP-1 to animals or humans, you affect endogenous GLP-1 secretion. So we think that there is cross talk, but not directly through L cells.

Q: Would SGLT-2 inhibitors have some cross reactivity on SGLT-1?

A: I haven’t seen the data. I would predict that if you partially block SGLT-1, then you could deliver more glucose lower along the GI tract, where there is a higher concentration of L cells, and have a pharmacologic bypass experiment.

Q: Since we estimate the half-life of L cells to be two to three days, would it be possible to drive stem cells to produce more GLP-1? And which transcription factors are involved?

A: There is a whole list of transcription factors in a paper we published. We did a microarray of these cells; we recognized a lot of ones that came up, and there were also a heap of ones we did not recognize. The conclusion was that we didn’t think it would be easy to generate more L cells without generating more I and K cells at the same time because it would be difficult to pull out a transcription factor that would actually specifically produce L cells and not other cell types. There is a lot of potential and we haven’t tried everything yet.

Q: When we eat, we are bombarding L cells with hundreds or thousands of ligands at the same time. So the way it integrates those nutrient responses is different from when we give a lot of glucose or a lot of peptides alone. Have you had a chance to look at L cell response  to a combination of nutrients? To enhance GLP-1 response, would it be best to stimulate with a lot of one type of nutrient, or would it be optimally increased in a combinatorial manner?

A: I think you need more than one stimulus to get a lot out of an L cell. We haven’t fully defined the synergy but it looks like more than an additive effect when you combine stimuli. At least in vitro you can start to add these signals together, and the very effective action of glutamine you see in vitro illustrates the strength of putting transporter systems together. I think elevating cAMP is the most obvious way to try to increase secretion in vitro. I think combined with food intake arriving at right place, you can have the best effect.


Diana Williams, PhD (Florida State University, Tallahassee, FL)

Dr. Diana Williams’ study focused on GLP-1 neurons in the nucleus of the solitary tract and investigated whether the projection of these neurons to the nucleus accumbens (NAc) mediates the anorexic effects of meal-related GI signaling. In her lab’s study, rats received intra-NAc injections of saline or exendin [9- 39] (a GLP-1 receptor antagonist) 15 minutes before a preload of saline or condensed milk, after which rats were able to feed ad lib on chow. Exendin [9-39] reduced food intake following the preload by decreasing meal size rather than the number of meals consumed or the frequency of meals. To Dr. Williams, this finding suggests that NAc GLP-1 neurons play a physiologic role in the control of food intake. Results also showed that GLP-1 appears to reduce the palatability of food, indicating a possible mechanism through which GLP-1 projections to the NAc may regulate food intake. Dr. Williams then detailed her investigation of GLP-1 action in the context of a high-fat diet. Previous data show that overconsumption of a high fat diet is due in part to endogenous opioid activity in the NAc, leading Dr. Williams to examine whether NAc GLP-1 receptors interact with μ-opioid receptors. GLP-1 agonists affected the consumption of a high-fat meal only when given in combination with a μ-opioid antagonist (Naltrexone), suggesting that endogenous opioid activity in the NAc may impair the anorexic effects of GLP-1 signaling.

Questions and Answers

Q: Do you think that rats are the same as mice in regard to this response? Are they the  same as larger animals and humans? Because rodents are completely different in terms of periodicity and their 24 hour cycle, and the way they eat, etc. Do you have any sense of how applicable the learnings are from the rodent studies to humans?

A: This is an issue with GLP-1 neurons because even mouse and rat are relatively different, although the neuronal projections are similar. We don’t know if the details, such as the behavior effects, are the same. We don’t know how GLP-1 affects palatability. As far as humans go, no one has had GLP-1 injected intracranially. We only have to go on peripheral injection. Anecdotally, there’s discussion on the effect of GLP-1-based therapy on food reward. People are saying that they’re less excited about palatable food such as junk food. But there has to be more studies done because it’s hard to tease out the details on what is happening. There is some reason to think that these types of effects are going on in humans, but we need more studies.

Q: Though I think you’ve looked at this before, you didn’t speak today on whether you can separate an aversive response from a more selective satiety response. How do you view the separation between a highly specific responses vs. just sensing a massive aversive response?

A: That’s certainly an issue with looking at GLP-1 in the brain. We know that global administration will give you taste aversion, such as nausea. I think that GLP-1 is part of the aversive response. That’s been found to be part of the response. I don’t think that’s going on a meal-to-meal basis. We have shown that for some brain areas, injection of GLP-1 at the same doses that affect food intake doesn’t produce a taste aversion. And injection of exendin-9 at this site doesn’t reduce nausea. So there are sites where you don’t get that effect. Injection of GLP-1 into the amygdala shows an effect. So I think there’s a lot of functional segregation throughout the nervous system.



Astrid Plamboeck, MD (University of Copenhagen, Copenhagen, Denmark)

Dr. Astrid Plamboeck opened by noting that because native GLP-1 is rapidly degraded by DPP-4 (only 10% of intact peptide reaches systemic circulation), it may act locally to confer its biological effects, possibly via the vagus nerve. Dr. Plamboeck’s lab studied participants who had undergone truncal vagotomy (surgical cutting of the vagus nerve) to investigate GLP-1’s effects on plasma glucose, gastric emptying, insulin secretion, and food intake. Ten vagotomized participants and controls received GLP-1 infusion or saline 30 minutes before a liquid meal and four hours before an ad lib meal (where participants could eat as much as they desired). Vagotomized participants exhibited accelerated gastric emptying compared to controls (when both were given saline) and GLP-1 had a slight effect on gastric emptying in vagotomized participants compared to a significant effect in controls. Vagotomized  patients also exhibited altered glucose homeostasis, with a higher level of glucose following the liquid meal compared to controls. Glucagon levels were similarly elevated in vagotomized participants. GLP-1 infusion lowered post-prandial glucose and glucagon levels to a much lesser extent in vagotomized vs. control participants, perhaps due to the loss of GLP-1 signaling via the vagus nerve. Furthermore, greater insulin secretion, decreased food intake, and less appetite sensation was observed in vagotomized participants, with GLP-1 infusion having no effect on these parameters.

Questions and Answers

Q: Regarding the pathways that you’re identifying that may be lost, do you think that they are highly selective GLP-1 pathways? Or are they relaying all kinds of information, and maybe the loss of GLP-1 responsiveness is secondary to multiple defects in these people? Because this is not a selective ablation of GLP-1 pathways, it’s a massive ablation of communication.

A: I think a value of this experiment is that we examined them with or without the GLP-1 infusion. I can’t speculate on the rest of the action.

Q: Did you have a chance to treat these people with a DPP-4 inhibitor, which we speculate is a more gentle method, but perhaps also very important for communication signals, perhaps from the same pathway? So you get much lower levels of GLP-1, but more direct engagement of that nerve communication.

A: We didn’t do it in this study, because it doesn’t look like DPP-4 inhibitors affect food intake or weight in a clinical setting. We did that in another study where we gave an OGTT, but we haven’t analyzed these data yet.



Julio Ayala, PhD (Sanford-Burnham Medical Research Institute at Lake Nona, Orlando, FL)

Dr. Julio Ayala presented results from his research on potential mechanisms of GLP-1 agonists on cardioprotection by examining the effects of exendin 4 (the active ingredient in exenatide) on cardiomyocyte loss in neonatal rat ventricular myocytes. Previous studies had established that incubating cardiomyocytes in high glucose promotes apoptosis. Dr. Ayala found that incubating cells with exendin 4 reduced hyperglycemia-induced cell death in a dose-dependent manner. He then investigated the mechanism for this cardioprotective effect. He first hypothesized that exendin 4 might mediate oxidative stress, one cause of cardiomyocyte death. However, exendin 4 did not reduce markers of oxidative stress. Endoplasmic reticulum (ER) stress occurs downstream of oxidative stress, and Dr. Ayala found that markers of ER stress (CHOP and GRP78) were down regulated with the addition of exendin 4. In examining exendin 4’s effects on ER stress, Dr. Ayala identified a potential therapeutic target: SERCA2a. He found that exendin 4 protected cardiomyocytes from thapsigargin-induced death; thapsigargin is a known inhibitor of calcium homeostasis that acts by inhibiting SERCA2a (a pump that moves Ca2+ from the cytosol into the sarcoplasmic/endoplasmic reticulum and is important for cell viability). Specifically, exendin 4 enhances the phosphorylation of phospholamban, a protein that, only when phosphorylated, enhances SERCA2a affinity for Ca2+. Additionally, SERCA2a is also inhibited by hyperglycemia; since Dr. Ayala had already demonstrated that exendin 4 protects cardiomyocytes from hyperglycemia-induced death, he concluded that exendin 4’s ability to enhance SERCA2a activity would be especially relevant to preventing loss of cardiomyocytes in the context of diabetes.

Questions and Answers

Q: You used a high glucose model, and for some of your stressors you don’t really need glucose. So do you have a sense of the interactions preserved under normoglycemic conditions, or do you require high glucose to get the good effects?

A: Yes, so we’re doing some of those controls right now to look under normal conditions. One of the key targets for GLP-1 receptor agonist activation is PKA and we don’t know if we want to be activating PKA under normal conditions. So this might be an example where hyperglycemia produces a defect that GLP-1 restores.

Q: You only showed results for exendin 4, so I would be curious to compare these effects to other GLP-1 receptor agonists.

A: We are interested in doing that, and there is even interest in using the GLP-1 9-36 molecule.

Q: I saw two or three posters at ADA that said many actions of GLP-1 on the heart in terms of glucose uptake or vascular reactivity were lost in older patients with diabetes. Do you have thoughts about making animals, even for 24 hours, hyperglycemic first, and then examining whether you get preserved responses?

A: We’re establishing methods for isolating cardiomyocytes from mice but that’s certainly an interesting study to see if this is a preventative or intervention type effect.



Daisuke Yabe, MD, PhD (Kansai Electrical Power Hospital, Osaka, Japan)

Dr. Daisuke Yabe presented data on methods to enhance GLP-1 secretion. He reported that metformin and alpha-glucosidase inhibitor therapy (specifically, Acarbose) increased both intact and total circulating GLP-1 post-prandially compared to control, but had no effect on levels of DPP-4. Therefore, Dr. Yabe suggested that use of metformin or Acarbose in combination with a DPP-4 inhibitor would be an especially potent combination for enhancing endogenous GLP-1 activity. Additionally, past research had indicated that A1c reduction with the use of DPP-4 inhibitors correlated with fish intake in Japanese patients. Efforts are currently underway to identify which nutrients in fish may potentiate this effect,  but Dr. Yabe hypothesized that the fish nutrients may actually potentiate GLP-1 secretion. In addition,   he found that when Japanese patients with type 2 diabetes eat fish before rice, total GLP-1 levels  increase more after the meal than if rice is eaten before fish, and post-prandial glucose excursions are also improved. Therefore, he concluded that the nutrients in fish enhance GLP-1 secretion and that metformin or Acarbose in combination with eating fish could be another viable treatment option.

Questions and Answers

Q: Your results about the order of fish and rice were quite intriguing. Did you consider the possibility that rice empties faster from the stomach than fish? Could that explain parts of results?

A: Yes, I believe some of the effect might be explained by that.



Juris Meier, MD (Ruhr-Universität Bochum, Bochum, Germany)

Dr. Juris Meier reviewed several studies in discussing GLP-1’s pancreatic effects, including its influence on beta cell function, insulin secretion, and glucagon concentration. He began by reasoning that since people’s capability for beta cell proliferation go down as they age, incretins likely preserve beta cell function by inhibiting beta cell death rather than stimulating cell replication. Native GLP-1 also promotes glucose homeostasis, primarily through increasing insulin secretion. Dr. Meier emphasized that GLP-1 also influences the “rhythm” of insulin secretion, namely the frequency and amplitude of insulin pulses. Dr. Meir then turned to GLP-1 agonists and explained that short-acting agonists inhibit gastric emptying, delay the absorption of intestinal glucose, and decrease post-prandial insulin levels while long-acting GLP-1 agonists stimulate insulin and reduce glucagon concentration, and thus act primarily on fasting glucose. Dr. Meier also highlighted that GLP-1 agonists stimulate insulin biosynthesis in addition to insulin secretion, and thus appear to help prevent a chronic degranulation of beta cells. GLP-1 agonists also increase the glucose sensitivity of beta cells and inhibit glucagon   secretion by increasing somatostatin. Dr. Meier closed his presentation by reviewing data showing that GLP-1 agonists’ effects on insulin and glucagon contribute equally to the drugs’ ability to lower glucose.

  • Dr. Meier explained that while several in vitro and in vivo studies of GLP-1 agonists and DPP-4 inhibitors indicate that incretins can stimulate beta cell proliferation, investigators have not seen a large clinical impact of incretins on the progression of type 2 diabetes. To Dr. Meier, the discrepancy is attributed to the fact that rodent studies have used young animals at an age equivalent to human infants. Subsequent studies that investigated beta cell proliferation in young and older animals found an effect of exendin-4 (the active ingredient in exenatide) on beta cell mass and replication only in the younger animals. Researchers have associated the decline in beta cell proliferation with the accumulation of the cell cycle inhibitor p16, which is increasingly expressed with age. Studies using pancreatic specimens from children show that Ki67 – a marker of beta cell replication – is abundantly expressed in pancreases from six-month-old infants but not 15-year-old adolescents, indicating that the rate beta cell replication declines drastically with age. The expression of p16 shows an inverse trend – nearly no expression in young pancreases but high expression in older specimens – suggesting that P16 may be one of many factors that restrict beta cell replication as people age.
  • In discussing GLP-1 agonists, Dr. Meier pointed out an interesting postprandial effect – acute administration of GLP-1 agonists in the postprandial state appears to lower, rather than raise, insulin levels. Dr. Meier explained this observation by noting that glucose  concentration drives insulin secretion and that several GLP-1 agonists delay gastric emptying and thus reduce the absorption of intestinal glucose. He stated that since the effect on gastric   emptying is dependent on tachyphylaxis (the decrease in response to a drug following its administration), long-acting GLP-1 agonists, which maintain a prolonged, elevated level of GLP-1 in the body, have little effect. Thus short-acting GLP-1 agonists inhibit gastric emptying, delay the absorption of intestinal glucose, and decrease post-prandial insulin levels while long-acting GLP-1 agonists act primarily on fasting glucose by stimulating insulin secretion and reducing glucagon concentration.

Questions and Answers

Q: You may have seen a paper published two weeks ago from Dr. Accili’s lab in New York. He used multiple different models – lineage tracing, genetic models, etc.– and found that it’s not apoptosis, it’s dedifferentiation. He doesn’t see apoptosis as a consistent procedure in diabetes, but he always sees dedifferentiation. So beta cells go away, but they go away to become other endocorine cells, and if we can switch that process back, that may be an effective approach. So we can look at these markers from the same sections of the pancreas and there’s data – it’s not great data–that suggests that GLP-1 affects the differentiation of beta cells. Do you want to comment?

A: It’s certainly intriguing, but it’s extremely difficult to prove or disprove in human models. We can’t do this experiment of GLP-1 stimulation, because we don’t have access to the pancreas. I think the story for the plasticity of beta cells is an old story, but there’s no proof in humans. What we do have is evidence of beta cell death in humans and animal models. Using markers of beta cell apoptosis, we do see apoptosis. That doesn’t exclude the fact that there might be some dedifferentiation. There’s also this hypothesis that some beta cells lose insulin expression. Maybe some beta cell death might be beta cell dedifferenition and the loss of insulin expression. Again, we can’t exclude this based on the models we use, but it’s an interesting thought.

Q: I wanted to challenge your view on the difference between short-acting and long-acting drugs. I think you neglected the glucose dependence of GLP-1 action in the way you looked at the data. There’s clearly a difference in their effect on gastric empyting. But I strongly believe there’s no difference in their effects on insulin and glucagon secretion. You have to take into account the effect of glucose. GLP-1 acts in a glucose-depedent manner. Fundamentally, the only difference I see is the effect on gastric emptying. All other effects seem to me to be exactly the same if you normalize for different glucose concentrations.

A: I don’t think we are in disagreement. I think that the difference in glucose concentration results from the difference in gastric empyting. That’s why I said that the major driver of insulin secretion is not GLP-1 itself, but the concentration of glucose. So the abiltiy of GLP-1 to stimulate insulin is glucose dependent. Clinically, what we see when we inject these drugs is a stimulation on one hand and a reduction on the other hand, but other than that, I agree with you.



Daniel Drucker, MD (Samuel Lunenfeld Research Institute, Toronto, Canada)

Dr. Daniel Drucker presented his view on various extra-pancreatic actions that have been proposed for GLP-1 agonists. Most of his presentation focused on the hot-topic of GLP-1 agonists’ potential cardioprotective effects. He reviewed several mechanisms for how GLP-1 agonists might exert positive cardiovascular effects; specifically, he identified blood pressure reduction as his favorite mechanism because it has been validated in humans. In turn, his favored putative mechanisms for how GLP-1 agonists reduce blood pressure are by diuresis and vasodilation. Dr. Drucker also discussed the lipid- lowering effects of GLP-1, but he cautioned that, based on currently available data, lipids could not be identified as an independent risk factor for cardiovascular disease that is modified by GLP-1. While preclinical studies have shown powerful cardioprotective effects of GLP-1 receptor activation or DPP-4 inhibition, Dr. Drucker stressed that little clinical data exists suggesting that these therapies will produce cardioprotection in elderly human patients with diabetes. He expressed hope that ongoing outcomes studies will be able to establish the cardiovascular effects of incretin-based therapies, but also warned it will be difficult to directly compare treatments due to differences in study design.

  • Dr. Drucker reviewed potential means by which GLP-1 agonists could act on the cardiovascular system; he identified blood pressure reduction as his favored mechanism by which GLP-1 agonists may provide beneficial cardiovascular effects. In humans, validated effects of GLP-1 therapy include weight loss, blood pressure reduction, heart rate increase, decreased post-prandial lipids, reduced post-prandial glucose, and improved vascular flow-mediated dilation (FMD). Dr. Drucker presented data demonstrating that both GLP-1 agonists and DPP-4 inhibitors are associated with rapid and sustained reductions in blood pressure (liraglutide produced ~2.5 mmHg reduction in systolic blood pressure within two weeks over 26 weeks).
    • Dr. Drucker believes that diuresis and vasodilation are the most likely mechanisms by which GLP-1 agonists reduce blood pressure. GLP-1 agonists and DPP-4 inhibitors both increase urine sodium excretion, though preclinical data suggest that DPP-4 inhibitors also act through a separate mechanism independent of GLP-1 that may involve the kidney. Dr. Drucker believes that endogenous GLP-1 does not act directly to induce vasodilation, but that this effect may be mediated through the GLP- 1 (9-36) cleavage product. With regards to potential GLP-1 effects on heart rate, Dr. Drucker believes that a substantial number of questions remain in this area including what parameters should be measured, whether the central or peripheral nervous systems are involved, if the effect persists over time, and whether mean heart rate or outliers should be the key endpoint.
    • Dr. Drucker also discussed the lipid-lowering effects of GLP-1, though he cautioned that lipids could not be identified as an independent risk factor for cardiovascular disease that is modified by GLP-1. In humans, DPP-4 inhibition reduces post-prandial total serum triglycerides (TGs) after four weeks of treatment. Additionally, in healthy humans, exenatide reduced the production of ApoB48, but not ApoB100 at a very low dose. However, he warned that the extent to which lipid lowering affects diabetes complications has not been well established, and the mechanisms coupling GLP-1 signaling to Apo-protein secretion are not well understood.
  • While preclinical studies have shown powerful cardioprotective effects of GLP-1 receptor activation or DPP-4 inhibition, Dr. Drucker stressed that there is little clinical data to suggest that these therapies will produce cardioprotection in older human patients with diabetes. An interesting incongruity in the argument that GLP-1 agonism produces cardioprotective effects is that when one compares the number substrates in the cardiovascular system that are responsive to DPP-4 to the number responsive to GLP-1, the  list is much longer for DPP-4, suggesting that DPP-4 inhibition might also act independently of the GLP-1 axis. Additionally, many metabolites generated by DPP-4 also have independent effects on the cardiovascular system (he did not specify what types of effects). Dr. Drucker stressed that, with regards to GLP-1 action in the cardiovascular system, we have very little idea of where the hormone works (whether the cardiomyocyte, nervous system, immune system, or blood vessel is the actual target of therapy). Furthermore, mice used in preclinical studies are often young and healthy as opposed to old and diabetic or atherosclerotic. He hopes that the ongoing long-term outcomes studies will be able to establish the cardiovascular effects of incretin-based therapies, but also warned it will be difficult to directly compare treatments due to differences in study design.

Questions and Answers

Q: Normally the active GLP-1 is a peptide sequence from amino acids 7 to 36, and the inactive one is amino acids 9 to 36. From your experience, is there a way to convert the 9- 36 peptide to the 7-36 peptide?

A: I’m sure you could do that chemically in the lab, but I’m not aware of a biologic process that would re- anneal those amino acids back to their previous form.

Q: If you have pharmacologically characterized some hints that there must be another receptor for GLP-1, why is it so difficult to nail it?

A: Let me clarify the concept of a second receptor: for years, people have asked me, “Is there a second GLP-1 receptor?” When we give a GLP-1 receptor agonist to our GLP-1 knockout mice, we never see glucose lowering, insulin secretion, or a reduction in food intake and body weight. I think those actions are mediated by the classic GLP-1 receptor. Once you start going into cardiomyocytes, liver, muscle, adipocytes and blood vessels, I think you can have GLP-1 dependent and independent effects, and I think some of those cells don’t express the classical receptor. I don’t think we know what that other mechanism might be. Others are trying to look at this to see what it is that recognizes the GLP-1 receptor agonist or a degradation product. It’s not an impossible task, it just hasn’t been done elegantly enough.

Q: Do you think GLP-1 is really the system that is mediating the complexities that result in metabolic  syndrome?

A: I think all benefits of GLP1 are pharmacological. You get a little improvement in insulin secretion at the upper ranges of the physiological range, but I don’t think it’s as important of an endogenous system that taking it away would cause dysregulation.

Q: What percentage of metabolic syndrome do you think is attributable to [endogenous] GLP-1 and its antagonists?

A: Close to zero.

Q: You emphasized the use of old and young animals. I’d like to know also when you continue with elevation of active GLP ligand, what happens to the receptor? It’s also important because as far as I understand, when we treat patients with a DPP-4 inhibitor, if active GLP-1 level goes up then total GLP-1 release might go down. When we continue with high levels of GLP-1 what happens to the receptor?

A: There are some very small limited studies saying that treatment of animals for a few weeks does not decrease receptor expression. I think most of those studies are done with antibodies that don’t recognize the GLP-1 receptor. I don’t think we have good data to answer that question. What chronic therapy in diabetic models does to receptor expression and signaling is not very well understood.

Q: Could you speculate about the potential cardioprotective role of reducing postprandial hyperglycemia? In ApoE knock out mice, if you produce intermittent hyperglycemia, you exacerbate atherosclerosis compared to those subject to chronic hyperglycemia. And alpha-glucosidase inhibitors appeared to be cardioprotective.

A: There is an abundance of biomarker data: post-prandial glucose correlates with ROS [reactive oxygen species], coronary artery calcification, and much more. To date, when you look at human intervention studies designed to target post-prandial glucose and then look at CV events, you don’t see convincing data. Acarbose does not only have one mechanism of action either – it increases GLP-1 in some studies and not in others, so we will have a hard time attributing its effects.



Michael Nauck, MD (Diabeteszentrum Bad Lauterberg, Bad Lauterberg im Harz, Germany).

Dr. Nauck, who has researched GLP-1 for the past 25 years and whose studies were repeatedly cited throughout the symposium, gave a straightforward overview of current and future GLP-1 agonists and DPP-4 inhibitors. He first compared short-acting vs. long-acting GLP-1 agonists, explaining that short- acting agonists have a greater effect on gastric emptying and post-prandial glucose while long-acting agents more significantly reduce A1c and fasting plasma glucose levels. Dr. Nauck then compared several GLP-1 agonists with respect to their differing effects on fasting glucose, A1c, post-prandial glucose, and weight. Specifically, he noted that long-acting GLP-1 agonists provide similar glycemic control compared to basal insulin, a fact clinicians should consider when choosing a diabetes treatment. In contrast, short-acting agents may be especially well-suited for combination with basal insulin therapy. Dr. Nauck then turned to DPP-4 inhibitors, first noting that some agents differ pharmacologically with respect to their pharmacokinetic profile, specificity, mechanism of DPP-4 inhibition, and interaction with CYP450. After comparing the effectiveness of DPP-4 inhibitors vs. sulfonylureas, Dr. Nauck ended by noting that not all of the effects of DPP-4 inhibitors appear to be mediated by GLP-1, leaving room for the existence of other mediators, which could be favorable targets for diabetes therapies.


2. Insulin and Insulin Therapy

Oral Presentations: What’s New in Insulin Therapies


Athena Philis-Tsimikas, MD (Scripps Whittier Diabetes Institute, La Jolla, CA)

Dr. Philis-Tsimikas detailed what she called the first study comparing a basal insulin (in this case, degludec) with a DPP-4 inhibitor (sitagliptin). Her group’s 26-week, open label study randomized 458 insulin-naïve patients with type 2 diabetes to insulin degludec (titrated to a pre-breakfast plasma  glucose of 71-89 mg/dl) or sitagliptin 100 mg, both on a background of one to two oral anti-diabetic agents (metformin, sulfonylurea, or pioglitazone). Baseline characteristics were similar between both groups, with an average age of 55-56 years, BMI of 30-31 kg/m2, A1c of 8.8-9.0%, and diabetes duration of 7.7-7.8 years. Over 26 weeks, degludec provided greater reductions in both A1c and fasting plasma glucose compared to sitagliptin (treatment difference of 0.43% for A1c and 39.1 mg/dl for FPG; p <0.001 for both). Nine-point self-measured plasma glucose data indicated that insulin degludec provided superior post-prandial glycemic control – we would have preferred to see “time in zone” data from CGM to assess this. Participants on degludec experienced weight gain while those on sitagliptin exhibited a slight weight loss (treatment difference of 2.75 kg [6.1 lbs]; p <0.05), and a significantly higher rate of overall hypoglycemia was observed in the degludec group (3.07 episodes per patient- year) vs. the sitagliptin group (1.26 episodes per patient-year). Both groups experienced comparable rates of adverse events.

Questions and Answers

Dr. Pablo Aschner (Javeriana University, Bogota, Colombia): We did a study called EASIE comparing glargine with sitagliptin and we found a greater A1c difference between glargine and sitagliptin. (For details, see page 36 of our ADA 2012 Insulin Therapies report at Does this have anything to do with titration? Did you have a committee monitoring the titration?

A: Are you wondering why we didn’t see as big of a difference in A1c? We had a nice reduction in A1c with sitagliptin in this population, more than what we’ve seen in the past. The titration was aggressive. Even the insulin reduction in A1c was higher than what we’ve seen in other studies.

Q: Regarding the titration scheme, in almost all the degludec studies, the target was a fasting plasma glucose of 70-90 mg/dl, which appears a bit scary. But the data indicate that you reached a glucose level of only 126 mg/dl. It looks like a failure of titration.

A: I agree. In almost all the studies, that’s approximately where all the fasting plasma glucose levels have been. I do think that the investigators backed off a bit and didn’t push to the maximum titration. I think that’s all we can conclude.

Q: Being able to take degludec at any time of day is great – it’s good for the administration by the patient and it gives them more freedom in the daily application. Is there a plan to combine the injection of the insulin with an insulin pen that records the last injection and the dose?

A: That would be fun. I’m not aware of that being the case. If you have a busy schedule and you still miss the dose, but you’re able to give it eight hours later, you’re confidence that it has the same effect without increasing the hypoglycemia rate.

Q: Your definition of confirmed hypoglycemia is someone with a glucose level below 56 mg/dl. Do you have any idea of how many people wake up during the night feeling low but not having a level below that? That’s a pretty rigorous standard.

A: An ideal way to measure this would have been to use a CGM measurement, and we don’t have those measurements. That might be the next study design, so we can get those measurements.



Jay Skyler, MD (University of Miami, Miami, FL)

Dr. Skyler reviewed the PH20 coformulation results in 117 type 1 MDI patients presented at ADA 2012   by Dr. Irl Hirsch – for the complete data, see page 210 of our ADA 2012 Full Report at The study compared a coformulation of lispro or aspart plus PH20 (hence referred to as analog-PH20) to lispro alone. Patients were randomized to one of the two groups for 12 weeks of treatment and then crossed over to the other group for another 12 weeks of treatment. The study met its primary endpoint for A1c non-inferiority. PH20 really shined in improving postprandial glucose excursions (an 82% overall reduction vs. lispro alone; p=0.004) and reducing hypoglycemia events ≤70 mg/dl by 5% (p=0.035) and events <56 mg/dl by 7% (p=0.045). Total daily insulin dose was similar between both groups (slightly trending in favor of PH20) and weight gain was negligible for both   groups (again slightly trending in favor of PH20). The safety profile looked quite good as well, with no apparent sign of increased injection site pain, reactions, or adverse events.

Questions and Answers

Q: It’s so interesting that the postprandial changes sound dramatic but A1c didn’t change. These were very well controlled patients from the get-go. But it’s interesting that you didn’t tease out a change in A1c.

A: The study was only 12 weeks long for each insulin, which might not have been enough.

Q: Did you see any local site reaction beyond the three-month treatment period?

A: There did not seem to be any increase during the course of treatment from the beginning to the end. We don’t know in perpetuity. But there wasn’t an increase during the study.

Q: Was there a change in the amount of basal insulin required when on hyaluronidase?

A: The two insulins had a one-unit change. One unit for basal and one for bolus.

Q: What about variability?

A: That hasn’t been analyzed.

Q: Was there a change in free fatty acids? That would reduce lipolysis.

A. It was not measured

Dr. Richard Bergenstal (International Diabetes Center, Minneapolis, MN): It was interesting seeing the difference in the ending ratio of rapid acting to basal. For a more rapid acting analog with less of a tail, you might need more basal.

A: It’s interesting. One of the things we need to be cautious about is that with the shorter duration of action with rapid acting analogs, it’s more important to be sure you have the basal insulin correct – the prandial won’t have a tail. As you’re implying, you need to pay attention to both simultaneously.

Q: What happened with hyaluronidase circulation after subcutaneous administration?

A: Subcutaneous tissue fairly rapidly corrects within 24 hours. There’s no permanent effect if one uses a different injection site.

Q: From a theoretical point of view, could you use hyaluronidase in an inulin pump?

A: The issue is more in terms of formulations at the right temperature and the coformulation of products. The company is about to embark on CSII study using hyaluronidase in an interesting way – before you change your infusion set. [Editor’s note: interim results were presented as a poster at ADA 2012; see page 217 of our report As we understand it, the new study will be a phase 3 500-patient study. At ADA were the results of the pilot.]

Q: Using a pump, could you have a system with basal infusion of just the rapid acting analog and the prandial bolus with hyaluronidase? I noticed you used twice daily glargine.

A: As I pointed out, the problem is that the tail of longer acting insulins contributes to basal insulin. Any time very rapid acting insulin is used, you must get the basal insulin correct. To prevent any vagaries in that, glargine was intentionally given twice per day in the trial so that at steady state there would be constant basal insulinemia.


Oral Presentations: Insulin: Beyond Traditional Delivery


Andreea Soara (University Campus Bio Medico, Rome, Italy)

Ms. Andreea Soara presented results on the use of Generex’s buccal spray insulin, Oral-lyn, in 11  patients with impaired glucose tolerance (IGT). Twenty-two participants were randomized to a treatment group (12 puffs of Oral-lyn at each meal, physical activity, and diet) or a control group (physical activity and diet). After six months of treatment, A1c declined by 0.4% in the Oral-lyn group (baseline: 6.2%) and remained stable at 6.1% in the control group. There was no increase in body weight, hypoglycemia, insulin antibodies, or other adverse events in the Oral-lyn group. Unfortunately, the A1c improvement was not sustained after a six-month post-trial washout period and the Oral-lyn group’s mean A1c returned to its pre-study baseline level of 6.2%. This was of course disappointing to see, though we think it would be interesting to follow the patients over a longer period and see if even  the six-month period of better glycemic control would help reduce progression to diabetes. Ms. Soara concluded that Oral-lyn “appears to be a valuable tool for subjects with IGT” and the product’s features and simple administration favor compliance and improved quality of life in subjects affected by IGT. Of course, since relatively few people with IGT develop type 2 diabetes (according to Ms. Soara it is one third), treating people with prediabetes with Oral-lyn will have to be either highly cost-effective or  based on other markers than just IGT alone. Further, we wonder if the company can concentrate the insulin to reduce the number puffs – 12 at one meal strikes us as on the higher side.

  • Generex’s Oral-lyn is a buccal spray insulin that is absorbed through the oral mucosa. Oral-lyn consists of human recombinant insulin (regular acting) that has been   dissolved in a buffer at neutral pH. Ms. Soara noted that it is identical to an injection but with the addition of absorption enhancers and stabilizers. She further explained that the particles are large and cannot enter into the lungs. Oral-lyn reaches peak efficiency after one hour and has a bioavailability of 15%.

Questions and Answers

Dr. Jay Skyler (University of Miami, Miami, FL): During the treatment period, was there a placebo in the control group?

A: No.

Q: You had the subjects take 12 puffs at each meal. That sounds like a large number of puffs. Are you doing any research to diminish the number of puffs?

A: In a previous publication, we already studied varied dosages. In these types of subjects, we compared four puffs, six puffs, and 12 puffs. We noticed that only with 12 puffs was there a reduction in plasma glucose and no hypoglycemia. That is why we started with 12 puffs. We have to make sure we have an improvement without causing hypoglycemia.



Andreas Pfützner, MD, PhD (Institute for Clinical Research and Development [IKFE], Mainz, Germany)

Dr. Andreas Pfützner presented interim results from an ongoing trial of the InsuPad, a patch-like device worn on the skin that provides localized heating after injection. Sixteen participants with type 2 diabetes injected rapid acting insulin analogs either with or without the device (each patient completed the test under both conditions) before a standard liquid meal tolerance test. Glucose excursion data over 300 minutes demonstrated a significant 23% reduction in maximal glucose excursion (106 mg/dl vs. 137 mg/dl; p <0.05) and a significantly lower glucose excursion from 55 minutes to 140 minutes. Dr.  Pfützner hypothesized that based on these initial findings, patients using the InsuPad should require less insulin (~20-30%) for postprandial glucose control in real world settings. A recently initiated trial in 160 type 1 patients is expected to complete by the end of this year and should shed more light on this hypothesis. We look forward to better understanding patient perceptions of the device, as we suspect many MDIs will perceive it as a hassle to wear and replace each day. Still, we see postprandial control  as an area deserving more innovation and hope the company continues perfecting the approach.

  • For background, the InsuPad device applies localized heating after insulin injection. It is comprised of a disposable pad (containing the insulin injection window, which has space for six injections; intended for one day use) and a reusable control unit (contains the heating block, electronics, and a rechargeable battery). Using the product has four main steps: 1) the patient connects the electronics to the disposable pad and places the pad on the skin; 2) the patient injects meal insulin through the opening in the pad; 3) InsuPad is activated automatically with each injection to deliver heat (39.5 degrees Celsius/103 degrees Fahrenheit) and stop after few minutes; and 4) the patient removes the disposable pad from the skin and the electronics are removed from the disposable pad and recharged. The product is being developed by InsuLine Medical and has already received CE marking.
  • Dr. Pfützner presented interim results from an ongoing study of the InsuPad device in 16 patients with type 2 diabetes using rapid acting insulin analogs. Participants had a mean age of 59 years old with a mean BMI of 29 kg/m2, and a mean A1c of 8.5%. The stated study objective was to test the effect of the InsuPad on post prandial glucose levels during a meal tolerance test and evaluate the safety of the device. Subjects injected 0.2 U/kg insulin before a standardized liquid meal and glucose and insulin measurements were taken from a venous line at various time points through the study. Each patient went through the protocol twice, once with the device (experimental group) and once without the device (control group).
  • Mean maximal glucose excursion was 23% lower in the InsuPad group (106 mg/dl vs. 137 mg/dl; p <0.05). 
  • Dr. Pfützner also showed the glucose excursion curve over 300 minutes post-meal consumption   to highlight that from minute 55 to minute 140, the difference between groups was significant at most time points. Area under the glucose curve was only significantly lower in the InsuPad group from 0-120 minutes (p <0.05; 0-60, 0-180, 0-240, and 0-300 were also assessed); however, Dr. Pfützner believes that once the full analysis is done, the study will show “significant differences all over the place.”
  • The InsuPad device was well tolerated by all participants, said Dr. Pfützner. There were no reports of skin burns, redness, or irritation.
  • An open-label, randomized, parallel design study of the InsuPad patch in 160 patents with type 1 diabetes has already initiated. The study is meant to test the effect of the InsuPad in a real world setting. Dr. Pfützner hypothesized less insulin will be needed to maintain the same post prandial control when using the InsuPatch, which would be welcome for patients and payors. The first results are expected by the end of the year and data will hopefully  be presented at ADA. Called the “Barmer Study,” it was designed with Barmer Health Care Insurance, one of the largest insurance companies in Germany. Pending positive results, InsuPad will hopefully be able to gain reimbursement for the device. We think this is an excellent example of a payor-industry collaboration that has the potential to benefit patients.

Questions and Answers

Q: What’s the temperature achieved?

A: We applied 39.5° Celsius. There is a trial seeing whether going up to 40° has an impact and it seems further increase has further impact.

Q: Can subjects feel the temperature increase?

A: Some patients realize that something is going on, but not all of them do. It does not at all disturb patient’s wellbeing.

Q: Did you use a control device?

A: We have not, but we will in the future.

Q: What was the baseline glucose before the meal?

A: It was in the range 120-160 mg/dl.

Q: Was it the same in both groups?

A: Yes, there was no difference in mean baseline levels.

Q: You commented that there could be a reduction in the dose needed for insulin. What’s the mechanism behind that?

A: Faster onset of action leads to the shutting down of hepatic gluconeogenesis. We can expect to see a 20- 30% reduction in required prandial insulin dose.

Q: Do you have information on the pharmacokinetics?

A: From this study we don’t because the study is still ongoing and we do those measurements at the end, but we know from previous studies that we see an acceleration – looking at t1/2 values – of approximately 20-25 minutes.

Q: If you didn’t use a dummy device, how could you ensure the depth of injection was the same?

A: The disposable pad is only a plastic window, so it is after the injection that the heat is applied. Basically, when you inject through the window without applying the heat it is like you are injecting without anything. And the same needles were used, so needle depth was not influencing the results.

Q: Why is there no pharmacokinetic data again?

A: We will measure that once we’ve completed all 20 subjects to make sure assay conditions for the insulin is standardized across all subjects.



Jitendu Vora, MD (Royal Liverpool University Hospitals, Liverpool, United Kingdom)

In this analysis of six phase 3 clinical trials of insulin degludec vs. insulin glargine, Dr. Jitendu Vora showed that few patients taking either analog experienced higher levels of insulin-related antibodies.  The titers of antibodies were measured using radio-labeled insulin; the result was expressed as the fraction of the total insulin that bound to antibodies. Levels of degludec-specific or glargine-specific antibodies were low (roughly 1% or lower bound in the radio-labeled insulin assay). The researchers  also assessed levels of antibodies that cross-linked with human insulin. The titers of these antibodies  were fairly low in insulin-naïve patients with type 2 diabetes (around 2%), while titers were higher in people who had been previously treated with insulin whether they had type 2 diabetes (around 5%) or type 1 diabetes (around 10-20%). Dr. Vora further showed that mean antibody titer did not increase beyond the first several weeks of a study, and that few patients experienced an antibody increase of  more than 10%. Some patients did experience significant increases in antibody levels, but Dr. Vora said that –encouragingly – antibody levels were not correlated with baseline A1c, change in A1c, insulin dose at the end of the trial, or injection-site reaction. This high-level analysis looked generally positive; we assume that we’ll get a chance to report on immunogenicity data in greater depth at degludec’s FDA advisory committee meeting on November 8.

Questions and Answers

Dr. Philip Home (Newcastle Diabetes Centre, Newcastle University, United Kingdom): Some people, as you would expect, had relatively high antibody titers. You showed that it didn’t affect A1c, but was there evidence of other immunological phenomena?

A: There was no evidence of injection-site immunological reactions.



Rebecca Owens (Lilly Research Laboratories, Indianapolis, IN)

In a comparative analysis of Lilly’s PEGylated insulin lispro (LY2605541, hereafter LY), Ms. Rebecca Owens showed that LY has lower in vitro binding to both isoforms of the insulin receptor (IR-A and IR- B), lower in vitro binding to the IGF-1 receptor (which can be involved in mitogenicity), and greater in vitro selectivity for IR-A vs. IGF-1 (roughly 100-fold greater than lispro). In assays of cell lines that overexpressed IGF-1, LY was also shown to have significantly lower IGF-1 binding compared to other insulin analogs. Ms. Owens further demonstrated that LY’s residence time on each IR isoform is similar to (and slightly shorter) than those of insulin lispro and insulin glargine, as opposed to the long residence time seen with the mitogenic insulin AspB10. As for what happens to LY once it has bound to the insulin receptor, Ms. Owens noted that the large PEG domain could theoretically interfere with the normal trafficking of activated receptors to the endosomes of the cell. An analysis suggested that low- dose LY has less trafficking than low-dose lispro, but the maximal trafficking response was similar for both analogs. This suggests that the PEG domain dose not prevent normal co-localization (though Ms. West acknowledged that the assay says nothing about how LY is processed once it’s inside the cell). As noted during Q&A, extrapolating in vivo activity from in vitro data is not always straightforward, so we hope that preclinical studies support LY’s safety and selectivity for the insulin receptor.

Questions and Answers

Dr. Philip Home (Newcastle Diabetes Centre, Newcastle University, United Kingdom):   Your in vitro affinities are down by 10-12 times. But in vivo clearance is through the insulin receptor, so it seems like circulating concentrations would be equivalently elevated by 10 times. At the same time, selectivity vs. IGF-1 is only fourfold lower than the other analogs. So won’t you get 2.5-fold more activation of IGF-1 in vivo? I’m not suggesting for a second that matters, but that is my interpretation of the data.


Oral Presentations: Insulin Action in the Liver


Bernard Zinman (University of Toronto, Toronto, Canada)

Dr. Zinman presented a 52-week head-to-head trial of insulin degludec vs. insulin glargine in patients with type 2 diabetes (n=1,030) on a background of metformin with or without DPP-4 inhibitor. Antihyperglycemic efficacy was similar (which makes sense given the treat-to-target design), but degludec led to roughly 20% lower rates of all confirmed hypoglycemia and nearly 50% lower rates of nocturnal confirmed hypoglycemia (based on the prespecified analysis of the “maintenance period” from week 16 to 52). Dr. Zinman attributed the benefits to reduced variability in action profile with degludec; unfortunately, data are not available to test whether time of glargine dosing affected hypoglycemia prevalence. He also noted that the small absolute rates of hypoglycemia in both study groups were due largely to strict exclusion criteria for people with multiple recent severe hypoglycemic episodes, implying that the real-world benefits would accrue to a large number of people. Indeed, he believes that the numbers-needed-to-treat – 50 patients for a year to prevent one instance of severe hypoglycemia, four patients for a year to prevent one instance of nocturnal hypoglycemia – are sufficient to favor use of degludec.

  • This 52-week, treat-to-target study enrolled insulin-naïve patients with type 2 diabetes and randomized them in a 3:1 ratio to take insulin degludec before the evening meal (n=773) or insulin glargine at a consistent time of day of the patient’s choice (n=257). The two groups were similar in mean BMI (roughly 31 kg/m2), disease duration (roughly nine years), and baseline A1c (8.2%); roughly 80% of both groups completed the study. Enrollment criteria permitted patients taking metformin (allowed to be maintained in the study) with a DPP-4 inhibitor (allowed to be maintained in the study; roughly 2% of patients were on a DPP-4 inhibitor), a sulfonylurea (discontinued prior to study start), or acarbose (discontinued prior to study start). Patients with notable historical duration of severe hypoglycemia were excluded for safety reasons. Each insulin was initiated at 10 U/day and titrated to a tight fasting plasma glucose target of 3.9-4.9 mmol/l (70-89 mg/dl).
  • According to the hypoglycemia definitions of degludec’s phase 3 program, degludec conferred statistically non-significant reductions in overall confirmed hypoglycemia vs. glargine, with statistically significant reductions in both confirmed nocturnal hypoglycemia and severe hypoglycemia. For the full 52-week study, degludec led to reductions in confirmed hypoglycemia (18% lower rate, not statistically significant), confirmed nocturnal hypoglycemia (36%, p<0.05), and severe hypoglycemia (86%, p<0.05). For the maintenance period (weeks 16 through 52), degludec’s relative advantages were larger still: 23% lower risk for all confirmed hypoglycemia, and 49% lower for nocturnal confirmed episodes
  • The absolute prevalence of hypoglycemia was quite low in both groups: fewer than two overall confirmed events per patient-year, below 0.5 confirmed nocturnal  events per patient-year, and below 0.05 severe episodes per patient year. However, during Q&A Dr. Zinman said he believes degludec is worth the switch based on the numbers- needed-to-treat: 50 patients for one year to prevent one instance of severe hypoglycemia (extrapolated from the full study), and four patients for a year to prevent one confirmed nocturnal hypoglycemic event (extrapolated from the maintenance period).

Questions and Answers

Q: Could you comment on weight change with degludec vs. glargine?

A: In this case there was absolutely no difference. Weight gain was a small number of kg in each group; I cannot remember specifically.

Q: I wonder whether you have any data on endogenous insulin production using C-peptide.

A: We did not measure endogenous insulin production, though that would have been an interesting sub- study. As you said, higher C-peptide levels would be predictive of easier control; though these were fairly advanced patients.

Q: Was time of injection of glargine associated with rate of nocturnal hypoglycemia?

A: Unfortunately in this large study, time of glargine dosage was not measured. In retrospect obviously we should have documented this. I assume it was mostly at night or in the evening since this is the general labeled recommendation, though clearly the rate of hypoglycemia changes based on when degludec is dosed, so it might be relevant for glargine as well.

Dr. Julio Rosenstock (Dallas Diabetes and Endocrine Center at Medical City, Dallas, TX): Why do you get a dichotomy of hypoglycemia when the A1c and insulin dosage is the same, and which only becomes apparent at the end of the study?

A: Tim Heise has several papers and posters showing major difference in variability between glargine and degludec. We all have patients who say that they do the same thing every night and eat the same things  but wake up with higher insulin. I think this phenomenon has to do with variability.

Dr. Rosenstock: Then why do we not see the difference at 12 weeks?

A: The doses were very low – patients started at only 10 units.

Dr. Geremia Bolli (University of Perugia, Perugia, Italy): I remember when we were comparing basal analogs to NPH, we had a similar percentage benefit on hypoglycemia.

But the absolute rate of nocturnal hypoglycemia was roughly tenfold higher in those studies compared to here. Could you comment?

A: Professor Bolli is pointing out an important aspect – that rates of hypoglycemia are low. In recruitment, if they had severe hypoglycemia of great duration, they were excluded. But we can still  project number-needed-to-treat from this study. You would have to treat 50 patients for one year to avoid one instance of severe hypoglycemia, based on these data. To avoid one confirmed nocturnal hypoglycemia, you would have to treat four patients for one year. To me that sounds like a good deal, even though those rates are low.

Q: Was there a difference in recovery time from hypoglycemia between arms.

A: I think that is a key question. It is good to have a stable insulin, but will you be able to recover? There is a poster comparing glargine and degludec response to induced hypoglycemia, and if anything the catecholamine response looks better with degludec.



Mary Moore, PhD (Vanderbilt University, Nashville, TN)

In this hyperinsulinemic/euglycemic clamp study in a conscious dog model, Lilly’s PEGylated insulin lispro (LY2605541, hereafter LY) was shown to have a greater impact on net hepatic glucose balance   and a slower effect on skeletal-muscle glucose uptake compared to regular human insulin (n=6 dogs for each insulin). Delivery of LY also led to a slower suppression of lipolysis compared to regular insulin (though both non-hepatic insulin response and lipolysis suppression were similar at the end of the study, the 420-minute mark). During Q&A, Dr. Moore noted that the study was too short to reflect any differences in hepatic triglyceride content but that buildup of liver fat – a potential safety concern given LY’s greater action in the liver – is being studied in ongoing human trials. She added that these preclinical data support phase 2 studies in which LY seemed to affect the liver to a relatively greater degree than other insulins, though she seemed unclear on the precise biochemical reason for LY’s mode   of action.

Questions and Answers

Q: Could you comment on the glycogen content of the liver?

A: We haven’t assessed glycogen content yet.

Q: Do you think that the delayed non-hepatic action reflects longer insulin action in the muscle fibers or is just a difference in the way the PEGylated insulin is metabolized?

A: It’s difficult to say. It may simply take this large molecule longer to reach those tissues; eventually we did reach similar concentrations of lipids and non-hepatic glucose uptake.

Q: In vitro, is there any difference in insulin action?

A: I don’t have any in vitro assays; some of my colleagues may wish to comment on this.

Q: Do you have any data for increase of de novo lipogenesis due to an increase of insulin signaling? Did you measure hepatic triglyceride content at the end of the study?

A: Given the acute nature of this experiment, we did not anticipate a change in hepatic triglyceride content and so did not measure it. In ongoing human studies the effect on hepatic triglycerides will be assessed. That is a very good question.


Stephanie Ros, PhD (Centre European d’Etude du Diabete, Strasbourg, France)

To explore an animal model for comparison of continuous intraperitoneal vs. continuous subcutaneous insulin infusion (CPII vs. CSII), the researchers used an osmotic pump with a single basal rate of insulin flow to test each delivery mode in male Wistar rats with streptozocin-induced type 1 diabetes. At one  and four weeks, rats were scarified and analyzed. The CPII group achieved better four-week weight    and better blood glucose control (measured by fructosamine), and it was superior to CSII in up regulating hepatic glycogen uptake while reducing hepatic oxidative stress and inflammation. Questioners proposed that the positive effects in the liver might have resulted simply from better glucose control rather than intraperitoneal delivery per se. However, Dr. Ros emphasized that the hepatic clearance of insulin seemed to be responsible for the liver effects, given that several non-liver oxidative stress measures were improved equivalently by both insulin delivery routes. She concluded by  suggesting that clinical intraperitoneal insulin delivery could hold promise for reducing complications rates. We lament that safe and patient-friendly options for CPII are not widely available, today, but we hope that translational research will accelerate due to the demand for faster-acting insulin in closed- loop applications (see Dr. Eric Renard’s presentation during the EASD 2012 Roche Corporate Symposium in our EASD Day #2 coverage at

Questions and Answers

Q: Do you have a difference in metabolic control between these groups? Could that be what was the difference between groups?

A: We do have a difference in glucose control. However, we can say that the research obtained is really   due to insulin bypass in the liver because in terms of oxidative stress we have it in cardiovascular systems, and CPII is not able to normalize this oxidative stress; so we have an important effect of insulin in the liver.

Q: I was interested in any effects that were potentially transient. If you look at human studies of CPII, you see a transient rise of triglycerides that eventually goes down with therapy. Did you look at other time points besides one and four weeks?

A: This was a very short study. We have a technical problem with the osmotic pump, which can last for six weeks and not longer. We might try to choose a different experimental setup in subsequent experiments.



Hans DeVries, MD (University of Amsterdam, Amsterdam, the Netherlands)

Dr. Hans DeVries described two studies comparing insulin degludec three times weekly (3TW) vs. once- daily insulin glargine; in these studies, degludec 3TW was shown to be inferior to once-daily glargine, which supported the ultimate decision to dose degludec once a day. Both the AM Trial (n=459) and the PM Trial (n=467) randomized insulin naïve type 2 diabetes patients formerly on oral anti-diabetic agents to insulin degludec 3TW or insulin glargine once-daily, both on a background of metformin ± a DPP-4 inhibitor, and both titrated to a pre-breakfast plasma glucose of 70-89 mg/dl. The trials followed the same protocol, but dosed degludec either in the AM or in the PM (Monday, Wednesday, and Friday). In both the AM Trial and PM Trial, insulin degludec 3TW failed to achieve non-inferiority in reducing A1c compared to glargine (glargine had a ~0.34% and ~0.26% greater reduction in the AM and PM  trial, respectively). Similarly, degludec 3TW failed to achieve non-inferiority in reducing fasting plasma glucose. A significantly higher incidence of nocturnal hypoglycemia was observed with insulin degludec 3TW in the AM Trial (11.5% for degludec vs. 7.4% for glargine) while in the PM trial, degludec led to a significantly higher incidence of overall hypoglycemia (32.2%) compared to glargine (21.4%). In the AM Trial, the incidence of hypoglycemia appeared to be higher the day of injection vs. the day after  injection, while the reverse pattern was observed in the PM Trial. The results support a once daily  dosing of degludec over a 3TW dosing.

  • Dr. DeVries opened his presentation by briefly reviewing the rationale behind the two phase 3 studies. Following its subcutaneous injection, degludec forms long chains of hexamers, with each hexamer centered on a zinc ion. As the zinc slowly diffuses, degludec monomers are released from the ends of the hexamer chains (for more information on degludec’s mechanism of action, please see page 38 of our EASD 2012 Day #2 report at Clamp studies indicate that a substantial amount of degludec remains in the body 48 to 72 hours after injection, and that degludec has a flat action profile with a half-life twice that of glargine’s. These preliminary data suggested a possibility for a less-intensive dosing regimen with degludec (dosed only three times weekly), which was supported by the results of a phase 2 trial.
  • The AM Trial and PM Trial had nearly identical study designs: The 26-week AM Trial randomized 459 insulin naïve type 2 diabetes patients formerly on oral anti-diabetic agents to insulin degludec 3TW dosed in the morning (between waking and first meal of the day) or to insulin glargine, both on a background of metformin ± a DPP-4 inhibitor. Both insulins were titrated to a pre-breakfast plasma glucose level of 70-89 mg/dl. Insulin degludec was dosed on Monday, Wednesday, and Fridays with a starting dose of 20 U, while glargine was dosed according to its label with a starting dose of 10 U. The 26-week PM Trial followed this same study protocol, but included 467 participants and dosed degludec 3TW with the main evening meal. Participants in both studies had similar baseline characteristics: mean age at baseline of 57-58 years, mean BMI of 32-33 kg/m2, mean A1c of 8.2-8.3%, mean duration of diabetes of eight to nine years.
  • DegLudec failed to achieve non-inferiority in reducing A1c and fasting plasma glucose. Over 26 weeks, degludec reduced A1c from 8.2-8.3% to 7.2%. Fasting plasma glucose levels were reduced from 167-178 mg/dl to 121-122 mg/dl. Data on self-measured pre-breakfast plasma glucose levels showed that on the first day following injection, degludec had a similar pre- breakfast SMBG (108 and 105 mg/dl for the AM and PM trial, respectively) compared to glargine (108 and 106 mg/dl). However, an increase was observed on the second day (124 and 117 mg/dl) and third day (135 and 127 mg/dl) after the injection.

Questions and Answers

Q: Regarding safety, did you see any significant difference regarding weight? Also, were the study discontinuations due to safety, a lack of efficacy, or to hypoglycemia?

A: There was no difference in body weight observed in these two trials. The completion rates were almost 90% in both trials, without any difference for the reasons for the discontinuations.

Q: You had quite an aggressive target for your insulin titrations – a fasting plasma glucose level of less than 90 mg/dl. During daily practice, patients sometimes go that low, but if you consider that your study population had a diabetes duration of nine years, would you consider that to be a really aggressive push, or do you think that it was realistic?

A: I think it was a reasonable approach. It’s been noted to be safe for glargine. These patients were insulin naïve. The titration target was identical for both arms in both trials. Plus, the target is only a piece of advice – the investigators can back off if they think the blood glucose is getting too low. I think in all the trials in insulin naïve patients, the targets were not met.

Q: Three times weekly is not optimal for this compound. How do these findings comply with the PK data that show that degludec is really in the circulation for a long time? Were the results a surprise, or were they expected?

A: I think with hindsight, we could have expected this results, although there is a substantial amount of degludec in the body 48 to 72 hours after injection. We were depending on a lot of endogenous insulin taking over during that time. But we were hoping for too much from a diseased body and pancreas.

Q: In one study presented earlier, you looked at three times weekly dosing. What does your data look like added to the phase 2 trial from last year?

A: We think the phase 2 trial showed that you could dose degludec three times weekly, but that there was no benefit in hypoglycemia – that benefit was lost. So it was an indeterminate outcome and we needed this larger phase 3 trial to draw real conclusions. We found the same thing with respect to A1c – no difference. But in these trials, you see more hypoglycemia with degludec, which was not seen in the phase 2 trials.

Q: I’m curious about the study design because you used metformin with or without a DPP-4 inhibitor and asked people to stop using their sulfonylurea. Why did you do that? If you analyzed people on just metformin vs. on metformin with a DPP-4 inhibitor, were there  any clinical differences in the results?

A: Are you asking whether people should stop sulfonylureas with insulin? This is one of the major unresolved questions with insulin therapy in type 2 diabetes. We had a trial a few years ago and after lots  of discussion, we decided to leave it up to the countries participating in the trial and 50% decided to continue sulfonylurea and 50% decided to stop it. One main consideration with stopping sulfonylurea is that in the trial, if you see hypoglycemia, you have to blame it on the insulin, not on the sulfonylurea. The 20 people on DPP-4 inhibitors came mainly from the US, and there was not a planned analysis to just look at that.

Q: I think the main conclusion you suggest is that degludec should be used for once daily injection.

A: Yes


Oral Presentations: Insulin: Beyond Traditional Delivery


Marinos Fysekidis, MD (Saint-Louis University Hospital, Paris, France)

Dr. Marinos Fysekidis presented results from the INSUlin regimens and VASCular functions study (INSUVAC), a randomized, open-label study that examined changes in cutaneous blood flow (CBF) in people with poorly controlled type 2 diabetes before and after insulin treatment. Inclusion criteria required participants to: 1) be aged 30-72 years; 2) have had type 2 diabetes for more than one year; 3) have an A1c between 7.1% and 12% (mean A1c was 9.0%); and 4) have been on metformin and sulfonylurea therapy for at least two months. Participants (n=42 included, 34 completed) were randomized to an insulin regimen of either 1) aspart before meals; 2) basal bolus insulin therapy (detemir and aspart); or 3) bed time detemir. Baseline characteristics did not differ significantly, except for gender (female/male: 8/6, 10/3, and 3/11, respectively [p=0.01]). INSUVAC assessed CBF and sympathetic activation at fasting, and one and two hours after a standardized breakfast (75 g of carbs). Measures of CBF and sympathetic activity were taken at baseline and after four to five weeks of insulin therapy. Overall, insulin therapy generally resulted in significantly (p=0.007) increased fasting CBF,  but there was no difference between insulin therapy arms. The magnitude of the increase in  postprandial CBF was similar pre- and post-insulin therapy for fasting CBF, suggesting that insulin does not amplify breakfast-induced increases in CBF. Following the intervention period, the increase between participants’ fasting and one-hour post-prandial CBF levels was only significant for therapy regimens that included aspart. Dr. Fysekidis interpreted these results to be consistent with the expected vasodilatory effect of insulin. The researchers also analyzed sympathetic activation; sympathetic activity was enhanced post-prandially before insulin treatment, but not after insulin treatment. The absence of a change in sympathetic activity post-breakfast after insulin therapy, led Dr. Fysekidis to suspect that other mechanisms, like improved endothelial or myogenic activity, may be affecting CBF.

Questions and Answers

Q: Are increases in blood flow a good or bad thing?

A: I would say we probably should not jump to conclusions and see what the prevailing comorbidities are. My guess is that it would be sometimes helpful and other times provoke more damage.


Oral Presentations: Profiling Glucose and Clinical Trials


André Scheen, MD, PhD (University of Liege, Liege, Belgium)

Monnier’s hypothesis states that postprandial hyperglycemia contributes more to overall A1c for  patients at lower A1cs (~70% contribution at <7.3% A1c), whereas fasting hyperglycemia contributes more to overall A1c for patients at higher A1cs. Dr. André Scheen stated that Monnier’s group based this conclusion on four-point daytime glycemic profiles, and that a more recent study using seven-point glycemic profiles over 24 hours reported a uniformly high contribution of fasting hyperglycemia to total baseline hyperglycemia (76-80%) across all A1c values (Riddle et al., Diabetes 2010). Dr. Scheen sought to challenge Monnier’s hypothesis by performing a retrospective analysis of the initiation phase of the DURABLE study. In the original study, 2,091 patients with type 2 diabetes were randomized to insulin glargine (n=1,046) or insulin lispro mix 25 (LM25; n=1,045). Patients performed a 24-hour seven-point SMBG profile at baseline and week 24, which Dr. Scheen used to determine the contributions of fasting hyperglycemia (FHG) and postprandial hyperglycemia (PPHG) to area under the curve (AUC). Patients were stratified by baseline A1c into quartiles: <8.0%, 8.0% to 9.0%, 9.0% to 10.0%, and >10.0%. Across these four categories, the contribution of FHG increased from 59% to 73%, and the contribution of PPHG decreased from 41% to 27%, from low to high A1c. This confirmed Monnier’s hypothesis that the contribution of PPHG decreases with increasing A1c and that the contribution of FHG increases with increasing A1c. However, even in the lowest A1c quartile (<8.0%), the contribution of FHG (59%) was greater than the contribution of PPHG (41%), confirming Riddle’s findings as well. Additionally, Dr. Scheen examined responses to two different insulin regimens based on baseline A1c. As would be expected, the proportion of patients achieving a 7% A1c target decreased as baseline A1c increased. For patients taking basal insulin glargine, the decrease in A1c was accounted for by decreases in FHG with no effect on PPHG. For patients taking LM25, reductions in both FHG and PPHG contributed to A1c reductions. A greater number of patients taking LM25 reached target than patients taking glargine, but, unsurprisingly, at the expense of greater hypoglycemia and higher total insulin dose.


Oral Presentations: Hypoglycemia


Anne-Sophi Sejling, MD (Hillerød Hospital, Hillerød, Denmark)

Dr. Anne-Sophi Sejling presented results from a study investigating the association between all-cause mortality and 1) number of episodes of severe hypoglycemia; 2) number of episodes of mild hypoglycemia; and 3) hypoglycemia awareness status. Two hundred sixty five patients with type 1 diabetes were given surveys to ascertain baseline characteristics and one-year retrospective hypoglycemia information. Then, patients were followed prospectively for 11 years through monthly questionnaires. At baseline, patients had a mean A1c of 8.6%, age of 45, and diabetes duration of 21 years. Forty four percent of patients were hypoglycemia aware, 44% had impaired awareness, and 12% were unaware. At the conclusion of the study 39 patients had passed away. In our opinion, this small sample size insufficiently powered the statistical analysis, and as such, results need to be cautiously interpreted. Dr. Sejling found no significant associations between all-cause mortality and severe hypoglycemic episodes, mild hypoglycemic episodes, or hypoglycemia awareness status. Mild hypoglycemic awareness status was closest to statistical significance with p=0.09. Some variables were found to be significantly associated with all-cause mortality: 1) male gender (hazard ratio: 3.66; p=0.005); 2) diabetes duration (1.03; p=0.04); 3) macrovascular disease (2.49; p=0.03); 4) diabetic nephropathy (4.31; p <0.001); and 5) baseline A1c (1.38; p=0.03).

Questions and Answers

Comment: I am always concerned by mortality studies with low numbers. It seems like you have low numbers across the board.

Q: Are we doing any better with mortality over the last 12 years? Do you have power on that?

A: We do not, sorry.

Q: The striking thing is the link between nephropathy and mortality. Were patients with nephropathy more likely to experience hypoglycemia; was this a confounding effect?

A: Good question. There were not many patients with nephropathy, only about 10%, so we don’t have power to investigate that.

Comment: It is well known patients with nephropathy have hypoglycemia unawareness and experience more hypoglycemia.

Q: Did you see higher rates of nephropathy in males?

A: I do not think so, but I will have to go check the data.

Q: Did you se any association of accidents with hypoglycemia?

A: We have not investigated that yet.

Q: What about different treatment regiments?

A: That is an excellent question, we haven’t looked at that yet.



Abd Tahrani, MD (University of Birmingham, Birmingham, UK)

Dr. Abd Tahrani presented results from The Global Attitudes of Patients and Physicians 2 Survey (GAPP2), which was designed to 1) examine the frequency of self-treated hypoglycemia in basal insulin analog patients with type 2 diabetes; 2) assess how patients respond to self-treated hypoglycemic  events; and 3) assess how HCPs advise patients after self-treated hypoglycemic events. Dr. Tahrani explained that global data on self-treated hypoglycemia is limited (hence, GAPP2); however,   considering the survey’s low response rate and the disproportionate number of US participants, it seems that these findings represent neither the global picture, nor any specific region. Additionally, as Dr. Tahrani noted, because the survey was conducted over the Internet, young, employed patients were  likely overrepresented in the global population of people with type 2 diabetes on basal insulin analogs. Based on GAPP2 analysis, Dr. Tahrani concluded 1) self-treated hypoglycemia was common in patients with type 2 diabetes and 2) these events impact diabetes management, sometimes in inappropriate   ways (like letting blood glucose get high before bedtime to prevent nocturnal hypoglycemia). While we felt the study was not robust, we appreciated that Dr. Tahrani interpreted the results as reason for educational initiatives that will better inform patients about the appropriate responses to hypoglycemia and the potential harms of inappropriate therapy adjustments in response to hypoglycemia.

  • Of the 1,034,363 patients and 36,240 HCPs invited to participate in the survey, only 101,499 (~10%) and 5,115 responded (~14%), respectively. Of those who responded, 3,042 patients and 1,653 providers fulfilled enrollment criteria and had complete responses. Patients with type 2 diabetes treated with basal insulin analog therapy (48% bolus; 52% basal/bolus) were included. Patients took a mean 1.4 basal injections per day via prefilled pens (61%), vial and syringe (33%), or refillable pens (11%). HCPs included primary care physicians (42%), nurses and diabetes educators (26%), and specialists (32%). On average, providers saw 97 patients with type 2 diabetes per month.
  • The US was disproportionately represented. Of the six countries included (US, Canada, Japan, UK, Denmark, and Germany), 61% of patients and 28% of providers were from the US.
  • Eighty percent of patients reported ever having experienced a self-treated hypoglycemia event and 36% of patients reported a self-treated hypoglycemic event in the last 30 days. Of the 80% ever experiencing an event, 24% were nocturnal events and 76% were diurnal.
  • After a hypoglycemic event, 46% of patients had ever increased their blood glucose monitoring and more than 10% had ever altered their basal insulin regimen. We think it is unfortunate that hypoglycemia precedes more regular blood glucose testing and hope, as Dr. Tahrani suggested, that improved, structured educational initiatives could proactively encourage better testing. To reduce risk of nocturnal hypoglycemia, 14% of patients reported letting their blood glucose go high and 16% of patients reported not taking their insulin as prescribed – a serious problem in our view, that again, could be improved through education (or perhaps, better communication between patients and providers).
  • Only some HCPs advise patients who report self-treated hypoglycemia to alter diabetes management: following an event, “most of the time” 32% of providers would advise patients to change blood glucose monitoring, 19% would temporarily reduce basal insulin doses, 16% would reduce basal insulin long-term, and 3% would split the basal insulin dose. Obviously, discussions about what preceded the hypoglycemic event and other behavioral change recommendations can help prevent hypoglycemia, and it was unfortunate that these types of discussions were not captured in the survey. Dr. Tahrani’s presentation also made us think about the Living Diabetes: Journey for Control survey finding presented on Day #4 at Merck’s press conference that there appears to be a discrepancy between what providers claim to have told patients and what patients remember hearing – we wonder how many of these HCPs’ recommendations were actually put into practice by patients and whether the necessary conversations are even being had between patients and providers about hypoglycemia (for our discussion on the press conference, see page 36 of our Day #4 highlights at

Questions and Answers

Q: You approached over one million patients in different countries. How did you get ethical approval to send the survey?

A: We approached each country separately. The US and UK required approval, but the other countries did not.

Q: By inviting such a huge number of patients and HCPs and ending up with small percentage in the study, can you speculate on how selection might have played a role?

A: We didn’t know anything about the initial patients invites besides that the patients had diabetes. The ~13,000 were selected based on having an age above 40 years, type 2 diabetes, and meeting other inclusion criteria. You are right though, only 10% responded from a million at first.

Q: You mentioned education is important. Do you have any information about education in these patients?

A: The survey did not include any questions about education. We’d be interested especially in differences between countries, but we don’t have data.

Q: Did you have different outcomes in patients on basal vs. basal/bolus therapy, because I suspect the latter probably more educated on this.

A: Hypoglycemia events were mostly during the day suggesting basal/bolus was probably causing the problems, but we have not analyzed the data in that way.

Q: Do you have information on different insulin regiments or different phenotypes of patients?

A: We have their age and gender, but not their BMI. We tried to keep survey as minimal as possible because we wanted patients to complete the survey.


Oral Presentations: Mechanisms of Insulin Action


Agued Gonzalez-Rodriguez (Spanish Biomedical Research Centre in Diabetes and Associated Metabolic Disorders, Madrid, Spain)

Dr. Agued Gonzalez-Rodriguez described an experimental model that tested whether the favorable effects of resveratrol in mice lacking the insulin receptor substrate 2 (IRS2) are mediated through an increase in SIRT1. Previous data show that IRS2 knockout mice develop hyperglycemia and that resveratrol increases insulin sensitivity by decreasing levels of PTP1B (which blocks insulin signaling). To determine whether these effects of resveratrol (an established SIRT1 activator) were due to its effects on increasing SIRT1, Dr. Gonzalez-Rodriguez’s group evaluated glucose homeostasis, insulin sensitivity, and insulin signaling in IRS2 knockout mice treated with resveratrol vs. IRS2 knockout mice generated to overexpress SIRT1. The results showed that resveratrol, but not the overexpression of SIRT1, decreased PTP1B levels and improved insulin sensitivity. The study indicates that in this particular mouse model, the effects of resveratrol on insulin action are independent of SIRT1 elevation.

Questions and Answers

Q: What do you suggest is the next step, and what is the translational relevance of where you see your experiments going forward?

A: I think the relevance of resveratrol is that resveratrol’s effects on insulin signaling is due to decreased PTP1b, not because of SIRT1 activation. This has relevance because the inhibition of PTP1b has been receiving increasing attention, potentially as a therapeutic target.

Q: This is an interesting paper. It’s very exciting. Resveratrol is very big in the anti-aging field and a lot of these pathways have been studied in simple organisms. Does this pathway explain any part, or all of, resveratrol’s anti-aging effects, or is it unrelated?

A: I think it could be due to the fact that the survival path is regulated through these pathways too. Tyrosine receptors or growth factors signal through IRS1, and this impacts survival.

Q: You said that you increased SIRT1 activation, but you only overexpressed it and assumed it was active. Did you actually have a positive control that shows that SIRT1 had the predicted effects in other pathways?

A: We are trying to perform the SIRT1 activity assay to reinforce these results. We are trying to do it, and we currently don’t have more information about it SIRT1 activation.


Posters: Novel Insulins


Douglas Muchmore and Daniel Vaughn (Halozyme Therapeutics, San Diego, CA); Linda Morrow and Marcus Hompesch (Profil Institute for Clinical Research, Chula Vista, CA)

This poster summarized the full results from a randomized, double blind, crossover study using Halozyme’s PH20 in insulin pumpers with type 1 diabetes. It tested both an aspart-PH20 coformulation (n=20; 16 completers) and pre-administration of a single PH20 injection at the time of infusion set change (n=25; 17 completers). Compared to aspart alone, both PH20 strategies significantly accelerated insulin absorption, reduced the duration of insulin action (i.e., a “fast on, fast off” profile), and improved post-prandial glucose control. Unfortunately, it’s hard to directly compare the pre-administration data to the coformulation data since each approach was tested in a different cohort of patients and the experimental designs differed slightly around mealtimes. Adverse events were comparable between all study arms, generally mild, and typical of the euglycemic clamp procedure. We look forward to seeing the results of Halozyme’s long-term phase 4 insulin pump study testing pre-administration of PH20 – it recently received FDA and IRB approval to begin enrolling patients (see our coverage of Halozyme’s recent 2012 Analyst Day at Additionally, given the excellent data thus far in type 1 and type 2 MDIs (see pages 210 and 219 of our ADA 2012 report at as well as type 1 pumpers in this study, we hope Halozyme can find a partner for its coformulation program.

  • This randomized, double blind, crossover design study examined use of Halozyme’s PH20 in insulin pumpers with type 1 diabetes. Both an aspart-PH20 coformulation and pre-administration of a single PH20 injection prior to infusion set insertion were studied. The study was divided into two cohorts:
    • Aspart-PH20 coformulation (Cohort 1): Twenty type 1 pumpers (mean age: 37 years, mean BMI: 26 kg/m2; 16 completers) used both Novolog insulin alone and an aspart-PH20 coformulation in the crossover design. New infusion sets (Quickset) were inserted upon admission (5 pm on day one) and study drug was infused for 3.5 days. Euglycemic clamps following bolus infusion of insulin aspart (0.15 units per kg) were conducted on the mornings of days two and four (0.5 days and 3.5 days of continuous use). Standardized solid test meal challenges were performed for all dinners to assess glycemic response. Patients were readmitted to the CRC 5-14 days after discharge and crossed over for repeat procedures with the alternate treatment.
    • PH20 preadministration (Cohort 2): Twenty-five type 1 pumpers (mean age: 34 years, mean BMI: 27 kg/m2; 17 completers) used both Novolog insulin/a single pre- administration injection of PH20 and Novolog insulin/a sham injection at the time of infusion set change in the crossover design. New infusion sets (Quickset) were inserted upon admission (5 pm on day one). Novolog was infused overnight and a second new infusion set was inserted ± PH20 the next morning (5 am on day two) followed by 3.5 days of Novolog infusion. Euglycemic clamps following bolus infusion of insulin aspart (0.15 units per kg) were conducted on the mornings of days two, three, and five (two, 26, and 74 hours of continuous use). Standardized solid test meal challenges were performed for all dinners to assess glycemic response with IV insulin pretreatment to stabilize blood glucose to 110 mg/dl. Patients were readmitted to the CRC 5-14 days after discharge and crossed over for repeat procedures with the alternate treatment.
  • The aspart-PH20 coformulation and pre-administration of PH20 both improved postprandial glucose (PPG) control over aspart alone. Although it appears that the aspart-PH20 coformulation had a more impressive postprandial benefit than PH20 pre- administration, we note that these results are not directly comparable. As noted above, the coformulation test meals were done without normalizing preprandial glucose, while the pre- administration test meals normalized glucose to 110 mg/dl using IV insulin. The latter is more robust and rigorous. We wonder how the results would have differed if both experiments had used the same methodology.


Aspart-PH20 Coformulation

Aspart Alone

PH20 Pre-


Aspart Alone

1 hr PPG

125 mg/dl

145 mg/dl (p=0.006)

135 mg/dl

140 mg/dl (p=0.37)

90 mins PPG

118 mg/dl

146 mg/dl (p=0.006)

119 mg/dl

133 mg/dl (p=0.055)

2 hrs PPG

121 mg/dl

146 mg/dl (p=0.017)

112 mg/dl

126 mg/dl (p=0.098)

Glycemic Excursion from Pre-Meal Baseline

1 hr

+2 mg/dl

+25 mg/dl (p=0.005)

+29 mg/dl

+40 mg/dl (p=0.077)

90 mins

-5 mg/dl

+27 mg/dl (p=0.017)

+13 mg/dl

+33 mg/dl (p=0.007)

2 hrs

-2 mg/dl

+26 mg/dl (p=0.044)

+6 mg/dl

+25 mg/dl (p=0.020)

  • Insulin aspart absorption and action profiles varied with infusion site age, a phenomenon that was eliminated with PH20 pre-administration. For aspart alone, the fraction of insulin exposure occurring within one hour following bolus infusion was 16% for a new infusion site and 27% after three days of use (p <0.0001). By contrast, one-hour insulin exposure with PH20 pre-administration was consistent over time: 32.1% with a new infusion set and 32.6% after three days of use (p=0.85). This is very encouraging from a patient perspective, since PH20- preadministration would provide more consistent and predictable insulin absorption over time. We suspect few patients are aware of these fairly big differences in insulin absorption over time, effects that change insulin action timing quite meaningfully.
  • The aspart-PH20 coformulation significantly accelerated both insulin pharmacokinetics (PK) and glucodynamics over the life of the infusion set. Both aspart alone and the aspart-PH20 coformulation had significantly faster absorption profiles on  day 2.5 vs. day 0.5. However, the aspart-PH20 coformulation was faster than aspart alone at both the beginning and end of the study.

PK and Glucodynamics of Aspart-PH20 Coformulation vs. Aspart Alone 

(Day 0.5 → Day 2.5)


Aspart-PH20 Coformulation

Aspart Alone

AUC0-1 hr

(% Total)

35.0 51.0 (p=0.002)

21.0 32.8 (p=0.007)

Early t50%


17.4 10.9 (p=0.015) 26.0 16.9 (p=0.02)

Duration of Insulin Action (mins)

146.5 132.6 (p=0.005) 163.6 147.5 (p=0.022)
  • PH20 pre-administration had consistently accelerated insulin pharmacokinetics (PK) and glucodynamics over the life of the infusion set. In other words, PH20 pre- administration started fast and stayed fast until the end of the study. By contrast, the absorption of aspart alone started slow and got faster over the course of the study; however, it was still slightly slower than PH20 pre-administration by the end of the study.

PK and Glucodynamics of PH20 Pre-administration vs. Aspart Alone

(Two  hours  → Three  Days)


PH20 Pre-


Aspart Alone

AUC0-1 hr

(% Total)

32.1 32.6 (p=0.85)

15.7 27.3

(p <0.001)

Early t50%


23.4 21.4 (p=0.35)

35.6 24.4

(p <0.001)

Duration of Insulin Action (mins)

138.6 145.8 (p=0.24) 180.4  156.1 (p=0.001)


Symposium: Michael Berger Debate: Early Insulin Treatment of Type 2 Diabetes: Benefit or Hazard?


Jack Leahy, MD (University of Vermont College of Medicine, Burlington, VT)

Dr. Jack Leahy presented a well-structured case in which he championed the use of early basal insulin treatment by making four arguments: 1) while all anti-diabetic drugs provide better glycemic control when used early during treatment, basal insulin provides greater glycemic efficacy compared to several other therapies – in particular, basal insulin as an add-on to metformin appears to be a powerful treatment that minimizes the concern of hypoglycemia; 2) early basal insulin therapy is safe: ORIGIN results indicate that the use of insulin glargine in people with early-stage diabetes does not confer a greater risk of cardiovascular disease or cancer; 3) basal insulin may provide beta cell protection; and 4) after starting basal insulin, HCPs can use alternatives to basal/bolus therapy to further intensify treatment – Dr. Leahy focused on the addition of GLP-1 agonists to basal insulin (which provides   further A1c reductions and improvements in weight, with little risk for hypoglycemia), as well as “basal- plus therapy”, the use of rapid-acting insulin at only one meal each day.

  • Dr. Leahy advocated for the efficacy of early insulinization by reviewing several study results. A systematic review of 29 trials found that 40% of patients on early basal insulin therapy achieve an A1c of <7%. Furthermore, a meta-analysis of GLP-1 vs. insulin showed that both treatments provide similarly effective glycemic control (Yang et al., Diabetes Obes Metab 2011). Dr. Leahy argued that patients experience greater glycemic control at a lower baseline A1c level (i.e., early on in the disease progression): a pooled analysis of 2,193 patients taking insulin glargine with oral agents found that 75% of patients with an A1c <8% reached their target. Furthermore, data show that patients with a lower baseline A1c level have less risk of weight gain with insulin, suggesting that the weight side effect of insulin is less pronounced with early insulinization (Leahy et al., EASD 2011).
  • Dr. Leahy noted that while all anti-diabetic drugs provide better glycemic control when used early in the treatment process, basal insulin provides greater glycemic efficacy compared to several other therapies. A pooled analysis of 11 trials that included over 2,000 patients treated with glargine and other therapies found that participants on glargine experiences a greater A1c reduction compared to those on zero or one oral agent. In addition, 70% of those taking glargine with metformin reached their target treatment goal. These patients experienced a low risk of hypoglycemia, which to Dr. Leahy illustrates that early insulin use with just metformin is a powerful treatment tool that minimizes many concerns typically associated with insulin therapy.
  • To finish his discussion on the efficacy of early insulinization, Dr. Leahy turned to the ORIGIN trial, noting that the when used in a study population with prediabetes and early type 2 diabetes (average A1c of 6.4%), intensive glargine therapy provided excellent glycemic control (in our view, some may argue that since standard care provided nearly as good glycemic control, basal insulin therapy may not be necessary in this population; for our detailed coverage of the ORIGIN results, please see our ADA ORIGIN report at
  • Dr. Leahy then argued that early insulinization is not only efficacious, but also safe. He again referenced the ORIGIN trial, noting that investigators found no influence of insulin glargine therapy on cardiovascular outcomes and cancer, and that the trial maintained a strong differentiation between the two treatment arms: at study end, 80% of people in the glargine arm remained on insulin while only 11% of those in the standard care arm had progressed to insulin therapy. We note that during his talk on diabetes and cancer at EASD 2012, Dr. Jeffrey Johnson (University of Alberta, Alberta, Canada) reasoned that because both arms included patients on insulin, as well as on a range of oral anti-diabetic agents, ORIGIN can only indicate that the specific combination of glucose lowering therapies used in the glargine arm (and not glargine alone) is not associated with an increased risk of cancer. To conclude his point on safety, Dr. Leahy noted that though insulin has been found to result in greater nocturnal hypoglycemia compared to exenatide (Heine et al., Ann Intern Med 2005), the clinical significance of this finding is currently uncertain.

  • Dr. Leahy proposed that insulin provides beta cell protection, since the insulin receptor lies upstream of a well defined signaling pathway that influences beta cell size, proliferation, and differentiation. Two studies in rats have illustrated insulin’s effects on beta cells (Jetton et al., Diabetologia 2001 and Jettton et al., Diabetes 2005). Dr. Leahy also pointed to a Chinese study which found that among ~400 patients with newly diagnosed type 2 diabetes who underwent intensive insulin therapy or oral agents for two weeks, the remission rate for those on insulin nearly doubled that for people on oral agents (Weng et al., Lancet 2008). Furthermore, ORIGIN found that insulin slowed the progression from prediabetes to diabetes compared to standard care.

  • Dr. Leahy’s fourth and last argument stated that after starting basal insulin therapy, HCPs can use alternatives to basal/bolus therapy to further intensify treatment. Dr. Leahy discussed two already-studied options: 1) adding a GLP-1 agonist, which has been found to further decrease A1c levels without weight gain, and with little risk of hypoglycemia; and 2) “basal-plus therapy”, where a single injection of meal-time insulin is given only for the largest meal of the day. Dr. Leahy’s discussion of these two approaches was quite well timed, as during Sanofi’s panel discussion, several KOLS noted that HCPs’ and patients’ negative views on insulin therapy (including a resistance to injections and the belief that insulin represents the “last resort” option) represent a significant barrier to early insulin use (details on the panel discussion are available in our EASD Day #4 report at



Guntram Schernthaner, MD (University of Vienna, Vienna, Austria)

In his fast-paced presentation, Dr. Guntram Schernthaner passionately argued that there is little to no added benefit with early insulinization, but several drawbacks. He began by pointing out that most people are already obese at the time of diagnosis, that insulin is associated with further weight gain,  and that weight gain may lower patients’ glycemic response to treatment. Dr. Schernthaner reminded the audience that insulin therapy leads to higher rates of hypoglycemia, which itself is associated with an increased risk for all-cause and cardiovascular mortality. In addition, longer periods of insulin therapy have been linked to increased risk for cancer, though a causal relationship has not been establish. In addition, Dr. Schernthaner argued that early insulin therapy would provide little benefit in glycemic control, and has not been shown to be superior to other diabetes therapies in reducing macrovascular  disease.

  • Dr. Schernthaner opened his presentation by referencing the ORIGIN trial. He noted that at year six of the ORIGIN trial, investigators observed only a small difference in A1c (0.3%, or 9 mg/dl) between the glargine arm and standard care arm. However, glargine therapy resulted in  a higher rate of confirmed hypoglycemia and a nearly three-fold increase in the rate of severe hypoglycemia compared to standard care.
  • Dr. Schernthaner noted that most newly-diagnosed patients are already obese, and that insulin therapy likely leads to further weight gain. Data indicate that obese diabetes patients have a reduced glycemic response to insulin compared to their non-obese counterparts.
  • Dr. Schernthaner reminded the audience that higher rates of sever hypoglycemia were observed with basal insulin in the DURABLE study (Buse et al., Diabetes Care 2011) and the ORIGIN trial. Previous data (including from those from VADT, ACCORD, ADVANCE, and ORIGIN) reveal a relationship between A1c level and severe hypoglycemia. Low baseline A1c levels are associated with a greater risk of hypoglycemia, which itself is associated with an increased risk for all-cause mortality and cardiovascular mortality, as shown in the NICE SUGAR study (our coverage of the results is available at
  • Dr. Schernthaner reviewed previous studies to demonstrate that early insulin therapy would provide little glycemic benefit. At study in newly-diagnosed diabetes patients found no significant difference in A1c, weight change, or hypoglycemia risk in people taking insulin and metformin vs. those on triple oral therapy. Furthermore, the triple oral therapy included glipizide, a drug associated with greater weight gain and hypoglycemia (Lingvay et al., Diabetes Care 2009). Dr. Schernthaner questioned the implications of the Weng study (where newly-diagnosed patients experienced remission after two weeks of intensive insulin therapy), stating that the study participants were not obese and that the follow-up period lasted only 12 months.



Jack Leahy, MD (University of Vermont College of Medicine, Burlington, VT) and Guntram Schernthaner, MD (University of Vienna, Vienna, Austria)

Q: What do you mean by early insulin therapy? Early could be taken as being when beta cell reserve is partial soon after diagnosis, or it could be taken as meaning before A1c deteriorates while having used three oral agents.

Dr. Leahy: For me, the definition is adding insulin before the beta cell failure progresses beyond an existing point with certain therapy. I think that in many ways, we can think about using basal insulin after metformin, but that’s certainly not appropriate approach for everybody. We live in a world with many doctors using different therapies and using them effectively. I see patients with an A1c over 10% on many oral agents and with other comorbidities. For me, it’s that you can try one drug or two drugs, but if you don’t get control, don’t wait until the A1c gets very high to start insulin.

Q: Dr. Leahy, what is the point in putting the patient on just basal insulin. In your study, when you started insulin, you either added a GLP-1 agonist or you put them on prandial insulin. Is there a point to starting basal insulin with oral anti-diabetic agents such as a DPP-4 inhibitor?

Dr. Leahy: Absolutely, yes. What I showed you were two topics of conversation in the diabetes world, but they’re not exclusive. So certainly, I don’t think we have a lot of information on adding DPP-4 inhibitors to insulin, and the information out there is disappointing. However, there are certainly many ongoing   studies to try to understand it as a first line, second line, and third line therapy. I think that going back,  oral anti-diabetic agents can be added to basal insulin.

Dr. Schernthaner: By diagnosis and early type 2 diabetes, most patients are obese, so we have to use insulin in non-obese patients. In general, the use of basal insulin very early would be against the new ADA/EASD guidance that clearly states that insulin should be used when A1c is higher. Before that you have no benefits, only costs.

Dr. Leahy: We can’t assume that there is one mode of therapy. The problem is that so many HCPs believe that the most important goal in diabetes therapy is avoiding insulin at all costs and in al circumstances, and they think that they are doing their patient a favor even if the patient’s A1c gets very high, as long as the patient isn’t on insulin. Would I start insulin as the first drug? No. Would I use it on top of   metformin? That’s not uncommon. You get excellence A1c lowering, fewer problems with hypoglycemia, and minimal weight gain. And it’s not a hard therapy for the average patient – one injection a day. Once it’s presented in a positive way, many patients feel good about it. I’m not saying it’s the only approach, but it should be offered to patients in a fair and balanced way.

Q: Your presented hinted that after a while, patients end up on a combination therapy. Do you have any evidence that says that if patients start basal insulin earlier, they end up using less therapies later on?

Dr. Leahy: I don’t think there is any evidence. I’m impressed with ORIGIN because after six years of therapy, the majority of people are still just on metformin with basal insulin with outstanding glycemic control. There was also outstanding control with oral anti-diabetic medications, but there was a little   bleed through into other agents. I agree that no DPP-4 inhibitors were used. In my consulting practice,   the people coming to me are people who are started on metformin and were then given a sulfonylurea and then a TZD and then a GLP-1. They make it to me with four, five, or six years of the disease and the A1c is above 9%. There are lots of those patient sin my country that could have been stopped in the course of diabetes if someone had thought about insulin, and that’s the message I’m trying to get across.


Symposium: The ORIGIN Trial: Effects of Basal Insulin Glargine and of Omega 3 Fatty Acid Supplements on Health Outcomes in People With Dysglycemia


Hertzel Gerstein, MD (McMaster University, Hamilton, Canada)

Dr. Hertzel Gerstein reviewed the results of the ORIGIN trial, highlighting the neutral effect of insulin glargine on cardiovascular outcomes and cancer over a median 6.2-year follow-up, the reduction of progression to type 2 diabetes with insulin glargine, and modest increases in hypoglycemia and weight with insulin glargine. He also noted that the trial showed that insulin therapy could be safely discontinued even after a period of extended use. For full commentary on the ORIGIN trial, please see our Closer Look at



Matthew Riddle, MD (Oregon Health and Science University, Portland, OR)

Dr. Matthew Riddle presented new (if minor, in our view) analyses from the ORIGIN trial on baseline characteristics of the subgroups of patients without and with diabetes, glycemic responses over time by subgroup, and predictors of maintaining A1c less than 6.5%. The main independent predictors of maintaining mean A1c less than 6.5% up to five years were diabetes status, baseline A1c, alcohol use (the positive effect was surprising), and glargine use. In conclusion, Dr. Riddle stated: 1) target-directed intervention early in dysglycemia can maintain baseline A1c levels for at least five years; 2) glargine- based regimen is more likely than standard care to keep A1c under 6.5%; and 3) more data and further analyses are needed to define the benefits versus risks of these approaches.

  • Patients without diabetes and patients with diabetes had similar age and BMI at baseline, but varied on a number of different characteristics, including A1c, and fasting plasma glucose.





No diabetes


No diabetes







Female (%)





Age (years)





BMI (kg/m2)





Median A1c (%)





Median FPG (mmol/l)









Hypertension (%)





Urine albumin/creatinine









Prior CV event (%)





Beta blocker use (%)









Statin use (%)





LDL (mmol/l)





Alcohol >2x/wk (%)









  • Dr. Riddle highlighted the changes in glycemic control over time by subgroup. Insulin glargine treatment brought fasting plasma glucose down to approximately 5.0 mmol/l (90 mg/dl) for patients without diabetes, and below 5.3 mmol/l (95 mg/dl) for patients with diabetes; A1c remained largely unchanged. A higher percentage of patients without diabetes at baseline achieved A1c <6.5%; a higher percentage of those on insulin glargine treatment achieved this target A1c.

Percentage of patients with A1c <6.5% at one and five years



One year

Five years

No diabetes





Standard care









Standard care




  • The main independent predictors of maintaining mean A1c less than 6.5% up to five years were diabetes status, baseline A1c, alcohol use, and glargine use. Diabetes  versus no diabetes was associated with an odds ratio of 0.31 (95% CI: 0.24-0.40; p <0.001), baseline A1c (per 1%) was associated with an odds ratio of 0.19 (95% CI: 0.18-0.21; p <0.001), modest use of alcohol (at least two times per week) was associated with an odds ratio of 1.61 (95% CI: 1.41-1.84; p <0.001) (this was characterized as puzzling), and glargine use versus standard therapy was associated with an odds ratio of 2.98 (95% CI: 2.67-3.32); p <0.001).



Thomas Pieber, MD (University Hospital of Graz, Graz, Austria)

Providing independent commentary on the ORIGIN trial, Dr. Thomas Pieber stated that insulin glargine treatment does not appear to prevent diabetes to a meaningful degree. This had never been disputed when the original results were given, so we were a little surprised by the emphasis by Dr. Pieber. In addition, no benefits were seen in morbidity, mortality, or quality of life with insulin glargine    treatment. As such, he thought there isn’t necessarily a good reason why patients with prediabetes  should use insulin glargine, when it doesn’t confer any clinical benefits for them, and likely causes severe hypoglycemia three times as frequently as standard care (information on the frequency of severe hypoglycemia in patients with prediabetes was assumed to be the same as in the overall study population). This, of course, had been the original question in the trial, and had been addressed in the original talks on ORIGIN at ADA. Dr. Pieber also emphasized the need to further examine the   unexpected high mortality rate in ORIGIN.

  • Dr. Pieber emphasized that the unexpected high mortality rates in ORIGIN urgently needs additional analyses. In the ORIGIN trial, the mortality rate was 2.57% per year, which Dr. Pieber noted was higher than the mortality rate in other large outcomes trials – the ADDITION, ADVANCE, PROactive, and ACCORD trials, had mortality rates of 1.27%, 1.69%, 1.41%, and 1.41% per year, respectively (mortality rates were not stated explicitly in the publications for ADDITION, ADVANCE, and PROactive; Dr. Pieber estimated them based on the available data in publications). He noted that differences in age, duration of diabetes, history of a prior cardiovascular event, and smoking status could not explain the observed high rate of mortality.
  • He highlighted that insulin glargine caused a threefold increase in severe hypoglycemia in ORIGIN; 6% of patients receiving insulin glargine experienced a severe hypoglycemic event, compared to 2% of those receiving placebo. Dr. Pieber noted that the number needed to harm was 25, that is, 25 patients would have to be treated with insulin glargine to cause one severe hypoglycemic event. In one study, it was shown that 94.7% of all endocrine emergency hospitalizations are caused by hypoglycemia. Dr. Pieber commented that there doesn’t seem to be a good reason why patients should use an agent that causes hypoglycemia when they don’t confer any clinical benefit in morbidity, mortality, or quality of life (and when they cause weight gain).
  • Dr. Pieber stated that insulin glargine use in patients with impaired glucose tolerance is not effective in preventing type 2 diabetes. He commented that though the 28% risk reduction in ORIGIN was quite impressive, it was relative risk reduction, not absolute risk reduction; absolute risk reduction was 6.5%. Dr. Pieber pointed out that 15 patients would have to be treated with insulin glargine for six years to prevent one new case of type 2 diabetes; the 14 other patients treated with insulin glargine for six years would have progressed to type 2 diabetes.


Symposium: Where Are the Insulin Resistance Gene Loci?


Ele Ferrannini, MD (University of Pisa School of Medicine, Province of Pisa, Italy)

After he provided a comprehensive overview of the in vivo effects on insulin, Dr. Ele Ferrannini argued that it is important to consider non-classical insulin effects when looking for genetic and environmental determinants in insulin resistance. He defined classical insulin effects as the action of insulin on glucose metabolism, and non-classical insulin effects as the action of insulin on other metabolic and hemodynamic processes. Through a series of clamp studies, Dr. Ferrannini demonstrated the many physiological effects of insulin on beta cells, thermogenesis, kidney function, endothelial tissue, midbrain processes, and dyslipidemia, to name a few. Certainly, Dr. Ferrannini made clear in his presentation  that insulin sensitivity (and insulin resistance) has whole body effects.


Symposium: Hypoglycemia and Mortality in Diabetes


Sophia Zoungas, PhD, MBBS (Monash University, Victoria, Australia)

Dr. Sophia Zoungas presented on severe hypoglycemia in the ACCORD, ADVANCE, and VADT studies. We appreciated that Dr. Zoungas concluded her presentation with a review of research insights from ACCORD, ADVANCE, and VADT analysis, but even more so, that she concluded with the clinical implications: 1) choose approaches to glucose lowering that minimize risk of severe hypoglycemia; 2) ensure patients are educated about avoidance and management of hypoglycemia; 3) experience of severe hypoglycemia should lead to an examination of comorbid diseases that may produce adverse outcomes; and 4) persistent attempts to intensify therapy in non-responsive high risk patients should be avoided. Dr. Zoungas’ presentation added to the discussion of the role of severe hypoglycemia in excess mortality in a thoughtful manner – she suggested that severe hypoglycemia could be both a direct cause of severe hypoglycemia and a marker for other comorbid diseases that cause the excess mortality. Certainly, the discussion of cause vs. association is far from over, but as Dr. Zoungas’ clinical tie-in at  the end of her presentation reminded us, the most important thing is that severe hypoglycemia is being seriously considered and appropriately addressed in the clinical setting.

  • Patients in the VADT had a longer duration of diabetes and had a greater percentage of patients entering the trial already on insulin therapy. It follows then, that by the end  of the study, in the intensive treatment by the end of the trial a greater percentage of patients   were on insulin therapy compared to in ACCORD or ADVANCE. Dr. Zoungas thought the differences in diabetes duration and insulin therapy could help explain the differing rates   observed in severe hypoglycemia between the trials.


Insulin Therapy at End

Severe  Hypoglycemia



Duration of Diabetes (years)

Insulin Therapy at Start






























  • Dr. Zoungas believes that the different titration patterns account for the differing patterns in severe hypoglycemia between the ACCORD and ADVANCE trials. Dr. Zoungas explained that in the intensive ACCORD arm, therapy was rapidly titrated. As such, the highest rates of severe hypoglycemia occurred early in the trial. In ADVANCE however, titration was much slower (i.e., started on oral agents, then basal, then more complex insulin regiments) and the higher rates of hypoglycemia were thus observed later on.
  • Dr. Zoungas argued that the majority of cases of severe hypoglycemia could be potentially be prevented with education. She began by reviewing the most frequent immediate antecedents to severe hypoglycemia recorded in ACCORD: 1) variation in food intake (i.e., missed or delayed meal) preceded 48% of events; 2) exercise (either unexpected or more vigorous than expected) preceded 15% of events; and 3) incorrect insulin use preceded 9% of events. Variation in food intake and exercise made for antecedents that could be mediated through education. Moreover, if patients were better educated on the frequent symptoms associated with hypoglycemia then they could know what to look out for and prevent the severe episodes.
  • Dr. Zoungas believes that severe hypoglycemia cannot be ruled out as a direct cause of increased mortality. ADVANCE showed no close temporal relationship between severe hypoglycemic events and adverse endpoints and no dose response (where patients with mild hypoglycemia would have had an increased risk of adverse outcomes) and ACCORD investigators did not attribute excess mortality to hypoglycemia; however, Dr. Zoungas argued that in these trials there is no way to account fully for the downstream cascade of pathophysiological consequences resulting from severe hypoglycemia.
  • Severe hypoglycemia could also be a marker of risk. Dr. Zoungas suggested that severe hypoglycemic events may reflect the effects of co-morbid diseases that increase a patient’s vulnerability to both severe hypoglycemia and adverse clinical outcomes. Importantly, Dr. Zoungas emphasized that viewing severe hypoglycemia as a cause of excess mortality and as a marker of risk for a comorbid disease causing excess mortality were not mutually exclusive.
    • Consequently, severe hypoglycemia should lead to an examination of patients for comorbid diseases. Providers shouldn’t just alter therapy in response to severe hypoglycemia, she said, but should first think more broadly about why the patient is presenting with severe hypoglycemia at that time. We very much appreciated this perspective.
  • Persistent attempts to intensify therapy in non-responsive high risk patients should be avoided. While she seemed to have disagreed with the ACCORD investigators on the role of severe hypoglycemia in mortality, on this clinical implication they certainly agreed.
  • The relationship between severe hypoglycemia and adverse events may be different for younger patients with type 1 diabetes than the comparatively older patients with type 2 diabetes analyzed in these studies. She suggested that younger patients with type 1 may be able to compensate better for the negative effects of severe hypoglycemia.

Questions and Answers

Q: Can you cut the data to look at the effect of sulfonylureas?

A: These were not monotherapy studies; the majority of patients are on complex algorithms so the proportion of patients on a single agent is small. We’ve looked at SFUs as an independent variable, but in fact when you put it in a model with other variables it seems to be accounted for by renal function and other factors. The SFUs did not seem to be associated with increased mortality

Q: In someone who is hypoglycemia naïve, is that first episode more dangerous?

A: In ADVANCE there were so few reported repeated episodes that there’s no sensitivity to examine  effects. So there are small numbers and we can’t make conclusions with any confidence about whether repeated or isolated event have any association on mortality. By talking with other trials, maybe we can do that in the future.

Comment: You didn’t consider the ORIGIN trial. Why not start early and keep A1c down from the beginning? In this case even treatment with SFUs does not increase hypoglycemia because patients will still have counter regulatory mechanisms.

A: I’ll take that as a comment, thank you.



Simon Heller, MD, FRCP (University of Sheffield, Sheffield, UK)

Dr. Simon Heller reviewed severe hypoglycemia and “dead in bed syndrome,” focusing mostly on cardiac-related experimental evidence. Dr. Heller emphasized that dead in bed is a clinically relevant problem – citing data from Thordarson and Sovik, Diabet Med 1995, dead in bed syndrome appears to occur in 6% of all deaths in people with diabetes younger than 40 years old. Indeed, Dr. Heller noted that his clinic has lost two young patients to nocturnal hypoglycemia. He spent most of his presentation discussing some of the experimental data linking hypoglycemia to death, including 1) a prolonged QT interval; 2) rises in circulating adrenaline concentrations and falls in circulating potassium levels caused by insulin induced hypoglycemia; 3) increased activity in sympathoadrenal fibers innervating the heart; 4) subclinical or overt autonomic neuropathy within the heart; and 5) mutations or polymorphisms affecting the relevant proteins concerned with the cardiac action potential. Interestingly, more males seem to die from dead in bed syndrome than females – Dr. Heller hypothesized that this was due to increased vagal tone in males. In closing, he reemphasized that dead in bed is relevant clinically, appears to be associated with nocturnal hypoglycemia, and warrants further  investigation.

Questions and Answers

Q: What is the effect of hypoglycemia on late changes in the autonomic nervous system? The sympathetic defect remains 16 hours after induction of hypoglycemia.

A: You’re talking about Freeman and colleagues in Boston. In non-diabetic individuals they induced hypoglycemia and depressed the sympathetic responses 24 hours later. You could imagine the vagal response might be unbalanced – that’s another potential mechanism that needs to be looked at.

Q: You talked about the importance of autonomic neuropathy. What do we know about interference with beta blockers?

A: Good question. It’s important to remember that these individuals have short duration diabetes. Fewer have evidence of neuropathy. We have done a study using beta blockers in people with diabetes, and in an experimental situation, we could obliterate the QT interval. On other hand, if bradycardia is causing a problem, it’s not a great idea. Before we advocate using them, we need more investigation.

Q: Regarding type 1 diabetes, is the use of human insulin as a basal insulin (vs. analogs) responsible?

A: In terms of causing this problem, it’s irrelevant as far as I can tell. We’ve had a 16-year old and a 22- year old die on human insulin. But we emailed two patients treated with analog insulin that also ended up dying.


Corporate Symposium: Realizing the Potential of Diabetes Therapy (Sponsored by Novo Nordisk)


Peter Kurtzhals, MD (Novo Nordisk, Copenhagen, Denmark)

Dr. Peter Kurtzhals, Senior Vice President and Head of Diabetes Research at Novo Nordisk, provided a concise and comprehensive overview of the thought process behind the development of insulin degludec. The Novo Nordisk team set out to engineer a novel mechanism of prolonging an insulin’s half-life, By fusing a fatty acid chain to the naturally occurring insulin hexamer, researchers were able to induce multihexamer formation, resulting in long strands containing thousands of degludec hexamers. Each hexamer is held together by a zinc ion at its core, and as the zinc ions diffuse from the ends of the multihexamer chains, insulin hexamers dissociate into monomers, allowing for a slow and steady release of insulin. Dr. Kurtzhals concluded by noting that while degludec’s unique design gives it an ultra-long, flat, and predictable PK and PD profile, degludec still has the same receptor binding properties and metabolism as human insulin.

  • Dr. Kurtzhals provided a detailed explanation of the biochemical design behind insulin degludec. Insulin degludec is injected subcutaneously in dihexameric units linked together by accessory fatty-acid side chains. This pharmacological formulation also includes carefully chosen levels of zinc and phenol. As phenol becomes quickly diluted following injection, the dihexameric conformation of insulin shifts such that long, multihexameric chains are able to form. This sudden drop in phenol concentration is followed by a slow and gradual decline in zinc levels. Zinc, which holds together individual hexamers, becomes diluted in solution, which causes the release of insulin monomers at a rate purportedly even slower than other available basal insulins.



Tim Heise, MD (Profil Institut for Metabolic Research, Neuss, Germany)

Acknowledging that many basal insulins are already on the market, Dr. Heise made the case for insulin degludec by stating that degludec offers the greatest potential benefit when designing basal insulin therapy in a variety of cases. Dr. Heise described three features of the ideal basal insulin: 1) longer duration of action to control fasting blood glucose with only one daily injection in all individuals; 2) flat time-action profile to lower the risk of hypoglycemia at the same glycemic level; and 3) less day-to-day variability to minimize glucose excursions and provide more predictable control, facilitating titration. Insulin degludec, according to Dr. Heise, has a half-life twice as long as insulin glargine, and consequently, plasma glucose levels decline more slowly over time. According to a poster presented at EASD this year, patients receiving insulin degludec reach steady state (90% of plateau) in fewer than three days (Coester et al., Poster 909). Ultimately, Dr. Heise concluded that the pharmacological characteristics of insulin degludec may facilitate titration to lower blood glucose targets, potentially allow for flexibility in dose administration time, and may lower the risk of hypoglycemia and hyperglycemia.



Chantal Mathieu, MD, PhD (University of Leuven, Leuven, Belgium) and Martin Abrahamson, MD (Harvard Medical School, Boston, MA)

During this team presentation, Dr. Martin Abrahamson discussed the medical consequences of hypoglycemia while Dr. Chantal Mathieu detailed its socioeconomic impacts. Dr. Abrahamson emphasized that both patients with type 1 and type 2 diabetes experience frequent hypoglycemic events, and that increasing the duration of insulin therapy in people with type 2 diabetes leads to a rate of hypoglycemia comparable to that observed in people with type 1 diabetes. Dr. Abrahamson then reviewed the morbidities of hypoglycemia, focusing on its effects on neural, muscular, and cardiovascular systems. Dr. Mathieu followed by reviewing studies that showed that hypoglycemia impacts patient quality of life and adherence to diabetes medications. Hypoglycemia also has significant healthcare implications – insulin and oral anti-diabetic medications are the second and fourth drugs, respectively, most commonly associated with emergency room visits, and ~95% of all endocrine emergency hospitalizations are due to hypoglycemia (Budnitz et al., N Engl J Med 2011). Dr. Mathieu concluded her talk by presenting data showing that hypoglycemia leads to increased treatment costs as well as reduced productivity (Brod et al., Value Health 2011).

  • After reviewing the clinical and biochemical definitions of hypoglycemia, Dr. Abrahamson highlighted that hypoglycemia remains a problem for both type 1 and type 2 diabetes patients. Data show that the rate of hypoglycemia (events per patient per year) reaches 43 events in patients with type 1 diabetes and roughly 16 events in those with type 2 diabetes. The same study found that the rate of severe hypoglycemia totals 1.2 events and 0.4 events among those with type 1 and type 2 diabetes, respectively (Donnelly et al., Diabetic Med 2005).
  • Dr. Abrahamson then reviewed the morbidities of hypoglycemia in diabetes. He explained that hypoglycemia can negatively impact the brain by causing blackouts, seizures,   coma, cognitive dysfunction, and psychological effects. Hypoglycemia is also associated with several cardiovascular events such as myocardial ischemia and cardiac arrhythmias. Furthermore, hypoglycemia can result in damage to the musculoskeletal system by increasing the risk for falls, accidents (including driving accidents), fractures, and dislocations. Dr. Abrahamson reminded   the audience that in ADVANCE, severe hypoglycemia was associated with an increased risk of several adverse outcomes. Severe hypoglycemia was also a major predictor of cardiovascular   death in VADT and, notably, provided a hazard ratio greater than that associated with  experiencing a previous cardiovascular event.
  • Dr. Abrahamson ended his presentation by explaining that hypoglycemia can lead  to cardiovascular consequences through several pathways. It promotes a sympathoadrenal response and an increase in adrenaline, which leads to heart rate variability, as well as hemodynamic changes (increased heart workload, contractility, and output). Hypoglycemia also increases inflammation and promotes endothelial dysfunction. Furthermore, low glucose levels can influence blood coagulation and result in abnormalities such as an increase in platelet and neutrophil activation, and an increase in factor VIII (a blood clotting protein).
  • Dr. Mathieu presented data from France, Germany, the UK, and the US, showing that hypoglycemia has significant socioeconomic consequences in type 2 diabetes by increasing treatment costs and reducing productivity (Brod et al., Value Health 2011). The study found that patients take 5.6 extra blood glucose tests within seven days after a hypoglycemic event. Furthermore, 25% of patients reduce their insulin dose and 25% contact a healthcare provider following a hypoglycemia episode. Notably, out-of-pocket costs associated  with hypoglycemia in people with type 2 diabetes amount to roughly £16.42/month  ($26.5/month; we’d be interested in knowing whether the per-episode health costs are higher for those with type 1 diabetes). Dr. Mathieu then reviewed data on the indirect social impacts of hypoglycemia which showed that the loss in productivity following a hypoglycemia episode results in a cost of £10-£60 ($16-$97) per episode. Following a daytime hypoglycemic event, 18% of patients lose an average of 10 hours of work time and 24% report missing a meeting or a deadline. Following an episode of nocturnal hypoglycemia, 23% of patients report arriving late to work or missing work altogether, 32% report missing a meeting or a deadline, and in total, roughly 15 hours of work are lost.



Chantal Mathieu, MD, PhD (University of Leuven, Leuven, Belgium)

In her presentation, session co-chair Dr. Chantal Mathieu walked the audience through insulin  degludec’s phase 3a program, which included over 10,000 patients in 40 countries and contributed to  the largest regulatory filing ever for a diabetes therapy. She pointed out four studies on the use of   insulin degludec as the first insulin therapy following oral anti-diabetic agents (LOW VOLUME, EARLY, ONCE Asia, and ONCE LONG). A fifth study (FLEX) investigated the flexible dosing of insulin degludec while a sixth study (BB T2) examined degludec as part of basal-bolus therapy. The phase 3a program also included three studies in people with type 1 diabetes – two investigating degludec alongside a rapid-acting insulin (BB T1 LONG and BB T1) and one examining degludec’s flexible dosing (FLEX T1). Dr. Mathieu ended by emphasizing two key aspects of degludec’s phase 3a program: 1) every study followed a “treat-to-target” principle; and 2) the same definition of hypoglycemia was used throughout the program: severe hypoglycemia was defined as an event in which a patient was not able to treat him or herself, while confirmed minor hypoglycemia was classified as an event in which a patient was able to treat him or herself and had a plasma glucose level below 56 mg/dl. Nocturnal hypoglycemia was defined as occurring between midnight and 6 am.



Bernard Zinman, MD (University of Toronto, Toronto, Canada)

Dr. Bernard Zinman reviewed three studies from degludec’s phase 3a program which compared degludec with insulin glargine: the ONCE LONG, ONCE Asia, and LOW VOLUME studies. In his compelling opening, Dr. Zinman described why, despite evidence from UKPDS and DCCT showing the remarkable benefit of early glycemic control, patients may take up to five years to initiate insulin after dual oral anti-diabetic (OAD) therapy, even with an elevated A1c levels. Dr. Zinman then reviewed the study designs for the ONCE LONG (1,030 T2DM patients; compared degludec vs. glargine plus metformin ± DPP-4 for 52 weeks) and the ONCE Asia study (435 T2DM patients; compared degludec vs. glargine ± OADs for 26 weeks). Degludec therapy was associated with a numerically lower rate of confirmed hypoglycemia (18% lower rate in both studies; not statistically significant) as well as a statistically significantly lower rate of nocturnal hypoglycemia (36% lower rate in ONCE LONG). Dr. Zinman then turned to the LOW VOLUME study (460 T2DM patients; compared degludec U200 vs. insulin glargine for 26 weeks). Those on insulin degludec exhibited a similar A1c reduction and daily insulin dose compared to their counterparts on glargine (0.62 U/kg for degludec vs. 0.66 U/kg for glargine). Furthermore, degludec provided a statistically significantly greater reduction in fasting plasma glucose levels, as well as lower, though not statistically significant, rates of both confirmed and nocturnal  hypoglycemia.

  • Dr. Zinman listed four main reasons why physicians hesitate to initiate insulin therapy. In a survey of physicians, 29% cited the fact that insulin causes weight gain, 55% cited pain from injections, 60% cited pain from blood tests (which increase in frequency with insulin use), and 85% cited a fear of hypoglycemia (Nakar et al., J Diabetes Complications 2007).
  • Dr. Zinman reviewed the trial designs for the ONCE LONG and ONCE Asia studies, both in people with type 2 diabetes. The ONCE LONG open label study enrolled 1,030 insulin naïve patients with an A1c between 7% and 10%, BMI ≤40 kg/m2, and aged ≥18 years. Participants were randomized 3:1 to insulin degludec or insulin glargine (both with metformin ± DPP-4 inhibitor for 52 weeks). The ONCE Asia open label study enrolled 435 Asian participants previously treated with OAD for 3 month with an A1c between 7% and 10%, BMI ≤35 kg/m2, and age ≥18 years. Study participants were randomized 2:1 to insulin degludec or insulin glargine (both ± OADs) for 26 weeks.
  • Insulin degludec and insulin glargine provided comparable reductions in A1c in  both ONCE LONG and ONCE Asia. In ONCE LONG only, degludec therapy provided a statistically significantly greater reduction in fasting plasma glucose, though Dr. Zinman noted that the difference was still modest. In both studies, degludec treatment was associated with a   18% lower risk for confirmed hypoglycemia compared to insulin glargine, though this different  was not statistically significant. However, degludec provided a statistically significantly lower rate (-36%) of nocturnal hypoglycemia compared to glargine in ONCE LONG, as well as a lower rate in ONCE Asia (38% lower rate, but not statistically significant due to smaller study size, shorter  study duration, and thus smaller event rate).



Melanie Davies, MD (University of Leicester, Leicester, United Kingdom)

In this presentation, Dr. Melanie Davies sought to answer the question: What can insulin degludec add  to basal-bolus therapy? To answer this, she reviewed the results of three 52-week basal-bolus insulin studies comparing insulin glargine to degludec: 1) the BB T1DM study; 2) the BB T1 LONG study; 3) the BB T2 study. The phase 2 BB T1DM study showed that degludec once-daily was well-tolerated and provided similar glycemic control to glargine once-daily. Furthermore, a lower rate of hypoglycemia was observed with degludec compared to glargine in individuals with type 1 diabetes. The phase 3 BB T1 LONG study found a 25% risk reduction in nocturnal hypoglycemia in individuals with type 1 diabetes taking degludec compared to glargine, and the BB T2 study revealed an 18% risk reduction of overall hypoglycemia in individuals with advanced type 2 diabetes taking degludec compared to glargine.



Bruce Bode, MD (Emory University, Atlanta, GA)

Dr. Bruce Bode reviewed the FLEX T1 and FLEX T2 trials, key elements of insulin degludec’s phase 3a program, focusing on the potentially favorable implications of degludec’s flexible dosing for patients  who struggle with regular dosing of basal insulin. Twenty-six week data from FLEX T1 and FLEX T2 showed that flexible administration of degludec (forced dosing once every 8, 24, or 40 hours, as opposed to dosing at the same time every day) was as efficacious as administration of both insulin glargine and degludec at the same time daily in reducing A1c and fasting plasma glucose levels. However, participants with type 1 diabetes undergoing flexible dosing of degludec demonstrated a 40% lower rate of nocturnal confirmed hypoglycemia compared to participants taking glargine and degludec at the same time daily. In people with type 2 diabetes, insulin glargine provided a 23% greater risk of  nocturnal hypoglycemia compared to degludec dosed at the same time once-daily. Additionally,  degludec same time once-daily was associated with an 18% higher risk of nocturnal hypoglycemia compared to flexible degludec dosing.

  • Dr. Bode gave the audience a preview of the GAPP2 (Global Attitudes of Patients and Physicians) survey in his presentation, the results of which will be presented this week at EASD 2012 by Brod et al. In the survey, 48% of respondents reported missing at   least one dose at any time during insulin treatment, with 51% reporting mis-timing at least one dose, and 38% reducing the dose at least once at any time during insulin treatment. To highlight the impact of inconsistent insulin dosing on glycemic control. Dr. Bode pointed to the fact that missing two basal insulin injections per week leads to a 0.2-0.3% increase in A1c.
  • The 26-week FLEX T1 study enrolled 493 people with type 1 diabetes and found that insulin degludec “fixed,” degludec “flexible”, and glargine same time once-daily provided comparable A1c reductions; however, degludec “flexible” led to a 40% lower rate of nocturnal hypoglycemia compared to glargine, as well as a 37% lower rate compared to degludec “fixed”. Participants on degludec “fixed” took their insulin at the same time once-daily, while those on degludec “flexible” took their insulin once every 8,24, or 40 hours per a pre-set dosing schedule.. The open-label study included participants who had been diagnosed with diabetes least 12 months prior and were over the age of 18. Average baseline A1c was 7.7% across all treatment arms. All participants had been previously treated with a basal- bolus insulin regimen. Nocturnal hypoglycemia was classified as confirmed if the patient was either unable to treat him or herself, or was able to treat him or herself and had a plasma glucose level below 56 mg/dl).
  • The 26-week FLEX T2 study in people with type 2 diabetes reported similar results, with degludec “flexible” providing 23% lower risk of nocturnal hypoglycemia compared to glargine same time once-daily. However, degludec “flexible” was associated with an 18% higher risk of nocturnal hypoglycemia compared to degludec fixed. The FLEX T2 study included 687 participants with type 2 diabetes with an average baseline A1c between 8.4 and 8.5%. Participants were over the age of 18 and had BMI <18 kg/m2. All participants had been previously treated with oral anti-diabetic drugs and/or basal insulin.



Stephen Gough, MD (University of Oxford, Oxford, United Kingdom)

Dr. Stephen Gough presented a prospectively planned meta-analysis comparing hypoglycemia rates of insulin degludec vs. insulin glargine. The analysis included two studies of degludec in type 1 diabetes patients (BB T1 LONG and FLEX T1), as well as five studies in people with type 2 diabetes (BB T2, FLEX T2, ONCE LONG, ONCE Asia, and LOW VOLUME). Pooled data from the five studies in T2DM demonstrated statistically significant reductions of 17% and 32% in the rates of confirmed hypoglycemia and nocturnal hypoglycemia, respectively, in individuals receiving insulin degludec compared to insulin glargine. Pooled data from the three studies in T1DM trended in favor of insulin glargine for confirmed hypoglycemia (10% greater rate with degludec) but trended in favor of degludec for nocturnal hypoglycemia (17% reduction), though neither finding was statistically significant. Combining the data from all seven studies revealed significant reductions of 9% and 26% for confirmed and nocturnal hypoglycemia, respectively, compared to insulin glargine. Dr. Gough then presented data obtained only during each study’s “maintenance period”, defined as the period after week 16. Compared to data from the full treatment period, data from only the maintenance period revealed a greater reduction in confirmed hypoglycemia with insulin degludec in both the pooled analysis of T2DM studies (25% reduction) and the pooled analysis of T1DM and T2DM studies (16% reduction). A similar result was observed for nocturnal hypoglycemia (38% reduction in the pooled analysis T2DM, 25% reduction in   the T1DM studies, and 32% reduction in the pooled analysis of T1DM and T2DM studies).



Peter Kurtzhals, MD (Novo Nordisk, Copenhagen, Denmark), Tim Heise, MD (Profil  Institut for Metabolic Research, Neuss, Germany), Thomas Pieber, MD (Medical University of Graz, Graz, Austria) Bernard Zinman, MD (University of Toronto, Toronto, Canada), Melanie Davies, MD (University of Leicester, Leicester, United Kingdom), Bruce Bode, MD (Emory University, Atlanta, GA), Stephen Gough, MD (University of Oxford, Oxford, United Kingdom)

Q: There are several audience members who worry about the concentration of phenol. Is that toxic in the body? What happens to the phenol?

Dr. Kurtzhals: The first thing is that phenol is not specific to the degludec preparation. It has been in any insulin preparation since 1946 when it was introduced in NPH insulin, and it is present in all human insulin preparations today. It’s not new and not specific for degludec. It’s used in low concentration, it’s a well-known and well-characterized substance, and it is eliminated within a few hours of being introduced into the circulation.

Q: So the phenol is not different from other preparations?

Dr. Kurtzhals: It’s the same as in all other insulin preparations.

Q: Are there side reactions?

Dr. Kurtzhals: The most important thing to remember is that in all the preparations that have been used, phenol has been there as well.

Q: Why did you choose to compare insulin degludec to insulin glargine and not insulin detemir?

Dr. Bode: It has been studied against detemir in type 1 and type 2 trials, so it has been studied against detemir.

Dr. Mathieu: Those results will be presented at later meetings.

Q: Do you have any data on type 2 diabetes during clamp studies with degludec?

Dr. Heise: I think I chose some data in type 2 patients. Basically, you see that, as you might expect, you have a very nice pharmacologic profile in both type 1 and type 2 patients I think I showed the data on a very nice dose response relationship. What we don’t have is a 42-hour clamp study in type 2 patients, but you would expect an even longer duration of action in type 2 patients because they can support basal insulin action with their endogenous insulin secretion.

Q: Regarding the adrenergic response to hypoglycemia with degludec compared to glargine, would this have any adverse consequences from the point of view of cardiovascular morbidity and mortality?

Dr. Pieber: Well, the most important issue around that is that in healthy subjects, who have a much stronger adrenergic response than people with diabetes, we have to look into cardiovascular side effects. But looking into the clinical program, there is no evidence of increased cardiovascular risk. I think the advantage of having a more pronounced hormonal counter-regulatory effect is that it’s easier for you to detect hypoglycemia, and we know that adrenaline, growth hormone, and cortisol are associated with the rate of glucose. If at all, I would see a benefit, not a disadvantage.

Q: What about weight gain and degludec?

Dr. Zinman: I think it is expected that any insulin that improves glucose control may be associated with weight gain. It’s interesting that detemir is actually a little different with respect to weight gain. However, there’s no difference when comparing degludec to insulin glargine: weight gain is identical.

Dr. Bode: Weight has been neutral between glargine and degludec. Against detemir, in the head to head study, there was a weight difference.

Q: In regards to exercise in patients with an ultra-long acting insulin, does anybody think that using an ultra-long acting insulin may increase risk for exercise-induced hypoglycemia?

Dr. Gough: There’s been no evidence in the clinical trial program that any episodes of hypoglycemia are related to exercise. This requires greater investigation, and there are studies on this that are currently ongoing. We have no reason for concern at the present time.

Q: Are you planning to study degludec in people who do sports?

Dr. Heise: We are just started a study last week where we’ll look at the effect of exercise on hypoglycemia and compare the effects of degludec with that of glargine. Hopefully next year I can report the results.

Dr. Zinman: Exercise is a particularly valued lifestyle for people with diabetes, so it’s important to document the responses with this basal insulin. If you go back many decades, people have studied basal insulin in response to exercise. When you infuse insulin in people with type 1 and you maintain the basal rate at a constant level, your ability to increase hepatic glucose production to match muscle utilization is not impaired. People who don’t have diabetes reduce their basal insulin by half and pumpers also reduce their insulin. Its only when you have very high levels of insulin that you suppress that hepatic insulin production and get a mismatch. This has been studied for many, many years and it’ll interesting to know – if you’re on a perfect basal insulin replacement, can you mount a perfect hepatic glucose response to exercise?

Q: I was wondering about the study where there was a fasting glucose level of above 9 mmol/l and the A1c of only 7.7%. How do you reconcile those two numbers? Is it post- prandial glucose?

Dr. Davies: In the phase 2 type 1 study, the A1c was 8.4%. In the second largest study, it was 7.7%. I don’t think it’s inconsistent with a fasting glucose of nine.

Dr. Mathieu: The big reduction in fasting glycemia and the similar A1c effect.

Dr. Davies: In that study, there was a trend in reduction of fasting glucose in both arms of the study as well as a greater, though not statistically significant, reduction with degludec. In the nine-point profile, there doesn’t seem to be a different in postprandial glucose exclusion in that study. It’ll be interesting to look at the CGMS data.

Dr. Zinman: I think A1c is an integrated measure of glycemia over a very long period. The lows contribute to the lows and the highs contribute the highs. If you reduce the number of lows and you reduce the fasting glucose level, it’s one canceling the other. That can explain why you have a lower fasting glucose and higher Ac, because you have a decrease in nocturnal hypoglycemia, which would decrease the lowering and increase the A1c. It’s a matter of balance.

Q: Will the use of ultra long acting insulins obviate the need for insulin pumps?

Dr. Bode: The benefit of the pump is that you can go from four shots a day to one shot every three days. And with a push of a button, you can give insulin. I think that degludec is as predictable as regular insulin and a pump. The variability is in the 20% range on a day-to-day basis. Control overnight is the same, based on the data that Tim Heise has shown. There are people who get tired of wearing a pump that will probably opt for a predicable basal. The reason why people with type 1 diabetes have so much trouble controlling their plasma glucose is a fear of hypoglycemia at night. So they do protective eating and other measures to get decent control at night. But you say that the fasting was 9 mol with an A1c of 7.7% - that’s common in type 1 diabetes.

Q: Why is there a different between the hypoglycemia risk of degludec flex vs. degludec fixed in the studies?

Dr. Bode: In the flex arm, there was a different in nocturnal hypoglycemia with the flex being significantly less than the fixed armed of degludec. Theoretically, if you take your insulin 40 hours apart, you’re going to get less hypoglycemia compared to taking it eight hours apart. I think this 40-hour buffer probably created less hypoglycemia during that time. That’s the only think I can think of. It was also the structure  of the study – you took degludec starting in the morning. You took it in the morning on three days and at night the other days. It’s hard to predict, but taking it that far apart probably resulted in less  hypoglycemia.

Q: For the hypoglycemia rates , you showed relatively risk reduction. Does anyone know the absolute risk reduction of hypoglycemia in the studies?

Dr. Mathieu: If we don’t know that’s fine. We’ll have to get that information out.

Q: How do you start this insulin?

Dr. Davies: I think if you’re talking about a type 1 patient, it appears to be a fairly equal move-over from   an existing insulin to degludec, so I think you can move patients across similar doses. In patients with   type 2 diabetes, there are two approaches. You can keep it simple and start with a dose of 10-15, and then move up. Or you can use an algorithm where you take into account fasting glucose and BMI and start with a higher dose, and that seems to be a very safe and effective approach. Moving forward, we will get better at getting the best out of this insulin. I think that we’re still learning about the titration.

Dr. Bode: On titration, it is a different insulin than NPH and glargine. As you saw from the data, it takes three days to get to steady state, so you don’t want to adjust the dose every day. It would be at best adjusted twice a week, maybe once a week. So it’s a little different.

Dr. Gough: I think it’s the whole point in type 1, learning how to titrate the insulin. From the meta- analysis, there doesn’t seem to be a benefit in confirm hypoglycemia, but there’s a benefit in nocturnal hypoglycemia and during the maintenance period. In the overall hypoglycemia rate, we’re seeing the effect of the bolus insulin, and we’ll learn how to best titrate that.

Q: Your phase 2 study showed the possibility a the three times weekly dosing – what happened to that dosing?

Dr. Zinman: Obviously, there is the potential with an insulin that has this long of a half life to administer   it less than once daily. And so that’s an exciting concept. Clearly it would be a paradigm shift on how we treat people with insulin. We completed a phase 2 study which demonstrated that you can administer it three times a week and get reasonable glucose control – the same seen with glargine – but in that study, you lost the advantage of a reduction in hypoglycemia. You got a similar A1c reduction, but you didn’t’ have the change in hypoglycemia that is so exciting about this insulin. So it was encouraging enough to do a large phase 3 study. I’m not going to give the results because they will be presented tomorrow. It’s quite interesting. It’s different than the phase 2 study, and I encourage you to attend that session. Also, if you’re going to administer it three times a week, you have to administer very large doses. But come to that oral presentation to see those results.

Q: Can you explain the fact that the rate of hypoglycemia is higher in patients with type 2 diabetes who have had a long duration of diabetes?

Thomas: There’s a clear explanation with longstanding type 2 diabetes: patients are losing the ability to secrete insulin and losing the ability to counter-regulate their own insulin secretion. So the longer you have the disease, the more prone you are to hypoglycemia. The UK study has shown that after 15 years, you have the same risk as a person with type 1 diabetes – that’s an important message. Anything that reduces the risk of hypoglycemia is important for both type 1 and type 2.

Q: With a higher concentration of insulin (the U200), is there any likelihood that there will be more injection site reactions because of the high concentration of zinc?

Dr. Zinman: I don’t think so. Many people do require more than 80 units, and the convenience of concentrated insulin. If anything, because the volume is smaller the site reactions would be les.

Dr. Kurtzhals: In the clinical trial program, the injection site reactions were very low. It was the same as the U100 formulation.

Q: If you have a more stable insulin, would that lead to less glycemic excursions and provide a cardiovascular benefit? Are there any long-terms studies or a meta-analysis of the cardiovascular events?

Dr. Heise: You would expect less fluctuation in the fasting plasma glucose levels because you’re reducing hypoglycemia. So you can titrate it better. If you believe the epidemiological data, that should reduce cardiovascular events, but we don’t have data that there is a causal relationship, so we can’t answer this question right now.

Q: Do we see any future for inhalable insulin? Also, the hexamers of degludec are released according to the zinc concentration. Do we see the possibility that these hexamers might be released relative to the plasma glucose concentration?

Peter: That’s a great question about the future. Regarding inhalable insulin, I personally don’t think there is a future. That’s also illustrated by the Novo Nordisk pulling out of the development of inhalable insulin five years ago. There may be a future for oral insulin. That’s what we are currently looking at in our labs and early clinical studies. We have an oral insulin in phase 1 trials now [Editor’s note: for further details on Novo Nordisk’s oral insulin, please see our Novo Nordisk 2Q12 report at]. With regards to zinc being responsible for the hexamers, that is true. We could make the insulin a glucose sensor relative to zinc. We could encrypt each molecule with a chemical glucose sensor. We did try to do that a few years back and published the data. We encrypted each insulin molecule with a glucose sensor and we encrypted each molecule with a glucose mimetic so they stuck together at low glucose concentrations but were released at high concentrations of glucose where the glucose in the circulation would compete for binding to the glucose sensor. It works well in the lab with sorbitol and works with high concentration of sorbitol. We’re not there yet, but the idea is neat. If we could do it, we would.

Q: What happens in patients with renal insufficiency?

Dr. Heise: There are data at this meeting, so you can look into a study where patients were treated with degludec and they were people with diff degree of renal insufficiency. You don’t see any difference in the PK levels. It doesn’t seem that renal insufficiency makes a difference with degludec. It’s cleared mainly by the insulin receptors, so you wouldn’t really expect anything.


Corporate Symposium: Diabetes Care Today: Individualizing Treatment Options (Sponsored by Lilly Diabetes)


Rury Holman, MD (University of Oxford, Oxford, UK)

Dr. Rury Holman gave an excellent review of the evidence supporting early use of insulin, arguing that since it is usually an inevitable part of treating type 2 diabetes, there is no reason to wait. Overall, he  was very positive on using insulin early in type 2 and highlighted that it helps achieve optimal glucose control quickly, may prolong beta cell function, and may enhance the effectiveness of other oral agents. Dr. Holman addressed some of the major barriers to using insulin early (lifestyle   convenience/injections, weight gain, hypoglycemia, cardiovascular risk, and cancer), highlighting how these are not really an issue, especially at low A1cs typically seen at diagnosis. In terms of how to use insulin, he discussed the major finding from the 4-T trial that early treatment with basal insulin is the way to go. Dr. Holman concluded with a fascinating Chinese study comparing two-week use of a sulfonylurea vs. intensive insulin therapy in newly diagnosed type 2s – after one year, insulin was associated with better beta cell function relative to the oral agent group. We found his presentation quite persuasive and wonder what percentage of clinicians would consider early insulin use, especially transiently at diagnosis.

  • “If we’re going to need insulin, why wait?” Dr. Holman emphasized that since type 2 diabetes is progressive, many patients will end up failing oral therapies and turning to insulin. He noted that there is no limit to insulin’s glucose lowering abilities, it has a virtually 100%  responder rate, and large doses can overcome insulin resistance. In his view, the major patient concerns associated with insulin use – lifestyle inconvenience and injections – are not a big barrier anymore due to modern syringes and delivery devices.
  • Hypoglycemia is “not a major barrier” to using insulin early. Dr. Holman reviewed UKPDS data comparing the risk of hypoglycemia in sulfonylureas and insulin. In patients with   low A1cs (and thus good beta cell function) using an insulin secretagogue brings a risk of hypoglycemia. However, insulin-using patients with low A1c levels in UKPDS had the lowest risk  of hypoglycemia. Dr. Holman explained that it’s the people who are difficult to control and those with a longer duration of diabetes that carry the highest risk of hypoglycemia. Data from   ACCORD also supports this notion: the increased risk of hypoglycemia seen in the intensive group was at higher, not lower A1c levels.
  • Weight gain associated with early insulin use is “modest” and offset by the reduced risk of complications that accompany improved glycemic control. Dr. Holman reviewed ten-year data comparing first line insulin therapy to metformin, and sulfonylureas. While metformin and sulfonylureas were weight neutral, insulin was associated with ~2.5 kg of excess weight gain. Dr. Holman conceded that this was “unwelcome,” but the reduced risk of diabetes complications makes up for it.
  • “Insulin does not increase cardiovascular disease.” Dr. Holman again referred to UKPDS data comparing insulin to conventional therapy. Insulin was not associated with an increased risk of cardiovascular disease as measured by any endpoint. He admitted, however, that the analysis was underpowered to detect statistically significant differences. Dr. Holman also reviewed the ORIGIN results, noting that there was “not a hint of benefit or harm” to giving low dose insulin early and it was associated with “excellent glucose control.”
  • There is “no data” to suggest that use of low doses of long acting insulin increase cancer deaths or cancer risk. Dr. Holman highlighted the neutral ORIGIN results in this regard – he noted that for cancer deaths and risks of developing any cancer, the hazard ratios fell right on or near 1.0. Compared to the benefits of reducing microvascular and macrovascular complications, he said, early insulin use doesn’t unacceptably increase cancer risk.
  • Dr. Holman covered the results of the 4-T trial, emphasizing the positive data on early use of basal insulin (Holman et al., NEJM 2007). As a reminder, this trial randomized 708 patients with high A1cs (7-10%) to receive biphasic insulin aspart twice daily, prandial insulin aspart three times daily, or basal insulin detemir once daily (twice if required). Patients were on max doses of metformin and sulfonylureas. At one year, there were minimal differences in A1c (though basal was slightly worse) and least weight gain was observed in the basal-only group. Relative to basal use alone, there was a six-fold increase in hypoglycemia with prandial insulin  and a three-fold increase with biphasic insulin. Dr. Holman also showed data on the likelihood of achieving an A1c <6.5% – there was a significantly worse response to basal insulin for those with   a baseline A1c >8.5%, while those with an A1c <8.5% were not significantly different. Over three years, the A1c lowering was identical between the three arms, though there was a net benefit in terms of weight gain, hypoglycemia, and waist circumference for those on basal insulin only.
  • Early use of basal insulin may also have benefits on beta cell function. Dr. Holman concluded his presentation with a review of a nine-center randomized trial from China comparing two weeks of intensive insulin therapy (pump or MDI) to orals agents in newly diagnosed type 2s (Weng et al., Lancet 2008). At one year, insulin treated patients continued to have a better acute insulin response than those in the oral agent group. Dr. Holman noted that the study is being repeated in the US.



Rury Holman, MD (University of Oxford, Oxford, United Kingdom); Tina Vilsboll, MD (University of Copenhagen, Copenhagen, Denmark)

Q: Would you start insulin with a basal or prandial insulin?

Dr. Holman: When starting early you should use basal insulin, but if it’s a rescue for high A1c, you’ll likely need MDI (multiple daily injections) anyway.

Q: At what level would you consider starting?

Dr. Holman: We’ve seen different targets for different people. If you think A1c is about to become too high, you should start the next level of treatment rather than waiting for the current treatment to fail. I agree that we shouldn’t maximize each treatment for incremental benefit, but move on. If I’ve persuaded you that early insulin is better, you should add it at the time when the A1c is going toward the target you’ve set.

Q: How many people need insulin and when?

Dr. Holman: That’s tough to say but UKPDS said that by 15 years after diagnosis, 77% of patients were on insulin alone or in combination with other agents. By modern standards with tighter A1c targets, I suspect it would be a lot more.

Q: Is there any reason to believe that analogs will have a better long-term impact on beta cell function compared to the older insulins?

Dr. Vilsboll: We don’t really know for sure. As long as patients are treated with GLP-1, they have improved beta cell function. That doesn’t tell us anything about the future of beta cell mass. Some patients tend to deteriorate while some react really well. We just learned from ORIGIN that insulin therapy doesn’t affect CV endpoints, but we’ll have answers regarding CV outcomes for GLP-1 starting in 2014.

Q: Do large doses of insulin boost cancer risk?

Dr. Holman: ORIGIN showed that six-year low dose early insulin use doesn’t enhance risk, but we don’t have long-term randomized controlled trial data. The evidence that insulin directly produces risk of cancer is quite small.

[Note: additional talks from this corporate symposium can be found in the “Incretin-Based Therapies” section of this report]


Corporate Symposium: A Comprehensive Therapeutic Approach to Diabetes Management (Sponsored by Sanofi)


Anthony Barnett, MD (University of Birmingham and Heart of England NHS Foundation Trust, Birmingham, United Kingdom)

Dr. Barnett emphasized the importance of individualizing therapy for the treatment of diabetes, discussing some of the considerations that should be taken into account. Tailoring treatment for patients with type 1 diabetes, he said, requires a personalized, patient-centered approach that accounts for changes in the individual through different developmental stages, individualizes treatment through the child-to-adult continuum, and limits hypoglycemia. Meanwhile, for patients with type 2 diabetes,   patient engagement should be prioritized in tailoring therapy, as adherence is of utmost importance. Other factors such as patient attitudes, age, diabetes duration, and comorbidities should also be taken into account.

  • Dr. Barnett noted that tailoring treatment for patients with type 1 diabetes requires a personalized, patient-centered approach that accounts for changes in the individual through different developmental stages, individualizes treatment through the child-to-adult continuum, and limits hypoglycemia. As an example, Dr. Barnett reviewed the individualized A1c targets for children and adolescents with type 1 diabetes outlined in the recent ADA guidelines (Diabetes Care 2012). Toddlers and preschoolers (ages 0-6) should target an A1c of less than 8.5% due to their vulnerability to hypoglycemia and  unpredictable dietary intake and physical activity. School-age children (6-12) should target an A1c of less than 8%, given their vulnerability to hypoglycemia. Adolescents and young adults (13-19) should target an A1c of less than 7.5%. Lower goals of less than 8%, 7.5%, and 7.0% for toddlers/preschoolers, school-age children, and adolescents/young adults, respectively, are reasonable if they can be achieved without excessive hypoglycemia.
  • He argued that for patients with type 2 diabetes, patient engagement should be prioritized in tailoring therapy, as adherence is of utmost importance. Dr. Barnett emphasized that non-adherence is a problem of epidemic proportions. The WHO estimated non- adherence in chronic disease to be about 50% by one year after therapy initiation. In Europe,  costs attributable to non-adherence are an estimated €125 billion per year, and contribute to 200,000 deaths per year. Three of 10 people stop taking their medications before their first  supply runs out, 25% take less than the recommended dose, and 33% of individuals don’t even fill the prescriptions they are given. Dr. Barnett noted that the most common factors related to non- adherence in diabetes are side effects (58%), difficulty in remembering doses (23%), and costs (8%) (Grant et al., Diabetes Care 2003).
  • He reviewed the recommendations of the ADA/EASD position statement, which outlined factors that should be considered in individualizing glycemic targets for patients with diabetes. The position statement proposes that individualized glycemic targets should take into account patient attitude and expected treatment efforts, risks potentially associated with hypoglycemia (and other adverse events), disease duration, life expectancy, comorbidities, established vascular complications, as well as resources and support systems; importantly, the achievement of glycemic control requires active participation and commitment from the patient (Inzucchi et al., Diabetologia 2012). Dr. Barnett noted that the ADA/EASD position statement allows virtually any medication to be used as second-line or third-line therapy, as long as it is appropriate for the patient.



Vivian Fonseca, MD (Tulane University Medical Center, New Orleans, LA)

After Dr. Fonseca highlighted early diabetes diagnosis as an unmet need, he provided rationale for timely insulin initiation, reviewing the clinical evidence in support of early initiation of basal insulin analogs in particular. He stated that: 1) early intensive control with insulin can maintain and improve beta cell function and reduce insulin resistance; 2) the early combination of basal insulin with metformin is associated with good glycemic control, as well as low rates of hypoglycemia and weight gain; and 3) resting beta cell function with early insulin therapy could reduce the need for other drugs.

  • Dr. Fonseca pointed out that the time to insulin initiation is increasing, and that a higher percentage of patients are experiencing complications prior to starting insulin therapy. From 2005 to 2010, the median duration from diagnosis until insulin   initiation increased from 1.4 years to 3.2 years in France, from 2.6 years to 4.2 years in Germany, and from 4.7 to 5.6 years in the UK. Disconcertingly, the percentage of patients who experienced  at least one macrovascular event prior to insulin initiation increased from 18.6% to 26.3% in France, from 5.3% to 50.5% in Germany, and from 27.4% to 31.7% in the UK over the five-year period (Kostev et al., Diabetologia 2011). Dr. Fonseca commented that it is unrealistic for people  to reverse their pre-existing complications when starting insulin so late. In an audience poll, 7% of respondents stated that they initiate insulin therapy for their patients less than one year after diagnosis, 15% said 1-3 years, 23% said 3-5 years, and 55% said 5+ years.
  • He emphasized that early insulinization can improve diabetes remission. In a study  in 382 patients with newly diagnosed type 2 diabetes, participants were randomized to receive continuous subcutaneous insulin infusion (CSII), multiple daily injections (MDI), or oral antidiabetic agents (OADs) until normoglycemia was achieved (therapy was given for two weeks, then stopped). At the one-year mark, 51.1%, 4.49%, and 26.7% of those receiving CSII, MDI, and OADs were in remission; beta cell function following two weeks of intensive insulin therapy was shown to increase the acute insulin response significantly after two weeks of therapy (Weng et al., Lancet 2008).
  • Dr. Fonseca emphasized that insulin glargine improves glycemic control rapidly  and that the effects are sustained over time. He noted that across various trials, insulin glargine treatment reduced A1c from an average baseline of 8.5-8.7% down to approximately 7% over 12 weeks. Citing 32-month extension data from a nine-month observational study (Schreiber et al., Diabetes Tech Ther 2008), Dr. Fonseca stated that the A1c-lowering effects of insulin glargine are sustained over time. Meanwhile, the 4-T study showed that basal insulin provides effective glycemic control with fewer hypoglycemic events than with biphasic or prandial insulin (Holman et al., NEJM 2009). Encouragingly, the AT.LANTUS study showed that patient-driven titration of insulin glargine is more effective than physician-driven titration (Davies et al., Diabetes Care 2005).
  • He stated that the early use of basal insulin in combination with metformin is associated with good glycemic control and low rates of hypoglycemia and weight gain. In a meta-analysis of pooled hypoglycemia event rates from trials adding basal insulin to metformin, severe hypoglycemia events were rare (DeVries et al., Diabetes 2012).
  • Dr. Fonseca noted that insulin glargine provided a significantly greater reduction in A1c versus sitagliptin in a recent head-to-head trial. In the EASIE study, insulin-naïve patients failing on metformin with baseline A1c between 7% and 11% were randomized to receive insulin glargine or sitagliptin treatment. Over the course of the study, insulin glargine and sitagliptin conferred respective reductions in A1c of 1.72% and 1.13% from a baseline of 8.5% (p<0.0001); a higher percentage of patients on insulin glargine achieved A1c less than 6.5% and less than 7.0% (Aschner et al., Lancet 2012). Dr. Fonseca noted that although the occurrence of symptomatic hypoglycemia was higher with insulin glargine treatment than with sitagliptin treatment, the absolute number of events per patient-year were low, and the occurrence of severe nocturnal hypoglycemic events was the same in the two groups (0.01 events per patient-year). Body weight increased slightly (0.44 kg [0.95 lbs]) with insulin glargine treatment, whereas  weight decreased slightly with sitagliptin treatment (average weight loss of 1.08 kg [2.4 lbs]).



Hertzel Gerstein, MD (McMaster University, Hamilton, Canada)

Dr. Gerstein reviewed the study design, baseline characteristics, and results of the ORIGIN trial, concluding that the early addition of insulin glargine for six years: 1) is possible and feasible; 2) has a neutral effect on cardiovascular events; 3) has a neutral effect on cancers; and 4) reduces progression to type 2 diabetes. He stated that with the ORIGIN results now out, insulin glargine has become the best studied of all glucose-lowering drugs. For more on ORIGIN, please see our full commentary at



Peter Boyle, PhD, DSc (International Prevention Research Institute, Lyon, France)

Dr. Boyle presented the results of a meta-analysis of 18 studies (including ORIGIN) that examined insulin glargine and its associated cancer risk, which showed no evidence to support the notion that the risk of cancer is increased with insulin glargine use versus other treatments (RR=0.93; 95% CI: 0.87- 1.00). There was no increased risk of colorectal cancer (RR=0.85; 95% CI: 0.76-0.94), prostate cancer (RR=1.08; 95% CI: 0.92-1.28), lung cancer (RR=1.04 (95% CI: 0.91-1.19), pancreatic cancer (RR=0.98; 95% CI: 0.84-1.14), or breast cancer (RR=1.10; 95% CI: 0.99-1.21). Moving on to other glucose-lowering therapies, Dr. Boyle noted that a meta-analysis of pioglitazone and its associated bladder cancer risk is  a little more worrying (RR=1.59; 95% CI: 0.97-2.59), especially when looking at patients treated for  over 24 months (RR=1.39; 95% CI: 1.08-1.80). Subsequently, Dr. Boyle commented that the concern  over pancreatic cancer with incretin therapies was the “biggest biological nonsense seen in recent years,” given that it takes more time for symptoms of pancreatic cancer to present than some of the incretin therapies have been available on the market.



Richard Bergenstal, MD (International Diabetes Center, Minneapolis, MN), Peter Boyle, PhD, DSc (International Prevention Research Institute, Lyon, France), Vivian Fonseca, MD (Tulane University Medical Center, New Orleans, LA), Hertzel Gerstein, MD (McMaster University, Hamilton, Canada)

Dr. Bergenstal: If you decided you’d like to start insulin therapy, does the possible risk of cancer that you’ve heard about influence your decision to start insulin or not?

Audience: 15% yes; 85% no.

Dr. Bergenstal: Can you hypothesize what that 15% is about? Is there still research that needs to be done?

Dr. Boyle: I think there are several issues. One of the issues that has been buried is of Lantus and long- acting insulins; I think the notion that they increase the risk of cancer compared to other insulin therapies has been buried. In total, we’ve got a million people that we’ve examined. Three million patient-years, which is an enormous epidemiological study, and yet we’re not able to find any [risk] that is elevated or significant. I really think it’s time to put that baby to bed, and think of other important issues. There are still some to be resolved with pioglitazone and Victoza.

Dr. Bergenstal: We talked about insulin and you [the audience] seem pretty convinced that [the possible risk of cancer] does not factor into your decision. Does the risk of cancer influence your decision at all about other agents?

Audience: 41% yes; 59% no.

Dr. Bergenstal: So it’s a little more split. Any comments clinically about this issue of deciding between different agents and the risk of cancer?

Dr. Gerstein: My only point is one we made several times. Diabetes is a risk factor for cancer, not just breast cancer, but for liver cancer, pancreatic cancer, and a number of other cancers. When people come to medical attention for a number of reasons and are identified as having diabetes, and are put on medications for diabetes, someone is going conduct an epidemiological study that says this drug is associated with cancer, when really the relationship is a spurious one. Unless there is a compelling reason to think the therapy is associated with cancer I don’t think we should make it. Guilt by association was a problem in the past. When you didn’t see it in [Dr. Boyle’s analysis], I think you need to be reassured.

Q: I’m really impressive how low the incidence of hypoglycemia was in the glargine group in ORIGIN. How can that be? They got up to 60 units of Lantus at night. How was the incidence so low with those doses, and decreased fasting glucose levels?

Dr. Gerstein: We haven’t published it yet, but people with IFG or IGT on Lantus had lower rates of hypoglycemia than people with diabetes. It sounds counterintuitive, but it’s not. Glucose levels in people without dysglycemia control glucose levels by both insulin and counterregulatory hormones that prevent glucose levels from going low. When the pancreas is fairly healthy, it can still produce insulin. So, for  those individuals, they are not just using exogenous insulin; rather, it is a partnership between exogenous insulin and the pancreas to control glucose.

Dr. Fonseca: It’s important to recognize that counterregulatory hormones and the responses of the body, including the autonomic nervous system response, decline as the duration of diabetes increases. So, you cannot just say the rates of hypoglycemia are the same in everybody with insulin – it’s different for those who’ve had diabetes for a long time versus those who’ve had it for a short time.

Dr. Gerstein: A classic example is for patients with type 1 diabetes in the DCCT. There were orders of magnitude more hypoglycemia for patients without any beta cell function.

Dr. Fonseca: Jay Skyler may talk about it in the afternoon. After islet cell transplantation, patients are not always insulin independent, but their rates of hypoglycemia fall, presumably because they have alpha cells that now function better.

Dr. Bergenstal: Here is another question to the audience. Once you’ve started a person on a long-acting basal insulin, do your patients self titrate?

Audience: 66% yes; 34% no.

Dr. Bergenstal: Vivian, in your talk we saw the data saying that self-titration was equally effective. Has it caught on? Do we have good tools, or good ways for patients to self titrate? What’s your experience?

Dr. Fonseca: I think this is a very important issue. We cannot see people as frequently as we would like to make the changes that are needed. I really liked seeing the results of the AT.LANTUS study – nothing is better than the empowered patient.

Dr. Bergenstal: Do you have rules on when your patients should call you, for example, when they get to 100 units of insulin per day?

Dr. Fonseca: Again, it depends on the individual patient. Some patients are comfortable making dose adjustments based on carb counting, whereas others are not. Self-titration should be a part of the discussion of individualizing therapy.

Dr. Bergenstal: How do you decide when it’s time to move on to add mealtime or rapid- acting insulin?

Dr. Fonseca: Since basal insulin targets fasting glucose, you know that if the A1c is not at target, then they have elevated postprandial glycemia. Also, if your patient is getting hypoglycemia at night, then you can’t titrate the dose up, so you could add a mealtime insulin to address daytime hyperglycemia. There are also physical dose limitations – each pen only has 80 units, and each syringe only has 100 units. If you have to go above that dose, what do you use? Basal-plus has done well in practice. There are a number of things you consider there, but if patients need high doses, you may need to use U-500 or other strong strengths being used in clinical trials right now.

Dr. Gerstein: I would totally endorse what Vivian said. When patients self titrate, they are really partnering with the healthcare professional team to do it. I completely agree that patients need to titrate their own insulin. In a recent trial in Canada, it was shown that self-titration of rapid-acting insulin actually yielded better results than physicians titrating insulin. If patients are taught to self-titrate using simple algorithms, it works.

Dr. Bergenstal: One more question on the cancer issue. What if a woman has history of breast cancer and needs to go on insulin?

Dr. Boyle: I don’t believe we have data to answer that question. It depends on the patient, and it is a clinical decision based on patient’s disease. Does the patient have stage zero breast cancer, or recurrent metastatic breast cancer? You have to look at everything and make a clinical decision; there isn’t an algorithm based on data you can use at present. One criticism I have of the clinical community is that when a woman comes in with breast cancer and has chemotherapy, I think endocrinologists are terrified of the chemotherapy and reduce the dose of insulin just in case. Conversely, when an oncologist sees a patient with diabetes getting a lot of insulin, they’re frightened to give too much chemotherapy. One in 10 women around the world are going to develop breast cancer. In ten years, one in 10 women will have diabetes. So, even if the two aren’t linked, one in 100 women are going to have both diabetes and breast cancer. I think there is a need to come up with some guidelines for both endocrinologists and oncologists on how to treat these women.

Dr. Bergenstal: One more yes/no question for the audience. Do you think that metformin reduces the risk of cancer?

Audience: 71% yes; 29% no.

Dr. Bergenstal: What’s the data?

Dr. Boyle: We have a paper in press looking at all observational studies of patients with diabetes using metformin, and the risk of developing cancer. It’s difficult to see anything concrete. Overall, the relative risk is not significantly reduced. The risk of taking metformin does not significantly reduce the risk of breast cancer. What is very interesting is the effect of metformin as an adjuvant therapy for breast cancer in current trials. Currently, 50 trials are investigating metformin as an adjuvant. One type of breast cancer, accounting for 20% of breast cancer, is nonresponsive to chemotherapy, but when you add metformin to the mix, you get significant responses. Certainly, there is a hope that metformin as an adjuvant to breast cancer treatment could have clinical benefit. We’re still in the early days, but there is hope.

Dr. Bergenstal: Back to insulin – one more question for the audience. If you have a patient on metformin and one other oral agent, and you’re ready to start them on Lantus, would you stop the other oral agent when initiating Lantus?

Audience: 31% yes; 69% no.

Dr. Bergenstal: What’s your take on this, Vivian?

Dr. Fonseca: I think that at least where I practice in the US, adding basal insulin is usual practice. It lowers fasting glucose, and postprandial excursions may remain but be attenuated by the fact that preprandial glucose is much better. Oral agents therefore seem to work better during the day; I continue with the agent no matter what it is. GLP-1 and insulin may be a very effective way of addressing both preprandial and postprandial hyperglycemia.

Q: For the ORIGIN trial, are we going to get more data going forward on this prevention issue? What are you predicting or hypothesizing, and how long do you think an effect will last?

Dr. Gerstein: Patients in the ORIGIN trial are now being recruited for a passive follow-up – ORIGINALE (ORIGIN And Legacy Effects). We’ll look at the durability of the effect on glycemic control and diabetes developments, and will provide more information related to that.

Q: Going back to the barriers of starting insulin, do you have any other advice on how to break down those barriers when talking to our primary care colleagues?

Dr. Fonseca: Again, it comes down to awareness, education, and having the right evidence base. I think ORIGIN helps to point out the advantages of early insulin use, and that it’s safe.

Q: Do the ORIGIN results apply only to Lantus, or can we assume it’s going to be similar for all other insulin analogs?

Dr. Gerstein: We get asked that question often. The short answer is that the study was done with glargine, so the results are most applicable and relevant to glargine. Everything else becomes an extrapolation – it just depends how much you want to infer on the biology of other drugs. I have tremendous confidence in applying the ORIGIN results to glargine. For other analogs, I think the results would probably apply, but I don’t have as much confidence. The reason that glargine was chosen was its time action profile, and its ability to achieve predictable glucose levels. We obviously would never have started ORIGIN using NPH – the day-to-day swings would have been an issue.

[Note: additional talks from this corporate symposium can be found in the “Incretin-Based Therapies” and “Type 1 Diabetes Therapies” sections of this report]


Corporate Symposium: The Challenge to Optimize Insulin Therapy: How New Diagnostic Concepts and Technology Can Support People with Diabetes and Their Healthcare Professionals (Sponsored by Roche Diagnostics)


David O’Neal, MD (University of Melbourne, Melbourne, Australia)

Dr. David O’Neal gave the Australian Perspective on insulin therapy, and argued that insulin therapy needs to be initiated sooner. Through the INITIATION pilot study, Dr. O’Neal explored whether general practitioner (GP) engagement can help address the imbalance between demand (i.e., the number of patients with diabetes) and resources (i.e., the number of endocrinologists and CDEs). Based on preliminary analysis he suggested that a GP and nurse practitioner (NP) team, with the support of a diabetes nurse educator (DNE), can safely and effectively initiate people with uncomplicated type 2 diabetes on basal insulin therapy. Recently, Dr. O’Neal initiated the STEPPING UP study, which is cluster randomized controlled trial that assesses whether the Stepping Up Program (a GP/PN training program) can enhance GP-based initiation of insulin therapy in people with poorly-controlled type 2 diabetes and lead to improved glycemic control. Additionally, Dr. O’Neal gave his perspective on methods for obtaining the necessary glucose information to optimize insulin therapy, concluding that structured seven-point SMBG offered the best compromise between glucose detail and cost.

  • In Australia, there are over 1.2 million people with diabetes, but only 490 endocrinologistsand 1,578 credentialed diabetes educators. While there are 10,759 primary care nurses and 24,720 general practitioners (GP), about 80% of insulin initiation is done by a diabetes specialist. Mean A1c prior to insulin initiation in patients with type 2 diabetes is 9.4%, which Dr. O’Neal said is suggestive of suboptimal insulin initiation. He reasoned that having more GPs effectively initiate and optimize insulin therapy could greatly improve diabetes care in Australia.
  • The INITIATION pilot study explored the feasibility of recruiting and training PCP/NP teams to initiate and optimize insulin therapy in patients with type 2 diabetes. Patients with uncomplicated type 2 diabetes and A1c >7.5% were initiated on basal insulin therapy by PCP/NP teams who had access to diabetes nurse educators (DNEs) as needed during the study Patients cared for by the PCP/NP teams were compared to a benchmark group comprised of patients with type 2 diabetes whose therapy was initiated by a diabetes specialist. An interim analysis found a non-significant difference in change in A1c from baseline between  patients initiated on therapy by PCPs (n=39) vs. patients initiated on therapy by specialists  (n=46). Dr. O’Neal concluded that basal regiment could be implemented in a safe, effective,   timely manner by a PCP/PN team with DNE support (pending study completion and full   analysis).
  • Qualitatively, INITIATION suggests that PCP/NP teams treating patients with basal insulin require less support from DNEs when they: 1) are treating patients well known by the staff; 2) have a strong interest in diabetes; 3) have good communication (between the PCP and PN); and 4) are teams in which the PN takes ownership of the cases. However, teams require   more support when: 1) the practice is a large corporate practice; 2) there are rapid staff changeovers; 3) the PN is overworked; 4) the team does not proactively manage the patient; and 5) there are barriers to communication between the PN and PCP.
  • STEPPING UP will randomize general practices to either the STEPPING UP training program or the control group to assess whether the HCP training program can lead to improved patient outcomes, as measured by A1c change from baseline. The STEPPING UP intervention includes: 1) 60-90 minutes of training; 2) ongoing support from endocrinologists and DNEs; and 3) one-on-one training for the NP with the DNE. Dr. O’Neal hopes that through this training program, GPs will initiate more patients in a timelier manner on insulin therapy, leading to improved glycemic outcomes.
  • Dr. O’Neal suggested that structured seven-point SMBG profiles were a good compromise between quality of glucose information and cost. Dr. O’Neal emphasized that the quality of glucose information was important for optimizing insulin therapy, however, he explained that CGM in Australia is not reimbursed and is unaffordable. He proposed three  glucose monitoring methods: 1) pre-breakfast fasting plasma glucose, which is simple and inexpensive for patients, but gives limited glucose detail; 2) structured seven-point profile, which gives detailed glucose data and is low cost, but is demanding on patients; and 3) CGM, which  gives highly detailed glucose information and is simple on patients, but is expensive.  Undoubtedly, seven-point profiles provide a more complete picture than a single fasting plasma measure, we believe that seven-point SMBG still falls short of the trending data provided by CGM and misses some of the glycemic variability CGM captures. Therefore, we hope reimbursement climates will change such that CGM can be viewed as a more realistic possibility.

Questions & Answers

Q: Do you have preliminary experience with use of the Accu-Chek 360° testing tool?

A: We have piloted the three-day seven-point profile and it has utility. The only thing one patient complained about was that the writing was on the small side, but really we’ve had no problem to date. The basal titration tool is very straightforward and there have been no issues with inappropriate dosing. The basal and prandial titration tools we have do allow HCPs some leeway as long as they can justify the deviation.

Dr. Richard Bergenstal (International Diabetes Center, Minneapolis, MN): I was struck by your data on A1c level at the time of insulin initiation; I have never seen it used as marker of success, but I like it. Are you going to track that?

A: We should be able to track it with time. There are a number of effectors for the A1c value. As far as healthcare resources are concerned, if you have an HCP give a patient a referral, it could take six months for the patient to be able to see a specialist during which time the patient could change their mind or come to think it must not be that important if it is taking six months for them to get to see the specialist.


3. Diabetes Technology

Oral Presentations: Devices, Algorithms and Their Application


Barry Keenan, PhD (Medtronic Diabetes, Northridge, CA)

Dr. Barry Keenan presented three computer simulations of Medtronic’s predictive low glucose management (PLGM) algorithm. The algorithm, which we believe will be incorporated into the  MiniMed 640G insulin pump, suspends insulin delivery based on whether hypoglycemia is predicted (30-minute time horizon) – Dr. Keenan emphasized that the advantage of PLGM is that it can reduce both the time spent in hypoglycemia (i.e., like the Veo, but more effectively) and the incidence of hypoglycemia (i.e., something the Veo cannot do since it only suspends once hypoglycemia is reached). Computer simulations compared measures of hypoglycemia in a control condition (no pump suspension), using the MiniMed 530G (suspending at 69 mg/dl), and using the PLGM algorithm. In one simulation using the FDA-approved UVA/Padova simulator (n=300), the PLGM reduced the number of hypoglycemia events by 18% and reduced the average duration of hypoglycemia by 50% (compared to reductions of 1% and 28% for the Veo algorithm). This is a significant difference and should really be appealing to patients and providers. The PLGM suspended insulin delivery at an average blood glucose of 83 mg/dl (vs. 69 mg/dl for the Veo) and the average max glucose after insulin resumption was 150 mg/dl (vs. 160 mg/dl for the Veo). As measured by YSI, the algorithm falsely suspended about 20% of  the time, but Dr. Keenan explained that the impact of these false positives “is minimal” due to the system’s logic and automatic pump resumption. Medtronic has also simulated poor sensor performance and the algorithm performed similarly. Overall, the results look encouraging and we look forward to seeing the clinical data currently being collected in Germany by Dr. Thomas Danne (we expect to see  this at ATTD 2013 in Paris) and by Dr. Tim Jones (Australia).

  • Medtronic’s predictive low glucose management (PLGM) algorithm suspends insulin delivery based on a 30-minute prediction of blood glucose. The algorithm will suspend insulin delivery if glucose is predicted and will automatically resume insulin infusion based on unspecified heuristics and logic. As a result of this predictive suspend, the PLGM can reduce both the incidence of hypoglycemia as well as the time spent in hypoglycemia. This builds upon the Veo, which only suspends insulin delivery once the hypoglycemic threshold is crossed (e.g., 69 mg/dl).
  • Medtronic tested the PLGM algorithm using the FDA-approved UVA/Padova computer simulator in 300 virtual patients. Hypoglycemia (targeting a glucose level of 60 mg/dl) was induced using a manual bolus. The PLGM algorithm was compared to the MiniMed 530G (i.e., the US version of the Veo that suspends at 69 mg/dl) and a control group (no pump suspension).
  • The PLGM reduced the number of hypoglycemic events by 18% and reduced the average duration of hypoglycemia by 50%. This builds on top of reductions of 1% and 28% for the MiniMed 530G (the reason for the 1% reduction in the total number of hypoglycemia events was due to simulated sensor error).
  • The sensitivity and specificity of the PLGM algorithm was modeled using data from the six-day Enlite accuracy study. The data set included 6,404 paired CGM-YSI points. The prediction horizon was set at 30 minutes and hypoglycemia was defined as <70 mg/dl. The algorithm had a sensitivity of 99.5%, meaning it detected nearly every hypoglycemia event. The tradeoff was a high false positive rate – 11% of suspensions were false alerts as measured by CGM and 20% were false alerts as measured by YSI. Dr. Keenan emphasized that the “impact of false positives is minimal” due to the algorithm’s intelligence in resuming insulin delivery.
  • To further test the PLGM algorithm, a range of good and poor performing sensors were modeled. MARDs ranged from 4% to 20% and true sensor noise and drift were acquired from the six-day Enlite accuracy study. Ten subjects were virtually evaluated and each repeated the study three times (i.e., three different sensors). Encouragingly, the data was in line with the numbers shown in the table above, suggesting that the PLGM algorithm is robust to poor sensor performance.

Questions and Answers

Q: What is the comparator in the clinical study?

A: There is a control arm in the study with Tim Jones. They are using exercise to create hypoglycemia.

Q: So it’s experimental and not in clinical events?

A: Yes, it’s in the CRC.



Dr. Emanuele Bosi presented accuracy results on the Glycolaser, a non invasive glucose monitoring device which uses two diode LED/lasers (the exact mechanism, he said, was an “industrial secret”). In our opinion, the picture of the Glycolaser revealed a form factor closer to hospital point-of-care meters than personal blood glucose meters. The device was tested in either fasting or postprandial conditions in 171 adults (31 healthy controls, 136 patients with type 1 or type 2 diabetes, and four patients with hypoglycemia syndrome). Only 49% of measurements were within ISO limits – 7.7% when glucose < 75 mg/dl and 52.5% when glucose >75 mg/dl. And if the proposed tighter ISO standards are implemented  in early 2013 as Dr. Lutz Heinemann suggested on Day #1 (see page 7 at, Glycolaser will have to make even more substantial accuracy gains before pursuing regulatory approval. Certainly, as documented in the  history of non-invasive methods that Dr. Bosi presented (“to date no one has been able to bring non invasive glucose measurement to reality”), achieving these improvements will be very challenging. Even Dr. Bosi (who was not involved in the development of Glycolaser) seemed relatively unconvinced about the meter’s viability during the Q&A that followed his presentation. He did note that if commercialized, the price could be low due to no “consumables” required.

Questions and Answers

Q: This seems to be interesting device. Some measurements were not acceptable, was this in specific patients? And does testing on different areas of the body play a role?

A: We didn’t identify what factor was making measurements be wrong. We need to work on the core technology. This kind of technology works only on the finger because it requires a lot of blood according to the inventor, who is a single and independent inventor, so there is not a great amount research behind it. I’m trying to convince the inventor to expand and get knowledge from more advanced researchers.

Q: You were smart enough to outline the critical history of non-invasive glucose monitoring. Unfortunately, all of these developments have failed. In a given patient, have you measured multiple blood samples in a row? Who has analyzed the data measured, and were patients aware of their blood glucose at the time?

A: The analysis was done separately. We are doing repeated measurements in the same individuals, and we have a few data, but more or less, the results are always the same. It’s relatively easy to reach the level of accuracy we achieved; the last gain in accuracy is always the most difficult to achieve.

Q: Can you elaborate on calibration and on the title? The title said in men – were results only in men?

A: Results were in humans in general – men and women. There was a slightly greater proportion of males. As to calibration, the inventor gave us the prototype, and according to him it was already calibrated, so there was not a calibrating phase. Apparently it doesn’t need it, but you know the results are still unsatisfactory. The matter of calibration has to be better investigated.

Q: What would the cost be, approximately?

A: That is impossible to answer. The cost is almost nothing, because once you have the instrument working, you have very little consumables. If it would work, the cost would be low. You’d still have to buy the machine, but I don’t know about the cost on that.

Q: So it’s not a Ferrari.

A: No, no.



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

Building on Dr. J. Hans DeVries’ intent-to-treat analysis of the CAT Trial at ADA 2012 (see page 86 of our coverage at, Dr. Eric Renard presented per-protocol results. As a reminder, the CAT Trial was a crossover-design comparison of 24 hours  under open-loop control, closed-loop control with the Padova/Pavia/UVa (iAP Consortium)’s MPC algorithm, and closed-loop control with the Cambridge University team’s MPC algorithm (CAM). Dr. Renard explained that the per-protocol analysis excluded time during technical difficulties due to pump, sensor, or operational issues. Overall, the per-protocol analysis showed similar results to the intent-to- treat analysis: 1) time in range was comparable among all three conditions; 2) time in hypoglycemia was significantly lower during closed-loop control; 3) mean average glucose was higher with closed- loop control; 4) time in hyperglycemia was significantly higher with closed-loop control. However, Dr. Renard seemed to believe that the higher average glucose and time in hyperglycemia were preferable to the hypoglycemia observed during open-loop control. Additionally, Dr. Renard stressed that patients were achieving comparable time in range during closed-loop without the management burden of an open loop. Automated control works, he concluded. We believe that as the “closed loop” becomes closer, some patients may prefer “their” control to “closed loop” and this will likely create some conflicts in how the control is assessed.

  • As a reminder, the CAT Trial assessed the primary endpoint of time in target range in 48 patients (47 completed) with type 1 diabetes; secondary endpoints included 1) time spent in hypoglycemia; 2) time spent in hyperglycemia; and 3) mean plasma glucose. The study was conducted in six centers (three of which were naïve to closed-loop systems), with eight patients at each center. Patients used the OmniPod, Dexcom Seven Plus, and either the UCSB artificial pancreas system or manual validation from a nurse. Patients were studied in three non- consecutive 24-hour periods under three different conditions (open-loop, closed-loop with CAM, closed-loop with iAP; order of the conditions was randomized). The 24-hour period included three meals and an exercise excursion meant to challenge the algorithm. For additional protocol detail, please see our ADA 2012 report.
  • The per protocol analysis excluded 0.4% of time in the open-loop arm, 13% of time in the iAP arm, and 17% of time in the CAM arm. Time was excluded for technical difficulties including pump, sensor, or operational failures.
  • There was no significant difference between time in range between open-loop or closed-loop control, which was also observed in the intent-to-treat analysis. Dr. Renard explained that this finding should be considered in the context of the reduced burden on patients during closed-loop control. Automated control works, he said, with the benefit of reduced hypoglycemia (see below). Thus, he argued constant validation of the algorithms’ suggestions is not needed.
  • In line with results from the intent-to-treat analysis, patients achieved significantly less time in hypoglycemia with iAP and CAM compared to open-loop control (p=0.001); however, this was at the expense of significantly higher mean blood glucose (p=0.01) and time in hyperglycemia (p=0.01). iAP and CAM were comparable across all measures. Dr. Renard stressed that insulin levels were higher during open-loop control and that open-loop patients required a significantly higher number of hypoglycemia rescues (p=0.001). During Q&A he observed that though the average blood glucose was better during open-loop control, one has  to ask whether mean blood glucose or time in hypoglycemia is more important. It seemed to us that Dr. Renard would take the latter view.
  • Dr. Renard argued that the trial demonstrated the feasibility and safety of using the closed-loop in a large number of patients. Notably, he said that the system’s performance would carry over to other centers irrespective of those centers’ previous closed-loop experience.

Questions and Answers

Q: Did you look at any changes in variability? We saw reduction in hypoglycemia but the same amount of time in target. Could the same glycemic results have been achieved in open-loop control by simply reducing the basal rate? And what was the algorithms’ target?

A: For patients in open-loop, insulin delivery was a patient decision, so no interventions in basal rate were made. It is common to see patients pushing too-high basal levels. For the algorithms, we targeted normal glucose control close to 100 mg/dl.

Q: Did you measure variability by standard deviation?

A: It has not been computed, but it seems that variability was reduced during closed-loop control. Another important remark – we wanted first to show the safety of the closed-loop, so we think that by choosing a better algorithm we could have less time in hypoglycemia and improve blood glucose control. The goal of the study was to show that this was feasible in a large number of patients. The next trial will do better at targeting the good control.

Q: Patients on open-loop achieved blood glucose targets better than the algorithms. Were you surprised?

A: The average glucose was better, but it all depends on whether you think what is important is mean blood glucose or time in hypoglycemia. To reach a good blood glucose with open-loop control means you are often in hypoglycemia. This is the difference with algorithms.

Q: There was no difference between SMBG and YSI calibration – was that retrospective?

A: It was randomized from the beginning. It was prospective. There were patients calibrating with YSI and others with SMBG.



Roman Hovorka, PhD (University of Cambridge, Cambridge, United Kingdom)

Dr. Hovorka presented results of a 36-hour crossover-design study that tested whether smaller meal- time boluses might reduce total insulin dose and risk of late postprandial hypoglycemia, without compromising glucose control. Adolescents with type 1 diabetes (n=8) spent alternating clinic visits under semi-closed-loop insulin delivery, with different amounts of prandial insulin (either 75% or 100% of the amount recommended by the pump’s bolus calculator). The smaller mealtime boluses led to   similar glucose control as measured by mean, time within the target of 70-180 mg/dl, time below 70 mg/dl, and time above 180 mg/dl – all with a total 36-hour insulin dose that was 10.6 units smaller overall, including lower basal rate. Dr. Hovorka hypothesized that acute postprandial insulin resistance may have been the reason that the larger size of the standard boluses did not translate to better control; he said that further studies were warranted to clarify why different insulin levels led to similar glycemia in adolescents (the finding was not replicated in adults in an AP@home study).

  • In this semi-closed-loop study, patients received mealtime insulin boluses of either 100% or 75% the dose recommended by the pump’s bolus calculator. Each patient in the crossover-design study underwent both test conditions, in randomized order. These eight adolescents with type 1 diabetes had mean age 15.9 years, mean A1c 8.9%, and median total daily insulin dose 0.9 U. The 36-hour study included two nights and one day, with moderate exercise  on a stationary bike in the morning and afternoon. Manual boluses were given for meals (50-80 g carb) but not snacks (15-30 g carb).
  • The closed-loop system consisted of an Animas 2020 insulin pump, a model predictive control algorithm, and a single FreeStyle Navigator sensor calibrated per manufacturer’s instructions; Dr. Hovorka noted that the Navigator’s accuracy during the study was about as high as he’d ever seen (99.8% of reference-paired measurements in the Clarke A zone, 0.19% in the D zone). Control actions were taken by the system every 15 minutes.
  • Smaller mealtime boluses led to similar glycemic control with significantly lower 36-hr insulin dose (61.9 vs. 72.5 U, p=0.1) – reflecting both a ~25% decrease in mealtime bolus and a slight decrease in closed-loop basal rate – and significantly lower mean plasma insulin concentration (186 vs. 252 pmol/l). The smaller-bolus approach led to a slight increase in percentage of time spent above 10.0 mmol/l (180 mg/dl), mainly due to a larger glycemic excursion after dinner on the study’s second night. However, the change in hyperglycemia was not statistically significant, mean plasma glucose was identical at 8.4 mmol/l (155 mg/dl), and the only instance of hypoglycemia below 2.5 mmol/l (<45 mg/dl) occurred under the standard-bolus condition. Total glucose appearance and glucose disposal, estimated using intravenous glucose with a stable radio-labeled tracer ([6,6-2H]glucose), were similar for each bolus strategy.


Smaller bolus

Standard bolus

p- value

Time within 3.9-10.o mmol/l (70-180 mg/dl)




Time below 3.9 mmol/l (<70 mg/dl)




Time above 10.o mmol/l (>180 mg/dl)




Total glucose appearance (umol/kg/min)




Glucose disposal (umol/kg/min)




Note: Percentages do not sum to 100% because they represent group medians rather than means

Questions and Answers

Comment: I think you are describing a clinically well-known phenomenon – that if your patients are slightly over-insulinized, then a slight reduction of dose does not lead to worse glucose control. As you showed, the incoming doses of 0.9 U/kg were high for adolescents. But I can offer no mechanistic understanding, of course.

Q: Do you think that this finding applies to other age groups, as well?

A: We tested this approach in the AP@home study and found that it does not apply to adults.

Q: Do you think that exercise might have been what reduced the need for basal in the afternoon or morning?

A: I don’t think there was interaction with the exercise – the observed drop in insulin concentration was already apparent overnight, after the first dinner. The exercise may have had an additional effect.

Q: Many reports show reduced accuracy of CGM in the hypoglycemic range. Was this something you found?

A: We were fortunate enough not to have much hypoglycemia, so I can’t really speak to that from this study. Our collective data suggest that coefficient of variation is greater at lower glucose – though this is only because you have a smaller denominator. We have found that Navigator accuracy is similar across the range.



Tao Yuan, MD (Peking Union Medical Hospital, Beijing, China)

Dr. Tao Yuan presented a small comparison of two ways for insulin-allergic diabetes patients to be desensitized to insulin. The patients were treated with diluted preparations of insulin using either an insulin pump (n=5; starting basal rate of 0.01 u/hr gradually increased to 1.4/hr over the course of three days) or subcutaneous injection (n=8; dose increased from 0.00001 to 20 u over the course of roughly seven days); treatment assignments were based on patient preference rather than randomization. Doses that provoked immune response were held steady until the immune response disappeared. Pump use was subjectively more convenient for the clinical staff, and it was associated with a slightly better success rate (100% vs. 87.5%). Rare though insulin allergy is, we are glad that attention is going toward optimizing its clinical treatment.

Questions and Answers

Q: Is insulin allergy more common in China than elsewhere?

A: It is rare, but my hospital is the best in China, so patients come from all over the world.


Oral Presentations: Profiling Glucose and Clinical Trials


Martin Prázný, PhD (Charles University in Prague, Czech Republic)

Dr. Martin Prázný presented data demonstrating that standard deviation, as a measure of glycemic variability, strongly correlates with the presence of microvascular complications (specifically, retinopathy, nephropathy, and peripheral neuropathy). In Dr. Prázný’s observational, case-control trial of 32 patients with type 1 diabetes, he compared 12 or 14 consecutive day CGM records of patients with microvascular complications (n=16) to those without complications (n=16). The CGMs used were the iPro2 and Dexcom Seven Plus and were used in masked mode. Standard deviation (SD), coefficient of variation (CV), and mean amplitude of glucose excursions (MAGE) of total CGM records were all significantly higher in patients with complications compared to those without complications (both when all complications were pooled and when each complication was analyzed separately). In a multivariate analysis, A1c, duration of diabetes, and age all failed to explain the presence of complications, but standard deviation did with an odds ratio of 4.53. Additionally, SD exhibited a significant positive correlation with vibration perception threshold (a measure of neuropathy), and exhibited a borderline significant correlation (p=0.056) with microalbuminuria. CV and MAGE exhibited a significant correlation with microalbuminuria. Neither mean absolute glucose (MAG) or continuous overall net glycemic action (CONGA) as measures of glycemic variability distinguished well between patients with and without complications. Dr. Prázný concluded that, since SD is very easy to calculate and is strongly associated with microvascular complications, it was the most efficient method of measuring glycemic variability. We believe that since this study correlates glucose variability with the presence of concomitant complications and not glucose variability as a predictor of complications, the clinical relevance is limited; the study assumes that, if glucose variability is to explain the presence of complications, that patients that had higher variability during the trial also had high variability before developing  complications.

  • Patients in the trial without microvascular complications (n=16) had similar characteristics to those with microvascular complications (n=16). Respectively, average age was 43 and 39; diabetes duration was 21 and 18 years; mean blood glucose was 9.2 mmol/l (~166 mg/dl); A1c was 8.6%; BMI was ~26 kg/m2.

Questions and Answers

Q: I agree that SD is fast and easy, but I’m afraid I have a few problems with the trial, especially with its size. When you did your correlation you found no relation between complications and A1c, which we know isn’t true. I think in the analysis of the DCCT cohort you’ll find that there was a relationship between SD and A1c, meaning higher A1c was associated with greater SD. Rather disappointingly to me, SD is not independently related to glycemic complications directly. But I think what really matters at the end of the day is A1c.

A: The patients in our study did have similar A1cs, but that doesn’t say anything about where they will be in five to 10 years from now.

Q: Speaking of the relationship between glycemic variation and complications, it seems like a chicken/egg question. Which comes first?

A: I must say it’s a question of believing or not believing. I think there are no clues that glycemic variability induces complications directly. Variation can induce oxidative stress, etc., which may explain its association with complications.


Oral Presentations: Profiling Glucose and Clinical Trials


Carsta Koehler, PhD (Dresden University, Dresden, Germany)

Dr. Carsta Koehler presented a study which investigated whether an individual’s CGM profile was reproducible after a four-year period during which glycemic control remained stable. Though well presented, the findings were not as meaningful as we had hoped, especially because the less-accurate  and higher-hassle Gold Medtronic CGM was used. The study included CGM data from 44 type 2 diabetes patients from Germany who, four years earlier, had participated in the 2007 ORIGIN CGM substudy  (the results of which are published in Hanefeld et al., Diabetic Med 2010). Participants’ characteristics remained stable during the four-year follow-up period, with the exception of a slight increase in fasting plasma glucose (mean BMI of ~30 kg/m2, A1c of 5.7-6%; FPG changed from 104 mg/dl to 111 mg/dl) The study collected 72-hour data on interstitial glucose (i.G.) and compared several CGM parameters to  those obtained during the 2007 ORIGIN CGM substudy (parameters included: average, fasting,  minimal, and maximal i.G.; standard deviation of i.G. [SD]; mean average glucose excursions [MAGE]; AUC for 24-hour i.G. and 2-hour post-meal i.G.; time in mild or severe hypoglycemia). Participants’ individual CGM profiles mirrored those obtained four years prior, and correlation analysis found a correlation across several parameters (including SD, MAGE, AUC two hours post-meal, and time in severe hypoglycemia) between baseline and follow-up.

Questions and Answers

Q: What were you expecting to find? You reported stability in terms of CGM and glycemic variability. What was your hypothesis? What did you anticipate to find over four years?

A: Our hypothesis was that if you have stable glycemic control in the patient (no change in treatment), the pattern of your glycemic profile does not change – if you have a rapid increase in blood sugar after the meal and a long decrease in the postprandial phase, you have the same thing two or three years later.

Q: CGM is quite expensive. Do you see subgroups which would have a particularly high benefit from using CGMS and where it is clearly realistic to use CGM?

A: I don’t think we can perform CGM in all type 2 diabetes patients, but I think that we have a chance to use this diagnostic method in patients with a risk for hypoglycemia on in patients with cardiovascular risk. Other studies have investigated these patients – patients with diabetes and cardiovascular risks – and the presentation of this data is tomorrow.



Bruce Wolffenbuttel, PhD (University Medical Center Groningen, Groningen, The Netherlands)

Dr. Bruce Wolffenbuttel and colleagues conducted a post-hoc analysis of the DURABLE study for two purposes: 1) to assess the relationship between A1c and self-monitored blood glucose (SMBG) across different racial and ethnic groups; and 2) to evaluate the use of estimated average glucose (eAG) values derived from A1c as an alternate way to report blood glucose levels. The study included data from 1941 participants from four racial/ethnic groups (64% Caucasian; 16% Asian 12% Hispanic; and 6% African descent; data from “other” not included). The investigators found that at each mean blood glucose (MBG) level (108, 162, 216, and 288 mg/dl), Hispanics and those of African descent exhibited higher A1c levels relative to Caucasians (difference of ~0.3-0.4%); Asians also exhibited higher A1c levels vs. Caucasians, though to a lesser degree. Calculating eAG values and plotting them against SMBG showed that at MBG levels ≤11.4 mmol/l (205.2 mg/dl), eAG overestimated actual blood glucose levels while at MBG >11.4 mmol/l (205.2 mg/dl), eAG underestimated true blood glucose levels. After citing possible reasons why the A1c vs. MBG relationship differs by racial/ethnic group (details below), Dr. Wolffenbuttel concluded that doctors should consider a patient’s racial/ethnic background when assessing A1c and that eAG has limited clinical value.

  • The study first assessed the relationship between A1c and self-monitored blood glucose (SMBG). Plotting baseline A1c vs. the mean of all SMBG measurements for each participant (one data point per person) yielded a rather diffuse scatterplot with several corresponding A1c measurements for each mean SMBG value; nevertheless a positive linear relationship was observed for each racial/ethnic group. Mean regression analysis allowed investigators to estimate how A1c levels differ by mean blood glucose levels (MBG) in different racial/ethnic  groups.
  • The investigators then calculated an eAG value for every participant based on  his/her baseline A1c, using an equation obtained from the A1c-Derived Average Glucose (ADAG) study (Nathan et al., Diabetes Care 2008): eAG (mg/dl) = [28.7 * A1c (%)] – 46.7.  Plotting each person’s eAG vs. his/her mean of all SMBG measurements yielded a regression equation that deviated substantially from the expected regression equation. The investigators   then calculated a “mean blood glucose index (MBGI)” for each participant using the equation [MBGI = eAG – {mean of all SMBG values}], with the expectation that MBGI would equal zero  (for every mean SMBG value) if eAG is a correct representation of mean SMBG. However, plotting MBGI vs. mean SMBG for each person yielded a linear regression equation with a strong negative slope instead of the expected horizontal line positioned at MBGI=0. The plot showed that at MBG levels ≤11.4 mmol/l (205.2 mg/dl), eAG overestimated actual blood glucose levels while at MBG > 11.4 mmol/l (205.2 mg/dl), eAG underestimated true blood glucose levels.
  • Dr. Wolffenbuttel cited several possible reasons why the relationship between A1c and mean SMBG differs by racial/ethnic group. He explained that several factors beyond blood glucose levels independently influence A1c, including age, gender, BMI, MCH, MCHC, smoking, and alcohol consumption. Furthermore, both nutrition and genetic differences between racial/ethnic groups may play a role in the observed discrepancy. Dr. Wolffenbuttel also noted that studies have reported glucose-independent differences in A1c between blacks and whites.


Oral Presentations: Can We Improve Outcomes in Diabetic Pregnancy?


Anna Secher, MD, PhD (University of Copenhagen, Copenhagen, Denmark)

Dr. Anna Secher presented a very disappointing study testing intermittent use of CGM in pregnant women with type 1 and type 2 diabetes. Patients in the CGM group (n=79) wore the Medtronic Guardian real-time CGM (Sof-Sensor) for six days at a time during weeks eight, 12, 21, 27, and 33 of pregnancy  and were compared to a non-CGM control group (n=75). Patients had a well-controlled A1c at baseline (~6.7%), though a high rate of severe hypoglycemia (~18%) that one of the doctors in Q&A pointed out probably reflects significant glycemic variability. Overall, there were no significant differences between the two groups in any glucose parameters at any point in pregnancy. Further, rates of macrosomia, pre-term delivery, and neonatal severe hypoglycemia were not significantly different between the two groups. We were quite surprised to see such neutral results though we think the study suffered from a number of major limitations: 1) the trial was done before wide trial use of the Enlite or Dexcom Seven Plus or G4 was widely available; 2) intermittent use of CGM for only six day periods following five   study visits (rather than continuous use); 3) low compliance with the prescribed intermittent wear (64%); and 4) high rates of baseline severe hypoglycemia that did not improve by the end of the study. Given these limitations, we are highly skeptical of these results and feel very strongly that CGM can be a very valuable tool for pregnant women with diabetes. We look forward to further studies of continuous, real-time CGM using next generation technologies.

  • This study assessed the use intermittent real-time CGM as part of routine pregnancy care in 154 women with pre-gestational diabetes. Women with type 1 diabetes (80% of the study population) or type 2 diabetes were included and were randomized to an intermittent real- time CGM group or a non-CGM control group. In the CGM group, the Medtronic Guardian CGM (Sof-Sensor) was prescribed to be worn for six days at a time after study visits at weeks eight, 12, 21, 27, and 33. Both groups saw a diabetologist every second week and seven-point SMBG profiles were analyzed. Women willing to wear CGM between study visits were allowed to do so – the speaker did not mention how many women choose to do this, though we suspect a low number did. Patients’ real-time CGM readings were used to adjust diet and insulin using locally developed guidelines,. No specifics were provided on the guidelines, though in Q&A, Dr. Secher suggested that severe hypoglycemia is a real problem at their clinic and this directs a lot of their focus. Patients were advised to conduct SMBG before and after meals (90 minutes) and at bedtime. Pre- prandial targets were 72-108 mg/dl and post-prandial targets were 72-144 mg/dl.
  • There were no significant differences at baseline between the CGM (n=79) and non- CGM (n=75) groups. Women were in fairly good control and had a baseline A1c of 6.6% in the CGM group and 6.8% in the control group. However, we note that the rate of severe hypoglycemia was quite high in both groups (18% and 17% respectively) at baseline. A questioner in Q&A  pointed out that this suggests a high rate of glycemic variability. Duration of diabetes was 10 years in the CGM groups and 12 years in the control group. Only 22% of type 1 patients in the study  were on insulin pumps.
  • There was no significant difference in A1c at any point in pregnancy between the two study arms. Patients in the two groups tested SMBG equally frequently throughout the study. In both groups, 16% of women had at least one event of severe hypoglycemia throughout the study. We were surprised and disappointed not to see a significant improvement in severe hypoglycemia in the CGM group, though given the intermittent use and low compliance, this was probably not surprising – technology has to be easier to use in order to expect positive results..
  • Rates of macrosomia, pre-term delivery, and severe neonatal hypoglycemia were not significantly improved with CGM use – surprisingly, these trended (non- significantly) in the opposite direction. In the CGM group, 45% of pregnancies had macrosomia, compared to 34% in the control group. Pre-term delivery occurred in 29% of CGM group pregnancies vs. in 22% of control pregnancies. Severe neonatal hypoglycemia occurred in 13% of CGM group pregnancies and 14% of control group pregnancies. All results were not statistically  significant.
  • A subgroup analysis of type 1s and those who followed the protocol did not change the overall findings. In women with type 1 diabetes, macrosomia occurred in 50% of CGM pregnancies vs. 36% in the control group. Pre-term delivery happened in 32% of CGM  pregnancies and 27% of control pregnancies. In a per-protocol analysis, the results were similar – macrosomia in 49% of patients for CGM vs. 34% in the control group and pre-term delivery/severe neonatal hypoglycemia in 24% of CGM patients vs. 22% of those in the control group. We would have been interested to see data for those patients that wore the CGM 24/7.

Questions and Answers

Q: I’m wondering about the type of population selected. I was surprised by the low A1c – the controls and study group were very close to 6.5%. The frequency of severe hypoglycemia was 16-17%. This was probably a population with lots of glycemic variability. That probably explains the low A1c and the frequency of hypoglycemia. Based on the results of CGM, did you intervene?

A: We did intervene based on the results of CGM readings. We analyzed these results together with each woman and we used them to adjust insulin therapy. With the present form of CGM, you must use it together with plasma glucose measurement.

Q: What was the time duration you used CGM – three days or six days?

A: It was used for six days.

Q: And it was the iPro2?

A: No, the Sof-sensor.

Q: Was it approved for six-day use?

A: Yes.

Q: Can you explain the large difference in macrosomia in spite of similar control? The CGM group numbers were higher than in the control group.

A: We were disappointed with the results and very surprised. We were very surprised that the factual numbers were higher, though they were not statistically significant. We have had many considerations about why. One reason is that in our center, we have a lot of data on severe hypoglycemia and pregnancy. We really want to avoid severe hypoglycemia. Perhaps if we’re focusing too much on severe hypoglycemia, we might pay a price on maternal hyperglycemic complications.

Q: This was a courageous trial. But I guess it was designed to fail from the beginning. You used intermittent real time CGM. Dr. John Pickup showed that you need to wear the sensor at least 75% of the time. Why didn’t you use it continuously or as continuously as possible?

A: We did indeed use it as much as possible. Women were encouraged to use it extensively and free of charge. The trial design was with intermittent use. That was for several reasons. We have much experience with CGM and we knew that numerous alarms and other limitations in the system limit compliance. By asking the women to use it continuously, it would have been hard to find an unselected population and sufficient numbers of women.

Q: With real-time CGM, the results are better in patients on pump therapy. They can take boluses easier and reduce basal rates. What proportion of your patients were on pumps?

A: In the women with type 1 diabetes, 22% were on insulin pumps, and the majority of them had insulin pumps that could be connected to the CGM system. We hoped for even better results in that subgroup of women. We did not find any improvement in that group either.


Posters: Pumps and New Devices


Whitehurst, A.E. Colvin, A.D. DeHennis, J.C. Makous, M. Mortellaro, S. Rajaraman, J. Schaefer, D. Smith, S. Tankiewicz, O. Tymchyshyn, S. Walters, X. Wang (Senseonics, Inc., Germantown, MD)

This poster detailed the design and performance of Senseonics’ (formerly Sensors for Medicine and Science) fluorescence-based implantable glucose sensor in 12 people with diabetes that participated in three 28-day pilot studies (the sensor is designed for implantation for at least six months). Sensors for Medicine and Science/Senseonics was founded in 1999 and they have worked for almost 15 years on the development of their implanted sensor system. This is the best human clinical data that they have presented to date. Sensors were subcutaneously placed in the wrist (n=4) or upper arm (n=16) and were powered by an externally worn armband reader or wristwatch. The studies included six clinic visits of eight or more hours each and sensors were calibrated with one blood glucose value at the beginning of each clinic day. A second blood glucose value was used for calibration after 12 hours if the clinic visit  was longer than 12 hours. CGM readings were prospectively determined and compared to YSI sampled every 15 minutes in clinic. Overall MARDs for the 20 sensors were 14.1% (upper arm), 16.8% (“improved” sensors in wrist), and 12.3% (“improved” sensors in upper arm), ranging from a low in one patient of 8.4% to a high in one patient of 29.6%. The “improved” sensors in the upper arm had 82% of points in the Clarke Error Grid A-Zone, 17% in the B-Zone, and 1% in the C- and D-Zones. This early feasibility data is encouraging, though we look forward to an upcoming six-month long trial in Germany and the UK that will help clarify some of the unanswered questions about the technology (see below).

  • Senseonics has developed a fluorescence-based implantable subcutaneous glucose sensor “designed to remain inserted for at least six months” (we note that this series of feasibility studies only tested the sensor for 28 days). The cylindrical sensor is 3 mm in diameter and 14 mm long – based on the poster’s picture, three of the sensors would fit on a 2-cent euro coin (quite small!). The sensors are inserted in the doctor’s office under lidocaine in an average of four minutes.
    • The implantable sensor uses an external reader (worn on an armband or wristwatch), which provides power to the implanted sensor, receives CGM data, alarms the user based on CGM values, and stores data for USB upload. The reader wirelessly provides power to the sensor using a wireless inductive link (13.56 Mhz). The reader also processes the implantable sensor data to determine sensor glucose values and rates of change. It can store up to six months of data, includes a beeper and a vibration motor to alert the patient (e.g., when glucose passes a threshold value), has a Bluetooth Low Energy link for wireless communication with a smartphone app, and features a USB port for charging and data exchange. The reader is pictured on the poster and appears as a small plastic box (about the size of a small matchbox) with a power button, a small screen (we presume sensor values and trends appear on it), and an elastic armband that secures the device to the user’s arm.
    • Senseonics has developed apps for the iPhone and iPod Touch that wirelessly operate as a user interface device for the reader, a display for CGM data, and a mechanism for providing user input. The poster noted that the apps are not used to process sensor data or store data long-term – this seems key from a regulatory perspective. The poster also showed color screenshots of the apps, which look quite   sharp: a large sensor glucose reading in big font, a trend arrow, colored graphs, and warnings (e.g., Caution: Glucose above target!). The apps also allow the user to enter data on daily events (e.g., meals, insulin, and exercise), which is immediately transmitted to  the reader.
  • A total of 20 implantable sensors were tested in 12 patients across three 28-day feasibility studies. Sensors were inserted in the wrist and upper arm and some patients wore two sensors at a time. Eight sensors were tested in an initial cohort of patients, followed by testing of twelve “improved” sensors in second and third cohorts (the poster notes a slight change was made to improve long-term stability). The 12 patients were between 22 and 65 years old, had type 1 or type 2 diabetes, an A1c <10%, and a BMI < 35 kg/m2 (mean values and further population characteristics were not provided).
    • The studies included six clinic visits of eight or more hours each (days three, six, 12, 18, 24, and 29 [post-implant]). During the in-clinic visits, blood samples were taken every 15 minutes and processed using a YSI blood glucose analyzer. A single blood glucose value from the beginning of each clinic day was used to calibrate each sensor for each session from a Roche Accu-Chek Aviva SMBG, and the sensor glucose values were calculated prospectively for the session. A second glucose value was used for calibration after 12 hours if the clinic visit was longer than 12 hours.
    • Two reader prototypes were developed for the pilot clinical studies: a wristwatch reader (designed for use with a sensor inserted subcutaneously in the dorsal wrist) and an armband reader (designed for use with a sensor inserted subcutaneously in the upper arm). The readers were each programmed to read the implanted glucose sensor every two minutes.
  • Combined MARDs for the 20 sensors were 14.1% (upper arm), 16.8% (improved sensors in wrist), and 12.3% (improved sensors in upper arm). MADs ranged from 12 mg/dl to 19 mg/dl. The highest MARD experienced by any subject was 29.6% and the lowest MARD was 8.4%. The highest MAD was 27 mg/dl and the lowest was 7 mg/dl. The worst overall accuracy came in the four sensors implanted in the wrist, though the highest single instances of inaccuracy came in the initial upper arm cohort. The distribution of glucose values in the study was not reported on the poster. There were some points as low as 50 mg/dl and some as high as 350 mg/dl, but the Clarke Error Grid Analysis presented contained most points in the euglycemic range.
  • The four improved sensors placed in the upper arm had 82% of points in the Clarke Error Grid A-Zone (n=892), 17% in the B-Zone (n=184), 0.18% in the C-Zone (n=2), and 0.5% (n=5) in the D-Zone. Clarke Grid analyses were not presented for the other cohorts. We note that this cohort had the best performance of the three studied.
  • Other companies have shown good implanted sensor data for up to 30 days, only to have problems with encapsulation over subsequent days and weeks. Performance, reliability, and sensor chemistry are certainly challenged in the body’s environment as wear time increases, and it will be valuable to see how Senseonics performs over time. Additionally, we look forward to learning more about explantation after the six-month implant. As we understand it, explant takes about four minutes and there have not been concerns with excising a large mass thus far.
  • Senseonics recently submitted a six-month clinical trial protocol to MHRA in the   UK and BfArm in Germany. The trial protocol was pre-approved by TUV SuD (the selected notified body) and it contains both accuracy (in-clinic) and home use components. As we understand it, if the end points are met, regulators have agreed that this is an appropriate amount of clinical data to support approval.
  • From a regulatory perspective, we wonder how long of a study the FDA would require for a six-month implantable CGM; we imagine approximately nine to twelve months. Senseonics plans to discuss the European clinical trial protocol with the FDA through the new “Pre-Submission” filing process that replaces the IDE. The hope is that the company can use the European data in the US submission, though this may be optimistic given the recent history with other devices (e.g., Medtronic’s Veo). Given the challenges in the US, we expect Senseonics will focus on Europe initially and then turn to the US at some point down the road.
  • It will be interesting to assess demand for implantable sensors from patients. Generally speaking, we imagine if reliability is high, the interest in implantable would expand the market to at least some degree. To what degree this cannibalizes the current market, which is only about 5% of type 1 patients and a very small minority (well less than 1%) of type 2 patients, will be interesting to see.
  • On the reimbursement front, the plan is to do insertions and explants in the doctor’s office – as a result, it will be key for the company to get those processes covered by insurance. Senseonics will begin those effectiveness trials after the pivotal trials are complete.


Posters: Blood Glucose Self-Monitoring


F. Kulozik, I. Platten, and C. Hasslacher (Diabetesinstitut Heidelberg, Heidelberg, Germany)

This poster, which was supported by Roche, compared the accuracy of 25 commercially available blood glucose meters in a wide range of glucose values (60-300 mg/dl) in 37 insulin-dependent patients with diabetes. All 25 meters were accurate within the requirements of the current ISO 15197 standards,  though 14 of the 25 meters failed to meet the proposed ISO standards (95% within ±15 mg/dl for <100 mg/dl and ±15% for >100 mg/dl). The five most accurate meters according to this study were: 1) the Roche Accu-Chek Compact and Bayer Contour Next USB (tie); 3) Roche Accu-Chek Mobile; 4) Abbott FreeStyle Lite; and 5) Sanofi iBGStar. Meters that failed to meet the proposed standards were mostly from smaller companies, though the LifeScan OneTouch Verio, Bayer Breeze and Contour Plasma, and Sanofi BGStar came up short of the new standards. Obviously, results vary based on strip lots, though we were encouraged to see fairly high sample numbers (250-300) in this study. It was positive to see a study with the newest meters included together, as some recent studies we’ve covered did not compare  all in one study (e.g., Freckmann et al., Journal of Diabetes Science and Technology 2012; Bayer’s North American Comparator Trial on pages 30-31 of our AADE 2012 report at The poster did not mention any study sponsor, surprisingly – we found out after our original viewing that Roche was the sponsor. Oddly, the Roche meter results were adjusted by 5% due to the strips’ hexokinase technology – we are in the midst  of trying to get background on the rationale for this. In any case, we’ll be interested to see what happens with the new ISO standards – if they are indeed tightened, it should help push companies to further improve accuracy and could even bode well for CGM (e.g., calibration).

  • This study compared the accuracy of 25 commercially available glucose meters in five different blood glucose ranges in 37 insulin-dependent patients with diabetes (n=24 type 1, n=13 type 2). Patients had a mean age of 50 years, a mean A1c of 7.5%, a mean duration of diabetes of 17 years, and no concomitant use of substances that could affect blood glucose readings. Blood glucose levels ranged from 60-300 mg/dl in the study. For each SMBG system, 230-300 paired values were obtained. Results were compared to an internally and externally validated laboratory reference method (Hitado Super GL). Results from the reference standard were converted from whole blood glucose values to plasma equivalent blood glucose values using the formula: plasma equivalent blood glucose (mg/dl) = 1.11 x whole blood glucose (mg/dl). Results were compared to current ISO 15197 standards (95% within ±15 mg/dl and ±20% for <75 mg/dl and >75 mg/dl) and proposed standards (95% within ±15 mg/dl for <100 mg/dl and ±15% for >100 mg/dl).
    • Since the Accu-Chek devices are calibrated by the hexokinase method, results from these meters were specifically adjusted by an increase of 5% (according to the poster, several studies have shown that blood glucose levels measured this way are 3.5-6.7% higher than glucose oxidase values). It struck us as somewhat odd that only the Roche meter results were adjusted. This also seemed somewhat unfair given that patients using the meters in a home-use environment would not see “adjusted” results on the meters.
  • The five most accurate meters were: 1) the Roche Accu-Chek Compact and Bayer Contour Next USB (tie); 3) Roche Accu-Chek Mobile; 4) Abbott FreeStyle Lite; and  5) Sanofi iBGStar. There was a wide range of overall accuracy between the glucose meters (80.4%-99.6%) and only three meters reached >98%. Accuracy was better at higher blood glucose values than at lower values. Data was collected and evaluated in the ranges of 50-99 mg/dl, 100- 149 mg/dl, 150-199 mg/dl, 200-249 mg/dl, and 250-300 mg/dl. For brevity in the table below,   we have omitted the breakdowns and only included the overall data. Results below are presented as the percentage of values that meet the proposed ISO standards: within ±15 mg/dl for <100 mg/dl and ±15% for >100 mg/dl. The bold divider line denotes meters that failed to meet the 95% threshold.


Overall Accuracy (n)

Accu-Chek  Compact

99.6%, n=282

Bayer Contour Next USB

99.6%, n=275

Accu-Chek Mobile

99.3%, n=294

Abbott FreeStyle Lite

97.7%, n=301

Sanofi iBGStar

97.6%, n=249

MyLife Pura

97.4%, n=271

Accu-Chek Aviva Nano

97.0%, n=303

LifeScan OneTouch Verio IQ

96.8%, n=278

LifeScan OneTouch Ultra Easy

96.8%, n=284

Bayer Contour USB

96.4%, n=248

LifeScan OneTouch Vita

96.2%, n=290

Beurer GL 40

94.9%, n=254

Bayer Contour Plasma

94.7%, n=284


92.6%, n=258

LifeScan OneTouch Verio

92.5%, n=265

A. Menarini Glucomen LX

92.0%, n=251

GlucoSmart  Swing

91.7%, n=253

A. Menarini Glucomen LX Plus

91.5%, n=246

Wellion Calla

91.3%, n=252

Sanofi BGStar

90.1%, n=302

SmartLab Mini

88.5%, n=253

Smart Lab Sprint

87.1%, n=280

Bayer Breeze

86.7%, n=249

Beurer GL50

85.5%, n=228

Omnitest 3

80.4%, n=245

  • All 25 meters tested met the current ISO 15197 requirements, though 14 of the 25 meters failed to meet the proposed ISO requirements. The investigators also looked at accuracy per the proposed ISO thresholds for the five different glucose range buckets evaluated in the study. Four meters met the proposed ISO standards for every single glucose range: Accu-Chek Compact, Accu-Chek Mobile, Bayer Contour Next USB, and the OneTouch Ultra Easy. Another six meters met the proposed ISO standards for four out of the five glucose ranges: Accu-Chek Aviva Nano, Sanofi iBGStar, Bayer Contour USB, FreeStyle Lite, MyLife Pura, OneTouch Verio IQ, and OneTouch Vita.
  • Roche gave financial support for this study.



Stephanie Roze (Health EVAluation SAS, Lyon, France), Mark Cook (Medtronic UK, Watford, UK), and Peter Lynch (Medtronic International Trading Sarl, Tolochenaz, Switzerland)

This study evaluated the lifetime health and economic benefits of using CGM vs. SMBG alone. The results were quite powerful in favor of CGM: an additional two years of quality adjusted life expectancy and an additional three years alive and free of diabetes complications. Moreover, CGM was cost effective, with an incremental cost-effectiveness ratio (ICER) of just £17,932 per quality adjusted life year (QALY) gained, well under the NICE threshold of £20,000 per QALY. We note that these cost-effectiveness   results were substantially better than those presented by Dr. Michael O’Grady at ADA 2012 (see pages 48-50 of our full report) – based on the JDRF CGM trial, he found ICERs ranging from a low of $57,170 (A1c <7% and wearing the sensor for seven days) and a high of $130,060 (A1c >7, age >25, and five days of wear). We believe much of this poster’s analysis was driven by the high baseline A1c used (10%) and the fact that wear time did not seem to be taken into account. Still, we think the results are powerful   from a payer perspective and hope cost-effectiveness studies for CGM continue to show benefits, especially with next-gen technology.

  • This study used a computer simulation model to estimate the lifetime impact of CGM vs. SMBG alone on health complications and economic outcomes in type 1 diabetes in the UK. The researchers used the Core Diabetes Model, an internet-based, “highly validated” computer simulation model. Inputs came from John Pickup’s meta-analysis on CGM (BMJ 2011), a real-life observational study of CGM (Rose et al., ISPOR USA Conference 2012), hypoglycemia and quality of life data from the JDRF CGM trial (Diabetes Care 2010), and quality of life data from Currie et al., Curr Med Res Opin 2006.
  • The following model assumptions and inputs were used: baseline A1c of 10% (we believe this is what was used though it was not explicitly clear from the poster; we note this is quite a high A1c); a 0.9% reduction in A1c with CGM use (again, not totally clear from the poster’s methods); mean age: 27 years; mean diabetes duration: 13 years; mean SMBGs per day: 7.1 (SMBG alone) and 4.4 (CGM); annual rate of major hypoglycemic events: 27.7 events per 100 patient years (SMBG alone) and 15 events per 100 patient years (CGM); an adjusted quality of life to account   for reduced fear of hypoglycemic events in the CGM arm (U=0.0152); diabetes related treatment and complication costs (UK specific) taken from various published sources; yearly discount rates of 1.5% (clinical outcomes) and 3.5% (economic outcomes). The analysis did not include indirect costs, which would have led to even better health economic outcomes.
  • CGM had an incremental cost-effectiveness ratio (ICER) of £17,932 per quality adjusted life year (QALY) gained, well under the NICE threshold of £20,000 per QALY. Quality adjusted life expectancy was 1.9 years higher with CGM (11.9 years with SMBG alone vs. 13.8 years with CGM). Total costs of SMBG alone were £59,193 vs. £93,523 for CGM. In  a one-way sensitivity analysis considering no effect on the rate of major hypoglycemic events for CGM, the ICER was £23,067 per QALY. The one-way sensitivity analysis based on no reduction of fear of hypoglycemic events due to CGM led to an ICER of £21,336 per QALY. By using an equal discount rate for both health and economic parameters of 3.5%, the ICER was £25,975 per QALY. By varying the treatment costs by ±10%, the ICER varied by ±17% (14,883-20,980). It’s good to see these sensitivity analyses did not drastically affect the main finding.
  • Use of CGM was linked to nearly three more years alive and free of diabetes complications vs. SMBG alone.

Time Alive and Free of Complications (Years)


SMBG Alone


Change with CGM

End-Stage Renal Disease




Myocardial Infarction




Severe Vision Loss












Average for all Complications





Posters: Continuous Glucose Monitoring


Price and K. Nakamura (Dexcom, San Diego, CA); T. Bailey (AMCR Institute, Escondido, CA), M. Christiansen (Diablo Clinical Research, Walnut Creek, CA), E. Watkins (Profil, Chula Vista, CA), D. Liljenquist (Rocky Mountain Diabetes and Osteoporosis Center, Idaho Falls, ID)

This poster gave a concise, useful summary of the accuracy and reliability improvements of the Dexcom G4 Platinum Sensor over the Dexcom Seven Plus. We’ve seen snippets of the new data at ADA 2012, EASD 2012, and ISPAD 2012, though this poster put it all together and did a nice job comparing a broad array of metrics from the two sensors’ pivotal studies. The short story is that the G4 Platinum has improvements in every measure of accuracy and reliability over the Seven Plus (with the exception of Day 1 accuracy, which is marginally worse with the G4 Platinum). Especially striking are the G4 Platinum’s improvements in the hypoglycemia range and the strong durability of accuracy at the seven- day mark. Moreover, this poster made it clear that the G4 Platinum pivotal study was more robust than the Seven Plus pivotal study in a number of ways – it included more patients (72 vs. 53), more and  longer in-clinic days for each subject (three 12 hours vs. one eight hour), nearly double the sensors (108 vs. 67), and ten times more points in the YSI range <55 mg/dl (361 vs. 33). For more details, please see tables below. As a reminder, the G4 Platinum was approved by FDA on the Friday of EASD (October 5, 2012) and will begin shipping in late October (see our coverage of the FDA approval call at

  • The G4 Platinum has reduced the number of outlier sensors vs. the Seven Plus. The poster statistically summarized this as follows: G4 Platinum – mean ARD: 13.6%, median ARD 12.5%, standard deviation: 6.7% vs. Seven Plus – mean ARD: 16.7%, median ARD 14.4%, standard deviation: 8.6%. A histogram also summarizes individual sensor MARD data, illustrating that the G4 Platinum has far fewer sensors with a MARD >20%. This is encouraging from a patient perspective and we believe it should encourage more patients to stay on CGM and wear it a   greater proportion of the time. The landmark JDRF CGM study (NEJM 2008) found that patients in all age groups (adults, adolescents and children) who wore CGM at least six days a week had a clinically significant reduction in A1c. We look forward to patients using the G4 Platinum more and, as in the JDRF CGM study, getting greater clinical benefit.

Dexcom Seven Plus vs. G4 Platinum – Accuracy vs. YSI


G4 Platinum

Seven Plus

Sensors in Pivotal Study



Mean Absolute Difference (MAD)

20.8 mg/dl

24.7 mg/dl

Mean Absolute Relative Difference (MARD)



MAD within Biochemical Hypoglycemia (YSI <70 mg/dl)

10.6 mg/dl

16.0 mg/dl

MARD within Biochemical Hypoglycemia (YSI <70 mg/dl)



Number of Samples in Severe Hypoglycemia (YSI <55 mg/dl)



Percentage of CGM values that were

<70 mg/dl when YSI <55 mg/dl



Clarke Error Grid (CEG) A-Zone Overall



CEG A-Zone in Hypoglycemia (YSI

<70 mg/dl)



Mean Paired Sensor Coefficient of Variation



Sensor Display Rate



Sensors Lasting Seven Days




Dexcom G4 Platinum – Durability of Accuracy


Matched CGM- YSI Pairs

MARD 40-400


Percent within 20% of YSI*

Percent >40% of YSI*

Day 1





Day 4





Day 7






Dexcom Seven Plus – Durability of Accuracy

Day 1





Day 4





Day 7





*For YSI <80 mg/dl, the absolute difference is presented as the difference between CGM value and YSI, rather than the percent.


Corporate Symposium: Dexcom G4 Platinum Continuous Glucose Monitoring: Revolutionary Technology Bringing Patient Care to the Next Level (Sponsored by Dexcom)


Thomas A. Peyser, PhD (VP Science and Technology, Dexcom, San Diego, CA)

Dr. Peyser gave an excellent data-filled presentation on Dexcom’s new G4 Platinum sensor (the first   time we’ve seen the branding “Platinum” to refer to the Gen 4 sensor) with one major takeaway: this is a whole new level of CGM accuracy for Dexcom – and patients will surely benefit. He gave a deeper dive into the pivotal trial data presented at ADA by Dr. David Price (see page 28 of our report at and presented some very unique cuts of the accuracy data that we hadn’t previously seen. We also appreciated his commentary on how current accuracy metrics fail to account for patients’ varied experiences with individual sensors – in short, there are limitations to presenting overall averages. Dr. Peyser emphasized that the G4 Platinum’s improved accuracy and ease of use will result in a more positive experience for patients wearing the sensor, which will hopefully result in more sustained use and a greater clinical benefit. Considering the low rates of CGM penetration, we certainly hope the Dexcom G4 Platinum and Medtronic Enlite sensors can usher in a new era of accelerated CGM adoption. Of course, it’s tough to disentangle how much of CGM adoption is related to accuracy vs. reimbursement/cost vs. hassle factor, though we have no doubt that the improved technology will help convince more patients that the technology is worth wearing 24/7.

  • Regarding the regulatory status of the G4 sensor in the US, Dr. Peyser stated, “I’m confident that it will be approved soon. It could be next week, it could be next  month, or it could be three months from now.” As a reminder, the system received CE Mark on June 15, 2012 and was submitted to the FDA on March 31, 2012. As of Dexcom 2Q12 (see our report at, approval was expected before the end of 2012.
  • Dr. Peyser reviewed the G4 accuracy data from the recent pivotal study, which he characterized as a “huge advancement” versus the Seven Plus and for the whole category. With the Seven Plus, 74% of values fell in the Clarke Grid A-Zone, compared to 80% with the G4. The overall MARD has also improved from 15.7% to 13.2%. Historically speaking,  this compares to 22% for the GlucoWatch, 20% for the Guardian Real-Time, and 26% for the Dexcom STS – very good for recent comparisons, though the Abbott Navigator did achieve 82% A and 12.6% MARD in a similar 2007 study. As a reminder, the pivotal study occurred at four sites in 72 patients over seven days. It featured three 12-hour in-clinic sessions with YSI on days one, four, and seven (>9,000 paired CGM-YSI points). For more background on the pivotal study, see page 28 of our ADA 2012 full report
  • The G4 Platinum features significant improvements over the Seven Plus in the hypoglycemia range. For values <70 mg/dl, the G4 Platinum has a mean absolute deviation (MAD) of 11 mg/dl vs. 16 mg/dl for the Seven Plus. MARD in hypoglycemia has improved to 19.1% with the G4 Platinum compared to 27.3% for the Seven Plus. Clarke A Zone readings (<70 mg/dl) were 80% for the G4 Platinum vs. 62% for the Seven Plus. The percentage of CGM values <70 mg/dl when YSI read under 55 mg/dl is also improved with the G4 Platinum: 88% vs. 73% for the Seven Plus.
  • Dr. Peyser discussed how current accuracy metrics “often ignore patients’ experience while using CGM.” He noted that patients’ confidence or lack of confidence in CGM is often based on their own experience with individual sensors; e.g., “I had a good sensor” or “I had a bad sensor.” As a result, accuracy metrics should capture this variation in individual sensor performance, something that averages fail to capture. It was great to hear this patient perspective. (to illustrate the point, Dr. Peyser jokingly referred to the old statistics joke that has a man’s feet in the refrigerator and his head in the oven – he feels “pretty good on average”).
    • The G4 sensor has reduced the variability in sensor-to-sensor performance. Overall, the Seven Plus had a mean ARD of 15.9% and a standard deviation of 8.6%, which has been narrowed to 13.2% and a standard deviation of 6.7% with the G4 Platinum. In Dr. Peyser’s words, “This has important consequences in terms of how patients experience the use of CGM” – essentially, fewer “bad” sensors. He also showed an interquartile analysis that breaks down individual sensor MARDs. This was a unique way to display the data that we had not seen before – it basically gives a broader picture of the full spectrum of sensor accuracy, which we appreciated instead of just a single mean. Dr. Peyser noted that Dexcom was “surprised and delighted by the results.” Indeed, the table below shows that 50% of G4 sensors have a single digit MARD – quite impressive indeed.


All Sensors

Top 75% of All Sensors

Top 50% of All Sensors

Top 25% of All Sensors

Seven Plus Average MARD





G4 Platinum Average MARD





  • Dr. Peyser discussed the performance of the G4 Platinum as measured by the Clarke Error Grid, including an interquartile analysis. While some have criticized the Clarke Grid for being too loose overall, too tight in hypoglycemia, or just unsuitable for continuous data, Dr. Peyser believes it’s a useful metric to put accuracy in the context of clinical decision-making. However, he cautioned that CGM companies have abused this measure in the past by providing A+B Zone data together, which masks A Zone performance. The table below presents A Zone data for quartiles of individual sensors – Dr. Peyser emphasized that it’s “pretty incredible given where this technology was a decade ago” and this higher accuracy means users will have a “positive experience from most sensor wears.”


All Sensors

Top 75% of All Sensors

Top 50% of All Sensors

Top 25% of All Sensors

G4 Platinum Percentage in A Zone





  • CGMs are very susceptible to calibration error from SMBG, especially on day one. Dr. Peyser showed an example of a G4 sensor with a MARD of 21.4% and 69% in the Clarke Grid A Zone. The low overall accuracy was due to a highly inaccurate first day, which improved to a MARD of 7.2% on day seven. It turns out that the blood glucose meter reading used to initially calibrate the Dexcom was off YSI by 54 mg/dl (!), which set the sensor session up for inaccuracy from the get-go. After simulating the correct YSI value retrospectively, Dr. Peyser noted that the accuracy would have been spot on for the entire first day. This underscores the importance of teaching patients that getting a good fingerstick is really important for calibrating CGM. Of course, Dexcom has a goal of eliminating or significantly reducing fingerstick calibrations in the future, so we hope this becomes less and less of a problem over time.
  • The precision of the G4 Platinum is on par with SMBG (Dexcom “is nipping at the heels of the BGM companies”). In the G4 pivotal trial, 36 patients wore two sensors (left and right abdomen). The coefficient of variation (CV) between the two sensors was 7%, on par with SMBG and a favorable comparison to previous CGMs (CV of 12-20%). Dr. Peyser explained that this is a “whole different level of performance and precision that has never been seen before in this field.”
  • The G4 Platinum’s improved performance reflects changes in the sensor, the transmitter, and the receiver. The sensor features a 60% reduction in volume and an improved biocompatible membrane (i.e., reduced wound response, more consistent performance across a wider range of patients). The transmitter is “a little larger” though “still very small” and has upgraded to a 2.4 GHz radio frequency. The transmission range has increased to 20 feet from just five feet on the Seven Plus. Dr. Peyser also highlighted some of the reliability data presented at ADA: 97% data capture over seven days (i.e., 279/288 possible readings per day) and 94% of sensors lasting seven days. Lastly, the new receiver has been redesigned to look like the original iPod nano. The screen is now color and the receiver incorporates new algorithms that adaptively adjust over time to account for changes in the environment of the sensor.



Martin Prázný, MD, PhD (Charles University, Prague, Czech Republic)

Dr. Martin Prázný gave the audience a first look at Dexcom’s new Studio software, which will accompany the new G4 sensor. The most significant change from the current DM3 software is the addition of a pattern recognition tool that automatically detects low and high patterns throughout the day – this was excellent to see and should really improve the value of CGM data. The approach reminded us of LifeScan’s OneTouch Verio IQ blood glucose meter, and in some ways it also resembled Medtronic’s CareLink Pro 3.0. Besides the new pattern tool, the software looks largely similar to the older version with sections for hourly stats, daily trends, distribution graphs, modal day reports, daily statistics, and a success report (based on the screenshots, we assume it is PC-only, which is disappointing). An example use of the pattern recognition feature is summarized in the table below – target ranges and nighttime/daytime periods can be customized. Like CareLink Pro 3.0, the new Studio software also graphically shades areas of hyperglycemia and hypoglycemia on the modal day report. Further, the pattern tool provides an associated column entitled “Some Possible Considerations” – these are fairly basic things like “If before meals, adjustment to basal insulin,” “If after meals, adjustment to meal time insulin,” “Review carb counting, effect of exercise, alcohol, and or food choices” – however, they are useful reminders for patients and clinicians. We look forward to hearing more about Dexcom’s plans for software, especially the ongoing integration with SweetSpot.

Pattern Insights Summary

Nighttime Lows

(0 Found)

No significant patterns detected.

Daytime Lows

(2 Found)

Most significant pattern of lows found between 6:45 pm and 7:40 pm.

Nighttime Highs

(0 Found)

No significant patterns detected.

Daytime Highs

(1 Found)

Most significant pattern of highs found between 8:05 AM and 11:35 AM.



Bruce Buckingham, MD (Stanford University, Stanford, CA)

Dr. Buckingham shared brand new data on a nocturnal remote monitoring study using the Dexcom G4 CGM in forty-one patients at two diabetes camps over one week. Using a USB cable, the G4 receiver was connected to an Android cell phone running the UVA Diabetes Assistant software. The cell phone sent   the CGM data to a central server and to doctors’ computers and iPads at the camp. Patients were randomized to either control nights (no remote monitoring and fingersticks to determine hypoglycemia treatment) or remote monitoring nights. Notably, remote monitoring led to a 79% reduction in events <70 mg/dl (seven events vs. 33 events), a 100% reduction in events < 50 mg/dl (zero events vs. nine events), and an improvement in attendants’ response time to nocturnal alarms (no p-values included). Positive data aside, Dr. Buckingham was also highly positive on the G4 sensor itself: “These kids really liked this sensor. They really found it to be accurate…Dexcom has a real winner here.” Part of the randomization involved treatment with either mini-doses of glucagon or carbs. Interestingly, glucagon was associated with more recurrent hypoglycemia within three hours (i.e., it failed more often relative to carbs). It will be interesting to watch this over time and see if future studies are consistent with this finding.

  • There are two strategies to prevent severe hypoglycemia: 1) suspend insulin delivery using low glucose suspend (LGS) or 2) actively intervene with fast-acting carbs using CGM and remote monitoring. We thought this was an interesting way to position CGM on a similar playing field to LGS.
  • This pilot study tested the impact of remote monitoring with the Dexcom G4 sensor on nocturnal hypoglycemia at two diabetes camps. Forty-one patients (n=29 at Chinnock and n=12 at De Los Ninos; mean A1c ~8.4%) were randomized to a night with remote monitoring or a control night. On remote monitoring nights, patients used the G4 sensor and receiver, the latter of which was connected via a USB cable to a Samsung Android cell phone running the University of Virginia’s Diabetes Assistant application (for more on the Assistant, see page 14 of our ATTD 2012 report at CGM data while patients were sleeping was sent from the cell phone to a server at the University of Virginia through a cellular network or local WiFi (one of the camps did not have cell service). The UVA server then transmitted the data to the doctor’s computers in camp cabins and to portable iPads. The camp’s doctors could view all patients’ CGM readouts at once and intervened once blood glucose dropped below 70 mg/dl. The master display uses the red-yellow-green traffic lights for hypoglycemia and hyperglycemia and displays the CGM reading and trend arrow for each patient. On control nights, patients also wore  a CGM, though it was not remotely monitored and their treatment was based on standard nighttime SMBG testing.
  • Remote monitoring decreased the number of events <70 mg/dl by 79% and events <50 mg/dl by 100%. The number of nights with remote monitoring (161) was comparable to the number of control nights (179). No p-values were reported on the slide. For more details please see table below.


Remote  Monitoring

Control Nights

<70 mg/dl



>1 hour



>2 hours



<50 mg/dl



>30 minutes



>1 hour



  • Encouragingly, remote monitoring also reduced the response time and increased the response rate to nocturnal alarms. Seventy-seven nocturnal alarm events occurred in the remote monitoring group, and 100% of alarms were responded to. By contrast, the control group had 119 events and only a 54% response rate. Dr. Buckingham also displayed a box plot graph (unfortunately not numerically labeled) to demonstrate that response time to nocturnal alarms was lower and less variable in the remote monitoring group, and outliers were less extreme – a max response time of 80 minutes in the remote monitoring group compared to 118 minutes in the control group.
  • “These sensors worked really well. We had a lot of faith in the data and the results we were getting. It was impressive….it was very rare we got called and it wasn’t  low. Notably, the Dexcom G4 had a true positive alarm rate of 79% in this study, substantially better than historical data from other CGMs: 60% for the Navigator, 54-67% for the Guardian RT, and 54% for the Dexcom Seven Plus. It’s great to see the Dexcom G4 has indeed made serious improvements in the hypoglycemic region and this data seems to hold in a more real world  setting. We believe detection/avoidance of hypoglycemia represents an important reason why many patients choose to go on CGM in the first place, and it’s certainly frustrating for patients when false low alarms occur. We would guess that improving the true positive alarm rate could have a noticeable benefit both on patients’ clinical experience with the device (i.e., reducing hypoglycemia) and potentially on CGM attrition.
  • Despite all the rigors of the camp (e.g., swimming, sweating, sports), 81% of the sensors remained on until the completion of camp (five to seven days). Dr. Buckingham noted that 5% of patients had mild erythema from the adhesive and there was no significant edema or inflammation at the sensor insertion sites.
  • Part of the randomization process included treatment with either mini-doses of glucagon or carbs – interestingly, glucagon failed much more often within three hours of the initial treatment. Mean glucose rose to ~180 mg/dl with mini glucagon vs. ~160 mg/dl with carbs (no significant difference). However, Dr. Buckingham noted that glucagon treatment was associated with many more recurrent lows within three hours.



Jay Skyler, MD (University of Miami Miller School of Medicine, Miami, FL)

Dr. Jay Skyler provided an exciting glimpse into Dexcom’s pipeline, which includes integration with SweetSpot Diabetes Care, insulin pump partnerships, the Gen 5 smartphone compatible sensor, involvement in several artificial pancreas projects, predictive algorithms, and remote monitoring. Notably, Dr. Skyler shared new accuracy data on the special edition of the G4 sensor designed for the artificial pancreas – an overall MARD of 11.3%, 96% of sensors with a MARD <20%, and in two   example sensors, a sub-5% MARD that even beats out fingerstick monitoring accuracy. We also appreciated Dr. Skyler’s discussion of remote monitoring and predictive alerts, which he believes is a valuable alternative to LGS for two main reasons: 1) less risk of a roller coaster pattern because insulin  is not being suspended; and 2) MDIs can use predictive CGM, unlike LGS systems that only pumpers can use. We thought these were both valuable points and it will be interesting to see how Medtronic and Dexcom position their respective next-gen products against each other in the years to come.

  • Dr. Skyler reviewed Dexcom’s partnership with SweetSpot Diabetes Care to build an Internet-based data platform. He first reminded attendees of the many advantages of the SweetSpot platform: it is cloud-based, compatible with several glucose meters and all insulin pumps (except Medtronic), integration with electronic medical records, and advanced glucose  data analytics. Dr. Skyler believes the SweetSpot system will help transform diabetes care by enabling new models of care and moving diabetes care from the clinic into patients’ homes.
  • Turning to Dexcom’s insulin pump partnerships with Animas, Insulet, Roche, and Tandem, Dr. Skyler highlighted the benefits of integrated pump and CGM data – a recent study published in Diabetes Technology and Therapeutics (Frontino et al., 2012) demonstrated a 1% improvement in A1c in pediatric patients (<7 years old) with an A1c >7.5% using an Animas pump and Dexcom Seven Plus CGM. These young patients wore the sensors 83% of the time, demonstrating their wide utility even in such a young patient population. Dr. Skyler noted that the Gen 4 sensor will be integrated into the OmniPod and Animas pumps and will hopefully “be available sometime soon” (as of Dexcom 2Q12, a PMA supplement for the Animas Vibe is expected to be filed before the end of 2012, while the timeline for the Insulet product has not been announced – see our report at Meanwhile, the Gen 5 sensor will be part of future Roche and Tandem pumps. Dr. Skyler also said that it will only be a matter of time before the “Paradigm people” (i.e., Medtronic) want to have a “real glucose sensor with their pump.”
  • The Gen 5 sensor’s direct smartphone connection “looks like a really good way to go.” Dr. Skyler showed a picture of both a smart phone and a watch displaying a Dexcom CGM reading in a sleek interface. He was most positive on the potential of the Gen 5 CGM data to go directly to the cloud from the smart phone, a particularly nice advance for parents. He also expressed optimism in referring to how far the technology has come since the first gen came out in 2006.
  • Dr. Skyler shared new data on the special version of the G4 sensor designed for use in the artificial pancreas (“a remarkable device”). As we heard at ATTD 2012, the G4 AP uses the same sensor, transmitter, and receiver as the G4, but it includes new algorithms for improved accuracy and reliability. It will be made available to closed-loop researchers under an IDE or equivalent. Dr. Skyler showed a few examples of the accuracy improvement on days one to seven – in both instances, the day seven MARD was better than fingerstick monitoring. For more details, please see table below.




Overall MARD



Percentage of Sensors with a MARD <20%



Example 1: Accuracy on Day One and Day Seven

11.1% to 4.7%

32.8% to 7.1%

Example 2: Accuracy on Day One and Day Seven

7.2% to 4.1%

12.7% to 5.3%

  • Dexcom is working with the University of Padova to develop a predictive algorithm and increase the warning time prior to hypoglycemia (<55 mg/dl). Dr. Skyler compared using a CGM threshold of 70 mg/dl alone to the threshold plus a prediction algorithm. The threshold alone resulted in a median alert time for hypoglycemia (<55 mg/dl) of 15 minutes, which increased to 20 minutes with the predictive algorithm. Additionally, the number of alerts that gave less than 15 minutes of notice for hypoglycemia dropped from 38% to 16% with the new algorithm. Encouragingly, the prediction algorithm only added an average of one nuisance alarm per week and 46% of sensors had no additional nuisance alerts. We're very glad to see Dexcom investing in this area given its potential to truly enhance patient quality of life and safety – as a reminder, Medtronic has had predictive alerts for some time in its Revel and Veo insulin pumps and is developing a predictive LGS system, the MiniMed 640G (see our Medtronic analyst day report at
    • The future of Dexcom technology will combine remote monitoring with advanced CGM – an “alternative to the need for LGS.” According to Dr. Skyler, using the most accurate and advanced CGM technology alleviates the need to suspend insulin delivery since the patient will be alerted in time to treat or prevent hypoglycemia. Indeed, he noted that shutting off the pump may lead to a rebound high, followed by overcompensation that drives glucose too low, followed by a rebound high, etc. The “best way to do it,” he believes, is to anticipate the hypoglycemia and intervene by taking appropriate carbs. Moreover, Dr. Skyler made the valuable point that LGS only works if you have a pump, while pre-warning with CGM would work in patients on MDI. Since ~90% of insulin requiring patients in Europe don’t use pump (~70% in the US), this is a huge fraction of the market.



Dorothee Deiss, MD (Endokrinologikum am Gendarmenmarkt, Berlin, Germany); Thomas Peyser, PhD (Dexcom, San Diego, CA); Martin Prázný, PhD (Charles University, Prague, Czech Republic); Bruce Buckingham, MD (Stanford University, Stanford, CA); and Jay Skyler, MD (University of Miami Miller School of Medicine, Miami, FL)

Dr. Boris Kovatchev (University of Virginia, Charlottesville, VA): Dr. Skyler, the outpatient study that you mentioned is now complete.

Dr. Skyler: Do you have the data? Are you presenting it here? Dr. Kovatchev: No.

Dr. Skyler: ATTD 2013? Dr. Kovatchev: Yes.

Dr. Gary Steil (Children’s Hospital Boston, Boston, MA): Dr. Skyler, you showed some Howard Zisser data on not using meal boluses. I agree that meal boluses can be problematic. You said you think it’s better to give carbohydrate than shut the pump off. I’m a bit concerned about weight gain from excessive use of corrective carbohydrate.

Dr. Skyler: Yes, the best thing is to avoid hypoglycemia altogether, and I think that can be done with some of the CGM algorithms in development.

Dr. Steil: I have more confidence than you in turning the pump off in a timely manner, as long as it’s turned back on in a timely manner. This is why I have more faith in a closed- loop system than simple LGS. It would seem that if you prevent hypoglycemia with pump suspension, and turn it back on, you could prevent hyperglycemic rebounds.

Dr. Skyler: I think you’ve got to do all those things. Bruce, you have looked at suspension.

Dr. Buckingham: I think even with the Veo, the blood glucose goes up to about 150 mg/dl after several hours. With our system the pump turns back on immediately past the hypoglycemic nadir. Even with that, we get a bit of rebound to the 120-140 mg/dl range, though I wouldn’t call this hyperglycemia. It depends on the duration of suspension.

Dr. Skyler: How long did you typically suspend for?

Dr. Buckingham: Usually about 30-40 minutes. There are limits: the system can’t turn off for more than three hours at night, for example. If someone takes a bedtime snack and over-boluses, there is so much insulin on board that suspension can’t prevent hypoglycemia. Once you are past bedtime insulin, then you can be very effective because you don’t have much insulin on board.

Dr. Skyler: My bias is I worry about shutting off the pump. It could take three hours before you re-initiate basal to get the blood glucose back up. That can be ages.

Dr. Buckingham: We find that upon restarting, you get back to pretty effective insulin after 60-90 minutes.

Dr. Skyler: But that’s for shutting off for a 30-40 minutes at a time. But a two hour suspend is a pretty long time.

Q: I’m from BD. I wanted to ask you about calibration on day one being the hardest. You said when it comes to the first calibration, you have to get it right. Does it become easier over the sensor wear because of some rating scheme or other processes?

Dr. Peyser: Calibration on day one is done anew. Each sensor is calibrated based on blood glucose meter readings that are input into the receiver. On day one, you begin with two initial calibrations, followed 12 hours later with another calibration. On days three or four, you have a cumulative history of readings. The sensor system is more susceptible to error from SMBG on day one than on subsequent days. We think this is a large source of error in clinical studies.

Q: I think that with current economic conditions, it’s very important to have studies that show the same benefit for patients with MDI or pump use. Maybe we can spend the money on CGM instead of pumps, because we don’t have them for both.

Dr. Skyler: I think you raise an important point: that CGM is inexpensive compared to pumps, and that having the glucose information is probably more important than having continuous insulin delivery. I agree with you that we need to do more studies to make sure that everyone gets the message that you can improve glycemic control by adding CGM in the absence of pumps.

Q: Some of my patients who have lost reimbursement continue to pay out of pocket for CGM: they say, “I can live without my pump but not without my CGM.”

Dr. Skyler: I think it’s a pity to have to choose, but I agree that if you have to choose, you are going to choose the CGM.

Dr. Nicholas Argento (Maryland Endocrine and Diabetes Center, Columbia, MD): Could the Gen 5 transmitter send signals to an alarm clock? For adults that are hard of hearing, those with sleep apnea, or just hard sleepers, hearing alarms is a major challenge.

Dr. Peyser: We are actively looking into that. If you look at sleep arousal thresholds, adults are around 65- 70 decibels. Children are 70-80 decibels. Many adolescents are 105 decibels. That’s basically a jet engine. We’re looking at louder alarms – certainly not a jet engine alarm – but something like that. It’s an active area of research and development at Dexcom.

Comment: Maybe for adolescents you need electric shock therapy. [Laughter]

Dr. Deiss: Patients are enthusiastic about mealtime data. Do your patients also use retrospective analysis? Do you think this new Studio software will help patients?

Dr. Prázný: If we educate patients well, they will be able to use this and profit from it. They can do retrospective analysis on their own, but they may need our help in that from the beginning.

Dr. Deiss: In your practical experience, how many patients are really downloading their data and looking at it at least once per week? In my experience, a very small number of patients are doing this.

Dr. Skyler: It’s more important for patients to be looking at the receiver all the time and letting it guide their life instead of looking at it at the end of the week. So I don’t get on them when they don’t download. I’m more concerned that they look at it on a frequent basis and take ongoing actions rather than after the fact.

Dr. Steil: Following Dr. Peyser’s answer about calibration – how far is Dexcom from a sensor that does not need calibration, or at least not daily calibration?

Dr. Peyser: I am hesitant to say an exact date. I will tell you that we have an active program and have done clinical studies with the Gen 6 sensor, which appears to be good enough to go with an initial calibration and last seven, 10, even 14 days. The issue is to develop a manufacturing process capable of producing sensors of that quality at the 5-10 million per year number. It is at least a few years out, but I am convinced that it is doable.


Corporate Symposium: The Challenge to Optimize Insulin Therapy: How New Diagnostic Concepts and Technology Can Support People with Diabetes and Their Healthcare Professionals (Sponsored by Roche Diagnostics GmbH)


Matthias Schweitzer, MD, PhD (Head of Medical and Scientific Affairs, Roche Diabetes Care, Mannheim, Germany)

Dr. Schweitzer emphasized that “glucose information is the key driver for any kind of diabetes management.” He pointed to the often-cited STeP and PRISMA studies to show the relationship between structured SMBG adherence and diabetes control, and he noted that CGM delivers even more and different quality information on glycemia. Calling for wider access to these technologies, he argued that “there is no justification to withhold patients with diabetes from access to glucose information.” Interestingly, in his conclusion he stressed that Roche Diabetes Care differentiates itself from other companies – especially from “low-cost (product-only) suppliers” – through constant investment in the development of medical concepts, clinical research, and new technologies. This seemed like a veiled reference to Walmart’s recent introduction of low-cost ReliOn Prime blood glucose strips (~$9 for a 50- count box of strips), which some industry players have pointed out will threaten R&D investment in the entire BGM field (R&D investment is generally determined as a percentage of profit, which stands to decrease with the additional pricing pressure from Walmart’s low-cost strips.)



Ralph Ziegler, MD (Dr. med. Ralph Ziegler und Kollegen, Muenster, Germany) and David Cavan, MD (Royal Bournemouth Hospital, Dorset, UK)

Dr. Ralph Ziegler and Dr. David Cavan provided additional detail and color on the first results of the automated bolus advisor control and usability study (ABACUS). As seen at Roche’s Media event in the morning, t he key finding of the study was that compared to the active control group (standard MDI), a greater percentage of patients in the intervention group (MDI with trained use of blood glucose meter containing a built-in bolus calculator) reached the target of more than a 0.5% A1c decrease from baseline. Both interventions showed a reduction in the number of subjects who reduced insulin to avoid hypoglycemia, although those who did so in the bolus advisor group had a greater A1c reduction than controls (0.8% vs. 0.4%). For our full discussion, please coverage of the Roche Diabetes Care Scientific Symposium below.

Questions and Answers:

Q: Do you have information on differences between those with type 1 diabetes and type 2 diabetes and are there differences between age groups related to different types?

Dr. Cavan: The short answer is no. There were relatively few patients with type 2 diabetes in study, but we haven’t analyzed differences between the two groups.

Q: Does the expert calculate insulin-on-board?

Dr. Cavan: Yes

Dr. Richard Bergenstal (International Diabetes Center, Minneapolis, MN): Is there any data on the percent who got to A1c less than 7.0%?

Dr. Cavan: Not yet, but there is more analysis to come.



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

Dr. Eric Renard gave a high-level, thoughtful overview of insulin pump therapy and a look toward the artificial pancreas – which he suggested should use intraperitoneal insulin delivery (he showed promising closed-loop data with the transcutaneous Accu-Chek DiaPort) and algorithms running on a smart phone. Drawing encouragement from the outpatient, ambulatory closed-loop studies that began in 2011 (including one run by his team in Montpellier), he said that the “closed loop is no more a dream…it can happen”).

  • Dr. Renard endorsed several design approaches for commercial insulin pumps. First, he spoke to the relative benefit of bolus calculators that target glucose levels at the midpoint of a patients’ target range as opposed to the margin of the target range. Second, he highlighted the value of pumps that allow users to change the profile of insulin boluses (i.e., single bolus, double bolus, square-wave bolus, or dual-wave bolus) according to the type of meal (i.e., high in fat vs. mixed carbohydrate and fat). Third, he suggested that infusion sets could be improved. While pump therapy has been developed and well researched for the past 30 years, he said that there   has been little research on the infusion set.
  • To demonstrate the benefits of intra peritoneal insulin delivery, Dr. Renard showed a subject’s glucose profile from the JDRF DiaPort Closed Loop Trial comparing intra peritoneal vs. subcutaneous insulin delivery using a model predictive control algorithm. Intraperitoneal delivery showed less glucose variability and greater time in zone (Eric Renard, Howard Zisser, et al., personal data).
  • To demonstrate the feasibility of translating CGM information into patient advice, Dr. Renard pointed to the DIAdvisor system, which is designed to predict forthcoming glucose profiles and provide therapy advice (EASD poster 1029: “Clinical assessment of DIAdvisor Device Shows High Accuracy in Glucose Prediction at 20-min Horizon and a Coherence of most advices on therapy in patients with type 1 diabetes”).

Questions and Answers

Dr. Richard Bergenstal (International Diabetes Center, Minneapolis, MN): Does data on intraperitoneal insulin tell us we need more rapid acting insulin? Does it work faster?

A: The difference is the efficiency when you infuse it towards the liver. With the intra peritoneal route, you modulate better glucose prediction from the liver.

Dr. Bergenstal: How many basal rates does a typical patient need to have effective control? Of course there is no such thing as a typical patient…but should we be worried if a patient has only one or two levels or should we be worried if they have 16?

A: I would be worried if there are more than three or four basal rates in a day. It would be a dream to think when we change the basal we will have immediate actions – it takes a   couple hours.



Richard Bergenstal, MD (International Diabetes Center, Minneapolis, MN)

The esteemed Dr. Richard Bergenstal concluded the symposium with his perspective on CGM. He opened with T1D Exchange data showing that despite high frequencies of severe hypoglycemia, CGM use has remained remarkably low. Encouragingly, Dr. Bergenstal believes that the recent 2012 ADA Clinical Practice Recommendations’ mention of CGM would advance the discussion around CGM use and CGM reimbursement. He further emphasized that facilitating discussion between clinicians and patients   about CGM would improve CGM use and patient adherence. Notably, Dr. Bergenstal called for a redefinition of “good” glucose control. He said that good control should be based on more than just A1c and consider time in range, hypoglycemia, and glucose variability – we agree! Dr. Bergenstal    concluded his presentation by presenting (very) preliminary findings from the REACT 3 study of CGM  vs. structured SMBG in type 2 diabetes. Both tools could improve glucose control, but CGM may be more effective in minimizing hypoglycemia at the same level of A1c reduction. Given the lack of consensus on the role of CGM in patients with type 2 diabetes, we are eager to follow up when the full REACT 3 analysis is completed and hope that it inspires more prospective studies on glucose monitoring interventions for behavior change.

  • Echoing previous presenters, Dr. Bergenstal emphasized the importance of using glucose information to inform therapeutic decisions. While he said that this point should be obvious, for many clinicians the importance of obtaining and using glucose information is not as engrained in practice as it should be. We appreciate that Dr. Bergenstal stressed the need to spread this message – it seems familiar to us, which reminded us how lucky we are to consistently have access to learning opportunities like EASD.
  • Dr. Bergenstal highlighted the mention of CGM in the 2012 ADA Clinical Practice Recommendations, which says that CGM use should be considered with intensive insulin therapy in adults with type 1 diabetes (A level evidence), some children (C level evidence), and in patients with hypoglycemia unawareness (E level evidence; Dr. Bergenstal noted that the low evidence grade reflects a lack of randomized control studies, but he said that in practice CGM use is well-established for this subpopulation). He also expressed his hopes that the evaluation in the ADA Clinical Practice guidelines would help get payers and providers thinking more seriously about reimbursement and prescription, respectively.
  • To improve clinical practice and patient adherence with regards to CGM use, Dr. Bergenstal emphasized the importance of: 1) explaining the connection between CGM and A1c; 2) setting clear goals; 3) giving a consistent message about CGM use (i.e., real-time vs. retrospective use); 4) teaching the difference between individual readings and patterns; and 5) teaching how to respond to the data. Interestingly, Dr. Bergenstal said that the question on when to use CGM (real-time vs. retrospective) was being hotly debated and that more studies were needed for a definitive answer.
  • Dr. Bergenstal asked the audience for help in expanding the definition of “good” glucose control. To demonstrate, he displayed the glucose profiles of two patients who, despite having identical A1cs had vastly different glucose variability. Glucose control needs to be defined by more than just A1c, he said, and proposed that the definition should also consider: 1) time in target range; 2) hypoglycemia; and 3) glucose variability.
  • Preliminary finding from the REACT 3 study suggested that in patients with type 2 diabetes both SMBG (collected and analyzed in a structured way) and CGM can provide similar improvements in glucose control, but CGM may be more effective in minimizing hypoglycemia whilst improving control. The study randomized patients to receive either real-time CGM or SMBG for 16 weeks. Every two to four weeks, patients met with physicians to discuss blood glucose data and make necessary adjustments in medication. Blinded CGM data was collected from both groups at baseline, eight weeks, and 16 weeks to assess the primary endpoint of time in range.
    • Both groups showed substantial reductions in A1c. The SMBG group’s mean A1c decreased from 7.8% to 7.0%, and the CGM group’s mean A1c decreased from 8.1% to 7.1%.
    • Both groups showed similar patterns in improvement for area under the curve, time in range, and percent of time >180 mg/dl; however, CGM showed greater reductions in time spent hypoglycemic. The SMBG group in fact trended towards increased percentage of time spent <70 mg/dl, <60 mg/dl, and <50 mg/dl over the 16 weeks, though the percentages themselves were very low (i.e., for <70 mg/dl, SMBG at 16 weeks was less than 2.5% vs. less than 1.0% for CGM; for <50 mg/dl, SMBG at 16 weeks was ~0.6% vs. ~0.2 % for CGM).

Questions and Answers

Q: In REACT 3, was the CGM used on a real-time basis for the patients in that arm or was it just used by the health care professional?

A: Patients had it real-time, but on monthly to two-week intervals, information was printed out and discussed with providers.

Q: Was there a difference in nocturnal hypoglycemia for SMBG patients, since they wouldn’t be getting 7-point profiles then?

A: I’m not sure yet. This is brand new data. I know a few instances of that, but can’t tell you statistically yet.


Scientific Media Symposium: Personalized Diabetes Management – Cutting-Edge Therapy Approach & Technological Innovation for Enhanced Patient Benefit (Sponsored by Roche Diabetes Care)


Luc Vierstraete (President Roche Diabetes Care, Roche Diagnostics GmbH, Mannheim, Germany)

Previously a bank in 1901, the Humboldt Carre served as the intimate setting for the Roche Diabetes Care Scientific Media Symposium. The room was set with eleven tables donning sparkling water, juices, and delicious German chocolate. Mr. Luc Vierstraete took the podium briefly to welcome the ~30 media representatives from across the globe to the early morning event that would feature a series of presentations on personalized diabetes management.



Antonio Ceriello, MD (University of Udine, Udine, Italy)

Dr. Ceriello reviewed (and encouraged) the ongoing trend toward personalization of type 2 diabetes treatment guidelines; however, he argued that clinicians could use more advice on establishing treatment goals and encouraging patient cooperation, relative to the open-ended ADA/EASD 2012 position statement. As a template for personalized diabetes management, he presented a six-step cycle based heavily on structured self-monitoring of blood glucose (SMBG) and collaborative review of the results: structured education, structured SMBG, documentation, analysis, personalized treatment, and treatment efficacy assessment – which feeds a new round of education to start the cycle anew (Ceriello et al., Diabetes Res Clin Pract 2012). We appreciated Dr. Ceriello’s frank acknowledgement of the difficulty of preventive medicine (“I am a diabetologist, and I am clearly overweight”), and we enthusiastically support the practice of ongoing, empirically driven diabetes management.

  • Dr. Ceriello presented a Personalized Diabetes Management Cycle based on what he called a “very naïve but important idea” – that a successful treatment regimen requires the patient’s cooperation and that measurement is needed in order to gain useful feedback for guiding that regimen. The cycle was heavily based around collection and analysis of SMBG data – the accompanying graphic included testing and/or a computer (to display results) in nearly every panel.



Ralph Ziegler, MD (Dr. med. Ralph Ziegler und Kollegen, Muenster, Germany)

Dr. Ralph Ziegler presented first results from the Automated Bolus Advisor Control and Usability Study (ABACUS). The study randomized patients with poorly controlled type 1 or type 2 diabetes on MDI to receive either standard MDI therapy or bolus advisor (BA) supported MDI therapy, which came in the form of the Accu-Check Aviva Expert, a BGM with an integrated bolus advisor. At six months, a greater percentage of patients in the BA achieved the A1c reduction target (>0.5% change from baseline) compared to the standard MDI group (p <0.01). However, the benefits were statistically significant only among patients with perfect baseline competency scores in MDI and carbohydrate counting, underscoring the need for interventions that work in less-competent patients. Another education-related finding was that 38.1% of patients (n=83) at baseline reported reducing insulin doses independently of insulin level due to fear of hypoglycemia: fortunately this fear declined (in both groups) without a rise   in hypoglycemia. We think that built-in bolus calculators certainly stand to add value by encouraging more appropriate bolus dosing. However, they seem likely to face serious FDA scrutiny – particularly   on the meters’ insulin-on-board calculations, considering that these depend on the user to accurately report his or her insulin doses.

  • The multi-national, prospective Automated Bolus Advisor Control and Usability Study (ABACUS) randomized 218 patients with poorly controlled type 1 or type 2 diabetes (A1c >7.5%) on MDI to receive MDI standard therapy (control) or bolus advisor (BA) supported MDI therapy (intervention). The latter arm used the Accu-Chek Aviva Expert, a BGM with an integrated bolus advisor (BA) BA that determines prandial and correction insulin doses. The primary endpoint assessed was six-month change in A1c from baseline. Secondary endpoints included additional measures of glycemic control, patient use of BA, and psychosocial measures including fear of hypoglycemia.
  • Baseline assessments showed that A1c was generally lower among patients with higher competency scores in MDI and carbohydrate counting. Patients with perfect scores in baseline assessments of both categories (n=66) had a significantly lower mean A1c than patients with no perfect score in either MDI or carbohydrate counting (n=79; 8.6% vs. 9.1%; p <0.01). Dr. Ziegler explained that this finding underscores the importance of structured education in diabetes management. Following competency assessments, participants received remedial education in areas where deficiencies were identified.
  • Overall, 38.1% of patients (n=83) at baseline reported reducing insulin doses independently of insulin level due to fear of hypoglycemia. Survey results showed: 1) 45.9% of participants (n=100) were worried about hypoglycemia; 2) 65.1% (n=142) engaged in hypoglycemia avoidance behavior (i.e., reducing insulin dose or eating additional carbohydrates independent of glucose level); 3) 37.6% (n=82) were worried and engaged in avoidance behavior; and 4) only 26.6% (n=58) showed no worry or avoidance behavior.
  • A greater percentage of the intervention group achieved over a 0.5% A1c reduction from baseline compared to the control group (56% [n=56] vs. 34.4% [n=32]; p <0.01); the average A1c reduction in the intervention group was 1.2%. However, when considered by baseline level of MDI/carbohydrate counting competency, the between-group difference (control vs. intervention) was significant only in the group with perfect competency scores. (First results were assessed only for completers.)
  • Both groups showed a significant decrease in the number of patients who reduced their insulin dose out of fear of hypoglycemia (p <0.05). Within the control group, the percentage of patients employing this avoidance behavior decreased from 37.5% (n=42) to 25.6% (n=23). Within the intervention group, the percentage of patients decreased from 39.8% (n=41)    to 27.4% (n=26). While the between group difference was not significant, Dr. Ziegler emphasized that the intervention group avoidance behavior change was accompanied by a greater reduction in A1c. And importantly, the use of the bolus advisor was not associated with increased frequency of severe hypoglycemia over the six months compared to controls.
  • We sense the FDA to be quite wary of meters with bolus calculators, not least because bolus calculators depend on patients to have comprehensively logged their insulin injections. According to, a UK ease-of-use study comparing Abbott’s FreeStyle InsuLinx with built-in bolus calculator to other glucose meters was recently completed (updated September 10, 2012; Identifier: NCT01432275) and we look forward to hearing results, which could potentially substantiate the value of the built-in-bolus calculator and help with FDA approval (Editors note: the study was originally slated to complete in January 2012). As a reminder, the FreeStyle InsuLinx is approved ex-US with a built-in bolus calculator, but the approval timeline in the US for this feature remains unclear (the FreeStyle InsuLinx meter   without the calculator was FDA cleared in March 2011). For additional discussion on the   FreeStyle InsuLinx, please see our Abbott 2Q12 report at Similarly, we hope to learn more about any wider launch plans for the Accu-Chek Aviva Expert at the exhibit hall.



Sanjoy Dutta, PhD (JDRF, New York, NY)

Dr. Dutta summarized JDRF’s groundbreaking efforts to encourage development of an artificial pancreas and outlined some of the main ongoing challenges. He said that the first generation of semi- automated-insulin-delivery products can be developed with today’s technologies but that the second generation will depend on better sensors and faster insulin action. He briefly mentioned longer-term efforts toward the third-generation artificial pancreas features (e.g., multi-hormone delivery, CGM sensors that are either implantable or on the same port as insulin delivery), and he noted that JDRF is also working on “out-of-the-box” approaches such as a self-regulating insulin that could recognize ambient glucose concentration and accordingly dose itself at a molecular level in real time. The overarching goal of JDRF’s Treat Therapies division, Dr. Dutta said, is to take a “holistic approach at achieving glucose control and overall metabolic balance” while keeping in mind the “changing landscape of type 1 diabetes and its etiology” (e.g., greater prevalence of overweight/obesity and concomitant insulin resistance).

  • Dr. Dutta said that the second generation of artificial pancreas products will require better glucose sensors and faster insulin action. To the former, JDRF and the Helmsley Charitable Trust to fund sensors with tighter accuracy and better form factor; so far they have partnered with Medtronic and BD. As for insulin delivery, Dr. Dutta highlighted Roche’s Accu- Chek DiaPort (a port that allows insulin pumps to reach the intraperitoneal cavity and thus act more physiologically than subcutaneous insulin), BD’s intradermal microneedles (which give “as low a pain level as you can imagine with invasive insulin delivery” but which are in the “very   early” stages of clinical testing [Pettis et al., Diabetes Technol Ther 2011]), and Insuline’s InsuPatch (a heating device to improve vascular flow of subcutaneous insulin that has been  studied at Yale University).



David O’Neal, MD (University of Melbourne, Melbourne, Australia)

Dr. O’Neal argued that in order to improve diabetes outcomes, primary care physicians must be empowered to take a larger role in insulin initiation. To this end he is helping to conduct the Stepping  Up study, a cluster-randomized trial of 58 primary care practices (290 patients with type 2 diabetes). The researchers will study whether A1c can be lowered more effectively than with standard care using insulin initiation and titration based on the Stepping Up protocol (which, in turn, is based on three days of seven-point SMBG profiles using the Accu-Chek 360-degree View paper-based tool). Dr. O’Neal noted that structured SMBG provides much more detail than “traditional” fasting morning glucose tests at a much lower cost than continuous glucose monitoring (CGM), making it viable for near-term reimbursement in Australia. He closed with an analogy of CGM vs. structured SMBG: “Instead of a Ferrari, we’ve got a Volkswagen; I think that’s what we need right now, and that’s why we’ve chosen it for the Stepping Up study.”



Michael Schoemaker, PhD (Roche Diagnostics GmbH, Mannheim, Germany)

Dr. Michael Schoemaker posited that the major limiting factor preventing widespread CGM use is lack of CGM accuracy and reliability, and he highlighted Roche Diabetes Care for their investigation into the sensor-to-tissue interface in order to improve both of these limiting factors. Details were scarce, but Dr. Schoemaker said that Roche’s sensor, in early-stage development, shows “outstanding” accuracy, precision, and reliability, especially in the hypoglycemia range. This update follows on the heels of the company’s 2Q12 announcement that it was restructuring Roche Diabetes Care division – a project that includes increasing R&D investment in insulin pump and CGM technologies. Undoubtedly, the shift in R&D focus is timely for Roche considering the persistent pricing pressures in the BGM market and growing recognition of the benefits of glycemic trend data. (For more discussion of Roche’s   restructuring initiatives, please see our Roche 2Q12 report at and Roche Investor Day report at Dr. Schoemaker concluded that CGM can become the standard of care if certain improvements are made, in particular: 1) better accuracy and precision; 2) translation of complex CGM information into medically relevant and actionable information; and 3) improved user- friendliness.


4. Novel Therapies and Basic Science

Oral Presentations: SGLT-2 Inhibitors


Agata Ptaszynska, MD (Bristol-Meyers Squibb, Princeton, New Jersey)

Dr. Agata Ptaszynska presented pooled data from 12 phase 2b/3 placebo-controlled, double-blind trials on dapagliflozin (DAPA) in patients with type 2 diabetes. Trials included in the meta-analysis examined dapagliflozin monotherapy, dapagliflozin use in initial combination with metformin, and use of dapagliflozin as an add-on to insulin, metformin, TZD, or sulfonylureas. In these trials, participants  were randomized to receive either placebo, dapagliflozin 5 mg, or dapagliflozin 10 mg therapy. Twenty- four-week data was derived from all 12 trials, while 102 week data was derived from six of the twelve. DAPA initiation resulted in a rapid decrease in estimated glomerular filtration rate (eGFR) in the first week that subsequently returned to baseline levels. Dr. Ptaszynska posited that the initial eGFR decrease likely represents a physiological adaptation, and was not associated with any adverse events. Renal adverse events were mostly similar between dapagliflozin and placebo-treated groups: at 24 weeks, 0.9% of patients in the placebo group (n=1393) reported an adverse event related to renal function, compared to 1.3% in the 5 mg DAPA group (n=1145) and 0.8% in the 10 mg DAPA group. At 105 weeks, 1.7%, 1.8%, and 1.9% of placebo, DAPA 5 mg, and DAPA 10 mg, respectively, had experienced such an event. Dr. Ptaszynska said it was, of course, to be expected that with more time there were a greater number of events. As assessed by electrolyte levels, the study did not find any major effect on other tubular functions; additionally, DAPA did not adversely affect albuminuria levels. However, a modest blood pressure decrease was observed in the DAPA group and while volume depletion events were uncommon, some non-serious adverse events of hypotension were reported with DAPA use. For additional discussion and detail on this study, please see our page 163 of our ADA 2012 coverage at

Questions and Answers

Q: Have you looked at pH? Was there any effect on bicarbonate? One of your tables showed that bicarbonate levels were low in two patients.

A: We did not analyze the pH, but patients are not losing bicarbonate in the urine.

Comment: You concluded long-term safety based on just two years and I really take issue with that. We need five and ten year data to make conclusions. You also have not showed us any tubular factors whatsoever.

A: First, you are right. Two years is short, but we have to start with something. We have plans for longer trials. We will have patients in this program that will be observed for several years. In terms of tubular function, we don’t see any affect on loss of either proteins or electrolytes in urine. So in the clinical program we don’t see any tubular dysfunction.

Comment: It is very easy to do small, short-term studies in humans looking in great detail at tubular function.

A: The short answer is the programs are finished so we are open to proposals.



Sunder Mudaliar, MD (University of California San Diego, San Diego, CA)

Dr. Sunder Mudaliar presented study results investigating the effect of 300 mg canagliflozin (CANA) on oral glucose absorption. In a two-period crossover study, 20 healthy men were assigned to receive  either a single 300 mg oral dose of CANA or placebo before a mixed meal tolerance test (MMTT) with  14C glucose in 75 g glucose solution, solid food, and 960 mg acetaminophen. Participants also received 3H-glucose IV infusion starting three hours before MMTT initiation and lasting through the six-hour post-MMTT observation period. The primary endpoint assessed was the rate of appearance in plasma  of orally ingested glucose. In the first two hours following MMTT, CANA reduced plasma glucose (~35% by area under the curve [AUC]) and plasma insulin (~43% by AUC), and reduced plasma levels of oral glucose tracer (14C); however, CANA had little effect on plasma levels of IV glucose tracer (3H) during this time period. Oral glucose absorption was delayed with CANA, which Dr. Mudaliar said was likely due to transient intestinal SGLT-1 inhibition; however, there was no meaningful effect on total glucose absorption over the six hours. Importantly, the reduction of postprandial plasma glucose could not be explained by increased glucose disposal, despite the increased urinary glucose excretion – the rate of disposal was actually lower over the first two hours post-MMTT with CANA. Additionally, over the first two hours post-MMT, plasma GIP was ~50% lower, plasma PYY was 60% higher, and plasma GLP-1 was 35% higher. Dr. Mudaliar noted that the reduced GIP was consistent with intestinal SGLT-1 inhibition, and suggested that the increased PYY and GLP-1 may be due to more glucose reaching the distal small intestine. As for general safety, CANA was well tolerated by participants with no gastrointestinal adverse events observed. For additional discussion on this study, please see our page  145 of our ADA 2012 coverage at

Questions and Answers

Q: As for GIP, PYY, and GLP-1 you might expect to see levels increase after the two hour period. Did you follow levels past two hours?

A: We only measured the first two hours.

Comment: I think it would be interesting to look at longer period.

Q: Was GLP-1 elevation in first two hours statistically significant compared to placebo?

A: Yes.

Q: And you didn’t measure beyond two hours?

A: No.



Apostolos Tsapas, MD, PhD (Aristotle University Thessaloniki, Thessaloniki, Greece)

This meta-analysis assessed the efficacy and safety of SGLT-2 inhibitors versus placebo or any other antidiabetic medication (as monotherapy or as an add-on therapy). The investigators searched a  number of sources (e.g., electronic databases, abstracts from EASD, ADA, IDF, and AACE meetings, and pharmaceutical companies’ websites) to find randomized controlled trials of 12 or more weeks in duration comparing any SLGT-2 inhibitor versus comparator for the treatment of type 2 diabetes to include in the meta-analysis. In total, 45 records on 39 studies were included. The mean difference in   A1c between SGLT-2 inhibitors and placebo was -0.73% (95% CI: -0.82%, -0.63%) and -0.12% (95% CI: - 0.23%, -0.01%) between SGLT-2 inhibitors and active comparators. The mean differences in systolic and diastolic blood pressure were approximately -4 mmHg and -2 mmHg versus both placebo and active comparator. The risk for hypoglycemia was comparable between SGLT-2 inhibitors and placebo. However, there was a clear and significant increase in the frequency of urinary tract infections and genital tract infection versus placebo or active comparator. In summary, Dr. Tsapas highlighted that SGLT-2 inhibitors have intermediate/high efficacy, low risk of hypoglycemia, and cause weight loss,  with the potential for urinary and genital tract infections.

Questions and Answers

Q: Why haven’t you looked at lipids in your meta-analysis?

A: Unfortunately, lipids were not in the outcomes we predefined in our protocol.

Q: How many urinary tract infections and genital tract infections would this translate to for each patient treated per year? What advice would we have to give our patients on how frequently they should expect to experience these events?

A: The risk ratio was about five.

Q: But, per patient per year?

A: I won’t be able to tell you right now.

Q: Did you also look at cancer incidence?

A: The incidence of cancer was one of the domains we extracted, but unfortunately, only a small number of trials provided data for cancer. I’m afraid it was just a specific agent [that provided data]; the majority of other agents did not provide data on the incidence of cancer.



Guntram Schernthaner, MD (University of Vienna, Vienna, Austria)

Dr. Guntram Schernthaner presented the results of a 52-week phase 3 study evaluating the efficacy and safety of canagliflozin 300 mg versus sitagliptin 100 mg in patients with type 2 diabetes on metformin plus sulfonylurea (first presented at ADA 2012 as a late-breaking poster). At the 52-week mark, canagliflozin brought about a 0.37% reduction in A1c beyond sitagliptin (1.03% vs. 0.66%), as well as significant improvements in fasting plasma glucose, weight, and systolic blood pressure. For more details on the study, please see our ADA 2012 report at

Questions and Answers

Q: Do you have data on parameters of beta cell function and beta cell stress, like proinsulin/insulin ratio, or HOMA-B?

A: Not in this study.

Q: You presented a comprehensive dataset this week, but there was no mention of malignancies or imbalances. If you have any imbalances, how many cancer events did you have?

A: Sorry, I can’t tell you that at this moment – the drug is in registration with the FDA.

Q: Would you expect other drugs in the SGLT-2 class to have the same advantage in A1c reduction versus sitagliptin? Or, is there something specific with regards to the efficacy of canagliflozin being better than sitagliptin?

A: From a scientific point of view, we can’t make comparisons if we don’t have head-to-head studies. So, I won’t make a comment [on the efficacy of other SGLT-2 inhibitors versus sitagliptin].

Q: Would you speculate on the mechanism of the blood pressure drop, in combination with body weight? Is it a result of negative sodium balance?

A: I believe the main mechanism is the diuretic effect. I do not know if there is an additional vasodilatory effect.

Q: Was there the same incidence of infections over time, or were there more at the beginning?

A: There were more at the beginning, and then they came down.

Q: What is the mechanism that led to an increase in LDL in the canagliflozin group? Were all these patients already on statins?

A: More patients in the sitagliptin group were on statins compared to the canagliflozin group. Also, at the end of the trial, more patients on sitagliptin were on statins compared to patients on canagliflozin. The mechanism of the increase in LDL is largely unknown. I personally believe it is a compensatory mechanism, and probably a drug class effect.


Oral Presentations: Novel Therapies


Richard DiMarchi, PhD (Indiana University, Bloomington, IN)

Dr. Richard DiMarchi discussed the potential of glucagon/GLP-1 and GIP/GLP-1 co-agonists to improve glycemic control, body weight, and other metabolic parameters, reviewing some of the animal studies conducted to date. He noted that glucagon/GLP-1 co-agonism could work synergistically to decrease fat mass, while GIP/GLP-1 co-agonism has the potential to have additive effects on insulin secretion. In preclinical studies, both have been documented to improve body weight, fat mass, blood glucose, insulin secretion, and blood lipid and liver fat content. In addition, GIP/GLP-1 could provide a potential basis for less nausea, emesis, and gastric stasis. Dr. DiMarchi noted that Merck is exploring glucagon/GLP-1 hybrids in clinical testing, while Roche is exploring GIP/GLP-1 hybrids in clinical testing – the latter, of course, is the compound that Roche purchased from Marcadia in 2011 for $300 million plus significant potential for additional milestone payments.

  • Dr. DiMarchi provided the rationale for developing a glucagon/GLP-1 co-agonist, explaining that action at the two receptors could work synergistically to decrease fat mass. Specifically, chronic glucagon action decreases fat mass by increasing energy expenditure via the glucagon receptor, while GLP-1 decreases fat mass by decreasing food intake via the GLP-1 receptor. In addition, a glucagon/GLP-1 co-agonist should minimize the diabetogenic risk of a  pure glucagon analog.
  • Glucagon/GLP-1 co-agonists have demonstrated promise for weight loss in preclinical studies. Initially, Dr. DiMarchi and his colleagues developed two chimeras – one with a low potency for the glucagon receptor and a high potency for the GLP-1 receptor (Chimera 1; 10.5% and 299.3% potency relative to the respective native molecules), and one with close-to- normal potency for both the glucagon and GLP-1 receptors (Chimera 2; 99.1% and 165%, respectively) (Day et al. Nature Chem Bio 2009). In DIO mice, treatment with a glucagon/GLP-1 co-agonist conferred significant reductions in body weight beyond GLP-1 alone, as well as improvements in blood glucose, lipids, cholesterol, and triglycerides. Food intake was comparable to GLP-1 treatment, while energy expenditure was increased with glucagon/GLP-1 treatment. In a subsequent study in a rodent model, it was shown that increasing the relative tone of glucagon versus GLP-1 (to an extent, without impairing glycemic control significantly) could provide   further weight loss than balanced glucagon/GLP-1 co-agonism (Day, DiMarchi, and Pocai, J Peptide Science 2012). These findings were confirmed in nonhuman primates (Dr. DiMarchi  noted that Merck will present the results at a future conference).
  • Dr. DiMarchi noted that he and his colleagues initially explored GIP/GLP-1 co- agonism because of their additive effects on insulin secretion (Nauck et al., JCE&M 1993). He and his colleagues developed a GIP/GLP-1 hybrid with a threefold increase in GLP-1 receptor activity, and a twofold increase in GIP receptor activity. Surprisingly, the GIP/GLP-1 hybrid (administered at 3, 10, or 30 nmol/kg/day) decreased body weight more than exendin-4 (administered at 10 or 30 nmol/kg/day) in a rodent model (Matthias Tschop & Research Group, University of Cincinnati). In nonhuman primates in which subjects were administered treatment eight hours prior to a graded dextrose infusion, a GIP/GLP-1 hybrid provided a greater increase in insulin secretion and decrease in glucose excursion than liraglutide.

Questions and Answers

Q: Is there a specific advantage to having the potential for action on two receptors in one molecule, or would it be equally good to combine two peptides at appropriate concentrations?

A: There are actually three different approaches. The physical combination of two separate drugs has its own complications, as you’d have to register two individual molecules. Heterodimers – stitching one nonnative peptide and another by a covalent bond, can be troublesome because of the potential to cross- hybridize. It’s an inherent challenge, but there is no data to say they would be any inferior. We have chosen to make one molecule like a master key.

Q: You showed impressive weight loss with both the glucagon/GLP-1 combination and the GLP-1/GIP combination. Do you have any comment on what you see in lean animals? When animals reach a normal weight, does weight loss continue?

A: Our emphasis was to decrease body weight by increasing thermogenesis with the glucagon/GLP-1 molecule, whereas our emphasis with GIP/GLP-1 was primarily on glycemic control. We still don’t understand why they lose the body weight that they do, but they do. It’s very consistent in rodents and in nonhuman primates. One wants to use these in the obese setting, where we’re only reducing fatness, and I think that’s possible.



Andre Scheen, MD, PhD (University of Liege, Liege, Belgium)

After reviewing cortisol’s role in hyperglycemia, Dr. Andre Scheen discussed the potential of 11B-HSD1 inhibition as a treatment for diabetes. He noted that while initial studies in rodent models appeared promising, clinical trials to date have shown lackluster efficacy. He noted that three 11B-HSD1 inhibitors previously in development – INCB13739, MK-0916, and MK-0736 – have all been discontinued from development after their phase 2 trials. Dr. Scheen pointed out that not a single abstract at EASD 2012 is devoted to 11B-HSD1 inhibition; he was nonetheless encouraged that there is still potential for these agents – research is ongoing to develop new, more potent 11B-HSD1 inhibitors (he stated that Lilly’s LY2523199 is in phase 2). Dr. Scheen believed that we have to wait for more selective and more potent 11B-HSD1 inhibitors to be developed before they will have clinical utility.

  • 11B-HSD1 is an enzyme found predominantly in the liver and adipose tissue that converts the less active cortisone into cortisol. Cortisol promotes gluconeogenesis in the liver, protein catabolism in the muscle, and lipolysis in adipose tissue, which ultimately can contribute to hyperglycemia. The 11B-HSD1 enzyme can be thought of as an intracellular amplifier of active glucocorticoid; as such, 11B-HSD1 inhibition could be a potential target to investigate as a treatment for diabetes.
  • In rodent models, 11B-HSD1 knockout/inhibition conferred improvements in metabolic parameters. 11B-HSD1 knockout mice were found to have reduced corticosterone (CORT) levels in fat cells, reductions of visceral fat (but expansion of peripheral fat), and protection against associated metabolic abnormalities, whereas 11B-HSD1 transgenic mice (able   to overexpress activity) had increased CORT levels in the liver and visceral fat, increased liver fat and marked expansion of visceral fat, less favorable adipokine/cytokine profiles, and metabolic abnormalities (including high blood pressure) (Seckl et al., Recent Prog Horm Res 2004). Similarly, chemical inhibition of the 11B-HSD1 enzyme was found to decrease body weight, reduce plasma glucose excursions, and decrease insulin secretion (that is, reduce insulin resistance) in a rodent model (Wang et al., PLoS ONE).
  • Dr. Scheen commented that 11B-HSD1 inhibitors have had mediocre performance in clinical trials to date, reviewing a number of studies for MK-0916, MK-0736, and INCB13739. In a 12-week trial (n=154), patients with type 2 diabetes and metabolic syndrome   on MK-0916 treatment experienced some positive effects on body weight, waist circumference, systolic blood pressure, diastolic blood pressure, and lipid levels; however, there was a slight increase in LDL and no improvement in HDL; MK-0916 had no effect on fasting plasma glucose   or two-hour postprandial glucose, and marginal benefit on A1c at the 6 mg dose (Feig et al., DOM 2011). In a 12-week trial in obese patients with hypertension (n=211), MK-0916 only had a mildly beneficial effect on systolic and diastolic blood pressure; these were accompanied by modest weight loss, and overall, no favorable effects on the lipid profile (shah et al., JASH 2011). INCB13739 was demonstrated to improve glycemic control in patients with type 2 diabetes inadequately controlled on metformin in a 12-week study (n=302) (Rosenstock et al., Diabetes Care 2010). Patients receiving INCB13739 experienced a 0.56% A1c reduction, and a 24 mg/dl improvement in fasting plasma glucose relative to placebo; HOMA-IR decreased 24%, and patients on study drug experienced an approximate 1 kg (2.2 lb) weight loss over the course of the trial. Dr. Scheen noted that it appears the clinical development of MK-0916, MK-0736, and INCB13739 have all been abandoned after their phase 2 trials; there are no trials listed on for any of the compounds. He pointed out that not a single abstract at EASD  2012 is devoted to 11B-HSD1 inhibition, but was nonetheless encouraged by the fact that there is still research ongoing to develop new, more potent 11B-HSD1 inhibitors (he stated that Lilly’s LY2523199 in phase 2).

Questions and Answers

Q: Since these compounds affect intra-abdominal fat, maybe it may just take longer than 12 weeks to show an effect?

A: I agree with you that we need longer studies to see what can happen.

Q: Do you think that targeting multiple tissues is one of the reasons for the lack of efficacy?

A: Type 2 diabetes involves many tissues in its pathophysiology, so I think it’s realistic to target several organs.



Christof Kazda, MD, PhD (Eli Lilly and Company, Suresnes, France)

Dr. Christof Kazda detailed a 12-week study of Lilly’s once-daily oral glucagon antagonist LY2409021  for the treatment of type 2 diabetes. Following a screening period and a one-week placebo lead-in  period, 87 participants were randomized to LY2409021 60 mg/day (n=26), 30 mg/day (n=34), 10 mg/day (n=17), or placebo (n=10) for 12 weeks, followed by a four-week washout period. Baseline characteristics were relatively balanced across the four treatment groups (mean age of 51-53 years, mean BMI of 32-34 kg/m2, and mean diabetes duration of four to five years), with the exception of a higher A1c level observed in the LY2409021 10 mg/day group (8.0% vs. 7.6-7.8% in the remaining groups) and a varying degree of metformin use among the groups (53-70%). At 12 weeks, all three doses of LY2409021 provided significantly greater improvements in A1c (a reduction of -0.66% for the 60 mg/day dose) compared to placebo (+0.11%); LY2409021 also produced dose-dependent increases in mean fasting glucagon and total GLP-1 levels (but not active GLP-1 levels), which returned to baseline during the four-week washout period. Furthermore, LY2409021 had no observed effect on body weight, blood pressure, and lipid profiles. While LY2409021 resulted in a dose-dependent and transient increase in mean aminotransferase levels, the investigators found no other signs or symptoms of liver injury,   and the incidence of treatment-related adverse events was similar across all four treatment groups. No participant experienced an episode of severe hypoglycemia and of the four confirmed hypoglycemia events, three were observed in the LY2409021 60 mg/day group and one was reported for the 10  mg/day group.

  • At 12 weeks, significant A1c improvements were observed with LY2409021 10 mg/day (-0.83%; p = 0.030), 30 mg/day (-0.65%; p = 0.042), and 60 mg/day (-0.66%; p = 0.051) compared to placebo (+0.11%); correcting for baseline A1c levels revealed that LY2409021’s effect on A1c was dose dependent. LY2409021 also lowered fasting plasma glucose (FPG) levels within one week of treatment initiation, with FPG levels remaining stable throughout the treatment period and increasing back to baseline during the washout period.

Questions and Answers

Q: Regarding the number of patients you included in the study and how they differed across treatment groups, how does that affect your conclusions about efficacy, dose dependency, and the lack of a dose dependent safety response?

A: We respected these different allocations and corrected for the differences in the calculations that I presented. This was all taken into account.

Q: Given some of the beneficial effects of glucagon, would you see any problem with a glucagon antagonist at this point? I’m thinking it could affect beta cell regeneration and alpha-beta cell interactions.

A: So beta cell regeneration has been shown, but only in animal models. If you’re referring to alpha cell hyperplasia, we have done extensive animal studies for up to 12 months and we have seen that there is a slight increase in the alpha cell mass in animals exposed to significant multiples of the clinical dose. This increase in alpha cell mass plateaus after one month of treatment and then stays stable in these high doses. There is no progression in the mass, but I agree that glucagon needs to be measured in the clinical program

Q: How do you read the transient nature of the glucagon concentration changes? Is there a feedback mechanism that you can’t know completely in this study?

A: We have had some analytical issues in the glucagon assay in this study. In other studies, the glucagon levels stay elevated for the course of the treatment and this is concentration dependent. As soon as we stopped treatment, the glucagon levels returned rapidly to baseline. And we have not seen any rebound hyperplasia when we stop treatment in these patients.

Q: Have you conducted any hyperinsulinemic clamp studies?

A: I can refer you to poster 818, which has just been presented in the past poster session. We did a hypoglycemic clamp study where we assessed the time to recovery from insulin-induced hypoglycemia, and we showed that clinically relevant levels of LY2409021 did not impact the time to recovery in type 2 diabetes patients.

Q: I noticed that you saw elevations of liver enzyme. I believe that’s also been observed with blocking antibodies to glucagon. What do you think the mechanism is in term of glucagon  antagonism?

A: That’s a very good question. With antibodies, its been reported that there are increases in the aminotransferase levels. It’s most likely a mechanism-related phenomenon and we are currently performing a study in which we use magnetic resonance spectroscopy to look for changes in glycogen and fat content in patients exposed to LY2409021 to further assess which is the mechanistic basis for this increase in ALT levels.

Q: Could you give us a mechanistic explanation for the increases in total GLP-1 levels and its clinical implication?

A: GLP-1 and glucagon are located on the same gene, so the alpha cells also are capable of producing GLP- 1 besides glucagon, and therefore we assume that most of the GLP-1 released here comes from the alpha cells.


Oral Presentations: Beta Cell Function in Vivo


Camille Vatier, PhD (INSERM, Paris, France)

This study aimed to evaluate the effects of metreleptin on insulin secretion in non-HIV-infected, lipodystrophic, leptin-deficient patients with metabolic complications. A total of 15 patients were  included in the study – four with generalized lipodystrophy (LD), and 11 with partial LD. The 13 patients with over one-year of follow-up data available decreased their BMI from 23.6 kg/m2 to 21.7 kg/m2, calorie intake from 2,000 kcal to 1,650 kcal, A1c from 8.8% to 7.6%, and triglycerides from 4.1 g/l to 2.9 g/l over the course of one year; the four patients with generalized LD had the best metabolic response. One month after therapy initiation, patients in the study experienced a significant increase in insulinogenic index (as assessed by a 75g two-hour OGTT). In conclusion, Dr. Vatier stated that these results suggest the improvement of insulin secretion contributes to the metabolic efficiency of leptin in patients with lipodystrophy.

Questions and Answers

Q: Did you have a way to check whether the efficacy was different based on degree of leptin resistance?

A: We only included patients with low amounts of leptin, so they were all leptin responsive.

Q: Was insulin secretion corrected for insulin sensitivity?

A: There were parallel improvements in both.

Q: The accumulation of lipids in beta cells is thought to maybe be the cause for deficient insulin secretion. It’s difficult to examine in your model, but some are using pancreatic fat as a proxy for islet fat accumulation. Are you planning to look at that?

A: No.


Oral Presentations: The -Omics Frontier: Applications of New Technologies


Ele Ferrannini, MD (University of Pisa, Pisa, Italy)

In an engaging lecture, Dr. Ele Ferrannini discussed early research on two potential biomarkers of insulin resistance. He and his colleagues analyzed two large databases of metabolic data on people without diabetes at baseline, the RISC cohort (n=1,261) and the Botnia cohort (n=2,580 with strong family history of diabetes) to find molecules that were correlated with insulin resistance (as measured by oral glucose tolerance test), and they investigated two in particular: alpha-hydroxybutyrate (aHB), which is positively associated with insulin resistance, and linoleoyl-GPC (LGPC), which is positively associated with insulin sensitivity. In both cohorts, aHB and LGPC predicted risk of dysglycemia even after controlling for age, gender, BMI, familial history of type 2 diabetes, and each other (i.e., aHB predicted risk independently of LGPC, and vice versa). The two molecules also make sense as risk markers from a biochemical perspective, since each is a metabolic byproducts of other known players in the metabolic syndrome: branched-chain amino acids and/or free fatty acids. (The predictive value of aHB and LGPC relative to free fatty acids and branched-chain amino acids was not clear from this analysis. Indeed, Dr. Ferrannini emphasized that analysis of aHB and LGPC is still too early-stage for either to be used diagnostically: “I am not ready to replace ‘diabetes’ with ‘alpha-hydroxybutyratitis.’”)

  • Dr. Ferrannini opened by reviewing the qualities of a useful biomarker: for example,  it is specifically and easily measurable in body fluids, it improves prediction algorithms, it tracks with the underlying pathophysiological mechanism, and it maps onto a disease pathway (possibly in ways that increase scientific understanding of that disease).
  • To see whether alpha-hydroxybutyrate and linoleoyl-GPC would track with improvement in insulin sensitivity, the researchers analyzed 26 morbidly obese patients before and one year after roux-en-Y gastric bypass. Mean BMI at baseline was 51 kg/m2, while one year after surgery mean BMI had fallen to 34 kg/m2. Insulin sensitivity doubled one year after surgery, and aHB concentration fell to half its baseline value – maintaining the inverse correlation seen in the RISC and Botnia cohorts. However, LGPC levels did not rise.  Dr. Ferrannini noted that both biomarkers have more complex roles in the body than he is aware; he said that both are doubtless involved in many processes beyond insulin resistance.
  • Dr. Ferrannini also presented in vitro analyses of beta-cell function in which aHB was associated with higher glucose-responsive insulin secretion, while LGPC was associated with lower glucose-responsive insulin secretion. During Q&A, he said that aHB is also associated with better in vivo beta cell function but that LGPC levels did not seem to correlate with beta-cell function.

Questions and Answers

Q: Branched-chain amino acids (BCAAs) have been shown to have protective effects on beta-cell viability. Might aHB be a marker of increased consumption of BCAAs and loss of this protective effect? Have you considered this possibility?

A: We have.

Q: You said these molecules might have a direct effect on insulin levels. Are they in any way related to beta cell function in vivo?

A: We find reciprocal associations in aHB and beta-cell function in vivo, but a relationship between LGPC and beta-cell function is not seen. Nor do we see LGPC changes after bariatric surgery. Biologically these molecules must be doing many different things of which I am aware of only a small portion.

Q: There are so many data showing that diabetes starts very early. Why do we not change the diagnosis of diabetes? Putting it at the onset of hyperglycemia means that half the beta cell is already destroyed. I propose that we all change the diagnosis of diabetes.

A: Yes, that is an interesting proposal. I think these are just preliminary data; I am not ready to replace diabetes with alpha-hydroxybutyratitis.

Q: You said that these could be independent predictors, but alpha-hydroxybutyrate is downstream of free fatty acids and branched-chain amino acids. So what is the added value?

A: I didn’t say that it was independent of those. I said that it was independent of the phenotypic predictors used in risk-stratification schemes such as FinnDiane. It’s also independent of glucose levels. If you control for amino acids in the analysis you reduce the predictive power. The predictive variable with the most power tends to butt out the others. The significance physiologically is a bit uncertain.



Maren Carstensen (German Diabetes Center, Dusseldorf, Germany)

Ms. Maren Carstensen presented a prospective genetic analysis of 513 patients followed over a seven- year period, of which 50 developed type 2 diabetes. Patients were given glucose tolerance tests at baseline in 1999-2001 and at follow-up in 2006-2008. Those who developed type 2 diabetes were older (65 years vs. 63 years), had a higher BMI (31 kg/m2 vs. 28 kg/m2), a higher prevalence of hypertension (72% vs. 49%), and a higher A1c (5.8% vs. 5.6%). The researchers examined differentially expressed transcripts associated with the 50 cases of incident type 2 diabetes. After adjusting for age, sex, and  BMI, the strongest associations were observed for eIF2 (eukaryotic translation initiation factor-2) signaling), eIF4/p70S6K (70-kDa ribosomal S6 kinase) signaling, and mTOR (mammalian target of rapamycin) signaling. There was also evidence for an enrichment of differentially expressed transcripts and pathways involved in endoplasmic reticulum stress, inflammation, immune responses, lipid metabolism, endocrine function, mitochondrial function and cell/tumor proliferation and apoptosis. Ms. Carstensen concluded by noting that this was the first prospective transcriptomics study on incident  type 2 diabetes and the results could be used in the future for prediction of type 2 diabetes.



Tina Rönn, PhD (Lund University, Lund, Sweden)

Dr. Tina Rönn’s presentation explored the effect of exercise on DNA methylation. Dr. Rönn reminded the audience that exercise has been shown to improve insulin sensitivity and weight control, but that the mechanism mediating these improvements is not fully understood. Striving to shed insight on exercise- mediated physiological improvements, Dr. Rönn conducted a study in 23 males analyzing DNA methylation and mRNA expression in adipose tissue biopsies before and after a six-month exercise intervention. The exercise intervention significantly improved participants’ VO2max, waist   circumference, waist/hip ratio, diastolic blood pressure, pulse, and HDL levels, but importantly, had no affect on BMI. Dr. Rönn found a global increase in DNA methylation in subcutaneous adipose following exercise, and, looking at specific CpG sites (i.e., sites within the genome that are often methylated) that showed significant differences in methylation pattern, she found that 90% of CpG sites had increased DNA methylation and 10% had decreased methylation. Additionally, when comparing CpG sites that fell within gene regions and showed significant methylation differences pre- to post-intervention, Dr. Rönn found that 34% of sites also showed a difference in the mRNA expression of that gene.

  • The study showed increased DNA methylation within the ITPR2 gene, which is associated with waist/hip ratio (one of the clinical characteristics that changed significantly following the intervention). At baseline, the CpG site within the ITPR2 gene body averaged 56.8% methylation. Following the exercise intervention, the site had 63.3% methylation (p <0.0001). While this was one of the larger percentage increases in DNA methylation, it begs the question of whether a more successful exercise intervention (i.e., one resulting in improved BMI, LDL, or triglycerides) would have had more pronounced results on DNA methylation. Of course, methylation is only significant in this context if it is linked to changes in gene expression, which in turn need to be linked to physiological changes.
  • Dr. Rönn’s investigation into the effects of physical activity on gene expression nicely complements Dr. Tim Frayling’s (Peninsula College of Medicine & Dentistry, Exeter, United Kingdom) Minkowski lecture, in which he argued that today’s obesogenic environment amplifies the effect of risky genes. For our discussion on Dr. Frayling’s lecture, please see page 23 of our Day #3 highlights at Certainly, both talks underscore the importance of addressing the obesogenic environment we currently live in.


Symposium: GLP-1 Beyond the Pancreas


Trine S. R. Neerup, PhD (Zealand Pharma, Copenhagen, Denmark)

In front of a packed conference hall, Dr. Trine Neerup (a substitute for Dr. Dorthe Almholt, who was unfortunately absent, but scheduled to give this talk) detailed the effects of Zealand Pharma’s GLP- 1/gastrin dual agonist ZP3022 on glucose levels and beta cell mass in db/db mice. ZP3022 combines the N terminus of exendin-4 with the C terminus of gastrin, and has a half-life in mice that supports once- daily dosing in humans. Over eight weeks in male db/db mice, ZP3022 provided similar reductions in A1c and in fasting plasma glucose (FPG) compared to liraglutide, and both drugs led to significant improvements in A1c and FPG relative to vehicle. While a statistically significant increase in beta cell mass was observed with both liraglutide and ZP3022 at four weeks (relative to vehicle), only ZP3022 maintained this increase throughout the eight-week treatment period, indicating that ZP3022’s effect on beta cell mass may likely be independent of its ability to lower glucose levels. Furthermore ZP3022, but not liraglutide, led to an increase in total islet cell mass at eight weeks that appeared to be driven by an increase in mean islet mass. Data also showed that ZP3022 significantly increased beta cell proliferation, but had no statistically significant effect on beta cell death, though ZP3022 trended  toward reducing apoptosis.

  • Dr. Almholt opened her presentation by briefly reviewing the rationale behind GLP- 1/gastrin dual agonists. She noted that GLP-1 agonists improve glycemic control in humans and rodent models, and increase beta cell mass in mice and rats. The peptide hormone gastrin is secreted by G cells and has been shown to stimulate beta cell regeneration, neogenesis, and expansion. In previous studies, the combination of GLP-1 and gastrin has been found to preserve and increase beta cell mass in mice models, and to induce beta cell regeneration in human islets transplanted into mice.
  • The GLP-1/gastrin dual agonist ZP3022 is formed by linking the N terminus of exendin-4 (amino acids 1-28), which binds to the GLP-1 receptor, to the C terminus of gastrin, which interacts with the CCKB receptor. The resulting compound has a bioavailability in mice of 100% (compared to 67% with liraglutide) and a half-life in mice of two hours (compared to 0.3 hours for exendin-4 and 4.3 hours for liraglutide), which supports once-daily dosing in humans. ZP3022 binds to the GLP-1 receptor with a potency eight-fold greater  than that of liraglutide, but similar to that of exendin-4 (not surprising, since ZP3022 includes the receptor binding domain of exendin-4). As expected, ZP3022 binds to the CCKB receptor at a potency over 60-fold greater than that of liraglutide and exendin-4.
  • Study design: Nine-week-old male db/db mice were stratified by A1c and fasting plasma glucose and at ten weeks of age, were given ZP3022 (50 nmol/kg twice daily), liraglutide, or vehicle for a treatment period of two, four, or eight weeks.
  • Results – glycemic efficacy: During the two-, four-, and eight-week treatment periods, ZP3022 and liraglutide provided similar reductions in A1c and fasting plasma glucose that were significantly greater than those observed with vehicle (the only exception was the A1c measurement taken at the end of the two-week period, in which only ZP3022 provided a statistically significant decrease in A1c vs. vehicle [p <0.05]).
  • Results – pancreatic effects: while a statistically significant increase in beta cell mass was observed with both liraglutide and ZP3022 at four weeks (relative to vehicle), only ZP3022 maintained this increase throughout the eight-week treatment period (beta cell mass with liraglutide at week eight was numerically greater vs. vehicle, but did not reach statistical significance). When coupled with the results on glycemic efficacy, this finding indicates that ZP3022’s effect on beta cell mass is likely mediated by mechanisms independent of reducing blood glucose levels.
  • Furthermore, ZP3022 therapy but not liraglutide resulted in a significant increase  in non-beta cell mass at eight weeks. Similarly, only ZP3022 provided a significant increase in total islet cell mass at eight weeks, that appeared to be driven by an increase in mean islet mass rather than an increase in the total number of islets. Data also show that ZP3022 significantly increased beta cell proliferation (as measured by Ki-67+ beta cell number and density), but had  no statistically significant effect on beta cell apoptosis, though ZP3022 trended toward decreasing cell death (as measured by caspase 3+ mass fraction).

Questions and Answers

Q: The data you show in beta cell mass are interesting. You see a very nice doubling of beta cell mass. What happens when you stop the treatment? Do you keep the normalization, or improvement in glycemic control?

A: Actually, we’ve just published an article in Diabetes, Obesity, and Metabolism where we stop treatment after several weeks. For some weeks, you still have an improvement in glycemic control, so it seems like there’s a great effect on beta cell mass.

Q: If I understood it correctly, you have an increase in non-beta cell mass. Previous studies have shown that GLP-1 is produced by a subfraction of alpha cells. Have you looked at GLP- 1 levels in plasma, or even better, extracted from islets? That can influence beta cell mass.

A: That’s a good question. Unfortunately, we have not measured that GLP-1, but surely we will do it later.

Q: Elevated gastrin level are associated with tumors in the stomach. I was just wondering, have you done any long-term studies in your mice to look for tumors?

A: We haven’t done it but of course we need to. We have some speculation. If you look at patients that get PPIs, which increase gastrin levels, they don’t get tumors.


Symposium: Oral Therapies: New Targets


Stefano Del Prato, MD (University of Pisa, Pisa, Italy)

Dr. Stefano Del Prato began this review lecture by noting that, despite the recent “renaissance” of type 2 diabetes drug development, the only available therapies that improve insulin sensitivity are metformin and the TZDs. He proceeded to discuss investigational agents that might also enhance the body’s response to insulin. One investigational strategy is to try to mimic metformin and act through AMP kinase. AMPK activators that have been studied for type 2 diabetes include members of the thienopyredone family and derivatives of D-xylose; Dr. Del Prato said that the mechanism was   powerful but is difficult to make tissue-specific. An alternative strategy is to improve upon the TZDs, as is being attempted with PPAR-alpha/-gamma dual agonists (e.g., aleglitazar), PPAR-alpha/-delta/- gamma triple agonists (e.g., indeglitazar), and selective PPAR modulators (SPPARMs, e.g., INT131). Dr. Del Prato suggested that the most promising approach of all is to develop agents acting along the intracellular insulin-signaling pathway. On this note, he expressed particular enthusiasm for FGF21- based agents and also mentioned insulin mimetics, PTP1b inhibitors, antioxidants, anti-inflammatory agents, and blockers of fatty acid oxidation. He said that one thing seems sure: that the complex pathophysiology of type 2 diabetes will lead to an increasingly complex set of therapies. (Roughly 150 type 2 diabetes drugs are in development, he mentioned). Dr. Del Prato’s prediction is thus that diabetologists in the future will need, more and more, to become specialists – a prediction that seems sadly incompatible with the enormous incidence of the disease and today’s poor reimbursement for diabetes care.

  • Inspired by the mechanism of metformin, several drugs in development activate AMP kinase in order to improve insulin sensitivity and reduce inflammation. Dr. Del Prato looked in particular at two families of AMPK activators. The thienopyredones include A- 769662, which was shown to have positive effects on blood glucose and triglycerides in preclinical research (Cool et al., Cell Metab 2006). He also mentioned derivatives of D-xylose, such as E-36, which improved insulin sensitivity and decreased blood glucose. Dr. Del Prato explained that the many potential benefits of AMPK activation include increased glucose uptake and lipid oxidation in the skeletal muscle, increased exercise capacity, improved vascular function, and improved  lipid profile. However, he cautioned AMPK activation is difficult to make tissue-specific, and potential side effects of global exposure cardiomyopathy and increased appetite. In this vein Dr. Del Prato briefly mentioned SIRT1 activation as an alternative way to achieve benefits on multiple organ systems, though the candidate that he seemed most excited about was resveratrol in the form of Pinot Noir.
  • The most advanced insulin-sensitizing drug currently in development is Roche’s aleglitazar, a PPAR-alpha/-gamma dual agonist that is being studied in a 7,000- patient cardiovascular outcomes study called ALECARDIO. The 150 ug dose of  aleglitazar causes similar glycemic benefits to 45 mg pioglitazone, with better lipid effects (Henry et al., Lancet 2009). Side effects like weight gain and edema were reduced relative to pioglitazone, but still present – and Dr. Del Prato said that he thinks the side effects will be important to keep   in mind. He seemed somewhat more optimistic about PPAR-alpha/-delta/-gamma agonists, such as indeglitazar, which has been shown to lower body weight while improving glycemia (Artis et   al., PNAS 2009). An alternative to achieve efficacy with minimal side effects is offered by selective PPAR modulators (SPPARMs); as an example, Dr. Del Prato described phase 2a data for INT131 showing improvements in general metabolic profile (Dunn et al., J Diabetes Complications 2011).
  • In the third stage of his talk, Dr. Del Prato consulted a diagram of intracellular insulin signaling and reviewed the variety of therapeutic targets along this pathway. He explained that PTP-1b inhibitors could theoretically remove a major barrier to insulin signaling, and many agents have been studied. However, as detailed in a recent review (Verspohl, Pharmacol Rev 2012), such inhibitors need to be highly charged molecules – making oral dosage  a “tremendous challenge.” Fatty acids are another big challenge to insulin signaling, and blocking fatty acid oxidation with teglicar improves blood glucose – but unfortunately also increases liver triglyceride content (Conti et al., Diabetes 2011). The mechanism for which Dr. Del Prato seemed most enthusiastic was FGF21 signaling, which involves the liver and potentially brown fat. In a db/db mouse model, FGF21-based therapy has been shown to confer significant metabolic benefit (Wu et al., Sci Transl Med 2011).


Symposium: Novel Mediators of Metabolic Stress


James Kirkland, MD, PhD (Mayo Clinic, Rochester, MN)

Dr. James Kirkland argued that aging is the single biggest risk factor for many of the diseases that   affect society and thus, interfering with aging mechanisms could help delay multiple diseases (like type   2 diabetes and obesity). He reviewed evidence suggestive of close links between fat tissue and aging – decreasing visceral fat surgically increases maximum lifespan and obesity. Furthermore,  lipodystrophies are associated with an increase in many age-related diseases. Dr. Kirkland believes  there is a link between fat and the life- and health-span and as such, his research has concentrated on cellular senescence in fat tissue. Dr. Kirkland reviewed a series of in vitro and murine studies to demonstrate that senescent fat cell progenitor cells (i.e., the preadipocyte precursor cell) accumulate  both with aging and with obesity and that preadipocytes (i.e., adipocyte precursor cells) interfere with normal adipogenesis, and have a pro-inflammatory secretory state that attracts macrophages. Dr. Kirkland then showed that by ameliorating the senescence-associated secretory phenotype, he could increase fat mass and decrease frailty in elderly mice. The next step, said Dr. Kirkland, is to extend these studies to explore adipose senescence in obesity and diabetes.



Rinke Stienstra, PhD (Radboud University, Nijmegen, The Netherlands)

Dr. Rinke Stienstra presented on the role of the inflammasome in insulin resistance and therapeutic potential of targeting the inflammasome. For background, the inflammasome, he said, is the “guardian of the intracellular environment.” Importantly, he argued that the inflammasome controls caspase-1 activation, which in turn enhances cytokine production and leads to insulin resistance. Dr. Stienstra suggested that during obesity, inflammasome-mediated caspase-1 is activated in adipose tissue. Thus,  he posited that 1) inhibiting caspase-1 could reduce inflammation and improve insulin sensitivity and 2) blocking downstream effectors of caspase-1 activation could similarly improve insulin sensitivity. He highlighted several agents that potentially acted through these mechanisms including: 1) glyburide (the most widely used sulfonylurea for the treatment of type 2 diabetes in the US), which been shown to block inflammasome activation; 2) diarylsulfonylureas, which inhibit cytokine release and are currently   being investigated in obese animals; and 3) anakindra, an IL-1 receptor agonist (i.e., blocks   downstream effectors of caspase-1).

Questions and Answers

Comment: Glyburide is an inhibitor of the inflammasome, but my understanding is that the doses necessary far exceed levels we can obtain in the clinic so perhaps we should not believe too much in this drug.


Symposium: microRNAs and Diabetes


Markus Stoffel, MD, PhD (Eldengössische Technische Hochschule Zürich, Zürich, Switzerland)

Dr. Markus Stoffel’s presentation investigated microRNA-mediated insulin resistance and brown adipose formation. He explored two microRNA in depth: miR-103/107 (miR-103 and -107 are homologous) in white adipose and miR-133 in brown adipose. Considering the former, Dr. Stoffel posited that miR-103/107 regulates insulin sensitivity by acting through caveolin-1, a direct target of miR-103/107. Considering the latter, Dr. Stoffel suggested that miR-133 inhibition stimulates brown adipose tissue formation and activation.

MiR-103/7: Dr. Stoffel’s exploration of miR-103/7 began with the finding that miR-103/7 is upregulated in the livers of obese mice. In an elegant series of murine studies that followed, Dr. Stoffel found: 1) overexpression of miR-103/7 induces hypoglycemia and leads to glucose intolerance; 2) miR-103/7 silencing normalizes blood glucose levels in diabetic mice; 3) overexpression of miR-103/7 in visceral fat pad only (which accounts for ~25% of total fat) was sufficient to increase blood glucose level and confer insulin resistance; 4) miR-103/107 silencing increased glucose uptake in primary adipocytes from obese mice in both subcutaneous and visceral fat; 5) miR-103/107 silencing reduces fat mass and adipocyte size via increased adipocyte differentiation; and 6) miR-103/107 acts on insulin sensitivity through its caveolin-1 target. These studies potentially position miR-103/107 and caveolin-1 as potential areas of investigation for therapeutic targets – we’re eager to follow this early-stage work.

MiR-133: In a similar step-wise progression, Dr. Stoffel showed that miR-133 inhibition is intimately involved in brown adipose tissue (BAT) formation and activation. Importantly, he presented results demonstrating: 1) miR-133 levels are reduced in brown adipose tissue after cold exposure; 2) miR-133 is regulated by transcription factor Mef2c; 3) miR-133 silencing increases Prdm16 expression (Prdm 16 is needed for BAT differentiation); and 4) miR-133 silencing increases BAT function. As such, Dr. Stoffel suggested that miR-133 inhibition could be a novel therapeutic approach to increase energy expenditure.

Questions and Answers

Q: Is miR-133 involved in the browning process?

A: In the interest of time, I only showed brown fat, but we also performed these tests on preadipocytes from white fat,. As you know there are brownish cells and they behave the same as brown fat. We find the same regulation, so I do believe it plays a role in the browning effect.

Q: Is miR-103/107 implicated in skeletal muscle?

A: I am not sure, I cannot exclude it. miR-103/107 is expressed in lower levels in the muscle than the fat. And caveolin-3 is the main caveolin in the muscle, which is not the target of miR-103/107.

Q: How do you feel about importance of microRNAs in terms of abundance?

A: My feeling is from all knock out models, is that low expression levels are less likely to have major functions.



Emmanuel Van Obberghen, MD, PhD (Université de Genève, Genève, France)

Dr. Emmanuel Van Obberghen presented on the role of microRNAs (miRNA) in insulin secretion. First, Dr. Van Obberghen detailed a series of mouse studies that illuminated a pathway by which endogenous glucose production affects miRNA-mediated insulin transcription. He showed that glucose: 1) inhibits miRNA-375 expression, which leads to the increased transcription of PDK-1; and 2) increases miRNA- 214 expression, which inhibits PTEN (a phosphatase and tensin humolog protein). The downstream  effect of both these cascades is increased PI3-K activation, which in turn leads to increased activation of the insulin gene. Then, Dr. Van Obberghen transitioned to the second part of his presentation on epigenetics, and argued that miRNAs were involved in the fetal programming of type 2 diabetes. Through a series of murine experiments, Dr. Van Obberghen reasoned that a low protein diet during pregnancy reduces beta cell mass development in progeny by affecting beta cell proliferation (but not beta cell apoptosis). Furthermore, he suggested that the low protein diet acts by increasing miRNA-375 (which decreases fetal beta cell proliferation by inhibiting PDK-1 expression), and that the fetus has a “metabolic memory” such that altered fetal miRNA-375 expression can contribute to the development of type 2 diabetes later in life. Thus, he encouraged the audience to seriously revisit the thrifty phenotype hypothesis (that adaptations made by a fetus in response to limited nutrient supply increases susceptibility for developing chronic conditions later in life) and consider the important role that miRNA plays.

Questions and Answers

Q: You showed a U-shaped curve for birth weight, such that low birth weight and high birth weight could have similar deleterious effects on insulin sensitivity. Are they triggering the same program? Or are they resulting in the same end, but by different pathways?

A: I have no data on that. I think the evidence on the low birth weight side is probably stronger than that on the high birth weight side. The reason is if you have diabetic mice that are overweight, you have confounding effects of the metabolic situation, so it’s harder to give clear evidence.


Symposium: Epigenetics and Diabetes


Klaus Kaestner, PhD (Perelman School of Medicine University of Pennsylvania, Philadelphia, PA)

Dr. Klaus Kaestner showed that it is relatively common (more often than for beta cells) for alpha cells in the islets to have genes that are both epigenetically activated and repressed by epigenetic factors (the repressive factor is typically dominant so that the gene is not expressed). He explained that this “bivalent” epigenetic marking is common in embryonic stems cells because it allows a gene to be more quickly turned on. Rather than having to remove the epigenetic repressor and add an epigenetic activator, the cell only has to remove the repressor to express the gene. Dr. Kaestner hypothesized that alpha cells might have a relatively flexible identity (though on a more limited scale than stem cells) and be capable of converting into beta cells. When he applied an inhibitor of epigenetic markers to human alpha cells in culture, he found that a small fraction of the alpha cells began secreting insulin in addition glucagon. He thus proposed that a drug that triggered this phenomena on a larger scale could potentially reprogram alpha cells to secrete both insulin and glucagon, or even fully convert alpha cells into beta cells. We think converting alpha cells into beta cells could be a potentially very potent  treatment for late stage type 2 diabetes, as it would address the person’s reduced insulin secretion and increased glucagon secretion. It could also be useful for type 1 diabetes, but like islet or pancreas transplantation, such therapy would require immunosuppressants, limiting its value for people with type 1 diabetes.

Questions and Answers

Q: Have you thought of knocking down an enzyme that causes epigenetic repression in alpha cells, as a way of producing a better drug?

A: That is a brilliant idea.

Q: On the reprogrammed alpha cells, have you had a chance to see if the secretory response became more like a beta cell function?

A: We would love to do that, but unfortunately there are not enough cells that change to do population studies.




Soren Toubro, William Cefalu, John Xie, Daniel Sullivan, Keith Usiskin, William Canovatchel, Gary Meininger

This study characterized the relative contributions of loss of fat mass and loss of lean mass to overall body weight loss with canagliflozin (CANA) treatment in subsets of two phase 3 studies in which DXA body composition measurements were performed. Loss of fat mass accounted for approximately two- thirds of the overall weight loss seen with canagliflozin treatment. The authors note that this proportion is similar to those for other diabetes medications that are accompanied by weight loss. In Study 1, patients experienced slightly more loss of visceral fat than subcutaneous fat at the 52-week mark, as determined by abdominal fat analysis. Patients on CANA 100 mg and CANA 300 mg experienced average 5.4% and 5.6% losses of subcutaneous fat, compared to 7.3% and 8.1% losses of visceral fat, respectively.

  • This study characterized the relative contributions of loss of fat mass and loss of lean mass to overall body weight loss with canagliflozin (CANA) treatment in subsets of two phase 3 studies in which DXA body composition measurements were performed. Study 1 compared the efficacy and safety of CANA 100 mg and CANA 300 mg  versus glimepiride (GLIM) as an add-on to stable metformin therapy in patients with type 2 diabetes between 18-80 years old, with baseline A1c between 7.0-9.5% and BMI between 22-45 kg/m2, over a 52-week treatment period. Study 2 evaluated the efficacy and safety of CANA 100 mg and CANA 300 mg versus placebo in older individuals with type 2 diabetes (aged 55-80), with baseline A1c between 7.0-10.0% and BMI between 20-40 kg/m2, over a 26-week treatment  period. At baseline, patients in Study 1 had an average weight of 84-86 kg (185-189 lbs), and patients in Study 2 had an average weight of 89-95 kg (196-209 lbs).

Study 1, Week 52 (LOCF):


CANA 100 mg (n=71)

CANA 300 mg (n=69)

GLIM (n=68)

Body weight % change





Total fat measurement change (kg [lbs])


-2.9 (-6.4)


-2.5 (-5.5)


1.0 (2.2)

Total lean measurement change (kg [lbs])


-0.9 (-2.0)


-1.1 (-2.4)


1.1 (2.4)

Percent total fat change (%)





Study 2, Week 26 (LOCF):


CANA 100 mg (n=63)

CANA 300 mg (n=73)

GLIM (n=75)

Body weight % change





Total fat measurement change (kg [lbs])


-1.8 (-4.0)


-2.5 (-5.5)


-0.3 (-0.7)

Total lean measurement change (kg [lbs])


-1.1 (-2.4)


-1.1 (-2.4)


-0.3 (-0.7)

Percent total fat change (%)






David Matthews, Greg Fulcher, Vlado Perkovic, Dick de Zeeuw, Kenneth Mahaffey, Julio Rosenstock, Melanie Davies, George Capuano, Mehul Desai, Wayne Shaw, Frank Vercruysse, Gary Meininger, Bruce Neal

This study investigated the efficacy and safety of canagliflozin in the subgroup of patients in CANVAS, the cardiovascular outcomes trial for canagliflozin, on at least 30 units of insulin per day at baseline. Participants were randomized to receive canagliflozin (CANA) 100 mg, CANA 300 mg, or placebo in a 1:1:1 fashion. Over 18 weeks, patients on canagliflozin experienced significant reductions in A1c, fasting plasma glucose, weight, and systolic blood pressure versus placebo. Specifically, patients on CANA 100 mg, CANA 300 mg, and placebo experienced -0.63%, -0.72%, and 0.01% changes in A1c over 18 weeks. Canagliflozin treatment was generally well tolerated, but was associated with higher rates of genital and urinary tract infections, as well as osmotic diuresis-related adverse events.

  • Baseline characteristics were similar across treatment arms. Key inclusion criteria were: 1) age ≥30 years; 2) a history or high risk of cardiovascular disease (≥30 years of age with documented, symptomatic, atherosclerotic cardiovascular disease, or ≥50 years of age with at least two risk factors for cardiovascular disease at screening); 3) A1c between 7.0% and 10.5%.


CANA 100 mg (n=566)

CANA 300 mg (n=587)

Placebo (n=565)

Age (years)




BMI (kg/m2)




Diabetes duration (years)







Duration of continuous insulin therapy (years)







Insulin dose (IU)




A1c (%)




Fasting plasma glucose (mmol/l [mg/dl])


9.4 (170)


9.3 (167)


9.4 (170)

Body weight (kg [lbs])

96.9 (213)

96.7 (213)

97.7 (216)

Systolic blood pressure (mmHg)







  • After 18 weeks of treatment, canagliflozin brought about significantly greater reductions in A1c, fasting plasma glucose, body weight, and systolic blood pressure versus placebo. Canagliflozin treatment showed a trend toward reduced diastolic blood pressure versus placebo, though not statistically significant.

Change at Week 18

CANA 100 mg

CANA 300 mg


A1c (%)




Fasting plasma glucose (mmol/l [mg/dl])


-1.0 (-18)


-1.4 (-25)


0.2 (3.6)

Body weight (kg [lbs])

-1.8 (-4.0)

-2.3 (-5.1)

0.1 (0.2)

Systolic blood pressure (mmHg)







*all comparisons p<0.001 vs. placebo

  • Canagliflozin treatment was associated with a higher incidence of genital tract infections versus placebo. The incidence of urinary tract infections was higher in the CANA 300 mg group versus CANA 100 mg or placebo. The overall incidence of adverse events was slightly higher with canagliflozin treatment than with placebo (64.0%, 65.1%, and 59.1% for CANA 100 mg, CANA 300 mg, and placebo, respectively).


CANA 100 mg (n=566)

CANA 300 mg (n=587)

Placebo (n=565)

Genital mycotic infection (male)


15 (4.0%)


32 (8.3%)


2 (0.5%)

Genital mycotic infection (female)


22 (11.8%)


20 (9.9%)


4 (2.2%)

Urinary tract infection

13 (2.3%)

20 (3.4%)

12 (2.1%)


Corporate Symposium: Delivering Innovation in Type 2 Diabetes – Tailored Approaches with SGLT-2 and Incretin-Based Therapies (Sponsored by BMS/AZ)


Kamlesh Khunti, MD, PhD (University of Leicester, Leicester, UK)

Dr. Kamlesh Khunti detailed to the audience that barriers to effective treatment of type 2 diabetes include: poor adherence to therapy, fear of hypoglycemia, physicians’ behavior, natural history of the disease, and weight gain. Particularly noteworthy was Dr. Khunti’s impression of the ADA/EASD position statement (for our full initial commentary on the ADA/EASD position statement see our April 20, 2012 Closer Look at and the UK’s National Institute for Health and Clinical Excellence (NICE) guidelines. Dr. Khunti believes that both sets of guidelines make the lives of HCPs easier and that the main takeaway from these guidelines is that HCPs should be individualizing treatment decisions. In a manner that we would expect would be very useful to non-diabetes specialists, Dr. Khunti then described his three major types of patients and how he would tailor the ADA/EASD guidelines to meet their individual needs: 1) the patient that needs to avoid weight gain; 2) the patient with high rates of hypoglycemia; and 3) the cost-conscientious patient. For a patient with the first goal, Dr. Khunti recommended using either a GLP-1 agonist or DPP-4 inhibitor as the second-line treatment after metformin fails. If the goal is to avoid hypoglycemia he said a thiazolidinedione was another option in addition to incretins. Finally, if a patient’s main goal is to minimize cost then he suggested using a sulfonylurea (SFU) as the second tier treatment. This recommendation sparked a lively discussion amongst the panelists; Dr. Edoardo Mannucci said that SFUs should not be used – saying he would “never” take an SFU due to their inferior efficacy and safety profile to other options and that the only reason they are still used is that they are cheaper. In response, Dr. Khunti stated that currently in the UK, HCPs are being pushed to use SFUs but that it is mainly due to cost. We hope that governments and payors will put patients’ health higher on their list of priorities. Additionally, we question if prescribing SFUs is a good long-term investment considering the hypoglycemia and weight gain associated with all SFUs and the beta-cell burnout associated with some SFUs (and therefore greater medical need) the class is known to cause.



Paola Fioretto, MD, PhD (University of Padova, Padova, Italy)

Standing in front of a small apple tree, Dr. Paola Fioretto began her presentation with a short video outlining the history of dapagliflozin, beginning with the isolation of phlorizin (an SGLT-2 and SGLT-1 inhibitor) from apple tree bark. The problem with phlorizin, Dr. Fioretto explained, was that it  interacted with both SGLT-1 (which is located in the kidney and intestine) and SGLT-2, causing malabsorption of glucose and GI side effects. Dapagliflozin (BMS/AZ’s Forxiga), however, is a selective inhibitor of SGLT-2, the enzyme responsible for 90% of the kidney’s glucose reabsorption (which totals  on average 180 g glucose/day). Dr. Fioretto explained that in a person without type 2 diabetes glucose transporters cannot reabsorb more than 200 mg glucose/dl, however, in people with type 2 diabetes there is a counterproductive up-regulation of SGLT-2 resulting in even higher levels of glucose being reabsorbed by the kidneys. Dapagliflozin has been found to cause the excretion of glucose (glucosuria) of at most 70 g glucose/1.73 m2/day, which corresponds to a caloric loss of 100-300 kcal a day or weight loss or about 3 kg (6.6 lbs.) total (which studies have found is maintained for at least two years). Interestingly, studies of familial renal glucosuria (the genetic inheritance of an absence or reduction of SGLT-2) have found that when glucosuria is <100 g/1.73 m2/day the condition is asymptomatic. Thus, Dr. Fioretto argued that dapagliflozin does not cause enough glucosuria to produce serious side effects, and that by working in an insulin independent manner (it inhibits SGLT-2 no matter how much insulin is present in the body) SGLT-2 inhibitors lower glucose levels without greatly elevating risk of hypoglycemia, and have a similar efficacy throughout the progression of type 2 diabetes.



Samy Hadjadj, MD, PhD (Poitiers University Hospital, Poitiers, France)

Dr. Samy Hadjadj began his presentation by asking the audience what degree of A1c lowering they would expect to see with dapagliflozin: A) <0.5%, B) 0.5-1.0%, or C) 1.0-1.5%. The majority of the audience, 71%, answered the question correctly: 0.5-1.0% (22% selected 1.0-1.5% and 7% picked <0.5%). Dr. Hadjadj then provided a comprehensive review of the efficacy data from clinical trials performed on dapagliflozin (BMS/AZ’s Forxiga), highlighting that an A1c reduction of 0.8% to 0.9% occurs relatively consistently no matter a person’s baseline A1c (potentially because it is insulin independent). He also emphasized that dapagliflozin is associated with an average of about 3 kg (6.6 lbs.) of weight loss (from  a baseline of 80-90 kg [~177 - 198 lbs.]). This weight loss was maintained for the course of the two-year study, as was a reduction in blood pressure. Overall, dapagliflozin’s phase 3 clinical development program had 5,693 enrollees with differing disease progression and treatment backgrounds.



Andreas Pfeiffer, MD (Charité University Hospital, Berlin, Germany)

Dr. Andreas Pfeiffer, Chairman of the Local Organizing Committee of this year’s EASD, continued Dr. Samy Hadjadj’s review of the clinical data for dapagliflozin (BMS/AZ’s Forxiga), but focused on its safety and tolerability results. Dr. Pfeiffer noted that bladder and breast cancers were somewhat more frequent in those receiving dapagliflozin (increased incidence compared to placebo of bladder cancer [0.16% vs. 0.03%] and breast cancer [0.40% vs. 0.22%] is one of the main concerns that the FDA has with dapagliflozin; see our January 19, 2012 Closer Look on the drug receiving a CRL from the FDA at and our April 23, 2012 Closer Look on the positive CHMP opinion at However, Dr. Pfeiffer stated that preclinical data found no SGLT-2 target expression in either bladder or breast tissues, decreasing the likelihood that dapagliflozin causes tumor formation in these locations. He also emphasized that the overall cancer rates were similar between dapagliflozin and the control, and that some types of cancers (such as renal tract cancer) were more prevalent in the control. Additionally, during Q&A he detailed that very few people actually developed bladder cancer (about ten) and that the majority of these cases displayed signs of the cancer having  been present before treatment with dapagliflozin began. Dr. Edoardo Mannucci concurred, arguing  that the duration of these trials was too short for a cancer to have formed in response to the treatment. During his prepared remarks, Dr. Pfeiffer also discussed the increase in genital and urinary tract infections associated with dapagliflozin. He emphasized that these infections can be treated with standard care and mainly occur during the first six months of treatment. During Q&A Dr. Paola  Fioretto also noted that these infections tend not to reoccur, so it is unlikely a patient will get such infections frequently due to dapagliflozin. Additionally, Dr. Khunti (a primary care physician) said that most people – particularly people with diabetes – have had a genital or urinary tract infection before and are comfortable identifying and treating it themselves over the counter.



Kamlesh Khunti, MD, PhD (University of Leicester, Leicester, United Kingdom); Paola Fioretto, MD, PhD (University of Padova, Padova, Italy); Samy Hadjadj, MD, PhD (Poitiers University Hospital, Poitiers, France); Andreas Pfeiffer, MD (Charité University Hospital, Berlin, Germany); Edoardo Mannucci, MD (Careggi Teaching Hospital, Florence, Italy); Laurie Baggio, PhD (Samuel Lunenfeld Research Institute, Toronto, Canada); Petra-Maria Schumm-Draeger, MD, PhD (Clinic Munich Bogenhausen, Munich, Germany); Jiten Vora, MD (Royal Liverpool University Hospital, Liverpool, United Kingdom)

Questions and Answers

Q: Are urinary tract infections more frequent early in the treatment than later on?

Dr. Pfeiffer: There are some increases in the frequency of genital infections – a few percent more. So there is a small increase, but it can be handled by standard treatment.

Dr. Fioretto: There is clear data showing that there is an increase in genital infections and it tends to occur early on in treatment. The good news is there is usually only one episode; reoccurrences are not very common.

Dr. Hadjadj: The risk for one of these infections is greatest at the beginning of therapy but it becomes equal to the control in the long term.

Dr. Vora: Do you think this increase in urinary tract and genital infections should have a clinical  implication?

Dr. Fioretto: HCPs should explain to their patients that there may be this problem and to come see them if they have symptoms.

Dr. Khunti: Many patients are able to handle these problems on their own with over-the-counter medications, so I think education is important. If over-the-counter medications are not adequate then they should go see their doctor.

Dr. Pfeiffer: Would you do anything preventive?

Dr. Khunti: Most people – especially if they have diabetes – have had a urinary or genital infection before in their life. So they know the symptoms and are comfortable treating it. I think one just needs to educate them that there may be a slight increase in these infections.

Dr. Hadjadj: People do treat themselves for these infections on their own. Our job as clinicians is to warn them that they may have genital infections and to ask them if they have these symptoms.

Q: Does dapagliflozin have any issues regarding cardiovascular safety?

Dr. Fioretto: There has been a publication pooling the data from several trials and it looks like there is no increase in cardiovascular risk. If anything cardiovascular risk is going in the right direction for people on dapagliflozin.

Dr. Pfeiffer: There is no signal at the moment but we need more data.

Q: In which patients would you recommend using dapagliflozin?

Dr. Khunti: I would say that dapagliflozin is a primary care drug. After metformin you could use this in pretty much any patient. Most patients with type 2 diabetes are obese and we want to avoid them gaining weight, and better yet help them lose weight. We also want to avoid hypoglycemia.

Dr. Vora: In whom would you not use dapagliflozin?

Dr. Khunti: Those with renal impairment.

Dr. Fioretto: I agree that dapagliflozin could be helpful at any point in the spectrum of diabetes. It is hard to imagine it as a first line given we have metformin, but maybe for patients who cannot tolerate metformin. As far as in which patients I would not use dapagliflozin, I would be cautious if a patient is using a diuretic – not that I would not use it – I just would be careful because of dehydration.

Dr. Pfeiffer: I would use it in any patient with a glomerular filtration rate (GFR) greater than 60 ml/min/1.73m2.

Dr. Fioretto: I think that you should not use it in a patient with a GFR less than 60 ml/min/1.73m2. The problem is that at the beginning of dapagliflozin treatment you see some drop in GFR, as is the case with many diuretics. So I think if you have a patient with an estimated GFR of 100 ml/min/1.73m2 you really don’t have to worry. In contrast if you a patient with an estimated GFR of 70 ml/min/1.73m2 then I would check their GFR every few months to make sure it does not drop bellow 60 ml/min/1.73m2, at which case  I believe the recommendation is to stop.

Q: Why do we still recommend SFUs? Is it just because of price? Should we not be using SFUs?

Dr. Mannucci: I think we are going to stop using SFUs once the cost is comparable with the other treatment options. I think there is no reason to use an SFU beyond cost. We know their long-term efficacy is inferior to other options; their safety is also far from proven, so we really should not be using them.

Dr. Vora: Would you take an SFU?

Dr. Mannucci: No, never.

Dr. Khunti: In the UK we are currently driven by costs. So we are being pushed to use SFUs but I think that is mainly due to cost.

Q: Do you believe the cancer signals can be dismissed as balanced?

Dr. Pfeiffer: There were nine cases of bladder cancers; six of them had some hematuria before the onset of the study. This is not what you would expect if dapagliflozin was causing the bladder cancer. At present it is difficult to say if you can dismiss it. There is no reason to believe it is due to a specific effect. I think we should wait for more data.

Dr. Mannucci: I totally agree. We are talking of ten cases of bladder cancers distributed in two treatment groups; ten cases is far from proving anything. The duration of these trials were rather short, so the presence of those cancer cases were certainly present before the beginning of therapy. So I think we are really speaking of nothing at the moment.

Dr. Hadjadj: What is the duration of the phase 3 trial you would want to see to prove or disprove dapagliflozin causing cancer?

Dr. Mannucci: That is not a question for a phase 3 trial. That question is more for epidemiological observation than a clinical trial. I would probably want to see ten years of data.

Q: Should dapagliflozin be avoided in patients who are symptomatic? Who have hyperglycemia and may be dehydrated?

Dr. Hadjadj: In those cases you should probably go ahead to insulin. If you think the patient has a clear deficiency in insulin you should then begin with insulin. After one year on insulin you could add then add dapagliflozin. If you think your patient is symptomatic don’t turn around, go directly to insulin.

Q: What do you think about using dapagliflozin in the elderly who are prone to dehydration?

Dr. Hadjadj: Lets try to be careful. We don’t have any warning so far from the data, but let us stay alert.

Dr. Vora: You do not see being elderly as a contraindication?

Dr. Hadjadj: No.


Corporate Symposium: The Kidney – A New Therapeutic Partner (Sponsored by Janssen Pharmaceutical NV)


Luc Van Gaal, MD, PhD (Antwerp University Hospital, Edegem, Belgium)

Dr. Van Gaal highlighted the dangers of three key challenges to good health for people with diabetes – risk of hypoglycemia, blood pressure control, and weight management. He said that hypoglycemia in type 2 diabetes is chiefly a side effect of some current therapies and that it is associated with a substantial number of hospitalizations (Budnitz et al., NEJM 2011) as well as long-term cognitive decline. For its part, obesity (especially excess visceral fat) is associated with both macrovascular and microvascular complications, as well as insulin resistance. Thirdly, he noted that hypertension is an independent risk factor for macrovascular and potentially microvascular disease. The risks are higher still in patients with additional comorbidities (e.g., dyslipidemia) or risky lifestyle choices (e.g., smoking). All in all, he made the case that good diabetes management involves much more than reducing A1c. He also noted that the ADA/EASD 2012 position statement recommends less-stringent treatment guidelines in people with comorbidities; thus, addressing comorbidities can clear the way for more aggressive glucose-lowering interventions.



John Wilding, DM, FRCP (University of Liverpool, Liverpool, United Kingdom)

Dr. Wilding explained that renal reabsorption occurs mainly via sodium glucose transporter 2 (SGLT-  2), and to a lesser extent SGLT-1, and he outlined the potential utility of SGLT-2 inhibition to lower blood glucose in diabetes. Encouragingly, Dr. Wilding said that abnormally low glucose reabsorption  threshold seems to cause minimal negative health consequences in people with familial renal glycosuria (a genetic defect in SGLT-2 inhibition). (As a side note, he mentioned that a rare mutation in SLC5a1 – which codes SGLT-1 – can lead to glucose/galactose malabsorption, which is fatal if not treated by removing glucose and galactose from the diet. Dr. Wilding said that partial inhibition of SGLT-1 might  be therapeutically useful, “but we certainly don’t want to block it completely.”)

  • Dr. Wilding reviewed the maladaptive increase in renal glucose reabsorption threshold that occurs due to hyperglycemia. People without diabetes tend to lose glucose through the urine at plasma glucose concentration of roughly 8.3 mmol/l (~150 mg/dl) or higher, but people with hyperglycemia from type 1 or type 2 diabetes tend to have a higher renal reabsorption threshold, due to adaptive upregulation of the glucose transport proteins SGLT-2 and GLUT2. Thus glycosuria might not occur in people with diabetes unless plasma glucose is around 12 to 13 mmol/l (~216-234) or more. Inhibiting SGLT-2 could cause the reabsorption threshold to decline to non-diabetic levels (or even lower), so SGLT-2 inhibition could be a good way to reduce blood sugar (and also body weight, given that excess calories would be expelled).



Jochen Seufert, MD (University of Freiburg, Freiburg, Germany)

In this review of late-stage clinical trial data on SGLT-2 inhibition, Dr. Seufert showed previously presented data on BMS/AZ’s dapagliflozin, J&J Janssen’s canagliflozin, and BI/Lilly’s empagliflozin. All three SGLT-2 inhibitors, whether as monotherapy or in combination with other drugs, moderately reduce hyperglycemia without causing hypoglycemia in type 2 diabetes. (They could theoretically work in type 1 diabetes also, Dr. Seufert noted, given the insulin-independent mechanism of action). All three agents also tend to moderately reduce body weight and blood pressure. Unfortunately SGLT-2  inhibitors as a class seem to dose-dependently increase the risk of genital infections (particularly in women) due to the increase of sugar in the urinary tract. However, Dr. Seufert said that the prevalence of genital infection tends to be low and that the infections tend to be treatable.



Guntram Schernthaner, MD (Rudolfstitung Hospital, Vienna, Austria)

Turning to the elephant in the room of every session on new treatment options, Dr. Schernthaner discussed the widespread, pernicious problem of clinical inertia and offered several possible solutions – most notably, combination therapy with relatively low doses of each individual agent. He defined  clinical inertia as the failure to initiate or intensify therapy when doing so is clinically indicated, but he noted that patients can contribute to the phenomenon (since if a patient is non-adherent, a clinician will be less inclined to intensify the treatment regimen). He noted that early combination therapy of several agents, with low doses of each, can be a good way to achieve robust glucose control (and the concomitant benefits on beta-cell function and insulin sensitivity) with minimal side effects. Dr. Schernthaner is especially wary of side effects that increase the risk of clinical inertia and/or directly worsen outcomes; he gave the examples of weight gain, risk of severe hypoglycemia, bone strength, and potential cancer risk. We agree that early combination therapy could potentially prevent the common negative pattern of stepwise, wait-for-failure treatment intensification; we look forward to studies on the long-term impact of combination therapies.

  • Dr. Schernthaner believes that diabetes-related hospitalizations are an underused opportunity for effective diabetes management consultations, if performed in a way that complements a patient’s outpatient care team. He said that outpatient educational programs also have potential to improve adherence, though the data he showed on efforts in this regard were not wildly encouraging.



Luc Van Gaal, Degree (Antwerp University Hospital, Edegem, Belgium), John Wilding, DM, FRCP (University of Liverpool, Liverpool, United Kingdom), Jochen Seufert, MD (University of Freiburg, Freiburg, Germany), Guntram Schernthaner, MD (Rudolfstitung Hospital, Vienna, Austria)

Q: In SGLT-2 inhibitors, combination with metformin and insulin, is there a difference in risk of genital infections?

Dr. Seufert: It’s a bit variable in the co-medication studies. There is a consistent finding, though. It’s probably not so consistent that you could say higher risk with insulin; it is found with all co-medications.

Dr. Schernthaner: At the moment, we are confronted by many classes of anti-diabetic drugs. The newest is SGLT-2 inhibitors. Where do you see the advantages and disadvantages of the new class relative to DPP-4 inhibitors or GLP-1 receptor agonists?

Dr. Wilding: I think you are asking where these drugs will fit into the armamentarium. I think the barriers we discussed means that these drugs have potential benefits across the spectrum. I was involved in the dapagliflozin study that Dr. Seufert presented. Many had poor control despite being on other agents. With dapagliflozin we were able to achieve reductions in A1c, weight, and blood pressure. We were able to see a real advantage here. We could also see options in patients on or failing metformin, even quite early in the disease. We could use DPP-4 inhibitors across the spectrum as well, but in the empagliflozin study we see greater A1c effect and weight-loss effect – and patients like to lose weight. We could use GLP-1 receptor agonists, but those are injectable, and some patients are resistant to injectables.

Dr. Schernthaner: Dr. Van Gaal, today we have the term individualization. This is a dream word, but it is hard to describe the optimal combinations therapy for patients. Which should be used and which avoided in obese patients?

Dr. Van Gaal: The problem is already as difficult as that of obesity itself. In one or your slides you showed Dr. Zinman’s recommendation of early intervention with multiple agents. This has been done for years with blood pressure. I think that we should do this for people with type 2 diabetes, as well. In the armamentarium, we have metformin, DPP-4 inhibitors, GLP-1 receptor agonists, and SGLT-2 inhibitors, which have weight advantage or neutrality, minimal hypoglycemia, and good effects on other barriers as well. You showed a slide of how fast weight gain occurs in diabetes. There is an old study that shows that avoiding weight gain within the first year of therapy will have enormous positive effects even on CV outcomes. Early on, with the drugs available now, combination therapies can be an elegant way to treat diabetes with an eye toward weight.

Dr. Schernthaner: The drugs show a clear increase in genital infection. Jochen, do we have any data on people who are at higher risk of this side effect?

Dr. Seufert: Female gender poses an elevated risk, which we for clinicians have known for decades about genital infections. So far we don’t have strict cutoff parameters with which we can identify every patient with higher risk of genital infections. A history of preceding genital infections may be a risk factor. In females, post-menopausal state is a risk factor, this has been looked at carefully in all the SGLT-2  inhibitor studies. Clinical relevance also should be studied. In the general type 2 diabetes population,  many patients have genital infections and never go to the doctor with symptoms. But this has to be looked at carefully, of course.

Q: Now that phase 3 data for canagliflozin is out, are there any signs or signals of imbalances in breast and bladder cancer or hepatotoxicity such as were seen with dapagliflozin?

Dr. Schernthaner: This was a scientific session, so it wasn’t about canagliflozin specifically: all the SGLT-2 inhibitors were shown. We can’t comment on unpublished data. Probably you can go to the Friday EASD session: two of the four SGLT-2 presentations are about canagliflozin.

Dr. Schernthaner: A clear advantage of this class is that they are independent of beta-cell function. So probably they can be used at any time in disease, with one exception. John, could you talk about this?

Dr. Wilding: As glomerular filtration falls, so will the efficacy of SGLT-2 inhibition. Thus one group of patients we are challenged to treat effectively is those with significant renal impairment. I don’t believe this is likely to represent a safety issue, but obviously one wouldn’t want to use a drug in a patient for whom it wouldn’t work.

Dr. Schernthaner: However, a large data cohort from Kaiser Permanente was recently published on patients with age over 75. The n was about 100,000. The vast majority had renal function preserved, indicating that most people would be eligible for this new class of drug.

Dr. Wilding: Why do you think we can achieve good glycemic control in clinical trials, but it seems impossible in the real world?

Dr. Schernthaner: In clinical trials you include only “nice” patients: clever people with high adherence  that pass through other exclusion criteria – and you have a lot of study monitors. This is a selected group of patients and does not reflect the real-world situation, in my opinion. I think selection bias is important.

Dr. Wilding: I think it’s an important point to make. RCTs provide evidence to support what we do, but the reality of translation is quite difficult.

Dr. Van Gaal: Obesity treatment is the same way; the real world may be different from trials. Besides patient selection, the frequency of patient-provider contact likely contributes to good compliance as well, across all disease states. Every physician is dealing with this problem, I think.

Dr. Schernthaner: A lot of patients came to their study doctor after the study and said, please, can you improve me again in a study. We do a lot of short-term studies, six months or fewer, and long-term studies of five years or so. Adherence rates tend to be high for the shorter-term studies, but for five years, getting adherence is difficult. Ongoing CV outcomes studies are difficult for this reason.

Dr. Seufert: It’s also a matter of reimbursement. If you demand frequent visits in a real-world setting, most healthcare systems won’t reimburse.

Dr. Schernthaner: To refer back to an earlier question, the problem of imbalance of cancer can be seen due to the small number of patients in clinical trials. Observational studies are still important because they have a larger sample size and so eliminate false signals that  are due only to noise. They also include many patients with comorbidities; this can illuminate some risk factors that are seen only in particular populations or only after long exposure. I think we need both types of studies.

Dr. Wilding: I would agree with you, with a qualification. This is getting completely off-topic, but with the glargine cancer story, a lot of observational data suggested a risk increase, but ORIGIN showed otherwise. Selection bias can affect observational data: perhaps higher-risk patients are more likely to get certain types of drugs. Observational studies are hypothesis-generating but they don’t answer questions.

Dr. Schernthaner: Yes, in more-obese patients, you tend to use high-dose basal insulin and then arrive at these results. You have selection bias of course.


Corporate Symposium: Asking The Tough Questions in Type 2 Diabetes Treatment! (Sponsored by Boehringer Ingelheim / Lilly Diabetes)


The conversation moved to a newer drug category, SGLT-2 inhibitors, and specifically empagliflozin (BI/Lilly). The panel described the mechanism of action of SGLT-2 inhibitors and why they make an excellent addition to the choice of therapies. Important reasons include their complementarity to other therapies and weight loss. They believed that the risk of clinically important side effects for the class was low, but should be followed closely.

What is the Current Clinical Evidence for SGLT-2 Inhibitors?

John Gerich MD (University of Rochester, New York, USA)

Why do we Need Novel Therapies?

  • We don’t yet have the ideal diabetes drug. The ideal drug would have a robust A1c response, no hypoglycemia, no weight gain, complementary actions with other drugs, durability, low side effects, long term safety, simple administration, and added value (e.g. improved blood pressure, lipids, beta cell function, cardiovascular risk).
  • The kidney plays an important role in glucose metabolism, along with the liver. Renal glucose reabsorption is actually equivalent to the total glucose production by the liver and kidney. This amount is 75% higher in people with type 2 diabetes.
  • The kidney reabsorbs glucose and returns it to the blood, up to a maximum rate, after which glucose spills over into the urine. This maximum rate is known as Tmax. The amount of glucose reabsorption is increased in diabetes. There is evidence that extra glucose reabsorption has negative consequences for kidney physiology and function. Reabsorption is controlled by the SGLT-1 and SGLT-2 transporter proteins. In diabetes there is an increase in SGLT-2 (and GLUT2) transporter proteins.
  • SGLT-2 inhibition results in urinary glucose excretion and many advantageous effects. These include an A1c reduction better than metformin or SU, loss of calories, osmotic diuretic effect, no change in hypoglycemia, and an effect that is independent of beta cell function. In animal studies there is also evidence for renal protection. But they also cause a two to three fold increase in the frequency of genital infections. SGLT-2 inhibitors can also lower blood pressure and are lipid neutral.

Ele Ferrannini MD PhD (University of Pisa, Italy)

  • In the new class of SGLT2 inhibitors, the three leading contenders are dapagliflozin (BMS/AZ), canagliflozin (J&J/Mitsubishi Tanabe), and empagliflozin (BI/Lilly). However there are many others in development. Canagliflozin is a little different because it also slightly inhibits SGLT-1 as well as SGLT-2.
  • Dapagliflozin reduces A1c in low dose monotherapy. There is a 0.7-0.9% A1c reduction in combination with metformin, SU, pioglitazone, and insulin independent of the complementary drug. For canagliflozin, the dose response is less evident, but the A1c reduction is similar to sitagliptin  monotherapy.
  • In a 90-week trial of empagliflozin, A1c was reduced by 0.7% versus 0.5% for sitagliptin. There was about a three kg weight loss, whereas sitagliptin was roughly weight neutral. Canagliflozin added to metformin reduces body weight – around two to four kg over six to 12 months. This is superior to sitagliptin, which is closer to weight neutral.
  • The only distinctive adverse effect of the class is genital/urinary tract infections, but the absolute numbers are quite small. It will take some time to figure out if this is of a clinical significance. Otherwise, the safety and tolerability profile of empagliflozin is comparable with metformin, and that of empagliflozin plus metformin is comparable with sitagliptin plus metformin. In trials, dapagliflozin showed an imbalanced incidence in breast and bladder cancer – leading to what appears to be a delay in approval. This result was not noticed with other candidates. The balance of benefit to risk is still under investigation.
  • Significant clinical trials of SGLT2 inhibitors are underway to establish the benefits and also adverse effects. Trials of dapagliflozin, canagliflozin, and empagliflozin collectively have enrolled/are enrolling over 30,000 patients. A cardiovascular trial for empagliflozin is currently recruiting.

[Note: additional talks from this corporate symposium can be found in the “Incretin-Based Therapies” section of this report]


Corporate Symposium: Diabetes Care Today: Therapies on the Horizon (Sponsored by Lilly Diabetes)


Benjamin Field, MD, PhD (Imperial College London, London, United Kingdom)

Dr. Benjamin Field presented on the potential of oxyntomodulin (OXM) as a GLP-1/glucagon receptor dual-agonist. OXM reduces body weight in overweight and obese people, increases energy expenditure, and decreases food intake. In preclinical studies it has also been shown to lower blood glucose. Because glucagon receptor activation should be expected to increase blood sugar, Dr. Field’s team sought to explain how this “glucagon paradox” could exist. Aided by Day et al.’s work in demonstrating that GLP- 1/glucagon receptor co-activation at the right proportions produces emergent effects of increased weight loss and reductions in blood sugar, Dr. Field’s team found evidence that GLP-1 agonism opposes glucagon receptor agonism to ameliorate the acute hyperglycemic effect of glucagon. Dr. Field mentioned during Q&A that the OXM analog developed in his laboratory was licensed to Wyeth (which has since been bought by Pfizer) and no new data has been reported on the molecule though he believes it is in phase 2.

  • Dr. Field opened by presenting data demonstrating that elevated glucagon levels can be associated with blood glucose in the normal range. In 1970, when healthy volunteers were asked to fast for six weeks, glucagon levels rapidly rose and remained elevated for the duration of the fast, while blood glucose remained ~65 mg/dl (on the low end of normal; Marliss  et al., JCI 1970). He implied that other physiological counter mechanisms exist to maintain relatively normal blood glucose levels even in the presence of elevated glucagon, which is important if one tries to use glucagon for the treatment of diabetes, and in light of the fact that glucagon treatment diminishes appetite and causes weight loss (Schulman et al., Journal of Applied Physiology 1957).
  • Oxyntomodulin (OXM), like GLP-1, is a cleavage product released from the proglucagon peptide and is released after meals in proportion to energy intake, slows gastric emptying, and is elevated after Roux-en-Y gastric bypass. Subcutaneous OXM (400 nmol) reduces body weight in overweight and obese people (2 kg placebo-adjusted weight loss after four weeks). OXM increases energy expenditure and decreases food intake, so it is could be of great interest to develop as a diabetes or obesity treatment. It activates both the GLP-1 receptor and the glucagon receptor, which baffled Dr. Field’s team for a period of time because OXM was shown, preclinically, to produce both weight loss and reduced blood glucose, whereas one would expect a molecule that activates the glucagon receptor to result in increased blood glucose. Dr. Field’s team speculated at first that an undiscovered OXM receptor must exist to modulate this distinct effect, but in 2009 when Day et al. reported that GLP-1/glucagon  receptor co-activation at a certain ratio actually produced weight loss and reduced blood glucose  in mice (Day et al., Nat. Chem. Biol. 2009). The authors of this study speculated that the glucagon paradox (wherein glucagon activation reduced blood glucose) might exist because: 1) the  metabolic benefits of weight loss outweighed the diabetogenic effects of glucagon receptor agonism; 2) the activation of GLP-1 opposed or neutralized glucagon receptor stimulation; or 3) there was an unexpected beneficial metabolic effect of sustained glucagon receptor stimulation.
  • Dr. Field’s team found evidence for Day et al.’s second hypothesis, that GLP-1 agonism opposes glucagon receptor agonism to ameliorate the acute hyperglycemic effect of glucagon. He conducted a four-way randomized crossover study in ten overweight and obese people without diabetes where participants were given an infusion of glucagon (50 ng/kg/min), GLP-1 (0.8 pmol/kg/min), both, or placebo. Energy expenditure increased with glucagon and combination treatment, as expected. Blood glucose rose acutely with glucagon treatment, as expected, and this was associated with a rise in plasma insulin. The combination treatment resulted in a much larger increase in plasma insulin, which resulted in an initial rise in blood glucose that was partially ameliorated by the rise in insulin. Therefore, Dr. Field believes   the mechanism might involve increased energy expenditure due to futile metabolic cycling of glucose to glycogen. The initial glucose spike, he believes, may have been avoided by chronic glycogen depletion – which might be achieved through degradation-resistant OXM analogs.

Questions and Answers

Q: Is the ratio of glucagon to GLP-1 in these studies skewed in favor of glucagon?

A: In the infusion study I showed, we picked a glucagon dose essentially just sub-nausea threshold to get a measurable effect on energy expenditure in a few volunteers. I don’t know whether that acute situation would be the same type of ratio that would be required for long-term therapy for type 2 diabetes and obesity.

Q: Does OXM actually have any influence on insulin secretion, and does it have CV effects? Can you elaborate on the GLP-1 effect of increasing heart rate?

A: In terms of insulin secretion, the few studies available say that plasma insulin is elevated without much effect on plasma glucose in an acute or sub-acute setting. When one achieves profound weight loss, then  all bets are off. In terms of CV effects, there are nice data in mice showing that high dose OXM will cause an increase in heart rate. That is consistent with our use of glucagon as an antidote to beta-blocker poisoning as it causes an increase in heart rate.

Q: When you put two of these agents together, do you get unforeseen effects such as tiredness or thirst?

A: Thirst is interesting because GLP-1 is likely naturally a uretic. We haven’t done long-term studies, and we really need these degradation resistant analogs to do this. The last I knew, the molecule my lab had been working with was in phase 2, but I haven’t seen any clinical data; I can’t wait to though. [Editor’s note: Dr. Field later elaborated that this molecule was sold to Wyeth, which has since been bought by Pfizer, so it has undergone a few code changes, and we would presume that with all of these ownership changes, it has likely not been prioritized.]



Mathias Tschop, MD (Helmholtz Center, Munich, Germany)

Dr. Matthias Tschop reviewed several fascinating potential peptide and hormone combination   therapies. Several dual- and tri-agonists have shown beneficial synergistic effects with regard to weight loss and blood glucose control in pre-clinical trials. With a GLP-1/glucagon receptor dual-agonist that activates both receptors in the right proportion in rodents, the combined effect is a net reduction in   blood glucose and greater weight loss than either one produces alone. GLP-1/GIP dual-agonists and GLP-1/GIP/glucagon receptor tri-agonists also produced synergistic effects in rodents. Dr. Tschop quickly mentioned that the first human data for a GLP-1/GIP dual-agonist from Roche did not demonstrate such promising results – if this is true, this would be the first we have heard about results from the phase 1 trial for RG7685 (Roche’s GLP-1/GIP dual-agonist, formerly Marcadia’s MAR701). Finally, Dr. Tschop discussed how the hybrid-peptide approach could have applications beyond creating synergies. One molecule could be used to localize delivery of another if selective, rather than systemic, delivery is desired. As examples, he reported that GLP-1/estrogen or GLP-1/dexamethasone conjugates could produce weight loss in mice by targeting only cells that express the GLP-1 receptor.

  • GLP-1/glucagon receptor dual-agonists: since the GLP-1 peptide has substantial structural similarities to the glucagon peptide, a single analogue molecule can be designed that activates both receptors. By itself, activating the GLP-1 receptor causes weight loss and decreases blood sugar. Additionally, solely activating the glucagon receptor activation causes weight loss (due to appetite suppression and increased energy expenditure) and an increase in blood glucose. Remarkably, though, when both the GLP-1 and glucagon receptors are activated in the right proportion in rodents, the combined effect is a net reduction in blood glucose and greater weight loss than either one produces alone (Day et al., Nat. Chem. Biol. 2009). Dr. Tschop’s group developed an analog peptide that binds both the GLP-1 and glucagon receptors, with the right affinity so that the proper proportion of GLP-1 and glucagon receptors are activated together to produce this effect in mice. Dr. Tschop explained that chronic glucagon receptor activation increases plasma FGF21 expression and that this has been demonstrated in healthy humans. In mice, FGF21 is necessary for weight loss mediated through glucagon receptor activation.
  • GLP-1/GIP dual-agonists: In obese rodents, the GLP-1/GIP dual-agonist exhibits synergistic effects. At doses where GIP alone has no effect, the combination with GLP-1 produces greater weight loss than GLP-1 alone in mice. The GIP/GLP-1 dual-agonist has also been tested in non- human primates; it resulted in greater insulin induction in response to dextrose challenge than liraglutide alone and also reduced blood sugar more than mono-agonism. Dr. Tschop explained that using a co-agonist produces a greater effect than simply increasing the dosage of one agonist alone because at some point, if you only target one receptor, it becomes saturated and adding additional drug will not result in additional activation of the receptor. Dr. Tschop quickly mentioned that the first human data for a GLP-1/GIP dual-agonist from Roche did not demonstrate such promising results – if this is indeed true, this would be the first information we learned about results from the phase 1 trial for RG7685 (Roche’s GLP-1/GIP dual-agonist, formerly Marcadia’s MAR701). The phase 1 trial was supposed to have finished about one year ago, but as far as we know, Roche has not released results [Editor’s Note: we recently learned that MAR701/RG7685 has been dropped from phase 1 development, and a new GLP-1/GIP dual agonist candidate, RG7697, also developed by Marcadia, has taken its place; for more details please see our Roche 3Q12 report at]. After the phase 1 trial ended in 4Q11, Roche announced preparation of a “follow-up” study instead of initiating phase 2 trials, and we have not heard more since.
  • Dr. Tschop then asked, “Why stop with two?” and discussed a GLP-1/GIP/glucagon tri-agonist. He believes that the hormone network is strongly based on patterns, so manipulating just one or two may not be enough. Dr. Tschop’s group developed a tri-agonist analog that could be delivered subcutaneously in diet-induced obese mice. Mice receiving the tri- agonist lost 30% more weight than mice receiving placebo over one six weeks. While, mice receiving just the GLP-1/GIP dual-agonist lost ~20% more weight than the placebo group.
  • This hybrid-peptide approach could have other applications beyond creating synergies – Dr. Tschop discussed using molecule to localize delivery of another. A GLP-1/estrogen conjugate could be used to selectively deliver estrogen to cells expressing the GLP-1 receptor. Dr. Tschop stated that estrogen may have many favorable metabolic effects but has not been utilized in the treatment of metabolic disorders because it is dangerous to deliver systemically. A GLP-1/estrogen conjugate that only falls apart inside a cell after the GLP-1 binds to its receptor and is internalized, reduced body weight in mice by about 25% without carcinogenic effects or affecting uterus size (whereas a control that allowed estrogen to circulate systemically considerably enlarged the uterus). Another example he discussed of using hybrid peptides to localize delivery was to use a GLP-1/dexamethasone conjugate (dexamethasone [dexa] is an anti-inflammatory glucocorticoid) to target delivery of dexa to the brain in order to reduce inflammation in the hypothalamus that may be related to diet-induced obesity. Dexa is not known to reduce body weight, but in mice the GLP-1/dexa compound decreased body weight more than GLP-1 alone did.

Questions and Answers

Q: When you combine peptides, what effect does it have on the side-effect profile, like gastric emptying, blood glucose, and insulin resistance? What about micro- or macrovascular   complications?

A: We haven’t studied complications because the mouse isn’t a good model for that. Regarding insulin resistance, nothing I’ve studied did bad things to insulin sensitivity and always majorly improved it. We haven’t seen any side effects indicative of illness. I’m not a fan of using mice to evaluate side effects  because many false positives and negatives come up. But when we think of synergies in benefits, then naturally the question about synergies in side effects arises. Those don’t seem to be synergizing that much, and we believe that has to do with the fact that we are at a lower therapeutic window.



Bo Angelin, MD, PhD (Karolinska Institute, Hagalund, Sweden)

Dr. Bo Angelin discussed preclinical and clinical observations about the promise of FGF-21 for type 2 diabetes. Preclinical pharmacology studies demonstrated that FGF-21 had a durable glucose and lipid lowering effect, ameliorated insulin resistance, improved beta cell function and mass, reduced weight, and reversed hepatosteatosis without hypoglycemia, fluid retention, pancreatitis, liver toxicity, or mitogenicity. Observations in humans are not completely consistent; the 10-year follow up to the Stockholm DPP study found that patients with elevated FGF-21 levels in prediabetes (those in the upper quartile) had a four times greater likelihood of progressing to diabetes. He proposed that high levels of FGF-21 may be a new indicator of risk for type 2 diabetes, or that FGF-21 might have been induced as a defense  mechanism.

  • FGF-21 is a circulating fibroblast growth factor involved in lipid and glucose homeostasis via the activation of PPAR-α. It is produced mainly in the liver, white adipose tissue, and the pancreas, and plays a role in lowering triglycerides and blood glucose, as well as clearing excess fat from the liver. FGF-21 is abundantly available in circulating plasma, but FGF- 21 receptor activation depends on the rate-limiting availability of beta klotho.
  • In preclinical studies, FGF-21 administration in animals had positive effects on insulin sensitivity, obesity, diabetes, and dyslipidemia. In ob/ob mice, FGF-21 produced a durable glucose and triglyceride lowering effect with no hypoglycemia and no effect on adipose tissue in normal mice. In db/db mice, FGF-21 produced islet/beta cell preserving effects. In diet- induced obese mice, FGF-21 increased energy expenditure and weight loss in a dose-dependent manner over 14 days. Finally, in a proof of concept study in diabetic rhesus monkeys, 30 µg/kg, 100 µg/kg, and 300 µg/kg of FGF-21 lowered LDL cholesterol, raised HDL cholesterol, and increased levels of adiponectin and ApoA1 (a major component of HDL). Little evidence for adverse side effects arose from preclinical studies – FGF-21 did not cause hypoglycemia, fluid retention, pancreatitis, liver toxicity, or mitogenicity (the development of a related molecule, FGF-19, was previously stopped due to mitogenicity concerns).
  • Observations regarding FGF-21 in humans suggest that the human response might differ from the animal response. Fasting plasma levels of FGF-21 vary widely (anywhere  from 20 pg/ml to 80,000 pg/ml), so variations in human levels of FGF-21 might not be physiologically relevant. FGF-21 levels have been found to increase moderately with starvation  and with PPAR-α activation. In some studies, elevated serum FGF-21 has been observed in people with metabolic disorders including impaired glucose tolerance, insulin resistance, type 2 diabetes, abdominal obesity, hypertriglyceridemia, and fatty liver disease; however, the levels of FGF-21 found in these patient groups were still within the normal range. Dr. Angelin reported that the 10- year follow up to the Stockholm DPP study found that prediabetes patients with elevated FGF-21 levels (those in the upper quartile) had a four times greater likelihood of progressing to diabetes. He proposed that high levels of FGF-21 may be a new indicator of risk, or that FGF-21 might be induced as a protective response.

Questions and Answers

Q: What is the incidence of cancer?

A: FGF-21 seems, for some reason, not to be involved in cancers.

Q: Does physical activity increase FGF-21?

A: There are studies that see a slight increase. But I think the problem with these new things coming up is that you find a lot of studies supporting the thought of elevated FGF-21 being beneficial, but the changes at these levels would not have any physiologic relevance.

Q: What is the situation in pure hypertriglyceridemia patients without diabetes?

A: What we report in our first paper was an increase in FGF-21 of about 50%, and what we’re looking at now is whether there are subgroups in which this could be relevant to pathogenesis. I would just add hypertriglyceridemia as one of a group of situations in which we have FGF-21, which would probably indicate they have other aspects of metabolic syndrome.



Bernard Thorens, PhD (University of Lausanne, Lausanne, Switzerland)

Dr. Bernard Thorens first reviewed preclinical data supporting the promise of glucagon receptor antagonists for the treatment of diabetes. Mice lacking the glucagon receptor have reduced glucose  levels and normal glucose tolerance. Suppression of the receptor has also been found to induce hyperproliferation of alpha cells and higher glucagon levels. Interestingly, STZ mice (a model for type 1 diabetes) lacking glucagon receptor expression show no increase in hyperglycemia compared to control mice (whereas STZ mice with intact glucagon receptor expression experience substantially greater hyperglycemia than control mice), suggesting that glucagon antagonism could also be used to slow type 1 diabetes. Clinical data for Lilly’s glucagon receptor antagonist LY2409021 showed that the compound reduced fasting and post prandial glucose, and reduced A1c by 1% after 28 days of treatment (for full details on this study, please see page 11 of the Novel Drugs section of our ADA 2011 report at

Questions and Answers

Q: What about the dangers of hypoglycemia when you’ve blocked the action of glucagon?

A: In these tests and in glucagon receptor knockout animals, even if they fast for 24 hours, they don’t develop hypoglycemia. It doesn’t seem to be a major risk. Even patients who have been identified with mutations in the glucagon receptor and have no glucagon receptor action have normal glycemic control. It is surprising, but I think with the evidence from animal models and human mutants, it doesn’t appear to be a major problem so far.

Q: How would patients recover if they developed hypoglycemia?

A: Catecholamines might play a role in the counter response. Epinephrine and norepinephrine don’t seem to be very different in knockout mice or mutant humans, so maybe that is another system that is not completely understood that helps prevent hypoglycemia.

Q: What are the changes in liver enzymes with glucagon receptor antagonists?

A: There is a slight increase in liver enzymes but I can’t comment much more on humans because there are so few data.


Symposium: Diabet-OMICS: Diabetes Research in the OMICS Millennium


Joseph Nadeau, PhD (Institute for Systems Biology, Seattle, WA)

Dr. Nadeau used evidence from rodent studies to illustrate hypotheses about why the risk of the metabolic syndrome has been so hard to localize to specific genes, even though this risk clearly runs in families. Heritability is often much more complex than simple Mendelian genetics, Dr. Nadeau explained. Sometimes many different genes can have the same effect (e.g., protection against weight gain), but someone with any one of these genes would get just as much benefit as someone who had multiple protective genes (i.e., “genetic buffering” makes for effects that are non-additive). Another wrinkle is that the role of an individual gene might depend on the context of the whole genome, so that the same gene might have different effects in people with different genetic background. Finally, Dr. Nadeau discussed epigenetics, heritable variations in the way that genes are processed. His group conducted mouse studies of epigenetic changes that affect eating behaviors; they found that the way an individual eats might depend on the environment that its grandfather was exposed to. We are glad to see promise for new therapeutic targets beyond the genome itself, but this field seems so complex that we could see decades go by before we understand systems biology well enough to apply it in diabetes.



Miriam Cnop, MD, PhD (ULB, Brussels, Belgium)

In this cutting-edge talk, Dr. Cnop discussed in vitro type 2 diabetes research that compared human islet cells to other tissues, in the hopes of identifying mechanisms specific to islet-cell dysfunction. Notable differences were seen in the islet cells’ methylome (which refers to an important modification pattern  that can vary in different body tissues). The most notable modifications in islet cells were localized to  254 genes, some of which have never before been strongly implicated in type 2 diabetes; Dr. Cnop and her colleagues are conducting follow-up studies to characterize the methylome in type 2 diabetes still further. She also said that research on the islet cell transcriptome – which refers to the genes that get expressed in a particular tissue – has shed light on how palmitate (a kind of free fatty acid) and cytokines can each negatively affect islet-cell gene expression.


Symposium: Rising Star Symposium


Henrike Sell, PhD (German Diabetes Center, Dusseldorf, Germany)

Dr. Henrike Sell gave us a taste of the complicated world of “crosstalk” between the adipose tissue and muscles, a communication that seems to involve adipokines (hormones released from the adipose tissue). She reviewed how one of the better-known adipokines, adiponectin, appears to exert protective effects on insulin sensitivity. Getting into uncharted waters, her group has used proteomic analysis to identify 44 novel adipokines. This list turned out to include DPP-4, levels of which are elevated in type 2 diabetes. Dr. Sell said that DPP-4 seems to be important not only because it breaks down GLP-1, the mechanism most commonly cited to explain the efficacy of DPP-4 inhibitors. Indeed, DPP-4 also seems to have a direct deleterious effect on insulin signaling – an intriguing prospect that Dr. Sell and her colleagues are currently investigating further.

Questions and Answers

Q: Have you tried sticking DPP-4 onto your adipocytes to see what happens to the inflammatory  profile?

A: Yes. We have seen that DPP-4 is working on the adipocytes, and it affects adipokine secretion, but the data are not yet well developed.


Sixth Albert Renold Lecture


Decio Eizirik, MD, PhD (Universitet Libre de Bruxelles, Brussels, Belgium)

Dr. Decio Eizirik gave this year’s Albert Renold Lecture to a packed hall. In his presentation, Dr. Eizirik reviewed his and his colleagues’ basic science research into the signaling cascade behind mitochondria- mediated beta cell apoptosis. Dr. Eizirik’s work has helped provide further details on the cross talk between individuals’ beta cells and the adaptive and innate immune systems that trigger beta cell destruction.


5. Obesity and Prediabetes

Oral Presentations: Impact of Bariatric Surgery


Dimitrios Pournaras, MD (Hospital Oswaldo Cruz, Sao Paulo, Brazil)

Dr. Dimitrios Pournaras presented results from a study assessing the use of duodenal-jejunal bypass liner (DJBL; GI Dynamics’ EndoBarrier) in 16 patients with type 2 diabetes and BMI between 23 and 36 kg/m2. As a reminder, the DJBL is a flexible, tube-shaped liner that excludes the duodenum and proximal jejunum from nutrient flow. BMI, A1c, glucose levels, and insulin sensitivity and secretion   were assessed preoperatively and one week, 12 weeks, and 52 weeks postoperatively. Results suggested that BMI, A1c, and insulin sensitivity significantly improved after the operation and significant improvements sustained through 52 weeks. Insulin secretion and insulinogenic index, however, remained unchanged following device implementation. Dr. Pournaras concluded that these findings suggest combination therapy of DJBL and GLP-1 could mimic insulin sensitivity and secretion changes observed in gastric bypass in patients with low BMI. This could introduce a new therapeutic option for patients with type 2 diabetes who either do not qualify for or would not consider weight loss surgery, and introduces a real possibility for GI Dynamics to consider an additional indication for the EndoBarrier in patients who do are not obese.

  • Importantly, this study investigated DJBL in patients with lower BMI (23-36 kg/m2) compared to previous DJBL studies. At baseline, the 16 patients were 35-61 years old; had type 2 diabetes for 2-10 years; had an A1c between 7.5 and 10.2%; were on metformin; and were not on an insulin, a GLP-1 analog, or a DPP-4 inhibitor. Given gastric bypass increases postprandial insulin and GLP-1 response, and DJBL effects on insulin secretion was not yet established, patients were not allowed to use insulin or incretin medications due to concerns  about  hypoglycemia.
  • Patients’ BMI decreased significantly, but not to the level observed in gastric bypass. Dr. Pournaras said, a drop of gastric bypass’ magnitude is not desired considering the patients’ lower baseline BMI. In the first week post-implantation, weight loss was insignificant, averaging ~1 kg (~2.2 lbs.). The decrease in BMI, however, was significant at 12 weeks and was significantly sustained at 52 weeks (p <0.001).
  • At 52 weeks, A1c averaged 7.5% – a significant decrease from the baseline of 8.6% (p <0.001). Additionally, 62.5% of participants achieved an A1c <7%. During Q&A, the question   was raised as to whether A1c decreases would persist after the DJBL was removed. Dr. Pournaras said that some pilot data suggest that A1c will increase, but not to pre-DJBL levels, saying “These are studies we need to do”. Additionally, we would be interested in investigations of the safety and efficacy of repeat procedures.
  • Glucose levels significantly decreased one week after DJBL implementation (p<0.001) and were sustained for 52 weeks (p <0.01). Glucose area under the curve decreased and fasting glucose similarly decreased from pre-DJBL levels (p <0.001 at one week and p <0.01 at 52 weeks).
  • Improvements in insulin sensitivity, as measured by HOMA-R, were significant   after one week and persisted through 52 weeks. As Dr. Pournaras pointed out, HOMA-R is not the gold standard for measuring insulin sensitivity. He said, however, that given the evidence for improved insulin sensitivity in this study, future studies of DJBL would be adequately   powered and appropriately designed to investigate changes in insulin sensitivity.
  • Insulin secretion rate and insulinogenic index remained unchanged over the study period, and no changes in fasting insulin or insulin area under the curve were observed. (C- peptide levels changed significantly at one week, but were not sustained.) The lack of effect on insulin secretion contrasts the enhanced post prandial insulin observed with gastric bypass, making DJBL and GLP-1 combination therapy a possibility, according to Dr. Pournaras. If combination therapy proves effective, it could potentially mimic the effects of gastric bypass, but be available to patients who do not qualify for bariatric surgery or who consider bariatric surgery too extreme.
  • For context, GI Dynamics’ EndoBarrier received CE Mark approval in 2010 and approval from the Therapeutic Goods Administration in Australia in 2011 for the treatment of type 2 diabetes in obese patients for up to 12 months. Additionally, in August of this year, GI Dynamics announced conditional approval to begin a US pivotal clinical trial of the EndoBarrier for the same subpopulation. For additional discussion on the EndoBarrier and the FDA decision, see our GI Dynamics 2Q12 report at We think Dr. Pournaras’ study may introduce a real possibility for GI Dynamics to consider an additional indication for the EndoBarrier in patients who do are not obese – it will be interesting to hear ruminations on this going forward as more data emerges.

Questions and Answers

Q: What metabolic surgeries are available to help people with diabetes who have a normal weight?

A: That is the target group of this study – those who are not morbidly obese. These patients did not quality for weight loss surgery. The door is open for bigger, longer-term studies in exactly this group of patients.

Comment: The change of calorie balance after bariatric surgery is the elephant in the room. Yes, I think negative calorie balance explains early changes.

A: Reduced food intake is important. Because patients are not morbidly obese, the reduction of food intake was minimal. This is not the case for gastric bypass, and I appreciate that.

Q: Can you comment on potential side effects and disadvantages of DJBL?

A: Three patients experienced nausea and abdominal bloating.

Q: It is well known that patients with diabetes in a study will improve A1c initially and reverse these changes over a year or two – did you have a control group?

A: We had a control group of patients in exactly the same center who had no difference in A1c, but who had same appointments and follow up from a diabetology point-of-view. If we showed differences in A1c at 12 months we may experience a legacy effect for years to come, this is where studies need to go

Q: Your patients were on metformin only? Or were there other medications at baseline?

A: All patients were on metformin and they did stop that; four were on sulfonylureas and two stopped that.

Q: So after the procedure they still needed metformin?

A: My view on this is that metformin is safe, not expensive, and if it offers a good option. I wouldn’t stop it unless necessary. Our policy was not to stop the metformin unless needed.

Q: How long can you keep the DJBL, and what happens when you remove it?

A: 12 months.

Q: And when you remove it?

A: We have some pilot data that we have not fully analyzed, and will present next year. The data shows that when you remove the device A1c does not go back to what you would expect. These are studies we need to do.

Q: What are the metabolic side effects of this device?

A: With gastric bypass there is no malabsorption of calories, but there is malabsorption of vitamins. This device is only in for 12 months. Metabolic side effects with repeat treatment is something we would need to explore, but we have not had the vitamin deficiencies that have been reported with gastric bypass.



Sarah Steven, MBChB (Newcastle University, Newcastle upon Tyne, UK)

Dr. Steven presented a retrospective analysis of 73 patients with type 2 diabetes (mean baseline BMI: 51 kg/m2, mean baseline A1c: 7.7%, and median diabetes duration of five years) that received Roux-en-Y gastric bypass between 2009 and 2011. The study examined the remission of diabetes depending on diabetes duration and degree of weight loss. Diabetes remission (post-op A1c <6.1%) occurred in 80% of patients with shorter duration diabetes (less than four years) compared to 38% in those with long duration diabetes (greater than eight years). Remission was also higher in those who lost more weight: 53% remission for a loss of 10 kg/m2, 73% for 10-15 kg/m2, and 76% in individuals losing >15 kg/m2. Dr. Steven emphasized that it was the degree of weight loss that was the main determinant of diabetes remission. She also highlighted that individuals with long duration disease required greater weight loss in order to reverse type 2 diabetes, but it’s still encouraging that long disease duration is not a bar to reversing type 2 diabetes. Given these results, she hypothesized that beta cell function can be regained at any stage of type 2 diabetes, perhaps upon removal of the toxic environment (e.g., saturated fatty   acids). The study did not measure insulin resistance, sensitivity, or calorie intake, though Dr. Steven noted that an upcoming prospective study will include more data on these factors.

Questions and Answers

Q: How long after the operation did patients have post op assessments?

A: Due to the retrospective nature of the data set, this was not defined, but it was a minimum of three months.

Q: Looking at the medication of subjects, you would expect that patients on insulin therapy would have more profound insulin deficiency and would be less prone to reversal. Your patients seem nicely compensated – were there not many on insulin?

A: There was no difference according to medications. Most patients remaining on medications just stayed on metformin.

Q: How many patients were on insulin?

A: There were 12 patients on insulin therapy.

Q: Steve Kahn had a talk yesterday on beta cell dysfunction and questioned whether it was death or just dysfunction. And is it ectopic fat in pancreas that causes it. You had 12 patients on insulin. What was the range of doses and did all come off insulin?

A: I cannot tell you the exact doses, but they were on considerable amounts of insulin.

Q: It could valuable to have some other measure or indicator or insulin resistance besides weight and BMI. You haven’t measured this.

A: This was a retrospective data collection. We’re doing a prospective study looking at insulin resistance using hyperinsulinemic clamps.

Q: All your patients were very fat. Do you have any information on the amount of food they were eating? Is it food eating or weight loss that is the determining factor.

A: This is an important point. Calorie restriction is very important in the mechanism for reversal of diabetes in these patients. To maintain the BMI of these patients, they must be consuming 3,200 calories. In the first few days of gastric bypass, intake is in the range of 400-600 calories. We are going to be looking at that in a prospective study.

Q: Did you look at the data the other way around? In other words, for the same degree of weight loss, was there anything that predicted remission of diabetes?

A: I haven’t looked for this in the data set. The degree of achieved weight loss is important – people that haven’t gained as much weight since diagnosis are more likely to be reversed.

Q: Did the duration of resolution link to weight loss?

A: Data so far from other long-term studies ten years out suggests that individuals who manage to keep their weight down and don’t regain weight maintain their resolution of diabetes.

Q: Did you do an analysis to tease out which factor contributed. In other words, putting all in the same model. Were they independent predictors?

A: We didn’t do that. Mainly because of the numbers and retrospective nature of the data.


Oral Presentations: New Modulators of Energy Expenditure


Henry Karlsson, PhD (University of Turku, Turku, Finland)

Dr. Henry Karlsson began his interesting presentation by explaining that people who are obese and people with drug addictions have similar neural circuitry, as consumption of both drugs and food activate the brain’s reward systems. Overeating likely results from a misbalance between the reward circuits activated by food intake and the networks that inhibit reward seeking. Dr. Karlsson then   focused on his research, which looks at the endogenous opioid system’s role in overeating, sharing how µ-opioid receptors mediate feelings of reward. Comparing position emission tomography (PET) scans of 17 morbidly obese women (mean BMI of 41.9 kg/m2) and nine non-obese women (mean BMI of 23.6 kg/m2), Dr. Karlsson found that obese individuals, on average, had 37% lower µ-opioid receptor binding potential within neural reward circuits compared to the non-obese women (p <0.05 in the ventral caudate nucleus and dorsal caudate nucleus). Dr. Karlsson suggested that targeting the brain mechanisms involved in overeating, such as the low µ-opioid tone in obese people, could be a promising approach for new pharmacological treatments for obesity.

Questions and Answers

Q: Can you tell us about the interaction between the µ-opioid system and the dopaminergic system since the dopaminergic system also plays a role in the reward system?

A: I am not very familiar with the dopaminergic system, but I am looking to learn more about it, so hopefully I can answer that question later on.

Q: Do you know what the genetic status for the µ-opioid receptor was in the women you tested?

A: No, we are now looking to do that.

Q: Would you get different results depending on if a person was hungry or full during the PET study?

A: That is not as much of an issue for PET studies as it is for fMRI studies. We do however try to keep the situation similar between participants, in terms of hunger level.


Symposium: Neuroendocrine Control of Glucose Homeostasis


Matthias Tschöp, MD (Technical University of Munich, Munich, Germany)

Dr. Matthias Tschöp argued that it is time to abandon the approach of trying to treat obesity and/or type 2 diabetes using only a single pathway in the central nervous system (CNS). He presented data from obese mouse models showing that a combination therapy of a GLP-1 agonist and leptin is more powerful then either agent alone. He reminded the audience that leptin alone typically does not cause weight loss, because obesity is associated with leptin resistance. However, when given to obese mice in conjunction with a GLP-1 agonist, he found that the GLP-1 reduced the mouse’s weight and corresponding leptin resistance enough for the leptin to become effective. As a result, mice administered this combination lost significantly more weight over 20 days (~40% of baseline weight) then those that received only a GLP-1 (~30% of baseline). Similarly, when he administered fibroblast growth factor 21 (FGF21, which he said has good impacts on body weight and glucose metabolism even on its own) with leptin he found that the mice lost ~40% of baseline weight (vs. ~30% in those only receiving FGF21).



Jens Brüning, MD (University of Cologne, Cologne, Germany)

Dr. Jens Brüning presented interesting results from animal studies suggesting that insulin and leptin action in a region of the hypothalamus (the arcuate nucleus [ARC]) controls not only food intake (what the hypothalamus is frequently associated with) but also locomotor activity and hepatic glucose production. Thus, this region of the brain may have a larger role in the pathogenesis of type 2 diabetes then was thought. Additionally, upon looking at the impact of insulin on different regions of the brain, Dr. Brüning concluded that obesity causes insulin resistance in regions of the hypothalamus close to the blood-brain barrier (such as ARC) where inflammatory signals due to obesity could potentially leak through, but increased insulin signaling (due to hyperinsulinemia) in other regions (such as the ventromedial hypothalamus and lateral hypothalamus). These effects appear to synergize, impairing energy and glucose homeostasis.

Questions and Answers

Q: Does every neuron in the brain express both insulin and leptin receptor?

A: The insulin receptor is pretty much in every neuron, but to varying degrees. The leptin receptor, however, has distinct regions of expression.

Q: Do you think selective insulin resistance could have anything to do with circulating insulin levels in the brain?

A: The nature of the ARC is that the blood brain barrier is very leaky in that region. The leaky blood  barrier makes those neurons particularly sensitive to signals spilling over from the barrier, while other areas are better protected from these signals and are protected from insulin resistance. I think differential accessibility across the blood brain barrier is probably one of the most logical explanations for what we are seeing.


Symposium: Gut and Tissue Microbiome in the Development of Metabolic Diseases


Oluf Pedersen, MD (University of Copenhagen, Copenhagen, Denmark)

Dr. Oluf Pedersen’s presentation explored the relationship between the gut microbiome and common metabolic disorders. He began his exploration with observational studies, which suggested that obese individuals have reduced bacterial diversity, a depletion of Bacteroidetes, and an increase in Actinobacteria. Additionally, metabolically unhealthy individuals tend to have a low gut bacterial gene count (LGC; defined by <380 K-genes) compared to high bacterial gene count people (HGC). The LGC group had a higher prevalence of five “potentially” pro-inflammatory bacteria than the HGC group, while the HGC group had a higher prevalence of four anti-inflammatory bacteria. Perhaps, posed Dr. Pederson, the gut metabolic potential of LGC is health damaging whereas that of HGC is health sustaining. Dr. Pederson went on to hypothesize that gut microbiota could be used to predict and distinguish between lean and obese individuals. Maybe, gut microbiota could even be a clinically useful classification tool; Dr. Pederson explained that people with type 2 diabetes have gut microbiota imbalances as well (in terms of both microbiota composition and function). While it is doubtful patients would prefer a microbiota test (which would invariably involve fecal testing) over OGTT or A1c testing, the point of the test would appear to go beyond what an A1c or OGTT would offer. An important remaining question in Dr. Pederson’s exploration is, of course, whether microbiota dysbiosis (imbalances) in obesity and diabetes is a cause or consequence of the disease, or perhaps these are not be mutually exclusive.

Questions and Answers

Q: You suggested a clinical consequence in the future, but if you take stool samples an hour later versus a day later, what are the effects?

A: A crucial comment. We are working hard in the scientific community to define a standard procedure for sampling, because if you don’t do it in a standardized way, you have major problems. We really need international  consensus.

Q: In your studies, were stools frozen after sampling?

A: Yes, one hour after.

Q: Have you had a chance to look at unusual patients with type 2 diabetes and compare them to common variety type 2 diabetes patients?

A: We haven’t.



Rémy Burcelin, PhD (Institute of Metabolic and Cardiovascular Disease, Toulouse, France)

Dr. Rémy Burcelin’s presentation explored the effects of the metagenome, specifically the genome from humans’ intestinal microbiota, that could potentially be involved diabetes. His presentation served as an important reminder that the gut microbiota itself is another organ that should be considered when investigating the pathophysiology of the disease. Potentially, he said, the microbiota could be another target to prevent diabetes.


6. Complications and Cancer

Oral Presentations: The Importance of Glycemic Control: Results from Large Scale Studies


Trevor Orchard, M.B.B.Ch., M.Med.Sci. (University of Pittsburgh, Pittsburgh, PA)

The highly regarded epidemiologist Dr. Trevor Orchard told the audience that the life expectancy for people with childhood onset type 1 diabetes has improved greatly in the past several years and is now only four years less than that of the general population. He had originally presented this finding to much acclaim at ADA in 2011. Similarly, glycemic control has improved, and frequency of renal disease has declined, but the incidence of coronary artery disease (CAD) has not fallen proportionally. Thus, Dr. Orchard hypothesized that there are risk factors for CAD in those with “better” glycemic control that are not being addressed. To study this, Dr. Orchard followed people with type 1 diabetes but no CAD for 18 years, at which point he defined them as having either had consistently high or low glycemic control. He found that people with consistently low glycemic control had a significantly higher cumulative incidence of CAD (31.9%) than those with high glycemic control (20.8%, p=0.03). Dr. Orchard reported that lipids, blood pressure, and estimated glomerular filtration rate (eGFR) were significantly associated with CAD in both glycemic control categories. Controlling for other factors, Dr. Orchard found that beyond diabetes duration, only hypertension was significantly correlated with CAD in the consistently high-A1c group (Hazard Ratio [HR]=3.6; p=0.001). However, in the consistently low group, low eGFR (defined   as <60 ml/min/1.73m2; HR=1.6; p=0.006), in addition to hypertension (HR=2.4; p=0.04), was a predictor of CAD. Thus, renal function appears to be relatively more important among people who have better controlled A1cs than those with high A1cs. Dr. Orchard said this is consistent with previous findings suggesting that baseline A1c increases CAD risk in people without (but not in those with) microalbuminuria (which is associated with low GFR). To us, this indicates that HCPs may still need to put additional focus on the cardiovascular health of their patients with well-controlled type 1 diabetes, especially those with signs of diabetic nephropathy. Of course, patients who have well-controlled type 1 are less likely overall to have signs of kidney disease, but the importance of ongoing assessment of eGFR appears to be quite important according to Dr. Orchard. We believe that many patients do not have any idea what their eGFR is. We will be looking into this in more detail with the dQ&A panel to try and  assess awareness of eGFR and kidney disease more broadly.

  • Incident cardiovascular artery disease (CAD; n=82) was defined as the first instance of CAD death, myocardial infarction, stenosis ≥50%, revascularization, ischemic  electrocardiography (ECG), or Pittsburgh Epidemiology of Diabetes Complications (EDC) study physician defined angina. Dr. Orchard and his team obtained medical records to confirm participant reported CAD events.
  • Participants’ glycemic control was defined as being consistently high (n=157) or low (n=154). Those considered consistently high had an A1c in the highest tertile at all exams, while those classified as consistently low had an A1c in the lowest tertile at all exams. A1c was measured at baseline and after two, four, six, eight, 10, and 18 years. Tertile cut-of points were 8.0% and 9.2% at baseline and 7.0% and 8.2% at the 18-year follow up.

Questions and Answers

Q: You have taken a very tight way of characterizing people’s glycemic control. Could you comment on that?

A: I agree. We did study the middle group and it was pretty much consistent with the findings from the high A1c group. .

Q: We see more, and more overweight type 1 diabetics, did you look at these characteristics like waist circumference?

A: Yes we have published extensively on waist circumference and markers of insulin resistance in our cohort and they have been predictors of cardiovascular and renal events. However, we could not do that in this particular analysis as a key component of our prediction model of insulin resistance, A1c, was used to identify the groups, and waist circumference is in the model.



Gregory Nichols, PhD (Kaiser Permanente Center for Health Research, Portland, OR)

This observational, longitudinal cohort analysis evaluated the relationship between glycemic control, cardiovascular hospitalization, and all-cause mortality in the Kaiser Permanente Northwest health system. The analysis included 26,673 individuals with type 2 diabetes who had no known prior cardiovascular hospitalization. Beginning in 2002, the investigators identified the earliest date each individual had an A1c measurement, and averaged it along with all subsequent A1c measurements to segment the study population into 0.5% A1c groups: <6%, 6-6.4%, etc. up to A1c ≥9%. Incidence rates of cardiovascular hospitalization and all-cause mortality were adjusted for age, sex, duration, and A1c.  The incidences of cardiovascular hospitalization and of all-cause mortality were the lowest in the 7.0- 7.4% group; as such, this was used as the reference group. Over a mean of 6.2 years of follow-up, patients with mean A1c <6% or ≥8.5% were significantly more likely to experience cardiovascular hospitalization compared to the 7.0-7.4% reference group; meanwhile, those with A1c <7% or ≥9% had a significantly higher risk of all-cause mortality than those in the 7.0-7.4% reference group. Dr. Nichols stated that, consistent with previous findings, these results suggest the association between mean A1c and risk of cardiovascular hospitalization and all-cause mortality is U-shaped; risk of cardiovascular hospitalization and all-cause mortality were minimized in the 7.0-8.0% A1c range.

Questions and Answers

Q: Do you think that you should put a caveat on your main conclusion that these results are for people who are being treated for diabetes? Clearly there is a linear relationship  between A1c and the risk of cardiovascular disease in non-diabetic individuals. It’s not A1c itself, it’s the means by which it is lowered.

A: That may well be true. I also didn’t mention that these patients were all relatively well controlled. We did have some patients up at 9% but most were well controlled. Maybe additional lowering beyond some level doesn’t buy you that much.



Katarina Eeg-Olofsson, MD (University of Gothenburg, Gothenburg, Sweden)

Dr. Katarina Eeg-Olofsson studied people (n = 12,359; age: 30-75 years; A1c: 7.0-8.9%) with type 2 diabetes in the Swedish National Diabetes Register (NDR) to assess the association between improved glycemic control, cardiovascular risk, and total mortality. After a mean follow-up period of 4.8 years, she found that people whose A1c decreased by at least 0.1% (average change in the decreased A1c population was -0.8% from a baseline of 7.8%) had a markedly reduced risk of coronary heart disease (Odds Ratio = 0.53, p <0.001) cardiovascular disease (OR = 0.53, p <0.001) fatal cardiovascular disease (OR = 0.57, p <0.001), and total mortality (OR = 0.59, p <0.001) than those with a stable or increasing A1c (average change in the stable/increased A1c population was +0.7% from a baseline of 7.7%). Dr. Eeg-Olofsson therefore concluded that people who respond well to glucose lowering treatment (as indicated by a decreasing A1c) will benefit from that treatment, but that efforts must be made to improve risk factor control in people who do not respond to such treatments. Notably, changes in therapy between baseline and follow up did not account for whose A1c decreased vs. whose A1c was stable or increasing, as treatment modifications were similar in the two groups. During Q&A Dr. Eeg- Olofsson called for greater focus to be placed on what factors are causing people to not-respond to treatment; we hope future therapies can address needs of patients who do not benefit from current therapies for whatever reason.

Questions and Answers

Q: Did you separate out stable and increasing A1c?

A: We did not do that, but we are starting to do that. We are doing further analysis on this material.

Q: A small change of A1c of 0.1% is barely associated with any changes in risk, did you try redoing this analysis with a larger change in A1c?

A: These were the changes that we saw in this material. We did not have any larger changes to study with. We did have a very narrow inclusion criteria, maybe we should expand our criteria to include people with higher baseline A1cs, and might have greater improvements in A1c.

Q: Have you tried to compare your results to that of a randomized control study?

A: No, we have not.

Q: There must be a reason that we find in every study people who are not responding to any therapy. If we look at ACCORD the highest mortality rate was in those under intensive control but not responding to therapy. So is it a good job just to increase the amount of insulin or just add another oral anti-diabetic, or do we have to look to other factors that influence this bad outcome?

A: I agree. One of our conclusions was that we have to focus more on this non-responding group, because you could see that they received the same amount of treatment as those responding, but they were not responding. Why were they not responding?

Q: Did you look at depression?

A: No we did not, but it is a nice idea.



Rebecca Simmons, PhD (MRC Epidemiology Unit, Cambridge, United Kingdom)

This study examined the impact of a population-based screening program on mortality among people aged 40-69 at high risk of undiagnosed diabetes. Thirty-two practices in the eastern region of England were included in the study; 27 were randomized to screening (n=16,047) and the remaining five were randomized to no screening (n=4,127). Of those in screening practices, 15,089 (94%) were invited for screening, 11,737 (73.1%) attended screening, and 466 (2.9%) were diagnosed with diabetes. Over the follow-up period (a median of 9.6 years), there was no significant difference in all-cause,  cardiovascular, cancer, or diabetes mortality between the screening and non-screening practices. Compared to the control group, screening attendees had a lower mortality rate, whereas screening non- attendees had a higher mortality rate. Dr. Simmons commented that these findings suggest the benefits from population-level screening may have been overstated. She suggested that the benefits of screening could be increased by detecting and managing related cardiovascular risk factors alongside assessment of diabetes risk, and by repeating screening on a regular basis.


Oral Presentations: What's New in the Treatment of Diabetic Nephropathy?


Per-Henrik Groop, MD (University of Helsinki, Helsinki, Finland)

In a pooled analysis of data for urinary albumin-to-creatinine (UACR) ratio from seven randomized, double-blind, place-controlled trials for linagliptin as monotherapy or add-on therapy (n=4113), linagliptin was found to reduce albuminuria from baseline to 12 and 24 weeks while placebo did not.  The effect on albumin excretion was not explained by linagliptin’s effect on glycemic control (average of 0.7% placebo-adjusted A1c reduction) – there was no significant difference in UACR between quartiles stratified by A1c reduction. Dr. Groop cautioned at the end of his talk that this was not a randomized control trial (and we add that it did not use a “hard” endpoint but a biomarker as an endpoint), but rather a pooled analysis, so it does not necessarily prove that linagliptin improves renal function. Further long-term, randomized control trials would be needed to substantiate that statement.


Oral Presentations: Insights in Diabetic Retinopathy


Helen Looker, MD (University of Dundee, Glasgow, United Kingdom)

Dr. Helen Looker presented an analysis of the first five years of Scotland’s nationwide diabetic retinopathy screening program. The program, which began in 2006, seeks to take annual photographs  of every diabetes patient 12 years and older. From 2006 to 2010, ~20,000 people with type 1 diabetes and ~170,000 people with type diabetes had at least one screening (with a median of three screens per person during the five years). The prevalence of referable eye disease among patients with type 1 and type 2 diabetes, respectively, was ~14% and ~6% in the first year, ~12% and ~4.5% in the second year, and ~10% and ~2.5% each year thereafter. Just under 24,000 referrals were made during the five years in total. Compared to type 2 diabetes, type 1 diabetes was associated with lower risk of referable eye disease among people diagnosed for fewer than 10 years (odds ratio 0.7) but a higher risk among people diagnosed for 10 years or more (odds ratio 1.3). This corroborates other evidence we’ve seen that many type 2 diabetes patients already have (or are well on their way to) complications by the time they are diagnosed. Dr. Looker noted that in both type 1 and type 2 diabetes, 60% or more of the referrals were due to maculopathy unrelated to diabetic retinopathy, and thus did not lead to treatment; she said the program directors are still considering how to improve the clinical flow of these individuals.


Oral Presentations: Diagnosing and Treating Diabetic Neuropathy


Niels Ejskjaer, MD, PhD (Aarhus University Hospital, Aarhus, Denmark)

Diabetic gastroparesis (delayed gastric emptying in the absence of obstruction in the gastrointestinal tract) is often observed in people with type 1 and type 2 diabetes, due to neuropathy of nerves involved in digestion. Symptoms range in severity from constant nausea and painful vomiting to weight loss, electrolyte derangement/dehydration, and even ketoacidosis. Dr. Niels Ejskjaer presented the results of  a phase 2 study supported by Tranzyme Ltd. investigating the potential therapeutic benefits of a ghrelin receptor agonist. Over 28 days of treatment, the ghrelin receptor agonist was found to improve gastroparesis symptoms in participants with type 1 and type 2 diabetes, but therapeutic benefits did not last the full 30 days of follow-up. Despite the discouraging durability data, Dr. Ejskjaer remains hopeful that the drug can become a novel treatment for symptomatic diabetic gastroparesis.

  • Due to a lack of effective pharmacotherapies available and the impracticality of subjecting patients to neurostimulation and surgery, a selective potent ghrelin receptor agonist was chosen as a possible treatment for symptomatic diabetic gastroparesis. Ghrelin, a naturally occurring gut peptide hormone, was chosen because animal and clinical data show that IV ghrelin enhances gastric emptying (Murray et al. Gut 2005;54(12):1693-8). Ghrelin has a short plasma half-life of 6 to 12 minutes.
  •  This phase 2, multinational, double-blind randomized control trial included 55 individuals with type 1 diabetes and 37 with type 2 diabetes. Participants with type 1 diabetes were slightly younger (46.4 vs. 55.2 years of age), had a longer diabetes duration (25.5 vs. 10.2 years), and had higher average A1c (8.9% vs. 7.4%). Gastroparesis Cardinal Symptom Index (GSCI) score was used to measure the severity of symptoms on a 0-5 scale with 5 representing the greatest discomfort. GSCI was measured at baseline, 8, 15, 28, 42, and 58 days, and was scored  for nausea, early satiety, bloating, and upper abdominal pain.
  • Symptoms improved for patients with type 1 and type 2 diabetes similarly after 28 days, with a mean GSCI score improvement over placebo between 0.7 and 1.2 across all categories. Unfortunately, these results were not sustained over the following 30 days. When comparing patients receiving the 20mg dosage with placebo, though symptoms improved with statistical significance across all GSCI symptom categories at day 28, there was no statistically significant difference between GSCI scores at day 58. In fact, early satiety and upper abdominal pain mean scores were numerically higher for participants receiving 20mg of the ghrelin receptor agonist compared to those receiving placebo.
  • There was no correlation observed between GCSI total score and gastric half- emptying time. Gastric half-emptying time was measured by 13C-octanoate breath test, and average gastric half-emptying time was 193 minutes on average for all patients at baseline with little difference observed between participants with type 1 and type 2 diabetes. Whether gastric half-emptying time decreased after 28 or 58 days of treatment was not addressed.

Questions and Answers

Q: I was wondering about the side effects of ghrelin, especially weight gain. Within 28 days, did you observe any weight gain or other side effects?

A: This is a highly appropriate question, as one would expect patients to eat more. When looking at weight and A1c and fasting blood glucose on these 92 patients over 28 days, we found no changes. These results will be published this autumn in Neurogastroenterology & Motility.

Q: There were no other side effects?

A: A few patients suffered pain in the abdomen. This has been reported and may be due to propulsion in the GI tract, but that is highly speculative. We have no other explanation for that.

Q: I understand there is no correlation between symptom improvement and gastric emptying. Is there an improvement in gastric emptying following ghrelin therapy?

A: In two previous studies, we have demonstrated that ghrelin receptor agonists, given intravenously, accelerate gastric emptying. We were unable to show this in this study. There is a clear relationship in animal studies for certain, and we did a proof of principle in ten severely ill patients, where gastric emptying increased for all participants.

Q: When you showed the different effects on different symptoms, at the end of your follow- up the placebo crossed the ghrelin and was actually superior to the ghrelin. Do you have an explanation?

A: Not directly. I can only say that the placebo effect is dominant in all the patients – we look after them for 28 days. It might have been interesting to follow them over the weeks after follow-up and look at them as closely as we did during the treatment period.


Oral Presentations: Type 1 Diabetes Mellitus: Acute and Chronic Complications


Aleksander Araszkiewicz, MD, PhD (Poznan University of Medical Sciences, Poznan, Poland)

Dr. Aleksander Araszkiewicz presented a study assessing the relationship between insulin resistance and carotid artery intima-media thickness (cIMT). The study included 79 patients with type 1 diabetes with a mean age of 37 years, a mean A1c of 8.3%, and a mean duration of diabetes of 13 years. Patients defined as insulin resistant (see definition below) comprised 36% of the study population and had significantly higher IMT than insulin sensitive patients (0.63 mm vs. 0.55 mm; p=0.004). In a multiple linear regression model, greater insulin resistance was associated with increased IMT (beta -0.40; p=0.003) and was independent of duration of diabetes, sex, and LDL. Significant positive correlations with IMT were also observed for body mass (r=0.26; p=0.025), BMI (r=0.28; p=0.015), waist circumference (r=0.34; p=0.003), and waist to hip ratio (r=0.24; p=0.041). The study’s major  limitation was an inability to measure insulin resistance using the gold standard hyperinsulinemic euglycemic clamp – in particular, audience members were concerned that the use of estimated glucose disposal rate (which uses A1c, blood pressure, and waist to hip ratio) to estimate insulin resistance may muddy the results.

  • Insulin resistance was estimated using the formula for estimated glucose disposal rate. The inputs are hypertension, A1c, and waist to hip ratio. Patients with an estimated glucose disposal rate <7.0 mg/kg/min were defined as insulin resistant.

Questions and Answers

Q: After 13 years of diagnosis, why were there such big differences in neuropathy? Is it neuropathy is under-diagnosed.

A: It looks like it. Diagnosis of neuropathy is not as perfect as retinopathy. We didn’t use the conduction velocity or nerve biopsy or skin biopsy. Those might have been used to diagnose differently.

Q: Are you following this cohort for fatal events, MIs, stroke? Just to confirm the observation that you’ve seen with IMT.

A: Yes, this is 13 years of diabetes. We have only one MI in this group. There are too few events so we assessed IMT. In the future, we will assess strong points.

Q: How about revascularization procedures, CABG, angioplasty?

A: We have not done that.

Q: What about normalizing your data with hypertension numbers? The IMT difference might be due to differences in blood pressure.

A: We have to do some analysis on that.

Q: Did you include A1c in your regression? Because A1c brings a high risk for IMT.

A: We didn’t. A1c is in the glucose disposal rate formula. We couldn’t do it separately.



Lisa Olsson, PhD (Karolinska Institutet, Stockholm, Sweden)

Dr. Lisa Olsson presented results on a study investigating the risk for and risk factors of mortality in latent autoimmune diabetes of adults (LADA) compared to type 2 diabetes. Participants were drawn from the Nord-Trøndelag Health Study (LADA n=208; type 2 diabetes n=2,425) and followed in the Norwegian National Cause of Death Registry. Information on A1c, metabolic, lifestyle, and socioeconomic risk factors were collected. At baseline, participants with LADA had a slightly better metabolic risk factor profile with significantly lower BMI (28 vs. 30 kg/m2; p<0.0001), waist-to-hip ratio (0.88 vs. 0.90; p=0.002); central obesity (70% vs. 80%; p=0.001); however, patients with LADA had a longer duration of diabetes (15 years vs. 11 years; p <0.0001) and higher A1c (8.3% vs. 7.7%; p <0.0001). Despite these differences, mortality was equally high (slightly higher even) in patients with LADA than in patients with type 2 diabetes. When analyzed according to risk factors, BMI, lifestyle factors, education level, occupational position, and duration of diabetes could not explain excess mortality. Dr. Olsson then looked to the impact of glycemic control: when participants were categorized according to A1c, patients (both LADA and type 2 diabetes) with A1c ≥7% had significantly greater all- cause mortality risk than patients with A1c <7% (HR=1.62 [95% CI: 1.49-1.76] vs. HR=1.23 [95% CI: 1.08-1.41]). Particularly for LADA cases, the mortality risk appeared to substantially increase when A1c exceeded 9%. Dr. Olsson concluded that the higher mortality risk in cases with poor glycemic control emphasize the importance of correct treatment of hyperglycemia to prevent complications and improve survival.

Questions and Answers:

Comment: The study underlines that A1c is an important risk factor. This was also shown in the DCCT study.

Comment: The fact that you get increased mortality emphasizes the importance of identifying patients at risk with high A1c, and provides justification for the identification of LADA patients. This is a consistent feature in a lot of studies – that the A1c is increased in LADA patients.



Eva Aguilera, MD, PhD (Hospital Universitari Germans Trias I Pujol, Badalona, Spain)

In order to evaluate the presence of early atherosclerosis in asymptomatic patients with type 1 diabetes, Dr. Eva Aguilera and colleagues compared results from three non-invasive screening tests: 1) carotid ultrasound to determine mean cIMT and the presence of atheroma plagues (carotid ecography); 2) myocardial perfusion imaging with single-photon emission computed tomography (SPECT); and 3) multiple slice computerized tomography (CT) with quantification of calcification (calcium score).   People, living in Catalonia, aged 20-50 years who had type 1 diabetes for more than 1o years and no previous history of clinical macrovascular or ischemic heart disease (n=150) were compared to non- diabetic age- and sex-matched controls (n=50). For background, a calcium score of zero indicates no plagues (very low risk); a score from one-to-10 indicates minimal plaques (low risk); a score between 11 and 100 indicates mild calcification (moderate risk); a score of 101-400 indicates moderate calcification (high risk); and a score >400 indicates significant calcification (very high risk). Of the people with type   1 diabetes, 82% had a score of zero vs. 92% of controls. Calcium score in people with type 1 diabetes was significantly associated with age, diabetes evolution, metabolic control (as measured by A1c), and the carotid ecography test (which was also significantly associated with metabolic control). However, SPECT showed no association with either carotid ecography or calcium score. Looking at the overall  risk in patients suggested by these tests and the concordance (or rather, lack there of) between the three tests, Dr. Aguilera reached three conclusions: 1) a small percentage of patients with type 1 diabetes  living in Catalonia with more than 10 years of disease evolution showed data suggestive of subclinical atherosclerosis; 2) universal screening of coronary disease in this population is probably not justified; and 3) carotid ecography should be recommended in the subset of patients with type 1 diabetes with other cardiovascular risk factors and long disease evolution.


Oral Presentations: Oral Therapies: New Targets


Wendy Boertien, MD (UMCG, Groningen, Netherlands)

Copeptin – a proxy for the anti-diuretic hormone vasopressin – was evaluated as a potential biomarker of kidney disease in a retrospective analysis of the ZODIAC cohort (n=1,325 type 2 diabetes patients). Copeptin levels turned out to be strongly associated with baseline urinary albumin/creatinine ratio (ACR) and baseline estimated glomerular filtration rate (eGFR). A strong association was also seen between baseline copeptin and long-term increases in ACR and decreases in eGFR, independently of baseline levels of either ACR or eGFR, while also controlling for established diabetic nephropathy risk factors including BMI, A1c, blood pressure, and smoking – and it was stronger for baseline copeptin than for the other risk factors. (The long-term correlation is lost in patients using RAAS inhibitors, however, indicating that vasopressin’s deleterious effects might be mitigated with RAASi therapy.) The next step in validating copeptin as a biomarker will be to test whether interventions that change vasopressin levels can slow the progression of diabetic nephropathy. Since vasopressin is involved in maintaining the body’s fluid volume, such an intervention could be as straightforward as drinking more water.

  • Vasopressin, also called anti-diuretic hormone due to its role in fluid retention, has been shown to cause kidney damage in rodent models of diabetes: e.g., glomerular hyperfiltration, albuminuria, and glomerulosclerosis. Vasopressin is difficult to measure due to its small size and short half-life, so the researchers instead used copeptin – a fragment of the  copeptin precursor that is more stable in the blood.
  • The researchers retrospectively analyzed baseline plasma samples and annual kidney-function tests from patients with type 2 diabetes who had enrolled in the ZODIAC study (n=1,325, mean age 67 years, median duration of diabetes four years, mean A1c 7.3%). At baseline, mean urinary albumin/creatinine ratio (ACR) was 16 mg/g, mean estimated glomerular filtration rate (eGFR) was 65 ml/min/1.73m2, and mean plasma copeptin was 5.2 pmol/l. Data going out three years or more were available for 1,184 patients (89%), who were followed for a median of 7.4 years. Kidney function declined on average during this time: mean ACR increased by 0.1 mg/g/year, and mean eGFR decreased by 1.2 ml/min/1.73m2/year.

Questions and Answers

Q: Did you correct the serum copeptin levels for serum osmolarity and glucose? Clearly acute changes in these could affect vasopressin content.

A: We did not do this, but I think that if we had made all the patients fast before taking blood samples then the relationship would have been even clearer.

Comment: I am not suggesting doing this, I just think you should correct for osmolarity.

Q: To put these values in context, what would be copeptin levels in well-hydrated individuals without diabetes?

A: The values in the study are actually slightly higher than would be expected for the normal population.

Q: Do you think the levels of vasopressin could be used as a measure of sufficiency of RAAS blockade, for use in titration?

A: That is interesting – perhaps, though such a measure already exists.


Oral Presentations: The Effects of Interventions in Reality


Nils Ekstrom, MD (University of Gothenburg, Gothenburg, Sweden)

This observational study evaluated the efficacy and safety of metformin in patients with varying  degrees of renal impairment in a large nationwide sample, using data from the Swedish National Diabetes Register from July 2004 – December 2010. Insulin monotherapy (n=12,291) was associated with higher risks of cardiovascular disease (HR=1.18; 95% CI: 1.07-1.29), all-cause mortality (HR=1.34; 95% CI: 1.19-1.50), any acidosis/serious infection (HR=1.28; 95% CI: 1.14-1.43), and fatal acidosis/serious infection (HR=1.45; 95% CI: 1.07-1.97) compared to metformin monotherapy (n=14,697). Other OAD monotherapy (n=5,171) was associated with a higher risk of all-cause mortality compared to metformin (HR=1.13; 95% CI: 1.01-1.27). Meanwhile, metformin in any combination (n=31,628) versus any other glucose-lowering medication (n=19,997) was associated with reduced risks of all-cause mortality for patients with normal renal function (eGFR >60; HR=0.87; 95% CI: 0.81-0.94) and patients with mild renal impairment (eGFR 45-60; HR=0.87; 95% CI: 0.77-0.99), as well as any acidosis/serious infection (HR=0.91; 95% CI: 0.84-0.98; HR=0.85; 95% CI: 0.74-0.97). There was no increase in risk even in patients with moderate renal impairment (eGFR 30-45). Dr. Ekstrom concluded that these findings support more liberal guidelines for the use of metformin, such as those recommended by NICE.

  • Key inclusion and baseline demographics: Key inclusion criteria were a diagnosis of type 2 diabetes, an age of 40-85 years old, maintenance of glucose-lowering treatment, and no missing data. A total of 51,675 individuals were included in the analysis, with an average follow-up period of 3.9 years. Individuals were on average 65 years old, with diabetes duration of 9.4 years, A1c of 7.3%, and BMI of 29.5 kg/m2. Those who were on insulin treatment had higher A1c, longer duration of diabetes, and a higher degree of morbidity, while those on metformin had higher BMI, and those on other OADs had a higher age. To account for these differences, the study  investigators adjusted for propensity score.


Oral Session: Hyperglycemia and the Brain


Leonid Strongin, MD (Nizhny Novgorod State Medical Academy, Nizhny Novgorod, Russia)

This study compared two strategies for insulin therapy in the first 24 hours after stroke: intravenous infusion and subcutaneous injection. The study included 73 patients with either type 2 diabetes or blood glucose >11 mmol/l (~198 mg/dl) at admission. The 37 participants in the control group received subcutaneous insulin injection (SCI) from admission to discharge with target blood glucose 7.8-10 mmol/l (~140.4-180 mg/dl). The other 36 participants received intravenous insulin infusions (IVI) during the first 24 hours, followed by subcutaneous insulin injection until discharge. No statistically significant differences were found in age, history of stroke, diabetes duration, A1c (9.4% and 9.5% for intravenous infusion and subcutaneous injection, respectively), blood glucose, or % already receiving insulin therapy between the two treatment groups at admission. In both groups, participants received human regular insulin, which was administered with a B Braun perfusor in IVI. During the first 24 hours of insulin therapy, IVI individuals had mean blood glucose levels at 8.7 mmol/l (~156.6 mg/dl) whereas SCI individuals had mean blood glucose levels at 9.8 mmol/l (~176.4 mg/dl) with p=0.025. 72.0% of the IVI group maintained target glucose between 7.8 and 10.0 mmol/l (~140.4 – 180 mg/dl), compared to 32.4% of the SCI group with p<0.001. BADLI and NIHSS scores, measures of quality of life following stroke, were statistically significantly worse in patients in the SCI group. Ultimately, Dr. Strongin concluded from the study that glucose control using intravenous insulin infusion has advantages over subcutaneous insulin in regressing neurological deficit, improving functional recovering, and decreasing risk of hypoglycemia, though the impact of routes of insulin administration on six-month survival wasn’t proved as those results were statistically inconclusive.

Questions and Answers

Q: You included patients with hyperglycemia, but it is very well-known that in acute stroke, there are many patients without diabetes mellitus but present with hyperglycemia. And, after recovery, blood glucose goes to normal levels. Did you distinguish between patients with a history of type 2 diabetes and those with so-called transient hyperglycemia?

A: Yes, the study includes patients who have a history of diabetes mellitus. We tried to exclude patients with regular levels of glycated hemoglobin on admission.


Oral Presentations: Modifiers and Markers for Cancer in Type 2 Diabetes


Iris Walraven (VU University Medical Center, Amsterdam, The Netherlands)

Ms. Iris Walraven described the Hoorn study, which found that high proinsulin levels (> 16.5 pmol/l)  are independently associated with a two-fold increased risk of cancer mortality. The study investigated the relationship between fasting proinsulin levels (which is highly predictive of type 2 diabetes) and 20- year cancer mortality. In 1989, the study measured the proinsulin levels of 438 individuals (mean age  64 years, 50% male) and followed the participants for a median of 17.3 years, during which 45% of the original study cohort passed away, 27% of whom died from cancer. In their data analysis, investigators stratified the cohort into tertiles for fasting proinsulin, corrected for age & sex, and categorized patients by glucose metabolism (normal glucose tolerance, impaired glucose metabolism, and type 2 diabetes). No relationship was observed between glucose metabolism and cancer mortality. However, the highest tertile of proinsulin (>16.5 pmol/l) was associated with a greater risk for cancer mortality compared to the first and second tertile across all categories of glucose metabolism. While high proinsulin levels (>16.5 pmol/l) were significantly associated with all-cause mortality, this relationship diminished and was no longer statistically significant when cancer deaths were removed from the data set.

Questions and Answers

Q: Can you show the risk ratio for the different causes of death? As far as I can see, the risk ratio for cancer mortality is not different from that of non-cancer mortalities. The 95% confidence intervals for both are so wide that there is not really a difference.

A: The relationship between proinsulin and cancer mortality is stronger than that of proinsulin and all- cause mortality not including cancer.


Symposium: Diabetes and Acute Coronary Syndrome


Peter Gæde, MD, DMSc (Copenhagen University Hospital, Copenhagen, Denmark)

Dr. Peter Gæde explored the topic of whether patients with diabetes and acute coronary syndrome (ACS) should be treated intensively with insulin. He argued that hyperglycemia (acute hyperglycemia more so than chronic hyperglycemia) contributes to morbidity and mortality associated with ACS, but that data from randomized control trials of diabetes patients experiencing ACS provide evidence that using intensive insulin treatment to control blood sugar does not decidedly improve outcomes. In multiple studies (e.g., DIGAMI and DIGAMI-2) patients assigned to intensive insulin therapy did not actually achieve better glycemic control than patients in the standard care arm, so results are difficult to interpret. However, Dr. Gæde argued that what has been established is that when studies find that patients with more intensive glycemic control experience higher mortality, it is primarily because intensive insulin treatment results in more hypoglycemia, and hypoglycemia has been recognized as a predictor of mortality (e.g., in NICE-SUGAR, the largest study to compare intensive vs. standard glucose control in the ICU, moderate and severe hypoglycemia were associated with increased mortality). He stated that the highest priority for diabetes control for patients with ACS should be avoiding  hypoglycemia.

Questions and Answers

Q: Glucose variability is a popular topic now – any comments? Is variability or absolute glucose level most important?

A: As I alluded to earlier, variability is very important. If you look at some observational trials, you see patients that have very high glucose excursions on two hour-post prandial tests, and these patients have a very high risk for complications.

Q: Given the risk of hypoglycemia and given that you have been speaking only of insulin treatment, do you see a role for alternative diabetes treatments under these conditions?

A: Yes. I always say we need a statin for glucose, because if we could just lower glucose without the side effects or hypos, things would be very different. I think GLP-1s may head in that direction but I think we don’t have the data to say that yet.



Guy Rutten, MD, PhD (University Medical Center, Utrecht, Netherlands)

Dr. Guy Rutten gave a compassionate talk emphasizing the need to treat not only disease, but patients as a whole. He noted that many patients with both diabetes and acute coronary syndrome (ACS) find it hard to manage both diseases at once, neglecting one disease if they prioritize the other. He called for a tailored approach in which patients and primary care physicians are proactive in consolidating treatments prescribed by cardiologists and diabetologists. During Q&A, he noted that glycemic control after discharge for ACS might realistically have to take a back seat to blood pressure and lipid control.

Questions and Answers

Q: Regarding the problem of adherence to treatment – especially with regard to handling both diseases – what can be done? Is it just a matter of awareness or support? Or is there a potential technical solution? Should we use a system to track use of treatment?

A: People who are not adherent are, on average, less well educated and less well informed. I would emphasize the ADA recommendations. If each hospital has a structured system so that the patient talks with a diabetologist or cardiologist upon discharge for even five to 10 minutes to emphasize the need for continuing use of medications and suggest that the patient go to their primary care physician with a discharge form, it would ensure there is not a gap in the continuation of care. Perhaps a technical solution might work, but I think a discharge consultation is very valuable.

Q: What kind of A1c target would you propose for a patient with ACS in the first three months of discharge? Should we be stringent or less so?

A: Of course A1c is important but I think also in the months after discharge, blood pressure and lipids are even more important. But of course, if they let diabetes slip then, it would not be good. It’s not realistic to strive for strict glycemic control in the first month. We should be glad if they just continue all medications and can cope with the two diseases and keep A1c stable. If you look at the stratification analysis of  UKPDS, I think between 7% and 8% would be a realistic goal.


Symposium: Insights in Diabetic Retinopathy


Gabriele Lang, MD (University Eye Hospital, Ulm, Germany)

Dr. Gabriele Lang presented 36-month data from the RESTORE extension trial for the long-term safety and efficacy Lucentis (ranibizumab [RBZ]) in diabetic macular edema (DME). RESTORE found that intravitreal injection of 0.5 mg Lucentis ± laser therapy was superior to laser therapy alone for the treatment of DME. For more details on the original trial, please see our January 1, 2011 Closer Look at Among the 303 patients that completed RESTORE, 240 enrolled in the extension study, and 220 patients completed the 24-month extension phase. All patients received RBZ on an as-needed basis (PRN), and 19-25% of patients across all treatment groups did not require additional injections during the third year. No new safety findings emerged, and mean best corrected visual acuity (BCVA) achieved during the initial 12 months were maintained out to month 36, with a mean of 2.4-2.9 injections required during year three of the study. Change in BCVA at month 36 was +8.0, +6.7, and +6.0 letters (or >1 line on an EDTRS eye chart) for the groups that had received RBZ, RBZ + laser, and laser, respectively, during the first 12 months. The extension study will continue out to a total of 48 months.

Questions and Answers

Q: I think the data can be read in two opposite ways, perhaps. One way would be to say – when we first got results from year one we thought laser could just be thrown away because injections were so much better. The follow-up is interesting because it suggests that you can always start with laser, add injections later, and still obtain good outcomes at the end of three years. What is your position? Which should you start with? Ranibizumab is an expensive drug.

A: If you look at the one-year data showing rapid improvement in visual acuity, I would consider ranibizumab as first line treatment. With consistent laser therapy, the patient develops retinal scars so   you do morphological harm to retina. We have different kinds of macular edema; ranibizumab is licensed for central involvement and visual deterioration. We also have non-centrally involved macular edema that is clinically significant, but in patients with central involvement I would suggest anti-VEGF treatment. I’m very pleased because, in contrast to wAMD and RVO where after six months you start to get worse, you have a longer time frame in DME to treat with anti-VEGF.

Q: Do you think that the decrease in number of injections required to maintain visual acuity was due to the treatment itself? Don’t you think it could be due to study effect? Patients who come in every month for three years must be very motivated. Don’t you think might have also improved glycemic and blood pressure control, which could explain resolution of macular edema after two or three years?

A: To my knowledge we did not find significant improvements in glycemic control or blood pressure control. Also in clinical practice, if you follow DME patients, you get the impression that it’s a different story from other diseases treated with ranibizumab. What we do is inhibit VEGF. Especially in diabetic retinopathy, there are lots of things that play a role in the pathogenesis. Perhaps VEGF is not so important in the whole course of the disease and is only important in a certain phase, but VEGF is certainly one of  the major players in causing DME. I think its treatment has a different time course compared to other diseases.

Q: In RESTORE and other ranibizumab treatment studies, what happened with the kidney? Was there any impact of diabetic kidney disease? What was the frequency of kidney disease? Did it have anything to do with DME? Is the institution of ACE inhibition during this study a confounding effect? These questions are not addressed in the papers.

A: Due to time we could not go into detail, but there was no confounding impact of kidney disease on the results.

Q: Do patients with DME have more diabetic kidney disease than control groups?

A: I do not know a study that has addressed this question.

Q: Do you think it would be important to study?

A: Yes, that would be important, because we are now more concentrated on just treating DME when we should not forget about the general microvascular complications and also the communication between diabetologists, general practitioners, and ophthalmologists.


Symposium: New Treatments of Macular Edema in Diabetic Patients


Per-Henrik Groop, MD, DMSC (Helsinki University Central Hospital and Folkhælsan Research Center, Helsinki, Finland)

Dr. Per-Henrik Groop summarized his opinions on the merits of current options for the prevention or treatment of diabetic retinopathy: glycemic control, blood pressure control, lipid-lowering drugs, laser therapy, and anti-VEGF. He relayed that evidence from DCCT showed that glycemic control is effective in type 1 diabetes, but that a Boussageon et al. study (BMJ 2011) pooling data from RCTs in type 2 diabetes show that intensive glycemic control is only effective in type 2 diabetes if practiced early on. He says that UKPDS showed tight blood pressure control reduced the risk of worsening retinopathy by 34% but in ACCORD, anti-hypertensive treatment did not seem to have an effect. When specific Ras blockers were investigated, he said enalapril and losartan reduced progression of retinopathy while lisinopril, indapamide, and candesartan were not as positive. He said that no statins have shown a significant effect on the progression or incidence of retinopathy but that the FIELD study showed that the use of fenofibrate resulted in the need for less laser treatment. He believes laser therapy is effective, but repeated laser therapy destroys the retina. He also stated that anti-VEGF treatments bear promise since they target the underlying pathogenesis of diabetic macular edema, but are expensive and too new to draw conclusions about their safety. In what he identified as the most important take home message of his talk, he stated that severe neuropathy and proliferative retinopathy is much more prevalent in patients with diabetic nephropathy and that these patients usually succumb to macrovascular disease. Finally, he concluded that diabetologists must play an active role in the treatment of diabetic   retinopathy because measures such as managing glucose control, hypertension, and lipids go a long   way to prevent other diabetes complications as well.

Questions and Answers

Q: Do you have a head to head for losartan and candesartan?

A: No, but I think it’s a genetic effect.

Q: What is the effect of GLP-1 on progression of diabetic retinopathy?

A: That’s a good question. I thought about it this morning, and I have absolutely no idea.

Q: Are there any anti-diabetic agents that worsen macular edema? For example, pioglitazone, which causes edema in other body parts?

A: A very good point. I would be very cautious in using TZDs in patients with complications, especially if you have patient with diabetic nephropathy, microalbuminuria, or diabetic retinopathy. And yes, the reason is that these drugs cause generalized edema. There are studies showing that TZDs are also effective in reducing albuminuria, but because of problems with edema itself, they are not recommended in these patients with existing renal dysfunction. I would be very cautious to use them if you have macular edema.



Pascale Massin, MD (University of Paris, Paris, France)

Dr. Pascale Massin reviewed the rationale for anti-VEGF therapy as a way to reduce leakage and improve visual acuity in diabetic macular edema, and she discussed the four therapies that are commercially available: ranibizumab (the only approved drug for DME), bevacizumab (not approved, but widely used in DME; much less expensive than ranibizumab but associated with septic risk due to repackaging from large cancer doses into small intravitreal doses), aflibercept (in late-stage trials for approval in DME; potentially dosable less frequently than ranibizumab), and pegaptanib (not approved in DME). She highlighted two questions that she said remain unanswered: how long anti-VEGF treatment needs to be maintained, and whether the required follow-up monthly monitoring with anti- VEGF therapy is really feasible. Whatever the answers, Dr. Massin said that anti-VEGF therapy leaves substantial unmet need for less-invasive therapies that address the 20-30% of patients that do not respond to anti-VEGF and/or that can be used earlier in the course of disease.

Questions and Answers

Q: From a cost perspective, I would be in favor of excluding non-responders from treatment.

A: We have not been able to predict this from clinical data to date. My feeling is that success is better   when systemic health is stable; my personal feeling is that poor blood pressure and glycemic control are predictive of poor response, but the studies enrolled only fairly well-controlled patients. But we don’t have a clear signal, including age.

Q: What could be done to improve visual outcome given that only those with worsened vision and edema are eligible for anti-VEGF therapy, even though we don’t screen for these factors?

A: We screen for macular edema, and we must associate with OCT and visual acuity. When visual acuity is low, predicting responder is difficult – the correlation with retinal thickness is not strong. We don’t know why visual acuity lessens with diabetic macular edema. It’s not just retinal thickening; glial cells appear to be involved. We hope that the new OCT devices will help us understand these mechanisms better.

Q: Anti-VEGF therapy is very expensive, and we don’t know how long is needed to treat. Do you ever see a need for recurrence of macular edema when the therapy is stopped?

A: Yes, for sure. That is why we need monthly monitoring – recurrence often occurs. Apparently we need fewer injections in the second and third years, at least in the studies. This treatment needs monthly monitoring; we must perform visual acuity tests once a month.



Jost Jonas, MD (Ruprecht Karls University of Heidelberg, Heidelberg, Germany)

As an alternative to anti-VEGF therapy, Dr. Jost Jonas promoted the use of the crystalline steroid triamcinolone acetonide, which was the first drug used intra-vitreally prior to the ranibizumab era. He also briefly reviewed other unconventional options such as inserting GLP-1 beads into the anterior chamber of the eye allowing for intraocular production of GLP-1; slow release of the glucocorticoid dexamethasone, human bone marrow mesenchymal cells, and steroid eye drops as other future alternatives.

Questions and Answers

Q: In reference to the GLP-1 receptors – what about DPP-4? It’s also found in the eye. Are any studies planned for a DPP-4 inhibitor in the eye?

A: No unfortunately the company that was working on the GLP-1 in Germany has changed its focus.


Symposium: New Treatments for Microvascular Complications


Merlin Thomas, PhD (Baker IDI Heart and Diabetes Institute, Melbourne, Australia)

In a straightforward, positive presentation on a vital topic, Dr. Merlin Thomas reviewed options for the treatment of diabetic nephropathy. He opened by presenting evidence that standard care (glycemic and blood pressure control) works very well, but that a large unmet need still exists for patients who still develop nephropathy regardless of diabetes and hypertensive control. He expressed optimism for established drugs for other indications (e.g. fenofibrate, vitamin D, linagliptin, exenatide, Viagra) that are currently being evaluated for effects on renal function. Finally, he reviewed promising new kidney treatments in development, highlighting bardoxolone methyl and H2 enriched water as particularly high-potential  options.

  • Old drugs, Dr. Thomas said, can still learn new tricks. Fenofibrate, originally developed as a lipid-lowering drug, has been found to reduce the progression of albuminuria. Thiazolidinediones have also been widely touted for decreasing albuminuria (e.g., pioglitazone in the PROactive study produced a mortality, myocardial infarction, and stroke benefit for those with eGFR <60), but Dr. Thomas believes that the unfavorable side effect profile of this class will probably prevent further pursual of this front. In a pooled analysis of linagliptin (Lilly’s Tradjenta) phase 3 trials, patients on a background of ACE/ARB saw a ~30% decrease in urine albumin to creatinine ratio (UACR) compared to ~8% decrease for placebo. Exenatide has also been shown to reduce albuminuria. Vitamin D supplementation has been found to reduce UACR, but there are concerns about giving a calcification-promoting agent to patients who already have problems with over-calcification of blood vessels. Vitamin D analogs that do not promote calcification have been found effective in reducing albuminuria in patients with type 2 diabetes (e.g., VITAL study, Lancet 2010), and Dr. Thomas thinks this will be worth expanding. Finally, even Viagra (sildenafil), has been shown to reduce microalbuminuria by half.
  • Dr. Thomas highlighted bardoxolone methyl and H2 enriched water as especially hopeful new treatments for renal disease. We learned that bardoxolone methyl acts by relieving imbalances between oxidants and antioxidants (a major challenge in diabetes management) by acting as a pseudostressor. Cells are thus tricked into releasing antioxidants. Bardoxolone methyl improves estimated glomerular filtration rate (eGFR) in patients with type 2 diabetes and chronic kidney disease by about 10 ml/min/1.73 m2, with at least 52 week durability and, furthermore, also produced a striking ~10 kg (~22 lb) weight loss (from a baseline BMI ≥35; Pergola et al., NEJM 2011); notably, Dr. Thomas said that this conferred an extra two to three years of dialysis-free life. Dr. Thomas also mentioned phosphodiesterase inhibitors and renal sympathetic nerve ablation as new possibilities in development. His favorite new approach is H2 rich water (HRW). H2 is a physiologically safe antioxidant against cytotoxic radicals; given its “fiendishly small” size, it easily traverses membranes. This treatment is easily produced by steeping a stick of magnesium into water, allowing the spontaneous re-dox reaction to bubble H2 into the water up to concentrations of 0.6 mM. In clinical studies HRW has been shown to decrease urinary isoprostanes (a marker of oxidative stress), increase superoxide dismutase (an antioxidant), and decrease TBARS (a reactive oxygen species).

Questions and Answers

Q: Should ACE inhibitors or ARBs be used earlier to prevent onset of kidney disease?

A: That has been studied and debated and is probably worthy of more debate. There’s lots of suggestions that the benefits in primary prevention of nephropathy from Ras blockade are no more than the effects on blood pressure lowering. When you eliminate differences in blood pressure, there are no differences in the effectiveness of primary prevention. I don’t think starting early is robustly supported by meta-analyses. However, we live in the real world where the way to best achieve blood pressure lowering is through the use of those drugs.

Q: Is there an independent effect of diet or exercise on renal function?

A: Calorie restriction trials in patients with diabetic kidney disease has been shown to be beneficial not  just because of weight loss, but actually because of beneficial effects it might have on the kidney. One  effect of the low calorie diet appears to be activation of the NRF2 pathway [the same pathway activated by bardoxolone methyl] along with a number of things that improve efficiencies. We argue in our recent paper in Diabetes that complications turn out the way they do because excess calories change the way the body functions. When you have no calories, the body is very efficient. The opposite is true in diabetes – a little bit of inefficiency doesn’t matter so if you lose a few mitochondria here or there, the body doesn’t  care because you have the resources to build another one. So you may not be able to look after things as well as you should. Maybe tricking the body into thinking it is on a low calorie diet using AMPK activators like metformin, or by using sirtuins would improve efficiency of glucose so that we get less AGEs and oxygen use so we get fewer reactive oxygen species.



Solomon Tesfaye, MD (University of Sheffield, UK)

As we have heard before, Dr. Solomon Tesfaye reiterated that current treatment options for diabetic peripheral neuropathy (DPN) are suboptimal and are not broadly effective, stating that right now “the best we can look for is 50% relief in 50% of patients.” He presentation focused on results from the COMBO-DN study, which compared duloxetine and pregabalin (the two most common first line drugs for DPN) for initial treatment of DPN. The study also examined whether intensification of monotherapy or combination of the two drugs at standard doses would be more effective for non-responders to first line treatment. Results demonstrated that initial treatment using duloxetine 60 mg/day provided better pain relief than pregabalin 300 mg/day (as measured by the Brief Pain Inventory Modified Short Form after eight weeks). For non-responders to initial treatment, there was no difference between escalating to high dose monotherapy compared to adding on the standard dose of the other drug in combination therapy. Treatment emergent adverse events were similar for both groups.

Questions and Answer

Q: Is there additional benefit for those with depression to use antidepressants to treat DPN?

A: Patients are often depressed too so it may make sense to use an anti-depressant. In a lot of studies people with depression were excluded.


Symposium: Therapeutic Applications of Stem Cells in Diabetes Complications


Timothy O’Brien, MD, PhD (National University of Ireland, Galway, Ireland)

Dr. O’Brien began his presentation on using stem cells to treat diabetic ulcers by reminding the audience that 15-25% of people with diabetes will develop foot ulceration, and that the current standards of care are not adequate. He explained that that people with type 1 diabetes have fewer endothelial progenitor cells (EPCs; cells from the blood or bone marrow that can differentiate into endothelial cells or support angiogenesis). Additionally, the EPCs that people with diabetes do have demonstrate decreased   function. He argued that EPCs are dysfunctional in people with diabetes because genes are dysregulated (>2-fold change in expression). He went on to present preclinical research (mainly in rabbit models) demonstrating that ulcers treated with a collagen scaffold seeded with EPCs exposed to osteopontin (a protein involved in wound healing that has a 28.54-fold reduced expression in people with diabetes) are more likely to close than those treated with a bare collagen scaffold. Similarly, he found that administration of mesenchymal stem cells (MSCs; progenitor cells isolated from the bone marrow that can become [among other things] adipose, bone, and cartilage) seeded on a collagen scaffold resulted in a significantly higher wound closure rate than a bare collagen scaffold, and that (in a mouse model) intramuscular delivery of MSCs may increase blood flow in the ischemic leg (a leg that originally had restricted blood flow). Dr. O’Brien concluded by saying that 67 people have received MSC injections for critical limb ischemia (mainly in phase 1 trials). So far MSC injection appears to be safe, and there is some evidence of its efficacy.



Young Sup Yoon, MD, PhD (Emory University School of Medicine, Atlanta, GA)

Dr. Young Sup Yoon continued the symposium, describing the potential for stem cells to treat diabetic neuropathy (a condition, he stated, that accounts for 20% of all hospital admissions among people with diabetes in the US). The majority of Dr. Yoon’s presentation focused on results from rodent studies (which he admitted have limitations), but demonstrated that, at least in rodents, endothelial progenitor cell (EPC) transplantation improved nerve conduction velocities in diabetic mice (p <0.05), and increased blood perfusion (p <0.001) and functional capillary density (~30 vessels in mice models treated with EPCs vs. ~20 in those treated with saline; p <0.001). It also appeared to restore neural vascularity (blood vessel development and functioning). Dr. Yoon said EPC treatment appears to work by EPCs differentiating into endothelial cells resulting in vasculogenesis (or formation of new blood vessels from progenitor cells), and increasing multiple angiogenic (promote the growth of new blood vessels) and neurotrophic factors (growth and survival of neurons), including VEGF-A. Similarly, Dr. Yoon presented research showing that injection of bone marrow mesenchymal stem cells (BM-MSCs) into rodent models improved motor and sensory nerve conduction velocities (p <0.05) and neural vascularity over saline injections. Additionally, rats with diabetes treated who received a BM-MSC transplantation had the same level of Schwann cell apoptosis as normal rats, and had restored levels of angiogenic and neurotrophic factors (including VEGF-A). He concluded by stating that stem cells could be an effective treatment for treating critical limb ischemia and diabetic foot, but that more work needs to be done to determine how well these effects are maintained long term.



Symposium: Diabetes and Cancer: Making Sense of the Data


Jeffrey Johnson, PhD (University of Alberta, Edmonton, Canada)

In his presentation, Dr. Jeffrey Johnson set a framework for the assessment of cancer incidence. He noted that the difficulty in studying the association between diabetes and cancer stems not only from the complex nature of both diseases, but also from a constellation of other factors including study biases   and possible confounding variables. After briefly comparing randomized controlled trials and observational studies, he listed several factors that investigators should consider in order to more accurately characterize the relationship between diabetes and cancer: 1) whether the study investigates all cancers as a whole or specific cancer types – Dr. Johnson believes that the latter is more favorable since the magnitude of risk differs by cancer type; 2) the possibility of reverse causality; 3) the biologic mechanisms underlying the relationship; 4) common risk factors between diabetes and cancer; 5) potential confounding by indication – while most observational studies use a “even/never” classification (categorizing patients by whether they have ever taken a drug or not), Dr. Johnson prefers accounting for the time-varying cumulative exposure of a medication; and 6) possible study biases, such as  detection bias and immortal time bias.

  • Dr. Johnson defined study bias as “any systematic error in a study that results in an incorrect estimate of the association between exposure and the risk of a disease.” Specifically, he noted that investigators should watch for possible detection bias (an artifact caused by the use of certain diagnostic techniques or the tendency on the part of the investigator to look more closely for an association in certain situations as opposed to others) and immortal time bias (when estimating drug exposure, the time between cohort entry and the person’s first drug prescription is described as “immortal” since the period is event-free [e.g., no death can occur for a patient to receive a prescription]. Incorrectly classifying this time period, or failing to include it in the data analysis, can inflate the perceived health benefits of a drug).
  • Dr. Johnson briefly discussed the ORIGIN trial results, which found no effect of insulin glargine therapy on cancer incidence (for our detailed coverage of the ORIGIN results, please see our ADA ORIGIN report at He applauded the authors for including cancer-specific data, noting that such information reassures that the risk of cancer did not differ between ORIGIN’s two treatment arms (standard care vs. insulin glargine). However,   he noted that due to the pattern of drug therapy used in both treatment arms, the determination  of cancer risk was neither clear nor simple. He reminded the audience that 11% of those on standard care received insulin (relative to 83% in the glargine group) and that participants in both groups used a range of oral anti-diabetic agents (such as metformin and sulfonylureas), each of which has been shown to have an association with cancer risk. Dr. Johnson stated that at best, the ORIGIN trial suggests that the specific combination of glucose-lowering therapies used in the glargine arm is not associated with a risk of cancer.



Andrew Renehan, PhD (University of Manchester, Manchester, United Kingdom)

After distinguishing between the two main types of mortality studies, Dr. Renehan emphasized that selection allocation bias and immortal time bias are major issues in many studies. In ever/never exposure studies (i.e., treatment vs. no treatment), the assumption is that patients in the analysis remain on the drug for the period of the protocol. However, in reality it isn’t that straightforward – some patients only have a short duration of follow-up, some have been on the drug beforehand, etc. Immortal time bias poses a challenge in ever/never exposure studies, as the studies do not account for duration of treatment.



Lonneke van de Poll-Franse, PhD (Tilburg University, Tilburg, The Netherlands)

Dr. Lonneke van de Poll-Franse presented a study on patient reported outcomes among colorectal  cancer (CRC) patients with diabetes. Data was obtained from the Eindhoven Cancer Registry, which collects information on patients’ sex, age, socioeconomic status, and co-morbidities at the time of cancer diagnosis, as well as data on the tumor stage and grade and the patients’ initial treatments. Her group’s study focused on patient reported outcomes (PRO) data from roughly 2,000 CRC survivors, among which 16% had diabetes at cancer diagnosis. To separate the effects of diabetes and CRC, the study included data from a normative population (760 people without cancer) matched for age, among which 12% had diabetes. Data revealed that CRC patients with diabetes had statistically significantly lower reported physical and cognitive functioning scores compared to people with only diabetes, only CRC, or neither disease (no significant difference was observed among the three latter groups for any PRO included in the presentation). Dr. van de Poll-Franse described the difference in physical and cognitive functioning as clinically significant. Similarly, people with both diabetes and CRC reported higher levels of fatigue, dyspnoea (difficulty breathing), erection problems, anxiety, and depression compared to healthy controls or those with only one disease. Dr. van de Poll-Franse ended her presentation by noting that greater investigation is required to determine the role of lifestyle and therapies for cancer and diabetes on PRO. Specifically, she mentioned that the recent inclusion of pharmacy data in the  Eindhoven Cancer Registry revealed that roughly 6% of CRC patients use metformin, and that such information will hopefully help investigators determine the impact of different diabetes therapies on  PRO in cancer patients.

  • Dr. van de Poll-Franse’s study included data from the Eindhoven Cancer Registry, which collects data from an area in the south of the Netherlands that hosts 2.4 million inhabitants and is served by ten hospitals and two radiotherapy institutes. Started in 1955, the registry tracks roughly 20,000 new cancer diagnoses per year, collecting data on patients’ sex, age, and socioeconomic status, as well as data on the tumor stage and grade and the patients’ initial treatments (e.g., type of surgery, radiotherapy, systemic therapy). Since the mid-1990s, the registry has collected information on patients’ comorbidities at the time of cancer diagnosis, and has gathered patient reported outcomes since 2004 from different cohorts of cancer survivals. Dr. van de Poll-Franse ended her introduction by noting that roughly 20% of CRC patients have diabetes at the time of cancer diagnosis, and that CRC patients with diabetes are treated less aggressively.
  • Dr. van de Poll-Franse concluded her presentation by noting that several questions remain, including those regarding the role of lifestyle and therapies for cancer and diabetes on patient reported outcomes (PRO). Specifically, she cited a recent article in JAMA reporting that metformin provides neurological benefits by encouraging new neuron growth and enhancing memory in mice (Hampton, JAMA, 2012). To Dr. van de Poll-Franse, such findings suggest a possible therapeutic value for metformin in patients with various neurological disorders. She then mentioned that the recent inclusion of pharmacy data in the Eindhoven Cancer Registry has revealed that roughly 6% of CRC patients use metformin, and that such data will hopefully help investigators determine the impact of different diabetes therapies on PRO in cancer patients.



Lucien Abenhaim, PharmD, PhD (London School of Hygiene & Tropical Medicine, London, United Kingdom)

Dr. Lucien Abenhaim presented results from the International Study of Insulin and Cancer (ISICA), a systematic case-control study assessing the relative risk of breast cancer associated with the use of individual insulins (glargine, lispro, aspart, and human), and the associated risk of insulin glargine with other insulin analogs and human insulin. The study found no difference in risk of breast cancer in patients using glargine or other individual insulins, and that the duration and dose of insulin glargine were not associated with increased breast cancer risk. Dr. Abenhaim noted that the study was powered to conclude that insulin glargine (for a mean duration of 3.2 years) had an odds ratio of less than 1.4; the power for subgroup analysis was limited.

  • This case-control study assessed the relative risk of breast cancer associated with the use of individual insulins, and compared the associated risk of insulin glargine with other insulin analogs and human insulin. Ninety-two large breast cancer centers  were identified in the UK, France, and Canada; of the 39,958 patients diagnosed from January 2008 to June 2009, 6.2% were identified as patients with diabetes. Of this subset, 41.3% were eligible to include in the study (age, cancer stage, insulin use, and glargine use were very similar  to patients who were not included in the study). Controls (with diabetes) were obtained from 582 general practitioner practices. Cases (n=775) and controls (n=3,050) were on average 52-53 years old; over 96% were on oral antidiabetic medications, and approximately 25% were on insulin (the average time to insulin initiation was 11.3 years). The proportion of patients in the case and control groups with A1c below 6.5%, from 6.6-8.0%, and above 8.0% were similar, as was the rate of complications.
  • Insulin use (of less than eight years in duration) was not associated with an increased risk of breast cancer. In addition, insulin glargine was not associated with an increased risk of breast cancer versus lispro (OR=0.85; 95% CI: 0.48-1.50), aspart (OR=1.10; 95% CI: 0.64-1.89), or human insulin (OR=1.29; 95% CI: 0.78-2.13).

Insulin <8 years

Adjusted Matched OR

95% Confidence Interval













Past use of any insulin (≥8 years)



  • The duration of dose of glargine use were not associated with breast cancer risk. Those who used glargine for less than four years had an adjusted matched OR of 1.15 (95% CI: 0.70-1.89), while those who used glargine for four to seven years had an adjusted matched OR of 0.94 (95% CI: 0.51-1.74). Those who used less than 27 units of glargine per day had an adjusted matched OR of 1.10 (95% CI: 0.61-1.97), while those who used at least 27 units of glargine per day had an adjusted matched OR of 1.02 (95% CI: 0.59-1.75).

Questions and Answers

Q: Was there any difference in outcomes in breast cancer patients treated with glargine and metformin versus treated with glargine but not metformin?

A: We did an interaction analysis, and there was no interaction with metformin. Metformin by itself was slightly protective; however, the non-effect of insulin glargine was the same in those with and without metformin use.



Helen Colhoun, MD, PhD (Public Health University of Dundee, Dundee, Scotland)

Dr. Helen Colhoun provided independent commentary on the International Study of Insulin and Cancer (ISICA), highlighting a number of strengths and weaknesses of the study. Strengths included: 1) the inclusion of 2,469 breast cancer cases of patients with diabetes; 2) good characterization of the tumor type; 3) the inclusion of reasonably extensive covariate data; 4) the attempt to model the cumulative effectives of exposure; and 5) contextualization of any effect of insulin glargine versus other insulins. Subsequently, she pointed out that the reliance on patient recall of prescriptions was likely to be highly inaccurate, especially in relation to the quantification of duration and dosage. In addition, she stated that random misclassification would bias the odds ratio toward the null. Dr. Colhoun also wondered  why study investigators went through such a laborious method to identify patients with breast cancer and diabetes, then obtain information on their medication use through recall, when there are currently so many good drug databases and disease registries. She stated that case-control studies have all the same problems as cohort studies: 1) those who are destined to receive the drug may be at higher or   lower prior risk of cancer; 2) negative allocation bias will hide a drug’s effect; 3) positive allocation bias will cause a drug to appear as though causally related to cancer when it is not; and 4) propensity scores and covariate adjustments cannot ever completely address this issue. Dr. Colhoun commented that it  was reassuring that there was no hint of any cumulative exposure effect, but the power to detect an  effect was limited; she considered the study as useful in adding to the body of evidence, but limited in the amount of new knowledge added.



Helen Colhoun, MD, PhD (Public Health University of Dundee, Dundee, Scotland)

Dr. Helen Colhoun proposed that epidemiological studies test for cumulative effects within users, since this method would not be subject to between-person allocation bias. She briefly reviewed a number of observational studies (Ruiter 2011, Lind 2011, Manucci 2010, Suissa 2011, Fagot 2012, Andersson 2012, and Boyle 2012) of insulin glargine and cancer that considered the effects of cumulative exposure, concluding that the studies suggest there is no cumulative effect of insulin glargine on total cancer. In closing, Dr. Colhoun stated: 1) there have been too few studies on the cumulative effects of insulin glargine on cancer risk to date, and that more data are needed; 2) the results are mixed, but mostly reassuring in terms of the effects of insulin glargine on colorectal cancer (nonetheless, it requires further examination); and 3) any future studies should report the effect of including time-updated terms for both ever use and cumulative use.

Questions and Answers

Q: I’m curious to hear your thoughts on some of the early signals and concerns for the new SGLT-2 class and its potential cancer risk. Insulin glargine has been out there for a long time, and that makes it possible to do a study with this type of rigor behind it. I just wanted to hear your thoughts on how we should think about a new class showing a potential link to cancer, and the importance of doing studies like this to give us confidence.

A: I think it’s really important. I think that it’s really difficult to ever get complete safety reassurance from RCTs, because most randomized trials of diabetes drugs are not going to have sufficiently long follow-up to be able to provide reassurance about every site-specific cancer. We will always need observational studies in addition. I do think that the three initial studies of insulin glargine in Diabetologia were flawed – they couldn’t look at cumulative exposure, and the effects were likely due to allocation bias. But, I think something useful they’ve done is to stimulate discussions about the appropriate methods in this field. What’s becoming a lot clearer now is the right way and the wrong way to analyze these data. We need to have multiple repeated studies in different populations to really show an effect as likely causal and real. As a community, we need to stop going off the deep end when someone reports an odds ratio above one – we need to take a step back and evaluate the methodology.


7. Type 1 Diabetes Treatments and Cure-Based Therapies

Oral Presentations: Intervention Studies in Type 1 Diabetes


Johnny Ludvigsson, MD, PhD (Linköping University, Linköping, Sweden)

Dr. Johnny Ludvigsson began his presentation on the two-year results from the Protégé study on the efficacy and safety of teplizumab by stating his beliefs that it is important to not throw away a potential treatment just because it does not meet its primary endpoint. As a reminder, the phase 3 Protégé trial failed to meet its primary endpoint (a composite of A1c and insulin dose), after promising phase 2 results. Encouragingly, Dr. Ludvigsson showed that while the 14-day teplizumab regimen (~17 mg; performed twice six months apart, n=207) had no significant impact on C-peptide preservation at one year (p=0.15) it was significant at two years (p=0.027). On the safety front, Dr. Ludvigsson presented data demonstrating that the 14-day regimen had fewer serious adverse events (SAE) than placebo  (11.1% vs. 12.2%, respectively) though substantially more people discontinued dosing due to an AE in the teplizumab arm than in the placebo arm (16.9% vs. 5.1%). During Q&A, Dr. Ludvigsson went into more detail saying that AEs, though not serious, were more common in children. Dr. Ludvigsson concluded Q&A by saying that while teplizumab is currently not the solution, anti-CD3s do appear to work, and that more research should be done on them. We are interested in seeing if other researchers  (particularly MacroGenics, manufacturer of teplizumab) agree with Dr. Ludvigsson and bring teplizumab “back to life” – during Q&A he did say that possible new studies are being discussed. The reality from our view is that primary endpoint does indicate success or failure of a trial and that companies are funded with this in mind; in this case, perhaps the dosing used was not aggressive  enough or the worry about the side effect profile was too much. We would certainly be interested in watching to see if this compound could be re-assessed, but believe the commercial reality will likely not allow this.

  • As a reminder, Protégé was a phase 3 trial (n=763) testing the efficacy of three different doses of teplizumab (14-day treatment, ~17 mg; 1/3 regimen, ~5.6 mg; and six-day regimen, ~4.6 mg) given two times, six months apart. Teplizumab failed to meet its primary endpoint, which was a composite of A1c and insulin dose (A1c <6.5% and insulin dose of < 0.5 U/kg/day) after one year (for more on these negative results please see our October 21, 2010 Closer Look at Since, these  results came out, there has been some debate about whether teplizumab truly ineffective, if the wrong endpoint was selected, or if the dose was lowered too much from the dose used in phase 2 trials (which was found to be efficacious but had a poor safety/tolerability profile). For an  example of this debate please see our 12th Annual Rachmiel Levine Diabetes and Obesity Symposium   report  at  (
  • Dr. Ludvigsson performed subgroup analysis by region, showing that people in the US had the best results (compared to those in Europe, Canada and Mexico, and India) and suggested it is likely because participants in the US on average had lower A1c levels (7.6% vs. 9.7% in India) and lower insulin doses (0.47 U/kg/day vs. 0.98 U/kg/day in India).
  • Dr. Ludvigsson then presented data from different subgroups showing that while  the 14-day regimen had no significant impact in adults (p=0.28), it was significant at year two in children and adolescents (age 8-17 years; p=0.03). Notably, the 14-day regimen was especially effective at year two in people who received the treatment within six weeks of diagnosis (p <0.01). We note that there is an ongoing study looking at using teplizumab preventively; that this subgroup did well gives us some (cautious) hope that this study will have positive results. Overall, significantly more people who received the 14-day regimen than placebo were receiving a low-dose insulin (<0.25 U/kg/day) while still having good glycemic control (A1c <7%), though this benefits appears to have declined over the two years.

Questions and Answers

Q: Did you monitor any markers to the immune system?

A: We did, and there were some correlations between teplizumab and changes in the immune system, but it is hard to draw any conclusions from the data.

Q: Are you doing any follow-up studies?

A: No. There are some discussions of doing some new studies, but nothing major.

Q: Were there no significant differences in adverse events among children?

A: Surprisingly, there was no difference in fever. If you look at skin reactions there were some differences, and there were some differences in the younger children, but nothing serious.

Q: Is this the way forward for treating type 1 diabetes?

A: This is a way forward, but this is not the final way, this is not the solution, but anti-CD3s seems to work. We know that you should be more careful about what patients you include. Participants should be younger, have lower A1c, etc. We need to know more about what is the right dose, if doses should be repeated, and if it should be used in combination with another therapy. We need to take the best from these results and move forward.



Mario Ehlers, PhD (Immune Tolerance Network, San Francisco, CA)

Alpha-1 antitrypsin (AAT) is an agent that has anti-inflammatory, anti-apoptotic, and potentially pro- tolerogenic effects that could be used as an immune-targeted intervention in type 1 diabetes. Part 1 of   the RETAIN study was a 12-month phase 1 safety and pharmacokinetic study including a preliminary examination of efficacy in 16 patients (eight adults aged 16-35 [mean age 21] and eight children aged eight to 15 [mean age 11]) with recent onset type 1 diabetes (<100 days disease duration). Patients were given six weekly infusions of AAT at a low dose (45 mg/kg), and after a three-week washout, they were given six weekly infusions of high dose AAT (90 mg/kg). For context, the approved dose used in patients with AAT deficiency is 60 mg/kg. At baseline, patients had an average of A1c of 7.9-8.0% with relatively well preserved C-peptide (0.9-1.0 pmol/l). The pharmacokinetic (PK) profile of AAT was similar to its previously published data, and no safety concerns arose. C-peptide did not significantly change from baseline to 12 months; this lack of decline (C-peptide in untreated individuals would be expected to decline about 0.3 pmol/year according to Dr. Ehlers) suggests that AAT might have a beta cell preservation effect. Part 2 of the RETAIN study will be a double-blind, randomized proof-of-concept  trial to further explore the efficacy of AAT.

Questions and Answers

Q: Is there a risk of developing antibodies against the protein that would prevent the full reaction?

A: We have samples for analysis but we haven’t done those analyses yet. In AAT-deficient subjects, they are rarely deficient to the point where protein is absent, so they will definitely have 50% AAT expression or more, and are given AAT treatment as augmentation. In those patients, years of treatment produces low incidence of anti-AAT antibodies. Although in the context of autoimmune disease, that is possible, and it would be bad to induce an immune response. We don’t have the answer.

Q: You are looking for a positive anti-inflammatory effect. Is that all? Is that better than any other anti-inflammatory treatments?

A: Good question, we know that anti IL-1 thus far has not worked. That presumably was going to be predominantly anti-inflammatory, and one could say anti-apoptotic, at the level of the islet. In some respects, AAT is targeting similar pathways. But what remains to be seen is if AAT can induce tolerance.

Q: Did you contemplate using anti-COX2 inhibition as another control?

A: No we didn’t.



Itamar Raz, MD (Hadassah Medical Center, Jerusalem, Israel)

Dr. Itamar Raz presented results from DIA-AID 1, a two-year pivotal phase 3 trial for DiaPep277. The study initially randomized 457 people with recently diagnosed with type 1 diabetes (within three months of screening) to receive either 1.0 mg DiaPep277 or placebo for two years (n=175 and 174, respectively   in the final intent to treat population). DiaPep277 treatment led to: 1) a statistically significant 23.4% relative preservation of glucagon stimulated C-peptide secretion over placebo (p=0.037) – the study’s primary endpoint; 2) no significant improvement in fasting C-peptide levels over placebo measured by mixed-meal stimulation; 3) a substantially (p<0.05) lower rate of hypoglycemic events in the per protocol population over placebo; 4) a significantly (p=0.035) greater percentage of individuals achieving a composite endpoint of A1c <7.0% and an insulin dose <0.5 units/kg/day in the completer population over placebo (41% vs. 30%); and 5) the number of hyperglycemic excursions/patient (blood glucose >140 mg/dl) was significantly (p=0.032) lower in the treatment group (11.5 vs. 14.4), as measured by six-day CGM at month 18 and 21. There was no significant difference between treatment and placebo group in incidence of treatment-related adverse events, overall adverse events, or serious adverse events. We were encouraged that the study met its primary endpoint of improving C-peptide response to glucagon stimulation, but we are concerned that there was no improvement in fasting C- peptide levels over placebo. We note that Diapep277 is now being further explored in a confirmatory phase 3 trial ( Identifier: NCT01103284), whose estimated primary completion date is December 2013.

  • Andromeda’s DiaPep277 is a synthetic 24 amino acid peptide derived from heat shock protein 60 (HSP60) that has anti-inflammatory effects. It was shown to prevent progression of type 1 diabetes in NOD mice, and Dr. Raz indicated that phase 1 and phase 2 clinical trials have been encouraging.
  • For background, we note that trials in children with recently diagnosed type 1 diabetes have found DiaPep277 to have no effect on the C-peptide decline (Schloot et al., Diabetes Metab Rev, 2007 and Lazar et al., Diabetes Metab Rev, 2007). This is probably because C-peptide levels decrease more quickly upon diabetes onset in children. Therefore, Andromeda has said that the current target population of the drug is newly diagnosed adult patients.
  • DIA-AID 1 is a two-year pivotal phase 3 trial for DiaPep277 in adults and adolescents recently diagnosed with type 1 diabetes (<3 months disease duration) with residual beta cell function (fasting C-peptide ≥0.22 nmol/l). The 457 participants were   randomized to receive nine quarterly injections of 1.0 mg DiaPep277 or placebo. In the final modified intent to treat population n=174 for placebo and n=175 for DiaPep277. At baseline, participants had an average A1c of 7.3%-7.5%, relatively high levels of autoantibodies at diagnosis, and an average insulin requirement of 0.42-0.43 units/kg/day.
  • The study achieved its primary endpoint; C-peptide area under the curve (AUC) declined less in patients given DiaPep277 compared to patients on placebo at month 24, when measured by glucagon stimulated test (change of -3 nmol/l vs. -4 nmol/l from baseline; p=0.037). Dr. Raz expressed surprise that the study did not achieve the related secondary endpoint of change in fasting C-peptide AUC from baseline to month 24 as measured  by mixed meal tolerance test (though it trended in the same direction). Additionally, 25% more patients in the DiaPep277 group achieved the combined target of A1c ≤7% and insulin dose of ≤0.5 units/kg/day than patients in the placebo group (p=0.035). A CGM substudy using the Medtronic iPro (n=78, performed at month 18 and 21 for six days each) found that the number of hyperglycemic excursions/patient (blood glucose >140 mg/dl) was significantly (p=0.032) lower in the treatment group (11.5 vs. 14.4).
  • Patients in the treatment group experienced less hypoglycemia than those in the placebo group (0.51 events/month/patient vs. 0.71 events/month/patient). There was no significant difference between treatment and placebo group in incidence of treatment-related adverse events, overall adverse events, or serious adverse events.
  • We note that Diapep277 is now being further explored in a confirmatory phase 3 trial, DIA-AID 2 ( Identifier: NCT01103284), which began enrolling patients at 100 medical centers in the US, Europe, and Israel in May 2010. The two-year study will ultimately include 450 people with recently diagnosed type 1 diabetes ranging in age from 20 to 45 years old. The study’s estimated primary completion date is December 2013.

Questions and Answers

Q: For those on less than 0.5 u/kg/day, how many injections were they given?

A: Most patients were either on a pump or four injections a day because they were newly diagnosed type 1 diabetes patients. What we measured was total daily insulin dose.

Q: Was the response of patients diagnosed at age 35 or older better or worse than the response of those diagnosed younger? How sure are you that you didn’t include LADA patients?

A: Very good question. I cannot say 100% that some patients were not LADA. We haven’t analyzed all the data yet, but we do find the response for those older than 20 was better than the response for those between the ages of 16 and 20. The group of those 35 years old plus was relatively small so that is hard to answer, and you’re right it might be that some of those patients have LADA.

Q: About the different effects on C-peptide AUC you got between glucagon administration and mixed meal test – perhaps beta cells in type 1 diabetes might have some functional defect e.g. incompetence toward the incretin effect that cannot be corrected by the therapy you tried.

A: I agree, and there is a recent publication showing that some patients with type 1 diabetes, especially young ones, have low GLP-1 levels, so the response to an oral test might be lower because of some defect in GLP-1. I agree with you, I think it has a lot to do with the GLP-1 system, which shows that we need to measure beta cell function in more than one way.



Odd Erik Johansen, PhD (Boehringer Ingelheim, Ingelheim am Rhein, Germany)

Dr. Odd Erik Johansen began his presentation by explaining that latent autoimmune diabetes in adults (LADA) is a slowly-progressing form of type 1 diabetes, and that adults with LADA are often misdiagnosed as having type 2 diabetes (it is estimated that 10-15% of type 2 diabetes cases are actually LADA). Dr. Johansen also reminded the audience that while current treatment options for LADA are similar to those used for type 2 diabetes, this treatment approach may need to be modified since insulin dependence occurs more quickly in people with LADA (due to a more rapid decline in beta cell function) than in people with type 2 diabetes. Dr. Johansen stated that linagliptin (Lilly/BI’s Tradjenta) could be a particularly effective treatment for LADA as preclinical research has found that the drug may slow beta cell death. Dr. Johansen therefore conducted a two-year, randomized study to compare linagliptin with glimepiride in people previously diagnosed with type 2 diabetes, who actually had LADA. Within his parent cohort (n=1,519 people with type 2 diabetes from 16 countries), Dr. Johansen found the  prevalence of LADA to be 7.8% (n=118). After 104 weeks, linagliptin was associated with a 202 pmol/l increase in C-peptide levels over baseline (944 pmol/l), while glimepiride was correlated with a 29 pmol/l decrease in C-peptide levels from baseline (1,374 pmol/l). Overall, the difference in C-peptide levels at week 104 between people on linagliptin and those on glimepiride was significant (p <0.001). Based off these findings, we wonder whether studies with investigate the use of DPP-4 inhibitors for people with type 1 diabetes, since both LADA and type 1 diabetes involve an autoimmune response against the pancreas.


Oral Presentations: The -Omics Frontier: Applications of New Technologies


Nancy West, PhD (University of Colorado, Aurora, CO)

Dr. Nancy West presented a preliminary study of DNA methylation analysis as a complement to genetic screening for type 1 diabetes. Using the DAISY database, her team retrospectively analyzed blood samples from 17 children who were diagnosed with type 1 diabetes shortly after the samples were taken and 17 age- and sex-matched children who did not get type 1 diabetes. The methylation changes  clustered around 16 genes – four that have previously been associated with type 1 diabetes or other autoimmunity disorders, two involved in insulin and/or glucose regulation, and 10 that may represent novel risk markers of type 1 diabetes. Dr. West was encouraged to find molecular markers of type 1 diabetes that precede diagnosis, though further studies will be needed to validate the results. In particular, the researchers plan to analyze older DAISY blood samples from the same 17 children with type 1 diabetes, to see whether the methylation changes are present at (or before) birth, or whether the methylation is a response to changes in insulin and glucose that precede type 1 diabetes onset by just a few months (in which case they would be much less useful for population-wide screening).

Questions and Answers

Q: The leukocyte population is known to be different between cases of type 1 diabetes and controls. Did you see whether the differences in whole-blood methylation might be due to this?

A: We have not studied this yet.

Q: We got a very messy picture with leukocytes, but we found a clearer analysis when we analyzed by type of monocyte.

A: Thank you for suggesting splitting monocytes into these groups.

Q: You said that preclinical changes in glucose control might confound the search for epigenetic predictors of type 1 diabetes risk. In earlier talks, the presenters on type 2 diabetes said something similar. Has anyone conducted a genetic comparison of type 1 and type 2 diabetes patients to try to isolate the methylation changes that are just due to glycemia?

A: That is an interesting strategy – thank you.

Q: DAISY has stored samples from further away in time, not so close to the diagnosis. Could you test samples from earlier on?

A: Yes – that is our plan. We have samples going back all the way to cord blood – before birth.


Oral Presentations: Role of the Immune System in Type 1 Diabetes


Antonio Toniolo, MD, PhD (University of Insubria, Varese, Italy)

Dr. Antonio Toniolo presented evidence that enterovirus infection might be a precipitating factor of type 1 diabetes. His group detected enterovirus (EV) genome fragments in peripheral blood leukocytes collected on the day of diagnosis in 76% of people with type 1 diabetes (n=119) compared to just 3% of controls matched for age and geographical location. Dr. Toniolo concluded that while this observational evidence cannot prove a causal relationship, the evidence is striking. However, during Q&A, audience members seemed a bit skeptical of the extremely low rate of EV infection in his control group. Whether a person develops type 1 diabetes due to an EV infection or not appears, to us, to have minimal implications on how the disease progresses; accordingly, no significant differences were observed between virus-positive and virus-negative patients with regard to age, sex, fasting plasma glucose,  blood pH, basal C-peptide, or insulin requirement at one month after diagnoses. Initially, virus-positive patients had a longer duration of pre-diagnosis symptoms, lower glucagon-stimulated C-peptide levels (suggesting they had poorer beta cell function), and higher A1c at the onset of disease, but one year after diagnosis, no significant difference in insulin requirement was observed, suggesting that there were no meaningful or durable consequences of these differences observed at disease onset Thus, we believe the role of EV infections in type 1 diabetes is probably most important for prevention rather than treatment.

Questions and Answers

Q: These results are rather different from what we’ve shown in Swedish studies. I have a question about controls because you had such low prevalence of EV positivity (3%). Who are they? How were they matched?

A: Of the controls, 60% were matched by age, geographic location and so on. These are children going to our pediatric endocrinology department for reasons other than diabetes or infectious disease. We selected those who were overweight and had short stature as controls. In addition we added another 20 or so  young blood donors (18-24 years old) to reach enough numbers for the control group.

Q: Don’t you find it remarkable that you found such low prevalence in the control group?

A: I find it remarkable compared to other studies, but we tested another 100 adult blood controls and could find only two with virus in their leukocytes.

Q: You are using a different sample of leukocytes than most previous studies, which use serum or plasma. Some previous studies have also shown high prevalence of EV at diagnosis. Have you compared, in your lab, the sensitivity of the assay that you are using depending on if it’s just serum or leukocytes?

A: Yes we did the comparison and feel that total leukocytes, not only mononucleocytes, is a more sensitive measure than serum or plasma. Plasma is also more sensitive than serum; serum is least sensitive.


Oral Presentations: Changes Over Time in Type 1 Diabetes


Olli Helminen (University of Oulu, Oulu, Finland)

Mr. Olli Helminen described his group’s study that tracked A1c and plasma glucose levels in children with beta cell specific multiple autoantibodies. The study included children with HLA-conferred genetic risk for type 1 diabetes who had developed multiple beta cell specific autoantibodies. The investigators compared glucose patterns between 195 children who had progressed to type 1 diabetes during the follow-up period to children who had not. A1c levels were measured every three months, and the children underwent an oral glucose tolerance test annually to determine their fasting and 2-hour post- meal plasma glucose levels. Data showed that in children who had developed diabetes, the mean A1c  and plasma glucose levels began to rise as early as one year before the time of diabetes diagnosis (data provided below); in contrast, A1c and glucose levels remained relatively constant in healthy children during the follow-up period. To Dr. Helminen, the results suggest that trends in A1c and plasma glucose levels may help predict the onset of type 1 diabetes in children with multiple autoantibodies.

  • Change in A1c level before diagnosis of type 1 diabetes in children with multiple autoantibodies

Time Period

Years prior to diagnosis

Mean A1c (%)


5.0-1.0 years



1.0-0.75 years



0.75-0.50 years



At diagnosis


  • Change in fasting glucose and 2-hour post-meal glucose before diagnosis of type 1 diabetes in children with multiple autoantibodies

Time Period

Years prior to diagnosis

Mean fasting plasma glucose (mg/dl)

Mean two-hour plasma glucose (mg/dl)


5.0-1.0 years




1.0-0.50 years




0.50-0.05 years




At diagnosis





Helen Colhoun, MD (University of Dundee, Dundee, United Kingdom)

Dr. Helen Colhoun presented a comparison (published recently in PLOS Medicine [])    of    mortality risk between Scottish adults with type 1 diabetes (20 years and older; n = 21,789) and the non-diabetic adult population of Scotland (n = 3.96 million) using the Scottish Care Information-Diabetes Collaboration database (years 2005-2007). As would be expected, she found that mortality rates are higher in people with type 1 diabetes than the general population (male incidence rate ratio [IRR] =  2.58; female IRR = 2.71), underscoring that people with type 1 diabetes continue to have increased type 1 diabetes despite improvements in treatment. When she broke the mortality IRR down by estimated glomerular filtration rate (eGFR), she found that (unsurprisingly) poor eGFR was associated with a higher mortality risk than people with better eGFR (IRR≈2 for eGFR >60 ml/min/1.73 m2  vs. IRR≈11 for eGFR<30 ml/min/1.73 m2; for context, >90 ml/min/1.73 m2 is considered a normal eGFR). Dr. Colhoun, however, argued that her finding that people with type 1 diabetes and healthier kidneys have a two-fold greater mortality risk than adults in Scotland without diabetes, counters Dr. Trevor Orchard’s (University of Pittsburgh, Pittsburgh, PA) hypothesis that excess mortality in people with type 1 diabetes is due to renal disease. This is because, she stated, if the excess mortality were due to renal disease, than people with type 1 diabetes who have healthy kidneys would not have such an increased mortality risk. We however, believe that people who have type 1 diabetes with eGFR >60 min/1.73 m2, probably still have microalbuminuria, which is what Dr. Orchard argues explains the excess mortality in people with type 1 diabetes. Though microalbuminuria is a marker for renal disease, it is also a marker for other factors such as general vascular health. Thus, looking only at eGFR (and thereby renal disease) does not account for his full hypothesis .

  • For her analysis, Dr. Colhoun matched data from the Scottish Care Information – Diabetes Collaboration database (SCI-DC) on estimated glomerular filtration rate (eGFR) and diabetes status of adults >20 years old with morality data from Scotland’s General Register Office (GRO). Data were collected from years 2005-2007. Unfortunately, she was not able to obtain eGFR data for the Scottish population without type 1 diabetes. Thus, she could not make a direct comparison between people with healthy kidneys and type 1 diabetes vs. people with healthy kidneys without type 1 diabetes, Instead, she compared incidence rate ratio (IRR) with the general Scottish population, regardless of the general population’s eGFR. The assumption made in this comparison (that the general population has a healthy eGFR) seems acceptable to us given it would skew data in a conservative direction, lending more strength to her argument: the IRR for people with type 1 diabetes and healthy kidneys is likely slightly lower than what she reported since the at large non-diabetic population will include people with renal disease, and therefore will have a higher overall death rate than if only people with healthy kidneys were used for the comparison.
  • When Dr. Colhoun broke mortality risk down by age she found that younger people with type 1 diabetes have a greater IRR than those who are older (20-29 years olds’ IRR ≈ 5.4 vs. ≥70 years olds’ IRR ≈ 1.8). She then broke mortality risk down by age and gender and found that the most at risk group was women 20-29 years old who had an IRR≈14 (men of the same age had an IRR≈3). Dr. Colhoun hypothesized that this difference was because these young women did not visit the doctor often enough, as they had unusually large time gaps between their last A1c test and when they died. We are curious about why young women would be so much less likely than men to see a doctor; in the US, conventional wisdom holds the opposite to be true, so we wonder what other explanations might explain this discrepancy.

Questions and Answers

Q: How would you compare this study to past studies of all the UK?

A: I think this demonstrates a real improvement in mortality risk.

Q: Interesting and intriguing data. Do you have plans for intervening in young women with diabetes?

A: We are doing a JDRF trial looking at adding metformin to usual care in type 1 diabetes patients. We would like to see if we can show changes in end organ damage. We hope that metformin will help patients lower their A1c with lower insulin doses. In terms of interventions for young females we haven’t studied any yet. I think one of the things this data suggests is that nationally we need to look at this population. It appears to me that this population may be removing themselves from getting care, based on a remarkably large distance back to their last A1c test.

Q: Your comparison of people with low eGFR was that against the whole background population or only people who had low eGFR?

A: The whole population, because we do not have the eGFRs of the background population.



Marie Louise Andersen, PhD (Herlev Hospital, Herlev, Denmark)

Dr. Marie Louise Andersen investigated how individuals’ rates of type 1 diabetes progression have changed over the past 30 years, as measured by stimulated C-peptide decline 2-15 months after diagnosis. To do so, Dr. Andersen compared rates of decline in four European cohorts (Wallensteen [years 1982-1985], Örtqvist [1995-1998], Hvidoere [1999-2000], and Danish [2005-2006]), and confirmed her findings against the TrialNet (2004-2009) cohort. (The TrialNet was only used for confirmation because it has stricter inclusion criteria than the other cohorts making direct comparisons more difficult.) Dr. Andersen found that the rate of stimulated C-peptide decline for a ten-year old child in the older cohorts was significantly slower than in the newer cohorts (p = 0.05 excluding TrialNet, p = 0.039 including TrialNet). Children experienced a mean 6.3%/month decline in stimulated C-peptide levels in the Örtqvist cohort, a 7.7%/month drop in the Wallensteen cohort, and a 7.8%/month reduction in the Hvidoere cohort; however, the rate of decline was 10.0%/month in the Danish cohort and 10.7%/month in the TrialNet cohort. These findings led Dr. Andersen to suggest that the disease’s initial rate of progression in individuals could explain the increasing incidence of type 1 diabetes globally (as a reminder, there is more than a 3% annual increase in the incidence of type 1 diabetes worldwide). We agree with Dr. Andersen that this may be a possible explanation, but note that in addition to being more recent, the Danish and TrialNet cohorts had higher C-peptide values at disease onset than the other cohorts. We think this and other inconsistencies between the cohorts (which Dr. Andersen did note)  could also explain the differences in disease progression that Dr. Andersen found.

Questions and Answers

Comment: When you compare these different cohorts, especially TrialNet, I don’t think you can use the TrialNet data because it is a very different type of cohort. It’s interesting to see what progress there has been, but I am not convinced we can say there is a difference.

A: I agree that it is a very different cohort. That is why we did the analysis both with and without it.



Valma Harjutsalo, PhD (Folkhälsan Institute of Genetics, Helsinki, Finland)

Dr. Valma Harjutsalo began her presentation by explaining that there has been an increase in the type 1 diabetes epidemic in Finland beginning in the 1990s. Around 1975, the incidence of type 1 diabetes in Finnish children younger than fifteen was “only” 2.8/100,000/year; however, around 2000, this rate jumped up to 41/100,000/year. Looking at data from several nationwide registers (The Hospital Discharge Register, The Drug Reimbursement Register, and The Drug Prescription Register) and  FinDM (the Finnish national diabetes research database) for the years 2006-2011 (n=3,302), Dr. Harjutsalo found the age-standardized incidence rate to be 61.5/100,000/year (67.6/100,000/year for boys and 55.1/100,000/year for girls). Dr. Harjutsalo noted that this represents 510 fewer cases in Finland between 2010 and 2011 than were projected using past rates. This drop in incidence was most pronounced for children between the ages of five and nine – in 2006 the incidence in this age group was estimated at 80/100,000/year but was 70.8/100,000/year in 2011. Dr. Harjutsalo concluded that this decrease in incidence could be because some environmental factors young children were commonly exposed to at the beginning of the millennium have abated. Another theory, proposed during Q&A, was that maybe environmental factors have not changed, but rather, the genetic susceptibility of the Finnish population has been met, such that everyone who is genetically susceptible of developing type 1 diabetes is actually developing it. We note, however, that this would only explain a plateauing of type 1 diabetes rates, not a decrease, unless one was to argue that genetic susceptibility for type 1 diabetes is  decreasing, which does not seem as likely to us as a reduction in environmental factors.

Questions and Answers

Q: Was there are also monitoring of other autoimmune diseases during the same period? Was this phenomenon that you so nicely described observed in other autoimmune diseases?

A: We have not studied this.

Q: We have recently analyzed the register from Sardonia, and the data is very similar. We have had a slight decline in incidence rate recently. Could we suggest that most people who are genetically susceptible to the disease have been affected by the disease, so that this is not a change in the environmental factors but a maxing out of people with genetic susceptibility?

A: Yes, I guess that could be it.

Q: What about obesity in children in Finland? There had been a weight increase year after year; is there any evidence that that has stopped?

A: No, there is no evidence of that.


Oral Presentations: Genetics of Type 1 and Type 2 Diabetes


Christiane Winkler, PhD (Helmholtz Zentrum Munchen, Munchen, Germany)

In this study, Dr. Christiane Winkler and her colleagues investigated whether single nucleotide polymorphisms (SNPs) in 12 type 1 diabetes gene regions could improve prediction of type 1 diabetes in first-degree relatives of individuals with type 1 diabetes (1,290 children from the BABYDIAB study, a study that followed children who had one parent with type 1 diabetes for a median time of 14 years for the development of islet antibodies and type 1 diabetes). Through the conduct of Receiver Operator Curve analyses, four genes (SH2B3, CTLA4, PTPN22, ERBB3) were found to be effective in discriminating islet autoimmunity, and five genes (INS, IL18RAP, IL2, CD25, IFHI1) were found to be effective in discriminating progression from islet autoantibody positivity to type 1 diabetes.

Questions and Answers

Q: Do you have any plans to replicate the study with an international cohort?

A: We are planning to replicate the study in other cohorts; maybe there will be some differences based on race and ethnicity.


Symposium: Should Healthcare Professionals Pursue a Mechanical or Biological Solution for Type 1 Diabetes?


Roman Hovorka, PhD (University of Cambridge, UK)

Dr. Roman Hovorka made a compelling case for the closed-loop, expressing his belief that it will revolutionize diabetes management and there will likely be widespread use of algorithm-directed   insulin delivery in the next one to two decades. The most notable update from his presentation was news that the Abbott FreeStyle Navigator II is now available in Europe – apparently, the device had a “silent launch” two to three weeks ago. We’re excited to hear this given the good accuracy of the original FreeStyle Navigator and look forward to hearing more and seeing how it compares to Dexcom’s G4 and Medtronic’s Enlite sensors. Also significant was an update on the three-week overnight closed-loop   home studies that are currently being conducted by the Cambridge group. The glucose traces from the first patients looked quite good and the study has recruited six out of a planned sixteen participants. Dr. Hovorka also gave broader thoughts on the state of the closed loop, highlighting that sensor reliability is the biggest limiting factor, followed by insulin absorption. He expressed concerns about  insulin/glucagon control and lamented that CGM reimbursement is “limited” in Europe. Dr. Hovorka also addressed the many advantages of closed-loop systems over a biological cure for type 1 diabetes, though he concluded that both approaches should be pursued and the closed-loop will be a bridge to a cure.

  • Notably, Dr. Hovorka announced that the Abbott FreeStyle Navigator II is currently available in Europe after a “silent launch” two to three weeks ago. He showed a picture of the device (“the first time it’s being shown”) and at first glance, it looked sleeker and smaller than the previous generation. We cannot find any information online about the new system other than the “completed” clinical trial listing on (NCT01455064), though we hope to hear more at the upcoming Diabetes Technology Meeting or perhaps ATTD 2013. We wonder 1) what the major changes are from the previous generation device; 2) what countries it is available in; 3) if the company plans to submit the device for FDA approval; and 4) why the launch was not announced.
  • Dr. Hovorka gave an update on the Cambridge team’s ongoing overnight home  study using the FlorenceD closed-loop system. Impressively, the study includes three weeks of overnight closed loop control. Six adolescents (12-18 years old) have been recruited out  of a planned 16. Dr. Hovorka showed graphs of the first couple patients’ glucose control, which were much flatter and more in range when the closed-loop system was used. However, there was one night where the sensor failed and the patient had to revert to open-loop control – this point  on CGM reliability was definitely a theme throughout his presentation. As a reminder, the FlorenceD system consists of the first generation Abbott FreeStyle Navigator CGM, the Companion (a device to assist communication with the Navigator), a small bedside tablet running an MPC algorithm, and a Dana R Diabecare pump (with a Bluetooth connection). The interface is quite simple and simply involves turning the system on and off. It targets a glucose range of 70- 144 mg/dl.
  • Dr. Hovorka believes the closed loop will revolutionize type 1 diabetes management in the next one to two decades, particularly because it has a number of benefits over biological approaches: 1) low biological risk because insulin is widely used, it has been tested for nearly 100 years, and can be used in all age groups; 2) the artificial pancreas has a scalability advantage – it is not limited by availability of islets; and 3) there is “enormous space” for innovation in closed-loop systems, as sensors could be reduced in size, systems could have more efficient wireless communication, better data sharing, etc. Further, changes in drug treatments require “tremendous effort,” while the innovation pathway for closed-loop systems is much cheaper, faster, and based on small iterative steps.
  • Compared to immunotherapy, there has been a more rapid flurry of recent closed- loop research and a higher proportion of these studies in humans. Dr. Hovorka plotted the number of immunotherapy and closed-loop research papers over the last 30 years, noting a few interesting trends. First, immunotherapy research has seen a slow increase over time, largely leveling off between 1990 and 2010 at about 80 publications per year. With the exception of a spike in the mid 1980s (the Biostator), closed-loop research (while numerically less) has accelerated exponentially in recent years, a trend Dr. Hovorka attributed to JDRF funding   starting in 2005. In terms of trials, he explained that the bulk of immunotherapy research is in vitro and animal models, compared to closed-loop research that is now largely occurring in humans. Dr. Hovorka also emphasized that the translation of closed-loop research to humans is much faster than biological research.
  • Closed-loop systems have a much lower development cost than immunotherapies and drugs. Dr. Hovorka noted that new drug entities have an average bill of ~$1 billion (Vernon et al., Health Economics 2010), compared to $30 million for reusing existing pump/CGM technology (Animas and JDRF) or developing a new CGM device for ~$50-100 million or a new pump for ~$20 million.
  • The biggest challenge in closed-loop control is the reliability and accuracy of CGM. Dr. Hovorka showed a couple examples where CGM sensors failed and closed-loop systems were forced to revert to manual open-loop control. On the accuracy side, he also showed the data we first saw at ADA 2012 (see pages 32-33 of our full report at on the frequency of large CGM errors. In that study, the FreeStyle Navigator came out on top in a comparison vs. the Dexcom Seven Plus. Fortunately, the three next-gen CGMs currently available in Europe – the Medtronic Veo/Enlite, the Dexcom G4, and the Abbott FreeStyle Navigator II – all have a MARD below 15%, “good enough” to achieve closed-loop glucose control.
  • Moving forward, the main focus is moving to home studies with portable systems. Dr. Hovorka asserted that the pathway is usually through hotel and camp studies, though the Cambridge team has bypassed that stage and gone straight into the home environment. State of the art computer simulations will be a key part of this transition. Two transitional studies have occurred in the past year from the DREAM group (see page 18 of our ATTD report at and the iAP group (see pages 14-15 of our ATTD report at Besides the Cambridge team’s portable system, there is the iAP group’s Diabetes Assistant and Medtronic’s Portable Glucose Control System (according to Dr. Hovorka, this is being ported from BlackBerry to Android) – for more details on both systems, see our DTM 2011 report at
  • Dr. Hovorka outlined the innovation pathway going forward, noting that a reliable sensor is the “enabling factor.” Boxes in the table below denote how important the innovation will be for each particular area – more boxes means that feature will be more important for that particular dimension. The column for user acceptance of a dual hormone system was colored red due to Dr. Hovorka’s view that such a system will impose additional patient burden.





User Acceptance

Reliable Sensor




Faster Insulin




Dual Hormone



** (red)

Single Port




Integration/Human Factors




  • Concluding, Dr. Hovorka summarized the key remaining questions in closed-loop development: performance at home (especially behavior); how accurate and reliable CGM should be for a commercial product; who benefits the most; and which control approach is the best. Jokingly, he also noted “what we do know that we do not know” – much less compared to immunotherapy or cell-based therapy.

Questions and Answers

Dr. Thomas Danne (Kinderkrankenhaus auf der Bult, Hannover, Germany): Congratulations on the home study. I know that you are not a friend of remote monitoring and watching the patient. You had a sensor failure. How did this happen and can you discuss whether it was safe or not?

A: We used the Navigator, which has a calibration period one and two hours after inserting the sensor. For the second calibration, there was early sensor attenuation. In the internal diagnostics of the sensor, it detected something wrong and the CGM stopped working. The sensor gave no data and the patient switched to open loop, so we don’t have data on this period. But even if there had been communication  and monitoring of the patient, there would not have been anything to do about it. This patient also tends   to measure at 4 am.

Q: What will be the first more widespread clinical application of closed-loop technology? Will it be used more in special patient groups like pediatrics, or perhaps overnight or in someone with unstable control?

A: We will see different developments, and the system that makes it will have results that show substantial improvements in outcomes. Overnight is the one that is economically feasible to justify this system. However, companies need to push through and it’s a challenge for many companies to do that. In terms of the population, Dr. Thomas Danne is expecting that those with poor control, children, will benefit the  most. That’s probably the case. It’s more likely that those with good control and motivation will be those most likely to use it. I don’t know for sure but it’s interesting.



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

In response to the symposium’s overarching question: “Should healthcare professional pursue a mechanical or biological solution for type 1 diabetes?” Dr. Kevan Herold exclaimed, “We want both of course!” Providing addition nuance, he explained that devices are unable to replicate the beta cells’  ability to perfectly regulate metabolism. He then reviewed researched immunotherapies, and hypothesized that antigen specific therapies (such as immunizations for alum GAD65 [Diamyd], insulin [including oral insulin], and proinsulin) have been unsuccessful since once people are diagnosed with type 1 diabetes they have many different autoantibodies, which he doubts a single antigen can address.  In a more hopeful tone, Dr. Herold said that some therapies have been successful including: abatacept (CTLA4Ig), teplizumab (anti-CD3 monoclonal antibody), and rituximab (anti-CD20 monoclonal antibody). We found his inclusion of teplizumab in this list to be surprising at first mention, as the failed Protégé study is thought to have “killed” the drug, but Dr. Herold highlighted that the high dose of teplizumab was found in Protégé to preserve C-peptide levels significantly better than placebo. It is now fairly widely felt, we believe, that the design of the trial was not ideal – this is perhaps most challenging to see for patients given the market is most unforgiving and that few second chances are given. Dr. Herold noted that none of these biologics have achieved sustained improvements in beta cell function.    He stated this could be because these therapies kill the present autoimmune cells but do not modulate the immune system and/or there is a programmed death pathway in people with type 1 diabetes’ beta cells that is unaffected by immune modulation. Looking to the future, he suggested research should focus on selecting the right patients for a therapy, designing combination therapies, and improving tolerance. Finally, he stated that preventing type 1 diabetes is “the ultimate frontier” and recommended treating people who are at high risk (~90%) for type 1 diabetes as if they already have the disease, since early treatment could prevent or slow progression. That was really interesting to hear – it reminds us of the debate on the type 2 front over when people with pre-diabetes should begin to be treated. We are curious about exactly what drugs Dr. Herold would recommend; GLP-1 agonists or very low doses of insulin, of course, came to our mind (GLP-1 isn’t yet approved in type 1 but it only seems a matter of time given the increasing numbers of patients of high-profile doctors who are taking the drug).

  • Dr. Herold exclaimed that both mechanical and biological treatments for type 1 diabetes are needed, and sympathized with the challenges of designing a device as well as nature has designed beta cells. He described (in a tone of awe) how the islets of Langerhans delivers insulin to the right place, at the right time, in the right amount, and stops when insulin is no longer needed. He stated that human innovation has not yet been able to replace what nature designed.
  • Dr. Herold detailed three biologics he believes have been successful: 1) abatacept (anti-CTLA4 antibody), 2) teplizumab (anti-CD3 monoclonal antibody), and 3) rituximab (anti-CD20 antibodies). All three therapies initially had statistically successful preservation of beta cell function but eventually beta cell deterioration occurred for all three, even when there was continued administration. Dr. Herold called this the “type 1 diabetes problem”  and proposed that it is due to a failure of these therapies to modulate or eradicate pathogenic cells and/or a preprogrammed death pathway being present in the beta cells of people with type 1 diabetes that cannot be modulated by an immune therapy.
    • Abatacept (which is approved for rheumatoid arthritis) has been found to delay C- peptide decline by about 9.6 months.
    • Teplizumab, Dr. Herold said, has shown success in five trials, and though the Protégé trial did not meet its primary endpoint, it did show that there was C-peptide preservation and a reduction in the need for insulin in people treated with the drug (for more on the Protégé study’s one year results see our October 21, 2010 Closer Look at and for more on its two year results see our EASD Day #5 Report at Dr. Herold said that the AbATE trial (a study of teplizumab in recent onset type 1 diabetes, whose results are being submitted for publication now) found that teplizumab delayed the decline of C- peptide levels by 15.9 months.
    • Rituximab, similarly, showed initial improvements in C-peptide preservation and insulin use but had a decreasing efficacy over time.
  • Interestingly, he detailed a new assay his lab has developed that measures beta cell death through quantification of demethylated insulin DNA. We are excited to see  further results using this assay, as we think it could mark a major advance in quantifying beta cell viability, which thus far has only been measured indirectly in humans.
  • Looking to the future, Dr. Herold proposed that more work needs to be done on matching people with the right therapy, as “nothing works on everybody”. An example, he provided of this was that teplizumab is substantially more effective in children than adults. This, of course, recalls all the recommendations on the type 2 front to “individualize” therapy – it’s easier said than done, but we’re happy to hear the push for greater personalization.
  • He also pressed for more research to be focused on developing combination therapies and drugs that increase tolerance. He said that when mice models are treated with a combination of anti-CD3 monoclonal antibody and either anti-IL-1β (an antibody that blocks the inflammatory protein IL) or IL-1RA (a natural inhibitor of IL) there is a synergistic reversal of diabetes, such that more reversal occurs when the drugs are given in combination than the sum of each of their effects as a monotherapy. On the tolerance front, he recommended trying to increase the number or activity of cells that regulate the immune system (Tregs).

Questions and Answers

Q: I like your death assay, it is very important. However, one of the main ways cells die is by apoptosis, which is a clean death. Does it involve the release in DNA?

A: That is a very good point. Clean death would not result in the release of DNA and therefore would not be detected by our assay, but beta cell death due to type 1 diabetes is probably not clean.


Symposium: Beta Cell Transdifferentiation and Neogenesis (Sponsored by the EASD and JDRF)


Simona Chera, PhD (University of Geneva, Geneva, Switzerland)

Dr. Simona Chera gave a clear and well-paced talk about several mouse experiments using toxic ablation of beta cells to simulate the beta cell loss of type 1 diabetes. When adult mice had their beta cells destroyed, a fraction of their pancreatic alpha cells were gradually transdifferentiated to beta cells (going from glucagon-secreting, to glucagon- and insulin-secreting, to insulin-only-secreting). By six months after ablation, enough transdifferentiation had occurred for the mice to survive independently   of insulin (~4% beta cell regeneration was sufficient for this). Some mice even had normoglycemia restored (possible with ~15% beta cell regeneration). The potential for transdifferentiation was similar at two months (early adulthood), one year, and 1.5 years, and it was not confined to the period immediately after ablation. When mice are only two weeks old, beta cells regenerate even faster but do not arise from trans-differentiation of alpha-cells. Rather, prior to expressing insulin, these new beta cells expressed Ngn3 – the marker of the earliest known stage in pancreatic endocrine development.  This finding suggests that the new beta cells either came from the same progenitor cells active in fetal development or de-differentiation and re-differentiation of non-alpha-cell pancreatic tissue. Dr. Chera and her colleagues are now studying beta cell regeneration in the two-week-old mouse model in greater detail, and they are also working to translate the research to in vitro studies of human tissue.

Questions and Answers

Q: As a pediatrician I am longing for beta-cell regeneration in humans. Can you speak to this?

A: I think it remains to see whether alpha-to-beta cell conversion is possible in humans. We are actively working on this in culture.

Q: Was the architecture of the beta cells that came back in the pumps normal?

A: The mixture of beta cells within the first few months after ablation is heterogeneous and non- geometric, as opposed to the natural well-ordered arrangement. I do not know whether the architecture might be reshaped after a year or so of regeneration.

Q: How is the population of alpha cells maintained?

A: Only about 5% of alpha cells are turned into beta cells, so not many alpha cells are really lost.

Q: In the bihormonal cells, did you use electron microscopy to see if there are separate granule pools for insulin and glucagon?

A: I am not sure if this was done; it would have been before I came to the lab. From what I have seen with confocal microscopy, they seem to be separately distributed within cells.


Symposium: Is Type 1 Diabetes an Enterovirus Disease? (Dedicated to the Memory of Professor Keith Taylor)


Noel Morgan, PhD (University of Exeter Medical School, United Kingdom)

Dr. Noel Morgan began the symposium by trying to answer the question “Can new insights be gained into the molecular events that cause type 1 diabetes in humans by analysis of autopsy pancreas   sample?” To answer this question Dr. Morgan and his research team studied autopsy samples from two cohorts (total n=89). They found viral protein 1 (VP1, the major structural viral protein) in 44 of the 72 samples they studied from children who had type 1 diabetes, whereas in 50 samples from children without diabetes, VP1 was found in only three samples . (However, on average only ~5% of islet cells in people with type 1 diabetes were positive for VP1.) Believing that viral induction and activation of the immune response would generate a specific, detectable, “foot print” within islet cells, Dr. Morgan looked at the expression of protein kinase R (PKR, an enzyme that protects the body against viral infections) and double-stranded RNA (dsRNA is produced by enteroviruses but not human cells). He found that  87% of the cells that were positive for VP1 contained PKR and that dsRNA is present in beta cells of people with type 1 diabetes. Dr. Morgan concluded that both these signals are consistent with the “foot print” of previous enteroviral infection in people with type 1 diabetes. During Q&A he explained that the cause of type 1 diabetes is probably not the virus itself, but rather an inappropriate immune response to the virus.

  • Autopsy pancreas samples from two cohorts were used for this study. One was from British children who died shortly after being diagnosed with type 1 diabetes (n=72, mean age 12.6 years, mean post-diagnosis survival ~8 months). The other, from the National Network for Pancreatic Organ Donors with Diabetes (nPOD), contained samples from Americans (mean age 25.7 years) who survived longer after diagnosis (mean ~11.6 years). Dr. Morgan said that one advantage of the latter cohort is that sample processing has been more strictly controlled.
  • Dr. Morgan identified several different markers that are consistent with beta cells sustaining an enteroviral infection. First, he found protein kinase R (PKR) in 87% of the cells that were positive for viral protein 1 (VP1), corroborating that these cells had been infected  by a virus. PKR fights viral infections by arresting translation so that viruses cannot replicate. But this also slows the cell’s own translation, which could contribute to beta cell death. Additionally, Dr. Morgan found that myeloid cell leukemia sequence 1 (Mcl-1), an anti-apoptotic (programmed cell death) protein is selectively depleted in islet cells expressing VP1 and PKR. He hypothesized that the increased levels of PKR in beta cells might degrade MCl-1, increasing sensitivity to pro- apoptotic stimuli and resulting in beta cell loss. Also found in the beta cells of people with type 1 diabetes were double stranded RNA (dsRNA) and melanoma differentiation-associated protein 5 (Mda5), a cellular sensor for dsRNA. Since viruses create dsRNA but human cells do not, these findings also indicate that the beta cells had been infected by a virus.

Questions and Answers

Q: There is always an argument that what you are showing is a very late-stage event. Have you looked at the pancreas of people who are at risk for type 1 diabetes, but do not have it yet?

A: We have not been able to do that yet, but I believe the nPOD [National Network for Pancreatic Organ Donors with Diabetes] collection will make that possible.

Q: Is it fair to say that the majority of people who have these viral infections will not develop type 1 diabetes?

A: That is correct. We are not saying that virus per se is the causal agent, but how people respond to the infection. This is also what genetic studies suggest.



Matthias von Herrath, MD (La Jolla Institute for Allergy and Immunology, La Jolla, CA)

Building on Dr. Morgan’s talk in the same session, Dr. Matthias von Herrath offered further perspective on the relationship between viral infections and type 1 diabetes. One objection to the relevance of viruses is that only ~5% of beta cells in people with type 1 diabetes are positive for viral proteins. But Dr. von Herrath proposed that type 1 diabetes might develop over the course of several relapses and remissions, with each relapse causing an increase in autoantibodies and more beta cell destruction. Complicating  the picture further, Dr. von Herrath described how some viral strains might cause diabetes while others might prevent it. He showed that in mouse models, more severe infections tend to accelerate the development of diabetes while less severe infections can abrogate diabetes. Based on these findings Dr. von Herrath said that a vaccine could be used to prevent serious infections that cause diabetes, potentially reducing their severity enough to even make them protective against diabetes. He also suggested that risk of autoimmunity could be reduced by exposing children to pathogens that are no longer common in our hygienic environment, such as parasitic worms – a bold approach to preventive care that we think would get a mixed reception from parents.

  • As Dr. von Herrath explained, viruses that infect beta cells can cause a strong inflammatory response, including the upregulation of major histocompatibility complex 1 (MHC 1). In mouse models with virally induced type 1 diabetes, the upregulation of MHC 1 is a prerequisite for beta cells to be recognized and destroyed by CD8+ (killer) T cells. In normal mice, MHC 1 levels go back to normal shortly after the mouse is infected, but in diabetes model mice MHC 1 levels stay elevated. Similarly, humans with type 1 diabetes can have elevated MHC 1 levels for up to eight years after diagnosis. However, this elevation is not always due to an inflammatory infiltrate (like a virus), which makes Dr. von Herrath wary of attributing causality.
  • Citing the hygiene hypothesis, Dr. von Herrath explained that the more infections   an immune system encounters, the more tuned it becomes to fighting off an  infection and then returning to its normal state. Upon being infected by a virus, a person’s immune system’s regulators (T regs) are decreased, and the immune system is ramped up to fight off the infection. Soon after, however, T reg levels are reestablished, reducing the immune response. This is necessary, because a long duration of high immune activity could result in autoimmunity. The immune system gets better at turning itself on and off the more times it goes through the process, Dr. von Herrath stated. This implies that modern hygiene, while generally a good thing, interferes with the immune system’s ability to train itself. He proposed that some things that have “fall[en] to the wayside” might need to be therapeutically brought back. In particular he suggested the use of parasitic worms that do not reproduce or grow into full parasites; these worms could potentially tune peoples’ immune systems (and prevent autoimmunity) without making them sick. Though we see the therapeutic value, we question if parents would be comfortable with giving worms to their children.

Questions and Answers

Q: I am a little confused. If viruses protect from diabetes, as you have shown, why should we have a vaccine against them?

A: You would have to develop a vaccine to those viral strains that replicate to higher levels or are  associated with diabetes in humans, but not against strains that are not associated with diabetes. Also, the vaccine could reduce the severity of the infection so that it will be less deleterious. A good vaccine could cause a severe infection to either be less deleterious or even protective.

Q: Brilliant as usual. You mentioned the hyperglycemia expression of MHC 1 in beta cell. What about MHC 2?

A: We have not done this in detail at this point. In mice it has been difficult to attribute any MHC 2 to beta cells. In the humans we have not looked at class 2. Right now we are looking at subsets of class 1.

Comment: We had the chance to look for MHC 2m and we found it expressed in some endothelial cells.



Heikki Hyöty, MD, PhD (University of Tampere, Tampere, Finland)

In his presentation, Dr. Heikki Hyöty sought to answer the question, “Is there enough evidence to start the development of an enterovirus vaccine against type 1 diabetes?” His answer: yes. As evidence that enteroviruses cause type 1 diabetes, he cited that the odds ratio for people with type 1 diabetes testing positive for an enterovirus is about ten. Further, Dr. Hyöty noted that the seasonality changes in type 1 diagnosis rates parallel that of enteroviruses. He admitted, however, that neither of these facts signifies causality. In Dr. Hyöty’s view, developing an effective vaccine against a type-1-diabetes-causing virus will require more research on what enterovirus serotypes (sub-species) cause diabetes – essentially, it is impossible to develop a single vaccine against the more than 100 enterovirus serotypes out there. By comparison, the polio vaccine is effective because it targets the three enterovirus serotypes that cause polio. Looking at 41 enterovirus serotypes in the Finnish Type 1 Diabetes Prediction and Prevention Project (DIPP; n=183 case children plus two matched control children for each case child), Dr. Hyöty found that children with the coxsackie B virus 1 (CBV1) had an odds ratio of 1.6 for developing type 1 diabetes (p <0.02). A similar study (VirDiab; n=454) performed in a broader population from the European Union corroborated that CBV1 is correlated with an increased risk for type 1 diabetes. According to Dr. Hyöty, CBV1 is the most common enterovirus serotype in the United States and results in severe systemic infections in newborns. He estimated, however, that <5% of children who have such an infection develop type 1 diabetes. Dr. Hyöty believes that a vaccine could prevent more than 50% of diabetes cases, though he did not explain how he reached this number.

Questions and Answers

Q: Is there an interaction between risk genotypes and being positive for CVB1?

A: We are actually working on this question. There are statistical problems as we divide groups into smaller and smaller groups. There is some evidence that there is an association but we need a much larger cohort

Q: Is the fetal pancreas more sensitive to CVB?

A: The young are very sensitive to CVB because their immune system is still developing.

Q: Is there any correlation between type 1 diabetes and mumps?

A: There have been some publications on this topic. Overall I don’t think mumps can play a major role. One argument against mumps causing type 1 diabetes is that the incidence of diabetes has not decreased since the mumps vaccine was started.


Corporate Symposium: A Comprehensive Therapeutic Approach to Diabetes Management (Sponsored by Sanofi)


Jay Skyler, MD (University of Miami Miller School of Medicine, Miami, FL)

After reviewing (mostly disappointing) results from recent trials for candidate type 1 diabetes therapies, Dr. Skyler provided his thoughts on what the future holds, discussing the potential of therapies to stop immune destruction, preserve beta cell mass, and replace/regenerate beta cells. He noted that at ADA 2012, both canakinumab and Anakinra were found to be ineffective; meanwhile, abatacept was found to sustain improvements in C-peptide out to three years. Dr. Skyler commented that although there has been a lot of fanfare in the media about BCG treatment (Faustman et al., PLoS ONE 2012), the study size was too small to make any conclusions; he seemed somewhat skeptical of Stem Cell Educator therapy as well (Zhao et al., BMC Medicine 2012). Dr. Skyler proposed a combination therapy approach, that might include an anti-IL1B or anti-TNF, along with a short-course anti-CD3 or anti-CD20 followed by a GAD vaccine and/or oral insulin to induce an antigen-specific protective response, and potentially add GLP-1 or human islet peptide 2b (HIP2b) to promote beta cell function. With regards to prevention, Dr. Skyler stated that oral insulin might hold promise for those with high insulin antibody levels. In  addition, TrialNet has an anti-CD3 prevention study underway, and is planning to conduct an  abatacept prevention study. Dr. Skyler mentioned islet transplantation, nanoscale encapsulation, xenotransplantation, transdifferentiation, beta cell neogenesis/proliferation, and stem cell therapy as interesting areas of exploration.



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

Dr. Edelman provided a brief overview of technologies for type 1 diabetes care (insulin pens, pumps, continuous glucose monitors, and mobile health), and discussed where we stand in the development of an artificial pancreas. Notably, he proclaimed CGM as the most important advance in diabetes management for patients with type 1 diabetes since the discovery of insulin, given its ability to reduce some of the unpredictability in diabetes management. Regarding the artificial pancreas, Dr. Edelman stated that there is currently a device that has the ability to stop delivering insulin when a patient has very low glucose (Medtronic’s Veo, using its Low Glucose Suspend), but looking down the road, we ultimately need to shoot for a fully automated, multihormonal (insulin, glucagon, and maybe amylin) closed loop. Dr. Edelman noted that we will likely need faster acting insulin than we currently have in order to close the loop, mentioning a number of companies (e.g., MannKind, Halozyme) with such products in development.

  • Insulin pens: Dr. Edelman noted that in addition to being convenient and discreet, they also protect insulin from light, heat and agitation.
  • Insulin pumps: Pumps provide features that multiple daily injections (MDI) do not. One advantage of pumps is that there is the capability to set a variable basal rate, and to deliver a bolus in a dual wave if so desired. Some pumps can now communicate with glucose meters, and can actually be manipulated through the meters. Dr. Edelman noted that pumps are getting smarter, smaller, and more convenient to use, briefly showing Tandem’s t:slim.
  • Patch pumps: Dr. Edelman noted that the fact that patch pumps don’t have tubing is a plus, since he believes tubing can introduce errors. He said that he is currently wearing the OmniPod. Other patch pumps include Valeritas’ V-Go (already approved; intended for patients with type 2 diabetes on previously on MDI), the Jewel Pump, and Cell Novo.
  • Mobile health: Dr. Edelman mentioned the iBGStar as a good example of how blood glucose meters can become smarter and more involved with mHealth. With the iBGStar, data can be sent by email, and the device seamlessly connects to iPod Touch and iPhone devices. Dr. Edelman noted that mHealthSys is working on device that can not only analyze blood glucose, but also use the technology included in mobile phones (e.g., GPS, an accelerometer, etc.) to record  information about location, activity, and time. He acknowledged that mHealth is still in its infancy, and that many unmet needs remain, including automatic food recognition (he noted that mHealthSys is working to provide such a tool).
  • Continuous glucose monitors (CGM): Dr. Edelman proclaimed CGM as the most important advance in diabetes management for patients with type 1 diabetes since the discovery of insulin (he said if he had to go back to using NPH or regular insulin in order to keep his CGM, he would do so). He acknowledged that the technology is not perfect, but nonetheless, it can take some of the predictability out of diabetes management. An important feature he highlighted was the trend arrow, which shows you whether your blood glucose is increasing or decreasing, so you have a better idea of how much insulin to take. Although he thinks all people with type 1 diabetes would benefit greatly from CGM, he recognized that some people may not be ready for the technology, but thought it should always be discussed as an option.



Geremia Bolli, MD (University of Perugia, Perugia, Italy), Thomas Danne, MD (Children’s Hospital, Hannover, Germany), Steven Edelman, MD (University of California San Diego, San Diego, CA)

Dr. Bolli: What information could CGM add for patients who consistently check their blood glucose day and night using traditional meters?

Dr. Danne: I think that they help only if they take away burden from the patient. The solution in my mind is the overnight artificial pancreas, which I think is going to happen in the near future. I think we are close to that.

Dr. Edelman: The technology needs to be improved, but I think it’s valuable for most patients. CGM not only tells you what your blood glucose is, but you also know the direction and can use that trend to adjust to have a better postprandial result.

Dr. Bolli: How do you maintain motivation in patients with type 1 diabetes to take control of their diabetes over the long term?

Dr. Edelman: Just like individualizing therapy, we should individualize our approach for each patient. Part of the onus is on us providers not to browbeat patients, and not view high blood sugars as bad patients. We should continually offer support, and encourage them to use their surroundings, but it’s not easy.

Dr. Danne: The most important thing is not to lose touch with the patient. If we look at public transportation in Berlin, two of us could discuss how to get to the Mitte, to the middle of Berlin. We could go different ways, but end up at the same place. That’s the same way you should approach a patient. You can go different ways, but should stay in touch and discuss along the way if we meet the goals, and if any adjustments are needed. Respect for the patient is number one.

[Note: additional talks from this corporate symposium can be found in the “Incretin-Based Therapies” and “Insulin and Insulin Therapies” sections of this report]


8. Regulations and Reimbursement

Oral Presentations: Profiling Glucose and Clinical Trials

Jennifer Green, MD (Duke Clinical Research Institute)

This study aimed to characterize the diabetes-related clinical trials registered in the dataset. Of the 96,346 studies in the dataset downloaded on September 27, 2010, 2,484 intervention trials were identified as diabetes-related studies. A majority of the trials were funded by industry (50.9%), followed by “other” (predominantly academic institutions) (41.6%), and the NIH (7.5%). The primary purpose of the majority of the studies was treatment of diabetes (74.8%); only a small proportion focused on prevention (10%). The majority of the trials assessed drug treatment (63.1%); a much smaller proportion investigated behavioral interventions (11.7%). Only a small percentage of trials targeted those who were 18 or younger (3.7%), or 65 or older (0.6%). Of all the registered trials, 40.5% of studies only had sites in the US, 49.7% only had international sites, and 9.8% of studies had both US and international sites. In closing, Dr. Green suggested that recently registered trials may not sufficiently address important diabetes care issues of involve all affected populations, and hoped these findings would be meaningful in making allocations for future research and resources.


Symposium: Management of Type 2 Diabetes: The ADA/EASD Position Statement


David Matthews, MD (University of Oxford, Oxford, United Kingdom)

Dr. David Matthews laid a foundation for the session on the new ADA/EASD position statement by discussing the document’s rationale, development, and process. He emphasized that the position statement is not intended to be algorithmic, but rather a resource to be used for support (“handrails for support,” “not railway lines”). The value of individualizing therapy was a big theme of this talk, though we most enjoyed his points about combining data from different trials (a bad idea) and the short- sightedness of targeting just A1c.

  • “Guidelines are not railway lines” that go in one direction and come to a decision point. Rather, Dr. Matthews urged that “they are better described as handrails to use as support” or “perspective lines” that give a three dimensional view from a position.
  • In discussing evidence based vs. eminence based guidelines, Dr. Matthews was critical of the American College of Physicians (ACP) guidelines for treating type 2 diabetesHe explained that the ACP guideline authors took a very strict, literature-review-based approach to writing the guidelines. Dr. Matthews facetiously boiled the wordy guidelines to three recommendations: 1) after diet use a tablet; 2) that tablet should probably be metformin; and 3) then add something else. This was followed by laughter and Dr. Matthews’ view that if you take a strict evidence-based approach, “you end up with things that aren’t helpful.” He emphasized that we must “use our brains as well as our computers” and combine the evidence base with expert opinion. We thought this was a somewhat ironic criticism of the ACP position statement, since  this three-step approach is effectively what is advocated for in the ADA/EASD position statement.
  • Algorithmic approaches have problems.” Dr. Matthews flashed the previous ADA/EASD algorithm and conceded that the just-released statement stands on the shoulders of preceding algorithms. However, he believes the mechanistic view taken by algorithms can lead to trouble. For instance, when one thinks in terms of combinations, testing all possible permutations for six different treatments would mean a 120-arm trial. Since data for all the permutations and combinations will never been known, it’s hard to take an algorithmic approach.
  • You cannot put these [trials] together in a huge pudding” – mixing trial data is “an absurd approach.” In what we thought was one of the best points in his presentation, Dr. Matthews showed a graph of A1c (y-axis) and years since diagnosis (x-axis). He then plotted UKPDS data and ACCORD data on the same chart – as a reminder, UKPDS started in patients at diagnosis while ACCORD started in patients 10 years after diagnosis. The A1c graph looked completely ridiculous, with steady upward lines denoting UKPDS A1c data from years zero to ten and then a straight downward line to an A1c of 6.5% once ACCORD started. Further, the approaches in these trials were completely different: UKPDS used monotherapy in patients that were newly diagnosed and complications free. By contrast, ACCORD targeted a very low A1c <6% and used high-risk patients. Dr. Matthews also emphasized that taking median values can be dangerous – for instance, it was critical to look at the interquartile range and spread in ACCORD, not just the mean and median.
  • Dr. Matthews questioned what to aim for when selecting pharmacotherapies. Lower A1c? – “but the ACCORD study showed that this is not always a good idea.” Preventing cardiovascular disease? – “loved by the regulators who like death and destruction measures” and “what about blindness, renal failure or stroke.” No hypoglycemia? – “but this is just a side effect.” Quality of life? – “but we give everyone metformin “which makes them feel unwell.”
  • In the spirit of the position statement’s overall message, Dr. Matthews concluded that an individualized approach to treatment is necessary. He showed the statement’s multi-colored graph that advocates choosing more or less stringent treatments based on a variety of factors: patient attitudes and motivation, risk of hypoglycemia, diabetes duration, life expectancy, comorbidities, and patient resources. Dr. Matthews underscored the cost issue using a wealth map of the world to demonstrate inequities in global resources – this was a key point emphasized throughout this entire session.
    • A “trial of one” approach may be valuable. Dr. Matthews believes this sort of trial and error approach can be valuable for deciding what therapies work in certain patients: drug A followed by drug B followed by a systematic assessment and discussion.



Silvio Inzucchi, MD (Yale University, New Haven, CT)

Dr. Silvio Inzucchi built on Dr. David Matthews’ presentation by more specifically discussing the position statement itself. Most interesting was his dialogue on the committee’s literature review process that concluded: 1) there is solid evidence behind metformin as a first-line therapy; 2) there is little evidence to support the choice of any particular therapy after that; 3) drugs generally lower A1c by the same amount (~1%); and 4) there are vast differences in cost and side effects between therapies. Dr. Inzucchi spent much of the presentation providing examples of how the individualized approach would work in practice – e.g., a patient with weight problems should be put on a GLP-1; a patient concerned about cost should use a sulfonylurea or insulin; etc. These examples were theoretically helpful, though their simplicity was put into a real-world context when Dr. Inzucchi described a “complex patient” with multiple comorbidities and health challenges. He concluded with some short comments on the GRADE trial, just funded by NIDDK to the tune of $134 million and slated to begin on October 1.

  • Dr. Inzucchi compared hypertension and diabetes drug development over the last 50 years. Interestingly, hypertension has had a steady stream of new drugs over the last five decades, with a new class of drugs every four to five years. By contrast, diabetes had only insulins and sulfonylureas between 1950 and 1995 (briefly biguanides before they were pulled form the market), with an absolute explosion of new drug classes in the last 15 years (approximately one new drug category every year or two). Dr. Inzucchi cautioned that we are still at a point in  diabetes therapy when newer drug categories are going to become available beyond the current 11 classes – in his view, there may even be another doubling of type 2 diabetes drug classes.
  • Dr. Inzucchi briefly reviewed previous algorithms, noting the most important milestonesThe 2009 ADA/EASD consensus algorithm was “a landmark algorithm because it encouraged metformin early in disease course.” However, it was criticized as being too algorithmic. Many felt the AACE algorithm was overly complex because it included too many medications; however, it was positive in that it channeled patients to use certain therapies based on A1c levels. The UK NICE algorithm, published in 2009, encouraged metformin as first-line therapy, sulfonylureas as a second line therapy, and then insulin. Interestingly, it also talked about the notion of individualization – e.g., for patients predisposed to hypoglycemia, DPP-4 is advantageous.