Keystone 2017 (Practical Ways to Achieve Targets in Diabetes Care)

July 13-16, 2017; Keystone, CO; Day #2 Highlights – Draft

Executive Highlights

The 27th annual Practical Ways to Achieve Targets in Diabetes Care conference, better known as “Keystone,” took place this past Thursday-Sunday in the breathtaking city of Keystone, Colorado. This document covers major highlights from day #2 of the meeting, a therapy-heavy day with one plenary on treatment algorithms and another on CVOTs. We also heard from nonprofit leaders on the role of their organizations in diabetes research and advancing care (including the Helmsley Charitable Trust’s Mr. David Panzirer and JDRF’s Dr. Aaron Kowalski), and we were treated to a most-valuable keynote address by ADA’s Chief Scientific, Medical, and Mission Officer Dr. William Cefalu.

Dive into all these highlights – 13 in total – below! In case you missed it, we also brought you a live report with our top ten highlights from Keystone day #1. See our full conference preview here for a look ahead at days #3-4 … we’ll be back with another highlights report on the final two days of the meeting before you know it.

Top 13 Highlights

1. In a rapid-fire overview of the most impressive new data from ADA 2017, Chief Scientific, Medical, and Mission Officer Dr. William Cefalu highlighted many exciting readouts, some of which included SGLT-2 inhibitors (namely, Merck/Pfizer’s candidate ertugliflozin), SGLT-1/2 dual inhibitors for type 1 (Lexicon’s sotagliflozin), the REMOVAL trial on metformin’s CV effects in type 1, biosimilar insulin, nasal glucagon, and Intarcia’s GLP-1 agonist candidate ITCA 650 (exenatide mini-pump).

2. What’s wrong with current diabetes treatment guidelines? According to Dr. Jay Skyler, an under-emphasis on lifestyle intervention, an over-emphasis on metformin, and insufficient differentiation between second-line therapies. He underscored that CV outcomes data should absolutely be incorporated into algorithms, especially with cardioprotective agents like empagliflozin and liraglutide on the market. This was a brilliant talk.

3. In his first Keystone appearance, Helmsley Charitable Trust’s Mr. David Panzirer made an impassioned plea for greater CGM adoption, suggesting that he would love to see 50% of type 1s wearing one by the time Keystone 2018 rolls around. Wow! No doubt ambitious, but Mr. Panzirer is of the mind that we have to shoot for the stars if we want to see change – bravo for this ambition and we hope the payers can be convinced.

4. JDRF’s Dr. Aaron Kowalski picked up where Mr. David Panzirer left off, with a strong push for greater CGM adoption, followed by a review of JDRF’s scientific priorities. From our view, we’d like to see even more evidence created on the impact of glycemic variability on long-term outcomes and longer-term, we think this can meaningfully change adoption.

5. In an emotional, much-anticipated evening hosted by Medtronic, we had the great privilege to hear from several patients regarding their experiences with the 670G. They gushed about how they can think less about their diabetes, spend more time in range, feel empowered, and no longer dread doctor appointments.

6. In his second talk of the day, Dr. Jay Skyler offered his view on how all the diabetes CVOT data published to-date should affect clinical practice: Stop using DPP-4 inhibitors and pay attention to surprise findings, including the amputation risk seen in CANVAS for J&J’s SGLT-2 inhibitor Invokana (canagliflozin) and the hypoglycemia benefit seen in DEVOTE for Novo Nordisk’s basal insulin Tresiba (insulin degludec).

7. Cleveland’s Dr. Steven Nissen gave a passionate defense of diabetes CVOTs, arguing that the 2008 FDA requirements led to “one of the greatest explosions of medical knowledge in the history of diabetes treatment.”

8. Through a review of ADA/EASD guidelines, Dr. Simon Heller concluded that written guidance is failing at its ultimate goal of helping people with diabetes reach their glycemic targets, although it’s a useful summary of the state of the field.

9. Dr. George Grunberger discussed the evolution of AACE guidelines leading up to their present emphasis on individualization of glycemic goals. He advocated strongly for earlier diagnosis and intervention to avoid later complications of diabetes.

10. Noting that CVD kills more than half of all patients with type 2 diabetes, Dr. John Buse recommended strategies for improving CVD risk factors in patients with type 2 diabetes, advocating for individualized glycemic targets, focus on tight BP control, and use of statins in those with overt CVD risk.

11. Dr. Simon Heller discussed the Lilly-sponsored IMPERIUM study, a phase 4 investigation of low hypoglycemia strategies vs. standard of care for elderly patients with type 2 diabetes, which was terminated early because of what may have been the wrong choice for primary endpoint.

12. The Barbara Davis Center’s Dr. Aaron Michels overviewed the strong association between type 1 diabetes and other autoimmune diseases, urging the audience of providers to screen for these accordingly. A whopping 25% of people with type 1 diabetes have at least one additional autoimmune condition, and notably, this statistic rises to nearly 50% in people type 1 diabetes for over 50 years.

13. BDC’s Dr. Paul Wadwa positioned telemedicine as the next major shift in diabetes management, but acknowledged possible barriers to implantation including reimbursement, provider skepticism, licensure across state borders, consent, and security.

Table of Contents 

Top 13 Highlights

1. “Best of ADA 2017” with Chief Scientific and Medical Officer Dr. William Cefalu

In a rapid-fire overview of the most impressive new data from ADA 2017, Chief Scientific, Medical, and Mission Officer Dr. William Cefalu highlighted many exciting readouts, some of which included SGLT-2 inhibitors (namely, Merck/Pfizer’s candidate ertugliflozin), SGLT-1/2 dual inhibitors for type 1 (Lexicon’s sotagliflozin), the REMOVAL trial on metformin’s CV effects in type 1, biosimilar insulinnasal glucagon, and Intarcia’s GLP-1 agonist candidate ITCA 650 (exenatide mini-pump). Clearly, the 77th annual Scientific Sessions were packed with exciting results on new therapeutic options, and Dr. Cefalu shared a positive outlook on how these agents will lead to near-term improvements in diabetes care. He articulated the “so what?” of all the clinical trial evidence presented on these therapies.

  • ADA 2017 featured several phase 3 readouts from the VERTIS program for Merck/Pfizer’s ertugliflozin. Dr. Cefalu acknowledged the SGLT-2 candidate’s compelling efficacy, but also suggested that due to the fact that other SGLT-2 inhibitors have shown evidence to reduce cardiovascular risk, it will have to show separation from other products in the class given the high hurdle the other products have established. Specifically, he pointed out that Lilly/BI’s Jardiance (empagliflozin) and J&J’s Invokana (canagliflozin) have already shown cardioprotection in the EMPA-REG OUTCOME and CANVAS trials, respectively. The bar for new diabetes drugs is higher than ever, and CV benefits are becoming increasingly important, especially for SGLT-2 inhibitors given speculation over class effects. The DECLARE CVOT for AZ’s Farxiga (dapagliflozin) is scheduled to complete in April 2019. VERTIS CV for ertugliflozin is also ongoing, and is expected to complete in October 2019 (Merck management shared during the company’s 1Q17 earnings call that the CVOT is now fully-enrolled, and is progressing on schedule). We’re excited by the prospect of a four-product SGLT-2 inhibitor class, as we think it will ultimately benefit patients by expanding choice. Moreover, given that canagliflozin was associated with a nearly two-fold risk for lower limb amputations in CANVAS, we’re eager to see continued development of agents in the SGLT-2 class. Additional data on other SGLT-2 inhibitors, including ertugliflozin, will help elucidate this potential safety signal – is it unique to canagliflozin, or does it have anything to do with the mechanism of SGLT-2 inhibition? And, if this amputation finding does negatively impact volume and sales of Invokana, we’ll be glad to see another product from Merck/Pfizer join the class, which has shown strong glycemic and weight loss efficacy overall. Dr. Cefalu noted growing enthusiasm for the SGLT-2 inhibitor class from patients/providers in the field, which was matched by the amount of data on these agents presented at ADA 2017.
  • Dr. Cefalu spoke about the data on sotagliflozin as an adjunct to insulin therapy in type 1 diabetes. He established that results from Lexicon’s inTandem1, inTandem2, and inTandem4 cumulatively support continued development of sotagliflozin. It is  particularly exciting that the agent may be one to be considered in the future as an adjunct therapy in individuals with type 1 diabetes.
  • In reviewing mixed results from the REMOVAL trial, Dr. Cefalu expressed his view that these findings shouldn’t and won’t meaningfully change clinical practice of using metformin off-label in type 1 diabetes. The study failed to meet its primary endpoint of mean carotid intima media thickness (cIMT), but did meet a tertiary endpoint of maximal cIMT. The significance of these findings on clinical markers aren’t really known as they represent surrogate markers of atherosclerosis. On the other hand, he emphasized that metformin promoted weight loss, reduced lipid levels, and modestly reduced daily insulin requirements, each of which contributes to the “glass half-full” perspective on REMOVAL. We found Dr. Cefalu’s commentary on this study reassuring. In concluding the ADA symposium on REMOVAL, Dr. Peter Rossing mentioned that the results do not support current treatment guidelines suggesting that metformin can improve glycemic control in patients with type 1 diabetes. However, metformin is arguably the most commonly-prescribed (off-label) adjunct therapy for type 1 right now, and we would hate to see this use drop off in a hasty reaction to REMOVAL, since many type 1 patients could still experience clinically-meaningful benefits from metformin therapy.
  • Dr. Cefalu underscored how biosimilar insulins offer similar benefits to branded products and the advantage would be a lower price. “It was encouraging to see  data on biosimilars being presented.” Given the growing controversy over soaring cost of insulin in the US, we definitely see an important niche for biosimilars to play on the market.
  • Dr. Cefalu established that treating severe hypoglycemia outside a hospital setting is a substantial unmet need in diabetes, and then discussed the ADA 2017 presentation on Lilly’s phase 3 nasal glucagon. He described the clunky reconstitution process of current injectable glucagon, which leaves huge room for error. Nasal glucagon offers a more patient/caregiver-friendly solution, and Dr. Cefalu reviewed efficacy data as well as positive feedback from caregivers on ease-of-use and satisfaction.
  • Dr. Cefalu captured the advantages of ITCA 650 over current GLP-1 agonist products by explaining how it circumvents the adherence issue (provides continuous, subcutaneous injection via an osmotic mini pump) and delays the need for other therapies. Two posters at ADA 2017 showed this delayed advancement to other diabetes drugs with ITCA 650 from analysis of the current clinical trial data. We’ve heard similar commentary from Intarcia CEO Mr. Kurt Graves, who underscored the added value that ITCA 650 could bring to the class by entirely avoiding the adherence issue that affects twice-daily, once-daily, and once-weekly injectable GLP-1 options. ITCA 650 is currently under FDA review, with a regulatory decision expected by December 2017.

2. Dr. Skyler’s Critique of Current Guidelines

What’s wrong with current diabetes treatment guidelines? According to Dr. Jay Skyler, an under-emphasis on lifestyle intervention, an over-emphasis on metformin, and insufficient differentiation between second-line therapies. In 1999, 44% of type 2 diabetes patients were at an A1c target <7%. By 2010, that number had grown only modestly, to 52%, despite advanced therapeutic options (40 new diabetes drugs since 1995) and evolving treatment guidelines. Moreover, Dr. Skyler shared that only 19% of patients between 2007-2010 achieved targets for A1c, blood pressure, and LDL cholesterol simultaneously. Clearly, there is much room for improvement in diabetes care, and that needs to start with better guidelines – we know all too well that busy healthcare providers (PCPs and endos alike) turn to professional treatment algorithms like the ADA’s Standards of Care to guide their practice, and it’s thus important for these algorithms to keep up with the latest clinical research. Dr. Skyler praised the ADA’s 2017 Standards of Care and the Canadian Diabetes Association guidelines for swiftly recommending GLP-1 agonist liraglutide (Novo Nordisk’s Victoza) and SGLT-2 inhibitor empagliflozin (Lilly/BI’s Jardiance) to diabetes patients facing high CV risk after the publication of LEADER and EMPA-REG OUTCOME. He argued that this is exactly the approach that guidelines should take, focusing on timely evidence and important clinical outcomes to rank the available treatment options, rather than listing six drug classes on one horizontal line for consideration after metformin (sulfonylureas, TZDs, DPP-4 inhibitors, SGLT-2 inhibitors, GLP-1 agonists, and basal insulin – as was the case in the ADA’s 2015 Standards of Care). Ideally, EMPA-REG OUTCOME data would have influenced algorithms even sooner, and CV effects would have been an even bigger factor in guidelines even earlier, but we’re glad to see these changes to match the growing emphasis on cardioprotection in the diabetes field. Dr. Skyler also spoke very favorably about the ADA’s and CDA’s recent emphasis on lifestyle intervention at all stages of diabetes management, simultaneous to any pharmacotherapy regimen. “The most crucial message I will tell you today,” he claimed, “is that lifestyle modification has to be the basis of all therapy. We do not put enough attention on lifestyle.” He discussed anecdotal cases in which patients prescribed lifestyle/metformin attribute all their glycemic and weight loss success to metformin, when in reality, healthy diet and exercise play a significant role. He went on to describe metformin as possibly “overrated,” and encouraged providers to support their patients in lifestyle modification at all stages of the disease. To drive this point home on the health benefits to lifestyle change, Dr. Skyler reminded everyone that UKPDS participants underwent intensive lifestyle intervention prior to randomization, and achieved a mean ~5 kg (~11 lbs) weight loss as well as ~2% A1c decline from a baseline 9% after three-six months.

  • Dr. Skyler criticized the American College of Physician (ACP) guidelines for the oral pharmacological treatment of type 2 diabetes, arguing that the 2017 iteration omits crucial CVOT evidence. This was a most-welcome critique, in our view. A number of thought leaders appeared to be uncomfortable with the 2017 revision when it was released in January, and we’re happy to hear from a well-respected thought leader that putting sulfonylureas and SGLT-2 inhibitors on equal footing is no longer acceptable, when there is compelling clinical information showing that SGLT-2 inhibitors can reduce a patient’s risk for CV death, effectively saving lives. Notably, sulfonylureas can cause CV harm, plus hypoglycemia, weight gain, and beta cell burnout, and we think it is quite negative for the field that the level of acceptability is still so high for this class, which could not come close to being approved today, in our view.
  • Dr. Skyler warned against relaxing A1c targets for older patients with diabetes – we were so happy to hear this since we worry about a “slippery slope” associated with such a recommendation. The ADA 2017 Standards of Care suggest a less stringent A1c goal for patients with limited life expectancy, but Dr. Skyler took issue with this recommendation: As we have questioned a number of times, what exactly does “limited life expectancy” mean? He alluded to 80-year-olds going skiing and mountain climbing, explaining how it doesn’t make sense – nor is it good diabetes practice – to determine glycemic control based solely on a patient’s age. “Let’s be careful where we decide to relax the goals,” he remarked, and we couldn’t agree more with this sentiment. While some have argued that elderly patients should aim for A1c >7% due to heightened hypoglycemia risk and frailty, we believe that for many patients, hypoglycemia risk can be managed with CGM, strong patient education, and other diabetes management support tools. No one, regardless of age, should be asked to compromise glucose control when there are strategies available to balance hypoglycemia risk (we particularly see a lot of untapped potential for CGM in the older patient population), and we’d love to see this language altered in the Standards of Care and other professional guidelines. We would also very much like to see more research on this front.

3. Helmsley Charitable Trust’s David Panzirer Calls for 50% CGM Adoption by Keystone 2018 in Impassioned Speech

In his first Keystone appearance, Helmsley Charitable Trust’s Mr. David Panzirer made an impassioned plea for greater CGM adoption, suggesting that he would love to see 50% of type 1s wearing one by the time Keystone 2018 rolls around. No doubt ambitious, but Mr. Panzirer is of the mind that we have to shoot for the stars if we want to see change - bravo! This portion of his talk is included in more details in the first bullet immediately below. Within a span of just over two minutes, he masterfully packaged T1D Exchange data showing that diabetes in the US is “neither safe nor managed”; called for CGM to be put on every single newly-diagnosed person; shared his hopes for the CGM landscape; compared diabetes in the US to Rwanda; and a whole lot more – definitely worth a read (or two!). We absolutely love the mindset shift he prescribed clinicians, suggesting that they ask themselves “why wouldn’t I put this patient on CGM,” instead of the traditional “opt-in” approach. Many KOLs appear to agree – CGM needs to become the default option in insulin users rather than the exception – and we believe that will happen as CGM becomes even easier to use, more accurate, etc. New technologies and drugs will only see adoption, of course, if they are prescribed. As an example of where CGM is going, Mr. Panzirer showed the slide of both generations of the Dexcom/Verily sensors, first shown at in Dexcom’s JPM 2017 presentation in January. Separately, Mr. Panzirer shared that Helmsley will be funding a prevention study in Europe beginning in late 2017/early 2018: 400,000 newborns will be screened at birth and, if found to be at-risk, put on a regimen of oral insulin beginning at three months of age for three years. As a reminder, the TrialNet oral insulin study presented at ADA 2017 demonstrated mixed results, with an overall negative signal, but a significant delay in diagnosis in a pre-specified secondary strata. This study will presumably have more statistical power, screening 400,000 babies, while the TrialNet study screened “just” 140,000, and will also (i) use much larger doses of oral insulin and (ii) intervene before autoimmunity. Mr. Panzirer added that the screening platform will eventually be used for many other diseases, not just type 1 diabetes – “This is massive and very expensive to do, it’d be insane to not think more broadly.” Read on for a history/description of the Trust since Mr. Panzirer became a Trustee, plus his view of his organization’s role in the diabetes ecosystem.

  • “Data from the T1D Exchange shows that we’re failing our patient population pretty miserably, yet we have access to a technology that can improve outcomes – CGM. But adoption remains at 24% in the T1D Exchange (and presumably far lower overall in the US and globally). T1D is the only disease I know of where patients and caregivers are making dosing decisions with a drug that could kill them.  They make these decisions 24/7 without the benefit of a clinician and if they get it right 60% of the time they are doing amazing! I don’t know of another disease that places so much burden on patients and clinicians. Think about how much things have changed in the past 10 years: When I first started, type 1 diabetes was completely misunderstood – people thought it was a safe and managed disease. But even today, 8.5% is the average A1c in the T1D Exchange population, and almost 10% of people have had severe hypoglycemia in the past year. That’s neither safe nor managed. And that’s in best clinics in US, with the best providers we have to offer. The common denominator to improve outcomes is CGM. CGMs should be put on every single newly-diagnosed person, full stop. Why? It’ll show them exactly what they need to learn. CGM’s not for everyone – we know about the issues related to form factor, cost – but seeing a naïve clinician should not be a barrier to adoption. We have proof that it works and just 24% of people wear it – that’s just not good enough. I get excited when I look at the future of CGM because a lot of the barriers, they’re coming down. Even FDA has realized these things are ready for primetime – Medtronic’s 670G automatically doses basal insulin off of it, and a fingerstick confirmation is no longer needed with Dexcom’s G5. We all need to be aligned and help get people on best tools. And the more people that get on devices, the more industry players will be brought in, and we will see innovation speed further. I recently traveled to Rwanda. In Rwanda, they couldn’t dream of  a technology like this – patients there have one strip per day, if they’re lucky, and they’re on premixed or human insulin. I could never go there and give a talk like this. But compare what we have here and we still don’t adopt it. At the end of the day, patients trust the person in the white coat – don’t let them down. Get them on the latest and greatest technologies. Next time you’re sitting next to a patient, ask yourself: ‘Why wouldn’t I put this person on CGM?’ Next year, I would love to see adoption at 50%. Everyone needs to join us and be advocates for these devices. Today we’re failing our patient population, but together we all hold the keys to change their lives.”
  • Helmsley is a $5.5 billion Charitable Trust employing 100 people that has given $1.8 billion to date and ~$400 million to type 1 diabetes. Mr. Panzirer shared the story of his daughter’s diagnosis and journey founding the Trust and T1D Exchange – see our 2014 interview for more on this incredible journey. Mr. Panzirer stated that one of the Trust’s goals is to improve the lives of all people living with diabetes. He continued: “Notice that the word ‘cure’ is not in the goal. We want a cure a much as anyone else, but cure is a fundraising word. We haven’t cured too many diseases, and if we do find a solution tomorrow, it’d still be 10-15 years until we could get it to patients. And it’s not for a lack of money, anyone who tells you that is full of it.”  So, our mission is to take better care of people now, so that when we do find a cure, they can participate.” (We love the alignment here with JDRF, who has also moved in this direction in the last decade, most notably with automated insulin delivery.) Mr. Panzirer emphasized that since Helmsley doesn’t have to fundraise, it is obligated to think differently, filling gaps in funding and taking on higher degrees of risk. Recent grants have included machine-learning predictive analytics for type 1 diabetes and a mental health provider diabetes education program.
    • Mr. Panzirer hit the road for 18 months of diligence before writing his first check at the Helmsley Charitable Trust. He spent time with members of the largely-successful Michael J. Fox and Cystic Fibrosis Foundations. He is particularly inspired by former CFF CEO Dr. Bob Beall whose work had a very tangible impact in his 30+ year stint at the foundation. In this time, the average lifespan for a person with Cystic Fibrosis swelled from ~7 to ~48! And in Mr. Panzirer’s view, almost every improvement in Cystic Fibrosis can be attributed to the Foundation, and this was accomplished by taking a business approach, aligning incentives for all stakeholders (payers, regulatory, and industry). See our 2014 interview with Mr. Panzirer for more perspective on this journey and the Helmsley Charitable Trust’s T1D program’s founding – for even more from the renowned Dr. Beall, see this interview with him and Ms. Margaret Anderson at Faster Cures last year.

4. Dr. Aaron Kowalski Commentary on CGM Adoption/Reimbursement; Reviews JDRF’s Scientific Portfolio

JDRF’s Dr. Aaron Kowalski picked up where Mr. David Panzirer left off, with a strong push for greater CGM adoption, followed by a review of JDRF’s scientific priorities. Mr. Kowalski noted that CGM adoption comes down to the age-old question: “If you build it, will they come?” In this case, it depends on the delicate balance between glycemic control and “diabetes happiness” (quality of life) – the Biostator, a bulky closed loop system from the 1980s may have driven perfect glycemic control, “but you can’t hike a mountain with it.” Of course, glucose monitoring has come a long way in the past four decades in terms of accuracy and form factor, stabilizing the teeter-totter of glycemia and quality of life more than ever before. Still, as Mr. Panzirer pointed out in the prior talk, only an estimated 24% of T1D Exchange patients in the US use CGM (far lower outside of the Exchange) – Dr. Kowalski asked that clinicians push for the newest and best advances for their patients so adoption of clinically proven technologies can increase. He also shared a unique perspective on the recent Medicare therapeutic CGM victory, acknowledging its monumental nature, but pointing out that for the population most vulnerable to severe hypoglycemia (seniors), CGM was not covered for more than a decade! Indeed, we are ecstatic that the Medicare population can now have access to potentially life-saving CGM, and we hope the access lag time shortens moving forward. On the topic of JDRF’s scientific portfolio, Dr. Kowalski focused on the automated insulin delivery program he spearheaded with former CEO Jeffrey Brewer. He enthusiastically exclaimed “we are here!” in reference to the MiniMed 670G launch, and told of amazing emails he frequently receives from patients detailing not only improved glycemic outcomes, but improved emotional wellbeing. Miniaturization/form factor and post-prandial control are JDRF’s top automated insulin delivery priorities at the moment, but multi-hormone is also notably high on the list – in Dr. Kowalski’s view, because hypoglycemia is so reduced in closed loop and postprandial excursions are “the biggest nut to crack,” amylin has tremendous potential. (Dr. Kowalski always emphasizes this dual-hormone combination more than glucagon.) As a reminder, Adocia hopes to enter clinical trials this year with a co-formulation of insulin lispro with pramlintide (see 1Q17). Lastly, he called the DIY community an incredible untapped resource of ingenuity in the type 1 diabetes community, and hinted that JDRF will be “weighing in more here” in the future – this caused some ears to perk up! We too believe that tremendous learning is happening in this community, and we hope the lessons and novel user experiences make it into commercial products! On the plus side, we’re starting to see this: Dexcom hired Ben West and Chris Hannemann from the OpenAPS community last year; Loop developer Nate Racklyeft now works on Apple’s Health team; Insulet shared at ADA that it has spent hours learning from the OpenAPS community; and Nighscout’s John Costik now works at Beta Bionics.

  • “We at JDRF often get criticized for working with industry. We can’t develop new products – we need companies to do this. We have to work with industry and see them take research developments and bring them to patients.” (This short-sighted JDRF-industry criticism was most recently brought up in Elizabeth Rosenthal’s book, American Sickness; see Jim Hirsch’s take in diaTribe here.) Dr. Kowalski pointed out in Q&A that many companies in diabetes take a long time – or don’t ever – turn a profit, and if they are driven out of business, patients lose. For automated insulin delivery specifically, Dr. Kowalski noted that many players are not on firm financial ground (e.g., Animas, Cellnovo, and Tandem come to mind, though he didn’t mention them), and part of the problem is that there are not enough people on devices – he firmly believes that outcomes on available systems (670G at the moment) will justify that more people should use diabetes devices. Many KOLs agree - we also hope that automation changes the conversation around diabetes technology from burdens to benefits.
  • Dr. Kowalski also touched on the glucose control (diabetes drugs), complications, prevention, beta cell replacement, beta cell restoration, and exercise (PEAK) programs in JDRF’s portfolio. Within these categories, he called glucose-responsive insulin the holy grail, spoke to funding into diabetic retinopathy and nephropathy, urged everyone to encourage participation in TrialNet, called the packaging of beta cells (for replacement) a big area of focus, noted that gene editing for beta cell restoration is “super hot right now,” and shared that the PEAK program (to help patients exercise successfully) has now rolled out 14 chapters across the US. There is certainly no shortage of opportunities to fund, and we’d have to guess JDRF’s job is getting more challenging every year!

5. Patients on 670G (and Two of Their Moms!) Share Heartening Stories and Impressions at Medtronic Symposium

In an emotional, much-anticipated evening hosted by Medtronic, we had the great privilege to hear from several patients regarding their experiences with the 670G. Diabetes CMO Dr. Francine Kaufman chaired the session, drawing invaluable insights from the panelists who differed by age, duration of diabetes, and use of different devices. One aspect they all shared in common, however, was their utter infatuation with and appreciation for the 670G. The impact of this device on both quality of life and glycemic control is undeniable, and we are SO happy to see this actionable technology making a difference in the hands real-life patients. With all the caveats we hear about managing patient expectations, we were particularly blown away by the recurring commentary that 670G allows them to not think about their diabetes for long periods of time, and then look down at their pump and be in range. Amazing! See below for some of our favorite quotes of the night. 

  • Quotable Quotes
    • The best part isn’t something you can quantify medically. The trust that I have since wearing the system has been life changing. I don’t think I realized how much of my life has been controlled by trying to be a well-controlled diabetic. My life has been drastically changed because anytime I want to do anything I know I can do it. I live alone and the only time I’ve ever had to call a paramedic for my diabetes was in the middle of the night, and it was the most terrifying experience. I haven’t had a low at night since I’ve put the 670G on. This pump is so well-tuned, I can eat dinner and 25 minutes later take my dog on a walk and not worry about it.” ­– Amanda (diabetes duration: 17 years; 670G use: ~3 months)
    • “We worry most about hypo. The first day I had the pump I was driving home and I was amazed at what was happening to my blood glucose. On my way home, I had to pull over because I was in tears because I didn’t believe I didn’t need to have this fear. My spouse has had therapy over my fear of hypo – to be able to have that reassurance, now my husband doesn’t even think about my diabetes.”– Dr. Gerard (diabetes duration: 30 years; 670G use: ~3 months)
    • “There’s something empowering about instead of going to the doctor to manage my diabetes, I’m finding out how to master it. – Amanda
    • I have no problem going to the doctor now. It’s hard when your A1c is not where it’s supposed to be and I think the doctor is going to yell at me. But even my mom can tell you my mood has changed about going to the doctor now.” – Ronny (diabetes duration: 16 years; 670G use: ~2 months)
    • “When I kept getting kicked out of auto mode in the beginning it was frustrating. After I figured everything out, it’s kind of second nature to get back into auto mode, it’s pretty easy.” – Kendrick (diagnosed 9 years-old; 670G use: 22 months)
    • “This pump has allowed us both to have some freedom and independence, you gain so much freedom and trust. She lives alone and my fears are my fears, but with this pump I can go to bed at night and know that she’ll get woken up with this low and I don’t have to worry about it, so thank you. [through tears]” – Ronny’s mother
    • Kendrick doesn’t just survive with diabetes, he thrives and I’m so thankful.” – Kendrick’s mother
    • “The single most important thing is that nocturnal hypo is no longer an issue. The control of hypo has greatly improved my quality of life. During the night, you’re flat.” – Dr. Gerard
    • “My kids no longer hear my pump beeping and say ‘mom are you ok?’ The 670G has changed not just my thinking but also my family’s. I don’t look at my pump most of the day because I’m not worried about my blood sugar.” – Shannon
    • “My overnights were perfect right away in the clinical trial. I have a perfect straight line at 120.” – Shannon (diabetes duration: 34 years; 670G use: ~2 years)
    • “It really eases the burden so much. At the end of the day we still have diabetes, it doesn’t go away without a cure. But this is definitely the easiest diabetes management I’ve had and I’ve been on four different kinds of sensors and two different pumps, I love it.” – Kendrick
    • I do a lot of my workouts at night so I have a lot of lows at night. Since I started the 670G, I’ve only had two to three lows, and it’s been huge.” – Ronny
    • “OMG you have to have this. #omg. I’m active on a lot of type 1 platforms and I’ve heard a lot that one device or another is the biggest innovation, that it will change your life. This is the only time I’ve felt like it really did. My brother hasn’t been on a pump, and this is the first he’s been really interested in.” – Amanda
    • “I come from using pumps and sensors and I didn’t like how they were constantly not in alignment, but with the 670 I’ve gone hours without thinking about diabetes and then I look and its 110. Last week I was at diabetes camp, and I was in auto mode 95% of the time and in range 83% of the time.” – Kendrick
    • “My third day with 670G in auto mode I was in range 100% of the time, I took a picture of it and put it on every social media post and thought: this was the day I’m never giving it up. – Amanda
    • “The first night in the hotel study [the pivotal trial] I ate an entire pizza and then I went, ‘oh wait this thing works.’” – Kendrick
    • I don’t have to pay any more attention during the day with the 670G than I used to, but because I sit all day at work I was having a lot more highs. After adjusting my insulin ratio it was fine. It took a little more tweaking, but almost every night is pretty perfect even if I’m high after dinner. I will wake up fine, which is remarkable.” – Amanda
    • “My patients need a refresher course on carb counting before using the 670G because the better you are at carb counting the better the 670G will work. It’s not a bigger burden than expected. There’s a little more room for error and we want to be more aggressive because of the reduced fear of hypo.” – Dr. Gerard
    • The visits to the doctor have become more manageable. I used to think it was such a hassle, but I think it goes more smoothly.” – Kendrick
    • “I like to talk to patients more about time in range than A1c.” – Dr. Gerard 
    • “The biggest hurdle for me is to figure out how much insulin I need when I have a hard workout. Leaving my pump on during soccer didn’t work out, so I’d resume it and immediately put it on right after the game. I have had a little trouble with exercise.” – Ronny
    • “I set the glucose target to 150 mg/dl before exercise and I’ve actually done extremely well with that. In the morning, I don’t need to do the temp, but you need to individualize this to your patients.” – Dr. Gerard
    • There have been times where I’ve told myself I don’t know if I can do this anymore, but I just kept going. I told my family I would do this for six months and I stuck through it and it’s been a lot better.” – Ronny
    • “I used to exit a lot, if I calibrate my blood glucose before bed I can’t remember the last time I exited overnight.” – Shannon

6. The CVOT Case Against DPP-4 Inhibitors

In his second talk of the day, Dr. Jay Skyler offered his view on how all the diabetes CVOT data published to-date should affect clinical practice: Stop using DPP-4 inhibitors and pay attention to surprise findings. Summarizing results from completed outcomes trials, he showed that no DPP-4 inhibitor agent has demonstrated cardioprotection, while some have shown a worrisome signal for heart failure hospitalization: In SAVOR-TIMI, saxagliptin (AZ’s Onglyza) was associated with a 27% increased risk of hospitalization for heart failure (p=0.007 vs. placebo). In EXAMINE, alogliptin (Takeda’s Nesina) reported a hazard ratio of 1.07 for heart failure hospitalization, which trended in the wrong direction even though it didn’t reach statistical significance (95% CI: 0.79-1.45). In TECOS, sitagliptin (Merck’s Januvia) reported a neutral hazard ratio of 1.00 for heart failure hospitalization (95% CI=0.8-1.20), but the FDA issued a Complete Response Letter (CRL) for inclusion of this data on the Januvia product label. A meta-analysis of these three CVOTs found an elevated heart failure hospitalization risk associated with DPP-4 inhibitors as a class (HR=1.13, 95% CI: 1.01-1.27), though this effect was clearly driven by the saxagliptin data. More importantly, in reviewing CVOT results on GLP-1 agonists (Novo Nordisk’s liraglutide and semaglutide) and SGLT-2 inhibitors (Lilly/BI’s empagliflozin and J&J’s canagliflozin), Dr. Skyler posed the question of why diabetes care providers would still consider DPP-4 inhibitors when our therapeutic arsenal now includes cardioprotective drugs. “Why use DPP-4 inhibitors at all when they show no CV benefit and have this issue of heart failure hanging over them? They’re not particularly cheap, either.” We’ve heard this position from Dr. Skyler before, at CMHC 2016, when Dr. Robert Ratner notably responded by defending DPP-4 inhibitors for their long history of safety/tolerability (which makes them especially useful for elderly patients and individuals with renal impairment). We see validity in Dr. Skyler’s argument, in that newer therapy classes like GLP-1 agonists and SGLT-2 inhibitors offer superior glucose-lowering efficacy vs. DPP-4 inhibitors as well as improvements to outcomes beyond A1c (CV morbidity/mortality, body weight, etc.). Some KOLs have pointed out that (i) DPP-4 inhibitors are oral agents and some patients may thus prefer them over injectable GLP-1 agonists, and; that (ii) generic DPP-4 inhibitors will likely become available prior to generic SGLT-2 inhibitors, making them a more affordable/accessible option for type 2 diabetes management. Still, Dr. Skyler’s emphasis on incorporating CVOT findings into real-world practice was invaluable commentary – after all, why conduct long, large outcomes trials if we’re not going to take that reliable evidence and use it to advance best practice diabetes care for patients? He added that providers, regulators, and guideline-writers should look to the (sometimes unexpected) non-CV results from these trials as well:

  • CANVAS, which just reported at ADA 2017, found a nearly two-fold risk of lower limb amputations associated with Invokana (canagliflozin). One of Dr. Skyler’s slides listed a hazard ratio of 1.94 for minor amputations of the toe or metatarsal (95% CI: 1.31-2.88) and a hazard ratio of 2.03 for major amputations at the ankle, below the knee, or above the knee (95% CI: 1.08-3.82). While he didn’t explicitly express an opinion on how Invokana should be prescribed from here on, he suggested that HCPs now have tough decisions to make surrounding SGLT-2 inhibitor starts and perhaps switching patients from Invokana onto Jardiance, which showed a significant CV benefit in EMPA-REG OUTCOME and which is indicated for the reduction of CV death. From our end, we’re terribly curious to see how CANVAS results will impact the Invokana business as well as prescription volume for the SGLT-2 class overall. Some experts, including Janssen’s Dr. Robert Cuddihy, have argued that the amputation risk will be manageable with proper foot care, more vigilance, and better education, and we hold out hope that future analyses of the holistic canagliflozin dataset may shed light on factors that mediate this risk so that patients broadly can still reap the cardioprotective benefits of canagliflozin medicines. In a way, we wish Dr. Skyler had shared his expert perspective on how HCPs should approach Invokana in the context of CANVAS results – the 14% risk reduction (p=0.0158 for superiority) for three-point MACE (non-fatal MI, non-fatal stroke, and CV death) vs. the hazard ratio of 1.97 in favor of placebo for lower-extremity amputations (95% CI: 1.41-2.75, p<0.001) – but perhaps this speaks to the incredible complexity of canagliflozin’s risk/benefit profile. For much more insight on CANVAS, read our coverage from ADA 2017.
  • A happier “surprise” finding appeared in the DEVOTE CVOT comparing Novo Nordisk’s Tresiba (insulin degludec) vs. Sanofi’s Lantus (insulin glargine). While Tresiba demonstrated CV neutrality, it showed a distinct hypoglycemia benefit vs. Lantus, reducing severe hypoglycemia by 40% (p<0.001) and severe nocturnal hypoglycemia by 53% (p<0.001). We thought these were very notable findings when presented at ADA 2017, and we were glad to hear Dr. Skyler endorse the importance of this clinical benefit. Hypoglycemia wasn’t the primary endpoint of this trial, but he advocated that the results are unambiguous and should be duly considered in treatment decisions. We agree wholeheartedly, and hope to see this take root in real-world clinical practice, just as it’s crucial for patients/providers to be aware when a diabetes therapy demonstrates significant CV benefit.

7. Dr. Nissen: CVOTs Offer an “Explosion of Medical Knowledge”

Cleveland’s Dr. Steven Nissen gave a passionate defense of diabetes CVOTs, arguing that the 2008 FDA requirements led to “one of the greatest explosions of medical knowledge in the history of diabetes treatment.” This extends not only to knowledge surrounding the CV effects of different drugs, but also to unanticipated benefits and safety signals – we were very moved by this characterization, and some have floated the idea that these studies shouldn’t really be called “CVOTs,” but rather “outcomes trials” more broadly. The renal outcomes data from both EMPA-REG OUTCOME (for Lilly/BI’s SGLT-2 inhibitor empagliflozin, branded Jardiance) and LEADER (for Novo Nordisk’s GLP-1 agonist liraglutide, branded Victoza) turned out to be real positives, providing important insight on the renal protective properties of these therapies. The nearly two-fold risk for lower limb amputations found in CANVAS (for J&J’s SGLT-2 inhibitor canagliflozin, branded Invokana) was a “particularly disturbing” unanticipated finding, according to Dr. Nissen. He reviewed the risk/benefit profile of canagliflozin as we currently understand it: In five years, we’d expect to see one fewer CV event per 43 patients treated vs. one excess lower-extremity amputation per 69 patients treated. Since these number-needed-to-treat values are in the “same ballpark,” Dr. Nissen explained that amputation risk is a serious consequence that should be very carefully considered by HCPs prescribing Invokana. He added that CANVAS and other CVOTs have refuted the idea of class effects. Invokana and Jardiance, for example, demonstrated exactly the same 14% risk reduction for three-point MACE (non-fatal MI, non-fatal stroke, and CV death), but only the former has a boxed warning for amputations on the US product label. Victoza has shown a distinct CV benefit, but two other GLP-1 agonists – AZ’s Bydureon (exenatide once-weekly) and Sanofi’s Lyxumia (lixisenatide) – have not. AZ’s DPP-4 inhibitor Onglyza (saxagliptin) showed a significant increase in heart failure hospitalizations, but this wasn’t seen for Merck’s Januvia (sitagliptin) and wasn’t significant for Takeda’s Nesina (alogliptin), even though it trended in the wrong direction (Dr. Nissen supported Dr. Skyler’s commentary on DPP-4 inhibitors from earlier in the day, emphasizing that they offer no CV benefit and are still rather expensive). Dr. Nissen established that all this accumulating knowledge from CVOTs since 2008 brings us to the “dawn of a new era in pharmacological management of diabetes,” in that we can now treat to reduce outcomes and prevent death. Evaluating therapies based on their ability to lower glucose is inadequate and unacceptable, Dr. Nissen argued, since A1c is only a surrogate biochemical marker for what we really care about (outcomes, mortality). We completely agree, and we echo Dr. Nissen’s call to the ADA, FDA, payers, and other players to recognize the critical clinical significance of cardioprotection and other benefits to outcomes beyond A1c. “Let’s integrate this knowledge into medical practice and overcome the residual inertia from decades of complacency.” Overall, this was a very important talk by Dr. Nissen and pointed out to many, including us, a number of the major advantages of this once-controversial group of trials (notably, for much time, we were particularly negative on the required timing).

8. ADA/EASD Guidelines: Failing to Help People Reach Glycemic Goals?

Through a review of ADA/EASD guidelines, Dr. Simon Heller concluded that written guidance is failing at its ultimate goal of helping people with diabetes reach their glycemic targets, although it’s a useful summary of the state of the field. Dr. Heller set the stage by describing the evolution of the ADA/EASD guidelines. The 2006 ADA/EASD Diabetes Treatment Algorithm recommended metformin as first-line therapy with later intensification using insulin or sulfonylureas – in 2009, second-line therapy options were updated to also include then “less well-validated therapies” like TZDs and GLP-1 agonists. In 2012, ADA/EASD pivoted from treatment algorithms to position statements, given the increasing array of available diabetes treatments. The 2012 ADA/EASD Position Statement outlined the risks/benefits of the major diabetes drug classes available at the time (metformin, insulin, sulfonylureas, TZDs, GLP-1 agonists, and DPP-4 inhibitors), emphasizing a “patient-centered approach” to decide which therapy would be most beneficial to an individual person, as diabetes is too complex and heterogeneous to be simplified into a universal algorithm. The Position Statement was updated in 2015, most notably to include SGLT-2 inhibitors and to recommend basal insulin/GLP-1 agonist combination therapy as an alternative to MDI. Dr. Heller surmised that the ADA/EASD’s guidance provides a comprehensive, up-to-date, and patient-centered summary of the landscape of type 2 diabetes treatment but does a poor job at actually producing improved outcomes for people with diabetes. Indeed, despite these many iterations of ADA/EASD guidance, people with diabetes are not doing much better today than they were previously, with 44% of type 2 diabetes patients at A1c goal <7% in 1999 and only 52% in 2010 – disheartening statistics, particularly considering the multitude of new diabetes therapies that have been developed alongside these more refined diabetes treatment algorithms. Dr. Heller speculated that this unfortunate trend may be due to the fact that most people with type 2 diabetes around the world are managed by busy primary care teams who may not have read the ADA/EASD guidance (or any of the multitude of other similar guidelines from other professional societies). He closed with a call for greater education of healthcare providers, particularly those in primary care. Many KOLs appear to agree – distilling the complexity and heterogeneity of diabetes management into an easily digestible form for busy PCPs is a daunting challenge, but a sorely needed task for educating the next generation of providers (who will encounter diabetes at an ever-increasing pace). 

9. Dr. George Grunberger Defends AACE Guidelines, Encourages More Aggressive Intervention Early On

The highly-esteemed Dr. George Grunberger discussed the evolution of AACE guidelines leading up to their present emphasis on individualization of glycemic goals. He advocated strongly for earlier diagnosis and intervention to avoid later complications of diabetes. In 2007, 567 evidence-based references were compiled into AACE guidelines that recommended an A1c goal as close to 6.5% (“normal”) as possible. The treatment algorithm divided patients into three groups according to starting A1c – 6.5%-7.5%, 7.6%-9.0%, or >9.0% –recommending more aggressive approaches for those with weaker glycemic control at time of diagnosis. In 2009, AACE added a caveat that the 6.5% A1c goal “may not apply to all patients,” and the 2013 algorithm added recommendations for weight management, lifestyle modifications, and consideration of CV risk factors. The most recent AACE guidelines feature further additions: the individualization of glucose targets taking life expectancy, disease duration, complications, and comorbidities into account; a goal of near-normal or normal A1c in recent-onset cases with no clinically-significant CV disease (if it can be safely achieved); and a less stringent 7%-8% A1c goal in patients with a history of severe hypoglycemia, limited life expectancy, renal or macrovascular disease, or longstanding diabetes in which they haven’t achieved glycemic goals. Moreover, the new 2017 treatment algorithm lists 14 principles, including personalization of both A1c targets and choice of therapies. Dr. Grunberger called attention to the growing importance of individualization – another new AACE guideline recommends target A1c ≤6.5% for patients without concurrent serious illness and at low hypoglycemic risk, but >6.5% for those with serious illness and hypoglycemia risk. Dr. Grunberger acknowledged that the ADA’s 2017 Standards of Care recommend <7% as a reasonable base A1c goal for non-pregnant adults, but defended the more stringent <6.5% as not too strict, especially for new-onset cases. Even patients with prediabetes, he explained, face increased risk for CV events, and diabetes complications progressively worsen in parallel to glycemic control, even if an individual hasn’t yet reached the A1c requirement for an official diabetes diagnosis. The DECODE study (n>22,000) found that 2-hour post-load glucose levels were associated with a linear increase in hazard ratio for all-cause mortality, while the Diabetes Prevention Program (DPP) investigation found diabetic retinopathy in ~8% of people with impaired glucose tolerance. Dr. Grunberger used this evidence to make an extremely compelling case for earlier diagnosis/intervention, and for more aggressive glucose-lowering early in the course of disease. This plenary on diabetes care guidelines (also featuring Drs. Heller, Skyler, and Umpierrez) ended with an engaging, insightful panel discussion – see our detailed discussion and commentary section below for a transcript.

10. Dr. Buse on Lowering CV Risk through Blood Pressure, Lipid Control

Dr. John Buse provided a comprehensive overview of the burden of CV disease in type 2 diabetes, discussing best practices for CV risk reduction and suggesting that hyperglycemia may not be as important as other risk factors (blood pressure, lipid levels, smoking). One study he presented showed that longstanding type 2 diabetes confers risk for MI equivalent to that of patients without diabetes who have already experienced an MI event. Noting that CV disease is the leading cause of morbidity/mortality for people with diabetes, accounting for 31% of diabetes-related deaths in 2013, Dr. Buse outlined the risk factors for atherosclerotic CV disease that are commonly comorbid with type 2 diabetes. CV disease can manifest as MI, transient ischemic attack, peripheral vascular disease, or even amputations, and risk reduction strategies should include lifestyle modification, platelet inhibition, management of dyslipidemia and hypertension, and individualized glycemic control. Smoking more than doubles CV risk in people with type 2 diabetes, and unfortunately, Dr. Buse shared that smoking rates in this patient population have not declined significantly in recent decades. He suggested that a risk calculator be used in prescribing antiplatelet agents to those facing high atherosclerotic risk, and also reminded providers to pay attention to aspirin (it isn’t prescribed and is often forgotten). Multiple trials have shown statins to reduce CV events in people with diabetes, and Dr. Buse explained that this benefit is seen with 30%-40% LDL-lowering regardless of baseline level. He articulated that lifestyle intervention should be a cornerstone of diabetes management, especially in the context of CV risk reduction, and that statins should be prescribed to anyone with high CV risk or overt CV disease. Turning to blood pressure, Dr. Buse described how UKPDS results showed significant reductions in mortality (32%), microvascular complications (37%), and stroke (44%) with tighter blood pressure control (≤150/85 mmHg). ACCORD indicated that a systolic blood pressure <140 mmHg confers near-maximum CV benefit, but Dr. Buse underscored that we don’t yet have the right clinical trials to know just how low blood pressure should be for optimal outcomes. Notably, a review of first CV-related hospitalizations for patients with type 2 diabetes found that only 6% had blood pressure under control. Considering all these results together, Dr. Buse urged HCPs to implement strategies to lower blood pressure and lipid levels in all patients with diabetes facing high CV risk, expanding the scope of treatment beyond A1c reduction.

  • Although it wasn’t quite the focus of his talk, Dr. Buse mentioned the importance of cardioprotective therapies in type 2 diabetes management. Speaking to the promise of newer agents with demonstrated CV benefit, he showed that only 39 people need to be treated with empagliflozin (Lilly/BI’s SGLT-2 inhibitor Jardiance) and 98 with liraglutide (Novo Nordisk’s GLP-1 agonist Victoza) for three years to prevent one death, compared to 100 people for five years with statins, 125 people for five years with anti-hypertensive agents, and 333 people for five years with aspirin. Dr. Buse also advocated for personalized A1c goals based on patient characteristics, preferences, and engagement. For individuals struggling with medication adherence, Dr. Buse suggested metabolic surgery for people with obesity not at glycemic goal and mentioned that Intarcia’s ITCA 650 (implantable exenatide mini-pump) as a good future GLP-1 option to guarantee adherence.

11. Dr. Heller on Low Hypoglycemia Strategies for Elderly Patients; No One Has to Compromise Glycemic Goal

Dr. Simon Heller discussed the Lilly-sponsored IMPERIUM study, a phase 4 investigation of low hypoglycemia strategies vs. standard of care for elderly patients with type 2 diabetes, which was terminated early because of what may have been the wrong choice for primary endpoint. Participants with type 2 diabetes age ≥65 (n=191) were randomized to one treatment arm with pharmacological interventions designed to avoid hypoglycemia (TZDs, metformin, GLP-1 agonists) or to another with standard of care (insulin glargine and a sulfonylurea, to start). After 24 weeks, the study found no significant difference on the primary endpoint of treatment success, or the proportion of patients in each group (n=98 in low hypo strategy, n=93 in traditional strategy) achieving A1c goal with no clinically-relevant hypoglycemia (defined as a period of low blood sugar that interrupts normal activity): 55% of the low hypo group and 65% of the traditional strategy group achieved treatment success (p=0.19). A1c decline was also similar in both IMPERIUM arms, dropping by a mean 1.1% in the low hypo group and by 1.2% in the traditional strategy group (from a baseline 8.3% across the study). That said, Dr. Heller pointed to impressive treatment differences on endpoints of asymptomatic hypoglycemia (only 8% in the low hypo arm vs. 32% in the standard of care arm, p<0.001), documented symptomatic hypoglycemia (5% vs. 39%, p<0.001), and total hypoglycemia (10% vs. 54%, p<0.001). No episodes of severe hypoglycemia occurred over the 24 weeks. Nocturnal hypoglycemia was numerically less frequent in the low hypo group (4% vs. 11%), although this association didn’t quite reach statistical significance (p=0.077). Dr. Heller remarked that IMPERIUM might have continued if the investigators had chosen a different primary endpoint, given that glucose-dependent therapies like TZDs, metformin, and GLP-1 agonists were associated with significant, meaningful reductions in asymptomatic, symptomatic, and total hypoglycemia. He advocated that these results still be used to inform diabetes care in older patients, a population that is particularly prone to hypoglycemia and to hypoglycemia unawareness. We’d love to see sulfonylureas eliminated from treatment regimens, definitely for patients at high-risk for hypoglycemia but also for patients with diabetes more broadly, given the weight gain and possible CV harm. We’re also curious how the low hypoglycemia strategies in IMPERIUM affected daily insulin requirements, and how insulin dose may have mediated the frequency of low blood sugar (although it also seems that older patients with longer duration of diabetes would be more likely to be on insulin therapy). Dr. Heller emphasized the importance of individualized glycemic targets, recommending to HCPs that they set an A1c goal <8.5% for elderly patients with frailty, but otherwise choose an A1c goal on par with what they would suggest to a younger patient (closer to A1c <7%). We were pleased to hear this clinical advice from Dr. Heller, since there have also been voices in the field advocating for less stringent A1c targets for elderly patients based on age alone. We’re opposed to the idea that any individual should have to compromise their desired glycemic control when there are glucose-dependent agents available (like those prescribed to IMPERIUM participants in the low hypo strategy), as well as CGM and educational tools that can help manage hypoglycemia risk.

  • Dr. Heller pointed to the irony that the vast majority of people with type 2 diabetes in the US are ≥60 years-old, and yet clinical research in this patient population is scarce. One literature search found that 289 of 440 investigations of type 2 diabetes treatments had an upper age limit (66% of studies), which Dr. Heller argued was rarely justified. Only six trials in this search were designed specifically to evaluate therapies in older adults. We certainly understand the major resources that go into conducting clinical trials, but we also appreciate Dr. Heller’s suggestion that the diabetes epidemic – costing the US healthcare system >$421 billion annually – is currently driven by older adults, and so more dedicated research in this population is more than warranted. Perhaps payers and regulatory agencies could better align incentives for industry to sponsor clinical trials specifically in elderly patients.

12. “No Disease is an Island”: Type 1 Diabetes and the Risk of Other Autoimmune Diseases

The Barbara Davis Center’s Dr. Aaron Michels discussed the strong association between type 1 diabetes and other autoimmune diseases, urging providers to screen for these accordingly. More than 25% of people with type 1 diabetes have at least one other autoimmune disorder, according to the T1D Exchange Clinic Registry. Thyroid autoimmunity (Hashimoto’s disease [hypothyroidism] and Grave’s disease [hyperthyroidism]) accounted for 68% of these comorbid cases, GI diseases (celiac disease and inflammatory bowel diseases such as ulcerative colitis and Crohn’s disease) accounted for 22%, muscle and joint conditions (rheumatoid arthritis, psoriasis, lupus, and scleroderma) accounted for 6%, skin diseases (vitiligo, alopecia, and dermatomyositis) accounted for 4%, and Addison’s Disease, a condition involving reduced cortisol secretion from the adrenal glands, accounted for 1%. Dr. Michels highlighted thyroid disorders and celiac disease as the “Big Two” that tend to cluster together with type 1 diabetes. He recommended increased screening for these autoimmune conditions particular (ideally, every two-three years in the absence of concerning symptoms). As for the likelihood of developing another autoimmune disease, Dr. Michels underscored that the biggest risk factor is time. Across the age spectrum, the incidence of autoimmune diseases in addition to type 1 diabetes steadily increases without an apparent plateau, rising from 20% in those age 0-13, to 35% in those age 26-50, to nearly 45% in those older than 65. Unsurprisingly, diabetes duration is also associated with more autoimmunity; 40%-50% of people develop another autoimmune condition after 50 years of diabetes, and those diagnosed before age 18 are at particularly elevated risk. Additional risk factors include female sex and white ethnicity. Notably, A1c is entirely uncorrelated with autoimmunity. Dr. Michels concluded by remarking that when it comes to autoimmunity “no disease is an island – if you look for more, you’ll find more.” We hope that Dr. Michels’ call for increased screening for other autoimmune diseases in people with type 1 diabetes takes root in real-world clinical settings, and that earlier detection will ease the burden of these conditions.

13. BDC’s Dr. Paul Wadwa Positions Telemedicine as the Next Major Shift in Diabetes

BDC’s Dr. Paul Wadwa positioned telemedicine as the next major shift in diabetes management, but acknowledged possible barriers to implementation including reimbursement, provider skepticism, licensure across state borders, consent, and security. Dr. Wadwa provided a useful breakdown of terminology: telehealth comprises the delivery of health-related services (e.g. electronic health records, phone consultations, etc.), while telemedicine uses information technology to provide clinical health care at a distance. While remote monitoring of CGM and closed loop devices are common in research, Dr. Wadwa believes we’ll eventually see these services in the clinic (beyond a single generous doctor tracking patients on Dexcom Follow). We find substantial potential for telemedicine in hard-to-reach rural populations, serving to reduce administrative and patient burden, improve cost-efficiency, and increase provider productivity. Dr. Wadwa discussed somewhat promising results from a Barbara Davis Center study last presented at the 2014 Keystone conference implementing telemedicine for type 1 pediatric patients across five active sites. Data from two sites showed an increase in the number of visits/year (2.0/year to 2.9/year) and in the proportion of patients visiting four or more times/year as per ADA guidelines (21% to 33%). However, A1c levels, which were already a high initial mean value of 9.0%, did not significantly change after one year. More recent two-year follow-up data (n=34) presented at IPSAD also failed to demonstrate A1c improvement. Dr. Wadwa cited variable glycemic control to explain these results, as A1c ranged from 7.1-13.7% with 50% of participants experiencing decreased or stable A1c. Those using an insulin pump were more likely to see A1c reductions, as were those with higher initial A1c values (not surprising). Now, Dr. Wadwa and the BDC team will turn their attention to improving outcomes, expanding the BDC telemedicine program, exploring other uses (they’re already piloting use of telemedicine to train patients on the 670G system), ensuring sustainability, and facilitating home use. Telemedicine is one of the big question marks in digital health, though it has shown promise in niche applications, such as early detection and treatment of retinopathy (see NIH’s Dr. Judith Fradkin’s talk from ADA). In its most ambitious form, Virta Health hopes to reverse type 2 diabetes in 100 million people by 2025 through its “online specialty medical clinic.”

Detailed Discussion and Commentary

Plenary Three: Diabetes Care Standards

Panel Discussion

John Buse, MD (UNC, Chapel Hill, NC); William Cefalu, MD (ADA, Arlington, VA); Simon Heller, MD (University of Sheffield, UK); George Grunberger, MD (Grunberger Diabetes Institute, Bloomfield Hills, MI); Guillermo Umpierrez, MD (Emory University, Atlanta, GA); Jay Skyler, MD (University of Miami, FL)

Dr. Richard Kahn (UNC, Chapel Hill, NC): Clinical trials need to be replicated before we believe them. So, going back to the UKPDS and the ~2% A1c drop with lifestyle intervention prior to randomization, has that ever been replicated? What did that lifestyle intervention actually entail?

Dr. Skyler: I can only tell you what I’ve heard. Drs. Robert Turner and David Matthews have talked about this, and have mentioned that participants were seeing a nutritionist and nurse on a regular basis. They also mentioned that participants were working hard. I asked how they got participants to stick to a diet. Their response: well, these were newly-diagnosed individuals, so they knew they had to pay attention.

Dr. Buse: I remember finding the drop pretty extraordinary and looking it up. I don’t think this has ever been replicated.

Dr. Skyler: But the ~5 kg weight loss (~11 lbs) has been replicated.

Dr. Kahn: Is there evidence that delaying intensive therapy for one year or two years makes a difference?

Dr. Grunberger: When you intensify treatment immediately after diagnosis, you can actually see diabetes remission in a subset of patients, but it only works initially. If you wait a longer time to intervene aggressively, it doesn’t work.

Dr. Kahn: But does that result in a long-term delay of complications?

Dr. Grunberger: I’m not aware of that data, because this study was focused specifically on the role of intensive insulin therapy at the time of diagnosis for remission.

Q: Cost has got to be a guideline for what we do. Compared to DPP-4 inhibitors, SGLT-2 inhibitors, insulin products, medical nutrition therapy is cheap – it’s walking and eating one less slice of bread per day. Can you comment on medical nutrition therapy guided by CGM for people with type 2 diabetes?

Dr. Skyler: That experiment is underway.

Dr. Satish Garg: The reason I set up this session was exactly what Jay pointed out – we have all of these enormous guidelines, but most of us who see patients don’t really get anything from those guidelines. What do you all think are the rational steps to improve outcomes for our patients with diabetes?

Dr. Heller: I don’t have an answer to that, but I will also agree with Jay, in that here we are 10 years later with new guidelines and the same average A1c. Somebody has said, “madness is doing the same thing and expecting to get a different result,” and we continue to do the same thing whilst expecting to see a different result. We have to change our paradigm. Our paradigm has failed. We have to experiment with different models, and I know this isn’t as sexy as developing new drugs (which we have plenty of). In our city, diabetes specialists are working in the community to support PCPs and teams. I don’t know if it’s going to work, honestly, but it has to be better than the current state of affairs. I’m sure there are other models, too. There are lots of approaches that might be working in the US – these need funding.

Dr. Cefalu: Guidelines are getting more complicated with new drugs. But let me underscore – no algorithm is going to tell you how to treat your patient. There’s still something called the art of medicine. What the guidelines tell you is to consider patient factors – consider weight gain, consider hypoglycemia – and your regimen will change. CV benefits and weight effects have to be configured into glycemic management. Above all, we still need education of the provider, because with more and more new drugs, it’s only going to get more difficult.

Dr. Grunberger: I agree with my esteemed colleagues. There’s no question that lifestyle modification is key for the management of diabetes. We live in an obesogenic society, and while physicians can’t fix society-wide issues, we can make sure politicians understand the value of sidewalks in neighborhoods. Telling a patient to lose weight unfortunately doesn’t carry as much weight as writing a prescription – it almost feels as if you haven’t done your job as a physician until you write a prescription. So, a long time ago, I used to actually write out prescriptions for exercise. I think we have to be creative. And, we should recognize that many patients come to our attention way too late, by the time their A1c is already at 8% or 9%. I agree 100% with Jay that metformin is probably overrated. The AACE guidelines clearly specify strength of recommendation. We rank medications to help providers according to the order we feel they should be prescribed. We decided to put GLP-1 agonists and SGLT-2 inhibitors as the top two choices (either as initial combination therapy with metformin or as the first choice after a maximum of three months on metformin) back in 2013, even before CVOTs came out.

Dr. Buse: I recommend we stop the conversation on differences between guidelines, because they’re actually 99% the same.

Dr. Larry Hirsch (BD, Franklin Lakes, NJ): Yesterday, I walked out of a session here convinced that lifestyle has not really shown to be effective (following Dr. Kahn’s talk). Jay, you were up on stage this morning championing lifestyle intervention. I respect both of you. So where does that leave the rest of us? I’m also having a hard time understanding why the ADA standards only give a B rating to CVOT data on liraglutide and empagliflozin. These were large, prospective, well-done RCTs – why are they not given an A rating?

Dr. Skyler: Well, they haven’t been replicated. There’s only one study on each agent.

Dr. Buse: I do think these deserve an A rating, and I’m sure the ADA committee will take this up. I’d like to defer the debate on lifestyle, because the truth of the matter is that diet/exercise work differently for different people. If lifestyle modification works in three months, great! If it doesn’t, move on. The issue comes in waiting 10 years, repeating “bubba, you’re too fat,” and classifying that as diabetes treatment.

Comment: I’d like to see a head-to-head study comparing the AACE vs. ADA guidelines in the real world. I’d like to see a cost-effectiveness guideline that helps me choose my medication based on cost criteria.

Dr. Grunberger: We’d all like to see this. One issue is in defining cost. Guidelines are sometimes criticized for favoring branded medications, not going for the cheapest stuff, but in response to that I ask: what is the cost of a patient’s life over a long lifespan of living with a chronic, incurable condition? Furthermore, cost is an artificial milepost since it depends on a specific insurance plan that month and also pales in terms of total cost of diabetes care over the patient’s lifespan.

Q: Is there a reason we so strongly recommend continuing metformin once we start patients on MDI?

Dr. Buse: There was a trial that randomized people to continue metformin or not, and while there was no difference in A1c, you did see less weight gain in the persistent metformin group.

Dr. Heller: There is some evidence from a very few hundred people (who were all overweight/obese) showing slightly less CV disease with metformin therapy.

Q: I don’t see TZDs prescribed as much anymore. Have these been replaced with SGLT-2 inhibitors?

Dr. Skyler: Low-dose pioglitazone (≤30 mg) does not have the side-effect profile it was once proposed to have. So, we continue to use that. I’ve sensed a renewal of interest in pioglitazone since the IRIS study in people without diabetes but borderline A1cs showed that the agent prevents stroke. I think we’ll see some degree of resurgence, particularly since it’s generic. People love to see generics. DPP-4 inhibitor sitagliptin is going to be generic soon.

Q: It seems like guidelines are designed toward physicians only. You don’t see a whole lot in algorithms that helps dieticians or psychiatrists, and that’s a bit frustrating. There are ways these major organizations could touch on that, and give some instruction to other allied health professionals. [Applause.]

Dr. Cefalu: I absolutely agree with you. One thing we’re doing in our PCP initiative is including nurse practitioners, educators, physician’s assistants. This information has to get to people at the primary care level. But as far as what to use for different patients, it’s going to be the same regardless of whether you’re a nurse practitioner, a PA, or a provider.

Dr. Buse: The groups that write these guidelines also include diabetes nurse educators, pharmacists, etc. A lot of language in the ADA Standards of Care incorporates behavioral approaches, medical nutrition therapy, exercise, etc.

Dr. Grunberger: AACE actually has a lot of these documents available – look it up! There’s a whole section on lifestyle modifications that discusses options and strategies. There’s also the ABCD initiative – adiposity-based chronic disease – taking the stigma out of the name “obesity.”

Q: My question is about ELEMENT 5 and biosimilars. That study was done in patients with type 2 diabetes, 18 years or older. What about patients with type 1 diabetes who are very insulin-sensitive? Are there studies that support that piece?

Dr. Buse: I think not, actually.

Comment: I think there’s something missing in guidelines. There are social issues here. I had one patient who was unemployed, uninsured, and unable to get medicine. Another woman was a hoarder, and wouldn’t go into her kitchen to eat so she wouldn’t take her insulin. As a care coordinator, what am I supposed to do with that? Another patient told me that if she died, people wouldn’t blame her – they’d blame her diabetes. My goal there was to connect her to her counselor. But a lot of patients don’t call back.

Dr. Buse: I really appreciate that comment. I think it’s spot on, and there are many hours of discussion we could have surrounding it. But, Satish is anxious for us to go outside right now and have a break.

Plenary Four: Role of Nonprofits in Diabetes Care and Research

American Diabetes Association (ADA)

William Cefalu, MD (ADA, Arlington, VA)

ADA Chief Scientific, Medical, and Mission Officer Dr. William Cefalu opened his presentation with a battery of impressive ADA statistics, referencing 4,600+ research projects and $770 million invested in diabetes research since the program’s inception in 1952. In 2016, ADA provided $34.5 million to more than 378 projects. Dr. Cefalu maintained that the operative word to keep in mind is “collaboration,” noting that diabetes is way too large a disease for any one organization to tackle and that ADA must be recognized as just part of the diabetes ecosystem. Dr. Cefalu envisions all organizations working together under one umbrella and mission, emphasizing the need to move away from silos. We agree, though the devil is always in the details when it comes to great partnership and collaboration. The ADA has had a strategic re-organization over the past six months to achieve these goals. Of note are the changes implemented to the ADA standards of careDr. Cefalu remarked that updating these guidelines once a year is far too infrequent and aims to convert the standards of care to a “living, breathing document” with new information clearly marked and incorporated in real time – how very exciting! We especially hope to see CGM more firmly recommended for all insulin users and those taking medicine that is associated with hypoglcyemia. Eventually, Dr. Cefalu hopes the standards will be available on an app, enabling providers (especially PCPs) to easily stay up to date. The Professional Membership initiative to further support PCPs is also noteworthy – as Dr. Cefalu commented: “I have three brothers in primary care and I can’t imagine them trying to keep up with all the available information.” We couldn’t agree more and applaud the ADA’s efforts to get actionable guidelines into the hands of primary care providers, where most of diabetes is handled. Dr. Cefalu ended his discussion on a bittersweet note, remarking, “the strategic realignment can only help us in this time of diminishing resources; we do have some very exciting initiatives moving forward.”

  • Dr. Cefalu seemed most excited about ADA Pathways to Stop Diabetes, a relatively new program initiated in 2013 geared towards accelerating diabetes research and supporting young investigators. We have been encouraged by the goals of this initiative ­– see our ADA coverage on Dr. Zhen Gu’s work on insulin patch delivery for just one example. Dr. Cefalu noted that he hopes to expand this initiative over the next couple of years, perceiving massive potential in programs of this kind, which provide early-career investigators typically overwhelmed with pressure to bring in grants a rare dose of freedom. We would love to a see a similar program to encourage medical students to go into endocrinology! (See our FFL 2017 coverage for Drs. Irl Hirsch and Bruce Buckingham on the concerning state of the field.)

NIH Perspective: Health Disparities

Pamela Thornton, PhD (NIDDK, Washington, DC)

NIDDK Division of Diabetes, Endocrinology, and Metabolic Diseases Program Director Dr. Pamela Thornton detailed current and future NIH efforts to combat health disparities in the diabetes landscape. In particular, she discussed the NIH’s emphasis on pragmatic and natural experiments aimed at fast-tracking translation and implementing effective intervention in diverse healthcare settings. Projects will be evaluated with an eye towards cost-efficacy, scalability, and sustainability – great metrics for sure! Dr. Thornton was especially proud of the time-sensitive obesity policy and program evaluation, a trans-NIH initiative comprised of six participating institutions NIH Institutes/Offices (NIDDK, NCI, NICHD, NIA, OBSSR, and ODP) dedicated to accelerating the time between grant submission and funding to only three months. She also highlighted several NIH studies and findings pertinent to diabetes, including SEARCH for Diabetes in Youth, FL3X, and the DPP. She characterized the SEARCH data as “extremely concerning and disturbing,” due to the demonstrated massive disparities by race and ethnicity (see our coverage from this year’s ADA examining pediatric patients with type 2 diabetes in a post-hoc analysis and the original report here). The ongoing FL3X study – run by PIs Drs. David Maahs, Elizabeth Mayer-Davis, and Michael Seid – aims to address poor glycemic control in the pediatric population with a coaching intervention, and, according to Dr. Thornton, the results will be available soon. In other news, Dr. Thornton mentioned that eight Centers for Diabetes Translation Research  grants were awarded in 2016, with the next competition expected in 2021 (see information about CDTRs at She concluded with a brief discussion of future directions and next steps for NIDDK, expressing an interest in work investigating social determinants, public health frameworks for obesity, and the outer rings of influence such as social context. These issues, as noted by Dr. Thornton, are multifactorial and extremely complicated, and we applaud NIDDK’s focus on understanding the reasons behind the slow uptake of interventions in real-world circumstances, which are often a far cry from the ideal environments simulated in research.

  • Dr. Thornton mentioned that an upcoming NIDDK workshop will focus on strategies to address obesity and type 2 diabetes disparities. Dr. Thornton indicated meeting logistics will be available on the NIDDK website soon and anyone interested in attending should contact her at Seating is limited. Dr. Thornton called for recommendations on research priorities focused on enhancing lifestyle and integrating social context into diabetes management. The meeting is intended to help clarify important contextual level factors and the role social determinants play in informing practice and adherence.

Meet-the-Peers Sessions: Adult

Transition of Care from Inpatient to Outpatient

Guillermo Umpierrez, MD (Emory University, Atlanta, GA)

Dr. Guillermo Umpierrez called attention to an important but seldom-discussed aspect of diabetes care: how to manage diabetes following hospital discharge. He reviewed new data just presented at ADA 2017 demonstrating the efficacy of DPP-4 inhibitor linagliptin (Lilly/BI’s Tradjenta) as an alternative to the typical prescription of basal insulin. Linagliptin resulted in equivalent glycemic control and lower rates of hypoglycemia – a key consideration in patients who are already at elevated risk of hypoglycemia from recent hospitalization. The study enrolled people with type 2 diabetes (n=140) on oral antidiabetic agents and low-dose basal insulin with A1c>7.5%. Participants were randomized to linagliptin or insulin glargine (Sanofi’s Lantus) for 24 weeks of long-term post-operative care. Mean daily blood glucose did not differ significantly between the linagliptin and insulin glargine groups (146 mg/dl vs. 157 mg/dl, p=0.06), but the frequency of hypoglycemia (defined as blood glucose <70 mg/dl) was significantly lower among linagliptin-treated patients (3% vs. 37%, p<0.001). This bolsters evidence for the utility of DPP-4 inhibitor therapy during the transition from inpatient to outpatient care. Similar results were presented at ADA 2016 for sitagliptin (Merck’s Januvia) in the Sitagliptin Discharge Trial. Dr. Umpierrez emphasized that the use of agents like linagliptin associated with lower hypoglycemia with no compromise in glycemic control should be preferred over basal insulin for post-operative management of diabetes, going so far as to recommend that insulin therapy should be entirely avoided in post-operative diabetes care unless the individual’s A1c is in the 8-9% range.


-- by Ann Carracher, Abigail Dove, Brian Levine, Payal Marathe, Maeve Serino, Adam Brown, and Kelly Close