American Association of Clinical Endocrinology 23rd Annual Scientific and Clinical Congress (AACE 2014)

May 14-18, 2014; Las Vegas, NV; Day #1; Highlights – Draft

Executive Highlights

Hello from Las Vegas and Day #1 of AACE 2014. Yesterday's pre-conference day was lightly attended, but packed plenty of insights from several key opinion leaders. Below, we bring you our top ten highlights, followed by detailed discussion and commentary.

1. A morning panel discussion with Drs. Timothy Bailey (AMCR Institute, Escondido, CA), Bruce Bode (Atlanta Diabetes Associates, GA), and Irl Hirsch (University of Washington, Seattle, WA) shared highly valuable and very opinionated insights on insulin delivery (MiniMed 640G in the US in 18-24 months), CGM, liraglutide and SGLT-2s for type 1 (Lexicon’s data is “a miracle”), the challenging reimbursement environment (“disgusting”), off-label prescriptions, and more.

2. Dr. Bode provided a comprehensive overview of current and future diabetes technology offerings, including his views on the advantages and disadvantages of different products and some new updates we had not ever heard related to Medtronic and Insulet.

3. Dr. Hirsch gave a case-study-based presentation on pumps and CGM and shared enthusiasm for standard deviation, optimism on the future of CGM (he believes CGM use will outstrip pump use in five years), and best practices for interpreting downloads.

4. Dr. Alan Garber (immediate-past AACE President; Baylor College of Medicine, Houston, TX) voiced several strongly-held opinions during his review of AACE’s comprehensive diabetes algorithm, including that “obesity is the fertile soil out of which beta cell failure and type 2 diabetes grows,” so the best treatment for prediabetes, after lifestyle, is actually anti-obesity agents rather than anti-hyperglycemic agents. He also pushed for adding an incretin to basal insulin treatment before adding a prandial insulin, severely spurned metabolic surgery, and made a compelling argument against use of SFUs.

5. We were quite surprised and disappointed, though, by a comment Dr. Garber made at the end of Q&A where he stated, “It’s my job to do the best I can for my patient, and, incidentally, it’s the patients’ job to find a way to finance it.”

6. Dr. Robert Zimmerman (Cleveland Clinic, Cleveland OH) reviewed a slew of new insulins in development, including Novo Nordisk’s, Lilly’s, and Sanofi’s next-generation basal insulins as well as a number of rapid-acting analogs and alternative delivery routes.

7. Dr. Jay Skyler (University of Miami, FL) discussed past and current efforts to stop or slow the autoimmune destruction of beta cells in type 1 diabetes. Although there are a few bright spots in the field of research, Dr. Skyler’s overview was quite sobering overall, as there has been a large number of disappointing results in recent years.

8. Dr. Ed Horton (Harvard Medical School, Boston, MA) discussed whether every CV trial that has been done in the last decade has been doomed based an event rate data that is antiquated – in short, he doesn’t know, but patients are getting better cardiovascular care overall.

9. Dr. Leigh Perreault (University of Colorado, Aurora, CO) advocated for aggressive action in prediabetes, with the ambitious goal of returning patients to normal glucose regulation.

10. Dr. W. Timothy Garvey (University of Alabama, Birmingham, Birmingham, AL) argued that anti-obesity medications should be used long-term (potentially life long), and that it appears safe to do so. Dr. Garvey also compiled a useful table of the different anti-obesity agents’ glycemic effects.

Detailed Discussion and Commentary

Initiating and Interpreting Meter/Pump/Sensor Downloads in Your Practice

Panel Discussion

Dr. George Grunberger (Grunberger Diabetes Institute, Bloomfield Hills, MI): If you had to guess a timeline, when do you realistically think we’ll see predictive low glucose suspend in the US? The expectation is in Europe by end of this year?

Dr. Bailey: I was delighted to see threshold suspend approved so soon after the data was presented. It looks like the FDA is getting a better handle on this. Bureaucrats, when you give them a whole lot they don’t understand, tend to do nothing. The difference between predictive and threshold is not big, but it’s important for the comfort of regulatory bodies. It’s a small, incremental benefit. I think within the next two or three years we’ll see predictive coming out.

Dr. Bruce Bode (Atlanta Diabetes Associates, Atlanta, GA): I think it will happen a lot quicker. FDA has opened up their doors for new technology after the MiniMed 530G approval. There are multiple closed loop studies from all over world going on now with many done in the outpatient arena. FDA is eager to get something approved if it will improve the lives with type 1 diabetes as long as it is safe and efficacious. In Europe, PLGM will be launched this year. Medtronic and Dexcom both have new sensors. It should be easy to control blood glucose overnight. Controlling meals is not going to be easy, so we’ll see some sort of hybrid closed-loop, either partial meal boluses or using other drugs to control postprandial excursions. I think it will be quicker than three to four years. Hopefully, we’ll see something on the market in the next 18-24 months.

Dr. Irl Hirsch (University of Washington, Seattle, WA): I think Dr. Bode is right. There are lot of ways to be creative with this. Novo Nordisk’s liraglutide for type 1 diabetes could potentially have a huge impact on this. What’s going to happen from a regulatory point of view with fingerstick meter accuracy is going to impact what happens with the sensor. I don’t think you can look at anyone thing. Predictive suspend that has come onto the radar, and it’s coming sooner rather than later. I don’t understand the politics, and I don’t want to understand the politics. The big issue in the coming years is the amount of increased hypoglycemia we’re going to be seeing – type 1 patients with 30-50 years of diabetes are doing well and living longer. In the T1D Exchange study we just completed, we took patients over 60 years of age old with over 20 years of diabetes and put them on blinded CGM. The amount of hypoglycemia we see daily is astounding. These are going to be our patients over the next decade. We need to reduce that hypoglycemia exposure.

Dr. Grunberger: This fall, AACE will conduct a consensus conference on the blood glucose monitoring morass. Stay tuned. Otherwise it will be totally crazy if we don’t get together. The aging of our patients is going to result in its own issues.

Q: Does the panel have experience using liraglutide and SGLT-2 inhibitors in type 1 diabetes? And what about using cheaper non-analog insulins?

Dr. Hirsch: We talk about taking CGM away – you could move to my part of country and patients have never had it. It’s brutal. Every single patient we have to fight for, and we’re getting beaten down. We hear, “You can use our product and just write a prior authorization for it.” It’s out of control. We just can’t do it anymore. This all gets back to the Affordable Care Act and ACOs. We’re being asked to see the same patients at lower costs. The liraglutide and GLP-1 data in type 1 is extremely encouraging. Novo Nordisk is starting the second registration trial for liraglutide in type 1. We will see FDA labeling for liraglutide in type 1 in the next couple of years. But our use of liraglutide, even in type 2 diabetes, is even limited in our part of the country. It’s because of the costs. For SGLT-2 inhibitors, it doesn’t matter what the pathogenesis is – they work. And they can work quite well. The question is with the side effect baggage, are they going to be worth the cost. The weight issues are important. To me, the biggest issue is hypoglycemia.

I have patients screaming at me because of what has happened to the cost of insulin over the last couple of years. For the typical patients I have seen, they are allowed one vial of insulin at a time. If it lasts 19 days, and they come in at day 18, they cannot get their next vial of insulin. It has made life so miserable. It all comes back to the cost of insulin, which has gone up 50% in the last 18 months. Some patients start using regular insulin in their pump, which is a complete disaster. One patient got creative and used half lispro, half regular insulin in the pump. Because of the cost! We have other patients using infusion sites too long and not changing them. The cost has become prohibitive. I appreciate that I live in a particular part of country. But never did I think I would have to do a prior authorization to get someone insulin. This is not a boutique product. It’s required for survival. And quite frankly, I’m disgusted.

Dr. Grunberger: Several companies are starting trials of SGLT-2s in type 1 diabetes.

Dr. Bode: Clearly SGLT-2 inhibitors are going for phase 3 trials in type 1 diabetes. All three companies out there are doing it, and Lexicon showed their phase 2 data recently. It’s a miracle. It beats anything out there. It’s amazing how well these drugs work in type 1. Novo Nordisk is going full tilt on liraglutide with the second registration trial. The benefit is less hypoglycemia and less hyperglycemia.

Dr. Bailey: One of the Helmsley trustees was quoted as saying every drug in type 2 should be tested in type 1. [Editor’s Note: See our interview with David Panzirer and Dana Ball.] I’ve started to prescribe GLP-1 off label in a lot of my patients with type 1. The benefit is the weight, but most of my patients have eventually run out of money. It’s expensive therapy. Even if it does get approved what is it going to cost? This all started with Amylin discussing the multihormonal basis of type 1. But it’s just a huge burden in terms of cost. Sorry to be so negative. But the trials are happening.

Q: What role, if any, do the PCP and internist have in dealing with pumps and sensors?

Dr. Grunberger: In our paper, which is in the May 2014 issue of Endocrine Practice, we have a Consensus Statement. Other countries have more regulated systems. In the US, anyone can prescribe a pump. We put together recommendations on what a good team should look like. But there is no certification system. Patient education is paramount. Accessibility is a must.

Dr. Bode: There are not enough doctors to go around. Easily 50% of endocrinologists won’t see diabetes and certainly won’t do a pump. Many PCPs like diabetes and like to do it. Obviously, we should find a way to support them. Industry is clearly out there to help support people on pumps and Medtronic is leading it. This requires a mixed approach, since not everyone has teams. Obviously if you need a transplant, not everyone can do it. That’s true for advanced heart failure care as well. Maybe that’s what we need to do in the long run. But right now we have both primary care physicians and diabetes specialists. The team might only be one person plus a doctor. It might be an NP in multi-specialty clinic.

Dr. Bailey: We want to have the best outcomes for patients. If you do one pump per year, you will never get good at it. You must involve lower cost people. Not every center will have that. But the mid-level people really take to this kind of thing. You must have someone who embraces this type of thing and earns less money than you do. Anything that loses you money, the more you do the more you’re going to have a problem.

Dr. Hirsch: I am the strongest proponent of mid-levels assisting in diabetes care. However, there is an inherent danger with this. We were the second largest DCCT clinic. Our model is partly based on the DCCT and partly on Dr. Jay Skyler’s clinic at the University of Miami. If you have mid levels doing state of the art diabetes technology, the providers/attendings/endocrinologists lose touch. That is what I see happening. Where I really see it is fellows that we have trained over the last 20 years. They are going out into practice and have mid levels doing everything. There is no reason for them to keep up with what’s new. How to balance it? I’m still not sure.

Dr. Grunberger: I still have fellows that tell me there is no exposure to pumps in training. They have never seen an insulin pump. What do we do about the future work force?

Dr. Bode: Academic centers also won’t allow reps to come in. How do they know about any of this new stuff? It’s amazing how much knowledge these people lack. They don’t even know a glucose sensor exists. The other big issue here is reimbursement. There are hurdles coming up in the near future. Major insurance companies aren’t going to allow CGM – they’re going to take the Medicare stance. When you’re having the patient pay everything, it becomes an exorbitant cost.

Q: Can the panel comment on the advisability or possibility of having a CPT code for downloading pumps remotely or in person? We’re having mid-levels doing this because reimbursement is not very good for it.

Dr. Bode: There are CPT codes for downloading out there. The problem is they don’t pay for it. So eventually we had to say forget it – we’ll just do it for free. You can bill for prolonged visit codes if a patient downloads at home and you interpret remotely. Some insurance companies will cover that. But getting reimbursement for the code specifically is not out there. JDRF and AACE could help us a lot – they certainly helped with CGM. ADA could help us, but I don’t think they will.

Dr. Grunberger: It’s on the agenda for AACE.

Q: Before you poo-poo written glucose records – when patients only download data and see it in the office, don’t they lose the opportunity to change their behavior when writing it down in a journal?

Dr. Bode: There is no question about what happens when people don’t look at anything. I think we’re getting to low energy Bluetooth – meters, CGM, pumps will all go to smartphones. Then we can develop apps that can tell them quickly what to do. It’s like when the iBGStar got launched, the coverage wasn’t there. But people loved it. It started growing dramatically and Sanofi stopped doing anything with it. People look at their cell phones all the time. They literally bump into people and get in car crashes all the time. Just think if you can put all their glucose, pump, and CGM data there.

Dr. Bailey: Until fully automated pancreas replacement technology comes along, some thought is going to be required. The more you look at the CGM, the better you do. The more you test your blood glucose, the better you do. There are meters that have more advanced features like pattern detection. Logbook technology is antiquated. It’s embarrassing.

Dr. Hirsch: It’s very patient dependent. Some people are doing well, with good averages and not many hypos. But that’s not the typical person. When someone is really not doing well and they track things for a few weeks, and actually write it down, it allows them to see things. Where I first got better attuned to this was when very first CGMS came out. We asked patients to track everything for three days. In those three days, patients had much better insight because they were tracking things. I had somebody on MDI and using the Dexcom. I have them track things – “I’m only going to ask you to write insulin doses and food and exercise.” It was extremely helpful for a more efficient clinic visit. If there was a really easy app on the phone to do this, I would love it. But I don’t know of one.

Q: When you’re talking about prescribing medications like GLP-1 off label, how do you approach that discussion with patients?

Dr. Bode: On this whole off label use – I mean most of what you do in your practice is off label. Giving insulin to a one-year old is off label. You must do what’s best for patients. However, you must play the coverage game. They were type 1, but now they’re type 2. It’s wordsmithing – it’s a hard thing to do.

Dr. Grunberger: Contraindications need to be taken seriously. But the FDA is still not in the business of regulating practice of medicine, even though they would like to. It’s in between you and patients. It should not be a different conversation. Off label will come into play based on cost. But I think the off label issue is overblown. If you have insulin resistant patients that could benefit from weight loss, I’m not sure it’s a problem as long as patient knows what you’re doing.

Q: It’s just that sometimes I find that if you say this has not been approved, that gets the patient worried.

Dr. Hirsch: There are two issues here. One is discussion with patients. The other is with payers. Metformin has a black box in type 1 diabetes. We’re not only using it routinely now, but the T1D Exchange is finishing a study in obese teens with type 1 diabetes. This whole issue of off-label is sort of ridiculous. It was less than 10 years ago that you were using IV insulin off label for the treatment of DKA. I had to declare in CME events that that was off label use. You shouldn’t take it as seriously. It’s not that big of a deal.

Q: The new MiniMed 530/Enlite is stopping the basal rate, but you’re still getting active insulin on board. Will it make that much of a difference?

Dr. Timothy Bailey (AMCR Institute, Escondido, CA): A lot of us who wanted an artificial pancreas 15 years ago thought that this feature wouldn’t do very much. We figured that just interrupting insulin for two hours would not do a lot of good. Your question is well intended. But it turns out that it works pretty well. The key thing at night is patients are not bolusing and not eating, so it’s a little bit simpler.

Q: What is your experience with islet cell transplants and pancreas transplantation?

Dr. Bailey: It would be a wonderful thing. In most of the cases I’ve observed, it’s almost a no brainer – they are getting a kidney and thus taking immunosuppressive drugs anyways. The problem is if you have type 2 or lots of insulin resistance. Also, there are not enough pancreases around, and you’re on immunosuppressive drugs. The best candidates I’ve seen are the people with severe hypoglycemia and renal disease. These patients’ lives can be transformed, as they have totally uncontrolled diabetes. But for the vast majority of patients who are healthy, I think technology will get us there sooner.

Dr. Bode: I agree. If you have type 1 and need a kidney, it makes sense to do the pancreas as well. Medicare will agree with that. The idea about islet cell transplants – they work, but they don’t work long-term. We don’t have better solutions. But JDRF is not giving up on this. You will also see bio-cellular type devices. I think there are solutions for everybody out there.

Dr. Hirsch: My hope is that next year at this time, there will be much more visibility about clinical trials in encapsulated islets. You will see more on encapsulated islets where hopefully no immunosuppression will be required. The other thing, and it’s not exactly islet cell transplant, is SmartInsulin. The company was bought by Merck and we didn’t hear anything for five years. Now they are starting clinical trials. That is also encouraging.

Current Pump and CGM

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

Dr. Bruce Bode provided a phenomenally comprehensive overview of current and future diabetes technology offerings, sharing his balanced views on the advantages and disadvantages of different products and disclosing some new details we had not ever heard (e.g., Medtronic’s hybrid closed-loop device is called the “740G”; the more modern Insulet OmniPod PDM in development will be rechargeable and include “mobile technology – iPhone like”). He also shared that the Accu-Chek Aviva Expert meter (built-in bolus calculator) is now available in the US (it was approved by the FDA in October 2013). Below, we’ve enclosed four tables with Dr. Bode’s views on the pros/cons of current pump and CGM products (Medtronic, Dexcom, Animas, Roche, Tandem, Asante, Insulet, Valeritas), as well as a summary of the upcoming pump and CGM devices he mentioned. All are listed in the order they were addressed in the talk. When applicable, we have added the latest timelines we are aware of.

  • The Accu-Chek Aviva Expert meter is now available in the US. This standalone meter has a built-in bolus calculator, a huge unmet need for MDI patients. The meter has been available in Europe, though there was no US timing update on this meter in Roche’s 1Q14 call – it was good to hear that this is now available stateside. As a reminder, the US version of Abbott’s FreeStyle Insulinx meter only has an insulin logging feature, as Abbott was presumably not able to get the European version’s built-in bolus calculator through the FDA.  

Table 1: Pros and Cons of Current Pump Offerings


Key Advantages

Possible Disadvantages

Medtronic MiniMed 530G

Quick/simple bolus programming; 300-unit reservoir; basal rate as low as 0.025 u/hr; wireless integration with Contour Next Link; easy download to CareLink; integrated CGM with low glucose suspend

Screen not color; not waterproof

Animas OneTouch Ping

Easy to see color display; basal rate as low as 0.025 u/hr; max bolus of 35 units; waterproof; meter remote; bolus calculator on remote and pump; web-based download software

Reservoir is only 200 units; hard-to-see screen on meter remote; more button pushes to complete bolus; no back button

Accu-Chek Combo System

Color screen on meter remote; reservoir holds 315 units (biggest on market); meter remote is fully functioning pump, allows for discretion

Bolus calculator only on meter remote; hard-to-see pump screen; not waterproof; software not compatible with Mac or Windows 7

Tandem t:slim

Color touchscreen display; lowest dose increment on the market (0.001 u/hr); site change reminder; 300-unit reservoir; waterproof; smallest pump on market; rechargeable.

No BGM integration; need to charge 10-15 min/day; food bolus only reduced if BG<70; nine minutes to change cartridge; BG meter must be downloaded separately to get on combined report.

Asante Snap

Large font on screen; simple menu structure, fast navigation; prefilled 300-unit cartridges; auto-prime feature; seven days between cartridge changes; no battery changes or charging; very LOUD alarms; lower upfront costs; waterproof

No glucose meter integration; not covered by Medicare; only Humalog right now; bolus calculator does not show specific amount of IOB; no reverse bolus correction if BG<70; download software with Diasend, with Tidepool planned for future.

Insulet OmniPod

No tubing; truly continuous insulin delivery; small size; auto-insertion; built in BGM; lower upfront costs

Reservoir is only 200 units; cannot bolus without PDM; not covered by Medicare; minimum 85-unit fill

Valeritas V-Go

Prefixed basal with simple on-demand bolus insulin; no need to look at device; easy to fill, apply, and remove, every 24 hours. Fully disposable.

Max 76 units per day. However, “80% of all US insulin users use <75 units a day.” No good randomized controlled trial data.

Table 2: Pros and Cons of Current CGM Offerings


Key Advantages

Possible Disadvantages

Medtronic MiniMed 530G/Enlite

Low glucose suspend; six-day wear; six-month warranty; predictive alerts; trends graphs; ~40 minutes of glucose information stored in transmitter; safe for use with acetaminophen; easy/automatic inserter; CareLink download; waterproof (hot water not suggested)

Sensor adhesion reportedly feels “fragile,” easily loosens; small screen can be hard to see; alarm can be hard to hear; same transmitter as Sof-Sensor – still loses contact with pump

Dexcom G4 Platinum

Seven day wear; one-year receiver warranty, six-month transmitter warranty; reported no discomfort when inserted; compact sensor/transmitter with excellent skin adhesion; 20-ft transmission distance; trend graphs; loud alarm; large, well-lit screen; download software offers progress report.

No pump integration; no data if receiver is forgotten; acetaminophen can influence accuracy.

Table 3: Upcoming Pumps



Latest Timeline


640G (predictive low glucose suspend)

740G (hybrid closed-loop)

Patch Pumps

EU launch by October 31, 2014; potential US availability by 2015.




Vibe with Dexcom integration

Artificial Pancreas Project with JDRF

LifeScan Reveal web-based software

Calibra Bolus-only pump

PMA filed with the FDA in 1Q13. JPM 2014 called for launch in 2014.



Last update called for additional studies to start in 2013


Accu-Chek Insight Pump

Launched in EU


t:slim with Dexcom integration

t:flex (480-unit reservoir)

t:dual (dual chamber w/ JDRF)

FDA filing in June 2014

FDA filing in mid-2014



OmniPod PDM with LifeScan Verio integration

Rechargeable PDM with mobile technology – iPhone like

OmniPod for use with U500 Regular insulin

Integrated CGM into pod

510(k) filed in February 2014; LifeScan evaluating feedback related to the strips.

On track to be developed in time for ADA 2015; FDA submission expected in 2015

FDA 510(k) submission expected in late 2014

Sterilization issues appear resolved. Compiling data from insertion study.

Table 4: Upcoming CGMs



Latest Timeline


Enlite Enhanced


Enlite 3

Harmony Sensors

Duo (combined sensor and insulin infusion)

Available in Europe; will be part of 640G



Oral presentation at AACE 2014 this Friday.



Gen 5 mobile platform

In final stages of FDA approval.

~2015 launch, though will likely be rolled out in stages

Questions and Answers

Q: Is there any way to get Medicare to approve coverage of CGM sensors?

A: No formal study is underway at this time; T1D Exchange and Helmsley Charitable Trust have been looking into a dedicated trial to approve CGM in the Medicare population. Currently, you can receive approval for CGM in this population if they have secondary insurance. Dexcom has been receiving approval via the pharmacy benefits portion of Medicare. Medtronic and others are also working on a dedicated CMS code for sensor augmented pump therapy that would also help with coverage for this population of Medicare patients. Based on our data from the T1D Exchange, patients older than 50 years old are at high risk for severe hypoglycemia and certainly would benefit from sensor augmented pump therapy and threshold suspend.

Case STudies: Interpreting Pump and CGM Downloads

Irl B. Hirsch, MD (University of Washington, Seattle, WA)

Quoting Steven Wright, “I was a peripheral visionary. I could see the future, but only way off to the side,” Dr. Hirsch defined the future of type 1 diabetes as “blurry.” However, his informative and engaging presentation suggested that pump and CGM will undoubtedly be cornerstones of future therapy. Dr. Hirsch emphasized the value of these device downloads, as the data “allows clinicians to see how patients act, react, and even think about their diabetes management.” He was particularly positive on the use of standard deviation, which provides a clear indicator of glucose variability and quick assessment of a patient’s “diabetes fingerprint.” Dr. Hirsch was very optimistic on the future of CGM – he expects that within five years, CGM penetration will outstrip that of insulin pumps, and patients may even be able to use CGM devices alone to make successful insulin decisions (the latter seems very likely to us, as a study is in the works on this front!). In addition to this CGM confidence, Dr. Hirsch also welcomed better tools for MDI patients and noted the recent US launch of the Accu-Chek Aviva Expert meter with a built-in bolus calculator (approved by the FDA in October 2013) as a great example.

  • Looking ahead, Dr. Hirsch anticipated that “probably within five years we’re going to see CGM out-using insulin pump therapy.” T1D Exchange data (n=~26,000 type 1 diabetes patients) provided an encouraging “reality check”: since enrollment to now (~2-3 years), overall CGM use in the Exchange has gone from 6% to 9%. Dr. Hirsch highlighted the trajectory in <6 year olds, where CGM use has tripled from 3% to 9%. Notably, CGM use is up to almost 20% in adults over age 26 in the Exchange. Additionally, the mean A1c in CGM users over age 26 is 7.2% vs. 7.7% in non-CGM users.
    • Though not as dramatic an increase, Exchange data suggests that overall pump use stands at 59%, up from 55% at enrollment. Dr. Hirsch noted that this data doesn’t reflect actually pump use in the US, which is known to be around 30%. Nonetheless, the study found that in every age category (<6 year olds, teens ages 13-26, and adults >26 years old), pump use increased. To the naysayers that doubt pump’s effectiveness, the Exchange data suggests that mean A1c in pump users is also better in every age category vs. MDI users: 8.2% vs. 8.8% (<13 years), 8.6% vs. 9.2% (13-26 years) and 7.6% vs. 7.8 (>26 years).
  • Notably, a future study will examine patients using a CGM device alone to make insulin-dosing decisions. We were very excited to hear this, since we know many patients are using CGM to dose insulin in the real world – this data would be incredible value for the FDA to see, as we have little doubt that dosing insulin off of current CGM is at least non-inferior to use of less frequent fingersticks. Additionally, CGM’s adjunctive claim is a limiting factor in securing Medicare reimbursement. In Dr. Hirsch’s opinion, “patients will be fine, but the only caveat is the first few hours when accuracy of the CGM isn’t that great.” He did, however, stress that patients are not using enough SMBGs to make insulin-dosing decisions, showing numerous examples of “blind boluses” that led to hyper and hypoglycemia.
  • One of the first components Dr. Hirsch looks at in patients’ downloads is standard deviation (SD), which has “become the key metric for the understanding of a patient’s glycemic fingerprint.” Dr. Hirsch finds that SD is easiest to see on the Dexcom download, but it is noted in most other device downloads. He emphasized the important of thinking about SD in the context of the mean (i.e., coefficient of variation, the variable of choice in his FLAT-SUGAR pilot study, which is “almost done”). Several quick and dirty calculations that Dr. Hirsch uses to get a sense of patients’ management:
    • For SMBG: SD x 2 < mean glucose. This first-line metric allows clinicians to see if patients are giving their insulin at the correct time and regularly matching their food with insulin. Even better, patients with SD x 3 < mean glucose are doing extremely well – Dr. Hirsch noted that patients who can get to this point are often still making some endogenous insulin.
    • For CGM: SD x 3 < mean. Dr. Hirsch was forward-looking, noting that current CGM downloads often have 3,000 data points, and we need better metrics. In his opinion, coefficient of variation (CV), TIR (time in range), TBR (time below range), and TAR (time above range) all need to be correlated with outcomes, as they are important pertaining to markers of inflammation, oxidative stress, and severe hypoglycemia. His FLAT-SUGAR study is looking at all of these parameters, though CV is the metric of choice.
  • Dr. Hirsch and his colleagues conducted a small survey with 17 providers to evaluate different manufacturers’ download software. He did not share specific results, but did say that Animas’ Diasend software “didn’t come in very high on our surveys… not at the bottom, but not near the top either.” He also mentioned that the logistics of Tandem’s t:connect software “hasn’t worked out the way Tandem wanted it to,” as patients are not downloading at home. This survey will be expanded in the future to T1D Exchange clinics around the US; with hundreds of providers in this potential study, we are eager to see how the different downloads stack up and what components are most useful. We universally seem to hear praise for Medtronic’s CareLink, though recent conferences have also suggested enthusiasm for Tidepool’s blip.
  • For the industry members in the audience, Dr. Hirsch presented his wish list for downloads. For pumps this included basic insulin statistics; TDD, % basal, % use of bolus calculators, time-specific bolus calculator over rides and under rides, and a daily summary to better understand ICR, ISF, basal rates, and if appropriate, over-rides and under-rides for trends. For CGMs, Dr. Hirsch requested basic blood glucose statistics, overall patterns, and daily decision making to best understand how patients think through challenges. As a patient and HCP himself, Dr. Hirsch has seen it all and truly understands the patient perspective – we appreciated hearing him say, “No matter what the NY Times says, this is a LOT of work!”
    • Dr. Hirsch finds that patients who over- and under-ride their bolus calculators ~20-25% of the time do the best. Those who never over- or under-ride do worse, as these patients are often not acknowledging the multitude of behavioral and environmental factors that affect glucose trends (e.g., about to exercise, stressed, etc). For those that are over- or under-riding 30-50% of the time, pump settings are most often wrong.
    • “I have patients in whom I don’t measure A1c, simply because they don’t work in some patients. However, to payers, I look like a bad provider. So now I’m measuring A1c so I don’t get put on a website that says I’m a crappy doctor. We have to do it, because we get dinged if we don’t.” Many nodding heads in the audience seemed to agree with this surprising take on the occasional lack of utility in A1c measurements. Dr. Hirsch cited that in a 600 patient study, 14% of A1c measurements are actually misleading. Some of the most common reasons for incorrect A1c values include erythropoietin therapy, anemia, liver disease, cystic fibrosis, and hemochromatosis.

Questions and Answers

Q: For a patient riding a bike to work in the mornings without eating breakfast and still having a hyperglycemic spike, would you suggest telling him to take a few units of insulin beforehand?

A: I would say he should continue watching his sensor and only take units if he sees his glucose rising.

Comment: At a Dexcom meeting last week, we heard that their new sensor has improved regarding the acetaminophen interference.

A: We’ve been told that for a long time, and we’re hoping this is true.

Q: When you encourage patients to over-ride their bolus calculator due to trend data, do you offer general guidelines?

A: As general guide, we told patients that if they have a 45 degree angle or single arrow up (depending on the device), to increase their bolus by 10%. If the arrow is straight up or a double arrow up, increase the bolus by 20%. Similar decreases if the arrow is pointing downwards. However, if you have insulin on board and the bolus calculator tells you not to give any more insulin, then we suggest not giving a bolus until an hour after the insulin was taken. Depending on the patient, this additional insulin can be 1 unit, 0.5 unit, etc.; there is no one size fits all. Also note that we don’t discuss this adjusted bolus information with patients in the beginning of CGM therapy because there is often too much information to digest at this stage.

Q: Have you had any patients who have gastric bypass either roux-en-y or gastric sleeve? I have a type 1 patient who had a gastric sleeve, and she tends to go really high after she eats. Do you have any suggestions for her therapy?

A: No, I’ve never had a type 1 patient with gastric bypass, so this is my first time hearing of such a situation. Maybe she might consider using an extended bolus or even using regular insulin. Believe it or not, because of the cost of insulin, patients are coming in with regular insulin in their pumps, which is a whole other topic. Additionally, consider if she might benefit from a trial of metoclopramide.

Q: You mentioned a patient eating a bacon cheeseburger. How do you use CGM data to work with patients who drink coffee with milk all day or regularly eat cheeseburgers?

A: We were told that as these tools came up (CGM and pumps), you can simply use these tools and eat what you want. However, this is not true. Patients that do the best, eat the best. Even if we get better insulin in the future, it’s still not the same as your pancreas directly releasing it into the portal system. It’s far from perfect. We don’t currently have tools to make it any better. My general advice is to eat small amounts of carbohydrates without huge amounts of fat.

Q: A confounding factor when we tried to translate data from MDI to pump, we found that many patients on MDI haven’t actually had good MDI education. So when we put them on a pump, the education is very intense. Once they learn a lot of the basics of insulin therapy and kinetics, they do better partially because of motivation factor, yet many become overwhelmed with pump therapy. A lot of issues are behavioral, like over- and underestimating. Before pump therapy, the emphasis on education hasn’t been hammered out, e.g. how to carb count, avoid highs and lows, etc. So when patients go back on MDI, they often do better because they now have education to get diabetes under control.

A: It’s hard to disagree. Some patients who come into our clinic go on pump and aren’t doing well. Pump patients may do really well with good training, but there is no infrastructure in the office for education; it’s cruel. The clinician spends 15 minutes with a patient and it’s just not enough time. At the end of the day, it comes down to face-to-face time. When you don’t have time to spend with a patient, you can’t do this therapy successfully.

Dr. George Grunberger: I would add that if we do spend a lot of time training, re-training, educating, and re-educating, who is going to pay for it? It is so important, but the infrastructure is as well.

Special Session: Re-Thinking Diabetes Care

AACE Comprehensive Diabetes Management ALgorithm

Alan Garber, MD, PhD (Immediate Past AACE President, Baylor College of Medicine, Houston, TX)

Dr. Alan Garber reviewed several of the different facets of AACE’s Comprehensive Diabetes Management Algorithm, including the recommendations on treatment of overweight/obesity, prediabetes, overt hyperglycemia, hypertension, and hyperlipidemia. As Dr. Garber emphasized, the comprehensiveness of the algorithm reflects AACE’s view that providers must treat diabetes patients’ risk factors holistically, and not take a glucose-centric view of only treating hyperglycemia. Dr. Garber voiced several strongly held opinions during this presentation: (i) that the best treatment for prediabetes, after lifestyle, is anti-obesity agents rather than anti-hyperglycemic agents (because they address the underlying etiology of the patient’s beta cell decline: said Dr. Garber, “obesity is the fertile soil out of which beta cell failure and type 2 diabetes grows”); (ii) that intensifying a basal insulin regimen should begin with adding an incretin, not prandial insulin; (iii) that surgery should be reserved for “patients who have an urgent need for a cosmetic result or truly serious complications” – he did not acknowledge the weight-independent diabetes benefit of metabolic surgery; (iv) that one should use sulfonylureas “particularly if you dislike the patient and want them to be out of your practice” (he also made a compelling argument during Q&A that even though sulfonylureas’ up front cost is cheap, the hypoglycemia that they cause ends up costing the system a significant amount of money in the end); and (v) that “to dispense with LDL targets in this day and age is irresponsible” (speaking of the new 2013 AHA/ACC lipid guidelines). Finally, when asked during Q&A about the absence of cost discussions in the AACE algorithm (in contrast to the ADA/EASD 2012 Position Statement), Dr. Garber made the completely fair argument that as a physician he does not have an obligation to increase insurance companies’ profits, but we were quite surprised and disappointed that he ultimately commented, “It’s my job to do the best I can for my patient and, incidentally, it’s the patients’ job to find a way to finance it.” We do think this is a sign of the times – more fatigue on the part of HCPs who are not necessarily trained or compensated to help patients on reimbursement.  

  • For more detail on Dr. Garber’s abovementioned views on treating prediabetes with anti-obesity medications: After lifestyle intervention, Dr. Garber believes that the more promising method for normalizing glycemia in people with prediabetes lies in pharmacologically assisted weight loss rather than anti-hyperglycemic agents. He remarked that new obesity drugs such as Belviq or Qsymia have produced “outstanding remissions” in prediabetes and prevention of conversion to type 2 diabetes. He stated that data has been published in the literature with either of these two drugs showing 80-90% reduction in risk of converting to type 2 diabetes (citation not provided) and that the same could be said for liraglutide 3 mg. Dr. Garber believes that since obesity is the underlying cause of insulin resistance and, therefore, beta cell insufficiency, that treating obesity is the way to address the underlying pathophysiology of prediabetes.
  • During Q&A, Dr. Garber made a  pretty compelling argument for why there are hidden costs associated with SFUs: he noted that the NEJM recently published a piece from the CDC that found that the second most frequent cause of ER visits and hospitalizations was antidiabetic medications, and that 95% of those were due to hypoglycemia. 40% of that hypoglycemia was from oral agents, which implied he thought most of the hypoglycemia due to oral agents was from SFUs. So, he argued, “yes you can buy a cheap SFU, but you’re also buying an increased risk of hypoglycemia” and, therefore, costly ER stays.
  • There were a couple of points during the presentation where Dr. Garber’s recommendations appeared slightly more academic than practical. For example, he commented that it typically takes three-to-five anti-hypertensive medications to get a person with diabetes to his or her blood pressure goal and that providers should be ready to pile them on lest they let the patient suffer the consequences of high blood pressure. Considering that the patient is likely taking three-to-five anti-hypertensive agents on top of multiple anti-hyperglycemic and lipid control agents, that comment seemed a bit insensitive to both providers and patients. In addition, when asked about why there is no discussion of costs in the AACE algorithm, he did not acknowledge that many people with type 2 diabetes come from disproportionately poorer backgrounds and may not have the means to afford good insurance that covers the gold standard in treatment.

Questions and Answers

Q: When you compare the AACE algorithm to the ADA algorithm, there are a couple differences. The ADA algorithm seems to put more emphasis on cost of medications, and there seems to be really no discussion of that issue in the AACE algorithm. The other major difference is that based on taking that into account, the ADA algorithm tends to have SFUs much higher up. Can you comment?

A: There are two kinds of costs for medications: out of pocket costs for meds or total cost to the system for using that medication. For example, in a recent publication in the NEJM from the CDC, they looked at reasons for adverse drug reactions leading to ER visits and hospitalizations. The number two cause of all adverse effects (AEs) were antidiabetic medications. 95% of those AEs were hypoglycemia, and 40% of that hypoglycemia was from oral agents. So yes you can buy a cheap SFU, but you’re also buying an increased risk of hypoglycemia. If you look at the gliclazide trial, published in NEJM a few years ago, they did an analysis of hypoglycemic frequency. Half the patients over a five-year period of time had hypoglycemia, and 5% of patients had severe hypoglycemia. That’s from a mild SFU. So this is not an inconsequential event. Then you have to get to perception of to whom do we have obligations as a physician? I don’t see on a personal level that I have any responsibility to a third party that’s trying to make a business out of insuring someone’s healthcare. My obligations are to my patients, first and foremost [audience applause]. If insurers can’t make a profit, they can be damned. They can figure out another business model. Or the politicians can stop making promises that they don’t want to fund. But it’s my job to do the best I can for my patient and incidentally it’s the patients’ way to find a way to finance it.

New Insulins on the Horizon

Robert Zimmerman, MD (Cleveland Clinic, Cleveland, OH)

Dr. Robert Zimmerman reviewed the slew of new insulins in development and discussed how they might be able to solve the current problems that insulin therapy poses: hypoglycemia, weight gain, peripheral action, injection burden, rapid-acting agents that are too slow, long-acting agents that are not flat, declining efficacy with injection volume, and concentrated insulins that are too slow. The data Dr. Zimmerman presented was a bit outdated across the board, but the sentiment behind his talk was in the right place.

  • In terms of basal insulins, Dr. Zimmerman highlighted Novo Nordisk’s Tresiba (insulin degludec), Lilly’s peglispro (PEGylated insulin lispro; LY2605541), and Sanofi’s new U300 insulin glargine formulation. He highlighted data demonstrating that degludec had demonstrated comparable A1c-lowering to Lantus with 25-40% lower rates of nocturnal hypoglycemia (but not overall hypoglycemia). He reminded the audience that the FDA had called for a pre-approval CVOT since pooled MACE events in degludec’s phase 3 trials showed a statistically significant 70% increased risk compared to Lantus. With regard to peglispro, he was very enthusiastic about the agent’s hepato-selectivity, which more closely mimics endogenous insulin secretion and should ablate some of the weight gain associated with current exogenous insulin secretion. He did mention that topline phase 3 data had been announced and that these data did not seem to replicate the weight loss with peglispro seen in phase 2 (full data to be presented at ADA). With regards to U300, he noted that the EDITION I trial found that U300 glargine had comparable A1c-lowering to U100 glargine with significantly decreased nocturnal hypoglycemia after six months.
  • In terms of rapid-acting insulins, he highlighted a range of new analogs and alternative routes of delivery.
    • Recombinant human hyaluronidase (Halozyme’s Hylenex) accelerates the absorption and action of co-injected rapid-acting insulin.
    • VIAject (Biodel’s Linjeta) has demonstrated much faster onset of action than current rapid-acting analogs (he did not mention, though, that Linjeta was discontinued long ago in favor of Biodel’s new candidate BIOD-123). BIOD-123 is currently awaiting a partnership deal to advance to phase 3.
    • Mannkind’s inhalable Afrezza delays the rise in post prandial glucose compared to aspart and has demonstrated comparable A1c-lowering to insulin aspart (Rosenstock et al., Lancet 2010). Dr. Zimmerman did not mention, though, that a major point of discussion at Afrezza’s most recent FDA advisory committee meeting was that pooled data suggest that it may not be quite as good as aspart for glycemic control. Dr. Zimmerman doesn’t believe that Afrezza’s negative effect on pulmonary function is too alarming – he displayed a graph showing that just having diabetes causes a marked decrease in pulmonary function, and the effect of using Afrezza on top of that is very small.
    • Finally, Dr. Zimmerman noted that Biocon’s oral IN-105 shows potential for being an agent of interest. It has demonstrated a dose-response relationship with plasma concentration in preliminary pharmacologic studies.
  • He noted that numerous other insulins are in development that he would not have time to discuss, including Novo Nordisk’s ultra-rapid acting FIAsp (NN1218), Biodel’s ultra-rapid acting U400 (BIOD-531), Thermalin’s rapid-acting concentrated insulin and heat-stable insulin, Novo Nordisk’s oral insulins, and buccal insulins.

Re-Thinking Diabetes Care

Treatment of Pre-Diabetes

Leigh Perreault, MD (University of Colorado Anschutz Medical Campus, Aurora, CO)

Dr. Leigh Perreault very enthusiastically discussed the importance of treating pre-diabetes with the goal of resuming normal glucose regulation, as well as the benefits associated with successfully achieving that goal. She referenced a retrospective re-analysis of the DPP Outcomes Study (DPPOS) that she presented at ADA 2012 (read our coverage) showing that patients that resumed normal glucose regulation at least once during the study were 56% less likely to advance to diabetes than those who did not achieve normal glucose regulation, regardless of the patients’ initial randomization to lifestyle or metformin. As opposed to the outcomes data in type 2 diabetes trials, Dr. Perreault suggested that there exists solid data connecting glucose-lowering drugs to improved cardiovascular effects in prediabetes patients, although much of the data she cited was from agents that are less popular today (TZDs and acarbose) – there is certainly more data needed in this area, especially in newer agent classes. We found Dr. Perreault’s enthusiasm for more aggressive action in prediabetes patients invigorating, even if her suggested goals were quite ambitious – most providers have a difficult time getting patients to halt the progression of their prediabetes, let alone reversing the progression. During Q&A, Dr. Perreault did acknowledge some of the challenges involved in using pharmacotherapy in prediabetes, such as the possibility of higher life insurance rates for patients on a diabetes drug (even those without diagnosed diabetes).

  • In an overview of the use of anti-hyperglycemic agents to halt or reverse prediabetes, Dr. Perrault noted that metformin alone has not yielded the most impressive results. TZDs have prevented the progression to diabetes with a higher level of efficacy, but their efficacy must be weighed against the increased risk of edema and other side effects. The CANOE Study found a 66% relative risk reduction of diabetes with a combination of relatively low doses of metformin and rosiglitazone together – although rosiglitazone (GSK’s Avandia) has fallen out of favor, more broadly speaking the results support the utility of combination therapy (an increasing therapy trend in type 2 diabetes) in the prediabetes setting. Dr. Perreault highlighted the results of the SCALE study of liraglutide in prediabetes, which she believes did not receive sufficient attention for how impressive they were (read our report on that data).

Questions and Answers

Q: What are the difficulties in getting coverage for medications for prediabetes? Does it affect patients’ insurability for life insurance if they happen to be on diabetes medications, even if they do not have diabetes?

A: In terms of reimbursement, most insurers will pay for metformin because it is so cheap. Other medications will start getting more coverage. I have talked to physicians all over the country, and some use newer drugs such as GLP-1 agonists in prediabetes patients, and some get it covered. It will depend on payer mix, will varies for all of you. In terms of life insurance, when you have a diagnosis of anything, it will mean a slight increase. It will likely impact life insurance, but I am a doctor, so my priority is doing the best thing for my patients’ health.

Q: Is there any data in patients with diabetes who then achieved normoglycemia? Should we aim for those same targets in patients with diabetes?

A: I don’t know if I’ve seen any cardiovascular benefit data in people with diabetes that have gone back to normal. The whole idea of setting normoglycemia as a target hasn’t been a target of our consciousness or the literature, although it is something that we know in our hearts.

Q: Do you think that DCCT/EDIC data showing a decrease in CV events in type 1 diabetes patients with excellent control shows that having good control in diabetes in general may have benefit in preventing or reversing CV risk?

A: I never really think of DCCT/EDIC data as relevant for type 2 diabetes patients. The diseases are very different in terms of their pathophysiology. I don’t know for sure – I suppose it is an interesting thought.

Immune Intervention in Type 1 Diabetes

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

Dr. Jay Skyler discussed past and current efforts to stop or slow the autoimmune destruction of beta cells in type 1 diabetes. Although there are a few bright spots in the field of research, Dr. Skyler’s overview was quite sobering overall, as there has been a large number of disappointing results in recent years. Primary prevention strategies are at a relatively early stage, though Dr. Skyler suggested that we might see clinical data from the TRIGR study of casein hydrosylate formula (as opposed to regular milk) at this year’s ADA. Most therapies being investigated to slow the progression of the disease post-diagnosis (such as anti-CD3 and anti-CD20 antibodies) have only been able to transiently show an improvement, before C-peptide levels resume their decline. As we have heard from him before, Dr. Skyler suggested that a combination of therapies (including agents to modulate the inflammatory response, antigen presentation through vaccines or insulin administration to tilt the immune system towards preservation, and even a GLP-1 agonist to improve endothelial dysfunction) could succeed where monotherapy has failed, although he acknowledged that administering a cocktail of three or more drugs to young patients may be a tough pill for many prescribers to swallow. Recently, at ATTD, Dr. Skyler also spoke on immunotherapy for type 1 diabetes, wisely noting that “hype” and “hope” differ by only one letter (see item #9 of our ATTD 2014 Day 3-4 Report).

  • Given that there is an extended sequence of events leading from genetic predisposition to diagnosed type 1 diabetes, there are a number of points where immune-based prevention interventions could be targeted. Dr. Skyler categorized possible therapies as primary prevention strategies (which would be targeted towards patients with genetic predisposition), secondary prevention strategies (which would be targeted towards patients with pre-type 1 diabetes short of fully diagnosed type 1 diabetes), and tertiary prevention strategies (which would be directed towards the recently diagnosed). He discussed proposed interventions from all three categories.
  • Many agents have shown promise for the primary prevention of type 1 diabetes in preclinical NOD mouse studies, but it remains to be seen if large clinical trials will support those results. A Finnish pilot study found that a hydrolyzed milk formula (relative to regular milk) delayed the initiation of beta cell autoimmunity, although it was a low-power trial. The full-scale TRIGR study is ongoing ( Identifier: NCT00179777), and Dr. Skyler shared that the results from the trials’ antibody endpoint might be presented at this year’s ADA – results of the type 1 diabetes endpoint are expected in early 2017.
  • Clinical trials of secondary prevention interventions have not yet yielded much success. A summary slide of clinical trials showed that the DPT-1 trial’s parenteral insulin cohort, the ENDIT trial on nicotinamide, the DIPP trial on nasal insulin, and the NIP trial on docosahexaenoic acid showed no lasting effect. The DPT-1 trial’s oral insulin cohort showed no significant benefit overall, but did show a benefit (~10 year projected delay of progression to type 1 diabetes) in a subset of patients with the highest baseline concentration of autoantibodies (Skyler et al., Ann NY Acad Sci 2009). Dr. Skyler stated that this section of the study is being repeated, as the ~10 year projected delay was found in a retrospective analysis.
    • Dr. Skyler presented an overview of the three ongoing studies (and one additional proposed study) in the TrialNet program, a network of researchers exploring type 1 diabetes prevention strategies. Dr. Skyler is the chairman of the TrialNet program, and put in a plug for interested prospective investigators to apply at or 1-800-HALT-DM1. He shared that TrialNet is looking to initiate a new early treatment study enrolling patients with only one autoantibody.
  • A theme among proposed tertiary prevention monotherapy interventions is that they are able to transiently delay, but not prevent, the progressive decline in C-peptide secretion. Anti-CD3 antibodies (teplizumab, otelixizumab), anti-CD20 antibodies (rituximab), T-cell co-stimulation blockers (abatacept), antigen-specific therapies (GAD vaccine, oral insulin), cytokine antibodies (canakinumab, anakira, etanercept), memory T-cell modulators (alefacept), and other compounds like thymoglobulin and Andromeda’s DiaPep277 have not shown conclusively positive prevention effects; most are able to prevent the decline in C-peptide secretion for a period of a few months to a few years, but following that period, secretion declines in parallel with the placebo group. A recent full-scale clinical trial on DiaPep277 found no effect on the long-term preservation of meal-stimulated C-peptide, but did demonstrate an effect on glucagon-stimulated C-peptide – Dr. Skyler withheld speculation on that specific therapy, although he expressed hope that it proves successful, as it is a relatively simple therapy to administer and appeared to have a durable effect in a pilot study. Early data on IL-2 therapy shows promise, but finding the right dose is very important (read our coverage of a presentation on IL-2 at EASD 2013 for more details).
  • Where individual preventions have failed, could a combination succeed? Following intriguing results from a controversial clinical study (Couri et al., JAMA 2009) using a combination of high-dose cyclophosphamide, ATG and AHBMT, Dr. Skyler suggested that combination therapy (albeit a less intense combination than the one tested by Couri et al.) could target the pathophysiology of type 1 diabetes from multiple angles. A combination could involve prolonged therapy with an agent to prevent the inflammatory component (anti-IL-1-beta or anti-TNF), cyclical long-term administration of GAD vaccine or oral insulin to tip the immune system scales towards protection, and (interestingly) a GLP-1 agonist to improve the endoplasmic reticulum stress that has been shown by Ferrannini et al. to accompany pre-type 1 diabetes. In addition to those long-term therapies, Dr. Skyler’s combination would start with a short course of anti-CD3, anti-CD20, or a co-stimulation blockade, as well as short-term administration of regulatory T cells or agents that would activate them.
    • Dr. Skyler acknowledged that his proposed combination might be perceived as very intense, but noted that the long-term components are relatively safe and well understood. One way to improve the risk/benefit balance of such a combination would be to identify better biomarkers for both the likelihood of progression to type 1 diabetes, and the likelihood of success on a given preventative therapy.
  •  See our Rachmiel Levine 2014 Day #2 Report for another recent presentation on immunotherapy for type 1 diabetes, by Dr. Carla Greenbaum (Benaroya Research Institute, Seattle, WA). In her talk, she cited the need for assays that can characterize pathogenic antigen-specific cells.

Questions and Answers:

Q: You presented data indicating that antibody administration had an effect over around six months, followed by a decline. Is there a reason why you can’t keep giving booster shots of the antibody?

A: With the anti-CD3 therapy, there have been studies of a six-month course and even a one-year course, and the long-term outcome at two years is identical. It hasn’t yet proven helpful to come back with a second course of treatment. We don’t know why that is. We have not been willing to give second doses of rituximab, the anti-CD20, because we have evidence of continued immune interference going out to 18 months following short-term treatment, and we don’t want to risk compromising patients’ immune systems.

Q: Even in the DPP, we saw elevated antibody levels in many patients. As many as 12% of patients that came in had antibodies. What is the phenomenon behind that occurrence?

A: When one looks at the rate of the development of type 1 diabetes in most studies, it is much faster the younger you are. Once you get to age 25-30, from Danish data, it appears that the rate of incidence remain fairly constant through age 90. You can have type 1 diabetes, but the loss of beta cell function is very slow when you get up into that age group. Also, there is nothing preventing people from having both type 1 and type 2 diabetes. We’re increasingly seeing kids with type 1 diabetes who become obese and insulin resistant. I call that type one-plus-two, because they really do have both diseases. You have to deal with both the immune component and the insulin resistance component in those cases. There is nothing about autoimmunity that prevents you from becoming obese, and there is nothing about obesity that protects you against autoimmunity.

Panel Discussion

Q: From the discussion, it seems to me that weight might be a bigger issue for the management and prevention of diabetes than blood sugar control. It’s an interesting paradigm shift that we seem to be observing, that seems to be occurring right now. Dr. Perrault, could you comment on that?

Dr. Leigh Perrault (University of Colorado, Aurora, CO): I think that weight reduction has pleiotropic effects that go beyond glucose lowering. As we all know, when you have a patient that successfully loses, weight, you see reductions in blood glucose, but also blood pressure, triglycerides, and cholesterol. If you can change the underlying physiology, you will change a lot of things that come with it.

Q: Dr. Garber can you comment on your view of why the Look AHEAD trial did not have cardiac benefits? Given that in your algorithm you are giving a lot of weight not only on just on preventing but treating diabetes with lifestyle.

Dr. Alan Garber (Baylor College of Medicine, Houston, TX): Weight is not going to solve the excess CV risk. Weight seems to have the greatest benefit on pulmonary complications of obesity. For patients who have failing beta cells, it seems to really do a lot of good for retention of beta cell function. I think there’s a whole separate pathway for cardiovascular disease (CVD). It’s not a central component of obesity. We already know if you get blood pressure under control, lipids under control, and blood sugar under control, most obese individuals have no major increase in CVD. To me the situation is fairly clear. If you want to treat CVD, treat their LDL, treat their blood pressure, do what you can for their blood sugar, at least improve it, and if you want to improve the blood sugar long term, if you want to improve the functionality of the patient, then you lose weight.

Q: At the Cleveland Clinic, we are being told by our bariatric surgeons that we can do all kinds of different things to get people to lose weight, but after a period of time, those patients will gain their weight back. There is a 90% failure rate in terms of achieving sustained weight reduction. After that, those surgeons say that everyone should get gastric bypass. You, Dr. Garber, don’t seem to be suggesting that everyone who is overweight should get gastric bypass.

Dr. Garber: Somehow, I can’t imagine all 80 million prediabetes patients all walking into the Cleveland Clinic and demanding to see Philip Schauer for gastric bypass operations. If you want to argue this point further, Dr. Schauer will be speaking at a symposium this afternoon on obesity as a disease.

Dr. Ed Horton (Harvard Medical School, Boston, MA): I agree essentially with what Alan has said. I think the lifestyle changes focusing on weight control and weight loss and increased exercise are a fundamental part of managing overweight patients with diabetes, but it’s not the only thing. We really have to take advantage of the meds we have plus the lifestyle. The results of look AHEAD really demonstrate that you can get to the same point by different routes. In the control group it was largely through use of more meds and better care in that regard. In the lifestyle group it was through more weight loss and more exercise. So there are different pathways to the same endpoint. I’m a strong believer of using them in combination. One thing we do at Joslin is we have a clinic designed for people who are significantly overweight with type 2 diabetes. It’s run by Dr. Hamby. One of the things we do is look at the medications we use to manage glucose control and pick ones that are weight neutral and help promote weight loss rather than ones that promote weight gain. One of the things we’ve always faced with in our office is telling the patient to lose weight, and you put them on an SFU or insulin, and they gain ten pounds. It’s a really difficult situation. I think you have to use both – good judgment in which meds you use and an intensive lifestyle program.

Q: Dr. Skyler, you presented some very interesting data looking at different interventions to try and prevent type 1 diabetes, some of which seemed to have major benefits. What I really didn’t see was any of the side effects that occurred with those interventions. Could you discuss some of them now?

Dr. Jay Skyler (University of Miami, Miami, FL): Let me confine my answers to the therapies that showed benefit, which were anti-CD3, anti-CD20, and abatacept. The nice thing about abatacept was that there were nearly no side effects, which is why we were willing to test that sooner in the progression. Anti-CD20 suppresses IGM and B-lymphocytes, and they remain suppressed for 9-12 months. IGM remains suppressed for at least two years, which is the big problem there. There were also some infusion reactions that occurred, but they were relatively minor. Infusion reactions were more common with anti-CD3, principally related to cytokine release leading to headaches, fever, and some postural hypotension. Those could be readily controlled, and were transient. I would characterize them more tolerability issues, as they weren’t long-term side effects. If you are giving a drug for 14 days and getting a benefit for much longer, you can live with those effects. You did get some palmar skin changes with the anti-CD3, but they dissipated over time. What we don’t know is what the effects will be if we move towards combinations of these drugs.

Q: Do you feel that within the next five or ten years we’ll have some way of actually preventing diabetes or do you think that’s longer term?

Dr. Skyler: I’ve learned over the years that one should not try to project timelines on things because as you were just pointing out everyone thought degludec would be on the market two years ago. I think timelines are difficult to project. I think we’re going to need to have a change to the regulatory approach to the way companies look at things to be successful. I think we will need combination approaches where we attack different components of things, meaning we’ll need to get different manufacturers together and get regulators to look at things differently. In animal models there are at least examples of one drug having no effect, but in combination it magnifies the effect of other drugs. In humans for example, leptin alone did not create weight loss but leptin and pramlintide resulted in 10% weight loss because pramlintide overcame the leptin resistance. They have to get over that relatively narrow view that each drug has to have an effect first before putting it in combination with something else. It’s not an easy task to fight all these battles.

Q: Dr. Garber, do you see other medications on the horizon to treat obesity that will have fewer side effects than the ones that have recently been made available?

Dr. Garber: Contrave is in development, and from my view of the phase 3 data, I think it will have about the same level of side effects as the two existing agents, lorcaserin and phentermine/topiramate XR. As you know, patient responses to individual medicines can vary greatly, so it always helps to have additional medicines on the market. So far, there is no panacea that will knock weight down by 10-15% by inhibiting weight centers in the hypothalamus without having adverse effects elsewhere.

Q: Dr. Garber, what are your thoughts on the liraglutide data?

Dr. Garber: Liraglutide is an interesting compound. For full disclosure I am the chair of the global advisory board for liraglutide. A number of years ago looking at data was abundantly clear that they had saturated glucose lowering effects at about 1.2 mg, 1.8 mg. That’s the does for diabetes, but weight loss kept on going. So Novo Nordisk did additional studies with 2.4 mg and 3.0 mg under some duress. In any event, they were amazed to see just continuing greater and greater weight loss. As far as I can tell, the 3 mg dose that’s in development is still not the top dose with regard to weight loss – if you look at the dose response curve they’re still climbing. They’re not at steady state at 3 mg. So I think you’ve got a drug with virtually unlimited potential.

Q: I was intrigued by Dr. Skyler’s data on responders versus non-responders. In the era of precision and personalized medicine, should there be a different strategy of looking for predictive biomarkers of responders? From the success of recent orphan drugs in other areas, could you see the possibility of finding subgroups of type 1 diabetes patients or individuals at risk for type 1 diabetes to target?

Dr. Skyler: I agree, and that’s what I was implying in my presentation. The big effort now is to identify the appropriate biomarkers. A vexing problem is trying to figure out what combination of genes play a role in this, because there are at least forty-something genes that influence type 1 diabetes. The bulk of the effect comes from the genes in the HLA region, but there are a number of others. A big effort on this front is coming from the University of Virginia. Hopefully we’ll start to see gene patterns that may predict which type of approach to use in which type of patient.

Q: When you do a study like that and you take out that group of patients that respond or don’t respond, could you do the same thing with a placebo group and say there was a group of placebo patients that seemed to not progress and another group that did progress?

Dr. Skyler: Figuring out the definition for responders is difficult. In the study I showed, the definition was people who did better than best patient in placebo. Therefore, by definition in that particular analysis there were no responders in placebo. Other ways we looked at it were greater than 7.5% fall from baseline C-peptide. In that case you do get placebo patients that respond obviously. It’s challenging to figure out those definitions. I’ve got one slide where you can show eight different definitions of responders. You have to have something to look at consistently. On the other hand, for a particular drug you could pick it out from responders within. It’s a rough area. Biomarkers in general are rough.

Diabetes & Exercise – Implications from the Look Ahead Trial

Ed Horton, MD (Harvard Medical School, Boston, MA)

Dr. Ed Horton provided an in-depth look at Look AHEAD results at years one, four, and 10 and a look to the future of the study. Dr. Horton ultimately believes that Look AHEAD’s neutral result is good news for everyone involved: he interpreted the failure to see a difference being due to the control group doing extremely well, indeed the CV event rate in the group was half of the 2% that was projected at the trial’s outset, and statin use in the control arm was very high. Look AHEAD participants are being followed through August 2015, and areas of interest that have emerged from sub-analyses include: obesity-associated malignancies, cognitive dysfunction, and mobility and physical function.

Questions and Answers

Q: Where do we go from here? Should we not be telling people to exercise or do you feel like the control group was such that they were all people that were willing to do an intensive exercise program had they been randomized to that arm so that the control group was not a typical control group that you would get in the normal population?

A: I think the press release when these results came out said this is a negative study. I think the opposite. I think it was a very positive study. While we didn’t see differences in the primary endpoint, which was CV events, there are many other positive findings that came out from the intensive lifestyle program. I think it would be a huge mistake to conclude that we should advise patients against weight reduction programs, exercise, and a healthy diet. We certainly should continue doing that. There were many other benefits besides this primary endpoint. It’d be huge mistake not to.

Q: Do you think that every CV trial that has been done in the last decade has been doomed because they’re based on event rate data that is antiquated?

A: That’s a good question, and I don’t know the answer. What we’re seeing in general is a reduction in the CV event rate because people are getting better care. You have to look at all the statin studies. Use of statins has gone up in our population. Everybody is paying much more attention to blood pressure control. I think we’re providing more global care to our diabetic patients. We’ve moved away from the glucocentric approach to looking at CV risk reduction in general, and that makes it difficult to really show a difference with any particular intervention. I think the other factor that I’ve thought about in this study is that all of the volunteers from Look AHEAD wanted to be randomized to the intensive lifestyle intervention program. Those in the control group were disappointed when they didn’t get randomized, and many of them got the message and started doing lifestyle modification on their own even though that wasn’t part of the protocol. The failure to find a difference was partly due to the control group.

Q: Do you have any intention of doing any economic analysis? You show that the intensive lifestyle had less use of medications, but this program also costs money. To me the more money we invest in time with patients costs us money, but I think down the road it costs less than other interventions. Just curious to know whether you saw reductions in costs in patients over time, especially as they begin to age.

A: The economic analysis is being done, and the lifestyle program, if you looked at translating this into a real-life situation, is certainly considered to be within the cost-effective range of treatment. You can do your arithmetic and say there’s a cost of lifestyle program, but it reduces the costs of meds, and it turns out to be pretty much a wash.

Bariatric Endocrinology – Chronic Disease Management of Overweight and Obesity

Weight Loss Medications

W. Timothy Garvey (University of Alabama, Birmingham, AL)

Dr. W. Timothy Garvey argued for the long-term (potentially life long) use of anti-obesity drugs. Dr. Garvey posed the question to the audience “Should there be time limits on the pharmacotherapy for obesity?” Overall, the audience seemed hesitant and uncertain answering the question. When an attendee did answer the question she suggested that she would probably only prescribe a currently available option for about two years. Her reasoning was that when older obesity drugs became available, they were considered to be safe; however, after about five or more years they were found to have negative CV effects (e.g., sibutramine, phen-fen). Given that only two-year data exists for Vivus’ Qsymia (phentermine/topiramate ER) and Arena/Eisai’s Belviq (lorcaserin), she does not think she would prescribe them for much longer than that. In response Dr. Garvey asked if her judgment was an “emotional decision, or an evidence based decision” and argued that Qsymia and Belviq are safe and effective. We did not think that this necessarily a fair dichotomy to raise, because, more than two years of safety evidence does not exist for either Qsymia or Belviq. Dr. Garvey proceeded to argue that the woman’s hesitation on the long-term use of an obesity medication was because the history of the anti-obesity field has “invoked fear.” He noted that Qsymia and Belviq have longer data than many medications for other areas – for example, anti-hypertension drugs are approved with only six-month of data. Dr. Garvey pressed that while long-term data and surveillance does need to be conducted and that doing no harm is the priority, he believes HCPs should commit to using them long-term. Additionally, Dr. Garvey reminded attendees to consider the consequences of not treating obesity.

  • Dr. Garvey also compiled data on the different anti-obesity pharmacotherapies’ impact on glycemic control into one succinct table. Without head-to-head trials one cannot truly compare the various agents’ impact, however, the data does suggest that the available options have relatively similar effects on A1c (placebo-adjusted).







Liraglutide 3 mg

% Weight loss













Baseline A1c







A1c change














% Reaching A1c ≤7%













Effects of drug on need for OADs







Press Conferences

Blood Sugar Basics: Get to Your Goals

Etie Moghissi, MD (AACE Board of Directors)

Dr. Etie Moghissi gave a brief 10-minute overview of ACE/AACE’s Merck-sponsored Blood Sugar Basics program, for which she is the physician advisor. The goal of the program is to empower patients through education, and the program’s newly-updated website ( features a number of patient resources, including handouts, low and high blood sugar checklists, questions that patients can ask their doctor or other HCP, and a daily management journal. New additions to the program in 2014 include a series of three missions for diabetes patients: (i) gathering intelligence by talking to their provider; (ii) finalizing a strategy by setting goals and committing to them; and (iii) regrouping by checking in again with their doctor and assessing their progress. The program’s outreach until now has largely focused on endocrinologists, but Dr. Moghissi shared that the program increasingly hopes to work with primary care physicians with in-office brochures.


--by Adam Brown, Hannah Deming, Jessica Dong, Jenny Tan, Manu Venkat, and Kelly Close