CC - ATTD 2016 (Advanced Technologies and Treatments for Diabetes) Day #1 Highlights

Executive Highlights

Ciao from Milano, where ATTD 2016 is in full swing at the bustling Milano Congresso (MiCo) Conference Center. There are delegates from 82 different countries attending the conference, with the greatest number coming from Italy (360), USA (331), Switzerland (203), Germany (182), and France (124).

Highlights from today’s program included a look at Medtronic’s Guardian Connect (its standalone, Bluetooth-enabled mobile CGM, competition for Dexcom G5) and Sanofi’s MyStar DoseCoach (BGM with integrated insulin glargine dose titration algorithm). Our glance around the exhibit hall also featured new news from CeQur – who shared timeline updates on PaQ’s launch in the US (on track for 2017) and EU (delayed from 2016 till after US launch) – and news that Unomedical plans to launch a new all-in-one infusion set-insertion device in 2016. Not to be outdone, Abbott’s corporate symposia drew a HUGE crowd to discuss FreeStyle Libre (no major updates) while a session of psychosocial aspects of the artificial pancreas shared new data reinforcing some of the key challenges for automated insulin delivery (provider time, reimbursement, varying patient expectations, lack of education). Please see below for our top eight highlights from day #1, along with exhibit hall coverage and detailed commentary of select sessions. Our preview is here for what’s on tap over the next three days.

1. In an exhibit hall surprise, Medtronic displayed Guardian Connect, its standalone, Bluetooth-enabled mobile CGM targeted at patients not on pumps. Signs indicated it is currently under CE Mark review, which could presumably place an EU launch sometime this year or early next year. This is a direct competitive answer to Dexcom’s G5. No EU plans had ever been disclosed for this device; as of JPM 2016, a US launch was expected by April 2017. Picture below!

2. Sanofi debuted its MyStar DoseCoach meter in its exhibit hall booth – the BGM device integrates an insulin glargine dose titration algorithm, providing guidance on insulin dose adjustments for patients with type 2 diabetes. We’ve been waiting forever for help on insulin dosing – this is so needed and we’re incredibly excited to see it. Great job to Sanofi for more on the COACHing front (as a reminder, Coach is a program used for Toujeo users).

4. Abbott’s standing-room-only corporate symposium shared continued enthusiasm for FreeStyle Libre – organizers were forced to set up an overflow room … even after a year on the market and a session with no major updates. We’ll hear REPLACE and pediatric data later this week.

8. During the moving opening ceremony, Dr. Antonio Ceriello (University of Udine, Udine, Italy) argued that glucose fluctuations cause greater oxidative stress and present an increased risk of complications relative to stable high glucose levels. To that end, he shared new unpublished data from his team’s study demonstrating that intermittent high glucose levels were associated with significantly lower levels of antioxidant GPx1 than constant high glucose levels, leading to increased cellular damage from free radicals. We’d love to get more information from “big data” sources on glycemic variability in the “real world” given that no one is going to fund a ten-year trial …

Table of Contents

Top Eight Highlights
Detailed Discussion and Commentary

Top Eight Highlights

1. In an exhibit hall surprise, Medtronic publicly displayed Guardian Connect, its standalone, Bluetooth-enabled mobile CGM targeted at patients not on pumps. Signs indicated it is currently under CE Mark review, which presumably means an EU launch could occur sometime later this year or early next year (our speculation, as reps would not comment on launch timing). The transmitter is the familiar clamshell design, though it will send CGM data via Bluetooth directly to a mobile app on Apple iOS at launch - cool. Medtronic does not plan to launch a standalone receiver, so patients will only get the transmitter and app (unlike G5, where they have the option of using a receiver or the app). We think that’s smart – simple = better. Interface-wise, Guardian Connect looks similar to the MiniMed Connect app, while retaining the text messaging notifications for caregivers and auto-upload to CareLink. We didn’t get to do a hands-on demo, but the app had the familiar glucose value, trend, and trace on the home screen, plus the obvious sub-menus for calibration and settings (see picture below). We see this as a key competitive answer to Dexcom’s G5 and Abbott’s LibreLink, plus an important Medtronic foray into MDI (Medtronic’s current real-time CGM requires a paired pump). Though we have known about this device since September 2014 (pivotal study completed in August 2015), this is the first time it has ever been on display in final commercial form in an exhibit hall. The As of JPM last month, Medtronic expected a US launch of Guardian Connect with Enlite 3 by April 31, 2017 (in FY17); there was no EU timing listed on the slide, which is what made today’s announcement unexpected. We’re elated to see the sensor field moving incredibly rapidly to make systems more convenient to use – who would have thought two years ago that we’d have all three players with smartphone apps to view data?! We hope data collection can move quickly so that access questions are at the forefront and broad access is possible.

2. Sanofi debuted its MyStar DoseCoach BGM in its exhibit hall booth – the BGM integrates an insulin glargine dose titration algorithm, providing guidance on insulin dose adjustments for patients with type 2 diabetes. This was also a big surprise! According to sales reps, the device received a CE mark this past December, and this ATTD display is its first exhibit hall promotion. We assume it could conceivably launch soon in Europe; unfortunately, there is no US timing to report. The MyStar DoseCoach comes with three treatment plans based on “well-accepted titration algorithms” – these are entered into key cards that can be inserted and activated into the device by healthcare professionals (the plans differ on aggressiveness and target range). After selecting the most appropriate treatment plan for the patient, the healthcare professional can then individualize the Dose Helper by entering the patient’s weight, starting dose, and usual dose time. The patient then uses the device like a normal BGM to measure fasting blood glucose levels and to record insulin doses; after the patient runs Dose Helper, the device suggests an updated insulin glargine dose based on the selected treatment plan. Notably, the device is currently only available with Sanofi’s Toujeo (insulin glargine U300) for type 2 diabetes – in our conversations with reps, the company first wants to see how this device will fare on a smaller scale prior to any considerations for expansion. We have expected Sanofi and its MyStar initiative (read more about the initiative from its launch at ATTD 2013) to launch such a device for quite some time and we are very excited to see this new class of devices take flight. We hope this system’s convenience and more individualized insulin guidance will bring more patients to goal and save HCPs time.

We also like that the MyStar DoseCoach prioritizes safety, with multiple built-in verification and human factors rules. For example, prior to a fasting blood glucose measurement, the device confirms with the patient that the blood glucose measurement is indeed a fasting one, even going so far to help define what a “fasting” definition entails. While such notifications can be annoying, accuracy is critical when systems like this are giving insulin dosing advice. The tradeoff is worth it in our view! Patients have a range of views, of course.
MyStar DoseCoach downloads to MyStar Connect, a management software platform that collects all of the patient’s MyStar Dose Coach data (i.e., graphs and logbook) along with other health records in one spot for healthcare professionals. This data unfortunately cannot be synced in real time as there are no cloud connections (it must be manually connected during a doctor’s visit). We wonder what the plans are to integrate with Diasend, Glooko, and Tidepool, as well as how this product could fit within the Verily (Google Life Sciences) partnership.

3. The first day of ATTD brought two major pipeline updates for CeQur’s second-gen PaQ insulin delivery device: (i) CEO Mr. Doug Lawrence confirmed that the company is on track for an FDA 510(k) filing in 2016 for a 2017 launch, while (ii) exhibit hall reps shared that the company has delayed EU commercialization until after the US launch (previous guidance put the EU launch first, in 2016). We are not sure what is responsible for the about-face though it sounds like the decision will enable CeQur to scale up manufacturing and seek out US reimbursement without the presumable distractions of commercialization in the price-sensitive EU market. Reading between the lines, we wonder whether the latter was proving especially difficult in the EU, given both that there is no predicate device [Valeritas’ V-go is covered by third parties only in the US and reimbursement is way more of a hassle than it should be given the major benefits it has shown in terms of A1c and adherence] and that the team still has limited clinical data. CeQur did recently complete its third small pilot study (see the data below) and while the findings have been very encouraging (improvements in A1c have been >1.5% across the board), it is may take much larger observational studies and RCTs of glycemic impact and cost-effectiveness vs. MDI to appeal to payers. On the US launch, we also learned that PaQ will be indicated for “diabetes” broadly, not specifically for type 1 or type 2; of course, it’s been marketed toward the latter though we think the labeling choice is smart considering many patients with type 1 could find the simple device convenient and easy to use, especially those transitioning from MDI to pump therapy. Fortunately, CeQur does have funds to work with – considering the impressive $100 million in Series C financing the company raised last September – though we imagine it’ll take much of that to fund clinical operations and a launch.

CeQur’s Dr. Julia Mader presented encouraging preliminary data from the company’s study of its insulin delivery device in patients with type 2 diabetes (n=15). Preliminary results indicated that PaQ treatment reduced A1c 1.5% from a baseline of 8.8% (p<0.0001). For context, the trial consisted of three study periods: baseline MDI (one week), transition from MDI to PaQ (1-2 weeks – to ensure that patients were on an appropriate basal rate), and PaQ treatment (12 weeks). Looking at individual data, it was great to see that every patient who transitioned to PaQ experienced an A1c reduction (i.e., 100% of patients saw some glycemic benefit) and given the short duration of the study, we would note that the 1.5% A1c reduction is very impressive. Secondary outcomes were also encouraging as total daily dose of insulin (TDD) and body weight were not statistically different from baseline – both metrics technically hinted at a slight increase though Dr. Mader attributed this trend to an artifact of the small sample. On the insulin dose front, 73% of patients successfully transitioned to PaQ with the first basal dose, which we hope is a good sign that HCPs will have an easy time transitioning patients onto PaQ. No severe hypoglycemic events or other serious adverse events were reported.

4. Abbott’s standing-room-only corporate symposium shared continued enthusiasm for FreeStyle Libre – organizers were forced to set up an overflow room … even after a year on the market and a big-time presence at three major EU conferences (EASD 2014, ATTD 2015, EASD 2015). This isn’t new but it’s encouraging to see just how much conference excitement the franchise continues to generate, an observation that perhaps stands as the greatest testament to the product’s transformative potential. The symposium featured no significant pipeline updates but we did hear from Dr. Gerry Rayman (Ipswich Hospital NHS Trust, UK) who presented three case studies from Abbott’s six-month outcome study – REPLACE (n=210 type 2s on MDI, A1c>7.5%) – that could support reimbursement for Libre in type 2 patients. All were success stories – one had a 1.8% A1c reduction, another had an ~66% reduction in hypo, and the third had both – though we’re looking forward to seeing the full data in an oral presentation on Friday. On accuracy, Dr. Rayman provided a stirring endorsement of the technology, noting that patients in all three cases went from multiple fingersticks per day to “virtually none” once they started on Libre – it’s always good to hear that the factory calibration lives up in practice. The remarks echo commentary on the excellent hypoglycemia accuracy we heard at last year’s ATTD and continue to hint at how this device is being used in the real world (confirmatory fingersticks are technically required in hypoglycemia, when symptoms do not match the readings, and times of rapid change). Broadly, the consumer experience was the most central theme throughout the session and we bring you some of our favorite quotes below:

“I had a family come in to see me that was using CGM. They saw Libre and went out and purchased it. They came back a few weeks later with huge smiles. They said, ‘I have a present for you,’ and handed me the CGM and said ‘I don’t need it any longer. You can give it to any patient you want.’”
“Patients really like Libre. I’ve had 50-60 people use it now. Not one has come back and said that this is not for them.
“Libre frees you to interrogate the device when you want to and not when the devices tells you to.”
“EVERY patient I’ve had on FreeStyle Libre is convinced that this is a major benefit to them.”
There was no update on the US regulatory timing, though we learned in the 4Q15 call that the product is “optimistically” expected to launch by the end of this year, which was the first news we had heard on this front. We’re not sure if it is at FDA or not, but would guess that it has not been submitted. We look very forward to the FreeStyle Libre Pro product coming to the US earlier than this. The ability to drive therapeutic change seems big though we’d love to find out more about how that is working in the real world.
Abbott reps informed us in the exhibit hall that the company is “coming along” in the process of developing an Apple LibreLink app for scanning FreeStyle Libre sensors without the need for a separate reader. Excellent! The timeline is not yet clear but we did learn interesting details about what the platform expansion might entail – the company is thinking about building an iPhone adapter to power the NFC-reading capability. See more below!

5. The psychosocial artificial pancreas session shared early data from a fascinating new HCP survey (n=370) and patient focus groups (n=60), illustrating some of the key challenges for automated insulin delivery (AID) – provider time, reimbursement, varying patient expectations, and the need for serious education. Most notable were the results from the HCP survey, which sampled an early adopter population (370 respondents out of 10,000 invited) that was clearly enthusiastic and comfortable with diabetes technology (e.g., 91% believed that AID will optimize diabetes control in type 1). A notable 66% of providers reported that they currently have difficulty getting insurance coverage for CGM in type 1s (only 33% for pumps), and the stat was equally concerning for automated insulin delivery: 74% of providers expect to have difficulty getting insurance coverage for AID once it is available. Also alarming (though to be expected) were perceptions about the time investment automated insulin delivery will require. A sound 63% of providers believe that AID will require more professional time than previous approaches to insulin delivery (at least at the start this is certain to be true – virtually all first to market devices do), and 95% believe that educators or trainers will be required to implement and manage AID in patients. Interestingly, there was a pretty even provider split on perceptions of patient cost-sharing: 39% of providers think patients will be willing to share the cost of AID with insurers and payers vs. 21% that were neutral vs. 41% that don’t think patients will share costs (in which they likely won’t have access, in our opinion). A notable 33% of providers – and these were individuals clearly comfortable with technology – said they will wait to prescribe AID until there is sufficient long-term evidence of success with the technology. Last, only 47% of providers trust AID systems to perform at the highest level of accuracy and precision – that was good to reinforce that challenges ahead, that is for sure. There were a slew of other questions on AID, pumps, and CGM, but overall, this survey reinforced the provider side of automated insulin delivery – industry must address concerns about reimbursement and time, even in leading clinicians familiar with the technology. We can’t imagine what the results would look like for PCPs. Below, we enclose some of the themes from the patient-focus groups, which reinforced the need for education and setting clear expectations about what this technology can and cannot do – there are clearly a wide spectrum of patients views on AID, and industry will have to wrestle with which groups to appeal to and with what terminology (e.g., pump/CGM users vs. MDI users).

6. In a rare pipeline update, infusion set juggernaut Unomedical disclosed plans to launch a new all-in-one infusion set-insertion device in 2016. The single-touch, fully automatic serter will completely hide the insertion needle and retract it, and the whole unit will automatically detach from the inserted site. It’s an especially great innovation for those afraid of needles, those with poor dexterity, or those who have trouble inserting sites in difficult to reach places. We also like that it is full disposable and self-contained, adding a lot of all-in-one convenience for patients. Presumably it could improve the consistency of infusion site insertion, potentially even reducing site trauma (our speculation). A poster in the booth also highlighted some earlier stage research at Unomedical: increased infusion set longevity (suppressing the inflammatory response at the insertion site, preventing cannula encrustation) and ensuring healthier skin throughout wear time (skin-friendly adhesive patches). There was no timing attached to these efforts, and a rep confirmed they are in the basic research phase. After years of little innovation in infusion sets, we are elated to see two new meaningful products coming to market this year – BD/Medtronic’s FlowSmart set (~March-April) and now Unomedical’s new automatic inserter! We expect these products to make pumping better for patients already on them, though we wonder about their long-term potential – can better infusion sets significantly enhance the clinical benefits of pumps/artificial pancreas systems, pushing more HCPs to prescribe them and more payers to encourage them? That would be the dream, of course.

7. Animas algorithm scientist Dr. Daniel Finan delivered a dry background presentation on the artificial pancreas, finally revealing in Q&A that Animas is currently planning its next study of the Hypo-Hyper minimizer system with the FDA. The study will include an ambulatory component, but it still sounds small, will include remote monitoring, and seems far from a pivotal trial. It’s unclear what the time to market is, but we assume two years or more. [Editor’s Note: On March 1, Animas contacted us to share that this study IS in fact a pivotal study! This was not evident in Dr. Finan’s remarks in Q&A. More details here.] The other sliver of an update was that Animas now considers “off-platform or ancillary software” as part of the artificial pancreas system, suggesting it may incorporate some kind of mobile app or remote monitoring system into the HHM. At least five minutes were devoted to Dr. Aaron Kowalski’s six-step artificial pancreas roadmap, which is now outdated as of his Diabetes Care update last summer. Dr. Finan placed Animas’ HHM as the precursor to a hybrid closed loop (e.g., Medtronic’s MiniMed 670G), suggesting it may not be as aggressive or burden alleviating as the 670G. Overall, the relatively pace of Animas’ automated insulin delivery program has been disappointing but we hope people remember the commercial realities of diabetes devices. As background, of course, the JDRF partnership was signed in January 2010, with the expectation it would be submitted to FDA within four years. We were all a bit more hopeful back then about timing. There’s no question Animas will be later to market than Medtronic’s MiniMed 670G (pivotal study ending in May; expected to launch by April 2017 in the US), and likely behind some others like Bigfoot (pivotal trial in 1H17) and TypeZero (IDCL trial commencing in 2Q/3Q, which could support a PMA) although that also assumes those timelines won’t change. It would obviously be fantastic if Animas could move faster, and we wonder about resources ahead.

Psychologist Dr. Katharine Barnard introduced Animas Academy, a new educational website with practical pump therapy tips for patients, caregivers, and healthcare providers. The website has launched in the UK, with further European countries coming. Animas has done a good job of making the content holistic (everything from basal rates to exercise to sleepovers), layering the content within five broad chapters (Get started with your pump, pump therapy essentials, day to day life, for your kids, HCPs), and refining the content down to key bullet points. The exercise section Dr. Barnard showed had a short summary at the top, “need to know” bullet points, “Do” and “Don’t” bullet points, FAQs, a short training video, and a quiz. The esteemed Dr. Barnard repeatedly emphasized that everything is evidence-based. There is also an HCP section of the website, which includes a virtual pump tour and downloadable PDF guides. Dr. Ralph Ziegler followed Dr. Barnard and walked through the CME-facing portion of Animas Academy, which revolves around four patient case studies and is highly interactive via a live voting system. Overall, Animas has clearly put a major investment in this educational program, and we salute the company for making the content comprehensive (not just about the pump), clear to navigate, and boiled down to key takeaways. We wonder how many patients, caregivers, and providers will actually use it – many pump and diabetes education programs already exist, though significant educational gaps still exist. What is most preventing patients from maximizing pump therapy – is it lack of education, challenges with the products, the difficult of dosing insulin, the food environment, etc.? We’re going to try to track down the very busy Dr. Barnard and get her worldview on some of these questions.

8. During the opening ceremony, Dr. Antonio Ceriello (University of Udine, Udine, Italy) argued that glucose fluctuations cause greater oxidative stress and present an increased risk of complications relative to stable high glucose levels. To that end, he shared new unpublished data from his team’s study comparing the effects of constant versus intermittent high glucose levels on antioxidant enzyme expression in human endothelial cells. The results demonstrated that intermittent high glucose levels were associated with significantly lower levels of antioxidant GPx1 than constant high glucose levels, leading to increased cellular damage by free radicals. Dr. Ceriello presented additional corroborating data from a series of studies stretching back to the early 2000s, highlighting the fact that greater glucose variability has been largely associated with higher risk of complications (we’re working to get details since we have never seen this exactly – we’ve seen poor interpretations of DCCT claiming to disprove this), and emphasized the importance of creating a standard for measuring glycemic variability (which is typically measured using standard deviation or mean amplitude of glycemic excursions [MAGE]). We see this as particularly valuable as therapies and technologies develop that improve time-in-range by reducing hypoglycemia at the (theoretical) expense of A1c. Dr. Ceriello was emphatic that well-designed clinical trials are sorely needed in the field but acknowledged that they are difficult to perform (and fund, we would add). He commented on the ongoing FLAT-SUGAR trial as an example, expressing concern that the study may not be successful long-term due to the complexity of glucose variability and its dependence on a broad range of uncontrollable factors such as blood pressure and cholesterol levels. (As a reminder, what was somewhat unclear from the preliminary FLAT-SUGAR results was whether the magnitude of glycemic variability reductions would be large enough to show long term differences in future outcomes.) We wouldn’t be so quick to dismiss the future potential of FLAT-SUGAR as there are still plenty of analyses to do on these data. It’s entirely possible different perspectives on variability will emerge over time and are hopeful we will learn more on Saturday afternoon when the brilliant Dr. Irl Hirsch is slated to present the next iteration of FLAT-SUGAR results.

Detailed Discussion and Commentary

Artificial Pancreas Psychosocial Measures Project

Introduction

Richard Bergenstal

Dr. Richard Bergenstal’s succinct opening introduction posed a couple of important and impactful questions for the session: Will the artificial pancreas help us reach the triple aim? Will it bring value to type 1 diabetes care? “If there’s only one word to emphasize,” he said, “it’s value. That’s outcomes/cost.” It’s the bar any device or drug must meet, but an especially challenging one to define for automated insulin delivery, given the slew of non-clinical benefits (peace of mind, burden, stress, worry, etc.). As Dr. Bergenstal challenged in Q&A, quantifying those aspects – and bringing them to payers in a compelling way – is not easy. Still, the team working on this project clearly has that end goal in mind, and we’re encouraged to see the progress thus far through both qualitative and quantitative research.

DATA FROM HEALTHCARE PROFESSIONAL SURVEY

Kellee Miller, PhD (Jaeb Center for Health Research, Tampa, FL)

Dr. Kellee Miller shared early results from a valuable HCP survey, which sampled an early adopter population (370 respondents thus far out of 10,000 invited) that was clearly enthusiastic and comfortable with diabetes technology (e.g., 91% believe AID will optimize diabetes control in type 1). A notable 66% of providers reported that they currently have difficulty getting insurance coverage for CGM in type 1s (only 33% for pumps), and the stat was equally concerning for automated insulin delivery: 74% of providers expect to have difficulty getting insurance coverage for AID once it is available. Also alarming (though to be expected) were perceptions about the time investment automated insulin delivery will require. A sound 63% of providers believe that AID will require more professional time than previous approaches to insulin delivery, and 95% believe that educators or trainers will be required to implement and manage AID in patients. Interestingly, there was a pretty even provider split on perceptions of patient cost-sharing: 39% of providers think patients will be willing to share the cost of AID with insurers and payers vs. 21% that were neutral vs. 41% that don’t think patients will share costs. A notable 33% of providers – and these were individuals clearly comfortable with technology – said they will wait to prescribe AID until there is sufficient long-term evidence of success with the technology. Last, only 47% of providers trust AID systems to perform at the highest level of accuracy and precision. There were a slew of other questions on AID, pumps, and CGM, but overall, this survey reinforced the provider side of automated insulin delivery – there are serious concerns about reimbursement and time, even in leading clinicians familiar with the technology. We can’t imagine what the results would look like for PCPs.

DATA FROM QUALITATIVE PHASE

Katharine Barnard, PhD (University of Southampton, UK)

In a bullet-point style presentation, Dr. Katharine Barnard listed the key themes about automated insulin delivery (AID) that have emerged following 60 focus groups of people with type 1 and their caregivers (n=26 adults with type 1; 11 youth/teens with type 1; 15 parents; and 8 partners). Ultimately, data from the HCP survey (above), patient focus groups, and previous evidence will be used to build five age-based measures to quantify the psychosocial impact of AID in type 1 diabetes. The goal is to help regulatory approval bodies, payers, HCPs, and patients effectively assess the psychosocial impact of this technology. Notably, the FDA is actively engaged with the program, and the team plans to meet with them on February 19. Once complete, the measures will be piloted in clinical research studies testing AID systems. As Dr. Aaron Kowalski pointed out in Q&A, this is about moving the conversation beyond A1c, and certainly, AID has a lot of benefits beyond glycemic control. We salute the Helmsley Charitable Trust for funding this valuable program and for researchers for working on it so diligently, in such forward-thinking, patient-friendly, value-oriented ways.

There is an overwhelming expectation that AID help manage diabetes. But what are the different levels of automation? (e.g., carb counting, levels of bolusing). Partners and children and teens had less understanding of AID systems than adults.
There is also confusion around the names artificial pancreas vs. closed loop and if they mean the same or different thing. This will be key once products come to market and for setting clear expectations. There is definitely a need to come up with uniform terminology.
Views on AID as a “cure” also differ. User input and visibility/wearing a device prevent AID from being a “cure.” Some believe that significant minimization of user input would be a cure, while others want to a see significant improvement in glycemic control. It should be acknowledged that even the word cure is in the eye of the beholder!
Hopes/Expectations for AID: reduced mental burden; improved glycemic control; fewer highs/lows and reduced variability; reduced risk of long-term complications; better quality of life due to less diabetes-related stress; less user input = less chance for human error; similar experience to current tech (pumps, CGM) – though MDI users are less enthusiastic about wearing devices; device aesthetics: sleek, modern, color choice; control from smartphone to aid discreet bolusing.
Potential downsides/barriers: Want more evidence that it works; becoming complacent or dependent on the system; forgetting how to perform routine diabetes tasks (similar to how patients forget how to do injections when their pump fails); concerns if insurance stops funding it; stigma associated with the visibility of devices; too many alarms; pain of cannula insertions (CGM, pump); system failure (what is the backup?); CGM inaccuracy; incorrect dosing of insulin; limited real estate on body of children for cannulas.
When is AID most useful? All the time; at night; mealtimes; exercise; when busy such as during work or school; when stressed; for travel or camp. People clearly want to use the technology for different things. We believe “in real life” the simplicity and ease of use will be key – we think patients will find the value at night to be significant, since there are so many fewer barriers to AID working well (no meals or exercise interruptions) and that although it’s important in the day, the value will be perceived as less by many early on since meals and exercise and stress will prevent most forms of AID from working as well as they will at night. Experiences will vary, of course.
Factors that would enhance trust in AID: thoroughly tested before market – reduce possible risks; reliability; experiential learning – trust builds over time; prior positive experience with pumps or CGM; positive testimonials from other users.
Factors that would impede trust in AID: Letting go control of diabetes after self-management for decades; prior negative experience with pumps or CGM. We would add early adopters who are used to “running things” and have a hard time ceding control may also be patients who feel less trust.
Relationship impact of AID: Positive – improved sleep for loved ones; decreased stress and improved quality of life for partners; avoid arguments about blood sugar numbers; less worry, think/talk about other things than diabetes; reduce tension between adolescents and parents and between parents. Negative – using system during intimacy.
Financing an AID system: Some (not all) would be willing to pay more if necessary. Youth may expect their parents to make sacrifices to pay for the system. Parents would be willing to pay more than currently. Partners would pay anything to make sure their partner received the technology. Most hoped/expected insurance companies to pay and hoped premiums would not increase. Some (not all), adults hoped health utility of the device (reduced long term complications) would motivate insurers to pay.

Corporate Symposium: Flash Glucose Monitoring – Clinical Use from Childhood through Adulthood

Use of The FreeStyle Libre System in Intensively Managed Adults with Type 2 Diabetes

Gerry Rayman, MD (Ipswich Hospital NHS Trust, UK)

Dr. Gerry Rayman presented case studies from Abbott’s six-month outcome study – REPLACE (n=210 type 2s on MDI, A1c>7.5%) – that will support reimbursement for FreeStyle Libre in type 2 patients. Dr. Rayman did not share specific findings or updates though we found it exciting nonetheless to get a sneak peak some glucose traces from a few patients. [We point out, of course, that these were likely carefully selected examples to demonstrate Libre’s efficacy.] As a reminder, the goal of the study was to show a change in A1c at six months (in patients with a high A1c >7.5%) and Dr. Rayman’s presentation focused on three “success” stories: (i) a patient (baseline A1c: 9.9%) who saw improved glycemic control at six months (A1c: 8.1%) with a non-significant change in hypoglycemia (0.84 hours/day [baseline] vs. 0.77 hours/day [six months]); (ii) a patient whose A1c remained stable (baseline: 8.6% vs. six months: 8.5%) but who saw significant improvement in hypoglycemia (1.65 hours/day [baseline] vs. 0.55 hours [six months]); and (iii) a patient who saw significant improvements in A1c and hypoglycemia – the “ideal” case. We can also imagine successful cases where A1c rose and hypoglycemia fell but nothing like this was shown. Full results from the trial will be presented Friday afternoon (see our preview).

Dr. Rayman shared a striking testament to Libre’s real-world accuracy, presenting SMBG data indicating that all three patients went from multiple fingersticks a day to “virtually none” once they started on Libre. The product label does recommend a confirmatory fingerstick when hypoglycemic, during times of rapid change, and when symptoms do not match reading, though we continue to hear that patients are pretty much eliminating fingersticks entirely. Abbott is presumably pursuing a similar insulin-dosing claim at the FDA though this has proven to be a regulatory conversation that has moved slowly (per 4Q15 commentary). Still, management optimistically expects FreeStyle Libre to launch by the end of the 2016 in US.

First Clinical Experience with the FreeStyle Libre System in Young Adults and CHildren with Type 1 Diabetes

Fiona Campbell, MD (Leeds Teaching Hospitals Trust, UK)

Dr. Fiona Campbell discussed the challenges of successful diabetes management in young adults and children by putting the disappointing status of glycemic control in this population in context. She shared 2013 data from the T1D Exchange, indicating that a majority of pediatric and adolescent patients did not meet ADA or ISPAD guidelines for glycemic control (ADA: <8.5% for those younger than 6 years of age, <8.0% for those 6-13 years of age, and <7.5% for those 13-20 years of age; ISPAD: <7.5% for all ages). Dr. Campbell stressed that the numbers reflect the underutilization of CGM in this population, stating aptly that, “Parents and families LOVE it … but the kids not so much.” She noted, however, that diabetes self-management can be significantly improved when complete glycemic profiles are available, noting that FreeStyle Libre is uniquely positioned to provide such data given: (i) its 14-day wear time; (ii) its factory calibration (no fingersticks); and (iii) the lack of alarms. Within this framework, Dr. Campbell presented case studies from the pediatric accuracy trial (BEAGLE), demonstrating that FreeStyle Libre use can help identify patterns of hypoglycemia, post-prandial excursions, and elevated glucose variability. She did not present specific findings – these will be shown in a Friday oral and perhaps during a press conference Thursday – though her enthusiasm for Libre was telling. As a reminder, Libre is not yet indicated for use in the pediatric population though it sounds like many pediatric patients are already using the system off label in Europe. That said, it will be excellent for Abbott to get the official label as the clinical value of Libre in this younger population seems quite clear.

Clinical Use of the FreeStyle Libre System in Adults with Type 1 Diabetes

Emanuele Bosi, MD (Vita-Salute San Raffaele University, Milan, Italy)

Dr. Emanuele Bosi provided a very straightforward review of the clinical use of the FreeStyle Libre in patients with type 1 diabetes. Through a series of cases, Dr. Bosi documented improvements in glycemic control, risk of hypoglycemia, glucose variability, and quality of life that he has observed over the first ~1.5 years of Libre’s availability in the EU. He did not share specific updates or new data, though the simultaneous improvements in A1c and hypoglycemia in his cases certainly piqued our interest about what data from Abbott’s six-month reimbursement study in type 1s – IMPACT (n=225 type 1s on MDI or pumps, A1c <7.5%) – will look like at ADA 2016.

Exhibit Hall

Abbott

Abbott reps informed us that the company is “coming along” in the process of developing an Apple LibreLink app for scanning FreeStyle Libre sensors without the need for a separate reader. The timeline is not yet clear but we did learn interesting details about what the platform expansion might entail – the company is thinking about building an iPhone adapter to power the NFC-reading capability. As a reminder, the iPhone franchise (up to and including iPhone 5S) has not had NFC built-in and it sounds like this solution would work for both existing iPhones and the upcoming iPhone 6 (that, in fact, does support NFC but – according to the rep – does not have the ability to read data from Libre.) As we understand it, the adapter would be plugged into the headphone jack, would scan data from the sensor, and would then send the data straight to the app. We certainly applaud management for identifying a workaround though it does sound like a less-than-deal solution – requiring an adapter certainly negates the seamlessness and patient convenience of scanning with just a phone alone. We’ve long wondered if Abbott would develop a Bluetooth-enabled FreeStyle Libre sensor, though there are obviously serious manufacturing tradeoffs to doing so. The mobile experience is of course critical when thinking about staying competitive with Dexcom’s G5 and Medtronic’s Guardian Mobile.

Ascensia

Ascensia Diabetes Care made its first public appearance as a stand-alone company though aside from the new branding it does not appear that much has changed. The BGM line is now being marketed under “Contour diabetes solutions” and as one rep put it, “The people are the same. The products are the same. The only thing that’s different is what we’re called.” Reps did allude to a new Bluetooth-connected BGM in the pipeline that is slated to launch later this year; they did not provide specifics though it’s great to see Ascensia is on course R&D-wise.

Dexcom

“Dose with Dexcom” was the theme of Dexcom’s G5-focused booth, marketing the company’s insulin-dosing claim in Europe (first shared at EASD 2015) – that has a very cool ring to it, doesn’t it! . Posters advertised the peace-of-mind remote monitoring advantages of G5 coupled with Share, something neither Abbott nor Medtronic has available in Europe (LibreLink does not have remote monitoring to our knowledge, and MiniMed Connect is only available in the US).

Medtronic

In an exhibit hall surprise, Medtronic publicly displayed Guardian Connect, its standalone, Bluetooth-enabled mobile CGM targeted at patients not on pumps. Signs indicated it is currently under CE Mark review, which presumably means an EU launch could occur sometime later this year or early next year (our speculation, as reps would not comment on launch timing). The transmitter is the familiar clamshell design, though it will send CGM data via Bluetooth directly to a mobile app on Apple iOS at launch. Medtronic does not plan to launch a standalone receiver, so patients will only get the transmitter and app (unlike G5, where they have the option of using a receiver or the app). Interface-wise, Guardian Connect looks similar to the MiniMed Connect app, while retaining the text messaging notifications for caregivers and auto-upload to CareLink. We didn’t get to do a hands-on demo, but the app had the familiar glucose value, trend, and trace on the home screen, plus the obvious sub-menus for calibration and settings (see picture below). We see this as a key competitive answer to Dexcom’s G5 and Abbott’s LibreLink, plus an important Medtronic foray into MDI (Medtronic’s current real-time CGM requires a paired pump). Though we have known about this device since September 2014 (pivotal study completed in August 2015), this is the first time it has ever been on display in final commercial form in an exhibit hall. As of JPM last month, Medtronic expected a US launch of Guardian Connect with Enlite 3 by April 31, 2017 (in FY17); there was no EU timing listed on the slide, which is what made today’s announcement very unexpected. We’re elated to the sensor field moving incredibly rapidly to make systems more convenient to use – who would have thought two years ago that we’d have all three players with smartphone apps to view data?!

Sanofi

Sanofi debuted its first-in-class MyStar DoseCoach in its exhibit hall booth – the device integrates an insulin glargine dose titration algorithm and BGM, providing guidance on insulin dose adjustments for patients with type 2 diabetes. According to sales reps, the device received a CE mark this past December, and this ATTD display is its first exhibit hall promotion. We assume it could conceivably launch soon in Europe; unfortunately, there is no US timing. The MyStar DoseCoach comes with three treatment plans based on “well-accepted titration algorithms” – these are entered into key cards that can be inserted and activated into the device by healthcare professionals (the plans differ on aggressiveness and target range). After selecting the most appropriate treatment plan for the patient, the healthcare professional can then individualize the Dose Helper by entering the patient’s weight, starting dose, and usual dose time. The patient then uses the device like a normal BGM to measure fasting blood glucose levels and to record insulin doses; after the patient runs Dose Helper, the device suggests an updated insulin glargine dose based on the selected treatment plan. Notably, the device is currently only available with Sanofi’s Toujeo (insulin glargine U300) for type 2 diabetes – in our conversations with reps, the company first wants to see how this device will fare on a smaller scale prior to any considerations for expansion. We have expected Sanofi and its MyStar initiative (read more about the initiative from its launch at ATTD 2013) to launch such a device for quite some time and we are very excited to see this new class of devices take flight. We hope this system’s convenience and more individualized insulin guidance will bring more patients to goal and save HCPs time.

We also like that the MyStar DoseCoach prioritizes safety, with multiple built-in verification and human factors rules. For example, prior to a fasting blood glucose measurement, the device confirms with the patient that the blood glucose measurement is indeed a fasting one, even going so far to help define what a “fasting” definition entails. While such notifications can be annoying, accuracy is critical when systems like this are giving insulin dosing advice. The tradeoff is worth it!
MyStar DoseCoach downloads to MyStar Connect, a management software platform that collects all of the patient’s MyStar Dose Coach data (i.e., graphs and logbook) along with other health records in one spot for healthcare professionals. This data unfortunately cannot be synced in real time as there are no cloud connections (it must be manually connected during a doctor’s visit). We wonder what the plans are to integrate with Diasend, Glooko, and Tidepool, as well as how this product could fit within the Verily (Google Life Sciences) partnership.

Senseonics

The implantable CGM company had a full-fledged booth, showing off Eversense: its 90-day implantable sensor (complete with implantation demos!), body-worn transmitter, and mobile app. Eversense is still pending a CE Mark, and the plan is still to launch in the first half of this year in Europe (consistent with our January update on the company when it filed to go public on the NYSE). The sensor implantation procedure took five minutes, and the demo made it seem pretty easy (one incision and then inserting the sensor under the skin with a placement device). While we’ve long characterized the on-body transmitter as a drawback to the implantable system, it was smaller than we expected in person. Still, we were reminded that the transmitter is not waterproof in the first gen version, which adds daily hassle for patients (who must take the transmitter and adhesive off when showering and then reapply them) – to what extent that is a deal breaker for Eversense remains to be seen – how the device is presented will be very important. A brief demo of the Senseonics mobile app revealed the expected home screen layout (glucose value, trend arrow, trace), though Senseonics has chosen to shade the area under the trace – this makes the screen look more crowded in our view without much additional benefit. Separate from the booth, Senseonics announced an agreement today with diasend to enable downloading of the Eversense system, an encouraging sign as it seeks to meaningfully commercialize this product. We hope this unique system can grow the underpenetrated CGM market – assuming pricing is competitive – and are glad to see the company pressing forward in a very competitive continuous glucose sensor market.

Unomedical

In a rare pipeline update, infusion set juggernaut Unomedical disclosed plans to launch a new all-in-one infusion set-insertion device in 2016. The single-touch, fully automatic serter will completely hide the insertion needle and retract it, and the whole unit will automatically detach from the inserted site. It’s an especially great innovation for those afraid of needles, those with poor dexterity, or those who have trouble inserting sites in difficult to reach places. We also like that it is full disposable and self-contained, adding a lot of all-in-one convenience for patients. Presumably it could improve the consistency of infusion site insertion, potentially even reducing site trauma. A poster in the booth also highlighted some earlier stage research at Unomedical: increased infusion set longevity (suppressing the inflammatory response at the insertion site, preventing cannula encrustation) and ensuring healthier skin throughout weartime (skin-friendly adhesive patches). There was no timing attached to these efforts, and a rep confirmed they are in the basic research phase. After years of little innovation in infusion sets, we are elated to see two new meaningful products coming to market this year – BD/Medtronic’s FlowSmart set and now Unomedical’s new automatic inserter! We expect these products to make pumping better for patients already on them, though we wonder about their long-term potential – can better infusion sets significantly enhance the clinical benefits of pumps, pushing more HCPs to prescribe them and more payers to encourage them?

--by Melissa An, Adam Brown, Varun Iyengar, Ava Runge, and Kelly Close

February 3-6, 2016; Milan, Italy; Day #1 Highlights – Draft