June 18, 2021; Virtual; Full Report – Draft

Hello from New York, Detroit, and San Francisco! Our tech team was spread out, but virtually got together at this year’s DiabetesMine D-Data Exchange summer gathering. See our top highlights from the meeting below.

1. Quotable Quotes from Smart Insulin Pen KOLs Sean Saint (Medtronic), Søren Østergaard (Novo Nordisk) and Challis Imes (Eli Lilly) on Integrated Systems and the Future of Smart Pens

Smart pens were the focus of a panel discussion between Medtronic/Companion Medical’s Sean Saint, Novo Nordisk’s Søren Østergaard, and Eli Lilly’s Challis Imes. DiabetesMine’s Amy Tenderich moderated the panel asking questions about the future of smart pen technology and diving into the capabilities of current connected smart pen systems. From our perspective, the highlight of the session came when Mr. Saint discussed the rationale behind developing smart pen technology recognizing that historically, patients and providers have turned to pump therapy when they want connected solutions. However, with the advent of smart pen technology, patients have more options to choose from to personalize diabetes management in a way that best fits their lifestyle and needs. We heard this message reiterated by Ms. Imes who highlighted the personalization possible with Lilly’s Tempo Pen and Tempo Smart Button, while we heard from Mr. Østergaard about the potential for connected smart pen systems to improve outcomes and reduce overall health system expenditures. Read on for a collection of quotable quotes from panelists in this exciting discussion:

• “Pumps are a fantastic technology, but the simple fact is that not everyone is using them, and from Companion’s perspective at the beginning, we were leaving a huge percentage of people without better technologies and telling them they needed to switch to pumps instead of offering them a technology they want to use. If people want to use pumps great, but if there are people who don’t want pumps, we wanted to make a solution to help them.” – Mr. Saint

• “What’s interesting about smart pens is that they can do a lot of great things like improving Time in Range, dose calculators that have been shown to help improve A1c – all these things that smart pens can do, the individual pieces have been shown to improve outcomes, so now seeing the additive effect will be really exciting.” – Mr. Saint

• “We’re very committed to integrating smart pen technology into many things that haven’t even been thought of yet. We’re also working on building into integrated diabetes management solutions and I think we will see more of this from all the players here – we’ll have to work together on some of these solutions even if we remain competitors on others. I envision a future where someone can be on both a Lilly and a Novo Nordisk insulin and get a full picture of their situation with integrated technology.” – Mr. Østergaard

• “We want solutions to work within people’s lives and the workflow of providers…For us the biggest part of connected technology are the insights derived from data and what we can get from smart pens is really only the first step.” – Ms. Imes

• “In terms of integration, CGM is a no-brainer here. I like to say that there are three legs to the diabetes stool: insulin, which smart pens cover, glucose, which can come from CGM, and carbohydrates and we need to bring all three together in connected solutions.” – Mr. Saint

• “We need to have reliable data on patient behavior, which smart pens can help provide, so providers and patients can have the right conversations about diabetes management and treatment.” – Mr. Østergaard

As a reminder, Companion Medical developed and commercialized InPen and was acquired by Medtronic last year. Novo Nordisk recently launched its NovoPen 6 and NovoPen Echo Plus smart pens in Sweden and Denmark, and Lilly recently announced new data integration agreements for its smart pen portfolio which includes the Tempo Smart Pen – an “enhanced” and “modified” version of Lilly’s existing prefilled and disposable Kwik Pen – and Tempo Button insulin pen attachment. Currently InPen is the only connected durable insulin pen available in the US and the system includes real-time data integration with Medtronic’s Guardian 3 CGM. Bigfoot plans to launch its recently cleared Unity system in the summer. InPen also recently received an updated CE-mark that included data integration with Medtronic’s recently approved Guardian 4 CGM and Medtronic aims to launch InPen in Europe this fall. Novo Nordisk’s smart pen line-up has started to roll-out in Europe and is available in Sweden and Denmark where both NovoPen 6 and NovoPen Echo Plus have been integrated with Roche’s mySugr logbook app. Lilly plans to submit its Tempo Smart Button for CE-marking “later in 2021” with launch following approval.

2. CNET’s Brian Cooley Takes a Look at Diabetes and Healthcare Technology from a Consumer Angle: Great Technology Should be Transparent, Intuitive, Intimate, and Constant

To kick things off at this year’s DiabetesMine, we heard a wide-ranging presentation from Brian Cooley (CNET). As a reporter at a consumer technology outlet, Mr. Cooley’s presentation was an interesting change of pace from the normally very med-tech (or even diabetes tech) focused speakers we see at DiabetesMine and other diabetes gatherings. While Mr. Cooley’s talk was very broad, touching on everything from AI to remote monitoring to 5G, we were particularly struck by his description of four traits that make great technology. In our view, these four things are applicable to both consumer technology and diabetes technology and should be in the minds of anyone developing diabetes technologies.

• Great technology is transparent. In Mr. Cooley’s parlance, when devices and software are at their best, they just “go away.” One obvious example of this is in the upgrade from fingersticks to CGM. For fingersticks, people with diabetes have to carry around a lancing device, fingersticks, and glucometer when they want to check their blood glucose. Performing a fingerstick also requires lancing the finger, a sometimes painful and highly visible activity. Of course, CGM devices still require patients to insert their CGM every 7-14 days. CGMs can also malfunction or require fingerstick calibrations. However, when at their best, these CGMs can also almost completely take away the burden of checking glucose. Looking towards the future, in some ways, smaller (or even implanted) devices and longer-wear can all be seen as efforts to make CGM even more transparent for people with diabetes.

• Great technology is intuitive. Mr. Cooley elaborated on this idea by noting that not only how to use a technology should be intuitive, but also why to use a technology should also be intuitive. For this feature, we immediately thought about Time in Range. As CGM adoption continues to grow, we’ve heard from more providers that Time in Range is easier to use with their patients than A1c. In our view, that’s because Time in Range is more intuitive than A1c. For a person with diabetes, knowing Time in Range is 70% is a lot more personally meaningful than knowing their A1c is 7%. That 70% Time in Range means they are spending ~17 hours per day in a normal glucose range; it means 70% of the day is spent without major concerns about hypo- or hyperglycemia and needing to adjust insulin doses. 

• Great technology is intimate. Intimate is Mr. Cooley’s word for “personalized” – in other words, great technology should know the user better than they do. Our case example for this from diabetes is automated insulin delivery systems. In the past, it was impossible to have a clinician or patient observe glucose data more than a few times a day and adjust insulin delivery. Now, with the advent of CGM and hybrid closed loop algorithms, technology can look at glucose data every five minutes and adjust insulin delivery accordingly. Looking towards the future, as we think about AID, new algorithms and sensors (e.g., exercise sensors, ketone sensors, etc.) can all be seen as an effort to help technology become more intimate with their user and learn their behavior better.

• Great technology is constant. Again, borrowing from Mr. Cooley, great technology should be present in “all of a user’s places.” There are a few ways to take this, but our case study for this feature is the availability of glucose data in all sorts of ways. For Dexcom CGM users, glucose is not only viewable on the smartphone or sensor reader, but also on a user’s wrist, if they have an Apple Watch. Dexcom also has Siri integration which allows users to get their glucose data by voice. Beyond the patient themself, most CGMs now also offer remote monitoring features for parents or partners or other caretakers to view glucose data remotely. For an example of how used to this technology’s constancy many have become, see our report on Dexcom Follow’s outage from back in November 2019.

3. Real-World Open-Source AID System Data: Time in Range +4.2 Hours/Day to 80% with DIY System; A1c Drops 0.9% to 6.2%; No Significant Differences in Users vs. Nonusers’ Ethnicities, Incomes, Occupational Status or Education

Out of Germany, Dr. Katarina Braune (Charité University Medicine Berlin) read out real-world data on open-source AID system use collected by the OPEN Diabetes Project, a patient-led, interdisciplinary, international collaborate that studies open-source and DIY AID systems. The collaborative has studied patients’ decision-making around DIY systems, clinical outcomes, lived experiences, and barriers to scale. One real-world study that Dr. Braune highlighted shows that participants’ (n=365) average A1c dropped by 0.9% to 6.2% and that Time in Range increased by +4.2 hours/day to 80% after initiating open-source AID systems, including Loop, OpenAPS, and AndroidAPS. This was true both among adults and among children. Notably, after initiating an open-source system, the range of participants’ A1c and Time in Range values shrunk, meaning that a greater proportion of participants achieved glycemic control. The study also showed no significant A1c differences based on age or gender, which is not true of commercially available systems.

• The OPEN Diabetes Project has conducted recent research on patients with diabetes’ reasons for using and for not using DIY AID systems. Based on data from 897 people with diabetes (722 adults, 175 adolescents and children), improvements in sleep and gaining back time are major benefits to the system. Another study under investigation found that the primary reasons why people with diabetes (n=648) reported not using open-source AID systems were: (i) challenges sourcing components and the associated costs; (ii) perceived limited support from HCPs; and (iii) low self-perceived levels of tech literacy. Perhaps unexpectedly, trust in medical technology, in open-source diabetes technology, and in unregulated medical devices were not commonly reported barriers to use. The study also compared the demographic features of non-users (n=129) to DIY system users (n=519) and found no “prominent” differences in ethnicity, income, occupational status, or education, which is notable given that access and privilege are often assumed to be associated with AID system use. There were two prominent demographic differences between users and non-users: the user cohort was more male (59% vs. 38%) and were significantly more likely to work in computer science/electronics or IT than non-users, signaling the potential importance of tech-savviness. An equal portion of users and nonusers worked in healthcare. We’ll be back with more interesting data from this presentation. Neither study has been published yet, but we look forward to seeing the full dataset and more details on participants’ decision-making.

• Dr. Braune also summarized the key points of an International Consensus on Open-Source AID that is currently under peer review with The Lancet: Diabetes & Endocrinology. The statement was written by 44 HCPs and four legal experts and is endorsed by ISPAD, IDF, ADCES, Diabetes Technology Network UK, and VDBD. The consensus has eight “key statements”: (i) DIY AID systems are a safe and effective treatment option; (ii) supporting the right of people with diabetes or caregivers to make informed decisions about their own care is essential; (iii) HCPs should support patients in using DIY systems or refer them to HCPs who can; (iv) help patients set realistic expectations; (v) patients own and should have access to their data and AID system operations should be transparent; (vi) patients should be informed of the benefits and limitations of DIY systems; (vii) do not violate any local law or organizational governance; and (viii) real-world experience should be considered by device regulators.

4. Product Demos: Bigfoot Biomedical, Waveform, Levels, Tidepool Loop, DiaMonTech, and Nudge BG

Bigfoot Biomedical

• Bigfoot CMO Dr. Jim Malone focused his D-Data product demo on the company’s recently cleared Bigfoot Unity System. Dr. Malone walked attendees through the system details including the two included smart caps compatible with disposable insulin pens (black for basal and white for bolus). Additionally, Dr. Malone outlined the ability of the white pen cap to act as a FreeStyle Libre scanner directly integrating CGM data into insulin dose recommendations. Bigfoot Unity received FDA approval in May for use in people with either type 1 or type 2 diabetes on MDI down to 12 years old. During Q&A, Dr. Malone explained that the lack of a pediatric indication for Bigfoot is because the company’s smart caps are currently only compatible with whole unit insulin pens, but that Bigfoot is working to develop a system that is compatible with half-unit pens like what many children with diabetes use. Dr. Malone also showed a 1-minute video outlining Bigfoot’s provider-facing “Bigfoot Unity Hub” that enables remote training for the Bigfoot Unity system and also allows providers to interact with patient data and make dose recommendations and adjustments.

Waveform

• Rick Valencia, Waveform CEO, provided an overview of the company’s needle free and minimally-invasive Waveform CGM. The Waveform CGM, marketed in Europe under the name GlucoMen Day, can be worn for 15-days, offers customizable alerts (low/high threshold, rapidly changing glucose, 15-minute high/low predictor) and sends minute-by-minute data directly to a connected app. One of the current draw backs of the system is that it requires 1 calibration/day. However, according to Mr. Valencia, Waveform is currently working to reduce the number of necessary calibrations down to once every 2-3 days and aims to have the next generation system factory calibrated. In exciting news, Mr. Valencia shared that Waveform is currently working to bring its CGM to the US and has already begun a pre-pivotal trial with another expected to start at the end of the summer and FDA submission expected in 2022 with approval to follow in 2023. Waveform is also working to extend its sensor longevity to 21 days, which along with its needle-free insertion would certainly help differentiate Waveform from other CGMs currently on the market. According to Mr. Valencia, Waveform still utilizes a small wire under the skin to measure interstitial glucose levels, but this sensor is inserted directly without the use of a needle resulting in less pain at the insertion site and enabling longer sensor wear time.

Levels

• Levels co-founder and CMO Dr. Casey Means discussed the company’s app-based wellness program that uses CGM data to provide metabolic insights for users. As Dr. Means described it, Levels is a “software overlay to contextualize and interpret CGM data stream to facilitate improved diet and lifestyle” and is not meant to be used to treat and/or diagnose medical conditions. Levels provides users with insights into how foods and behaviors impact glycemic levels allowing users to compare events and set challenges for themselves to improve glycemic control. Additionally, Levels allows for food logging via photos, can alert users to changes in glycemic levels, and tracks sleep and activity and their impact on blood glucose levels. Through these insights, Levels is working to provide, “not a judgement on health, but an educational opportunity to learn” – a valuable approach in theory, although all the devices are clearly being used off-label. Though we know this is done at various points in diabetes, a business built on off-label use doesn’t rest easy. Dr. Means said that Levels is “device agnostic” and can integrate with multiple CGMs, but the majority of current users are on the FreeStyle Libre system. Users access CGM via Level’s telehealth network of providers which boasts providers in every state who can then provide prescriptions to eligible patients. Currently, Levels is still in a beta testing phase, but plans to have a full launch in the third quarter of 2021. We first learned about Levels in November 2020 when the company closed a $12 seed funding round and it is one of a growing number of companies working to expand the CGM wellness market with others, including Supersapiens and Nemaura which are targeting athlete and prediabetes populations, respectively.

Tidepool Loop

• Mr. Howard Look, founder and CEO of Tidepool Loop, updated us on the company’s early 2021 FDA submission seeking 510(k) clearance for its interoperable automated glycemic controller. As a reminder, Mr. Look announced at DiabetesMine #DData Exchange 2018 that Insulet’s Omnipod would be the first official pump partner for Tidepool Loop, and at the same conference in 2019 revealed that Tidepool had also partnered with Medtronic and Dexcom. While Mr. Look was restricted in what he could share due to the terms of FDA premarket approval, he assured attendees that the FDA has been “fantastic to work with” and that the review process thus far has been “thorough and totally reasonable.” Mr. Look reiterated the company’s commitment to transparency throughout regulatory review, pointing listeners to a repository of documents containing all interactions with the FDA and sharing that Tidepool will make all approval documents available to the public as soon as possible. While he was unable to say more at this point, Mr. Look hopes to share updates in the Fall – and we’ll surely be on the lookout. From our perspective, approval could come any day at this point as the FDA continues to work diligently and through both COVID-19 and other regulatory submissions.

• On that note, there are a range of tools that people with diabetes in the US are waiting for, including, but not limited to Tandem’s app for the Control IQ that will enable insulin bolusing from the app (no active timelines as of 1Q21), closed loop capabilities for Abbott’s Libre 2 (no word on this or the Libre 2 app since FDA approval in 2Q20). As well, people with diabetes are certainly looking ahead to US regulatory approval of Insulet’s Horizon closed loop system (expected later this summer but could be later), Medtronic’s 780G (submitted in February 2021 and under active review with no stated timelines), etc. Additionally, while Medtronic’s next-generation CGM, Synergy, is slated for FDA submission by October 30 of this year, this submission, like others, even when submitted on time may experience later approval than possible given all the resources FDA has put in and around the COVID-19 pandemic. Medtronic had originally slated Synergy’s FDA submission for summer 2021 but it would certainly be understandable to see this later as meetings with FDA must be challenging to schedule. Finally, as of April, the FDA had assigned a reviewer to Senseonics' 180-day Eversense CGM with a review time of ~six months expected.

DiaMonTech

• CEO Thorsten Lubinski presented DiaMonTech’s non-invasive glucose monitoring technology which measures glucose via photothermal detection using a quantum cascade laser (QCL). While the company has currently only released D-Base, a shoebox-sized, multi-user monitoring device, the company hopes to make the technology more compact with: (i) D-Pocket, a pocket-sized device for end consumers (coming 2022); and (ii) D-Sensor, a wristband CGM equipped with data logging that syncs with iOS and Android devices (coming 2024-25). D-Base received CE-mark approval in 2019 and DiamonTech has submitted a pre-submission application for the device to the FDA as of September 2020. While we are excited to see another wearable glucose monitor potentially enter the marketplace in the future, there are a number of groups working to develop similar technology (e.g. LifeLeaf’s LifePlus, Movano) and clinical testing of D-Base revealed a MARD of 11.3% which is significantly higher than other BGMs and CGMs currently on the market and well-above the accuracy needed to qualify for an iCGM designation or run an AID system. During Q&A, Mr. Lubinski addressed some common concerns with wrist-worn glucose technology discussing potential interference from swear, hair, or dirt expressing that the D-Sensor will be designed to warn users when external conditions compromise the accuracy of a reading so that the user can obtain the best possible measurement. Additionally, Mr. Lubinski addressed concerns about skin color impacting accuracy explaining that skin pigments lie below the 100-micron depth of the QCL. Of note, Q&A moderator Ms. Amy Tenderich made an important clarification that while DiaMonTech’s end-product with a wrist-worn sensor will enable continuous glucose monitoring, the current D-Base and upcoming D-Pocket devices require the user to place their finger on a sensor meaning the data collected more closely resembles BGM data instead of that from a CGM.

Nudge BG

• Diabetes champion Lane Desborough presented the final D-Demo of the conference providing an overview of his new company Nudge BG that we first learned about in October 2020. Mr. Desborough has an extensive history in the diabetes technology world working at Medtronic and helping to found Bigfoot Biomedical and his most recent efforts have been directed toward developing an easier to use automated insulin delivery algorithm. Mr. Desborough has depicted Nudge as a fully closed loop system that requires minimal user input allowing patients to choose their level of engagement. Specifically, Nudge does not require users to input carbs or pre-bolus for meals which Mr. Desborough has recognized as major challenges in his son’s experience with diabetes, especially as he moved into young adulthood inspiring Mr. Desborough to develop the Nudge system targeted toward patients who want more freedom from constant interactions with their AID system. The Nudge algorithm is designed to integrate CGM and pump data dosing insulin every five minutes based on CGM readings. To date, Nudge has been tested in silico due to costs and risks associated with human clinical trials, but Mr. Desborough is hopeful that Nudge can move into clinical testing in the near future. During Q&A, Mr. Desborough also emphasized that Nudge BG as a company is an “engineering factory” for AID systems allowing for user innovation on top of developed systems.

--by Armaan Nallicheri, Hanna Gutow, Katie Mahoney, Albert Cai, and Kelly Close