ADA Day 2 Highlights – Spotlight on DIY Community and Hypoglycemia
By Jeemin Kwon, Adam Brown, Emily Fitts, Payal Marathe, Brian Levine, and Emma Ryan
Read about combo drug Xultophy, Toujeo vs. Tresiba, diabetes apps, the do-it-yourself community, FreeStyle Libre real-world data, and more!
Day #2 of ADA was another packed day about the newest diabetes drug studies and technology that make a difference in daily life and overall health outcomes. In particular, we noticed a strong focus on hypoglycemia (low blood sugar), suggesting broader recognition of the Beyond A1C Movement. We are excited to see such widespread acknowledgement of the importance of addressing hypoglycemia, along with other glycemic outcome metrics such as time-in-range. Read below for day #2 highlights!
Xultophy is a single injectable drug combining Tresiba (insulin degludec) and Victoza (GLP-1 agonist). While already known to be more beneficial than Lantus in lowering A1C, weight maintenance, and less hypoglycemia (low blood sugar), Xultophy demonstrated similar results when used in type 2 participants already on an SGLT-2 inhibitor (Farxiga, Jardiance, Invokana). On average, the 420 participants saw:
1.9% drop in A1C on Xultophy compared to 1.7% drop on Lantus
58% lower rate of severe hypoglycemia on Xultophy compared to Lantus
No change in body weight on Xultophy compared to 4.4-pound weight gain on Lantus (2% weight gain from baseline)
While there was no meaningful difference overall between the frequency of hypoglycemia while taking Toujeo versus Tresiba, BRIGHT study researchers observed that hypoglycemia (below 70 mg/dl) was 26% less likely on Toujeo during the initial titration period (adjusting insulin dose to meet the needs of each participant). Keep in mind, however, that after the titration period, the two basal insulins perform similarly in this particular study.
The TEDDY study hypothesized that there are “triggers” that cause a child with high-risk genes for type 1 diabetes to actually get type 1, based on the observation that not all children who are at higher risk develop the disease. The trial enrolled nearly 8,000 participants. To the disappointment of the presenters and the audience, the study results concluded that these two factors that are not triggers of type 1 onset: childhood infections and maternal use of vitamin D and omega-3 supplements.
The EDICT study treated people newly diagnosed with type 2 diabetes with metformin, pioglitazone (TZD), and Bydureon (GLP-1 agonist). After six years, participants still in the study who stayed on the triple therapy had an average A1C of 5.8%, while those receiving the conventional treatment had an A1C of 6.7%. Additionally, there was a ten-fold difference in hypoglycemia between the two treatment groups over six years, favoring the combination therapy. It’s important to note, however, that the sample size was smaller at year six (about 100 people). That said, early combination therapy does sound promising, and the audience was eager to know whether other triple combination studies would be done using more advanced diabetes drugs, such as SGLT-2 inhibitors.
diaTribe’s Adam Brown, author of Bright Spots & Landmines, spoke about two key features of useful diabetes apps: 1) radically convenient, frictionless tools and 2) compelling decision support. View the slides here. Here are the apps that Adam believes save time, money, and subtract hassle from people’s lives:
- Web apps like Glooko and Tidepool replace the inconvenience of plugging in devices to download data
- Mobile CGM apps (Dexcom G5/G6, Guardian Connect, FreeStyle LibreLink, Senseonics Eversense) remove the need for a CGM receiver
- Dexcom’s Clarity app offers CGM data insights without having to log in
- Automatic test strip delivery plus remote coaching (mySugr, One Drop, and Onduo)
- Apps that display pump data via Bluetooth are coming soon from Omnipod, Tandem, and Medtronic
Real-time, in-app, automated coaching (WellDoc’s BlueStar)
Glucose prediction and pattern recognition from Sugar.IQ (Medtronic)
OpenAPS and Loop, both products of the DIY community, are automated insulin delivery (AID) systems that are not recognized by the FDA. (For background on AID, click here.) Yet, the number of DIY-product users continues to grow, suggesting that these products are serving an unmet need in the diabetes community. Dana Lewis, founder of OpenAPS, shared some exciting updates:
Diabeloop, a French company, will be incorporating features from the OpenAPS system into its own closed loop product. Diabeloop aims to bring the first version of its AID system to France, the Netherlands, and Sweden in late 2018.
Ms. Lewis introduced the OpenAPS Simulator, which will allow users to simulate the effect of changing behaviors or other parameters on their blood sugar levels. For example, “What would happen to my glucose after a meal if I stopped pre-bolusing?”
OpenAPS Data Commons is a data-sharing platform; DIY users can donate data, and researchers interested in analyzing it can access the database by agreeing to 1) share insights back with the community and 2) publish their findings in an open manner. So far, 121 people have joined the platform.
A new study called OPEN will bring together traditional experts and people with diabetes (who are also experts!) to examine clinical and quality-of-life outcomes for DIY closed loop users; it will also seek ways to further improve DIY AID technology and access. This international study could help legitimize the use of these systems in the eyes of the clinical community, and perhaps even make a case for safety and efficacy to the FDA. Ms. Lewis did not announce when this study will start.
Ms. Lewis also expressed interest in seeing studies validating features such as “eating soon” (a mode that smooths out blood sugar levels by dosing a small amount of insulin 45 – 60 minutes before eating) and “autosensitivity,” (an OpenAPS feature that allows for real-time assessment of insulin sensitivity and adjusts pump settings accordingly).
Real-World Data from FreeStyle Libre Users Suggests Link Between Scanning Frequency and Blood Sugar Variability
Data from over 470,000 users – 4.8 billion glucose readings – showed that A1C ranges from 8.2% in those scanning about 4-6 times a day compared to 6.9% for those scanning 20 times per day. Additionally, the data found that time in hypoglycemia (below 54 mg/dl) drops as scanning frequency increases. As Dr. Irl Hirsch quipped in a separate presentation, “It’s no added cost or pain; it’s all the same.” This data is representative of 72% of FreeStyle Libre users.