ADA’s 80th Scientific Session wrapped up today, and there was a lot of ground covered, despite the fact that we all watched from our chairs at home and daily step-counts were way down from last year. Here are my top five takeaways and themes from this year’s (virtual) conference, with a few bonus round items at the end:
1. Health plans must support broader CGM use… ASAP.
Bergenstal et al. 69-OR: Freestyle Libre System Use Is Associated with Reduction in Inpatient and Outpatient Emergency Acute Diabetes Events and All-Cause Hospitalizations in Patients with Type 2 Diabetes. Diabetes Jun 2020, 69 (Supplement 1) 69-OR; DOI: 10.2337/db20-69-OR
As I wrote for CNBC last year, with the current generation of cheaper, more accurate, more connected CGM devices (Dexcom G6 and Freestyle Libre), CGM is ready to leap from use limited to people with type 1 diabetes to a much broader population. In 2019, we began using CGM in a very large population of people with type 2 diabetes in our UCSF Diabetes Clinic. There is increasingly evidence to back this up. Health plans should start covering CGM for an expanded set of people with type 2 diabetes, including CMS eliminating its requirements for test strip use and multiple daily insulin injections.
Numerous presentations and posters at ADA20202 showed benefits to CGM use in an expanded set of people with type 2 diabetes and also cost-savings. In France, Dr. Roussel et al studied nationwide data and found a drop of 47% in hospitalization rates for DKA in people with T2D after starting Libre. Notably, rates fell for people both who were using 0 test strips per day vs those using >5 strips per day.
Dr. Richard Bergenstal presented data (See Figure 1 on the right) showing a 50% reduction in acute diabetes events- defined as hospitalizations or ER visits for things like hypoglycemia and hyperglycemia- for people with T2D who started Libre. Again, these massive reductions in serious events held true whether people were using many strips or no strips.
Dr. Eden Miller et al showed that the same type of acute diabetes events fell by 30% in people with type 2 diabetes not on bolus insulin, after they started Libre. Dr. Miller et al also showed that in people with T2D, after starting Libre (like the others, this was also in retrospective data) A1c fell by between 0.5 %to 0.8% with actually the higher A1c drops seen in non-insulin using patients. In Spain, Dr. Gomez-Peralta showed significant cost-savings with use of Libre. I could go on, but I won’t. CGM use must be expanded.
Brown S et al. 984-P: First Home Evaluation of the Omnipod Horizon Automated Glucose Control System in Adults with Type 1 Diabetes. Diabetes Jun 2020, 69 (Supplement 1) 984-P; DOI: 10.2337/db20-984-P
2. Closed loop systems are getting incrementally better and better
Initial data was presented on the Omnipod Horizon system, which showed extremely low rates of hypoglycemia, very high time in auto mode, and time-in-range above 70% in a small number of at-home users, in preparation for a pivotal study. Numerous studies were presented on Medtronic’s next-gen 780G system which reached time-in-range above 80%, and similarly higher than 95% time in auto mode. Real-world data from Control IQ’s first 30 days in use similarly showed nearly 78% time-in-range, around 1% hypoglycemia, and 96% time in auto mode. The progress, though incremental, is very real and steady. You know things are improving when a common refrain from experts is that the “major Achilles heel of systems is now the infusion sets.”
Several presenters showed very troubling data about disparities in the use of diabetes technology, including CGM and closed loop therapy. Dr. Jill Weissberg-Benchell from Lurie Childrens in Chicago was a tour-de-force in her talk describing the many levels on which these disparities occur. These include differences in prescribing of technology, of access to technology, and even of inclusion in clinical trials, where she showed that almost no diverse populations have been included in closed loop clinical trials. Several presenters showed data that people from all ethnic and socioeconomic backgrounds benefit from use of these technologies, refuting the previous guideline notion that providers should prescribe to “ideal candidates.” Bottom line: there is no “ideal candidate” for technology. As a physician, I will be looking much more closely at my subconscious bias in who I prescribe diabetes technology to.
Weissberg-Benchell, ADA2020
4. You cannot treat type 2 diabetes without thinking foremost about cardiovascular disease
We have long known that the majority of mortality in type 2 diabetes is related to cardiovascular disease, but we were limited to focusing on “risk-factor” reduction, like controlling blood pressure, lowering lipids, eating healthier, and smoking cessation. With the emergence of anti-hyperglycemic drugs like SGLT2 inhibitors and GLP1 agonists that have proven risk reduction for cardiovascular disease, Endocrinologists and Primary Care providers must be focusing on a much more aggressive, proactive, cardiovascular risk reduction strategy. Outside of “diabetes” medications, there was data presented about other lipid-lowering and atherosclerotic cardiovascular disease risk reducing drugs like PCSK9 inhibitors and icosapent ethyl (REDUCE IT trial). People were starting to use the term “Cardio-Endocrinologists,” and I can see a future where these specialties either partner or combine to treat people with metabolic disease.
4. But, just as for technology, access and disparities are also an issue for the newer (and most effective) type 2 diabetes medication classes, SGLT2 inhibitors and GLP1 agonists
Many presenters showed data on this topic, and as with technology, it is not just one factor. Prescribing rates of these agents to non-white patient populations are lower (Dr. Winn et al with poster 1175-P), there are access issues for under-insured populations, there are cost issues (in the US), and more. One poster (144-LB) showed that in states with Medicaid expansion, use of SGLT2 inhibitors and GLP1 agonists rose faster than in non-expansion states, suggesting that one policy intervention might be broader insurance coverage.
5. Coaching, education, feedback, and support work in treating type 2 diabetes… but, how will business models support them?
Many posters from digital health companies showed a variety of benefits from their digital coaching solutions. Glooko, Onduo, Welkin, and other companies presented data. The common theme is that coaching is an effective intervention in type 2 diabetes management (Disclosure: I am an advisor to Steady Health). No surprise that the tricky part will continue to be finding an effective business model for how to sustain a more extensive, ongoing diabetes coaching model in the fee-for-service environment.
Bonus Round of Things That Surprised and Delighted Me:
Dr. Kelilah Wolcowicz presented data on a point-of-care insulin immunsensor. While this was tested with the intention of optimizing closed loop therapy in type 1 diabetes, there is a potentially huge market opportunity in POC insulin testing in metabolic syndrome.
A group from Stanford presented data showing their use of machine learning on EHR and CGM data trying to optimize which patients should get phone calls in a population health approach.
Dr. Diana Isaacs from Cleveland Clinic shared her experience in setting up shared medical appointments to leverage CGM for a small group of patients.
Laurel Messer, RN, MPH, CCRP, CDE, from the Barbara Davis Center shared some excellent resources for training and educating people with type 1 on the use of closed-loop insulin delivery systems.
What did you find most interesting at ADA this year? What important themes did I leave out?