Activity Transcript

Richard M. Bergenstal, MD: Welcome, everyone. Thank you for joining us today. Today, we're going to talk about the impact of new innovative technology on optimizing the management of type 1 diabetes. I'm Rich Bergenstal, Executive Director at the International Diabetes Center, and I'd like to welcome you to this program titled “New Automated Insulin Delivery Recommendations: Coming in Hot!” Joining me today to discuss this important topic are Dr Moshe Phillip, who is director of the Institute of Endocrinology and Diabetes at Schneider Children's Medical Center of Israel. Thank you for joining us, Moshe.

Moshe Phillip, MD: Thank you, Rich, and I'm extremely happy to be here with you today.

Dr Bergenstal: We also have Dr Revital Nimri, who is Director of the Scientific Technology Diabetes Service, also at Schneider Children's Medical Center of Israel. Welcome, Revital.

Revital Nimri, MD: Thank you, Rich. I'm super excited to be here and talk about the big thing in diabetes care.

Dr Bergenstal: Great. It's going to be great. And Dr Tadej Battelino, who's Professor of Pediatrics at Ljubljana University in Slovenia. Pleasure to have you with us today, Tadej.

Tadej Battelino, MD, PhD: Thank you very, very much. I am really happy to be with you all and to share together what we designed.

Dr Bergenstal: It's going to be great. And there are just times in history of a given disease, like type 1 diabetes, where it seems appropriate to just pause for a minute and review the latest advances in therapy. We paused all together in 2021 to celebrate the 100th anniversary of the discovery of insulin, and automated insulin delivery (AID) systems may be the biggest advance since the discovery of insulin with the potential to improve the lives of the 1.5 million people with type 1 diabetes in the US and the over 9 million with type 1 diabetes, globally.

The landmark clinical trial DCCT ended 30 years ago and showed a remarkable benefit of good glucose control, as you know, in HbA1c near 7%, yet just 4 years ago, few with type 1 diabetes in the US or globally were reaching this goal. We've had major advances in glucose monitoring, culminating with CGM and, most recently, advances in insulin delivery with better and better insulin pumps. And in the last 5 years, technology systems called automated insulin delivery or hybrid closed loop or often referred to in the press as ‘artificial pancreas’ has been commercialized worldwide. They combine CGM, an insulin pump, and an algorithm that automatically adjusts the insulin delivery based on the glucose readings from the CGM.

ATTD felt it was time to convene experts from around the globe and make recommendations on the effective use of AID systems, and the report of this consensus group was published in Endocrine Reviews. We're here today to review with you the collective wisdom of this group, and I'm extremely pleased that 3 of the key authors of this consensus have joined me today: Drs. Nimri, Phillip, and Battelino, who you'll hear from just momentarily. I wanted to have a special nod to Dr. Aaron Kowalski, who in 2009 laid out a seminal roadmap for the artificial pancreas and revised this roadmap slightly in 2015 to reflect the progress that was being made by the research community. So, thank you, Aaron, for your contributions.

This summary report of the AID field includes what an AID system is, who should use an AID system, how to teach patients and clinicians about the most effective use of AID systems, and, finally, where the field of AID therapy is headed. I couldn't think of a better group of clinicians, researchers, and friends to join me than Revital, Moshe, and Tadej to help sort out the what, who, how, and where of AID systems. Let's dive into this consensus report now with my colleagues. Moshe, I'll turn it over to you.

Dr Phillip: Thank you, Rich. Actually, the concept of automated insulin delivery system, what we call the AID system, which is based on continuous glucose sensing, a sophisticated algorithm, and an insulin pump was already published in the medical literature in the '60s, but it took several decades until real-time CGM technology was approved for clinical use and enabled the development of control algorithm that continuously adjusts insulin delivery in response to real-time sense of glucose levels.

Today, years later, AIDs are here and became an integrated path of the management of people with type 1 diabetes. Clinical evidence from randomized controlled trials of the efficacy and safety of AID systems has grown over the last 6 years with the introduction of commercially available AID systems. AID use increased time in range by 9% to 16% while the HbA1c levels decreased by 0.3% to 0.5% with either no change or even a reduction in time spent in hypoglycemia.

Newly released real-world data of about 13,000 users of 780G, the advanced hybrid closed loop system, showed that almost all participants achieved the recommended glycemic target of time in range between 70 to 180 mg/dL. In this cohort, time in range increased by 11% after initiation of 780G with reduction in both time above and below range.

Another study recently published the real-world data from the Control-IQ system, where almost 20,000 people with type 1 diabetes of all ages also showed an increase in time in range of almost 13% with no clinical difference in hypoglycemia. 

Omnipod 5 system also increased time in range significantly in children, adolescents, and adults with type 1 diabetes. 

Another 6-month study showed nicely that the use of AID improved time in range for any baseline HbA1c level. However, the greatest improvement was indeed seen in people with high initial HbA1c level. 

Based on the data we already have today, the consensus panel advocated that all people with type 1 diabetes should be considered for AID use, especially those experiencing suboptimal glycemic control and/or problematic hypoglycemia.

Lifestyle and quality of life should also be discussed when considering the use of AID. The panel acknowledged that there are racial and social inequality in access to technology and recommended to ensure fair access to AID systems. Solid evidence exists of the benefit of AID systems for school-aged children, adolescents, and adult with type 1 diabetes. Special populations with insufficient evidence are the preschool children and adults >65 years old.

However, recently some new evidence of safety and efficacy of AID use in these age groups has been accumulating. Data supporting AID use in populations with diabetes other than type 1 are limited and should further be studied. The experts participating in the consensus document felt that, in people with type 2 diabetes on intensive insulin therapy, they also could benefit from the use of AID systems, even though robust data is still unavailable. Rich, back to you.

Dr Bergenstal: Moshe, thank you very much. And now I turn to Revital. How do you actually use the AID systems effectively? What did our consensus group think? Walk us through that, if you will.

Dr Nimri: Thank you so much. So now, we heard what kind of devices there are and who can benefit the most, and it is time to get down to clinical practice. Let's talk about the how: how should we translate this technology into our routine clinic, how to specifically adjust education, training, and support for each AID system, and how to use the data to maximize its clinical benefits for AID success.

Let's talk a bit about preparation before initiating AID. The healthcare team should be equipped with tools to provide competent clinical assessments of AID use. The first thing is to set realistic expectations for the team of what AID system can and cannot achieve. Then the team should become familiar with each system’s specific features and can use the very helpful CARES structure framework on how each system calculates insulin delivery, which parameters can be adjusted, when should user revert to an open loop, what are the most critical education points and key aspects of sensor use. Specifically, the team should learn how to assess glycemic information and understand CGM data, how to explain to patients on time in range, time below range, time above range, and desired targets.

A vital consideration in starting an AID system is for the team to proactively engage with the person with diabetes to set realistic expectations for what the AID system requires from the user to optimize outcome. Acknowledge that there are some burdens, but the benefits outweighed the hassles or burdens. A comprehensive pre-AID education should be provided. Switching to an AID is not only a technical matter and learning how to use it cannot be separated from the overall management of diabetes. We are actually only adding the tip of the pyramid of education. The rest of the pyramid is the core diabetes knowledge and management, then the CGM, pump, and on the tip, the AID.

So where to do the training, either in clinic, virtual, or both. Training should be tailored to the individual based on current therapy and personal family skills. Special attention should be given to populations with longstanding diabetes and/or suboptimal control for potential transient worsening of retinopathy and microvascular complications. Early follow-up is needed with periodic adjustment of pump parameters. Detailed information in how to initiate AID is provided in this valuable table from the consensus, I encourage you to read it, including preparation, data needed to start each AID system, available initial setting, and also the factors that can be modified by the user and healthcare provider to optimize AID for each system.

Now, let's look at some highlights from the clinical recommendation with some case examples. All systems typically perform best and are safest when they are used according to the manufacturer-specific instructions. Encourage the user to trust the system. Attempts to trick the system by entering fictitious carbs, for example, should be avoided, as it can lead to an increase in glucose liability and decreased system performance. Here is a case of a user who used manual suspension when the system had already suspended delivery, causing reactive hypoglycemia, which could have been avoided.

In all currently available AIDs, the user is still required to give a pre-meal bolus. So, carb ratios are important and should be evaluated routinely. Many of these systems will benefit from a more aggressive ratio, say, 10% to 20% more insulin to carbs ratio as compared to open loop setting. This means, numerically, lower carbs ratio. Actually, the ability of the AID to modulate the basal insulin rate allows the user to give more insulin for meals. You can see in this case how nicely this well-controlled boy got even better shortly after starting the AID as the carb ratio was reduced. That means that more insulin was given to cover the meal with no subsequent increase in hypoglycemia. The change in carb ratio before and after AID initiation is presented in the red box.

The timing of pre-meal bolus is super, super important in the context of AID. And this is, in my opinion, the most important take home message. AID will automatically increase insulin delivery after initial rise in glucose level, thus, a bolus delivered either during or after the meal could lead to insulin stacking and hypoglycemia. See the following case: a meal was consumed, but the user bolus was given half an hour later, while the system already increased insulin as a response to the rise in glucose, and that caused hypoglycemia. The system-initiated bolus is presented in the red dotted box. Therefore, it is usually recommended to give a meal bolus 10 to 20 minutes in advance. In a situation where a meal bolus is missed or delayed, consider reducing the meal bolus or even skipping the bolus altogether.

Physical activity quickly increases insulin sensitivity, and several actions can be taken to reduce the amount of insulin during exercise. A higher glucose target should be set up to 1 to 2 hours in advance and even maintained post-exercise. Also, if prolonged exercise is anticipated, the pre-meal bolus dose can be reduced. One may also disconnect from the pump; however, it is very important to remember to stop delivery or suspend. What about the ingestion of carbs before exercise during AID? Advise against consuming carbs 15 to 60 minutes before exercise, as the system will respond by increasing insulin delivery and this may increase the risk of hypoglycemia. Carbs can be consumed if glucose is below 120 mg/dL or doing prolonged physical activity.

Another important point to consider in treating hypoglycemia during AID use, the traditional amount of carbs should be reduced as the system will reduce or stop insulin delivery upfront, even prior to overt hypoglycemia. This will help to reduce glucose fluctuation, as in the following case: the system stopped insulin delivery an hour and a half for predicted hypoglycemia. The user also consumed carbs and the glucose jumps to over 400 mg/dL. In case of hyperglycemia, be aware of DKA and don't forget the basics of treatment. Infusion set failure can occur in the same way it happens with insulin pump alone, and the same goes for sick days. Consider stopping AID and moving to open-loop sick day rule, especially in the presence of ketones.

And finally, a word on quality of life. AIDs improve glycemic control, but our patient’s quality of life is just as important. This is a case of an 8-year-old girl that was treated with sensor-augmented pump therapy. The parents kept her glucose level high at night to avoid nocturnal hypoglycemia. See the top panel. After switching to AID, nocturnal hypoglycemia was significantly reduced with stable glycemia during the night. And I got a happy text message from the parents of 100% time in range during the night, having a good night's rest.

Dr Bergenstal: Well, thank you so much, Revital. It's so amazing, your real-world examples, your pearls. Trust the system, get the pre-meal boluses, figure out your exercise, how many carbs to take. Anyway, thank you so much for this real-world evidence and data about how to actually use the systems. I think our listeners are going to gain a lot by digging into that section of the consensus report.

I just think Revital showed us so nicely the importance of really getting into those reports and digging into them and going through them with the patients. And it's interesting, though, when you look at these reports, they're all different, from all the different manufacturers. And the reports are all good, they're just all different. And should we become experts at each one of these to take care of our patients? Well, the consensus group felt it really would be important to strive for, at least, a standardized report, something we could start with, much like we do with our CGM data today.

So, the group worked to generate a consensus AID report as a starting point. I'm not going to go through it in detail today because it's in the nice document report, but I'll just say it's a two-page report and it starts off with the top panel that is the glucose and insulin metrics. The glucose metrics you're very familiar with and look a lot like the CGM metrics. The insulin metrics give you those key parameters that you need, the total dose, how much is with bolus, how much is a basal, what's automated, what's user-delivered. Then the middle panel gives you that AGP profile many people are used to looking at, and right below it are the key settings, the insulin, the carb ratio, the corrections factor, the set point, and they're lined up with the profile so you can get a sense what you might need to change.

And then the bottom panel on the first page is the bolus insulin assessment. I really like this component because all the meals are lined up according to when the bolus was given and the insulin to carb ratio sits right there in the picture. So you can say, "Whoa, this meal really had a spike" and "What's my insulin to carb ratio? It probably needs to be tightened," as opposed to maybe another meal where things seem to work. So that's the first page. The second page and third page, if one desires, are the daily profiles, and Revital showed us so nicely how the daily profiles are what we really need to dig into to make a lot of the setting changes that she so nicely reviewed with us.

So that's the standard report. And what I'm asking now is for everyone listening to continue to work together along with our industry colleagues to improve and refine this standard report even further. And then we should all work to figure out a way to get it integrated into the electronic record so it just shows up when we're seeing our patients. So that's our current state and our wish for a standard report. Now, I'll turn to my friend, Tadej, to tell us where he thinks the field is going in the future with AID.

Dr Battelino: Thank you very much, Rich, for this. And I so much agree that we need one standard report, perhaps even more important for people with diabetes than for us, so that we can communicate with them in a, if you want, universal way, wherever we practice our diabetes management. But then we also need to dream a little bit, and this consensus group allowed a little bit of dreaming. The dream concept was very important during the development of the artificial insulin delivery.

I think the first thing that we need to dream and turn into reality is broader adoption, particularly equitable adoption. So, we need more affordability. We should actually reduce the current disparities that are so obvious, the differences from more affluent regions to less affluent regions, because people with diabetes are the same wherever they live with this disease. So equitable adoption and affordability is an obvious priority that we all have to work on.

And then, of course, when we turn to the technical part, we all want smaller devices, perhaps even nicer or more lovely devices, more trendy, if you want, as everything that we carry on us, and then perhaps smarter algorithms that would learn from us and be personalized so that the person would actually have a personalized automatic insulin delivery system that could be much more helpful and perhaps less burdensome for the diabetes management.

And then, on the other side, we can imagine faster insulins. There are some faster insulins on the market, but they are more in the pipeline. And perhaps we will get an insulin so fast that we could mimic what we believe is physiological or come very close to it. And then co-formulations with other agents that could actually increase the absorption from wherever we inject our insulin could help bring us closer to what we then would refer to a full closed loop, so a closed loop that would work completely independently of whatever we do.

Another option, of course, is a dual hormone option that used to be quite popular, but I do believe now is less and less of an issue with the really smart algorithms that we have, excellent sensors, and perhaps also faster insulins. I would like to acknowledge the DIY community, the do-it-yourself community, that contributed considerably to the development and also to equitability, to the affordability of the automated insulin delivery. There is a beautiful consensus statement they published and I really thank this group and the JDRF for supporting this group, actually, to bring this option of automated insulin delivery also to people that cannot afford, for different reasons, what is available commercially in more affluent societies.

And finally, as Rich mentioned so nicely, we do need a system that would go automatically in our electronic medical health records that would be unified, that would be readily available and also, not just to our electronic health record, but also to a personal webpage that a person with diabetes may want to have so that we could communicate we did exactly the same way of data presentation and analysis. We can, of course, dream also of implantable sensors and implantable devices so nothing would have been on our bodies. And this is a beautiful dream and I do believe that perhaps, at some point, we can even imagine such a device. And with all these dreams, Rich, back to you.

Dr Bergenstal: Thank you, Tadej. I love to hear you dream about the future and reflect some of the dreams we heard in our expert consensus group. That was really a great gathering and synthesis of the data into this report. Well, we're coming close to a close, but I just have this opportunity with three of the great minds in the field joining me today to ask you, just in closing, if you'll just spend a minute each and just say what's one personal takeaway from this consensus group or your hope for the future for type one diabetes in general, or for AID? Let's start with Revital. Give us one personal thought or key takeaway from this whole experience.

Dr Nimri: We have come a long way with diabetes-related technology. It is so exciting for not only our AID system in routine use, but people with diabetes can actually choose the AID that is most suitable for them. As a society, we should make every effort to make AID systems available to anyone who wants to use it, while failure to achieve equity may translate into a two-tiered system of diabetes care based on who can and cannot use diabetes technology. So, my biggest hope is to have fully automated AID system soon and this will be a bridge until we find a cure for diabetes, hopefully.

Dr Bergenstal: Great. Thank you, Revital. It's great to hear you put those words into motion and see all the great work you've been doing to make these things happen. So, Moshe, let me turn to you next.

Dr Phillip: When we first met in the first ATTD meeting in 2008, the idea of having a closed loop system, automated insulin delivery system was a dream and we were not sure whether we will be able to achieve it, to get it. And now, with that consensus, I feel like we have closed the loop or closed the circle, and we're able to show that if we as a community focus on an aim or on a mission, we can take technologies that were developed for other purposes and turn it to the use of people with diabetes. And that's, I think, the take home message that I take from that lesson.

Dr Bergenstal: Great. Thank you, Moshe. And it's so great to have you offering this perspective, since you've been such a leader in the field, to look back and look forward. So, thank you for your insights. Tadej, one more reflection before I close it off here?

Dr Battelino: Yes. Thank you very much. It was such a joy listening to all this. And I agree, my wish is that the automatic insulin delivery could finally reduce the burden, the day-to-day burden, that people with diabetes have with their disease. And I have a feeling, and this is just an assumption, that we are closer to this wish, to this goal on one hand and, on the other hand, there is published tens of thousands of data showing that the time in range is really high with these devices. So my second hope is that also the complication rate will finally go down for people with diabetes so that they can live a high-quality life comparable to people without diabetes, as Revital said, until we find a cure, and I am a full believer we will. Thank you, Rich.

Dr Bergenstal: Thank you, Tadej. It's great to hear what you're thinking and I've observed enough that some things you start to think about, actually, you make happen, too. I want to thank all of the faculty: Moshe, Revital, Tadej. It was such a valuable experience to get your wisdom on this consensus paper and meeting and looking forward to the potential of even new technologies that are coming. So we appreciate all that each individual with type 1 diabetes contends with every day, and it is our sincere hope that this minute-to-minute, day-to-day, and week-to-week burden will be reduced somewhat by AID therapy, and there will be new innovations coming along to even further improve the data and the dreams that these people have. We hope this is all coming soon.

Thank you to our faculty and, to the audience, thank you for participating in this activity today. We hope you enjoyed it, took something away that will be helpful for you. Please continue on and answer the questions that follow and complete the evaluation. Have a good day, everyone.

This transcript has not been copyedited.