Melanie Cole, MS (Host): Welcome to Better Edge, a Northwestern Medicine Podcast for physicians. I'm Melanie Cole. And we have a thought leader panel for you today with three experts from Northwestern Medicine to discuss a recent clinical trial they led, published in the New England Journal of Medicine. Their trial aimed to understand the efficacy and safety of treating patients with type 2 diabetes with an automated insulin delivery system, which has been well-established in patients with type 1 diabetes.

Joining me in this panel are Dr. Grazia Aleppo, she's a Professor of Medicine in the Division of Endocrinology; Dr. Kasey Coyne, she's an Assistant Professor of Medicine in the Division of Endocrinology; and Dr. Jelena Kravarusic, she's an Assistant Professor of Medicine in the Division of Endocrinology.

Doctors, thank you so much for joining us today. And Dr. Aleppo, I'd like to start with you. Can you provide us with an overview of your recent study on automated insulin delivery systems for patients with type 2 diabetes? What prompted the need for this research and why is it so important, Dr. Aleppo, to evaluate automated insulin delivery systems in type 2 diabetes at this point in time, given their established use in type 1?

Grazia Aleppo, MD: Well, first of all, thank you so much for having us. My wonderful colleagues and I are very happy to be here today. So, as you know, type 1 patients, you know, they need to go on insulin from the beginning of their diagnosis. They lack insulin from the get-go. And patient with type 2 eventually will run out of insulin as well. And even though we have this fantastic medication, the GLP-1 receptor agonist, SGLT2 inhibitors eventually, or at some point, or sometimes even earlier than we think, these people with type 2 do need insulin. And it's so challenging and a lot of cumbersome also to do all these injections per day, and sometimes they forget them. There's no modulation of the insulin dose through continuous glucose monitoring. So, the study was made to say, do we think that people with type 2 requiring insulin can get better result, glycemic outcome when we have them use an AID system. And so, this was the principle from which we decided to have this study prepared. We were very happy to be chosen as one of the sites to do this study.

Melanie Cole, MS: Dr. Coyne, what methodology did you employ to ensure reliable and valid results in this 13-week randomized control trial? Tell us a little bit about that.

Kasey Coyne, MD: Sure. So, there were several methodological steps that really helped ensure reliability and validity of the results. I think that two of them in particular are worth mentioning today. First and foremost, like you mentioned, this was a randomized controlled trial, which is, as we all know, the gold standard for medical research, especially when we're thinking about things like ensuring internal validity of results.

So for this study, the participants were randomized in a 2:1 ratio to receive either the automated insulin delivery system or for the control group to continue their pre-trial insulin regimen. And the outcomes that we were looking at, the primary efficacy outcome was the hemoglobin A1c level at 13 weeks. And there were also several secondary efficacy outcomes, primarily looking at different continuous glucose monitoring metrics, also at 13 weeks, as well as some safety outcomes regarding hypoglycemia or prolonged hyperglycemia episodes. And so, by using such a rigorous study design, that really gives us confidence that, yes, we did everything we could to minimize bias to account for potential confounding factors. So when we're making these comparisons, we can trust that they are really meaningfully looking at the relationship between the automated insulin delivery and these outcomes of interest for our patients with type 2 diabetes.

Grazia Aleppo, MD: And Kasey, remember when we chose the patient, the FDA was very strict. We had to have specific kinds of patients. You remember, we had to have people on GLP, SGLT2, more than 100 units of insulin per day, not too good of an A1c, only 10% could be less than 7, more than 8. So, it really was meant to be a study to match and mimic the real population in real life. And the visits were not so many either. So, we didn't have so many visits, thankfully. But when we did, we were trying to truly mimic in a way what we would do in our usual care. Of course, there's more visits in 13 weeks that we usually do in clinic, but still...

Kasey Coyne, MD: Yes, absolutely. And I think in addition to all of the different parameters that you've already mentioned, you know, the recruitment goals also ensured that there was a wide variety of age ranges and different durations of diabetes and the type of insulin that people were using beforehand. So to your point, it really did try to kind of mirror the patients that are walking into our clinics every day. And not just some kind of isolated study group that, you know, this would apply to, you know, here and now in this trial, but maybe not in the real world. There's really an emphasis on ensuring that not only was there a robust methodology used to ensure internal validity, but that the patient sample, or participant sample rather, was reflective of the general population as well.

Grazia Aleppo, MD: That's right.

Melanie Cole, MS: Dr. Aleppo, tell us a little bit about user acceptance and ease of integration, especially among older adults and those that are less tech savvy. You and I have talked about this before, and this would seem to be important in this trial.

Grazia Aleppo, MD: That's right. We wanted to understand younger people are more sort of at ease with technology. Can older people do that? And so, we were making sure we had a very long training session about two, two and a half hours per patient. Sometimes having them come back for the first change of the insulin pump site. But the challenge was are you able to maneuver all the pieces? And so, we made them do sort of water or saline trials with the pillow so they will feel comfortable. What I'm trying to say is it wasn't just an easy training. And just remember, again, people with type 1 are more used to use technology, but we wanted to see can people with type 2 use technology? And some of them did so fantastic with very little training. The other one, they need more help, still were able to continue and do it, but needed a bit of a longer training and be aware of their challenges. So, we're following them a little bit more closely to make sure they didn't feel sort of confused. They also could get in touch with us and say, "Hey, I have a question. What do we do?" And then, we have this system with texting messages, it's okay because it's encrypted. And so, all this message would come in and it works okay. It's okay to have that. The participant was feeling comfortable. But you like to remember, not everybody can do it.

One of our participants, a very intelligent gentleman, was having such difficulty with putting things together. We thought maybe he shouldn't continue, and he was fine with that. He understood that that was not the tool for him, even though he was using CGM, but the pump for him was too much. All the people, even older than this gentleman, no problem whatsoever. So, it depends on the individual.

Kasey Coyne, MD: But also, to your point, I think that also just speaks to the fact that even with those limitations, I think that the fact that so many people were able to go through the study and successfully use the pump really does highlight that this can be an option for a wide array of patients and maybe some that we might not necessarily even inherently think about.

Grazia Aleppo, MD: Exactly. And they don't want to give it back. I said, "No, no, no. You got to give it back. This is a research study. You can't keep the pump." And it's an excellent point and we thought, "Oh my gosh, I never thought this patient would be a perfect candidate." And here he is or here she is enjoying using this tool and doing so much better from the get go.

Melanie Cole, MS: Really good points. And Dr. Kravarusic, how do the key findings of your study regarding glycated hemoglobin levels and time-in-target glucose range compare to the existing data that we have on AID systems in type 1 diabetes? What new insights do these results add to the conversation?

Jelena Kravarusic, MD, PhD: Thank you for asking that question. The results were actually even more impressive than in participants with type 1 diabetes. Not only did the hemoglobin A1c go down, but there was actually a greater reduction in hemoglobin A1c in participants with type 2 diabetes. These results show us that not only AID systems should be offered to people with type 2 diabetes, but that they're a safe and effective way of managing type 2 diabetes in patients who require insulin.

Grazia Aleppo, MD: And remember, Jelena, so, you know the time-in-range we saw, is always going to be going up and up, and it went quite a bit, 16%. But the other group, 1%. So, you can tell those were using CGM, we gave them CGM, but the difference is the modulation of the insulin and the use of insulin pump in our patients. What do you think that was? I mean, I'm sorry, I'm asking the question here, but because we work together in so many studies and with patients, what made the study so much better then for this type 2 patients?

Jelena Kravarusic, MD: You mean in regards to when you compare it to MDI?

Grazia Aleppo, MD: Yeah. Multiple daily injections.

Jelena Kravarusic, MD, PhD: I really do think it's the algorithm of the pump that is able to decrease and increase the insulin according to the blood sugar of the patients, which, as you know, the injections can't do. So, that offers so much more flexibility in management of type 2 diabetes that it's impressive.

Melanie Cole, MS: Dr. Coyne, what do you consider the strengths of the study? And conversely, please discuss any limitations you faced, how they might influence the interpretation of these results.

Kasey Coyne, MD: Sure. So, I think there are many strengths of the study and not to sound like a broken record, but I do think a lot of it is rooted in the really strong methodology that we've talked about. So again, the fact that this was a randomized controlled trial, and also just the heterogeneity of the population that was sampled from. I think that really gives us confidence that our results can speak to the true relationship between the benefit and the use of AID for patients with type 2 diabetes, and that it is reflective of the patients that we see in our clinic every day.

Additionally, you know, this study, there have been other smaller trials looking at similar questions, but I think the sample size for this trial was a huge benefit. It was, I believe, around over 300, 319.

Grazia Aleppo, MD: Three-nineteen.

Kasey Coyne, MD: Yeah. And so, while there have certainly been other studies looking at this in the past, I think the sample sizes were much smaller, something under like 100 participants. And so, I think just the sheer number of people we were able to enroll across all of the sites really just lends a lot of validity and statistical power behind the findings.

Grazia Aleppo, MD: And Kasey, the thing that also want to point out is the strength is we think or the audience at large might think, you know, why do you have to have insulin pump when people are using GLP. But the people using GLP and SGLT2 did better with the pump than those who were just on the regular multiple daily injection. So, I think that also is a major strength. Because here, you have a person using all this recommended medication, yet they're not getting full benefits unless the insulin is sort of modulated to a more physiologic delivery.

Kasey Coyne, MD: Yeah, absolutely. And I think especially for our patients with type 2, you know, usually the first therapies that we think about are non-insulin agents. And so, I think having this data from this trial, and these really impressive findings, I think that should hopefully be a signal to a lot of clinicians that we should not only be thinking about insulin for these patients, but thinking about insulin pumps. So hopefully, just changing a little bit of maybe some of the clinical inertia that's out there when we're seeing these patients in clinic.

Yeah, I definitely think that the fact that there was such an improvement in glycemic control for patients who were on, like you said, GLP-1s, SGLT2s, or even both, I think really speaks to sort of the power of this system for caring for these patients.

Melanie Cole, MS: Dr. Aleppo, I'd like you to take us from bench to bedside and interpret the changes in terms of clinical relevance for your patients based on the results of the trial. What potential impacts do you foresee for integrating AID systems into treatment plans for patients with type 2 diabetes?

Grazia Aleppo, MD: So, the study answered a lot of questions and that's why there were so many secondary outcomes. We know the A1c went down in the group of AID, 0.9; in the group of control, 0.3. So, a 0.6 difference of hemoglobin A1c, but also the time-in-range. So when we think about patients with diabetes, we try to always minimize the hyperglycemia. And no matter how much we try, absorption of insulin if given in larger doses might not be optimal. And so, the insulin delivery that is sort of infused in a particular steady way. But also, with the ability to increase and decrease or suspend the insulin removes, number one, the fear of hypoglycemia, so people feel that they can reach the target and have more courageous doses of insulin without having to suffer from hypoglycemia.

Another thing study showed was how the hypoglycemia was also so limited and decreased because, of course, the system sort of suspends the insulin. If you then think about the issue with the meals, there is always a maybe misunderstanding of pump therapy. I don't know if I should call it that. But people sometimes think if you don't count carbs, you're going to be on a pump. But actually, in this study, everybody did better. Whether they counted the carbs, they put small, medium, large or fixed doses. We gave the patient options, "What would you like to do?" And they were trained either way and they thought, "I'm going to do it this way." And so, we left the patients sort of able to choose how to manage their meals. And what we saw was they didn't have to necessarily count the carbs, they did well no matter what.

Another thing that was very important, we actually looked at C-peptide. Why is this important? Because C-peptide is the third portion of the insulin molecule. When it cleaves A and B, the active C is left over. And there is this challenge for us taking care of patients on Medicare, that if they have a too high of a C-peptide level, they may not be eligible for a pump. Well, the studies show that whether the C peptide wasn't measurable or not, everybody did well in the AID portion. So again, it shows that there shouldn't be so much limitation to access to AID systems. I see this in our practice so well already. We have patients on AID systems. But now, these are approved, so we have more, I don't know, ease of use or we feel more comfortable. We did assess, so we're definitely comfortable with these pumps. And we actually offer it to our patients sooner, because remember, even if you use all these new medication as you do. And GLP, if the glucose levels are high when they're fasting, the body cannot handle, they need insulin.

And so, better than forgetting the injection, I went to the restaurant, I didn't bring the injection with me. And so, I say, "You got to push a button. Everybody can push a button. It's on your body." And so, by just promoting this sort of ease of use and better glucose, it really engages the patient. What do think, Jelena?

Jelena Kravarusic, MD: I think, this is a very important point to bring up because I don't know how the other people think. But typically, people with type 2 diabetes are thought of as older and maybe a little more inert in how they approach their control of type 2 diabetes, as opposed to, you know, a younger population who focuses on type 1 diabetes, when they only have, you know, typically that condition. But people with type 2 diabetes are offered this and do extremely well, just like the people with type 1 diabetes. So, that offers me confidence to say, "Yes, we can do better." We don't have to give you a ton of insulin and have you gain weight. We can just offer you the insulin pump and help you when you're already working so hard. You're already taking the SGLT2 inhibitors. You're already taking GLP-1 receptor agonist. You're already modifying your diet. And we can do better than just, you know, a hemoglobin A1c of 8 and, you know, keep struggling.

Grazia Aleppo, MD: And the weight didn't go up much, maybe a couple of pounds. So, that was such a big difference, and the insulin dose went down. So, we know that when you inject insulin by injection, there is a less optimal absorption, I said this before. But with the pump, it's so much smoother that they really cut down by 20% or so insulin. And this has been shown in all the clinical trials. Everybody goes on a pump, type 1, type 2, always needs less insulin.

Kasey Coyne, MD: Yes. And I think that in particular is really going to appeal to patients exactly who are, you know, understandably always looking to for ways to sort of minimize medications or minimize the amount of insulin. So, I think that'll be a really appealing feature for a lot of folks.

Additionally, I think that studies like this will just really open up kind of a whole new kind of realm of of tools for patients with type 2 diabetes. And I think it kind of shifts almost the tone of the conversation, whereas a lot of appointments might to them, I'm supposing, feel like, "Okay, well, we have to keep going up and up and up on your insulin," rather, we can say, "Well, we have this other new tool that can reduce your insulin that has been proven to help improve blood sugar and rather than just sort of feeling like we're piling on more and more and more, it's kind of like a shift in, in paradigm, but it's a new tool. It's new technology, and it's proven that it'll make a difference for you.

Grazia Aleppo, MD: I totally agree.

Melanie Cole, MS: This is fascinating conversation. And Dr. Kravarusic, what changes in practice do you anticipate as a result, and how might these changes and these findings influence clinical guidelines for insulin-treated type 2 diabetes over the next few years?

Jelena Kravarusic, MD, PhD: So, I think I'm going to repeat what we just sort of discussed in a shorter way. I think, in patients with type 2 diabetes who are already taking GLP-1 receptor agonists and SGLT2 inhibitors, we can see that the AID systems are not only safe and effective, but that they offer a higher quality of life in patients with type 2 diabetes in their daily interactions with self-management of glucose.

Grazia Aleppo, MD: We hope to see them in the guidelines. You never know.

Jelena Kravarusic, MD, PhD: Yes, we are really hoping.

Grazia Aleppo, MD: So, the guidelines have to be written with good data. So, this is excellent data, and so we hope that this will be in the guidelines in the future.

Jelena Kravarusic, MD, PhD: I do too.

Melanie Cole, MS: Well, these results are so exciting. What are the next steps, Dr. Aleppo?

Grazia Aleppo, MD: So, we always want more for our patients. And why not, right? I would love to see more automation and this is coming. So, there are studies that are being done or will be done very soon. We try to minimize the role of the user in remembering to take in the dose of insulin for the meal. So, the algorithms have been modified to be understanding a bit more of what we call the disturbance in terms of glucose fluctuation and capture that with insulin dose so the patient doesn't even have to take a meal dose eventually for their meal. And so, that will be the next step.

You want to also optimize physical activity, because you want to promote physical activity. And even though there is less hypoglycemia in general in people with type 2 diabetes, there is still hypoglycemia in people with type 2 diabetes who use insulin because insulin is exogenous. So, it cannot be modulated by the body. It needs to have this algorithm do it for them. So, more responsiveness in terms of on the high glucose and the potential increase for low glucose risk like activity. Smoothening these two things will be truly a greater impact, not only for type 2, but also for type 1 participant patients in the future.

Melanie Cole, MS: I'd love you each to just give us a final thought. And based on this trial, what's one key insight that you'd like to share with other providers and your colleagues that are working with patients that are hesitant to adopt the AID technology in their type 2 diabetes population? And because we talked about the patient population and the technology, why don't you each answer what you would like to see happen and what you would like to tell your colleagues about this fascinating technology?

Grazia Aleppo, MD: I guess I'll start. Be familiar with the tool, you as a provider. There are simulator ops. You can ask the rep to bring you a pump and play with it. Wear it for a little while and see that it's not really that difficult.

The other thing is I tell my patients, "Why do you have a phone?" "Oh, well, because it's so easy. It's convenient." Well, you had to learn how to use your phone, so let's show you something that is a few buttons here and there, but you can still do it. So, you don't make it really scary because it's not, but also make it appealing and simple and convenient and an improvement in their lives.

Kasey Coyne, MD: Yeah, I think one of the takeaways that I gleaned from the study was just the additional benefit of using the AID system in patients who-- you know, both groups had the CGM. And so, this is kind of an additive benefit. So to your question about sort of approaching patients who are hesitant about using the AID system, I think, as with anything, approaching it in a stepwise manner. So, you can introduce technology stepwise and one at a time, or one piece at a time, I should say. I think it's a lot of kind of meeting the patient where they're at right now. And then, kind of trying to take one baby step and then say, "Okay, in the future, there's also this other part of this tool that could be really helpful." So, I think using technology in a patient-centered way and in an incremental way when appropriate, even if the goal is ultimately to use this fantastic tool that the data tells us can really help folks.

Jelena Kravarusic, MD: I wanted to add one more opinion from me that is a little bit different than what you both just said, and that is to call on what you said in the very beginning, when we didn't really know which patients are going to actually be able to take the pump. And it surprises me that sometimes patients that I think will do well sometimes get overwhelmed with technology, but sometimes the patients who struggle with taking multiple daily injections actually do excellent with technology.

So, I think my point is be open-minded when you approach a patient that may have a high A1c, maybe they're struggling not because they need more insulin, but because they cannot just get on top of five injections or six injections or four injections, but the pushing of the button and changing of the site every three days will actually do good. So, I think offering them earlier in patients who are struggling rather than waiting years and years is what I have taken away from this.

Grazia Aleppo, MD: It's a great point because people who had the highest A1c did the best.

Jelena Kravarusic, MD, PhD: Yes.

Grazia Aleppo, MD: And so, thank you and it's a great point. We forget sometimes that we think the ideal candidate without seeing the person in front of us that says, "I'm struggling. What can you do to help me? What can I do to help myself?" So, I thank you.

Melanie Cole, MS: Great points. What an enlightening and engaging discussion this was. Thank you so much for joining us. And to refer your patient or for more information, please visit our website at breakthroughsforphysicians.nm.org/endocrinology to get connected with one of our providers. That concludes this episode of Better Edge, a Northwestern Medicine Podcast for physicians. I'm Melanie Cole.