Good afternoon, and welcome to the clinical endocrinology update. Today's topic is evaluating advanced technology for patient-specific management in type 1 diabetes. This is a very relevant topic because it has really impacted many lives of people with diabetes. We have a wonderful faculty, all happen to be from the Barbara Davis Center for Diabetes, myself being the chair. And the other two faculty members are Dr. Akhtaruk, he's the Associate Professor of Medicine and Pediatrics at the University of Colorado, and Greg Ferlanza, he's Associate Professor of Pediatrics, and he's the Director of Pediatric Diabetes Technology Research at the Barbara Davis Center. So that's myself. I'm Editor-in-Chief of Diabetes Technology and Therapeutics, and also Professor of Medicine and Pediatrics, and Director of the Adult Diabetes Program at the Barbara Davis Center for Diabetes, which is part of the University of Colorado Medicine. These are my conflicts of interest, and none of these really conflict with any of the parts that I'm going to present in the next 10 minutes. This slide I included for a specific reason, even though most of the data on this slide highlights only the association of people with type 2 diabetes, more than 1.5 million people across 19 high-income countries. As you can see on the left side, on the upper graph, every decade, if you postpone the onset of type 2 diabetes, you actually advance the total life lived. For example, if somebody is diagnosed in the first, let's say second or the third decade, 20-30 years of age, they're going to lose every decade about 10 to 15 years. So keeping that in mind, our goal should be, can we postpone the diabetes onset for type 2 diabetes? And now you all know, all these new therapies are showing that might be the case with GLP, but we are not going to discuss that in this. But why this is relevant is because today, 60% of the new onsets who are diagnosed with type 1 are actually adults. It's no more a pediatric disease. And part of that might be because many people with type 2 diabetes were actually misdiagnosed. George Eisenbart and I did about 20 years ago a study at the Barbara Davis Center. Anybody who came into the clinic with type 2 diabetes, we found about 15 to 20% actually had positive antibodies. So part of that might be this higher new onsets in type 1 adults might be because they truly were type 1 all along. And secondly, both type 1 and type 2 diabetes are increasing at a phenomenal rate. So it doesn't matter. These two references down at the bottom are very important for you to take a note, both on High Impact Journal, NEJM, as well as the Lancet Diabetes and Endocrinology. So this is the state of affairs when you talk about the diabetes. On the left side, this is the data from T1D Exchange that shows that on the five-year follow-up, if many of you remember who might be part of the T1D Exchange, which included more than 80 centers and more than 35,000 patients across the U.S. with type 1 diabetes, it was a registry program led by JAIP Center and funded by Helmsley Foundation. At that time, the initial study and the follow-up study, this is the follow-up one, which was published in Diabetes Technology and Therapeutics. You can see the one in the follow-up is blue. That people on the follow-up in the five years after the initial report actually did worse. Despite all the new technologies, their A1Cs were actually getting worse by at least half a percent. However, more recently, on the right side, this is the data from T1D Exchange QI Collaborative. That's slightly different, meaning thereby it is not really, it is all done online. It is more of a QI Collaborative project, and all the data may not be exactly verified, but clearly shows that now with the newer technologies in the last few years, the trend is headed downwards, indicated in the blue or the green line. I'm just going to show you two slides on the continuous glucose monitoring. Since both Dr. Apturuk and Greg is not planning to show anything on the sensors, since I've worked on the sensors for the past 30 plus years, I still remember giving talks all over, including ADA and Endocrine Society, 30 years ago, everybody used to say, if many of you recall when 35 or 40 years ago, SMBG became available, we were all questioning who, why, and then, why would people ever check the finger stick? I remember when I presented the data on Glucowatch or the Dexcom implantable sensor study more than two decades ago, everybody was asking who is ever going to do these, nobody's ever going to use these tools, and the rest is history. You can see all the different technologies, how it has evolved now. We got Dexcom G7, you got Abbott Freestyle Libre 3+, pretty soon, sometime early next year, you'll have 3K+, which is also going to measure the ketone measurement, along with the glucose measurements, and of course, we recently, you might have read, the Eversense also got approval for one year, so now you've got an implantable, small implantable sensor where you still have to change the transmitter every day, every 24 hours, but just approved by the FDA two days ago. So we have come a long way, just to give you the history, when I did studies on Glucowatch, it was not uncommon to see MARD, which is mean absolute relative difference, which is an indicator of how accurate the CGM is as compared to the gold standard, which in those case was YSI, Yellow Springs Instrument, and now more commonly, we use either Novoprime or we use Radiometer. These are the two which are commonly compared with, those are called the gold standard. Going back to time, 20 years ago, it wasn't uncommon to see a MARD of 26 to 28%. Look at it now. All of these on the right lower end, pretty much all A1Cs are now, sorry, almost all of the MARDs are 7% or 8%, so we have come a long way in terms of accuracy, which is pretty much equivalent to what you see with the blood glucose meters. The data and the literature also shows that irrespective of, this is a cross-sectional study where Chantelle Mathieu was one of the senior authors, she's the president of EASD. If the time and range is higher, which is represented here in the yellow bars, any time the time and range is more than 70%, both acute and chronic complications, both micro, microvascular, severe hyperglycemia, and decay are less common as the time and range, which is 70 to 180 milligrams per deciliter is achieved. So clearly shows that you don't even need to look at the A1C, time and range is enough of a thing that you can see on a day-to-day basis. So this is a study I just showed at the EASD last week in Madrid last Thursday. This is from a large database, and it was just published on September 12th, 2024, from Optum Deidentified Market Clarity data from 79 million patients EMRs. After cleaning out all the data, we needed patients who should have had no CGM before the index period, which was between 2019 to 2021. And they had to have no CGM at least for six months before the index phase. And then we had to have data for at least a full one year of the A1C values evaluated every three months. And you can see ultimately we ended up with nearly 75,000 patients who use CGM with type 2 diabetes. And then what we wanted to look at is how does the use of CGM impact with all-cause hospitalizations, acute diabetes-related hospitalization, which is decay or hyperglycemia or hypoglycemia and acute diabetes-related ER visits. You can clearly see that in all the groups there was a significant reduction of all-cause acute diabetes-related or ER visits. In all three groups, we subdivided these 75,000 patients into three groups, NIT, BIT, and PIT. NIT is non-insulin treated group, where they're receiving no insulin, BIT is all people with type 2 diabetes, BIT is basal insulin, and PIT is prandial insulin therapy. In all the groups, the most interesting part is the very first bar, there was somewhere between 10 to 31% reduction of all-cause hospitalizations or acute diabetes-related and acute diabetes -related ER visits, even in individuals who are not using any insulin, who are non-insulin therapy treated group. They could be on metformin, they could be on GLP analog or SGLT. So clearly shows the CGM has an important role in day-to-day management. My last slide highlights a point that we don't have time to discuss further. The prevalence of overweight and obesity in adults in the well-established economies is really going up, both in the US as well as Western Europe. You can see on the left side is the same database from the DTT, the T1D exchange, that about 67% of adults are either overweight or obese. On the right side is comparing the data in patients with type 2 diabetes and type 1 diabetes. In type 2 diabetes, as you all know, more than 85% of people with type 2 diabetes are either overweight or obese, but when you look at people with type 1, it's a similar data as I showed you on the left side, about 60 to 70% of people now getting overweight. This is one thing, maybe in the next time when we do a session, what is the role of GLP analogs in people with type 1 diabetes? With these introductory remarks, let me now hand you over to Dr. Halis Khan Akhtaruk. He's an Associate Professor of Medicine and Pediatrics, and he's going to walk us through automated insulin delivery systems. Khan, please. Okay. Thank you very much, Dr. Garg. So my part will be the automated insulin delivery systems today. So these are my disclosures, all of them through University of Colorado. And so we have to start with some terminology. So first the FDA started with an artificial pancreas, so we got this first terminology then when first things started, and we are trying to stay away from this terminology and we are moving forward to the automated insulin delivery and hybrid closed-loop systems. And in my opinion, the automated insulin delivery is a little bit more, is more explaining to even the patients and the population, and hybrid closed-loop is more like a terminology that we can use. So basically these devices are, we have an insulin pump and it communicates with the continuous glucose sensor and it gets the feedback for the sensor glucose and with an algorithm inside the pump or, you know, an algorithm on the phone or algorithm on the pump, and then it just adjusts the insulin basically and prevents hyperglycemia and tries to prevent hyperglycemia also when increasing and decreasing the insulin. So why the AIDS systems are so important? Because we are getting so many, the evidence that AIDS systems are changing the lives of people with typhoid diabetes from the childhood to the adulthood and even for elderly patients. So I would like to highlight two studies here, both are from Barbara Davis Center. One is with the children with typhoid diabetes. This is a study we compared the patients are coming with the 2016, 2017, and between the 2020 and 2021. And this study is we did with over almost 1500 people. And then the results were excellent that we showed that AIDS systems are already, you know, improving the A1c and decreasing hyperglycemia. But why I chose this slide is because these are exactly the same people. In 2016 to 2020 and change their system or CGM and so whatever, and then you can see here. So if the people are changing from pump and the BGM, blood glucose meter, the hybrid closed loop, these are exactly the same people. You see that there is a significant decrease in their A1c. And the people are with the pump and CGM users, and they are losing the CGM excess and they are going back to the glucose meters and then they increase their hemoglobin A1c. And most of these people lost their insurance or lost follow-up and didn't get a prescription for any reason. And so they couldn't use their CGM. And if you look at that, the other group is the pump and CGM users. If they turn to a hybrid closed loop, so they were using independently a pump and the CGM, it's not connecting. And when we connect them together and you see that it's also helps a lot in hemoglobin A1c. So you see here that the power of the CGM actually in this slide, in this study, and then the AI, the systems is related that how the AI, the systems are improving the lives with the typhoid diabetes. In another study we did with the adults, actually we looked at more than 15,000 visits actually for the 4,000 patients between the 2014 to 2021. And then you see that our technology use, what we defined as the CGM or AID use. So we didn't intentionally define the technologies as an insulin pump because we have a strong belief that CGM is changing everything. And then related to CGMs, the AI, the systems are making a big difference. And if you see that our CGM or AID use was in our clinic was more than 80%. And hemoglobin A1c in the clinic decreased from 7.7 to 7.5. And the meeting hemoglobin A1c less than seven increased from 32 to 41%. And you can see that that's not a big A1c decrease. By the way, this was a statistically significant. So in this kind of longitudinal studies, you have to look at really how many people are changing their life, especially with the meeting, the people are changing the hemoglobin A1c less than seven or meeting the target more than 70 and less than the 4% for the time in reach and time below reach goals. And that's a very incredible number in terms of that. And if you look at that in the independently, the CGM and AID use decreased the hemoglobin A1c 0.5% and 0.74% for the AID use in the same population. So if you look at that, which pumps are FDA approved as of today, and then we have two pumps from the Tandem, Tandem Control IQ and Tandem Mobi. And we have Medtronic Minimed 780G from Medtronic. And we have Beta Bionics Islet, and we have Insulet Omnipot 5, and we have Sequel Twist, and which they are all FDA approved pumps. Sequel Twist is not in the market yet. It will be available probably end of this year. So that's a very important slide that I just want to summarize everything because I have a very limited time to talk and which this is a talk that I can talk on this slide probably another half an hour. So what can you adjust in AID systems in automated mode? So I think that's a very important thing to understand as a provider and also explain to our patients and families. So when they are high or when they are low and what they need to change. So other than looking at individual devices, I always try to look at this way. If you look at the basal insulin, for example, about all the pumps, which can you change? And then I will just explain them, all of them briefly. So I summarize them, all of them here. We already have Minimed, you know, Tandem pumps, Omnipot and Islet in the market. ChemAPS FX is FDA approved, but it's an algorithm only. So we don't have a pump yet that will be in the market in the United States. And it's also FDA approved for the pregnancy. As I explained, we have Twist. It's already FDA approved, but it's used as the type of algorithm, but it's not in the market yet. It will take another couple of months at least. So if you look at the basal insulin here, that you can only change in the Tandem pumps and the Twist for the basal insulin. For the CARB ratio, and you can change almost all of them except the Islet, because in Islet there is no CARB counting. You just announce your meals. So it's a small meal, medium or large meal, kind of the usual, more than usual and less than usual. You just, you know, tell the pump and it gives you whatever it's necessary. For the correction factor, you can only change in the Tandem pumps and the Twist pump. In Omnipot 5, it's not in the automation, it's in the user boluses only. Same thing for active insulin time is you can only change that you can change the algorithm is the 780G. And in Omnipot 5, it's only for the user boluses. In targets, in all pumps has different names for the different targets. So you can just change the target in all of these pumps. And some of them as a target from 110 to 150, you know, 50, some of them as a range target like the Tandem 100 to 160. And there is a sleep mode you can change in target 110 to 120. And in the Islet you can also as a usual target, less than usual, you know, and higher target. So all of them, it's somehow a target change, you can do that. And exercise. So it's an important thing in our patients say you can also have some sort of a temporary target for the exercise. It has different names in different pumps, but you can change in all of them except the Islet. In Islet exercise, you have to remove the pump and you can just put it back for the Islet for the exercise. And that's an important thing to summarize. You can get the slide and then you can use which one we use. If we try to change, for example, basal insulin in Minimax 780G or the Omnipot 5, the system is not going to do anything in the automated mode. So we have to really know that when the people are complaining about lower highs, that we have to teach our patients and families and we need to know what to change. So if I summarize also the each devices with the registration trials and also the real life studies. And this is I started the Medtronic Minimax 780G with one of the registration trials. and that's a comparison of the run-in phase which is we can consider as you know sensor augmented pump which they actually didn't use the autocorrection feature of the pump and during that duration and then the next 45 days and they used as an either hundred twenty target or with the hundred target and compared with the autocorrection using with these targets and if you can see that in this you know in the adolescents and adults and the results are significant and in overall the time in range increased from 68.8 percent to 74.5 percent and you see that in parallel in the adolescents and also in the adults are it's increasing and they are meeting the 70 percent target goal and if you look at the time below range metrics that for them the hypoglycemia especially you can also see that it's also going down so it's not only you're increasing the time range also you are decreasing the hypoglycemia so the results are also confirmed for the same device for the in the real-life study as you can see that here this there is a lot of countries here and most of them are from Europe and you can see that US is not here because at that time US in FDA problem was not done yet and was not enough obviously people were using that only we tried this in the research not in the real life so you can see that the pre-activation of this autocorrection system and then going to the advanced hybrid closed loop and that time in range was 63 percent and in the 780 G with the autocorrection that is 75 percent so you can also see that this is a slight decrease in the time below range but there's a significant decrease in the time above range with the hyperglycemia if you look at the countries which have different food cultures different times that people eat right for example in the south of your Europe you know people eat tends to eat a little bit late but you can see that the results are very similar despite this are the different people different you know food habits and they may be using different settings so the results are all over the 70% which meets that our target for the time in range and the time below range you can see it's very similar also less than 4% so if you look at the tandem control IQ so tandem control IQ I chose this study this is the combination of pulling data of the couple registration studies and if you look at that again is the comparison is either is a pump is using with a CGM system and then most of them are the sensor augmented pump therapy if you look at that again that the differences are significant the people are when they used in the control group and then continue the sensor augmented pump their time and range didn't change but if you look at that there's significantly for the and then with the control IQ they increased their time in which is 70% there is an 11% absolute increase in their time in range and their time above range metrics also significantly decreased and as well as you can see that hemoglobin A1c of course in parallel it decreased significantly and if you look at the tandem control IQ the real-life studies and in in all age groups you can have a look at then we have adults and more than 2,000 patients we have adolescents which is the you know highest A1c in the type on the exchange you know population and the age 6 to 13 you know we have children and you can see that all of them is a combination or separately and then you know their time and range is significant and you can see for the adults are meeting the 70% and then there you know the time below range metrics are also meeting our targets in the one year of the real-life use of this almost more than 3,000 patients in all age groups. So if you look at the Omnipot 5 and we have similar results this is also a registration study and then you can see that the comparison this is in the pediatric population if you compare this in the baseline standard therapy phase and then with the automated insulin delivery and you can see that the change is very similar almost 11% absolute increase and statistically significant and making almost 70% in the time and range and there's an hemoglobin A1c we have almost 0.6% decrease and then that's you can see that all this metrics in parallel to each other and all of the pumps I don't want to read line by line so even the variability is going down in all of this automated insulin delivery systems. And if you look at the Omnipot 5 there for the real-life study and separately for the children and for the adults and you can see that these are the people are used almost a year and more than 70% of the data and this is Dr. Forlanza's study and then you can see that the time in range and then you know it's in the real life also is almost 70% for adults and then again this is the real-life studies has no control that you know people are using their own settings and other things and it's a completely real you know data is retrospectively acquired in all this pump studies and if you can look at that another thing is that you know you can see that I want to show you here that the people are meeting the goals that you can see more than 70% the people are meeting the target so increase from almost 35 to 50% the people are meeting both of the goals for the less than 4% hypoglycemia and more than 70% and both of them together and you can see from 28 to 44% so that's a great number almost half of the people that we want this to be in this range and meeting their goal with this AID system in the real life. So the islet if you look at the islet so and then this is one of the registration studies and then it's also if you compare this in the 13 weeks of the results and then for the baseline is the comparator group was anything was allowed actually in that comparator group and in the study design and you can see that also the time and range increase from 53 to 65% and hemoglobin A1c decreased and then you can see that and there's a slight decrease in the also time below range for the hypoglycemia metrics and so the people are you know doing excellent in the AID studies but some people are still may choose to use insulin and also some CGM and all together so there is a recent study that we did and it was the results are announced in the ADA 2024 and this the paper is under review right now and then you can see that I can explain a little bit more in this one in the next couple minutes for the inhale-3 study. So inhale-3 study is a study that comparing to the AID systems to do an insulin degredate group we added to do inhaled insulin and which is the TI which is a technosphere insulin so basically we randomized people to two group one group was using insulin degredate with the TI is the inhaled insulin and other group use the usual care of the group either they were on multiple daily injections with the CGM or they were using sensor augmented pump actually majority of them were using AID system of any AID systems and then they also you know everybody started using the extended phase in the degredate and the TI group. If you look at these results for the 17 week study and then you can see that this is a TI group with the insulin degredate compared to standard of care group and you can see that there was no difference and it was a non-inferiority met which was the cutoff the 0.4% and then the difference was 0.11% so in 10 to 30 analysis shows the results were really good actually and some of the patients did actually excellent some of them individual you know numbers did really good some of them you know I didn't do well in this study so it's all it can be explained by that if somebody is bolusing already giving the bolus you know and then regularly and then inhaled insulin is also another option. The people that couldn't do well in the inhaled 3 study because the people were not bolusing much and the AID people was you know trying to you know do the job. So if you look at the post meal and group is that you know AID and our other things were similar we look at the post meal test and it was almost 1 to 2 ratio in overall the study it was almost 1 to 3 ratio and you can see that inhaled insulin group did really well especially in the first 120 minutes after the meals. So and the most common side effect was the cough and then no other severe hypoglycemia or decay was reported. So this is the end of my part then I will just introduce Dr. Forlanza and then I will answer the questions at the end. Thank you very much. Hello everyone my name is Greg Forlanza. I'm a pediatric endocrinologist at Barber Davis Center in Denver Colorado. It's a pleasure to be talking to you today about a topic I'm very passionate about technology use timing and overcoming barriers to improve technology use. These are my disclosures and I usually tell folks I work with every company in the United States that makes technology for people with diabetes and my bias is always towards how to optimize its use and people not favor one specific design or another. And so Dr. Garg outlined for you a lot of the problems that we face with both increasing diabetes incidence and prevalence around the world as well as suboptimal control being something that we know we're seeing across our patient populations and Dr. Arcut outlined the benefits of novel therapeutic advancements from inhaled insulin to the area that I do work in the automated insulin deliveries technologies. And so one of the questions that I'm most passionate about is the question of why isn't everyone doing this. I think that the data that we were just going over shows the benefit of these technologies for glycemic control and when this question has been studied by colleagues in both the adult and pediatric realms and the literature they've identified several major barriers that exist which I want to go over kind of thematically here. And these barriers are the major barrier to technology use is the provider perception that certain individuals should be ideal candidates for certain technologies and that ideal candidates I think is a big term that we're going to be talking about. This concept by us as providers that we should try and figure out who the right candidates are for technologies has led to something that I think we as a field should be embarrassed about which is that we see a gradient where people from poor social determinants of health, lower socioeconomic status, marginalized communities have lower rates of tech use than people from more favorable social determinants of health who have much higher tech use. And I very much believe that the solution to this is for us to stop being gatekeepers of technology access but rather facilitators of successful technology use. The society guidelines, I'm just going over them briefly here, very much support this idea. So to me it's very exciting that over the past three or four years we've seen organizations like the American Diabetes Association and the International Society for Pediatric and Adolescent Diabetes shift to saying that CGMs and automated insulin delivery systems are the preferred methods of therapy for people with type 1 diabetes and in some situations people with type 2 diabetes requiring insulin and that the data that Dr. Arndt just went over outlines that this is based on very strong randomized control trial evidence and so this is actually a grade A recommendation and something that as providers should be motivating to us to wanting to use these technologies. We went over a little bit on data from our center showing the benefits of automated insulin delivery technologies. This is an analysis that was kind of a precursor to the one Dr. Arndt showed you a minute ago where it was looking at what we saw within the automated insulin delivery population and what we see is between 2016 and 2017, the very dawn of this technology, hemoglobin A1c across our pediatric population was about 8.2% and then with more widespread AID use with more robust systems and with better provider understanding of how to use these novel technologies, we saw the average A1c dropping down to 7.6% and I know a lot of folks in the audience are adult endocrinologists but bear in mind the bulk of the people in this analysis are adolescents. As Dr. Garg was showing earlier, that's the lifetime peak in hemoglobin A1c for the population and just to be honest with you, seeing kids with an average A1c of 7.6 is phenomenal. The story for that average was as you're seeing up here like 9 to 10 so that was really exciting for us to see that that was possible. So a big question as we've seen these technologies emerging is when is the right time to start them and this analysis that came out from the adult group at our center was looking at the timing of CGM initiation and this actually drove a change in our practice, the results of this analysis. So what this is looking at is 400 people with type 1 diabetes who were followed for seven years and I know a lot of people in the audience may read or review or be conducting drug studies. In technology studies, you don't have to do seven-year trials because the tech moves faster than every seven years so to have a seven-year analysis is pretty phenomenal for a technology approach and so the question here is kind of dividing that up into groups looking at people who started CGM within the first year after diagnosis and comparing that to people who did not start CGM and what you see is that the CGM use group had hemoglobin A1c values that were consistently between 0.5 and 2.5 percent lower than the non-CGM use group and what's really cool is if you look at the people who then began CGM, so the crossover group, they shifted down and became more like the CGM group and within our pediatric clinic at Barbara Davis Center, the findings from this analysis moved us to introduce CGM before insulin pumps which was a change. CGM was originally viewed as sort of adjunct insulin pump therapy and now we've sort of inverted it to where insulin pump therapy is adjunct to CGM use and the real key concept that we arrived at as providers is moving to an opt-out approach rather than an opt-in approach and that's the idea that our default should be to have people using technology and we needed to justify why they weren't using it rather than our default being, you know, why do they qualify for it and little changes like that with our practice models can make a big difference in what we're seeing for people in terms of how readily they're using technology and so when we look at technology and glycemic outcomes, the big thing that we're seeing here is that as we see increasing technology use, we're seeing decrease in hemoglobin A1c because our technology is now automating and is able to drive better control as we were talking about previously and so we over the years have seen within our both adult and pediatric clinics big increases in people using technology as well as among technology users the percent achieving goal hemoglobin A1c and so to me this is really the thing that as a provider should be motivating is 47% of our technology users are meeting ADA goal for hemoglobin A1c and that means for me as a provider obviously I'm very biased overall majority of my patients are using technology but that means my average patient on a given day in clinic is meeting goal for hemoglobin A1c and that's one of the things that I like to kind of convey is we've seen this shift over the last decade we've seen as providers that we are having our visits routinely be focusing on maintaining goal A1c rather than we in the South call come to Jesus talk where you have to figure out how to change everything to get to goal our average person is meeting goal and that brings us to the timing of automated insulin delivery and so this is data from a Venn JDRF now breakthrough T1D sponsored study we conducted where we actually started automated insulin delivery in children with type 1 diabetes antibody confirmed at the time of diagnosis we were doing this in the study to try and see if starting technology at the time of diagnosis could preserve islet function and for those of you that may have seen the results from the study before you're aware that did not do that however it did do something that is still very valuable which is among the people that we started the technology on they were able to achieve a time and range of 81% over the first year of diagnosis which was significantly improved by about 14% from the people who were just using CGM or using sensor augmented pump therapy CGM and an insulin pump without automation and to me I think this is also a motivator to us as providers demonstrating that even in a period of time where people are not traditionally starting technology the first month after diagnosis we were achieving 85% time and range and through one year we were achieving an average of 81% time and range not only is this giving good metabolic control to these patients and their families but it's also setting them up to demonstrate that this is possible this is the norm for diabetes is having this level of glycemic control we already kind of reviewed this data this is my own kind of living summary slide of the results from the different pivotal trials I want to include it here for your benefit and so the main thing I like to emphasize is there's no best device all the devices are showing benefit in the populations in which we're looking at them in terms of minimizing hyperglycemia improving time and range and getting people to goal time and range and so I wanted to share a story with you of a patient who I started on technology very quickly after diagnosis because I think patient stories can help us remember things in a way that sometimes just blasts of data on a table which I just showed cannot and so this is a kid that I saw a 15 year old who was diagnosed with type 1 diabetes presented to the ER in moderate DKA he corrected over about two days I saw him and his family in the hospital and then he came over to our Center for Education he describes himself as a straight-A student interested in engineering and technology and when I was discussing with him what he wanted to do for his therapy he told me that his paternal grandfather so one of Dr. Garg's you know adults with type 1 diabetes has type 1 he's had it since childhood and is currently doing well with the Medtronic 780G system with the Guardian 4 sensor he's very close to his grandfather spends a lot of time with him I do a lot of different activities and he wanted to be on the same system as his grandfather and so I disrupted our usual process a little bit which I'm known for doing and within three weeks after diagnosis we had him started on the 780G system which is what he had requested and we transitioned him from MDI where his dosing was coming down as he was in his honeymoon period to the 780G and here demonstrating that we started him on the recommended targets of a target of 100 and an active insulin time of two hours and I decreased some of his settings because again as he was entering the honeymoon his required dosing was coming down and here is his data 15 year old boy two weeks after starting the 780G roughly six weeks after diagnosis and we see here is time and range was 96% with 0% time below range and he was having a very vigorous appetite as teenage boys are known for doing eating you know over 150 grams of carbs a day, bolusing about three or four times a day. So not someone I would classify as a super user and not needing to really do any finger sticks or any sensor calibrations. And I know that Honeymoon contributed to this. I'm not trying to say the system is doing this on its own, but my point is that with use of an automated insulin delivery system, he was over 90% time and range. And this was his data roughly three months after diagnosis. So a little bit of a waning as the honeymoon period started to abate a little bit, but still 84% time and range, 0% hypoglycemia, bolusing four times a day, and eating over 150 grams of carbs a day. And then this was his data, oh, I skipped it, but his data's most recent visit was very similar to that. Now over a year after diagnosis. And I share this story because I think it's important for people to understand that this is not something that should be an outlier. This is what we should all be aiming for doing. And so I think the main point about this case is that this was someone who wanted to engage with technology that early and wanted to be wearing it and using it. So it was started very, very early. He has been successful with it and we should be facilitators of that, not people who are limiting it. And then the last thing I'm gonna mention very briefly and then I'll wrap up is we are now moving to the next frontier in technology, which is automated insulin delivery via AI and fully closed loop systems. And so I'm gonna skip over a little bit of this explanation about what neural networks are, but basically we're now taking the same stuff that some of you may be using with chat GPT, some of the things you may be using in for image recognition or image generation or other areas of your life and applying these ideas to diabetes. And with use of these systems, we are now able to develop automated insulin delivery that can detect when someone has eaten and automatically dose them for meals, basically removing the ability of people to need the carb count at meals and just relying on AI to do it. We're running a study literally today on this technology and it's something we hope to see come into the real world within the next three to five years. And so to kind of wrap up, automated insulin delivery benefits children with diabetes across different ages and durations of diabetes and baseline control. There is no best candidate. There is no best system. We should be advocating for technology use and facilitating its success in our patients and their families. So thank you very much for listening. I look forward to getting to some of the questions. But I want to take this time for the Q&As. We have too many questions. I'm going to answer some, but it'll be tough for me to answer and include in the next 10 minutes all the questions. The first question was regarding the pregnancy. Is there any system that's approved for the pregnancy? Yes. The only system that was recently approved in Europe as well as in the U.S. In the U.S. it was approved two weeks ago called the CAM APS system. It is their algorithm that's approved because they can go down to 63 to 90 milligrams as their targets. However, the pump that it works with is not approved in the U.S. So unfortunately not. To the same extent, what is our experience, Dr. Polsky here at the Barbara Davis Center has done two or three studies. In fact, the majority of the people who go through pregnancy use this on a regular basis. So I think it's an off-label use as far as it comes to the U.S., but a lot of the practice unfortunately happens to be off-label. Now, one of the questions, maybe I'll have Greg answer. What is your experience of patients that report frequent and regular differences over 15% between sensor and capillary readings? Yeah, so that's a great question. The first thing, and I'm actually giving a lecture to our local school nurses about this tomorrow, is that the blood sugar meters are not and have never been incredibly accurate. And so the assumption people have always had is that the meter is providing truth, the CGM is providing an error, and so the CGM must be off. If you actually study it, fictitious hyperglycemia is much more common with blood sugar meters than it is with CGMs. And so what we usually recommend is people managing to what their symptoms are tending to show. So we have a little bit of a recall bias where when the sensor is reading a value that is different from what someone is feeling and they do a finger stick, and then the finger stick is closer to what they're actually feeling, they think the sensor is off, but that's because they're never doing the opposite, which is doing a random finger stick when the finger stick is off. They're reporting what the sensor value is. So if we see a discrepancy between what the sensor is showing and how someone is feeling, we do recommend following the finger stick value. But when you actually look at the accuracy data, the sensors are generally providing less inaccuracy than the finger stick meters are, but there's a little bit of randomness to that. So either one can have periods of inaccuracy, and so that's why you've got to be thinking and evaluating how you're feeling in addition to just relying on the electronics because both can have their own errors. Well, thank you, Greg. Since the questions are so many, the next one is for Khan. Could you comment on some pitfalls of AID systems? I mean, in the AID systems, obviously there is nothing one fits for all. So, you know, it depends on the person, like if somebody should be motivated and if somebody doesn't want any tubing, if somebody doesn't want any alarms, if somebody, you know, doesn't want a feature, and then I think this can be a burden. So, and sometimes people are, you know, want to be more independent, obviously something will be attached to you 24 seven most of the time, and it can be another issue. And then sometimes there's an alarm fatigue because, you know, your blood sugars are going up and down and you can get some notifications, and then it can be a pitfall in, you know, some of the cases. And it's not that, you know, everybody is, will be getting one thing. I think that's what I would suggest for discussing the pitfalls, but also highlighting that what's good for the AIDs and try to choose the best one for the patient. I think that can remove that barrier from using AIDs. So just to add to that, to this particular one, there are two other pitfalls we need to keep in mind, cost of AID systems. Many people in the U.S. might have a high deductible cost or 20% co-pay, and that may be important. And then in kids' case, Medicaid does pay. And when they become adults, many of those people are no more reimbursed or they may have a very different plan and they may not be on their parents' plan. And the second part is the ketosis. The pump therapy still continues to be associated with ketosis because anytime you have a clogging or catheter blockage, you have, you know, the ketosis. I give this example to all the people. Recently, we did a study with Dexcom, you know, G7 sensor, nine out of the 60 patients came in with ketosis early morning and the ketone levels were somewhere in the range of 0.9 millimoles and clearly close to one, which is really high, which means we couldn't induce hyper and hypoglycemia. And thus, those are some of the pitfalls. But when we look at DKA rates in the pediatric population, the rate of DKA in MBI users is higher than the rate of DKA in AID users. And so it doesn't prevent it. Sure, I mean, that is true. There is another question, Khan. These are expensive technologies. Can you speak to the cost-effectiveness of using these for type 1 diabetes versus type 2 diabetes? And recently, there is only one of the AID systems abroad, which is Omnipath 5 for the type 2 diabetes. And type 1 diabetes, they are mostly being used. And I think that, as Dr. Garg mentioned, that if obesity are expensive, there is an opportunity for that to get it with an insurance system. I think that's great. And sometimes people can take advantage of enrolling a study and testing these things because you make a commitment for five years and you may not change that. I think that's the best way probably to get it. And then with the insurance and then to get the AID system because it makes a very big difference in the people's life. I mean, type 2 diabetes, we have only one pump just approved recently, so we don't have any real-life outcomes yet. So probably it will be a question will be answered probably next year. So just to add to that, to Khan's comment, I personally have a little bias. I think when you add fuel to the fire, which is what you're doing in people with true type 2 diabetes, if you add insulin by giving more pumps, you're actually adding fuel to the fire. They're already obese, they're gonna get more obese, it's not gonna answer the question. So to me, the first step in these uncontrolled people with true type 2 diabetes, your antibody negative should be, have you tried the GLP analog? Once they lose decent amount of weight, I own clinical practice, there is hardly anybody who needs other than a small amount of basal insulin. They don't need pan-deal insulin. Thank heavens, they never need insulin pump therapy. So I don't think that is the answer. Let me ask Greg to answer, because there are so many questions, Greg. What is your experience with using Dexcom Stelo? I know it's not approved for kids, but it's approved for anybody. Have you heard any calls? We've started to see patients, so of our patients, we've started to see their parents and siblings use it as part of screening. We didn't talk on topic today about TZL or the pluzumab, but that's something our clinic is very engaged in. And so with the increased focus on that and early detection, we're seeing more and more parents and siblings and probably broader world children of people with type 1 diabetes, where it's being used as an over-the-counter affordable screening tool to try and see if they need to be worked up as stage two. That's the main place where we're seeing it. Within the PEDS world, I imagine it'll have use in people who are overweight or obese, don't qualify as sort of pre-diabetes to get their stuff covered, but they wanna see what's going on for lifestyle modification. But normally in practice, it's mostly been as a screening tool. Well, thank you, Greg, for that wonderful update, but it might not be fair to evaluate Stelo or for that matter, the two products that are about to be launched by Abbott, partly because it's only been the last two weeks. The few patients who have called us to the clinic, some say their readings are reading higher or lower. Some have to go into a doctor's office to get an A1C and realize that A1C was only 4.9. And many people say it doesn't really last for even leave alone the 15 days, may not even last 10 days. So I think it may be too early to say what's happening in the real world with Dexcom Stelo sensor. Okay, for Khan, what do the sensors, why do the sensors read lower glucose values at night due to pressure on the sensors at night? So basically this measures the interstitial glucose and then you're just decreasing that surface too when you are just pressuring on that too. And then you can see there is a typical like going down and going back up perfectly there. You can just recognize that there is no trend like going down. So basically it's a physiologic thing that you are just basically decreasing the duration that sensor is between the interstitial fluid. Just to add to that Khan's comment is partly because you're limiting the amount of oxygen going into the sensor. That's basically a hexokinase technology. So since you are limiting oxygen because of the pressure, thus it tends to read low. All right. You're sitting on your foot. That's what I always tell the kids. Sure. Like when you sit on your foot. Okay, Greg, for you, what are your suggestions? You may not have an answer to this. What are your suggestions for patients who use Humulin R U500 and want to use a AID system? No, I'll be happy to answer this, but go ahead. So the issue with Humulin R is that it's really, really slow. And so when you put it into an AID system, it's too slow to behave the way that the assumptions about PK and PD are within the system. And so it won't onset fast enough. And more frighteningly to me, it won't offset fast enough. And so I would not recommend using something that lasts that long within an AID system. I still get asked me about that cost. These are high total daily dose. So I would be afraid of it though, because the system will assume, even if you set the IOB to five hours in the system where you can set it, the system will assume it's gone before it's really gone. So let me add to this. I have two patient examples who clearly depict how you can use those who need large amount of insulin. What I did for these two individuals, one added GLP, either semaglutide and more recently terzapatite. That will definitely reduce their dose. Secondly, might also help in losing weight. Because their insulin intake, whether it's type two or type one, who are really overweight and big time obese, in those individuals, if you have to use Imulin R, what I did is I gave them a long acting insulin injection of Treceba, let's say U200, about 80 to 90 units. That reduces the amount of insulin they will need through the pump and then you can try it. I mean, I could go on and answering this question. I think we have gone over. I want to ask Brandon, do we have time to go over? We have too many questions to answer. Is somebody from the end of society there to guide me if I can go on or should I just? Okay, while he's typing the thing, I'll have one more. Could you comment some of the pitfalls of AID, Dr. Akhtaruk? Sorry, what's the question? I couldn't hear. Could you comment on some of the pitfalls of AID? I think AID is the most important thing Comment on some of the pitfalls of AID. We did that. I think we answered that question, yeah. Oh, we did, okay. Here's another one. What is your experience of patients that report? Oh, no. Is there a head-to-head study between pumps looking to see one system outperforms the other? Greg? People ask that question a lot and I always try and discourage them from thinking that's a solid idea. And so the issue is that any study that was attempted to be done like that is gonna suffer very heavily from how you design the baseline group and how you design the comparators. And it would essentially just be whatever company was sponsoring that trial, the study was rigged in their favor and the results wouldn't be valid. There's not a best system. There's not a universally superior system. There's just systems that would work better in certain cases versus systems that work better in other cases. And that's why I think that slide that Khan showed earlier, some slides that we have on the Panther Program website, demonstrate what the features of each pump are. And so I think it's really about, in pediatrics, I go to Harry Potter and say it's the wand picks the wizard. There's not one best design, it's just the right match for the right person. But there's not a universally best design. I wish there were, but there's things from each system that are superior versus inferior to others and it's about finding the right match for that individual. Fairly explained, Greg. There are two more that I'm gonna take and this is a good one for Greg. How do we learn about loop techniques that the patients use? Do you think they are good for patients who loop themselves? So Sage is baiting me because we published a commentary earlier this year that was a debate between myself and Dana Lewis, who's the designer of one of the DIY, or sorry, open source APS, she doesn't like the term DIY, builds of the Android APS. And so the debate that we published in DTT earlier this year, you can look up my last name or you can look up Dana Lewis, my fellow I-Dean was also an author on that, summarizes a lot of the research and the sources in the parts that Dana contributed and summarizes a lot of the counterpoints in the part that I contributed that people got mad about. And so there are a lot of resources for learning about it, but as I kind of outlined in the debate, there's also a lot of reasons why as licensed credential providers, we have to be very careful when dealing with things that are unregulated. Well, since there are still a lot of people online, Khan, could you please take the last questions? Many of the sensors fail and have errors. And the lag period is another issue. What are your thoughts? Patients with hypoglycemic unawareness get frustrated and usually wanna give up the technology. I think that hypoglycemia doesn't happen in five minutes. So I always try to explain that it's a process. So it starts and ends at some point. So if it's in the CGMs that are, and I wish I can show that it's in the five minutes, it's go down and going to the same level when they are 120, that's definitely not a low, it's a compression low. So it needs to be a trend. So I always tell my patients that they should also not reading the numbers, they should be also reading the arrows, also looking at the trends too. And they need to be managing things before ahead of the time. And I think that CGM is not only we should be looking at an absolute number thing and it should guide us what we need to do. And you are in doubt. If you have a good meter, you can always to check finger sticks and then nothing will be 100% accurate. But all the studies show that people that are using the CGM system significantly decreases their hypoglycemia. So I would disagree on point that I think we should give some reassurance of the accuracy and then most people are doing well. So very good points, Khan. Just to summarize what he was trying to highlight, rather than telling people to look at individual numbers, ask them to look at rate of change. And if they can't calculate rate of change of glucose, they can look at the trends. Because most of these sensors are giving you glucose values either every minute or every five minutes. You have enormous data and that's what people should follow, not an individual number. With those closing remarks, please scan this QR code so that you can get the credit for the CME. And thank you for attending the CME session. I'm so sorry that we went over by about eight minutes, but I really wanted to take this opportunity to thank Dr. Akhtaruk and Greg Ferlanza for joining me and they did an amazing job. Please give them a big applause. Thank you guys. Take care. Thank you.

type 1 diabetes

continuous glucose monitoring

automated insulin delivery

patient-specific management

Barbara Davis Center

A1C trends

FDA-approved systems

technology access disparities

glycemic control

artificial intelligence