Tuesday , October 17 2017
Home / Resources / Exclusive Interviews / Transcript: Dr. Richard Bergenstal on Hybrid Closed-Loop Technology

Transcript: Dr. Richard Bergenstal on Hybrid Closed-Loop Technology

See video here.

bergenstal-m-richardSteve Freed:  This is Steve Freed with Diabetes in Control and we’re here today with a very special guest, Dr. Richard Bergenstal, who’s an endocrinologist and an executive director of the International Diabetes Center at Park Nicollet. In 2010, he was ADA’s President of Medicine and Science and in 2007 he was ADA’s outstanding physician clinician. Dr. Bergenstal joined the National Diabetes Center in 1983 and has served as a principal investigator for several very important NIH grants including the DCCT, the EDIC, the Accord, the Grade, and now the NIDDK grant to evaluate the hybrid closed-loop technology and that’s really what we’re here to talk about today. So I want to thank you for your time. This is probably one of the most exciting things that’s happened in the field of diabetes from my experience, you’ve been involved a little bit longer, and I’ve only been involved since 1995. Of all the breakthroughs that we’ve seen, this is truly something that everybody’s been talking about for years and years and it actually got FDA approval within the last couple of days. So maybe you can describe the hybrid closed-loop system, the one you’re working with and how it works the one from MiniMed.

Dr. Bergenstal:  Right, thanks Steve, I would have to agree with you, this is one of those landmark occurrences out of 35 years I’ve been telling my patients, it’s coming in the next 5 years. To be honest, I was never sure I was going to see it in my career. But here it is, the first approval of a hybrid closed-loop and that probably is worth a little discussion of what that means. So, what the hybrid closed-loop means is we have a glucose sensor. In this case, it’s made by Medtronic, it’s their 4th generation sensor, the Elite 3 sensor, which is not available on other products yet, but that senses glucose every 5 minutes. It’s sending that glucose value over to their insulin pump and their insulin pump again is a redesign pump with slightly new user-interface, it’s pretty slick and pretty easy and inside this pump, it’s called the 670 G pump, there is a software program that looks at these glucose values and adjusts the rate of insulin infusion if your sugars are going up, it ramps up the infusion rate, if you sugars are going down it decreases the infusion rate. Overnight it takes over and just tries to bring you down to a glucose of 120. That’s the default set, try to keep the glucoses at 120. Now the word hybrid means that the patient, the individual with diabetes, still has to interact with it, when it comes into mealtime. They have to enter into their pump, how many carbs am I eating, because it then gives a bolus to get a little head start. The program on the pump is pretty smart, but when you start eating, when you’re catching up to all those carbs and the blood sugar’s heading up, the pump has a hard time rapidly going up, but if you give a bolus before the meal, then the closed-loop portion can take over and start to refine your guess on how much insulin you thought you needed for that meal, it can a little less, it can give a little more and in between meals, it works great. Overnight it works great.

Steve Freed:  So what were the endpoints in what you were doing?

Dr. Bergenstal:  So we did the first, what’s called the pivotal trials. The FDA said if we’re going to approve this, you’ve got to show us it’s safe and effective, but particularly safety the FDA’s concerned about. So we did, we were part of a trial with 10 centers in the world, 9 in the US, 1 in Europe. We did 124 patients. Our center turned out to have the most patients in the study. We had 15 and they were individuals from age 14 to there was no upper limit, but it turned out to be 14 to 75. We took people with Type 1 diabetes. They already had to know how to use a pump. Some knew how to use CGM and some didn’t. We did a run in for a few weeks and got where their baseline condition was and then we went for 3 months on the closed loop and we said, how does their A1C change, how does their timing range, the number of blood sugar readings between 70 and 180, the number of low readings the number of high readings. Of course, since we went for 3 months, we could do A1C. So the baseline A1C was pretty good, it was 7.4 on average. It went down to 6.8, 6.9, but it not only went down but the amount of hypoglycemia, so the values under 70, or if you take 60, or if you take 50, they all went down by as much as 20, 30, or 40%. So, improved A1C, less hypoglycemia, and that said this is pretty effective. This looks pretty good. Was it safe? We had no severe hypoglycemia, and we had no ketoacidosis, which were two parameters that the FDA said, if you had dramatic lows or dramatic highs, then we’re worried about this system. We had none of those, so it met the safety criteria and in this population it looked pretty effective.

Steve Freed: So, you had 124 patients total in the number of sites. Did you have instance where any of the patients needed emergency medical assistance medical assistance?

Dr. Bergenstal:  No one needed emergency medical assistance related to the device. There were couple of people who got bad colds and flus and went to the doctor, or a heart condition.  But nothing related to the medical device, which was remarkable, which was good. Now, there were times, remember the hybrid closed-loop system a sensor a pump, and a pump that delivers by an infusion site. There are still the possibility, no matter how sophisticated your program is, and your pump and your technology, it’s still delivered by an infusion catheter, so if that catheter gets blocked, gets twisted, gets kinked, pulls out, you can have a high blood sugar, and the pump will be pumping in more insulin, but it’s not going anywhere. So we had some instances of some high blood sugars that even reached 300, but you could almost predict, check your site, and the site was bad and they’d have to plug in a new site and let the program take over again.

Steve Freed: This study was more of a safety study, to make sure that it was safe. Why wasn’t it designed as a randomized controlled trial?

Dr. Bergenstal: I think it was thought that what’s the best way we can get this to people, we’ve been studying these closed loop systems now in academic centers and this system had been studied to show what it could do in small sets of patients so the notion here was, let’s do a before and after study, show that it’s safe and show what it can do. The FDA would say that’s what we need to get it approved, now post-approval, we’re going to go back and do longer term outcome studies in different age groups, pediatrics, and the elderly, those with hypoglycemic unawareness. So there’s a whole series of outcome studies that have to be done, but this was what was necessary for approval, show it’s safe and show that it’s effective but you don’t need a control group, you don’t need a comparator. We’ll take these people who’ve had diabetes for 27 years and in your 10 practices, in a sense, as your baseline group.

Steve Freed: So you had 124 patients. How long did they wear the pump? How long was the study?

Dr. Bergenstal: Three months.

Steve Freed:  So you’re talking about 90 days, 124 patients, and the FDA approved it because it’s going to be available for patients this spring. My big question is: it’s one thing to have 124 patients wearing this, but let’s move that number up to 5,000 patients wearing this over a year. Is there a current concern for the safety? Because nothing in life is perfect and if it’s mechanical it certainly can fail in some way or another that we haven’t even thought of. So, if I’m going to be connected to a machine that can actually kill me, technically, I want to know that it’s 100% safe and how do you overcome that with patients? Because that’s going to be a big question for people.

Dr. Bergenstal:  I think, it’s a really good question. I think you look back and say now what are we using here. Right now, patients are using both of these elements, the pump and the sensor and they’re trying to do their best and their guessing how much insulin to tell it to give and their setting their rates, their basal rates on the pump. What this system does is really just use those same features but it adds an additional feature, a little more or a little less depending on the trends of the glucose. If the program, if you jump out of closed-loop for some reason, for instance, let’s say your sensor comes out, so you don’t have the sensor signal, then the pump just reverts back to its usual basal rates that you were using prior to starting the study or however you set it. So there is sort of a failsafe that if the program isn’t working for some reason, if it doesn’t have a signal getting to it, it reverts back to the usual basal rates. I think compared to where people were, all this adds is an additional safety feature of shutting off that insulin if you’re low, giving a little more when you’re high. It does it in a pretty conservative manner, to just keep you settling back not to a blood sugar of 80 or 90 but to 120 as the set point.

Steve Freed:  One of the interesting things in science right now. They had a whole TV program on it last Sunday. I think it was on 60 minutes about artificial intelligence. Does that come into play here? Does it learn from the person’s physical activity and the type of the foods they eat and actually make adjustments on its own? Or is it strictly the numbers that you put into it and that’s what it works off of?

Dr. Bergenstal: Really good point, I should have mentioned that. Yes, there is some learning that goes on. What it’s learning and it actually gets updated every 24 hours. What it’s learning mostly is what’s the phase that you’re in right now of how much insulin that I see that I’m needing to deliver? In a sense, it learns your insulin sensitivity, it learns how much the insulin, a given number of units is lowering your blood sugar in any instance, so it’s constantly updating. Let’s say you get the flu and you’re just insulin resistant and you got all this stress that’s going on, it can sense that and say “boy, he’s needing more insulin, we’re going to ramp up the program a little bit. Oh, he’s getting better, I’m going to ramp it back a little bit.” So it’s learning that kind of individualizing for each person. Their total insulin needs and a little bit around the time of day. It’s not quite specific to say the last time you exercised you got this many units. If you want to talk about exercise I can tell you how we deal with that too.

Steve Freed: I remember, I wore a GlucoWatch, when it first came out. I had to keep calibrating it and if I sweat it would shut off. How much have we changed the technology so sweating doesn’t shut it off and we don’t get an error. According to what I’ve read, the CGM from Medtronic, is about 9% sensitive for basically that you have a 91% accuracy. So that obviously over time, those things improve. The first model cars are different from what we’re driving today and so the GlucoWatch was the first. This would be along those same lines. Where do you see it going eventually because I would think the next thing, you still have to worry about your boluses and your carb ratios. I would assume if there are a couple of these pumps out there. I know there’s one that uses two enzymes, or two hormones. Glucagon and insulin so that if you are going too low or too high, you got two different variables there. I just wanted to get your opinion on what you think of that system.

Dr. Bergenstal: Let me go back to your first part of the question and I’ll come to the dual hormone one in just a second. I completely agree this hybrid closed-loop, the patient has to interact. It’s first generation. It’s come a long way, so the sensors have gotten so much more accurate compared to where they were initially. Initially, when we had a sensor, we’d say do I want that sensor to be driving my insulin pump? But now they’ve gotten much better and we did some careful MARD, these Mean Absolute Relative Difference. It’s right around 10% for this system. So, every time you improve the sensor, a little bit that makes the system even more accurate. So I see that coming. Let’s say you had a faster acting insulin. That was put into the system. All the sudden the meals would be that much better. Let’s say you had a better infusion set that didn’t block up or kink and you probably know those are coming, which would make the system better. Let’s say you decide to get a little more aggressive with your programming and say I want to give a little more insulin faster when I’m going up, instead of first generation that was kind of conservative so it didn’t give too much. So every mathematics, infusion sets, faster insulins, little better sensor are all going to make the current insulin only system better. So I see generation 2, 3, 4 just getting better and better but all based on these same principles. Now the point you asked about the two hormones, wouldn’t it be nice to have insulin to keep the blood sugar from going too high, but if you were going too low, too fast, squirt in a little glucagon, which raises the blood sugar as a safety catch, that’s what a group in Boston and Canada and Seattle, different groups are working on two hormones in the pump. I think it’s exciting, we’ll have to see if it’s better, but they have to first do some studies to prove that infusing glucagon over the long-term is safe. We know when you use it in the short term to rescue you from a low, it does fine. There doesn’t seem to be any adverse problems from that. It does just what it’s intended. Nobody’s infused it for a year. They are undertaking some long-term studies that show that infused glucagon is safe for you and in the meantime, they’re designing up a pump with two little ports, with two syringes. I think that will be another option a few years from now possibly. But we’ll have the insulin only systems to start off this new adventure.

Steve Freed: It’s interesting, if you think back 1950 to 1995 we had one oral drug, sulfonylurea. It’s not the best drug in the world, but that’s what we had, that’s what the physician used. From 1995 through today, we’ve got over 11,000 possible combinations of drugs and now it’s actually confusing for the physicians, because he’s not sure which one is going to work and now we do trial-and-error. The science has actually exploded which actually I think it’s making it more difficult for physicians, because when there was only one drug, it was kind of simple, stupid. Now we have all these triple combinations, dual combinations, and over the next couple of years, we’ll have all these insulins to add that to the equation. So, everything, we like to think it’s getting simpler but actually I think it’s becoming more difficult with all the new science. What are your thoughts on that?

Dr. Bergenstal: I agree completely. It takes now, when you have these options, how do you individualize it for a given person and figure out what’s right for them. Steve, I’m predicting that we’re going to see more and more use of this continuous glucose monitoring, even in Type 2, even if it’s maybe done intermittently. To sort out your profile. We’re learning how these drugs work, which ones work postprandial, which ones work between meals, which work on the meals. I’m predicting we’re going to be profiling individuals and saying, here’s what my profile looks like, which of these medicines would be best for me. Or I’m at a particularly high cardiovascular risk, so I’m reaching first for one that has a cardiovascular improvement, then I’m going to look at my glucose profile and pick the medicine to match me. I think we are going to get to that level, just like we’re doing with this artificial pancreas, or this hybrid closed-loop. Then, to your point, I think the days of poking your finger, we’re going to be using CGM in place of getting red dots all over your fingers to check your blood sugar.

Steve Freed:  How often do we have to calibrate this?

Dr. Bergenstal:  It’s twice a day it’s calibrated. People will do it occasionally at other times, when their blood sugars are stable and they’re checking before a meal for instance, you could add in another calibration, but if you want to just do it a couple times a day, that’s all it takes.

Steve Freed:  You see a lot of patients. What kind of patients would you recommend this to?

Dr. Bergenstal: I think the ones, and this will be up to the company how they release this, but I think we are going to start off with people who are comfortable with pumps and maybe even those who are comfortable with pumps and sensors. That education component is removed a bit and then you teach them the new concepts with this. There are some new concepts. A little bit of letting go.  A lot of patients right now check their blood sugar or look at their sensor all the time and take a couple extra units if it’s going a little high. In this case you’ve got to step back, let the program work. So I’m going to take some people who are 1) comfortable with technology, comfortable with pumps and start with those and then I’ll eventually start working my way back to say, you want to learn a sensor and you want to learn a pump, let’s do the sensor first in my opinion and get you used to your glucose data, let’s put in a pump, now let’s start a closed loop system after that.

Steve Freed: What were your patients’ response and family members to this device? From all the different people. Were they allowed to keep it if they wanted to keep it? Did they want to get rid of it?

Dr. Bergenstal:  It’s amazing. So this was a 3 month study, right? Even within one month into the study the patients were lobbying the FDA to say “don’t take this away, this is the first time I’ve ever been able to go to bed at night and have a smooth blood sugar.” We heard it from the parents of the 14 year olds and 17 year olds. They’ve been getting up every night for the last 10 years since their child was diagnosed at 4, at age 7, checking their blood sugar a couple times a night and now they just know with this system, their sugar’s just come down and start off good every morning. The parents can sleep, the patients can sleep, although a lot of them were sleeping right through dangerous lows anyway. The parents, they petitioned the FDA to continue. The FDA talked with Medtronic and they said “yeah, we’re happy to do an extension study”. So the patients have continued on it now. They’ve been on it now, some of them for over a year and 80%+ of the patients continued on with the study. Most of them have really gotten used to it and said “wow, I’m not sure what I’m going to do if somebody were to take this away, because I’m finally seeing smoother, flatter blood sugars.”

Steve Freed:  When does the next study start for the younger age group, and obviously the older. There’s a lot of parents out there with kids that are 5, 6, 7, 8, 9, 10, 11, 12 that really would be looking forward to this and they don’t want to wait another 5 years.

Dr. Bergenstal:  You’re right. The pediatric group are the people using pumps the most in the country now. Our pediatric clinic, Park Nicollet, with the International Diabetes Center, the pediatric clinic has 60-70% of their kids on pumps. In most adult clinics it’s 40%. Right now, the pediatric study has started. So as I said, this one is approved by the FDA for 14 and up. The age 6-13 study is underway right now, with this device. Younger than 6 is going to have to planned and be done. That’s underway, but the initial approval will be 14 on. I don’t think there’s an upper limit. As I said, we went up to 75. Sometime in 2017, it’s going to be ready. It’s up to Medtronic when they decide to actually market it. To put it out there. I know they want to get all their educational materials and get everything ready, so as you said, people can get an efficient, effective, walk through the steps.

Steve Freed:  First they have to educate the educators and then they can educate the patients. One of my last few questions is how much education was necessary in trial for someone that wore a pump before and for someone never wore a pump before?

Dr. Bergenstal:  I can answer how much was required. So, in this study, everybody wore a pump before, so they all were pump users but they all weren’t CGM users, so some of them had to be educated on the CGM part. But they were all pump users in this study. The education really was not that extensive if they knew how to use a pump. It was almost I won’t say “unlearning” but it was saying you know how your pump works now with your various basal settings, let’s get those right for these first couple weeks so we are comfortable that these are good basal settings for you, because maybe you’re a new patient to me and I just want to check them out. But then, we let the program work and it’s really teaching them a little bit more about take your insulin, take it before the meals, let it work, it might drift up a little bit, but let it work so it will learn how much insulin you need and let it bring you back down. So, it’s a little more, if they’ve been pump users saying we know how you’ve done it in the past but here’s how we want you to try with this system to let it learn, let it work. So I think it’s very doable, but it would be nice to have the steps all laid out of the teaching points and the tricks and tips that we learned.

Steve Freed: You set the basic at 120 mg per deciliter. That was your border that’s what you wanted to accomplish. Obviously you accomplished that. As you feel more secure with the whole mechanics and new technology do you foresee that we can actually lower that to something maybe a little bit lower than that, maybe 110 or even 100?

Dr. Bergenstal:  Yes, that’s a really good point Steve. I think it’s good thinking. That would be a next generation. So in this particular pump, you cannot vary it. It’s set. 120 is the set point. Eventually, when people are comfortable that that didn’t cause too many lows. That got the right amount of reduction of highs, I would guess further versions going forward will allow that to be reset to different numbers. But the current one only has one number. Except, there is one other variation you can make with this that I think is worth mentioning, because you mentioned about exercise. Most people know when you exercise, you’ve got to be careful because it can do two things.  It can lower your blood sugar pretty darn quickly because you are burning up all that glucose and moving glucose into cells. So there is a switch on this that you can say I’m going to exercise. I’m going to reset the target to 150. So there is an exercise setting of 150. You probably know, Steve, with a pump now, if you have a pump and you are about to exercise, you can decrease the insulin rate for 2 hours or for 4 hours during or after exercise. In this case we do it just the opposite. We don’t change the insulin rate. We change the target up a little bit to 150. Then let the program work during exercise not to keep you so tight that you risk going low. That’s seems to work very well.

Steve Freed: That’s really a benefit. It takes out all the complicated math out of the equation. You exercise, you press a button and it takes care of itself. For my last question, because I don’t want to keep you too long. There was another product that was approved by the FDA just recently and supposedly will be available this coming spring, that’s Abbott’s Libre which has been in Europe and I have been told that you can’t keep it on the shelf long enough. When you were talking about replacing the finger sticks, even though the first amount is going to be professional. So you stick it on a patient for a while, they come in you get the readings, you sit down with them, you look at their food log and you can make all kinds of recommendations. Eventually, they’ll have one for patient use. What are your thoughts about that system? Have you even seen it?

Dr. Bergenstal:  I have not only seen it. We’ve used it. I think we’ve been the only US research centers. We’ve used it in a research study that we’re going to be publishing pretty soon through the Type 1 diabetes exchange. We actually used it in people to check out how the A1C correlates to the average glucose and such. I think it’s going to be an amazing edition to our material. Even a professional version is really nice to put something on that lasts for two weeks, that’s factory calibrated and gives you a nice picture. We’re pleased that with the Libre, the output from the Libre is something called an ambulatory glucose profile, just a way to display the data. The International Diabetes Center has helped work on that display. I think there’s going to be a great learning tool for people to learn where their blood sugars really are. How it affects their meals, how it affects overnight. I think it’s great to have one more addition to our tools.

Steve Freed:  Is all this new information in science making your life easier or harder?

Dr. Bergenstal: It’s a good question. We’ve been waiting so long for this and the information is so amazing that I’ve got to embrace it even though we’re all struggling to say how much time I can spend looking at all this data. Is it reimbursable the time that we’re spending? Most of us are just doing it because it really helps and it’s good for the patient so we do it. I think we’re also lobbying to say, some of this monitoring should be reimbursable because a lot of this we can do remotely. A lot of people come in for visits to have their blood sugars looked at. I can sit here just like we’re doing today and I can look at someone’s blood sugars and so we’re working on codes for remote monitoring to save patients so they don’t have to take off of work to have their blood sugars looked at.

Steve Freed:  The irony of all this science and technology is they’re working very hard for a cure, especially for Type 1 diabetes. We keep getting closer and closer and they’re working on a vaccine nowadays. It’s hard to believe that all this technology will be useless in the possible years, if they come up with a possible cure for Type 1. There’s 2 elements to diabetes, especially Type 1, and that’s a cure and newer technologies and newer drugs. Where do you think that the technology is actually going to happen?

Dr. Bergenstal: I don’t know. You’re making a really good point. It’s one of those things that if I could put myself out of business, so be it. There’s other things to do and we’ll be working. For the last 15 or 20 years, this has been sort of a race. Islet cells versus mechanical and they’ve been going hand-in-hand and right now the mechanical approach to, I don’t know if you want to call it a cure, but the mechanical approach to control has taken the lead, but the islet cell people are really close. If they come up with something and it’s available to the masses and this starts to get less I think that would be fine. Right now it’s limited because you have to take immuno-suppressants, by in large, until they get one more breakthrough. I think let the competition come. There will be more people making artificial pancreas or closed-loop devices and they’ll push each other. I think that’s terrific. There will be islet cell work going and that will keep everybody on their toes.

Steve Freed: Our goal for DiabetesinControl.com is to better educate physicians, because every physician we can touch is worth 5,000 patients. That’s our philosophy and giving them the information in a quick precise manner that they can use and I think you get our newsletter also.

Dr. Bergenstal: I do.

Steve Freed: Again I want to thank you for your time.