The Sansum Diabetes Research Institute has been focusing on research using an artificial pancreas in collaboration with the ultra-rapid inhaled insulin to attempt to gain tighter control when people with diabetes eat.

By: John Parkinson, Clinical Content Coordinator,

The goal of the artificial pancreas (AP) is to develop a fully-automated system to help improve peoples’ glucose control on a daily, moment-to-moment basis, keeping people in a tighter range so they don’t suffer either acute hypoglycemic episodes or complications from years of hyperglycemic numbers. And while researchers are improving upon this process, one area that still presents a challenge is during mealtimes.

Even in a fully-automated AP system, using subcutaneous bolus insulin causes inherent delays for the insulin to get absorbed into the body and begin working. This can cause mealtime spikes that can throw off peoples’ numbers.

Understanding these delays with subcutaneous insulin, Howard Zisser, MD, director of research and technology at the Sansum Diabetes Research Institute, in collaboration with the University of California, Santa Barbara (UCSB), decided to see if there was a way to decrease the postprandial glucose spike by introducing ultra-rapid inhaled insulin in collaboration with a artificial pancreas in a clinical study.

Zisser (pictured, lower right) and colleagues used the Dexcom Seven Plus CGM System, the Omni-Pod insulin pump, a computer-based program controlling insulin delivery, and MannKind’s Afrezza inhalable insulin for the study.  

The study was funded by JDRF, and the results were promising according to both JDRF and Zisser. JDRF’s Aaron Kowalski, PhD, vice president of treatment therapies, stated, “the early results of this exciting study are compelling. While more data are required, this protocol represents a potentially revolutionary way of combining the artificial pancreas with a simple-to-use inhaled insulin to significantly improve blood sugar levels. spoke to Zisser about the study’s results, his characterization about the progress of the AP development, and how this research is leading to serendipitous discoveries that will eventually be used to improve upon diabetes management. Can you provide an overview of how the artificial pancreas and rapid-acting inhaled insulin trial worked?

Zisser: To give you a little background first, what we are trying to do with the artificial pancreas system is automate it in a way that we can eliminate the excursions of glucose. We want to get rid of the high and low blood sugars. In essence, compress that excursion into a narrower band so you don’t have any acute problems with acute hypoglycemia, or have any long-term complications from hyperglycemia.

Initially, we thought we could have very tight control. However, the whole system is sluggish. The first sensors we used were not as accurate as we would have liked them to be. And when you inject subcutaneous insulin there are some time delays. It’s not a system set up where you see something coming at you and you can make a reaction and the glucose comes back right away.

Imagine if you are driving a car and you see a traffic light, but you can’ tell if the light is red or green—that is the sensor part of things. You put your foot on the brake, but the brakes don’t kick in for another five minutes—that’s the subcutaneous insulin. Imagine if you could see the light clearly and you could get an immediate reaction and result when you put your foot on the brakes?

Because we are always making these micro adjustments in managing diabetes, we decided to go to a concept called control to range. We really wanted to get rid of the highs and the lows to get people into a normal range throughout the day. It would be mild excursion after meals, but after a certain amount of time you would be in this optimal range.

For this inhaled study, we wanted to see if we could get the insulin to act a little quicker so the subcutaneous insulin doesn’t have to work as hard. The inhaled insulin gets into the circulatory system very quickly. It mimics first phase secretion of insulin. For example, if you go into a bakery and you don’t have diabetes, you smell this food, and your body thinks it’s getting ready to eat so it secretes a little bit of insulin into your body. If you don’t eat, your body’s glucagon kicks in and everything is fine. We are trying to get it so that right before the meal, we can get insulin working in the body. We are not trying to cover the whole meal; we are trying to get a system in place so that the patient knows he or she is eating and will start storing sugar instead of releasing the sugar. Is the theoretical idea of how this would work is that the patient would rely on the AP outside of mealtimes, then suspend the AP usage at meals and manually takes the inhaled insulin?

Zisser: I tend to think of the two working as a hybrid system. With the fully automated AP system, the insulin is responding to the glucose needs throughout the day, and right before a meal, we will give them a small dose of inhaled insulin that the AP system doesn’t even know about.

What we expected to see was that the postprandial slope after a meal was not as steep, and the subcutaneous insulin delivery system doesn’t have to work as hard. As the glucose won’t go up as fast, we don’t have to use as much subcutaneous insulin to control the meal. It sounds like they are not working exclusively outside of each other. You are still introducing the inhaled insulin even when the pump is working, so you are not turning anything off?

Zisser: Yes. So the inhaled insulin takes away the large glucose swings of the highs and lows?

Zisser: To lower the postprandial peak so you are not going up quite as high. You have said the preliminary results of this trial are promising. Can you explain your characterization?

Zisser: I can’t go into too much detail because it is unpublished data, but what we were seeing in the trial is what we expected. The inhaled insulin will lower the postprandial excursions.

Here is a graph showing the difference in a study participant`s postprandial glucose using inhaled insulin and without the use of it. How did patients respond to the rapid-acting inhaled insulin?

Zisser: Overall, great. It was the first time they were using it, and they just had to learn how to use the device, which took just a couple of minutes to learn.  And, there were no side effects or complications. What are the remaining challenges for using an AP in collaboration with inhaled insulin?

Zisser: We gave everyone just a very small priming dose, as opposed to a dose that would be used for a meal. We could probably optimize that a little bit better based on some modeling, but I think it works just as well as is. Are you looking to do follow-up trials with this as well?

Zisser: Yes, we have ongoing AP trials as well as looking to start new ones, and the best thing to do is go to our website to find out about how to participate.  

[Editor’s note: For anyone interested in possibly participating in research, go here.] Can you tell us about other AP-related research Sansum is doing?

Zisser: We have some new work going on where we are moving to outpatient trials. We are getting to the point where our sensors are more accurate, and patients systems are more reliable. So we are taking patients basically from the clinical research center and out into the real world with systems. It’s liberating for both the clinical staff and patients.  Of course, we are doing this in a judicious way, and most of the groups we are treating in these outpatient studies are like camps. How would you characterize these various aspects of diabetes care you have been studying right now?

Zisser: The inhaled insulin gives us another arrow in our quiver—so to speak—to manage diabetes. I think it will be a great drug, especially for type 2 diabetes. Glucose monitoring has gotten much better and people can see how their decisions and activities affect their glucose in real time. Not only the value, but what direction it’s heading in, and with this learning, make better decisions in their care.

A fully-automated artificial pancreas is something we hope will be a reality, but before we get there other discoveries will be made. It’s like the space program in the 1960s. The goal was to get to the moon, but along the way, we got Kevlar, space blankets, Tang, and other things we would have never thought about before.

I think one example of this in diabetes research is remote monitoring. We are going to be able to send all this information anywhere we want to. We have started making these systems for the artificial pancreas where we are importing the CGM signal into a computer. Once we realized that we had a AP system, we could use this in our clinical trials. If the patient is getting ready to go low, we can send a message to the patient, to the researchers, and the clinical staff, so they all know—no matter where they are. Eventually, the parent of a child with diabetes will be able to see what’s happening with the child’s glucose and his or her insulin delivery while the child is in school. I think it’s one of the Kevlar, or pleasant discoveries. We really never thought about it, but once we started building these systems and monitoring people, we realized we could do bigger things.