Skip to Main Content
Research & Innovation, Family Health

An 'Artificial Pancreas' Eases Burden for Type 1 Diabetics


Research and clinical trials at Yale helped Medtronic develop its new personal diabetes management device.

Hundreds of mothers have told Yale Medicine’s Stuart Weinzimer, MD, that they don’t sleep at night. They wake frequently, frantically checking the blood sugar levels of their children with juvenile (type 1) diabetes to make sure they haven’t dropped low enough to cause seizures, unconsciousness—or in rare cases—death.

This whole part of my brain that was devoted to thinking about my diabetes all the time suddenly got freed up to do other things.

Diabetes patient

That’s why some parents cried with relief when Dr. Weinzimer, one of several Yale Medicine physicians at the forefront of diabetes research and treatment internationally, presented information about a new device under development at a national JDRF (formerly known as the Juvenile Diabetes Research Foundation) summit back in 2006. They knew that someday this new piece of technology would better control their children’s blood-sugar levels and release them from those nighttime fears.

Medtronic’s MiniMed 670G system is now on the market. The device, which was tested in clinical trials at Yale, continuously monitors blood glucose levels and adjusts insulin delivery automatically based on those levels. In quality-of-life measures, it’s a game changer, especially for those parents who weren’t sleeping. Seventy-five percent of life-threatening low blood sugar reactions happen overnight.

Until now, people with juvenile diabetes­—which affects more than 1 million Americans—have worn sensors and pumps, but the two have never worked automatically together. This device is often called an “artificial pancreas” because it mimics the glucose-regulating system of a healthy pancreas.

Rachel Hicks, a participant in a clinical trial of a new diabetes management device, tries out a climbing wall. The device doesn't interfere with most activities.

The device integrates the two separate jobs of measuring glucose and delivering insulin into one regulated self-contained system. “To be able to deliver insulin automatically is a tremendous advance in clinical care,” says Dr. Weinzimer, a professor of pediatrics (endocrinology) at Yale School of Medicine. “A significant portion of our population will do amazingly well with this. For them, we will be eliminating or reducing diabetes-related complications.”

While the MiniMed 670G is a major step in improving care for people with diabetes, it is not a panacea. For example, blood glucose levels still need to be checked several times a day, such as before meals or when calibrating the sensor. Dr. Weinzimer calls the device, “evolutionary, not revolutionary,” adding, “It’s important that people understand this is an automated device, but it’s not fully automated. We have to manage expectations.” Medical science, he says, still has a long way to go to develop a cure for diabetes.

Other researchers didn’t want to explore an artificial pancreas because it wasn’t transformative enough, Dr. Weinzimer says. “But our philosophy is that if we can improve our patients’ health and reduce the burden of care, they’ll be in better shape when a cure comes along,” he says.

The system was approved by the U.S. Food and Drug Administration in 2016 for adults and children over the age of 14. At least four other trials are underway for systems with slightly different components, including ones that can be used by children as young as 6. Yale is participating in two of these trials.

Dr. Weinzimer continues to participate in ongoing clinical trials related to diabetes management. We caught up with him to have him explain what this device on the market means for people with juvenile diabetes.

How does this device work?

The MiniMed 670G consists of an insulin pump, glucose sensor, transmitter and a glucose meter. The sensor is inserted just beneath the skin’s surface in the abdomen, and continuously measures glucose levels in tissue fluid. The lightweight transmitter attaches to the sensor. Every five minutes the transmitter wirelessly sends data to a small insulin pump, which is worn on a belt under clothing or in a pocket. The pump delivers insulin, as needed, via a small catheter.

Produced in the pancreas, insulin is a hormone needed to move blood sugar, or glucose, into the body’s cells, where it is stored and used for energy. For people with type 1 diabetes, the body produces little or no insulin. This means glucose builds up in the bloodstream and the body cannot use the glucose for energy.

Is any surgery required to use the device?

No, patients can insert the sensor themselves with a spring-loaded device. You pop in the needle and pull it out. The thin sensor filament stays in place. The data transmitter, a little radio shaped like a clamshell and a little bigger than a thumbnail, sticks out. To shower or swim with it, you can tape it down.

The sensor should be changed about once a week.

How does this differ from previous medical devices for people with type 1 diabetes?

Two of its main pieces—the continuous glucose monitor and the pump—are existing medical devices. Sensors have been used for about 10 years, and insulin pumps have been around for decades. But this is the first time they have combined into an integrated system that enables them to communicate with each other to dynamically adjust the delivery of insulin.

How will this device change the lives of diabetic patients?

In this study we found that they were spending much less time thinking about their diabetes. They were checking their blood sugar levels before meals, entering their carbs into the pump, and then going on with their day. They weren't worrying about whether the dose was right, or whether they would drop suddenly at night. People often get up several times a night to check their sugar, take more insulin if needed or eat a snack if they are low. That just didn’t happen anymore. People were sleeping better, their parents and spouses were sleeping better and, even with less thinking about diabetes, everyone’s blood sugar levels still got better. One person told us: “This whole part of my brain that was devoted to thinking about my diabetes all the time suddenly got freed up to do other things.”  

What are the limitations of the device?

You still have to test your blood sugar at mealtime and enter the information into the pump, as well as track the amount of carbohydrates you are eating, as people with diabetes are already taught to do. If what you put in is too much or too little, the system will adjust to give you more or less insulin, based on whether the glucose levels are going up or down. It’s like driving a car on auto pilot. You still have to steer and you still have to hit the brakes if another car pulls out in front of you, but the car maintains your speed pretty well as you go up and down hills. I also tell patients it’s like a safety net: The system is always working in the background while you are awake or asleep, to help keep the glucose levels in a safe range.

Another issue has been that the sensors have to be calibrated, and sometimes don’t last all week. Since this is the first of these systems on the market, it was designed to be conservative. So the glucose levels come down too slowly for some people. There can be malfunction issues, like with any device, and it goes through batteries very quickly.

Some people may be disappointed in it, which goes back to managing expectations. You have to under-promise and over-deliver. After about a month, people get the hang of using it and learn how to make it work best for them.

Are there any side effects?  

The only real side effects are those from having diabetes. Blood sugars can still go too high or too low, even if you are using the system correctly. In our recent trial, though there were no serious or unexpected side effects.

Does insurance cover it?

Most of the time, insurance companies cover pumps and sensors, and I wouldn’t expect it to be any different with this system. The device is considered durable medical equipment. Coverage will likely hinge on a doctor writing a letter of medical necessity for it. The insurance company will request documentation of medical necessity, which is pretty standard

Why is this device only for people age 14 and over?

We rolled the study out to adults first, which makes sense because kids are our most vulnerable population. However, kids stand to benefit the most. Their burden of living with the disease and its complications is greater than for adults. We are now testing devices for kids as young as age 2, but this will likely take another few years.

Meanwhile, we are part of a large-scale study examining the effects of diabetes on the developing brain, starting with children diagnosed with diabetes at a young age and following them through puberty. We want to see how cognition and structure of the brain are affected by diabetes. The rationale is that, historically, clinicians have been primarily concerned about low blood sugar levels, and have set very conservative blood sugar targets. As a consequence, many children have had too much exposure to high blood sugar levels, which, as we are learning, may also be bad for the brain.

What was Yale’s role with the development of this device?

Yale has long been a pioneer in the field of diabetes care and research. The first insulin pump was tested here in 1979 during a clinical trial. Yale Medicine’s William Tamborlane, MD, was on that trailblazing team.

Dr. Tamborlane recruited me to Yale in 2002 and, along with industry partners and researchers at other academic centers, we began working on a project to develop the artificial pancreas, which drew support from the JDRF. A number of studies and clinical trials at Yale and other institutions followed, paving the way for last week’s device debut.

What’s next in terms of such technology?

This has been like the race to the moon, and Medtronic got there first. Several other academic institutions are also working on hybrid closed-loop systems, similar to the Medtronic device—but with slightly different components. One institution is working on something called the bionic pancreas, which uses two pumps and two hormones. These are all in trials in one stage or another.  And we are working with the School of Engineering at Harvard on a system designed for very young kids.

Five years from now, I think most doctors will be able to say 80 percent of their patients are on an automated insulin delivery system, one that will be incrementally better than what we have today. Maybe you won’t have to do the mealtime calculations, and the wearable devices will be smaller. That is what the landscape will look like. Next will be biological replacement of the patient’s stem cells or islet cells, which are found in the pancreas. We would transplant the cells into the body. These transplants may require some degree of immune-suppressing medications to prevent rejection, or people may need to have repeat transplants every so often to replace these cells.

Could this device work for people with type 2 diabetes?

Theoretically, anyone who requires insulin should benefit from this technology, but the insulin dosing calculations will be different. I suspect we have a way to go before we have something like this for type 2.

To learn about diabetes clinical trials at Yale, click here.