Mass General Leads Pivotal Trial of Bionic Pancreas
In This Article
- Massachusetts General Hospital was instrumental in developing and investigating a medical device that uses next-generation algorithm technology to control diabetes
- A new study shows that the novel iLet® Bionic Pancreas simplifies diabetes management and improves glucose control in a wide variety of patient populations
- The device requires very little input from healthcare providers and patients, replacing complicated calculations and carbohydrate counting
- The lead investigator believes the device has the potential to markedly improve A1c control in populations not meeting goals for therapy and overcome challenges related to a scarcity of diabetes specialists
- The FDA is currently reviewing the device for approval
A multicenter trial led by Massachusetts General Hospital may soon lead to the approval of a novel insulin delivery system using a special algorithm to automate and simplify diabetes management.
Subscribe to the latest updates from Endocrinology Advances in Motion
The iLet® bionic pancreas could be a game-changer in diabetes care, says lead author Steven Russell, MD, PhD, an endocrinologist in the Diabetes Unit at Mass General. He and his colleagues worked with researchers at Boston University to develop the technology, which was then licensed out to a company named Beta Bionics that manufactures the device.*
"It's really fundamentally different—much less is demanded of the provider, much less is demanded of the user than with any of the other artificial pancreas or hybrid closed-loop systems currently on the market," says Dr. Russell. "The goals are to expand access, improve glucose control and reduce the burden of diabetes. I can't wait until it gets approved by the FDA and we can offer it to people living with diabetes, and get the word out to their primary care providers."
Diabetes Control Is a National Problem
Fewer than 20% of patients with diabetes in the United States meet the treatment goal of less than 7% glycated hemoglobin. That means that 80% of people with diabetes are at risk for serious complications, Dr. Russell stresses.
Part of the reason is that calculations are complicated, and existing diabetes management devices require significant input from patients and providers. A specialist must make the initial calculations and conduct regular follow-up, while the patient must count carbohydrates multiple times a day, which is burdensome and leaves room for error.
"It's really hard to figure out all the insulin doses you should take on a minute-to-minute, hour-to-hour basis. And even as we've gotten better tools like continuous glucose monitoring and even partially automated insulin-delivery devices, the number of patients meeting treatment goals has stayed stubbornly low," Dr. Russell says.
Another compelling reason for a more streamlined device is the lack of diabetes specialists nationwide. Current insulin devices must be set up by an endocrinologist or diabetologist who programs the device to account for basal and bolus insulin delivery, carbohydrate ratios, correction factors, and glucose targets.
The specialist also conducts regular follow-up testing and adjusts insulin regimens. But there's a scarcity of specialists throughout most of the country, and many primary care physicians (PCPs) don't feel comfortable taking on that role, Dr. Russell says.
In contrast, the iLet device uses a novel algorithm to determine insulin delivery. The system is started using just the person's body weight. The user is required only to make a rough estimate of carbohydrates consumed during each meal. Dr. Russell hopes that PCPs will feel comfortable prescribing and using the new system.
The iLet Bionic Pancreas: No Complicated Calculations
The bionic pancreas is designed to eliminate the complicated calculations and need for specialist assistance that make it hard for many people to control their blood glucose. "It seemed to us that we needed a better tool that did things a little bit differently," says Dr. Russell. "We wanted to create a system that doesn't rely so heavily on the person having to pay attention to their diabetes."
The device is about the size of a credit card—only about 15 millimeters thick and very lightweight. It contains a tubed insulin pump and a continuous glucose monitor. Users wear the bionic pancreas clipped to their belt, in a pocket, or otherwise tucked into their clothing.
Setup is simple for the provider. "You put in the person's weight and press go. It starts with an initial dose scaling based on the weight and it immediately automates glucose management," Dr. Russell explains. "It autonomously and continuously adapts to the person's insulin needs over time."
Patient use is also much less complicated. The device's touch screen shows a little icon of a knife and fork. The user touches it to "announce" a meal—breakfast, lunch, or dinner. Then the system asks whether the meal contains the usual amount of carbs, fewer carbs, or more carbs.
The algorithm learns over time how much insulin a specific user requires, automatically adjusting to provide less or more insulin as needed. This method requires a lot less input from the user and no quantitative measurements or calculations. Even manual insulin corrections (via insulin bolus) are automated.
Clinically Significant Reduction in A1c
Dr. Russell and his colleagues recently published a pivotal trial of the bionic pancreas in the New England Journal of Medicine. The study, mostly funded by the National Institute of Diabetes and Digestive and Kidney Diseases, involved 16 sites throughout the country and was led by Mass General's Diabetes Clinical Research Center.
Study participants with type 1 diabetes who used the bionic pancreas had a rapid, clinically significant reduction in blood sugar. The device also kept A1c levels within normal ranges more effectively than standard-of-care approaches. In fact, this is the first trial to show that a device could produce what many would consider a clinically significant reduction in A1c in patients with type 1 diabetes.
Those who used the bionic pancreas had:
- Lower glycated hemoglobin level
- Lower mean glucose level
- Increase of 2.6 – 3.2 hours per day in the target glucose range
- Less time in a hyperglycemic state (without an increase in the incidence of hypoglycemia)
"The effect was very, very rapid, and then extremely stable. So although we didn't program it with any information about the person—we just put in their body weight, and we let the adaptive algorithms do its thing—it gave them a new, improved level of glucose control within 48 hours and they stayed there for the 13 weeks of study," Dr. Russell says.
Almost Everyone Is a Good Candidate
The researchers were careful to include a patient sample representative of the actual diabetes population, Dr. Russell adds. And the technology worked well in all groups, including:
- Both pediatric and adult cohorts, ages 6 to 83
- Ethnically and racially diverse patients
- Lower-income, underserved populations
- Those previously using a broad range of other kinds of diabetes technology (pump versus no pump, with or without automation, multiple daily injections, other hybrid closed loop or automated insulin delivery devices)
- Those with varied educational backgrounds
However, Dr. Russell acknowledges that some people may not be entirely comfortable giving up control to an automated device. "Imagine this device like a self-driving car, and the person in the passenger seat keeps reaching over and grabbing the wheel when they get nervous. That leads to trouble," he says. "They might take extra insulin that they don't tell the device about or tell it they ate a meal when they didn't so they can get some extra insulin. Some of the people in the study did those kinds of things," he continues. "That interferes with the adaptation of the algorithms and leads to counterproductive changes. So people who really aren't able to take their hands off the wheel are better off with another device."
Next Steps for a Breakthrough Device
The iLet bionic pancreas received the FDA's Breakthrough Device Designation in 2019. Now, this pivotal trial provides essential data for the agency to review. As the researchers await approval to market the device, they are continuing to explore and expand its utility.
A separate trial will test a bi-hormonal version of the bionic pancreas that uses both insulin and glucagon. The goal is to get even lower average glucose levels with less hypoglycemia.
Dr. Russell and colleagues are also exploring the possibility of setting bionic pancreas users up remotely. Because some of their research was conducted during the COVID-19 pandemic, some patients had to be set up via telehealth visits. That did not appear to be a factor in the device's success. This could greatly affect accessibility after the device is approved.
*At the time that this research was conducted, Dr. Russell was a paid consultant to Beta Bionics, in a role that was consistent with Mass General Brigham policies. Since this research was concluded, Dr. Russell has become Chief Medical Officer of Beta Bionics.
Explore research in the Diabetes Unit
Refer a patient to the Diabetes Unit