Artificial pancreases were a breakthrough for type 1 diabetes. The research in type 2 is just beginning

For the first time, a team of researchers has published research into how a closed-loop insulin delivery system might help patients with type 2 diabetes manage their disease at home.

The study, described in a new paper published Wednesday in Nature Medicine, included 26 patients who tested out a so-called artificial pancreas, which works by pairing a glucose sensor with an insulin pump to automatically adjust doses of the drug. The system improved the time participants spent in their target glucose range by about 15%; its users also managed to avoid bouts of severe low glucose while reducing overall glucose levels.

Those results are particularly meaningful for the pool of patients included in the study: people who need regular dialysis for kidney failure, which can dramatically impact blood glucose levels. Many of the study’s participants also saw benefits in their everyday routines to manage their conditions, said Charlotte Boughton, lead author on the paper.


“A lot of people who are on lots of different injections found it much, much easier,” said Boughton, a clinical lecturer at the University of Cambridge. That included reducing the number of diabetes-related tasks they had to complete in a day — several participants with neuropathy had trouble injecting insulin manually — and providing more peace of mind as an algorithm calculated their insulin doses.

Now, the question is how broadly the benefits extend to at-home use for type 2 diabetes patients. The systems were first approved in 2016 for the management of type 1 diabetes, which affects about 1.6 million Americans. But researchers are now exploring whether the systems could be used to treat the millions of type 2 diabetes patients who use insulin.


“There are lots of caveats for type 2, and we just need more data and more studies,” said Bruce Buckingham, a pediatric endocrinologist at Stanford University School of Medicine who conducted early trials of the systems in type 1 patients and was not involved in the research.

Researchers, including Boughton’s colleagues in Roman Hovorka’s lab at the University of Cambridge, have already demonstrated that closed-loop systems can be beneficial for type 2 patients while hospitalized. The sicker someone is, the more vulnerable they are to the harmful swings in glucose that the systems are designed to prevent.

“There’s a real role for automated insulin delivery in the hospital. It really became obvious during Covid,” said Buckingham, as higher glucose levels have been associated with more severe disease outcomes. During the pandemic, such systems also allowed providers to get glucose data and deliver insulin without going into a patient’s room.

While the new study published results on at-home use for the first time, it was just a stepping stone to the unassisted use that patients with type 1 are used to, said Boughton. Because the patients were visiting dialysis centers frequently, they received regular help refilling the insulin in the pump and attaching it to their body every three days.

For the study’s positive results to translate to broader type 2 populations, patients will need to be able to use closed-loop systems successfully without supportive care. “It’s not connect and forget,” said Freimut Schliess, director of science and innovation at the diabetes contract research organization Profil. “You have to use it in an informed and educated way.”

But as with other diabetes technology that first gained approval in type 1 patients — continuous glucose monitoring — demand for the artificial pancreas may be increasing before the evidence is firm. “I think there are probably people self-funding closed-loop systems with type 2 diabetes out there,” said Boughton. “Lots of our participants ask us if it’s an option to do after they finish in the trial.”

As researchers continue to investigate outcomes in type 2 patients using closed-loop systems, they’ll need to ask practical questions, including those related to efficiency of care and costs.

“The question is how well would this work in a broader type 2 population,” said Buckingham, “and can you actually show outcomes that would make it reasonable for if you’re an insurance provider to invest in providing this type of technology for type 2s.”

In particular, insurers may want proof that a closed-loop system is more valuable than continuous glucose monitoring alone. In this study, the closed-loop group had access to their real-time CGM data; the control “usual care” group didn’t get to see their glucose numbers. While the evidence on the value of CGM in type 2 diabetes is still limited, recent studies have suggested the sensors can improve glucose control on their own. A good control arm to pull apart the differences, said Buckingham and Schliess, would be a group basing their daily insulin injections on real-time CGM data.

The Hovorka lab is already recruiting for its next trial, in adults with type 2 diabetes who don’t need dialysis, to answer some of those questions. “We feel quite confident that if automated insulin delivery can be safe and effective in this population, it would be beneficial to a wider population with type 2 diabetes who don’t necessarily have kidney failure,” said Boughton. “The desire is definitely there. It’s just a huge limiting factor in terms of the cost of the technologies.” An insulin pump costs about $6,000 upfront, with thousands of dollars more to keep it running over a year.

“You could imagine there would be a commercial motivation for the companies that make these systems,” said Steven Russell, associate professor of medicine at Massachusetts General Hospital and Harvard Medical School. In the U.S., Tandem and Medtronic’s devices are approved, and Insulet has a system currently under consideration by the Food and Drug Administration. “It would substantially increase the population of people who could potentially use their product.”

As the research expands in scope, it may become apparent that automated insulin delivery could help the United States address the surge in under-managed type 2 diabetes and its devastating complications.

“My belief is that in the next five years, we see type 2 diabetes patients treated with closed loop metabolic control systems, but not all of them,” said Schliess. People with kidney failure or multiple comorbid conditions could be among the first to access the devices, he said.

“You have to start with more critical patient groups where you can really demonstrate cost-benefit ratios and risk-benefit ratios which are attractive for payers,” Schliess said. One group is currently conducting a clinical trial of closed-loop systems in that population.

“For type 1 diabetes, automated insulin delivery is really an important advance,” said Russell. “And I think that there is a similar unmet need for improved glucose control and reducing the burden of diabetes management in people with type 2 diabetes who have complex insulin regimens.”

In diabetes care, though, the biggest goal is to help patients before they get to the point of those severe comorbidities — one that current practices are failing to meet. In a recent study in the New England Journal of Medicine, researchers followed adolescents with type 2 diabetes over nine years. “They had a pretty significant number of complications for being pretty young people,” said Buckingham, including renal disease and retinopathy.

To prevent those kinds of outcomes, diabetes care providers need better strategies to improve glucose control across the board — whether that includes closed-loop insulin delivery or other strategies. “I think there’s a lot to be done in type 2,” said Buckingham. “This is very early days, but it shows it can work. And there’ll be a role, it’s sort of when it happens — at what stage and at what age.”

Source: STAT