Miriam E Tucker

September 18, 2015

STOCKHOLM — A closed-loop insulin-delivery system, dubbed the 'artificial beta cell,' improved glucose control and reduced hypoglycemia compared with sensor-augmented insulin-pump therapy in patients with type 1 diabetes in real-world settings, new research shows.

The findings were presented September 17 here at the European Association for the Study of Diabetes (EASD) 2015 Meeting by Lalantha Leelarathna, PhD, consultant diabetologist and honorary senior lecturer at Central Manchester University Hospitals NHS Foundation Trust, United Kingdom.

The findings were also simultaneously published online in the New England Journal of Medicine.

The data come from two 12-week multicenter, crossover, randomized controlled studies, one in 33 adults who wore the systems continuously for 12 weeks, the other in 25 children and adolescents who used the systems only at night.

The trials represent several firsts in "artificial-pancreas" research: they are the longest conducted to date, and they are the first in which patients wore the systems while completely unsupervised and unrestricted geographically, with no remote monitoring of their blood glucose levels and no limits on their food or physical activity.

Unlike the so-called "bionic pancreas," which incorporates both insulin and glucagon, this "artificial beta-cell" system uses only insulin, with insulin pumps and sensors that are commercially available in Europe (Sooil and Abbott, respectively).

"Our main aim was to be able to show that closed-loop application works in the day and night period, especially in the long term, and without supervision….We are now at a stage where we've shown that prolonged 3-month, day/night use of closed-loop control in adults is feasible, and 3-month overnight control in children is feasible," lead investigator Hood Thabit, MD, staff diabetologist at the University of Cambridge, United Kingdom, told Medscape Medical News.

Asked to comment on the new findings, "bionic-pancreas" investigator Steven J Russell, MD, PhD, of the department of cellular and molecular physiology, Joslin Diabetes Center and Harvard Medical School, Boston, Massachusetts, told Medscape Medical News, "It's a very strong paper. It's a well-done study, and I think the results are impressive."

He did suggest, however, that the long optimization times (4 to 6 weeks for adults, 2 to 8 weeks for children), the need for carbohydrate counting (as opposed to merely estimating whether a planned meal is the same, more, or less than usual with the bionic-pancreas system), and the lack of glucagon as a safety mechanism might represent barriers.

Real-World Use

Both studies used identical, individually adapting algorithms that rely on a dynamic model of glucose regulation to calculate insulin delivery predicted to achieve desirable glucose levels. Every 12 minutes, the control algorithm calculated an insulin infusion rate that was automatically sent wirelessly to the study insulin pump. The adult study used a "hybrid closed-loop" approach, in which participants estimated their carbohydrate consumption and dosed premeal insulin using the standard bolus calculator.

Dr Thabit noted that most artificial-pancreas systems in development still require users to calculate carbohydrates, estimate meal size, or at least signal that they are about to eat. However, he said, "The closed-loop is in the background.…If the patient [estimates incorrectly], the system will compensate."

The adult study was conducted at three European centers, one each in the United Kingdom, Germany, and Austria, and at baseline, the 33 participants had HbA1c levels of 7.5% to 10%. The children, aged 6 to 18 years, were seen at three UK centers, in Cambridge, London, and Leeds. They had baseline HbA1c levels below 10% on insulin-pump therapy.

For both studies, the same patients used the closed-loop and sensor-augmented pump therapy for 12 weeks each, with washout periods in between.

The primary end point was the proportion of time that the glucose level was between 70 mg/dL and 180 mg/dL for the adults and 70 mg/dL and 145 mg/dL for the children and adolescents. (The upper target was lower for the children/adolescents since postprandial levels weren't included.)

HbA1c Values Relevant for the First Time

There was a statistically significant 11-percentage-point increase in the number of adults using the closed-loop system who met this end point (68% vs 57% with the insulin-pump therapy; P < .001), while in the children there was a significant 25-percentage-point improvement (60% vs 34%, respectively; P < .001).

Mean glucose levels were 157 mg/dL with the closed-loop vs 168 mg/dL during the control period among the adults (P < .001), and 146 mg/dL vs 176 mg/dL, respectively, in the children/adolescents (P < .001).

HbA1c values — relevant for the first time in an artificial-pancreas study due to the long duration — were significantly reduced in the adults with the closed-loop system, from 7.6% to 7.3%, while remaining unchanged following the control period (P = .002). For the children, the drop from 7.8% to 7.6% with the closed-loop system didn't quite reach statistical significance (P = .17).

The proportion of time spend with blood glucose levels below 50 mg/dL was significantly lower when using the artificial pancreas than with the control among the adults — 0.3% vs 0.4% (P < .001), but this difference was not significant in the children/adolescents (0.3% vs 0.6%; P = .31).

Three severe hypoglycemic episodes occurred during the closed-loop phase, but they happened when the closed-loop system was not in use.

The Glucagon Question

Dr Thabit told Medscape Medical News that these data add further validation to the concept that closed-loop systems can be achieved without incorporating glucagon as a safety feature. And because they are less complex than the bionic pancreas, which requires its own dedicated dual-chamber device, single-hormone systems may be poised to reach the market sooner.

"At the moment, we're looking at having a system available as soon as possible, as we're using off-the-shelf items. The pump and [continuous glucose monitor] CGM are already available in Europe," he noted.

He pointed to a study published in June in Lancet Diabetes & Endocrinology and reported at the American Diabetes Association Scientific Sessions showing that although closed-loop systems that use glucagon do further reduce the risk for hypoglycemia, mean glucose levels and time in target range don't differ.

With the bionic pancreas, "you get further protection from hypoglycemia, but it comes at a price of further device burden," Dr Thabit said, also noting that stability and cost of glucagon are issues, as is the uncertainty about long-term safety of chronic glucagon exposure.

However, Dr Russell said that he has seen no evidence for toxicity with glucagon in the bionic pancreas thus far, but of course that will be investigated further in the pivotal trial of their system. He also said that his team is exploring the use of more stabilized formulations of glucagon and glucagon analogs that are in development.

But he also told Medscape Medical News that the differences in the two types of closed-loop systems might mean more alternatives to meet individual patient needs.

"I'm not sure it's either/or. I think there may be room for multiple products in the marketplace."

Funding for the adult artificial beta-cell study came from a European Union grant, while the study in children/adolescents was funded by the Juvenile Diabetes Research Foundation. Both studies had additional support from the National Institute for Health Research Cambridge Biomedical Research Centre and Wellcome Strategic Award. Drs Thabit and Leelarathna have no further relevant financial relationships. Disclosures for the coauthors are listed on the journal website. Dr Russell is an advisor and/or consultant to Tandem Diabetes Care, Companion Medical, Sanofi, Dexcom, Eli Lilly, Abbott Diabetes Care, Insulet, Medtronic Minimed, and International Biomedical.

N Engl J Med. Published online September 17, 2015. Article


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.