How Many Hormones Make an Ideal 'Artificial Pancreas?'

Miriam E. Tucker

July 01, 2020

Some automated insulin delivery systems currently in development add glucagon and/or pramlintide to insulin, but whether the extra hormones are worth the additional cost and effort is a subject of debate.

Also called closed-loop or artificial pancreas systems, they are comprised of an insulin pump and a continuous glucose monitor (CGM) that communicate via a built-in algorithm to deliver insulin based on glucose levels. Currently available systems are called hybrid closed-loops because they still require user input for meals, exercise, illness, and other special circumstances.

Two hybrid closed-loop systems available in the United States, the Medtronic Minimed 670G and the Tandem Control-IQ, as well as the Medtronic Minimed 780G that was just approved in Europe, use insulin only.

Of all ongoing active closed-loop clinical studies, 44 involve insulin-only systems, as of May 2020.

However, two such systems in development add a glucagon analog to insulin in the same pump (in separate cartridges), with the aim of minimizing the risk of hypoglycemia. And four investigational systems combine insulin with pramlintide (Symlin, AstraZeneca), an amylin analog that reduces post-meal glucose spikes. Three systems in development combine all three hormones. 

In a debate during the virtual American Diabetes Association (ADA) 80th Scientific Sessions, Roman Hovorka, PhD, of the University of Cambridge, UK, argued in favor of insulin-only systems on the basis of efficacy, less burden and complexity, and lower cost.

But Steven J. Russell, MD, PhD, of Massachusetts General Hospital, Boston, countered that glucagon adds safety and value to the system by allowing for more aggressive insulin dosing with lower hypoglycemia risk, benefits which he said would overcome any downsides.

Insulin-Only Systems Are Good Enough

Hovorka began by defining a "good" artificial pancreas as one that produces consensus time-in-range targets of at least 70% of glucose values between 3.9 to 10 mmol/l (70-180 mg/dL) and less than 3% below 3.9 mmol/L (70 mg/dL). At the same time, he said, the burden should be low, which he suggested means no more than 10-20 minutes a day spent managing the system, low "alarm burden," and minimal technical issues.

"We need to balance glucose control and the burden. For some people, reducing the burden is sometimes even more important than the glucose control," Hovorka commented.

He pointed out that in addition to Medtronic's and Tandem's systems, two other insulin-only hybrid closed-loop systems are marketed outside the United States. These are the CamDiab system, available in the UK, which uses his group's Cambridge control algorithm in a Dana pump with the Dexcom G6 sensor, and the Diabeloop algorithm, available in Europe, that combines a patch pump with the Dexcom G6.

"Lots of energy and resources are going to taking [insulin-only] systems into clinic use," Hovorka observed.

He reviewed recently published data for both the Tandem Control-IQ and the Cambridge control algorithm showing similar results meeting the "good artificial pancreas" definition.

In his current clinic population of 160 patients aged 2-80 years using the Cambridge algorithm, 69% of users have achieved 70% or greater time in range and 28% have achieved 80% or greater time in range.

"So, the insulin-only system can achieve acceptable, and in some instances very good, glucose control," Hovorka said.

He acknowledged that there are still challenges with insulin-only systems, including exercise-related dysglycemia and postprandial hyperglycemia related to slow insulin absorption, missed or incorrect boluses, or large meals.

But, Hovorka said, downsides of dual-hormone systems include the need for room-temperature stable glucagon and for dual-chamber pumps with two cannulas and two infusion sites (in addition to the sensor site), and the unknown long-term biological risks of chronic subcutaneous glucagon or pramlintide delivery.

Moreover, he said, costs are expected to be higher for a two-chamber versus single-chamber pump, as well as for the second hormone, reservoir, and infusion set.

Data thus far from short-term studies suggest that insulin-only systems are sufficient in eliminating nocturnal hypoglycemia, while the addition of glucagon potentially reduces daytime hypoglycemia, especially during exercise.

However, longer-term head-to-head studies will be needed, Hovorka said, noting, "Comparative benefits of the single- and dual-hormone system for improving hemoglobin A1c and preventing severe hypoglycemia remain unknown."

He suggested that glucagon dual-hormone closed-loop systems might be suitable for patients who are particularly prone to hypoglycemia, whereas pramlintide dual-hormone systems have the potential to more fully close the loop when used with ultra rapid-acting insulin analogs.

Nonetheless, he said, "Many if not most users may achieve acceptable control with insulin-only systems."

Dual-Hormone Systems: Extra Benefit Worth It?

Russell, who is an investigator in multicenter pivotal studies of both insulin-only and bihormonal configurations of the Beta Bionics iLet bionic pancreas, began his debate presentation by endorsing the effectiveness of insulin-only hybrid systems and stating that he encourages his patients with type 1 diabetes to use them.

But, he said, adding glucagon can allow for better automation of hypoglycemia prevention and treatment in situations such as exercise.

"A bihormonal system achieves lower glucose, higher time-in-range, and less hypoglycemia than a well-functioning insulin-only system."

Moreover, Russell said, "Glucagon reduces the need for medicinal carbohydrates, promotes satiety, and increases energy expenditure...Combined, these three factors may oppose weight gain or encourage weight loss as opposed to a system that uses insulin only."

He pointed to a 2017 meta-analysis that showed improved time-in-range and greater reductions in hypoglycemia with dual- versus single-hormone systems. 

And, in unpublished data from a randomized random-order crossover study of 23 patients with type 1 diabetes who each spent a week with usual care (insulin pump with or without CGM), insulin-only iLet, and bihormonal iLet, mean glucose levels were 165, 148, and 139 mg/dL, respectively. Time in range was 60%, 72%, and 79%, and median time with glucose below 54 mg/dL was 0.6%, 0.6%, and 0.2%, respectively.

Russell also addressed each of the arguments made by Hovorka and others against glucagon use.

Regarding the need for a stable glucagon formulation, he said that the analog being developed for the iLet, dasiglucagon (Zealand pharma), is stable for more than a month at 40ºC, with higher bioavailability and slightly slower absorption than glucagon.

And while he acknowledged the need for two separate hormone cartridges, Russell said that the Gen4 version of the iLet is fairly simple and intuitive, and the device itself is about the same size as the Tandem t:slim.

Use of glucagon didn't increase insulin use in iLet trials, nor was it associated with increased reported nausea or liver glycogen depletion.

And users universally reported preferring the bihormonal system.

Long-term safety of chronic glucagon exposure has yet to be addressed, but animal data are reassuring, Russell said.

Regarding increased cost, he pointed to 2018 data showing that the incremental improvement in glycemic control from no automation to single-hormone automation is the same as from single to dual (mean glucose reductions of 7.4 and 13.6 mg/dL, respectively, and decreases in time spent in hypoglycemia of 1.28% vs 2.95%).

"I would argue that if one can justify adding automation, one could justify some additional expense to add the cost of glucagon." And, he said, the cost would likely be based on a negotiation around the extra value offered by the dual hormone system.

"The addition of glucagon, I believe, will be justified by the improved outcomes and improved quality of life," he concluded.   

Hovorka has reported receiving research support from MiniMed Medtronic, Abbott Diabetes Care, and Dexcom; being a speaker for Novo Nordisk, Eli Lilly, and Dexcom; holding license fees from B. Braun and Medtronic; and being director of CamDiab. Russell has reported holding patents on aspects of the bionic pancreas; receiving honoraria, travel expenses, and/or research support from Dexcom, Eli Lilly, Tandem Diabetes, Sanofi, Novo Nordisk, Roche, Ascensia, Zealand Pharma, and Beta Bionics; being a consultant for Flexion Therapeutics, Senseonics, and Beta Bionics; and participating in scientific advisory boards for Companion Medical, Tandem Diabetes, and Unomedical. 

For more diabetes and endocrinology news, follow us on Twitter and Facebook.


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.