First bi-hormonal home-use trial starts today

Beta Bionics, using its iLet system, started the first at home trial using both insulin and glucagon to control blood glucose. Here’s the opening paragraph of the press release.

Boston, MA and Copenhagen, Denmark, May 22, 2019 – Beta Bionics, Inc. and Zealand Pharma A/S (NASDAQ: ZEAL) announced today the first patients dosed in a clinical trial investigating the home-use of dasiglucagon in the iLetTM Bionic Pancreas System. The iLet, developed by Beta Bionics, is the world’sfirst autonomous bionic pancreas device — a bihormonal system leveraging lifelong machine learning and artificial intelligence to deliver insulin and glucagon analogs for the autonomous treatment of type 1 diabetes. In addition to dosing insulin, the iLet doses dasiglucagon — a glucagon analog developed by Zealand Pharma with a unique stability profile in a ready-to-use aqueous solution.

It’s promising that a shelf-stable glucagon is finally in the mix. This might be the “killer-app” of the automated insulin dosing category. We shall see.


I was asked to put in for the iLet trial last week, but didn’t qualify: my A1Cs are too good! Kinda disappointed, kinda relieved. I was curious, especially about the glucagon analog, but it was surprisingly short notice and there’s a lot of interviewing and returning for testing and whatnot. I work right down the street from MGH, where the trial is taking place, but it was still going to mean a lot of hours away from my desk. Would have been interesting to be in on it, though, I was really of two minds, but they made the decision for me. Curse you, 5.8 A1C!!!


It’s unfortunate that most studies on people with diabetes require a high A1c so that when measuring study outcomes, the effect will be more dramatic. I wish they would find a way to use volunteers who already achieve good glucose control.

I’ve reviewed many studies that excluded me based on my normal A1c. But now they’re excluding me since I’m older than 65. Oh well!

1 Like

Yes, welcome to the clinical trial world where those of us over 65 are often excluded. And yes, it is a double whammy being over 65 with a good A1c! I would really like to test the iLet!


Not only is the effect more dramatic, but it’s also easier to prove it’s actually doing something. If you think about it, how much are you going to improve an A1c of 5.8, which is considered a normal range?

Besides the fact that they stack their trials to get the results they want. Which in this case is people they can look at and say, hmmm, we can give part or all of them diet advice etc and use this new tool and look what amazing results we’ll get. And they can even say to promote it at the end, with just a little advice and this devise we got amazing results.

Yeah, I was kinda anticipating that, so I went ahead and gave them my number along with a few other basic data points so I wouldn’t waste hours going in for the initial interview if they were going to bounce me out anyway.

The general public has no idea how bogus, corrupt, and self-serving medical studies are. My wife saw it happen first-hand when she worked for a doc who was getting bucks to conduct a study. He skewed things in order to prevent losing out on the study continuing. Lots of shady stuff goes on, folks. She worked for him for a very short time, thank goodness.

1 Like

Oh it’s awful. I don’t think people realize the scope of how you can pull the data results that you want to get from just about everything.

1 Like

I agree with your point. I think a narrow focus on A1c outcomes alone, however, miss many other measurable outcomes that could lend statistical heft to the study and eliminate the need to exclude many study candidates. Things like increasing time in range, reducing BG variability and reducing time spent hypo are all possibilities. I think fixating on the A1c outcome is intellectually lazy.

1 Like

@Terry4, yea, but think how much better all that can be “solved” by someone that isn’t controlling it well. And the A1c test is just one way that lets them filter that.

There are also quality of life measures that are under-utilized as study outcomes. I think anything that measures better quality of sleep due to fewer out-of-range glucose interruptions would make sense. How about the number of times you have to manually correct errant BGs, interrupting your work and family time.

I understand the perception of the A1c as a proxy for overall control but, as you well know, improvement of blood glucose levels affects so many facets of our lives. I realize that my position here is again in the minority and study originators are not standing in line to seek my guidance! My argument is more rhetorical.


Most doctors and patients are concerned with lowering the patient’s A1c as long as it doesn’t create more dangerous hypos. I wouldn’t characterize the weight that is given to the A1c results as “intellectually lazy”. Seems a bit harsh! Of course there are other considerations, but the A1c is regarded as having a threshold number at which far fewer complications occur and I’d say that is why we all hope to stay under that “magical” number. I also agree that having a high TIR would be far preferable to having frequent roller-coaster bg’s.


This really worries me, @Dave44. Coming from the harder sciences, we were taught that we would go to hell for falsifying data. I might be asked to do that now. I sorta feel like if you are increasing access to care - like, bringing care to people who currently have none, then maybe it can’t really hurt anybody. In some ways, care is currently so unethical that I wonder how you can make it any worse. It would be super embarrassing, though, if someone combed over your results and was able to challenge them.

Its been my experience, so far, that employers give a very short amount of time to conduct a lot of analysis. The tools that they use are embarrassing. This results in analysis that is just outright wrong. They seem OK with that. I dont understand it. That seems so embarrassing.

Of course they might be interested in finding out whether the thing could maintain that high a standard in someone who is already achieving it. The description of the test said something to the effect that they wanted you to go about your daily routines as normal to see how the system performed under real-life conditions. Having tried the 670G, which in my case failed to meet that test, I was curious to see if the Insulet could pull it off.

What’s the point of a trial for people with high A1C’s? If they have hypo problems it is likely because they aren’t doing a lot of things they should be, like using CGMS etc.
Why would they ever need Glucagon?

Someone trying for an A1C near 4 would be far more likely to benefit from it.
I can only get into the 5% range without it and without excess hypos.

1 Like

@jhe6 That’s exactly the point. You can give someone counseling with the trial and that way make sure the results look impressive. " I went from 10% to 7% with this new system and some minor adjustments, see how good it is. Versus I went from 5.9% to 5.2% with this new system. Say what?? They didn’t even have a problem in the first place so who cares would be the common response I believe for a lot of people.

Regardless of the in/outs of trial selection, this is exciting news!

As someone who has a lot of variability due to wildly different activity levels each day, having a system that can proactively respond to lows beyond suspending insulin delivery, seems very desirable.


Same here, that’s why I was really interested in trying it with my “normal daily routine” to see if it could really handle it. Like the ridiculously high DP in the a.m. and the nasty hypo if I’m bike commuting in the afternoon, stuff I’ve been trying to even out for decades. I’ve gradually gotten better and better on my own—5.8!—so can these sorts of systems do as well? I know the 670G could not—big part of that was the MedT CGM not catching my DP—but could a dual system using the more reliable Dexcom?

One thing I didn’t mention is that they’re running the trials with the G5, not the G6, which is too bad, as I’m finding the G6 significantly more accurate. Except in the first 24 hours, like all CGMs. And how do any of them deal with that problem, BTW?

1 Like

Agree. A1C is and always will be the most meaningful measure of complication risk. All the other numbers are tools to use day to day, but a1c is a result and therefore will always be the gold standard in clinical studies, and generally the bottom line in almost every discussion in medicine…

:slight_smile: except for those who are focused on TIR.

Personally, I wouldn’t my A1c to suffer because I got my TIR to be better, but it was at the higher end of “the range”. I don’t know about the rest of you, but MY range is set by ME–not some industry number.