Grand Prize Winner Pancreum: A Small, Modular Artificial Pancreas

The Pancreum closed loop (automated insulin + CGM + glucagon) system that won a Grand Prize in the DiabetesMine Design Challenge this year may look like a pipe dream, but designer Gil DePaula assures us it is “visionary but real.”

Have a look at the video, below, and also Gil’s company website.

“The glucagon part is definitely a futuristic concept — because there’s no predicate device for glucagon delivery with the FDA, so that’s a huge question. But the insulin pump and CGM are as real as OmniPod was when I joined Insulet in 2002 (he worked there 5 years). And now that’s real and on the market; there are predicate devices out there. So if I want to introduce a new disposable tubeless insulin pump, I can do it. I won’t have to climb mountains.”

— Gil DePaula, Pancreum LLC

Based in Fort Lauderdale, FL, Gil DePaula is a software and electronic engineer who previously worked on the Guardian RT CGM system at Medtronic Diabetes, and also helped launch OmniPod with Insulet Corp. He went out on his own a few years ago to pursue his dream of developing a flexible, open-platform, high-performance closed-loop system for diabetes.

We interviewed Gil about Pancreum this past winter, before anyone was thinking in design contest mode; he hadn’t even heard of the competition back then. Once he learned about the contest, he decided it was worth “going public” with his only-partially-patented ideas, because above all, he wants to see the diabetes device market evolve. “I think we’re stuck with too many old ideas,” he says.

“When I saw all the other videos and entries this year I was nervous. I wondered if we had a chance. I still can’t believe we won!”

We talked with Gil again last week, to get some more detail on where he is now in the process of developing the Pancreum system. One thing he was excited to share was that he’d connected with last year’s DiabetesMine Grand Prize winner Mauro Amoruso, a professional designer based in Turin, Italy, who has helped Gil refine the form and function of the Pancreum system.

DM) I think the big question for many people looking at your design is, how do the cannulas work?

GdP) The pump cannula is similar to OmniPod, with automatic insertion, except that OmniPod has two springs and we hope to need only one.

In the OmniPod, the cannula is wrapped around the needle. It inserts, and a second spring fires and pulls the needle out and leaves the cannula. We’re working to make this a one-step process.

For the CGM, if the needle is the cannula, which is currently the case with most models, then you also need just one spring. But current CGMs have manual insertion devices. We hope to do away with that and offer automatic insertion there too.

Right now our whole circular unit with two wedges (the insulin pump and CGM portions) is about the size of the OmniPod.

If it’s so small, how can it house enough insulin to be useful for most patients?

We think the reservoir size is going to be 400 units, so it’s actually 2x the capacity of the OmniPod reservoir.

We’ve designed it mechanically — a 400-unit reservoir and drive mechanism — that fits inside our little ‘wedge.’ It’s a narrow and long device.

So you already have working prototypes of all the pieces of this system?

I have a physical prototype that I’m bringing to the ADA conference this week.  It’s not the black one you see in the video — that looks like Darth Vader! — but a cooler color.  The prototype is now much smaller too. It’s half the size of a cell phone.

The only thing that hasn’t been touched in real development work yet is the mechanical part — the drive mechanism and reservoir I mentioned. We’ve designed it, but haven’t created the necessary plastic pieces yet.

For me to have someone design an electronic board that fits inside the footprint we’ve created, I need money. That’s why we’re looking for investors now.

What will the system offer patients, beyond connecting the pump and CGM? Like automatic shut-off of insulin if you dip too low?

This is not just sticking several devices on a patient. We’d like to take the functionality a step further. For example, if you’ve had to correct several times in a row, should you not increase your basal rate? In other words, if the system has to suggest correction boluses frequently, it’s also gonna suggest a basal rate increase as well. It will also suggest decreased basal rates, stopping basal rates, and will take insulin on board (IOB) into account — multiple features like that.

And also automatic shut-off should definitely be part of the system once the FDA will allow it!

So that and glucagon delivery are the biggest FDA hurdles?

Yes. I don’t understand why FDA picks on glucagon so much. Unlike insulin, if you mess up with glucagon (or automatic shut-off) all you’re gonna do is raise somebody’s BG levels for a while. But if you mess up with an overdose of insulin, you’re gonna do some serious immediate damage.

Isn’t the Artificial Pancreas Project already testing with glucagon delivery in the mix?

We’ve gotten a lot of information from different people, including Dr. Russell from Massachusetts who’s done testing using all three devices on pigs.  He’s doing a lot of great work, but still there are a lot of questions to be answered:

• How much glucagon do people generally need for six-day wear?
• Does glucagon’s viscosity effect how wide the cannula needs to be?
• What is the glucagon ratio? As in how much does one unit raise your BG?
• Do we need basal rates for glucagon or just boluses?

These are all questions to be answered, and it will take time.

So what’s your next step? What will you do with the prize money, for example?

We’re developing an app that runs on Android right now. We’re doing this first because of Apple — they require too much before you can develop something that can connect with their platform.

I’ll use the prize money towards improving the Pancreum Android App and building a real-size electronic prototype for the CoreMD and wedge(s).

We’d like to get in touch with others who did CGMS software entries (like mobiLIFE) to discuss possible collaborations.

That’s great, Gil! Anything to add?

I wanted to say that as soon as I saw the BLOB entry, I thought from the start this design will win! I don’t have diabetes, but one of the oldest technologies in the field is insulin pens. The BLOB idea totally revolutionizes that space.

We hope so, Gil. And as one of readers commented on the winners’ announcement: Go Pancreum!