Omnipod Feasibility Study Seeks to Perfect Formulas That Will Make Artificial Pancreas Run

For years, researchers around the globe have raced toward the holy grail in type 1 diabetes (T1D) care: the artificial pancreas, a device that delivers the right insulin dose without patient interaction, without extreme highs or lows in blood glucose, and with limited fear of overnight episodes of hypoglycemia that keep loved ones sleeping with one eye open.

Last month’s FDA approval of Medtronic’s MIniMed 670G was the first acknowledged by JDRF as worthy of the “artificial pancreas” label, although it is a hybrid closed loop system, not what JDRF calls “the end goal.” Other manufacturers have devices at various stages of development, which advocates for the T1D community say is a good thing—more competition means innovation, choice, and, they hope, downward pressure on prices.

In mid-September, Insulet Corporation announced that the first patient had completed a feasibility study for its Omnipod Horizon Automated Glucose Control System. Insulet’s entrant in the artificial pancreas market is an advance of its existing Omnipod tubeless insulin pump, which is compatible with different continuous glucose monitoring (CGM) systems—consisting of a sensor, a transmitter, and a receiver that allows the patient to read data. Mobile CGM systems can send data to a patient’s smartphone, which makes monitoring their blood sugar discreet and gives them one less device to carry.

Not everyone with diabetes uses a CGM, but those that do have ongoing information about where their blood glucose levels have been and where they are headed, which allows more precise insulin dosing and can let patients (or parents) take action before a hypoglycemic episode ends with a trip to the emergency department.

So, just what is an artificial pancreas? As researchers explain during updates each year at the American Diabetes Association, closing the loop adds a third element to the CGM and insulin pump—the algorithm. These are the event-driven formulas written into software, which will allow the devices to make the multitude of decisions about insulin delivery—based on what patients eat, how much they sleep, how much they exercise, or if they’re running a fever—that would otherwise be made by a healthy pancreas.

According to Shacey Petrovic, president of Insulet Diabetes Products, the current feasibility study is about perfecting the algorithm—exploring its limits under a variety of conditions. The Omnipod Horizon system will be a hybrid closed loop system, Petrovic said, which means patients will still interact with it at mealtime to “announce” how many carbohydrates are being consumed. After 20 adults complete the feasibility study, she said, “we will test larger patient populations based on different scenarios.”

The next step would be a pivotal trial. Based on how quickly the FDA acted on the MiniMed 670G, Petrovic said, she hopes that the Omnipod Horizon system could reach the market by 2019.

The device’s niche among competitors for a share of the artificial pancreas market is its tubeless design—the waterproof “pod” affixes to an arm or other part of the body with 3 days’ worth of insulin, making it ideal for people who are active, including children.

“The pod is doing amazing things for kids and for their families,” Petrovic said. Kids can wrestle, swim, and play without worrying about tubes being tangled, and parents are not faced with multiple daily insulin injections on their young children, she said.

Omnipod’s users tend to be younger: according to data from Insulet, the average age of current active patients is 36, and 29% of the direct active patients are age 20 or younger. Separate data from the Diabetes Connections USA Patient Panel Report show the Omnipod is preferred by 31% of the pediatric segment, defined as age 18 or younger.

But besides the design, there’s another reason why Omnipod users skew young: So far, Petrovic said, Medicare will not pay for the device. This occurs even though recent evidence shows that beyond T1D patients, those with type 2 diabetes (T2D) who use insulin might benefit from switching to a pump instead of injections. Fear of needles has been cited as a reason T2D patients resist moving to insulin therapy. (Data from Insulet show 11% of users have type 2 diabetes, including a fraction of pediatric patients—1% of users below age 18.)

Payers’ willingness to fund diabetes technology—and give patients choices—has been top of mind in the advocacy community. In July, an FDA panel supported approval for Dexcom’s G5 Mobile CGM system to be used for insulin dosing; this is a change that many see as the first step for Medicare to pay for CGM. But advocates were taken aback in May when UnitedHealthcare reached a deal with Medtronic for most adult patients to transition to Medtronic’s MiniMed pumps.

Petrovic said that Insulet is watching all these developments. “There is great concern about this," she said. It’s one thing for a payer to have a preferred drug, but “people don’t view interacting with a device the same way.”

Fortunately, she said, early talks with payers suggest that improving quality of life issues will matter, along with improved glycemic control, when it’s time to make coverage decisions on artificial pancreas technology. “We don’t want to stifle innovation that has finally started to accelerate,” she said.