There are two areas for improvement in insulin pump technology – automation and miniaturization. Cam Med, a medical device startup, is working to design a pump that makes improvements in both areas.
Five years ago, Cam Med CEO Larry Alberts and his colleagues began chasing an ambitious milestone in type 1 diabetes device design: a flexible insulin delivery device no bigger than a Band-Aid. The goal would be for the device to become part of an automated insulin delivery system.
If successful, Cam Med’s device could become one of the smallest, most user-friendly insulin delivery devices on the market, which could substantially ease the burden of type 1 diabetes management for its wearers.
The thin, flexible nature of the device would mean it would be invisible under clothing, Alberts explained.
“You have a lot of flexibility of where you can put it on your body,” he said. “It also has many individual micropumps that have the ability to store and deliver medication.”
Building through research and innovation challenges
By 2016, the company had started to show off its prototype as it participated in the Diabetes Innovation Challenge, presented by T1D Exchange and M2D2. From there, Cam Med landed a major funder in JDRF, and has since gone on to partner with large medical manufacturers, including Johnson & Johson on the development of what they now call the Evopump
“We have been able to make quite a bit of progress in the past couple of years,” said Alberts. “JDRF provided us with significant grant funding that has helped us take our design through an ‘alpha’ prototype to what we call the ‘beta’ prototype.”
Changing the form factor for delivery devices
Though insulin pumps now come in an increasingly diverse range of form factors and sizes, they remain relatively large and bulky, especially for children. Their size stems from the fact that the component parts are difficult to jam into a small package.
Consumer electronics – like computer chips and cell phones – are relatively easy to miniaturize because the individual parts are typically made of hard, stable materials and the processes are well understood. Medical device technologies that contain and transfer liquid, however, are much more difficult to miniaturize.
“We have this array of micropumps that are embedded in flexible materials,” Alberts said, and the whole device “is electrochemically driven.”
The Evopump contains four key layers, each with a different function. The top layer holds the system together, while a subsequent layers serves as a piston, of sorts, deploying the insulin as needed.
To do this, the designers utilized microfluidics, a complex science that combines physics, chemistry, and nanotechnology. The ability to control tiny amounts of liquid could one day lead to the miniaturization of a wide range of drug delivery devices, including insulin pumps.
“A lot of our secret sauce lies in being able to identify the various materials that will go into this membrane. It’s a very thin membrane, but it [includes] a number of different layers,” Alberts said.
Building software for innovative hardware
Cam Med faces another major scientific challenge when it comes to building a software package to handle this new type of delivery device. In addition to grappling with the material science aspects, Cam Med must also dial in the exact features of an algorithm that will reliably deliver insulin on demand and over extended periods of time. Any insulin delivery system needs to be both precise and reliable, able to deliver exact amounts without malfunctioning over days and weeks.
“It has taken us most of the past couple of years to work out these material property issues, as well as the algorithm development to maintain precision of delivery,” he said.
The Evopump still needs additional validation, testing, and tweaking before it’s ready for prime time. Currently, the device is being tested on pigs, and Alberts said the company is still working to miniaturize the circuit board and develop the user-interface that patients will use to interact with the device. Soon, Alberts hopes the prototype will be ready for human testing.
“Today we can produce 8,000 to 10,000 [pumps] per year with the equipment and processes we have,” he said. But in order for mass commercialization, the pumps have to be able to be manufactured at scale. “With the next version, we would be able to produce let’s say 100 thousand per year,” he said. “We can do this at about the cost of an injection pen.”
Ultimately, Alberts hopes that Evopump can be used for a wide range of medical applications, but the development of an insulin delivery device remains the company’s highest priority.