NIH backs smartphone-based artificial pancreas trials with almost $13M

All these academic efforts are on behalf of startup TypeZero Technologies, which in-licensed the artificial pancreas prototype tech–dubbed inControl AP–from the University of Virginia in 2013 and has since been testing and developing applications for it.  The recent NIH award was for an academic team led by UVA and Harvard University.  A 240-patient, 9-site U.S. and European 6-month trial is slated to start early this year and lead into a second trial in a subset of 180-patients that will be followed for an additional 6 months to further test the algorithm used.

The first trial will compare the artificial pancreas to a control of a standard insulin pump to comparatively assess how well blood-sugar levels are controlled and whether the risk of hypoglycemia was reduced.  “To be ultimately successful as an optimal treatment for diabetes, the artificial pancreas needs to prove its safety and efficacy in long-term pivotal trials in the patient’s natural environment,” said principle investigator Boris Kovatchev, director of the UVA Center for Diabetes Technology, in a statement.  “Our foremost goal is to establish a new diabetes treatment paradigm:  the artificial pancreas is not a single-function device; it is an adaptable, wearable network surrounding the patient in a digital treatment ecosystem.”

There are several marketed systems for Type 1 diabetics that combine the continuous monitoring of blood glucose levels with the ability to use insulin to manage them.  However, they rely upon patients ultimately to take action on insulin usage, rather than automating that process like an artificial pancreas.  In addition, almost all of these glucose monitor/insulin pump combos require routine, daily finger sticks for external calibration via a more traditional monitor.

The artificial pancreas that will be tested is a closed-loop system that includes an insulin pump, a continuous glucose monitor implant and an adaptive control algorithm software that runs on a smart phone.  The system is based on a zone model-predictive control algorithm that is based upon identifying and maintaining a personalized blood glucose range for each patient, rather than linking the glucose goal range to specific, pre-identified points.  The application takes into account variables including meals consumed, physical activity, sleep, stress and metabolism.

In 2014, the NIH started a program with $20 million in funding for artificial pancreas advanced clinical testing.  This award is the largest it’s made thus far under that initiative, which previously has funded other comparable efforts by the University of Cambridge with $6.4 million, the Diabetes Wireless Artificial Pancreas Consortium (DREAM) of diabetes centers in Germany, Israel, and Slovenia with $2 million, as well as Boston University and Massachusetts General Hospital with $1.5 million.

“The biggest challenge in the design of the artificial pancreas is the inherent uncertainty in the human body,” noted co-principal investigator and engineering lead on the project Francis Doyle III, dean and Professor of Engineering & Applied Sciences at Harvard.  He added, “Day to day, hour to hour, the various stresses that impact the human body change the way it responds to insulin-controlling glucose. Physical stresses, anxiety, hormonal swings will all change that balance.  To be able to control for those factors we need to see longer intervals of data.  This is the first trial where we’ll be looking at multi-month intervals of time with cohorts of subjects where we can actually see a long enough window to learn those patterns, to adapt and fine-tune the algorithms, and to improve the overall level of glucose control.”

NIH isn’t alone in its support of TypeZero’s work on an artificial pancreas; nonprofit Type 1 diabetes research organization JDRF and the Virginia-oriented seed-stage investor CIT GAP funds them as well.

REFERENCE:  Fierce Medical Devices; 05 JAN 2016; Stacy Lawrence

Leave a Comment