Conference Report:
Obstacles and Opportunities on the Road to an Artificial Pancreas

National Institutes of Health
December 19, 2005

On December 19, JDRF staff and volunteers convened with basic scientists, clinical investigators, and experts from industry and government at a conference focused on prospects for an artificial pancreas. Held at the National Institutes of Health in Bethesda, Maryland, "Obstacles and Opportunities on the Road to an Artificial Pancreas: Closing the Loop" was devoted to evaluating the current state of development for the device and determining what still needs to be done.

JDRF has made the artificial pancreas one of its six cure therapeutic goals and recently increased its commitment by authorizing more funding and resources for this objective. It is estimated that without our leadership, development of such a device would take more than 10 years, but JDRF hopes to dramatically accelerate that timetable.

The overriding sentiment at the meeting was clear: An artificial pancreas is within reach soon if enough researchers and agencies make it a priority and various constituencies work together to address the remaining hurdles. Technology already exists for the two main components of a "closed loop" systemglucose sensing and insulin deliveryboth of which have proven effective in everyday use. The key is to link the two components into a reliable device and ensure that it is available to most type 1 patients.

At the meeting's outset, Michael White, Chair of JDRF's Research Committee, reminded participants of why they were there and why an artificial pancreas is desperately needed for type 1 patients: Management of diabetes is difficult and the majority of people are not under proper control. The main barrier to reaching good control is fear of hypoglycemia. This fear is so intense that patientsor those caring for themallow blood sugar levels to go too high, leading to devastating complications. Mr. White explained that the objective of JDRF's initiative is a device that will change the way diabetes is managed.

The goals include:

•  Technology that improves glycemic control;
•  Devices that have broad patient access (and not restricted to a small number of patients due to high price, lack of insurance reimbursement, or limited capabilities);
•  A thriving, competitive market with lots of choices, similar to the range of insulin pumps now available;

•  Current Care is Insufficient.  Michael Brownlee, M.D., a JDRF-funded researcher at Albert Einstein College of Medicine, detailed the extensive complications faced by people with type 1, including data showing that by their early forties, 100 percent of people with type 1 have macrovascular damage.  He noted that even those people who achieve recommended hemoglobin A1c levels spend too much time in glucose ranges high enough to cause serious complications.  Reducing glycemic variabilityeven without reducing A1c levelscould reduce retinopathy by two-thirds. 
•  Technology can make a huge difference, even if short of an artificial pancreas.  Even short of closing the loop, use of insulin pumps and continuous sensors can improve glucose control.  William Tamborlane, M.D., a JDRF-funded researcher at Yale, said, "We can't go directly from the relatively poor control we have now to perfect control, but an improvement will make a big difference." The first artificial pancreas is likely to be an "open loop" system where the patient uses information from a continuous sensor to adjust insulin delivery.  
•  An external 'closed loop' is already proving itself in the lab, even in children. Dr. Tamborlane presented new data from studies his group has conducted at the JDRF Hypoglycemia Center at Yale University using a completely "closed-loop" system in kids (a Medtronic continuous glucose sensor and insulin pump). Results were very promising. Overnight control was excellent, and some post-meal glucose increases were addressed by some insulin boluses prior to meals.

But technical challenges remain. Several scientists discussed difficulties of devising algorithms accurate enough to rely upon in an automated, closed loop system. Regulating glucose levels around meals still presents the biggest challenges for automated devices. Another challenge is that sensors tend to "drift" away from actual values and need to be recalibrated - as a result, the first sensor components of an artificial pancreas will need to be calibrated with fingerstick glucose readings periodically.  And finally, researchers discussed how the pancreas does much more than simply produce insulin, and that future generations of the devices should monitor and respond to multiple metabolic factors so they mimic a real pancreas even more closely.

Overall, this NIH-JDRF-FDA workshop was an important step in JDRF's efforts to help accelerate the availability of an artificial pancreas to people with type 1 diabetes.  This effort will include not only research to test the effectiveness of these devices and solve scientific challenges but also efforts to speed regulatory approval and obtain health insurance coverage for these technologies.