Skip Navigation

Diabetes Introduction Treatments Complications Statistics Clinical Trials
NDEP Awareness and Prevention Series Resources Order About NDIC Informacion en Espanol
A to Z list of Diabetes Topics and Titles Easy-to-Read Publications Spanish-language Publications Awareness and Prevention Series
Email to a friend  Email to a friend icon
Print this page    Print this page icon
  Home : About NDIC : Diabetes Dateline : Fall 2008

Diabetes Dateline
Fall 2008

NIH Hosts Artificial Pancreas Workshop

Photo montage of two doctors consulting while looking at a chart and a close-up of a hand holding a glucose monitor.Leaders in the effort to make an artificial pancreas a reality for the millions of people who take insulin for their diabetes convened at the National Institutes of Health (NIH) in July to discuss how much closer they have come to achieving that goal and the major obstacles that stand in the way.

The National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), in collaboration with the U.S. Food and Drug Administration (FDA) and the Juvenile Diabetes Research Foundation International (JDRF), has been instrumental in speeding progress toward an artificial pancreas.

As conceived by researchers, a “closed-loop” artificial pancreas is a system of devices that would be able to monitor blood glucose, also called blood sugar, and administer the correct amount of insulin, and possibly glucagon, to maintain safe and healthy blood glucose levels—all without intervention from the person using it.

Since December 2005—the last time the NIH sponsored a meeting about developing an artificial pancreas—“the NIDDK has continued to support cutting-edge research for the development of minimally invasive glucose monitors, long-term implantable glucose monitors, new glucose-regulated insulin delivery technologies, technologies to increase the biocompatibility of the devices, and the development of the algorithms that mimic the physiologic homeostasis,” said NIDDK Director Griffin P. Rodgers, M.D., M.A.C.P. “All these were eventually to be components of a future artificial pancreas.”

In 2006, the JDRF launched the Artificial Pancreas Project, making perfection of a closed-loop system one of its six therapeutic goals. “We believe we have the core technologies that are necessary to build a safe artificial pancreas,” said Larry A. Soler, the JDRF’s vice president of government relations. “It’s going to take all of us working together to put these things together and make a difference for people with diabetes.”

The FDA continues to include the artificial pancreas as part of its Critical Path Initiative—a program to bridge the gap between basic scientific research and medical product development. “This initiative was not meant to be just an FDA project,” said Janet Woodcock, M.D., acting director at the FDA’s Center for Drug Evaluation and Research. “This [initiative] was meant to call upon the scientific community to band together to find new ways of advancing the movement of innovative science on the critical path to becoming actual products.”

“It has been very satisfying to see how the NIH, the FDA, the JDRF, academia, and industry have collectively catalyzed very intensive and productive activity in the development of an artificial pancreas,” said Rodgers.

Lessons Learned

Workshop participants summarized lessons learned from studies of continuous glucose monitors, closed-loop systems, and open-loop systems, which require some intervention from the user.

Controlling post-meal glucose spikes has been a major challenge for closed-loop systems, which suffer from inherent delays between when a change in glucose is detected and when insulin infused through a pump takes effect. Unlike the pancreas, which delivers insulin directly into the bloodstream, most pumps deliver insulin just beneath the skin.

To keep glucose in check, the closed-loop system must somehow recognize when a meal is occurring and increase insulin infusion. Too much insulin, however, can cause low blood glucose, a condition called hypoglycemia. Turning up the insulin pump 30 minutes before mealtime helps control glucose spikes but requires action from the user.

Another big issue for closed-loop systems is adjusting for exercise, said Nelly Mauras, M.D., chief of endocrinology at the Nemours Children’s Clinic in Jacksonville, FL, and a principal investigator for the Diabetes Research in Children Network (DirecNet). DirecNet is an NIH-sponsored, multicenter study group evaluating continuous glucose monitoring devices in children. Physical activity can cause rapid drops in blood glucose and has residual effects on blood glucose levels for up to 16 hours afterward. Based on DirecNet studies, glucose levels measured at the start of exercise are the best predictor of hypoglycemia occurring during physical activity and at night following physical activity, according to Mauras.

Counter-regulatory responses to repeated drops in blood glucose are severely blunted—epinephrine and growth hormone do not prevent hypoglycemia, and glucagon and cortisol do not rise during hypoglycemia. Mauras said closed-loop systems should be equipped with alarms to alert users of severe hypoglycemia and should ultimately be able to predict and avert post-physical activity drops in blood glucose.

Infusing glucagon, a hormone that counters insulin by freeing glucose stores in the body, is a strategy that would help avoid sudden, severe hypoglycemia, said Edward Damiano, Ph.D., a biomedical engineer at Boston University. Damiano presented data from a diabetic pig model showing infused glucagon’s ability to avoid hypoglycemia without the ingestion of rescue carbohydrates.

Although some at the meeting believe glucagon adds an unnecessary layer of complexity—requiring a dual chamber pump that infuses insulin and glucagon—Damiano said that simply stopping insulin infusion is not enough to avoid dramatic drops in blood glucose. “I think it’s difficult to imagine a system that could be effective without some counter-regulatory effect,” he said. “Glucagon is very fast.”

The performance of a fully implantable closed-loop device was reviewed by Howard Zisser, M.D., director of clinical research at the Sansum Diabetes Research Institute in Santa Barbara, CA. The device, about the size of a hockey puck, is implanted under the skin on the abdomen. Similar to a heart pacemaker, leads are tunneled under the collarbone and a glucose sensor is inserted in the vena cava.

Zisser said intravascular sensors require only once-a-month calibration, whereas subcutaneous sensors must be calibrated once every few days due to protein buildup. Although internal systems require surgeries with significant recovery times to insert and remove the pump and are subject to many of the same issues as external systems, “with these devices patients report they forget they have diabetes…once patients had them in, none wanted to take them out,” said Zisser. Research on fully implantable systems is currently on hold due to a halt in manufacturing the device.

Perfecting Algorithms

Perfecting algorithms may be the biggest hurdle to a fully closed-loop artificial pancreas. Algorithms are the brains of the artificial pancreas that translate glucose monitoring data into appropriate insulin dosing. Performing an infinitely complex task, algorithms must account for factors that result in steep climbs or drops in glucose levels, such as eating, physical activity, and sleep.

A quick lesson in Algorithms 101 was presented by Francis J. Doyle III, Ph.D., an engineer at the University of California, Santa Barbara. Akin to the way the chess computer Deep Blue predicts an opponent’s strategy based on past moves, a suitable algorithm would facilitate insulin dosing based on past responses, according to Doyle.

A perfect system would recognize when a meal is taking place and then, based on the algorithm, deliver the appropriate amount of insulin to maintain glucose control. But knowing which parameters to include in the algorithm, such as the rate of change in glucose levels, is complex and is the focus of computer-based “in silico” testing—a process whereby mathematical models based on real patient data are used to test algorithms.

Patient Considerations

Tim Wysocki, Ph.D., a psychologist at the Nemours Children’s Clinic in Jacksonville, FL, outlined possible behavioral impediments to patient acceptance of an artificial pancreas, such as aversion to being “controlled” by a device or, in the case of a child, parents’ unwillingness to entrust a device with their child’s safety. Wysocki envisioned possible negative repercussions of the artificial pancreas, such as unwanted weight gain.

“For people who have lived with diabetes for a decade or more with continuous constraints on what and how much they can eat, a device that responds automatically and instantly with an appropriate insulin dose in response to whatever is eaten might create a significant source of temptation for many patients,” he said.

Diabetes advocate Kelly L. Close, founder of the health care information firm Close Concerns, Inc., and editor of the patient newsletter diaTribe, put a human face on the disease. Close, who has had type 1 diabetes the majority of her life, said the goal of the artificial pancreas should be to reduce but not necessarily eliminate severe hypoglycemia—often the biggest fear for people with type 1 diabetes and their families. Close said setting unrealistically high expectations for the artificial pancreas may unintentionally deprive people of a lesser device that could save lives: “If we aim to get rid of [hypoglycemia] completely, we’re in danger of moving too slowly or of research coming to a complete standstill.”

Curing Diabetes

Workshop attendees seemed unanimous in the view that in some shape or form the artificial pancreas is technologically feasible and must remain aggressively pursued. However, it remains only part of the bigger picture.

As Aaron J. Kowalksi, Ph.D., director of metabolic control research at the JDRF, put it, “the ultimate goal is to walk away from devices and cure diabetes.” For more information about diabetes, visit


NIH Publication No. 09–4562
December 2009



Diabetes Home | Diabetes A to Z | Introduction | Treatments | Complications | Statistics | Clinical Trials | NDEP | Awareness and Prevention | Additional Resources | Order Publications | About Us | Información en Español

Contact Us | Health Information

The NDIC is a service of the National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health.

National Diabetes Information Clearinghouse
1 Information Way
Bethesda, MD 20892–3560
Phone: 1–800–860–8747
TTY: 1–866–569–1162
Fax: 703–738–4929

Privacy | Disclaimer | Accessibility | PDF versions require the free Acrobat® Reader® software for viewing.
H H S logo - link to U. S. Department of Health and Human Services
NIH logo - link to the National Institute of Health
NIDDK logo - link to the National Institute of Diabetes and Digestive and Kidney Diseases