Mass General clinicians, scientists and engineers from Boston University are building a breakthrough tool to save type 1 diabetics from the dangerous highs and lows of their disease.
An artificial pancreas system, currently in the first phase of clinical trials at Mass General, holds tremendous promise.
By the time David Damiano was 13 months old, he was hooked up to his first insulin pump. Now, 10 years later, the four-ounce device has only rarely left his side, continuously pumping insulin to his body through a thin, plastic tube inserted just under the skin on his abdomen. Without uninterrupted access to the insulin in his pump, David cannot survive.
David has type 1 diabetes, a chronic condition that develops when the body’s immune system destroys its own pancreatic beta cells — the only cells that produce insulin, a hormone needed to absorb nutrients, such as sugar, into cells and convert them into energy. In a person without diabetes, the beta cell watches carefully over the glucose levels in the blood. As the body digests a meal, the level of glucose — the simple sugar found in food — rises in the bloodstream. The pancreas normally responds by releasing insulin into the bloodstream which transports the glucose into the body’s cells. These cells depend on the energy they derive from glucose, without which they would starve and die.
The stakes of this simple physiological process are high, as any person with diabetes will attest. Without enough insulin, hyperglycemia (high blood sugar) can lead over time to complications causing kidney damage, cardiovascular disease, blindness and amputation. Too much insulin, on the other hand, causes hypoglycemia, triggering seizures, loss of consciousness and, rarely, brain damage and even death. For most of us, a normally functioning pancreas protects our bodies from these dangers, ceaselessly measuring our blood sugar while we work, play, eat and sleep, and giving us just the right amount of insulin.
For the more than 23 million people in the U.S. living with diabetes, five to 10 percent of whom have type 1 diabetes, walking the tightrope between too much and too little blood sugar is a full-time job. Fortunately for him, David’s parents have taken extraordinary measures to help him in his struggle to stay healthy.
Between the two of them, David’s pediatrician mother, Toby Milgrome, MD, and his biomedical engineer father, Edward Damiano, PhD, have checked their son’s blood sugar at least once every two hours — night and day — since he was diagnosed at age one. In addition to his battery-operated insulin pump, 11-year-old David totes around a comparably sized continuous glucose monitor, or CGM. The CGM device, also connected by a tiny catheter to the wearer’s abdomen, measures the body’s glucose level every 10 seconds and calculates an average for these measurements every five minutes for up to 72 hours.
Yet David’s parents still worried they weren’t doing enough to maintain his health, dreading the day when their son heads off to college with no one to rouse him for a quick blood test in the middle of the night.
Approximately eight years ago, Dr. Damiano faced these fears for David head on. Collaborating with colleague Firas El-Khatib, PhD, he began developing an artificial pancreas for persons with type 1 diabetes, first at their lab at the University of Illinois and soon after that at the Boston University (BU) Department of Biomedical Engineering, where both researchers specialize in the field of biofluidics. Biofluidics is the study of the fluid mechanics of blood vessels, the inner ear and other physiological structures.
Today, their artificial pancreas system — comprising a laptop, a CGM, an insulin pump and another pump for glucagon, a hormone secreted by the pancreas to raise blood sugar levels — is making automatic blood glucose control a reality. In its present state, as a jumble of interconnected medical devices made by different companies, the system itself is not much to look at. By the time it hits the market, the entire device will most likely be about the size of a cell phone. But, it’s the elegant algorithm driving the artificial pancreas system that has the type 1 diabetes community poised to celebrate this project as the next big breakthrough in diabetes care.
“Our system takes each component of current technology and gets them talking to one another directly, closing the loop and relieving the user of the burden of calculating and deciding when to pump insulin,” explains Dr. Damiano. “The user can override the system, but barring that, the artificial pancreas regulates blood sugar independently and far more frequently than a user would.”
Adds Dr. El-Khatib, “The algorithm — equipped with a mathematical expression for estimating the rate at which insulin is absorbed into the blood — will drive the device’s decisions.” Dr. El-Khatib, who wrote the software for the artificial pancreas as his doctoral thesis, has worked to refine the program over the last few years as a research fellow funded by the Juvenile Diabetes Research Foundation. “The algorithm is objective, relentless and watchful on the user’s behalf; it will never get tired or irritable like people do.” And for people with type 1 diabetes and their families, that means that each calculation is more accurate, each therapeutic decision is safer and overall control of blood sugar levels is tighter in the long run.
By 2006, Drs. Damiano and El-Khatib knew that their system worked in diabetic pigs, but neither researcher had been involved in human clinical trials before and the pair weren’t sure how to move their project to the next stage. Dr. Damiano went public with their preliminary results at a conference at Boston’s Joslin Diabetes Center, where he captured the attention of young endocrinologist Steven J. Russell, MD, PhD, from the Massachusetts General Hospital Diabetes Center.
“When I heard Dr. Damiano’s talk, I just knew that this could really work,” recalls Dr. Russell. “I didn’t waste any time — I walked right up to him and said, ‘I want to help you move this toward human trials immediately.’”
Boston is rich in diabetes care and research centers, but Dr. Damiano recognized that the Mass General diabetes team had what it took to catapult the artificial pancreas forward. The BU researchers and MGH clinicians partnered and began the complicated preparations for human safety trials a mere few weeks later. With the guidance of David M. Nathan, MD, director of the MGH Diabetes Center and the MGH Clinical Research Center, the team mapped out the clinical trial protocol and launched phase I trials by mid-2008.
“Mass General offers something truly unique: a vast, university-sized set of laboratories filled with investigators conducting basic research right next door to top-notch clinician-scientists with access to large patient populations,” says Dr. Nathan. “Because these two groups of experts are so seamlessly integrated at MGH, this is the perfect setting for testing and refining homegrown devices, getting them into humans first and revolutionizing care.”
“Mass General provides the perfect clinical research setting,” agrees Dr. El-Khatib. “Without the MGH team, we could only take our hardware and algorithm so far. This collaboration makes the perfect match, and we’re like family now.”
The Mass General-Boston University partners wrapped up phase I trials a year ago, then published the results last spring — declaring the test run a success. The artificial pancreas system achieved near-normal blood glucose levels for a group of 11 patients over a period of 27 hours, while competing devices from around the country left some subjects dangerously hypoglycemic.
Two unique characteristics of Dr. Damiano’s device contributed to its ability to steer patients safely through the ups and downs of the body’s sugar processing. First, in order to protect patients from insulin-induced hypoglycemia, Drs. Damiano and El-Khatib incorporated glucagon into their system. In a healthy pancreas, glucagon serves to counter the effects of insulin — raising blood sugar levels when insulin depresses them by too great a degree. Though type 1 diabetic patients are able to produce glucagon, their pancreas often does not secrete the hormone or enough of the hormone when the body needs it, and diabetic patients regularly carry emergency doses of glucagon in case of insulin overdose. By building glucagon into their artificial pancreas, the engineers have established yet another line of defense against the hypoglycemic states so common in type 1 diabetic patients.
Also setting Dr. Damiano’s system apart is the nature of the algorithm at its heart.
“Unlike other prototypes out there, this algorithm does not have to be calibrated to each individual user — you simply plug in a person’s weight, and it’s ready to go,” explains Dr. Russell. More importantly, the algorithm is extremely sophisticated in the way it doses insulin, allowing the artificial pancreas to accommodate huge fluctuations in insulin absorption in real time.
“We were surprised to discover the enormous range of insulin absorption,” says Dr. Damiano, “not just among our group of volunteers at Mass General, but within each individual.” Indeed, analysis of the subjects’ blood-insulin levels revealed a fourfold difference in the rate at which individuals absorbed insulin and a twofold difference on an individual basis across the experiment. Yet the algorithm was robust enough to adapt to these unexpected findings, making safe dosing decisions for each subject and reducing the risk of hypoglycemia.
Until a cure for type 1 diabetes is discovered, eliminating hyper- and hypoglycemic episodes will take the teeth out of the disease. According to a landmark study led by Dr. Nathan in the 1980s and 90s — the Diabetes Control and Complications Trial (DCCT) — keeping blood sugar as close to normal as possible can mean all the difference to long-term health. It’s Dr. Nathan’s DCCT statistics, in large part, that motivate the Damianos and other diabetic patients and their families to search tirelessly for tighter and more precise methods of blood sugar control.
“The type I diabetic patients I treat now are incredibly healthy compared to the previous generation of sufferers,” says Dr. Nathan. But their health, he adds, comes at the cost of constant attention to their disease. And even with the best care, many persons with type 1 diabetes experience at least one episode of mild hypoglycemia each day. “The artificial pancreas is a way to keep them healthy and give them back a more carefree lifestyle, to minimize the daily inconvenience of self care and reduce the risk of hypoglycemia and of complications they’ve coped with since childhood.”
Phase II clinical trials were begun in the summer. Over the next year, they will test the effectiveness and safety of the artificial pancreas in a larger, more varied patient population at Mass General. Future phases of the human trials will see the development of a portable, wearable version, controlled wirelessly by a handheld computer device, as well as nationwide outpatient experiments allowing subjects to go through their daily routines at home while attached to the system. By the time the team’s artificial pancreas is available for purchase, it will be a small device, weighing only slightly more than the average insulin pump, with a digital screen indicating blood glucose levels, hormone reservoir status, and a map of regulatory actions taken by the system.