Balancing Act

Stuart Weinzimer majored in molecular biophysics and biochemistry as a Yale undergraduate and attended the Albert Einstein College of Medicine in New York. In 1995 he accepted a fellowship in pediatric endocrinology—the study of the body’s hormone-secreting glands—at the Children’s Hospital of Philadelphia. “I liked the science of endocrinology,” he says. “It’s beautiful and mathematical and makes perfect sense.”

But at the University of Pennsylvania School of Medicine, where he spent several years doing research after the Philadelphia fellowship, Weinzimer found himself intrigued by juvenile diabetes (now called type 1 diabetes) precisely because, as he says, it “makes no sense. You can’t reduce it to easily digested treatments.” The more insoluble the problem of insulin management seemed, the more Weinzimer wanted to crack it. When he treated young patients in the clinic, he saw firsthand the outsized impact the disease had on their day-to-day existence. What if, he mused, someone could come up with a reliable way to automate diabetes management—to free up patients’ time and energy for the business of living life?

Yale has a long record in diabetes research. The first insulin pumps for children were tested in the late 1970s at the Yale School of Medicine by William Tamborlane and Robert Sherwin. In 2002, Weinzimer joined Yale as an assistant professor of pediatrics in the medical school and began working with Tamborlane. (Weinzimer is now an associate professor.) Around that same time, he heard that Medtronic, which manufactured the first commercial insulin pump, was developing a closed-loop artificial pancreas device for adults. His brain started to race. Could the device work for his under-30 patients?

The closed-loop system didn’t work very quickly, and Weinzimer worried that time lags in the absorption of insulin could put his young patients in danger. “One of the technical problems with the first generation” of the device, he explains, “is that because you’re delivering medicine subcutaneously, you’re always going to be a little behind.” Insulin delivered through the skin takes longer to absorb than insulin delivered directly into the bloodstream, and the resulting delay—as much as an hour—could potentially affect children and teenagers differently from adults. But Weinzimer thought he could address the problem by allowing patients to give themselves small “priming” doses manually, using a button on the insulin pump, just before eating. The closed-loop system would manage the rest.

Working with Garry Steil, a Medtronic engineer now at Boston Children’s Hospital, Weinzimer designed a clinical trial, and in 2008, they published data from the first 17 participants in the journal Diabetes Care. (The trials are funded by the Juvenile Diabetes Research Foundation and the National Institutes of Health and conducted in partnership with Medtronic.) Teenagers and young adults who wore the artificial pancreas achieved better control over their blood glucose levels than they’d had before trying the device. Some participants’ glucose levels fluctuated after they ate, but in the trial group that received doses before meals, the fluctuations were smaller.

Media outlets all over the world showered Weinzimer with attention after the data appeared. But he knew he still had to deal with the problem of exercise. The artificial pancreas is programmed to cut back on its normal background rate of insulin delivery whenever a patient’s glucose levels fall below a predetermined level. This helps ensure that wearers’ blood sugar never falls dangerously low during sleep, for instance—when one in three diabetes patients suffer from low blood sugar. But type 1 diabetics’ blood sugar often plummets after exercise, and the plunge can take place just 15 minutes after the start of biking, swimming, running, or other intense activity. Weinzimer was concerned that the detection mechanism might not work quickly enough to compensate. He designed a second trial to find out—the trial that had brought Ben Yarmis to Yale–New Haven. This study, which includes adults up to age 30, is still under way.

One of the first patients to sign up for the exercise study was Tyler Wolf, a 27-year-old financial analyst at Google. Diagnosed with diabetes as a young teen, Wolf initially rebelled against the regimen the disease imposed on him. “I was very stubborn. I would even deny friends when they told me, ‘You should check your blood sugar,’” he says. “There were a few instances when I was out of body, almost delusional, especially when I was under a lot of stress at school. One instance, I was driving home and got pulled over two blocks from my house. The police thought I was intoxicated.”

Wolf took part in Weinzimer’s second trial in the spring of 2009. Like Yarmis, he made two visits to Yale–New Haven. On the first, he followed his usual maintenance routine so the doctors could get an idea of his typical glucose levels. On the second, he wore the artificial pancreas for three days, exercising intensely for brief periods. The effortlessness of his 72 hours with the device thrilled Wolf. “It was an amazing respite from all of the day-to-day things I have to deal with,” he says. “The control during the night was excellent. I woke up feeling perfect.”