For millions of diabetes sufferers, life is a constant battle to keep their blood sugar balanced, which typically means they have to test their glucose levels and take insulin throughout the day. A new generation of “artificial pancreas” devices may make tedious diabetes micromanagement obsolete. In healthy people, the pancreas naturally produces insulin, which converts sugars and starches into energy. People with type 1 diabetes, however, do not produce any insulin of their own, and those with type 2 produce too little. All type 1 and many type 2 diabetics have to dose themselves with insulin to keep their bodies fueled—and doing so properly requires constant monitoring of blood sugar because appropriate dosages depend on factors such as how much patients eat or exercise. Stuart Weinzimer, an endocrinologist at Yale University, has devised an artificial pancreas that combines two existing technologies: a continuous glucose monitor, which uses an under-the-skin sensor to measure blood glucose levels every few minutes, and an insulin pump, which dispenses insulin through a tube that is also implanted under the skin. The glucose sensor sends its data wirelessly to a pocket computer a little bigger than an iPhone that is loaded with software developed by Minneapolis-based Medtronic. The program scans the incoming data from the glucose monitor and directs the pump to dispense the correct amount of insulin.
At an American Diabetes Association meeting in June, Weinzimer and his colleagues reported that 86 percent of type 1 diabetics they studied who used the artificial pancreas reached target blood glucose levels at night, whereas only 54 percent of subjects who had to wake up to activate an insulin pump reached their target levels. Other, similar systems are in development at Boston University, the University of Cambridge and Stanford University.
Some technical glitches still need to be worked out. For example, the device occasionally has trouble adapting to drastic changes in glucose, such as those that occur after exercise. And it will have to go through more rounds of vetting, which could take years, including large-scale patient trials that will be required before the Food and Drug Administration can approve the device. Nevertheless, Weinzimer says that the enthusiastic responses he has gotten from his trial participants remind him why the long slog toward commercialization is worthwhile.
This article was originally published with the title Putting Diabetes on Autopilot.