The first human trials of the latest design of an artificial pancreas for people with type 1 diabetes found the device worked without causing low blood sugar (hypoglycemia).
Ideally, this type of automated device would finally free people with type 1 diabetes from the insulin injections that many require each day, while relieving them of the constant need to check blood sugar levels and monitor the food they eat accordingly.
The device, produced through a collaboration of experts from Boston University, Massachusetts General Hospital and Harvard Medical School, delivers two hormones that are deficient in type 1 diabetes — insulin, which keeps blood sugar levels from going too high after a meal, and glucagon, a naturally occurring hormone that prevents blood sugar levels from dropping too low.
Because the device doesn’t rely on human input, it’s called a “closed-loop” system.
“A bi-hormonal closed-loop system is feasible and it can give you good average blood sugar readings,” explained one of the device’s designers, Edward Damiano, an associate professor of biomedical engineering at Boston University, and the father of a son with type 1 diabetes.
“What we’ve developed is automated decision-making software that uses a mathematical formulation to infuse varying amounts of insulin and glucagon when needed,” he explained.
Type 1 diabetes is a disease in which the body’s immune system — which normally protects you from infections and other diseases — turns against healthy cells. In type 1 diabetes, the immune system attacks beta cells in the pancreas, effectively destroying the body’s ability to produce insulin and control blood sugar levels.
What many people don’t realize, however is that beta cells aren’t the only ones damaged by the autoimmune attack. Alpha cells, which produce the hormone glucagon, are also damaged. Damiano’s colleague, Dr. Steven Russell from Harvard, said, “There is a functional deficiency in alpha cells in type 1 diabetes and they don’t work properly. They don’t secrete glucagon as they should, so an extra level of security is lost and you can wind up with hypoglycemia that can be scary and even life-threatening.”
That’s why they decided to add glucagon to their artificial pancreas to give it an added level of protection, said Russell.
In the current version of the device, the researchers tracked blood glucose via a special sensor placed into a vein. Future versions of the device will use currently available continuous blood glucose monitors (CGMs), but for this trial the researchers wanted an extremely accurate way to measure blood sugar levels so that the only variable was the mathematical formulation used to program the delivery of insulin and glucagon.
Eleven people with type 1 diabetes were included in the initial tests, and were studied in 27-hour experiments. During that time, they were hooked up to the artificial pancreas and given carbohydrate-rich meals (carbohydrates are transformed into glucose in the body).
The device responded to the rise in blood sugar levels by administering insulin. In six people, the device achieved an average blood glucose level of 140 milligrams per deciliter (mg/dl), which is well within the American Diabetes Association guidelines for care. However, five people absorbed the insulin much slower than expected, and ended up with low blood sugar levels serious enough to require intervention with additional food.
The researchers were surprised by the significant difference in blood sugar absorption rates, but went back and adjusted the mathematical formulation, and retested the device in a second experiment. This time, they achieved an average blood glucose level of 164 mg/dl, which is slightly higher than the ADA’s goal. However, there were no instances of hypoglycemia that needed intervention.
The researchers said that people using the pump would rid themselves of the need for daily injections. Instead, they might just need to change the pump site every three days, and the glucose-monitoring site once per week. No one-site integration of hormone delivery and glucose monitoring has been developed yet, although that’s the ultimate goal.
In the next trial, the researchers hope to deliver at least some of the insulin prior to a meal, which is the standard treatment. Damiano said this may be accomplished with a pre-meal button, and the user could just choose whether they were having a small-, medium- or large-carbohydrate meal.
The next set of trials will also test a device that includes insulin only, because it would likely be available faster. The reason is that glucagon is currently only FDA-approved in a freeze-dried form as an injectable rescue medication. It’s not FDA-approved yet for delivering through an insulin pump in tiny doses, as it would be in an artificial pancreas. Both Damiano and Russell think it could be possible to have an insulin-only closed-loop system available for use by patients within five years or so.
“The goal of an artificial pancreas is to try to restore normal physiology as closely as possible, and this study demonstrates that this technology is real, and it’s good in real people,” said Aaron Kowalski, assistant vice president for glucose control research and research director of the Artificial Pancreas Project for the Juvenile Diabetes Research Foundation (JDRF). “We’ve talked for many years about the theoretical potential of a closed-loop system, and now we see the real potential. These technologies are going to be built into real systems and will have the potential to transform the management of diabetes.”
Results of the study were published in the April 14 issue of Science Translational Medicine.
* * *