Research Centers Report “Encouraging Results” in Tests of Artificial Pancreases
Even as diabetes researchers worldwide strive for total control over-or even an outright cure of-type 1 diabetes via gene therapy, altered cells, or surgical intervention, other researchers continue to press toward creation of a functional “artificial pancreas.”
The goal is a machine that, like a natural pancreas, adjusts a person’s glucose levels subtly and seamlessly throughout the day, free of the need for constant manual adjustments and fears of over- or under-dosing.
The elements of an artificial pancreas are simple: You take a fail-safe insulin pump and connect it via a fail-safe communication method with a fail-safe continuous glucose monitor. When the CGM detects a too-high blood sugar level, it tells the pump to inject X mg of insulin, calculated by means of a fail-safe algorithm.
Aside from the technology, which has to be virtually perfect and isn’t yet, a big drawback to the concept is that every person with diabetes has a different metabolism and a different threshold of insulin tolerance. Consequently, current CGM/insulin pump partnerships can’t really “know” the exact dose a patient needs. Instead, the patient must tediously monitor readings throughout the day and manually adjust the pump dosages.
Progress is being made, however. Results of an experiment with five type 1s at the University of Virginia (with the results duplicated on three patients at the Padova University in Italy) hold out the promise that the artificial pancreas may be a big step closer to reality.
The University of Virginia scientists, led by Boris Kovatchev, an associate professor of psychiatry and neurobehavioral sciences and systems and information engineering, created a “smart program” based on an algorithm that allows for the differences in each patient’s metabolism.
The resulting system, which links a very smart CGM to an insulin pump, provided “very encouraging results,” said Kovatchev, who enjoys global renown for his ability to apply sophisticated computational methods to diabetes research. The system efficiently maintained each patient’s blood glucose at optimum levels throughout the day and achieved excellent overnight control with no incidences of hypoglycemia.
The two universities are among seven sites worldwide where the Juvenile Diabetes Research Foundation is funding research into advanced algorithms for controlling blood glucose levels that could lead to a workable artificial pancreas.