'Closed-loop' insulin delivery is within reach.
While "open-loop" insulin pumps have been on the market for more than 20 years, the first continuous glucose monitors (CGMs) to display real-time values to patients--Medtronic's REAL-Time and DexCom's STS--only became available in 2006. Although Medtronic's system allows the sensor readings to be displayed on the insulin pump, the two technologies remain essentially separate. A reliable connecting algorithm that allows the output of the sensor to control the pump will be required to develop a true "artificial pancreas" that maintains consistently normal glucose levels with little or no input from the patient.
That's still a challenge, but progress is being made, Dr. Christopher D. Saudek and Dr. Barry H. Ginsberg said during a workshop on diabetes technology at a meeting on diabetes sponsored by Ohio State University.
"For several decades, pumps have been available. Sensors have been the holdup. In the last 5 years, the CGM has finally come to function," said Dr. Saudek, director of the diabetes center and the inaugural Hugh P. McCormick Family Professor of Endocrinology and Metabolism at Johns Hopkins University, Baltimore. "For the first time, a closed loop system is a realistic possibility."
Dr. Ginsberg, formerly vice president for worldwide medical affairs for the diabetes division of Becton, Dickinson and Co. and now a private consultant in Wyckoff, N.J., quoted one estimate from a group of independent experts who predicted that a functional closed-loop system may be available as soon as 2013.
But he believes that projection--contained in a 2005 report on CGM technology from the New England Healthcare Institute (www.nehi.net)--"may be a hair optimistic for a full system. Partial systems may be available by then," Dr. Ginsberg said in an interview.
Indeed, said Dr. Saudek, sensor accuracy and robustness will need to improve first. Beyond that, two issues still to be resolved are whether the system would require the patient to somehow signal before eating a meal (thus technically leaving the loop slightly open), and whether a true artificial pancreas would need to secrete glucagon in addition to insulin.
A major challenge to development of the connecting algorithm between the sensor and the pump concerns the lag time between interstitial fluid (ISF) glucose that is being sampled by the CGM, and the plasma glucose. Although some reports have estimated that lag to be around 20 minutes, a study by Dr. Saudek and his associates suggested that while the ISF glucose does significantly lag behind the plasma glucose regardless of glycemic direction, the time may be only around 4-10 minutes (Diabetes 2003;52:2790-4). "Lag may be an overrated problem in driving a closed-loop system," Dr. Saudek commented.
But Dr. Ginsberg said other data have shown that the lag may differ, depending on whether the glucose is rising or falling. Not only is the lag typically much longer when glucose is rising than when it's falling; when glucose is dropping, the ISF may actually lead the blood glucose. "Any algorithm would have to account for that."
First-phase insulin response is another critical issue. A normal pancreas immediately secretes insulin into the blood after a meal when there is a rise in glucose, but with current insulin, the insulin would not reach the blood in substantial amounts for 60-90 minutes. Data from several studies suggest that a so-called hybrid system, in which a small amount of anticipatory insulin is given prior to a meal, results in more normal postprandial blood glucose values than do fully closed-loop systems in which insulin is secreted in response to a rise in glucose. "I think that the first-phase insulin is going to have to be reproduced some way.... It may end up being a signal to the device that you're about to eat, so the pump can begin delivering insulin a little bit earlier," Dr. Saudek said.
As for the glucagon issue, Dr. Ginsberg noted that about half of approximately 400 experts who were surveyed during the recent diabetes technology conference said they believed a closed-loop system would need to include glucagon, while the other half thought it wasn't necessary. Dr. Ginsberg believes it is. "The body uses a dual system with insulin lowering blood glucose, and glucagon, epinephrine, growth hormone, and cortisol raising it. Although good control can probably be reached without glucagon, it will be needed for very good control and for safety."
But Dr. Saudek is on the other side. "I don't think the glucagon in a pump is really a big issue .... I think it's unlikely to be a needed or realistic factor."
Exercise is another variable that the system will have to take into account, although it is probably an easier one than food. The patient could either signal that he or she was about to exercise or wear a device that could detect body motion, Dr. Ginsberg noted.
Insulin pharmacokinetics is another issue. The current rapid-acting analogs on the market--lispro, aspart, and glulisine--are still too slow, not peaking until 60-90 minutes, with no first wave. A never generation of ultrarapid-acting analogs currently under development could overcome this problem by acting so quickly that they actually replicate first-phase insulin secretion. The first of these to hit the market is likely to be Viaject, made by Biodel Inc. It is currently in phase III trials, with an anticipated New Drug Application filing in 2008, according to the company's Web site. Dr. Ginsberg disclosed that he owns a small amount of stock in Biodel.
Injection with microneedles of 1.5 cm or less--also under development--could speed insulin action too, he added.
Dr. Saudek noted that while an external artificial pancreas is indeed within sight, an implantable system is not. He and his colleagues at Johns Hopkins had worked with implanted insulin pumps in the 1990s and a group in France is still implanting the devices, which were made by Medtronic. Although more than 300 patients in Europe currently have the implants, few have continued to use them in the United States. For a variety of business-related and logistical reasons, Medtronic has decided not to pursue implantable pumps at present, instead choosing to focus on closing the loop externally, according to a Medtronic spokesman.
BY MIRIAM E. TUCKER Senior Writer
|Printer friendly Cite/link Email Feedback|
|Author:||Tucker, Miriam E.|
|Publication:||Internal Medicine News|
|Date:||Jun 1, 2008|
|Previous Article:||Adherence to guidelines improves melanoma outcomes.|
|Next Article:||Fetuin-A is risk factor for type 2 diabetes in elderly.|
|Driving innovation in diabetes care: new diabetes treatment and monitoring technology means opportunity for both patients and medical device...|
|Stepwise progress made toward artificial pancreas.|
|Level of success for diabetics.|