CHICAGO (Reuters) - It's a dream of medical science that looks tantalizingly within reach: the artificial pancreas, a potential breakthrough treatment for the scourge of type 1 diabetes.

Meant to mimic the function of a real pancreas, the artificial version is a complex device that combines a pager-sized continuous glucose monitor and sensor that tracks blood sugar with a pump that automatically delivers the correct dose of insulin at just the right time.

That technology could make a major difference to the three million Americans with the disease who must vigilantly monitor their blood sugar, even at night, and risk deadly consequences if they are slow to notice a dangerous change.

But it is caught up in America's long-running tug of war between supporters of more rapid medical innovation and those who seek better safety for new devices. A fresh confrontation is about to break open this week as the U.S. Food and Drug Administration lays out a path toward regulatory approval for such devices, expected as early as Thursday.

Advocates of the artificial pancreas fear the bar will be set too high in terms of how the devices can be tested in patients, to what lengths companies will need to go to prove they are safe and whether a change in even one of its components will require a whole new round of testing.

Based on the FDA's track record with an early version of these devices, they expect the guidelines could delay access to the technology for many years.

People on both sides of the debate have a personal stake in the outcome.

Jeffrey Brewer, president and chief executive of the Juvenile Diabetes Research Foundation, a powerful advocacy group, has a teenage son with type 1 diabetes.

He notes that medical device giant Medtronic has a very early version of an artificial pancreas called the Paradigm Veo insulin pump, sold in 50 countries but not in the United States.

The pump has an automated safety feature, called low glucose suspend, that shuts off the insulin flow when glucose falls dangerously low.

The Veo, and newer devices being developed, are meant to be worn outside the body, but are connected to patients through a tiny catheter placed just under the skin.

That device could have prevented Brewer's son from overdosing on insulin and spending two days in the hospital. Brewer says cumbersome FDA regulations in June requiring proof of safety and effectiveness for low glucose suspend, already widely used elsewhere, will set back U.S. approval for at least 2-1/2 years.

He is worried the agency will take a similar approach when it issues final guidance, expected this week, on the more complex artificial pancreas.

"The biggest bottleneck right now is the FDA," says Brewer, who charges that the agency is slowing down progress.

Charles "Chip" Zimliki, who heads an FDA initiative to speed up availability of an artificial pancreas, defends the agency's caution over a device on which many lives will depend. A type 1 diabetic himself, Zimliki says he wants to see these devices approved.

But the FDA has been sharply criticized for approving devices that proved unsafe, such as artificial hips recalled last year by Johnson & Johnson. The agency is unlikely to skip any steps to get a device as complicated as an artificial pancreas to patients.

"The FDA wants the artificial pancreas on the market as much as anyone else does. We just have to operate within U.S. laws to make sure it is safe and effective," Zimliki said in an interview earlier this year.

A Reuters analysis of significant complaints to the FDA over the Medtronic Paradigm Veo found about two dozen reports connected to the device, including one incident involving an overdose of insulin that was described as life-threatening and at least two others that sent patients to the hospital with high blood-glucose levels. The reports, dating back to July 2009, mostly did not specify the cause of malfunction.

Medtronic said it was not aware of any problems specifically related to its low glucose suspend feature. It said the devices have a high safety profile and that two dozen complaints would amount to one per 3,000 pumps.

GETTING IT RIGHT

Brewer's passion is born out of the fear and frustration shared by many parents of children with type 1 diabetes, in which the immune system destroys cells in the pancreas that make insulin.

Type 1 diabetics currently must monitor their blood sugar and take insulin several times a day. Getting it right is tricky. Meals, exercise, stress, even a growth spurt, can alter blood sugar. And mistakes are costly.

Taking too little insulin can cause high blood sugar, increasing the risk of long-term complications such as eye damage, kidney failure and heart disease. Taking too much can cause blood sugar to drop, causing hypoglycemia, which can result in seizures, unconsciousness, brain damage and death.

That was the case with Brewer's son Sean. About 16 months ago, the 14-year-old decided to have a late-night snack.

"He forgot he had given himself insulin and gave himself another dose," Brewer said. "Unfortunately, he was behind a locked door, so the fire department came with an axe and broke the door down. He spent the next 36 hours in the intensive care unit."

For about a day, Brewer's son could not recall his name.

Sean recovered fully from this serious episode of hypoglycemia, but many families are not so lucky.

"People are hurting themselves and killing themselves all the time with insulin," Brewer says. "The data says one out of every 20 people with this disease might actually die with hypoglycemia."

That's where a small cellphone-sized device like Medtronic's Veo would help, Brewer said. The JDRF organization has been pushing hard to get the FDA to allow these devices to enter the U.S. market without a lot of additional testing. The group, along with other diabetes advocates, have been ramping up the pressure.

In October, the American Diabetes Association, the American Association of Diabetes Educators, the Endocrine Society and the American Association of Clinical Endocrinologists sent a letter to FDA Commissioner Margaret Hamburg urging the agency not to repeat the low glucose suspend decision.

In early November, they sent a petition to the FDA with 100,000 signatures from lawmakers and diabetes experts. They followed up with ads in the New York Times and the Washington Post featuring a young girl named Piper who, like other type 1 diabetics, is at risk of dying from hypoglycemia.

The ad said the new artificial pancreas guidelines are the FDA's chance to show the world the United States "is leading in medical innovation, not standing in the way."

SLOWING DOWN RESEARCH

It's not just the JDRF that is complaining.

A survey of 150 firms from the National Venture Capital Association, released in October, shows that venture capitalists have been cutting back on their investments in medical-device companies, blaming regulatory hurdles at the FDA.

Investment in medical devices last year fell to $2.38 billion from $2.62 billion in 2009 and $3.52 billion in 2008, according to the survey.

While U.S. researchers have been allowed to test artificial pancreas systems in carefully supervised hospital studies, they need to be tested outside of the clinic in real-world settings to see how well they work. That has forced many to seek research partners abroad.

"We have needed to collaborate offshore and continue to do so just because it was taking us so long to get these regulatory approvals," said Dr. Howard Zisser of the Sansum Diabetes Research Institute in Santa Barbara.

Zisser and Frank Doyle, a chemical engineer at University of California, Santa Barbara, have developed the "brain" portion of an artificial pancreas system -- a software program that uses blood sugar readings sent wirelessly from a continuous glucose monitor and tells the pump how much insulin to give.

So far, the software from this team can work with pumps from Insulet Corp and Animas, a unit of Johnson & Johnson. And it works with sensors from Abbott Laboratories, Dexcom Inc and Roche Holding. That allows components from different device makers to become part of one artificial pancreas system, a new approach for medical technology.

In October, Zisser, Doyle and other members of an international consortium won a $4.5 million National Institutes of Health grant to test the technology in challenging settings. The group has run two pilot studies in Italy and in France.

In these trials, study volunteers placed the control of their disease in the hands of the device while they dined in a restaurant and slept overnight in a hotel. Research teams huddled in nearby rooms wirelessly monitoring their progress.

Zisser says it will be challenging for the FDA to feel confident with these kinds of trials. But there has been progress. The agency last month gave the green light for Medtronic to begin an in-home clinical trial of a low glucose suspend system in the United States.

Dr. Francine Kaufman, former president of the American Diabetes Association and chief medical officer for Medtronic's diabetes business, says the company worked hard with the FDA to reach agreement on the trial design.

It will not be easy. The company must find a group of diabetics who are especially prone to hypoglycemia and they will try to show that the device is not only safe, but that it reduces such episodes.

"We've agreed to do it. We'll start that trial and get our way down the path," Kaufman said.

STUMBLING BLOCKS

The potential stumbling blocks go beyond the design of clinical trials. Researchers and medical device executives say they still have plenty of work to do to make sure they have a system that functions properly.

While the technology has come a long way from the refrigerator-sized machines developed in the 1970s for hospital use, many of the systems are still using laptops to run the software program instead of more portable devices.

Medtronic Chief Executive Omar Ishrak said the company is committed to developing the technology, but there are many technical and clinical hurdles yet to cross. He says it will take a decade or so before a system is ready.

"It's early enough for me to not even think about products," Ishrak said. "We can do demonstrations and we've done clinical studies here and there, but we don't have a product."

Boston University's Ed Damiano is part of a team developing an artificial pancreas system that uses both insulin and glucagon, a hormone released by the pancreas to raise blood sugar levels when they drop too low. He says blaming the FDA will not help scientists work out the kinks in the devices.

Damiano, like Brewer and Zimliki, has a personal stake in getting an artificial pancreas approved.

His son David, now in middle school, developed type 1 diabetes when he was 11 months old. Since he was a baby, Damiano has crept into his son's room several times a night to check on David's blood sugar.

David's blood sugar levels have been in the normal range for most of his life. Damiano wants him to have a system that helps ensure his good health when David goes off to college.

"We've got seven years," Damiano said. "It's important."