Artificial pancreas has a way to go
As Medtronic prepares to begin the first U.S. trial of a low-glucose suspense system, and as diabetes device companies, researchers and advocates look to the U.S. Food and Drug Administration for guidance, one thing is obvious: The United States is way behind in giving patients access to the Holy Grail of type 1 diabetes treatment, the artificial pancreas.
Ideally, the artificial pancreas would be a small, portable, easily accessible and easily disguisable closed-loop system comprising a continuous glucose monitor, an insulin pump capable of delivering precise amounts of insulin and a computer algorithm to tell that pump just how much insulin to deliver based on blood sugar levels.
The good news is, we already have most of the pieces. Dexcom, Medtronic, Abbott and others already make continuous monitors that alert users when blood sugar levels are becoming too high or low. Medtronic just announced Wednesday that it was enrolling patients in a U.S. trial for its Enlite sensor, a smaller product with supposedly painless sensor insertion and better accuracy. (Sidenote: It received the CE Mark in April.) Animas, Roche, Insulet and Medtronic are just some of the marketers of sophisticated insulin pumps.
But, according to JDRF, continuous glucose monitors are U.S. Food and Drug Administration-approved for use alongside finger sticks, rather than in place of them. And, although some implantable insulin pumps exist, they are expensive and high maintenance, according to pediatric endocrinologist Dr. Stephen Ponder in a column for the Corpus Christi Caller.
“In the near term, the pieces absolutely exist,” said Dr. Aaron Kowalski, vice president for treatment therapies at JDRF. “At least the way I see it, the technology exists today to partially close the loop. We need to add some degree of automation, so there are some technological advances that need to happen.”
Medtronic has the closest thing to an artificial pancreas in the U.S. with its MiniMed Paradigm REAL-Time REVEL System, which comes with an insulin pump, a built-in continuous glucose monitor and personal therapy management software.
In Europe, Medtronic markets the MiniMed Paradigm Veo — an even more intricate device that uses low-glucose suspense, which stops the pump from delivering insulin for two hours when a patient’s blood sugar level gets too low. But the Veo isn’t yet available in the U.S. — in fact, the FDA just approved Medtronic for a new device study for the product less than two weeks ago.
“The obvious next step is to add the algorithm,” Kowalski said. “Data absolutely shows that the computer can control blood sugar better than a human can.”
And researchers are working on just that. Mayo Clinic was scheduled to begin running trials this month on a closed-loop system that uses an abdominal patch to measure blood sugar continuously, a pager-sized pump to deliver insulin beneath the skin and a sophisticated software algorithm using unique patient information. Mayo’s system also takes into account physical activity and metabolic response, which is a critical variable in insulin action.
Similarly, Massachusetts General Hospital researchers are running phase 2 clinical trials on an artificial pancreas system using a continuous glucose monitor, two pumps hooked up to a laptop. The researchers hope to replace the laptop with a computer chip as part of the pump, so that the whole system is the size of a cellphone. This system is unique because it also incorporates a pump for glucagon, a hormone that protects the body from hypoglycemia by raising blood sugar levels when too much insulin is produced or delivered.
Technology-wise, the issue of hypoglycemia is likely the biggest challenge for artificial pancreas devices. Hyperglycemia — or high blood sugar — can lead to complications in the organs if sustained over time. If the low-glucose suspense system, like the one being tried by Medtronic, were to malfunction and shut off the insulin pump too often, hyperglycemia would be the result.
“Why wouldn’t you have a low-glucose suspense?” Kowalski questioned. “If (the pump) was turning off too frequently, glucose levels could get worse, but there’s absolutely no acute risk.”
But allowing a computer to dose insulin means risking over-delivery resulting in too little blood sugar, which manifests more quickly and can cause seizures or brain damage.
“The risk with dosing is significantly more than the risk of turning off a pump,” Kowalski said. “Dosing too much insulin can be life threatening; hypoglycemia can kill them. The concern from the FDA and clinicians is what happens in the closed-loop sensor if the computer is wrong and it sends too much insulin?”
Kowalski added that the first real-world trial of a closed-loop system that involves dosing insulin is taking place in Europe.
The success of the artificial pancreas relies heavily on collaboration between academia and medical device companies. “In the work that we’ve been funding in academia, they’ve been using every different brand of insulin pump and glucose monitor, from Insulet, Roche and even some small companies,” Kowalski said.
GlobalData estimated the type 1 diabetes treatment market at $2.7 billion in 2009, with an expected annual growth rate of 6.1 percent. An estimated 3 million Americans have type 1 diabetes, putting them at risk for long-term complications if they don’t seek treatment.
But that treatment comes at an awfully high cost, which could be another barrier to the ultimate integration of the artificial pancreas for type 1 diabetes treatment. An insulin pump runs anywhere from $5,000 to $7,000 on average, plus about $2,500 in disposable components, according to Ponder. Glucose sensors for continuous monitors cost about $35 to $55 and need to be replaced every several days.
The regulatory barriers
Kowalski said the biggest thing holding back progress now is the U.S. regulatory environment. The JDRF has been working with lawmakers, advocates and healthcare professionals to push the FDA to implement policy changes and guidance that would speed the time frame for devices to go to market and to allow for research into new products.
Meanwhile, the FDA is preparing draft guidance for companies to get the artificial pancreas approved, which Kowalski expects to see released in the next month.
“The pathway to the market is going to be driven by what the FDA asks for on Dec. 1,” he said.
Kowalski expects the low-glucose suspense technology to be available in the U.S. within two years. Globally, he said, a functional artificial pancreas could be just a few years away.
“It’s not helpful for the U.S. to be behind — not just from a humanitarian perspective, which of course is the most important — but also from a business perspective,” he said. “This is very challenging because this is a major market for many of these companies. To invest in research, we have to have access. The FDA has to see that there is a pathway, that these devices can be helpful, and give companies the opportunity to drive innovation.”