What is holding pump design back?

[MarkMunday]

I am a veteran T1 and I have been considering switching from MDI to a pump. But, after having done some research, I was dissapointed in the limited functionality of current insulin pumps.

If I am going to have a gadget attached to me 24/7, it must do everything. The way pumps work at the moment, you still have to test your blood sugar. And you still have to give yourself a bolus infusion if it is too high.

Why shouldn't the device that infuses insulin also decide on how much to infuse and when? A sensor implanted below the skin could regularly send BS readings to the pump. And, based on parameters that are set, it would respond by infusing a suitable amount of insulin.

The fact that there is a time delay between infusion and action of insulin complicates things. You would still want to be able to give yourself a bolus before meals. But, for the rest of the time, the pump should be able to actively manage your blood sugar level.

What is preventing this sort of feedback driven system from being used? Developing a suitable algorithm and programming a microprocessor is not difficult. Please let me know if I am missing something.

Cheers,

Mark

[Belinda]

Mark, When I was looking to upgrade from my older pump to a new one last month I checked on variety of pumps but my best information came from the members here that wear the ones I was looking into. Tony sent me this info about the progress of the pump. Hope it helps.





Research Presented At ADA Annual Meeting Demonstrates Accuracy and Feasibility of Artificial Pancreas Components


Investigational device mimics a pancreas, brings an artificial pancreas one step closer to reality


SAN FRANCISCO - June 17, 2002 - An artificial pancreas has long been considered the Holy Grail of diabetes management, yet many people with diabetes and their physicians believe that it is decades away. However, Medtronic MiniMed, the diabetes management business of Medtronic, Inc. (NYSE: MDT), today announced results of research indicating that a medical device mimicking the human pancreas is closer to becoming a reality. The study, conducted by investigators at the Department of Endocrine Diseases, Lapeyronie Hospital in Montpellier, France, and sponsored by Medtronic MiniMed in Northridge, Calif., was presented during the annual meeting of the American Diabetes Association (ADA).
Long-Term Sensor System Evaluated

The study presented at the ADA annual meeting and conducted in France, "Accuracy of Real- Time Blood Glucose Measurement by Long-Term Sensor System Allows Automated Insulin Delivery in Diabetic Patients," was designed to assess the sensor accuracy over a six-month period and to test automated insulin delivery according to the sensor's signal. Five Type 1 diabetes patients were implanted with a long-term glucose sensor and an implantable insulin pump, which together comprise Medtronic MiniMed's Long-Term Sensor System™. In this system, an implantable sensor, which measures glucose levels using an enzyme-based electrochemical technology, was inserted in the bloodstream near the right atrium of the heart. An abdominal lead was used to connect the implantable sensor to a Medtronic MiniMed Model 2007 implantable pump, which can be programmed to deliver insulin to a patient using a hand-held remote programmer.

At weekly intervals during the six-month study, glucose readings taken by the implantable sensors were compared with the results of at least six daily capillary blood glucose (CBG) measurements, more commonly known as "finger stick" tests. Cumulative data from the implantable sensors strongly correlated with the CBG results, with an r-value of 0.83 to 0.93. The closer the r-value is to 1.0, the higher the direct correlation is between the device and the results measured in blood sugars.

Over a two-day period, investigators also tested the system in a "closed-loop" setting, in which continuous glucose data from the sensor was used to automatically regulate insulin delivery from the implantable pump. In this system, the amount of insulin required to manage a patient's diabetes is calculated using a mathematical algorithm. The researchers found that glucose levels of patients using the implantable sensor were maintained in a near-normal range (70 to 120 mg/dL) more often (42.3 percent of the time) than patients whose insulin requirements were determined by CBG measurements (21.6 percent of the time).

"We are extremely pleased with the accuracy of the implantable sensor over the course of the study, and the promising results of our closed-loop trial," said Eric Renard, M.D., Ph.D., lead investigator of the study and Professor of Endocrinology, Diabetes and Metabolism, Montpellier Medical School, Montpellier I University; Senior Medical Practitioner, Department of Endocrine Diseases of Lapeyronie Hospital in Montpellier, France. "In using a sensor-augmented system, we are essentially creating an artificial pancreas. This type of system is intended to help people who are unable to control their diabetes despite intensive insulin therapy."

"While implantable sensors are still being tested, Medtronic MiniMed's Continuous Glucose Monitoring System, which uses an external sensor, is being used today by clinicians for improved diabetes management," added Alan Marcus, M.D., F.A.C.P., associate clinical professor, University of Southern California Keck School of Medicine, who has extensively studied and authored many articles using the system. "The external sensor, which is generally worn by a patient for up to three days, can provide up to 288 glucose values retrospectively over a 24-hour period, and offers far more comprehensive information than nominal glucose measurements obtained by patients using traditional strips and meters."

Latest Closed-Loop Results

In the fourth and most recent closed-loop experiment, the company combined an algorithm for calculating insulin requirements with a pre-meal bolus (extra insulin delivered by the implantable pump) to address high blood sugar that is normally associated with meal consumption. By delivering a pre-meal bolus, the magnitude and duration of post-meal hyperglycemia (high blood sugar) were greatly reduced without increasing hypoglycemia (low blood sugar). In diabetes patients, hypoglycemia can occur when a patient attempts to lower elevated blood sugars by delivering extra insulin. Over a 24-hour period, hypoglycemia was reduced by more than 50 percent. Moreover, the patient spent nearly 50 percent more time between 70 and 120 mg/dl (the study's stringent target range), resulting in overall improved glycemic control.

"We are delighted with the positive results of this fourth closed-loop study using our Long-Term Sensor System," said Scott R. Ward, Senior Corporate Vice President and President, Medtronic Neurological and Diabetes. "Medtronic MiniMed has a vast amount of experience in conducting closed-loop studies, and we continue to refine our algorithms to further strengthen our knowledge of glycemic control. Our product pipeline includes advanced algorithms that work with both external and implantable sensors and insulin pumps. We are committed to developing an artificial pancreas, in which a system continuously records glucose and delivers insulin automatically, without the patient intervention that is necessary today."

Medtronic MiniMed (www.minimed.com) designs, develops, manufactures and markets advanced infusion systems with a primary emphasis on the intensive management of diabetes. The company's products include external pumps and related disposables; a continuous glucose monitoring system; an implantable insulin pump, which currently is approved for distribution in the European Union, and is under clinical investigation in the United States; and an implantable sensor, which is in clinical trials prior to FDA submission.

Medtronic, Inc. (www.medtronic.com/diabetes-ada), headquartered in Minneapolis, is the world's leading medical technology company, providing lifelong solutions for people with chronic disease. More information about pump therapy and continuous glucose monitoring is available online at www.medtronic.com/newsroom/media_kits_Diabetes.html.

[am1977]

I understand how you feel. I think currently they are working on a sensor type of device being connected to the pump. Actually, I do know that the implanable pump (the one I think will be closest to a cure) will be able to do this.

Truthfully, I knew ahead of time that the pump was not going to be able to measure glucose levels. The thing that bothers me about the pump is that it didn't stabelize my blood sugar more. That's what I was counting on.

I am hopeful that there will be a lot of good things to come for Diabetics...it's just a matter of being patient.

[MarkMunday]

Thaks for the info. It looks like they have the technology. They just have to prove that it works and adds value in a compelling way. Which could take a while.

I think I will persevere with Lantus and wait for the next generation of pumps.

Cheers,

Mark

[MALACHY]

I've been looking into the status of continous monitors. I found this to be very interesting.
http://www.chemlin.de/news/mar04/20040316e01.htm

I think the key points to a automatic pump(meaning automatic insulin dosing) are a reliable accurate continous sensor and a control system that can reliably deliver the right amount of insulin.

[MarkMunday]

It is an exciting development. But, as the article points out, there are a number of challenges that still have to be overcome. A continous monitoring device would be a boon to MDI and pump users alike.

Developing an insulin infusion control mechanism that is driven by feedback from the monitoring device, shouldn't be too difficult. A limitation to the responsiveness of the system, is the length of time it takes for insulin to start acting. But insulins that are currently in the pipeline will have a suitable rapid action profile for this application.

They are definitely on the right track with this. It just seems to be taking a long time. There is huge money at stake. And whoever gets there first will be a big winner.


Mark

[Dobson]

what would happen when the implant uses the insulin it has? does it have the technology to produce insulin with the implant?

[MarkMunday]

My understanding is that the pump would still be worn externally. But it would be sent bg readings from an implanted sensor. An implanted pump would have to have a reservoir and pipe that protruded through the skin, through which it is filled from an external source. Gives me the shivers!

[Belinda]

Originally posted by MarkMunday
An implanted pump would have to have a reservoir and pipe that protruded through the skin, through which it is filled from an external source. Gives me the shivers!


Yeah and at my rate I would have it filled with lots of air bubbles!!!

[JasonSmithMT]

There is some information about the minimed implantable insulin pump here (minimed link). (http://www.minimed.com/patientfam/pf_products_implantpump_eu.shtml)

It is currently only approved for use in European Union countries. It will be great when they get the Continuous Glucose Monitoring System working effectively and tied in with the implantable pump.

Jason

[TonyAlmeida]

I was reading where there is a woman in CA that already has this system (or a more advanced one) surgically installed - the pump going in her upper abdominal region and the sensor (also surgically implanted) is attached to a main artery. A 2 hour surgery. Sensor uses RF to continually feed the pump BG information. Refill - you need a needle/syring to pierce your skin to some port on the pump inside you to refill the resevoir. The catheter is only a drain for the implanted pump.

FDA has not let her activate it yet.

This stuff is kinda scary though. I'm on my second pump because of a BIOS error they fixed with a later pump of the same model - imagine if the implant fails, another surgery to replace it. Can't just have them FedEx you one overnight. Special Insulin. And what if something else around you emits RF that confuses the pump and you suddenly get a burst of 5 units in error? And they still have to replace the batteries - yet another surgery.

[MarkMunday]

Interesting. There is obvious ly a lot of work going on in this field. The logical approach, to my mind, would be to keep the pump external at least until the technology is perfected. The only benefit I can see of an implanted is comfort and appearance.

I can see how relying on RF messaging from the monitor could be problematic. But programming tolerance levels into the pumps algorithm shouldn't be too difficult.

It sounds like a robust biofeedback driven mechanical system that controls T1 could still be a way off. And I find the prospect of regenerating beta cells using stem cell technology a lot more exciting.

Mark

[Shalyndria]

I'd be interested to know how the absorption is with the insulin being introduced into the peritoneal cavity (FYI that's a membrane that lines the abdomen and internal organs) with this implantable pump.
Can't say I'd be too thrilled at the idea of having a metal hockey puck in my abdomen myself, though. Mind you any technology that may make things easier is worth the trouble :) and I, personally, find it all exciting! And heck 40 years ago a person had to wear a pump on their back, so we've come a long way!

Shy

[TonyAlmeida]

The woman who has the implant said she did it for appearances. there are others, supposedly in France and Germany, who have been using the implant pumps and RF BG Sensor for a few years now.

I still want the adult stem cell research though ..... a real fix. I want to be my old happy self again - not paranoid like I am now. :(