Many in the Diabetes Community are anxiously awaiting the release of the smaller and sleeker 2nd generation OmniPod, but already we know what a third generation won’t have:
A Pod that’s integrated with a Dexcom CGM sensor.
Yep, it’s true: the two companies say they have abandoned the integration deal they had in place since 2008.
Two executives at Massachusetts-based Insulet Corp., makers of the OmniPod, say they’re no lon
ger working with California-based Dexcom on integration, and Dexcom’s CEO Terry Gregg confirms this.
They’ve essentially broken up after dating for five years, deciding marriage wasn’t in the cards. Yes, the relationship’s been rocky for at least a couple years, and so this parting isn’t completely surprising. But both are finally admitting they can’t keep going on the same path together.
“The official agreement isn’t actually terminated, we’re just not moving forward at this time and we don’t expect to in the future,” Gregg said. “I believe Insulet is making a mistake, but it is their mistake to make.”
Rather than stay with Dexcom, Insulet’s already courting a new CGM partner. In the past two weeks, Insulet has signed a development agreement with an unnamed private company to develop an OmniPod that would incorporate the insulin-infusing Pod with a CGM sensor in one single unit — eliminating the need for a second site on the skin.
Insulet’s CEO Duane DeSisto says, “Our goal is pretty simple: Have a product that keeps the patient out of trouble, doesn’t interfere with their life, and is just one thing on the body. That is where we are going to go. We think we can put one thing on the body that has a sensor and delivers insulin all in one space.”
Why the Split?
Insulet’s leaders say their original agreement with Dexcom was to create an integrated product that would have still required both a Pod and CGM sensor/transmitter to be placed separately on the body; it would have only eliminated the need for two handheld receivers, dropping the Dexcom receiver in favor of all data being displayed on OmniPod’s handheld Personal Diabetes Manager (PDM).
But now it’s clear that “one-site integration” is the future, Insulet’s CEO DeSisto says.
And with Dexcom’s plan for its newest Gen5 that will bypass any receiver and send data directly to a smartphone, both DeSisto and Insulet’s chief financial officer Brian Roberts say that continuing to co-develop with Dexcom doesn’t make sense anymore.
“They’re obsoleting what this integration was always meant to be: to integrate these two handhelds into one,” Roberts said of Dexcom. “There’s nothing for us to be working on with them at this point. We’re just on slightly different paths.”
Roberts says Insulet is interested in “leapfrogging” the simple integration of two handhelds or two on-body parts, instead looking ahead to a single unit that bring patients real gains in quality of life.
At Dexcom, Gregg says his company does not share the vision of a single-infusion-site system for several reasons — for one, he’s not sure it’s even technically feasible. Dexcom has been studying the concept with an international academic research center, and finds there are too many open questions and concerns to justify this “same site” concept, at least for now. Gregg believes the timeline may be five years or more for what Insulet is working on, since nothing like this is currently developed and ready for regulatory review — which means validating it through the FDA process would likely cost hundreds of millions, potentially leaving OmniPod behind as new dual-site integrated devices begin to come to market.
Competitor Medtronic already offers an integrated pump/CGM with its next-gen model on the way, both Animas and Tandem will likely have a sensor-integrated pump here in the U.S. within the next two years, and Roche Diagnostics is also expected to have an integrated device at some point in the next few years. But none of these currently aim to send the data to a smartphone the way Dexcom is planning.
Gregg’s vision for Dexcom’s future is to create an extended CGM sensor life of up to 10 days that’s completely compatible with an insulin pump, where data’s sent directly to a smartphone. The new system would ideally be accurate enough to eliminate the need for fingersticks altogether.
But Insulet has a different plan in mind.
Forging New Relationships
So what is Insulet’s next move? In early January, in a presentation to investors at the 31st JP Morgan Healthcare conference in San Francisco, Insulet’s DeSisto announced a new CGM development partner. (Listen in starting at the 16:20-minute mark in the news conference.)
He didn’t name the new partner, nor would Roberts in a call with the ‘Mine late last week. All the two will say is that this other player has been developing new sensor technology for about eight years now and has been in the glucose monitoring business for much longer than that, so it’s “in the position to compete” with others in the CGM and pump market.
Insulet and this mystery company have spent the past year or more studying multiple sensor concepts to determine what could be the best product for the OmniPod platform. DeSisto says the concept is to create an 80-hour sensor that would last as long as the Pod on the body (since insulin stays stable for about 80 hours), and the patch pump would infuse insulin from one side while sensing glucose levels from the other.
They now have what they believe to be “a viable sensor option” and the hope is to have a prototype designed by the end of 2013, with possible human studies outside the U.S. in early 2014. They’ve tested the sensors in pigs and DeSisto says the results are promising, comparable to other existing CGM devices and traditional blood sugar checking. An actual entrance to market could be three or four years out, the Insulet execs said.
On top of all this, DeSisto told investors that his company has also signed on with another company developing closed-loop technology, giving Insulet the option to license both safety and predictive algorithms. What this means is that with its integrated pump/CGM technology and necessary algorithms, Insulet could potentially develop its own artificial pancreas product (!) Again, Insulet’s mum on the partner company’s identity.
Could one of these mystery partners be Abbott Diabetes, which has been quietly working toward a next-gen CGM to follow on the Freestyle Navigator it pulled off the U.S. market in 2011?
No one will say officially, but the rumors are flying in the diabetes device industry.
There may be some clues… Last summer, Insulet and Abbott announced expanding an agreement for Freestyle glucose monitoring technology to be used in future Pod versions through 2013 — an announcement that came about six months after Insulet agreed to incorporate LifeScan OneTouch Verio glucose meter technology into future Pod versions. And with Abbott testing the Freestyle Navigator 2 in clinical studies here in the U.S., the stage could be set for these two D-device makers to hook up. Maybe the writing’s been on the wall for keen eyes to see.
Jeff Christensen, public affairs director at Abbott Diabetes, would not comment, other than to say the Insulet agreement from last summer “was specific to blood glucose monitoring test strips.” Of course they won’t confirm or deny any rumors about business development.
So, only time will tell who Insulet’s mystery dates turn out to be.
Risky Business?
Just one infusion site on your skin for both pumping and CGM sounds like a pipe dream to many a PWD. And maybe it still is…
The reliability of this kind of “all-in-one” unit raises questions. Keep in mind, Insulet is basically starting over from scratch with a new partner that hasn’t yet proven itself in the CGM market (let alone showing this single-attachment concept will work!) rather than sticking with Dexcom, which already has an established and successful product available.
Naturally, Insulet allays those fears. Roberts says switching the CGM sensor supplier now will NOT impact the timing of a next-generation Pod that would be integrated with CGM technology, adding that developing a third-gen product would not have begun any earlier anyhow, due to the wait on FDA clearance for the new OmniPod that just came in December.
Plus, Roberts believes people are willing to wait for a single infusion site, which is what they want even more than device integration. Market data they’ve been gathering shows that 90% of type 1s are willing to live with one device attached to their body. But that number drops down to 40% when you’re talking two attachments.
Seven-year-old OmniPod now captures 10% of the pump market share in the U.S., and 70% of its customers are brand new to the pump market. A third are younger than 18. With the new smaller Pods coming to the market soon (late February to end of March), it expects to double or triple growth in the coming year.
But even with that popularity playing out as Insulet predicts, is it realistic to sacrifice the marriage of two FDA-approved devices for a relationship that includes an unknown partner with yet-to-be-proven technology?
Patients are currently being advised that pump infusion sets and CGM sensors should generally be placed a couple of inches away from each other. In fact, even the Dexcom G4 instructions warn: “Locate the sensor at least 3 inches from the insulin pump to ensure accurate readings.” Hmmm. Makes you wonder. So their brand new sensor technology will solve this issue completely?
Even if the device worked, an 80-hour sensor would have a life much shorter than that of other CGM devices already out — such as the G4 that’s FDA-approved for seven days but can often be worn (off-label) for three weeks or more. Medtronic’s current integrated pump-sensor is approved for three days, but can last six days or longer; the next-gen pump and sensor will likely be approved for a week and practically be worn for longer than that.
Assuming the future OmniPod will function the same way it does now, shutting down automatically after its use period, will users be confined to the whole integrated system lasting just over three days? Wow, that could really bring up your cost of diabetes supplies…
It may be a long wait to find out. DeSisto said at the JP Morgan conference that it might have taken a year or year-and-a-half to integrate with Dexcom, but it wasn’t worth that time or cost if Dex was going to bypass the receiver altogether. Now, the timing with a new partner could be three or four years to get through development, research and regulatory phases… Do the math.
Why Insulet wouldn’t continue working with Dex to create an interim integrated Pod on the way to its more fascinating future generations is puzzling. Seems like people who’ve been anticipating this integration might like to have it soon, rather than being forced to just hold out hope for the future while other D-device companies bring integrated products to the market.
Of course, that’s the $64 million question: Are users willing to wait? Is the allure of this new gadget several years down the road enough to convince people — especially younger PWDs brand new to pumping — that it’s better to deal with essentially four devices (tubeless OmniPod, PDM, G4 sensor, and Apple-esque Dexcom receiver) while they wait, or opt for a line of CGM-integrated traditional pumps that are already available now?
Seems like a risky gamble to take. But maybe the captivating concept of an integrated, one-site, tubeless CGM-Pod system will trump all in the end. After all, it’s the allure of a sexy and seductive new partnership that beat out an established five-year relationship with Dexcom.
Thanks for a very comprehensive and interesting article. For those who can’t do the math, 80 hours is 3 and 1/3 days. Given a 2 hour warmup period it’s really 3 1/4 days. For anyone who hasn’t used a Dexcom that might be reasonable, otherwise….I wonder what this would do to the footprint of the proposed OmniPod.
I do hope that Insulet and they anonymous partner will publish the data format for their new device(s). There’s a lot of interest in open data format support. I’m working with a number of people on this issue and we hope to put serious pressure on device makers to have open data formats as a requirement for new devices. Thanks DiabetesMine for also working on this issue.
Thanks for the story Mike.
It will be interesting to fill in the blanks. Among those is will the Pod people remain loyal to the pod or will integration tempt them to try other pumps. They seem a very loyal bunch.
I would be curious to know how many podders currently use cgm. I recall reading some research someplace that suggested that CGM users were more willing to consider giving up their pumps than their sensors. That seems to suggest that there is a risk for Pod as a business but IMHO what really matter is the health of PWD. Slowing the advances of integrated care I see as unfortunate.
Best
Bennet
Intriguing. As some of the earlier commenters noted, not only is the “usable” sensor time diminished (with all the warm-up periods), but this system removes choice from the patients. Assuming the only pods will be those with sensors, that would not only drive up the cost but, if someone prefers a different CGM, they’d be forced to wear two sensors (unlike the Medtronic system which doesn’t require a sensor with the Revel/Enlite). And since I can’t see an integrated cannula/sensor wire, that’s a needless hole through one’s flesh.
Good reporting Mike, but not really so surprising given the way smartphone penetration has gone in the last 5 years.
I was struck by this parenthetical in your story “(since insulin stays stable for about 80 hours)” – can you elaborate or provide more context?
Thanks for the comment & question, mike! Here’s what Insulet said: “When the OmniPod was originally developed and cleared by the FDA, rapid acting insulins were labeled to be stable for three days unrefrigerated. The size of the Pod reservoir and the duration of its functionality (80 hours) were designed based upon this labeling.”
That is interesting, and of course leads to more questions than answers. Things like:
WHEN did they first go to the FDA?
Is their design concept of 80 hours/pod outmoded when today’s insulins last longer?
Put another way, are they too invested in the current design to see a more innovative method for tubeless delivery (beyond integrating a CGM?) that is more patient and environmentally friendly?
This also tells me that Dexcom made the right choice in ending the relationship. Sigh!
I don’t know about anyone else. But I have issues with Omnipod insertions anyways. I’ll be lucky if I can last 3 days without bruising, bleeding, failed pod, etc. One more reason I am looking at switching.
What I also find odd is that they specify that they want the insulin to be room temperature before it’s put into the pod. So what? They expect me to take the vial out of the fridge, wait for it to warm up, put it in the pod, and then put the remaining vial back in the fridge? Weird.
Insulet is missing the ball here. Even though the likelyhood of a workable 2- in -1 device is small, getting a pump and a cgm to BOTH read on a smartphone is more important than reducing 2 sites to one for us.
My son wears both the omnipod and the dexcom. frankly the navigator was never a viable alternative when it was on the market and is still off the market here.
until a different tubeless pump is on the market we will stay with omnipod, but ditching dexcom is a BAD idea. Getting the data on a smart phone for a pump and the CGM is going to be where people go.
too many issues with 1 device IMHO. tossing a perfectly good sensor every time a cannula kinks or you have a small 200 unit device is not a great plan.
Insulin in one spot for more that 48 hours isnt a great idea to begin with. I dont understand what they are thinking at this point. Maybe this is there way of saying “we are stuck and there will not be integration of any cgm in the forseeable future.”
IF they refuse to work on this , I would bet the podders will leave as soon as a viable tubeless pump is on the market with a integration plan that makes sense.
Everyone would love a 1 device on the body but when you look at the logistics of it, and really think about what it would mean, I think people can understand WHY it isnt a great idea with the current technology.
My daughter uses both the Omnipod and the Dexcom G4. While she would prefer one site, the truth is the Dexcom Sensor lasts ( with restarting ) sometimes 3 weeks. The accuracy of the G4 is unbelievable. From my daughter’s perspective, there’s no way she’s going to switch from the omnipod to another pump because she prefers the lack of tubing.
The lack of integration on the handheld side in the short term is really no big deal to her. The PDM woudn’t be a great place to see the graphs anyway.
I see the lack of integration as a non-issue given news that G5 will transmit to the smartphone.
If folks at insulet are reading, my daughter has one BIG request of you. The first gen and second gen PDMs have hard to push buttons that make a noise to her that is like nails on a chalkboard. Go to a touch interface or at least make the buttons quiet. The G4′s are fine.
my daughter is 12 and also a podder/ G4 customer, she has same sentiments as your daughter, doesn’t give a crap about the two devices, a non issue, dex going to cell will bypass integration efforts, she doesn’t mind the two ” things on the body’ at all as I expected she would ( as our endo told us she might)
she is just so damn happy to have the numbers all the time and to not get high and low so much and FEEL BAD.
Isn’t it great these companies are fighting so hard for our business??
Well … all of the above comments are true .. but you should think out of the box …. what if the canula and the sensor electrode can use one penetration spot … what if you do not need to warm up anymore .. what if the sensor and the pod combined are cheap enough so you do not need to worry …. The thing that bothers me is about the problem that the pods got now .. if insertion fails you lose the pod .. same goes with the new device .. if for some reason one of the TWO components fail … you lose the pod …. Thats the real troublemaker for me… but after all it’s all about the money … if its cheap enough .. everything will be OK
GL
What are Amy’s thoughts on this, since she wears both a Pod & G4?
@Judi – thanks for asking. Yep, I’m an OmniPod and Dex G4 user. Frankly, I’m bummed that my two favorite (and IMHO the two best) diabetes technology companies out there are no longer collaborating. That duo would have been fantastic!
That said, I’m sure not willing to switch to any tubed pump just because it’s paired with a CGM — even if it is a great Dexcom product.
So I guess I’m forced to wait to see what kind of integration Insulet might come up with. In the meantime, I am wearing these 2 devices and struggling to find enough “real estate” to accommodate both.
I would place my bet on Insulet partnering with one of the next gen non-invasive sensors that are using spectroscopy to read BG as opposed to the current sensors that have a wire filament sitting under the skin. Possibly C8 medisensors, which is under regulatory review in Europe. If this new technology truns out to be accurate, it would be a great fit with Omnipod, as the sensor can be easily moved around without wasting supplies and the pod could easily house the large battery that is needed to power the C8.
Thank you for posting this today. I shared it with my daughter and she’s beyond sad. We both find the idea of an all in one site less than appealing. Integration with a smart phone – sure that sounds great. She’d like that as she moves toward HS and college. However, a two in one site that results in a cgm life of 3 days and the necessity of changing both one’s insulin site and cgm site should one of the two fail? Not sure who thought of that, but we both think they completely missed the boat. Too bad. I’m sure my kiddo isn’t the only one who’s decided she’ll now wait for the other tubeless pumps she saw at Expo to hit the market. Omnipod is, according to her, irrevocably off her list.
Could be good news for Roche Solo and CellNovo but they may delay their launches in order to integrate Dexcom *sigh*.
As for using Raman Spectroscopy–expect skin irritation!!!
The news is not necessarily a bad thing. Insulet produced an innovative product, that was quite an engineering feat. The design is also refined and there is relatively low failure rate for the pods, when it comes to occlusions and other mechanical malfunctions. Aside from a tendency for the pods to fall off, I doubt the failure rate is more than a “traditional” insulin pump.
I trust Insulet to make a better product than most other diabetes device manufacturers, including an integrated CGM/Pump–a CGMPod.
There are all sorts of cool things that can come out, with the physical integration of the two devices. Such as, if you were going to change your old CGMPod at 80 hours and it logged your sensor BG continuously for the past X hours (i.e. probably around 3 hours) prior to the change, you could change to a new CGMPod and use those past X hours from the old pod as an “instant” calibration along with a meter BG (or more) to get “instant” and “continuous” readings from the actual CGM without disruption.
Their strategy could also be a good way to get JDRF funding.
But, Insulet’s software, is horrible, along with any software that deals with diabetes medical device data, including software using mobile devices and the cloud. Insulet’s software is only capable of operating on Windows, due to the .NET framework. Even if you were to get it running it in Parallels or VMWare Fusion on a Mac, the drivers would never work.
The only current device for uploading diabetes device data that is powerful and efficient is the Accu-Chek Smart Pix. I can upload my data to my computer using the software stored INSIDE the Smart Pix USB device. I then open up the file it produces (which opens up in any web browser), save it as a TXT file, and I then run my data through MatLab on my netbook. It takes way less time than any other type of diabetes related software.
Anyways, although the Accu-Chek Combo is not cosmetically attractive, it is appealing. Accu-Chek is on my short list for my next insulin pump.
Anyways, if the Artificial Pancreas project is going to be truly successful, a medical device data storage AND communication is going to have to be used. If the JDRF funds such projects, a precondition should be made that ISO/IEEE 11073 must be used for all data and communication between all devices used in the Artificial Pancreas projects.
Otherwise, the organization is just wasting donated money. Because, it is much harder to program and deal with proprietary data structures and communications–and it certainly costs more money. Also, analysis of the various AP clinical trials cannot be done as easily in Tandem.
I moved to Omnipod from Medtronic 4.5 years ago (after 5 with Medtronic) and will never go back to a convention (tubing) pump. I was really looking forward to the next generation integration, reducing the # of things in my pockets is a big sell. But my long term vision has always been an all in one device (one handheld and one site on the body). I just started on the Dex G4 (my first CGM ever) last month and can already say that while I love it I’m going to have to start getting creative with places to put these 2 devices on my body. One device would be a huge improvement. Additionally, I lead a very active lifestyle and even with using taping products like IV3000 or Tegaderm both my Pod and my sensor have a lot of issues with staying stuck on me. Long workouts and hot summer days play havoc with how long I can functionally use things, so to me the 80 hour lifespan isn’t really a big deal. Ultimately, with a potential 4 year development horizon, who’s to say that they might not even decide and develop the ability to transmit data to a smartphone as long as they’re at it.
[...] The two companies say they have abandoned the integration deal they had in place since 2008. Read more var dd_offset_from_content = 45; var dd_top_offset_from_content = 0; Category : Featured, [...]
When I heard of this in the Insulet presentation at the J.P. Morgan Healthcare Conference in January, it actually made sense the way the company presented. Why bother working on integration when a bluetooth version that anyone can use will be in a future generation of the Dexcom? If you’re a company like Insulet, you obviously need to prioritize which projects you’ll invest time and resources into. Given Dexcom has been fairly public about the direction it plans to go, its a wonder other pump companies that have partnered with Dexcom aren’t already moving in the same direction.
You mention four things to carry with current OmniPod and Dexcom CGM; pod, sensor, pdm, and cgm handeld or an intergrated pump. If you have an “intergrated” pump would you not have an infusion site w/tubing, a pump, a sensor, and a finger stick meter, which is required to bolus as well as calibrate your cgm? Also, if dexcom does go to an app for cell phones your back to tubless vs. tubed. Good article, just some thoughts.
I would like more a partnership with Echo Therapeutics. Think of a pod with a Simfony sensor on is flat side, which you insert as seen in the following video. The sensor works “down” while tha cannula inserts as usual on a side. I think it lasts around the 80 hours of the Pod.
Otherwise, with a needle CGM, you could think to use a CGM Pod and after 3 days retract the canula, keep it going as CGM only and put a “simple” Pod for the other 3 days. You would have 2 Pods toghetrer on for 50% of the time, but it would work for sure.
http://www.youtube.com/watch?v=_MtOVQGOEI4
[...] – read the story [...]
[...] Mine Insulet & Dexcom Break Up Over Plans for Next-Gen Integration (Kind of a bummer! Click over and read all the great [...]
My understanding is that controlling a life critical medical device, such as an insulin pump, through a smartphone would never be approved by the FDA. How could they test cellular interference, communications on planes, and interaction with the numerous other pieces of software on a phone? I imagine that an additional friction point with Dexcom was determining whose customer the integrated device user would be.
Hopeful that the interest of the diabetic will lead the technology development at Insulet.
Very interesting … The OmniPod was my first pump, and I had major issues with both failed pods and infusion site absorption and (most importantly) allergy to the Pod adhesive, at least during the warmer weather here on the East coast. A 3-day sensor is ridiculous, in that I get 6 days out of my Medtronic sensors and note that the Medtronic sensors are becoming better. The past two I’ve used gave me very accurate readings after the first calibration! My endocrinologist does not allow any of her pumping patients to keep their infusion sites in for more than 2 days. That’s an individual things, but in my case it’s good advice. All the more reason for me to stick with Medtronic, despite the tubing.
It’s not strange: it all depends on I much time it takes to start working well after insertion and how much does it cost. It will be a new sensor so we cannot say anything now.
As for Medtronic, here in Italy for the “new” Enlite sensor I read many nightmare experiences …..
i disagree, try inserting the sensor 24 hours in advance (make sure the top part is protected in the correct position using med tape.