By Chris Leach
Few, if any, people can match Terry Gregg’s perspective on both the business and technology of continuous glucose monitoring. As President and CEO of MiniMed from 1996 to 2002, he was instrumental in the acquisition of that company by Medtronic in 2001. He joined Dexcom as an independent member of the board in 2005, and as CEO in 2007. After the first three-day glucose sensor was approved in 2006, Dexcom went on to develop its seven-day sensor technology and moved from that to its fourth generation of continuous glucose monitors, with a fifth and sixth generation in development. IN sat down with Gregg for this interview at the Consumer Electronics Show in Las Vegas in January 2013.
IN: Dexcom’s G4 Platinum is your latest product release. What’s different about it, compared to the Seven Plus?
TG: Lots of things, but three stand out. First, the sensor is 20% more accurate from 40 to 400, and 30% more accurate in the 70 milligrams-and-below area of hypoglycemia detection. Second, is its range of connectivity. The Seven Plus had somewhere between 5 and 8 feet of connectivity and the G4 Platinum is labeled for 20 feet. We’re hearing that it’s closer to 50 feet, especially in line of sight. So, if you think about a patient’s mobility at home, you might be able to leave it on a nightstand, go down to the kitchen, come back up, and not lose data packets. Third, it’s about one-third the size of the old receiver. You can slip it into your pocket or put it in your purse. We actually didn’t think too much about the size, but customers really like that improvement.
IN: You’ve completed your pediatric trials. When do you think the new device will be approved for kids?
TG: Right, the pediatric trial is done. We’re scrubbing the data, and we expect to submit it in the next few weeks. Then it’s the statutory 180-day review, so we’re looking at mid 2013 to actually expand into the pediatric market. The increased connectivity range we talked about is important for parents, because they can monitor their child on a playground, on a sports field, without having to strap the receiver to the child and have them come to the side to see it. So, for us, it was like, “Wow, we didn’t really design it with that intent,” but that’s been an ancillary benefit.
IN: Approval for kids might increase your sales significantly, right?
TG: Yes. We don’t call on pediatric endocrinologists today, and between 20-25% of the Type 1 population are 17 and younger, so it will expand our market very quickly.
IN: What’s possible, on the low end, in sensor accuracy? You’ve already said that you have a G5 version in development, and that it will be more accurate than the G4.
TG: The G4 Platinum has a 13% MARD (Mean Absolute Relative Difference) overall. But the device gets accustomed to you and your fluctuations and it becomes more accurate over time. If you look at it on a per day basis, Day 1 is 16% MARD and then it drops down to around 11-12% MARD, and actually closer to 10% by Days 4, 5, 6, and 7, just bordering on that single digit. Gen 4+ will have a new algorithm that will reduce the overall MARD by another 2%, so we’ll be around 11% at the beginning and drop to the 10% – 9% range in days 4 to 7. So that’s Gen 4+. There is a Gen 5, which I’m actually wearing this week, which is mostly about hardware change. I port my information directly to an off-the-shelf Android phone.
IN: So, that takes care of wireless connectivity. Is there a Gen 6 on the drawing board?
TG: Yes, and it’s a very exciting potential improvement. The expectation on Gen 6 is significant calibration reduction to the point that it’s done at the factory and doesn’t have to be confirmed or re-calibrated by finger sticks. We can calibrate it at the factory, you insert it, and that’s the last thing you have to do for 10 days. That’s the future, that’s what we believe, and it’s single-digit MARD, down around 9%. We believe that is the technology that FDA will get comfortable with, to allow us to dose insulin just based on readings of interstitial fluid.
The expectation on Gen 6 is significant calibration reduction to the point that it’s done at the factory and doesn’t have to be confirmed or re-calibrated by finger sticks.
IN: That’s very exciting. When do you think it might be available?
TG: Probably around 2017.
IN: Progress takes a long time in this space, doesn’t it?
TG: It does take a long time; this is truly still an R&D project in the context of a small company with a budget. We could accelerate, at a cost. There are some not-for-profit organizations and some high-net-worth individuals with extensive diabetes in their family, who have seen the early data, as part of the artificial pancreas consortium, and have asked us if we’d develop a special team to help accelerate that. We’ve said, “Yes, and this is the budget that it would require.” So we’ll see.
IN: You bring up the artificial pancreas. Can you talk about Dexcom’s involvement with the various groups that are developing them?
TG: There are approximately 25 artificial pancreas projects being conducted around the world. We’re in 22 of them, so we’re highly invested in all of that. We believe it’s an important technology, but I always caution the audience that these are systems. So, if you think about an artificial pancreas, it’s at least five components the way I describe it. It is certainly a pump and the disposables. If they’re using a durable pump, it involves insulin, it involves glucose sensors, and it involves an algorithm. All of these have to work flawlessly and in concert. If any one system should fail or one component fail, the system fails.
With that in mind, we’ve made tremendous progress, particularly over the last two years, where we’re allowing patients to be remotely monitored. That’s what Ed Damiano’s Beacon Hill project got approval for in November. The folks at UVA have done a pilot project for 30 hours where they’ve remotely monitored patients using our technology and the bi-hormonal two-chamber Tandem pump. (See “Building The Bionic Pancreas,” Insulin Nation, Issue 5, and “The Artificial Pancreas Hits The Road,” Insulin Nation, Issue 6.)
Today, we can make exquisite sensors. Pumps can deliver the right amount of insulin. Algorithms can detect glucose levels and dose accordingly. So, a lot of the components are in place. I think the insulin is the Achilles heel right now. We need faster-acting insulin and one that is more predictable in its duration across all patients.
IN: You’re also involved in a number of collaborations with other equipment manufacturers, such Animas, Tandem, and Roche. Will you comment on the nature of these relationships?
TG: Animas, Tandem, and Roche are our three pump partners. Animas has incorporated our G4 technology into their Vibe pump, which has been available in Europe since 2011 and has a good chance of being approved for sale in U.S. in 2013. Tandem originally wanted to use our G5 technology, but after studying the G4, decided to use it in a combination pump/CGM device that we hope to have available early in 2014. Roche is still working with our G5 platform and that product launch is currently set for 2015, but they could decide to use G4 instead, and have a product ready by the end of 2014. These are all non-exclusive deals.
There are approximately 25 artificial pancreas projects being conducted around the world. We’re in 22 of them, so we’re highly invested in all of that.
IN: You were the COO of MiniMed when Medtronic. acquired it. Is there any satisfaction in taking away business from your old company?
TG: I did spend a lot of time there, but I don’t have any emotional tie to Medtronic. I want to do what is best in the interest of the patient. I think the world of Medtronic as a company. I think they do a lot of good. I think we build a better sensor. In fact, I offered our sensor technology to Medtronic so they could include it in their pumps.
IN: What was Medtronic’s response?
TG: They declined.
IN: CGMs produce a lot of data. How are you making the numbers easier to understand, particularly by the wearer?
TG: Here’s the challenge for Dexcom. We’re not doctors, so as much as we would like to say in a report to a patient, “We think you need to adjust your insulin.” We can’t say that. We’ve tried with the FDA and the FDA has said that’s the line that we can’t cross, and we’ll respect that. Our Studio software download, which is free, provides what we call a Dexcom Portrait, where you can see, on one screen, seven different reports that visually summarize your glucose patterns. We think the best practice is to try staying in your target range, and avoid either hypo or hyper periods as much as possible. If you have to adjust your insulin, do it in small increments. We’re constantly looking at ways to simplify the information so that individuals, providers or patients, can immediately look at it, recognize, and make modifications.
IN: In CGM use, what’s more important — the number, or the trend?
TG: The trend. If we had the ability to remove the number and show them trend arrows, it would be the best scenario. But in the nomenclature, if you think of CGM being a meter on steroids, meters have numbers, physicians are comfortable with numbers, and so that’s why those numbers exist. I will tell you that our attitude is very much trend first. Look at the trend, because you need to know where you’ve been and where you’re going.
IN: CGMs and pumps are often lumped together, as though they were part of the same market. But your “meter on steroids” description shows you don’t feel tied to growth in pumping. What is your long-term vision for Dexcom?
TG: The pump market is somewhat limited, worth about 2 billion dollars. Glucose monitoring, or glucose sensing, is a 10-billion-dollar market. We’re going to stick to making glucose sensors.