Memorandum

Dexcom 3Q16 – Record sales of $149 million grow 41%; insulin dosing claim likely in early 2017; Android G5, new inserter submitted; G6 and first-gen Verily in 2018; competition heats up – November 2, 2016

Executive Highlights

  • Dexcom’s global sales hit a record high $149 million, growing 41% year-over-year (YOY) and 8% sequentially. This marks all-time high sales in six out of the past eight quarters, and 16 straight quarters of 40%+ YOY growth. The law of large numbers is kicking in, however: for two straight quarters, Dexcom has seen its lowest YOY growth since 3Q12 – pretty expected, given the steadily increasing base.
  • Following FDA approval of the MiniMed 670G and FreeStyle Libre Pro, management referenced marketplace “confusion” and “noise,” expressed confidence in the pipeline, and highlighted Dexcom’s advantages. Although it’s never been a more competitive time, the FreeStyle Libre Pro will likely have a positive impact if any in our view (introducing CGM to more people), at least until the FreeStyle Libre consumer product is approved.
  • FDA approval of a non-adjunctive (insulin dosing) label claim for G5 is “more likely an early 2017 event.” G6 will likely launch in 2018 (management characterized it extremely favorably, especially on the comparison to earlier generations - “in many respects it's a smaller, better performing version of the Libre”), and the first Verily product will launch in 2H18 – we also learned it will be factory calibrated. Several products are under FDA review: Android G5, next-gen one-button insertion system, 50% smaller transmitter, and the touchscreen receiver. Big!

Dexcom reported 3Q16 financial results yesterday afternoon in a call led by CEO Kevin Sayer. It was another quarter of record-high revenue, but also a defense of the company’s long-term vision in an increasingly competitive environment (670G, Libre). We enclose the top business and pipeline updates below, followed by a pipeline summary and a very insightful Q&A.

Financial Highlights

1. Sales reached an all-time high of $149 million, growing 41% year-over-year (YOY) on a very tough comparison to 3Q15. Sales grew 8% sequentially from 2Q16. This marks record-high sales for Dexcom in six out of the past eight quarters, and 16 straight quarters of 40%+ YOY growth since the launch of G4 in late 2012. The law of large numbers is kicking in as the base grows, however: for the second straight quarter, Dexcom saw its lowest YOY growth since 3Q12.

2. The international business saw a slowdown in the quarter, with sales of ~$18 million flat sequentially and rising 29% YOY – the lowest YOY growth Dexcom has seen OUS in over two years. Management cited FreeStyle Libre traction in countries without CGM reimbursement.

3. Full-year revenue guidance from 2Q16 was maintained – $550 million-$575 million (37%-43% YOY growth) – and notably, management expects sales in the “mid to upper end” of the range.

4. The call extensively discussed the US competitive landscape following FDA approval of the MiniMed 670G and FreeStyle Libre Pro in September. Management referenced marketplace “confusion” and “noise,” expressed confidence in the pipeline, and highlighted Dexcom’s advantages. The “us vs. them” discussion does obscure a key point: CGM is still a tremendously underpenetrated market globally.

5. Cash-based net income was $15 million in 3Q16, declining 25% YOY, though a 37% improvement sequentially. In addition to the receiver recall, management cited a few drivers of lower-than-expected profitability, including customer support investments and increased direct-to-consumer marketing (strong conversion rates so far).

Pipeline Highlights

6. FDA approval of a non-adjunctive (insulin dosing) label claim for G5 is “more likely an early 2017 event.” Medicare coverage is still expected in 2018 (“a really big deal” – we agree), and CMS is already in discussions with Dexcom and the FDA. Talks remain focused on the additional training and post-market study. FDA is also working to establish standards for a non-adjunctive sensor, including accuracy, manufacturing, and performance. This has implications for Libre too.

7. G6 “will more than likely be a 2018 launch.” The pivotal trial has started as expected, and to date, already 40%-50% of patients have completed the study. Initially, G6 will be 10-day wear and one calibration per day after startup, though management said it will be the foundation of Dexcom’s “no-calibration technology,” both with the first Verily product and G7.

8. The Verily partnership will launch a first-gen in the second half of 2018, with the smaller bandage-like sensor to be available “as early as 2020.” Management confirmed for the first time that the first-gen sensor will have no calibration, 14 day wear, Bluetooth (real-time), a smaller form factor than Libre, and non-adjunctive labeling – excellent to hear.

9. Dexcom finally filed an Android version of the G5 mobile app with the FDA in 3Q16 (PMA supplement); launch is expected in late 2016 or early 2017, depending on the FDA review. Notably, Android G5 launched “in several key markets internationally” earlier in 3Q16.

10. The next-gen one-button insertion system and 50% smaller transmitter were filed with the FDA in 3Q16. Dexcom is now calling this “G5x”. No approval timeline was shared – there will be significant enthusiasm for this.

11. Dexcom recently received FDA approval for a new version of the iOS G5 Mobile app and a transmitter firmware update, adding a mute override (when the phone is on silent/do not disturb, G5 will still audibly alarm), improved Bluetooth and Apple Watch experience, and more. The app update launched on Monday, though reviews are still critical – expectations are unbelievable (in our view) high for apps these days.

12. Dexcom’s next-gen color touchscreen receiver remains under FDA review (PMA supplement filed in 2Q16). No approval or launch timeline was shared.

Financial Highlights

1. Dexcom reported all-time high revenue of $149 million, representing a 41% year-over-year (YOY) gain on a very tough comparison to 3Q15. Sales also grew 8% sequentially from 2Q16. This marks record-high sales for Dexcom in six out of the past eight quarters, and 16 straight quarters of 40%+ YOY growth since the launch of G4 in late 2012. It also beat the previous high revenue of $137 million last quarter (2Q16) handily. With the growing base of sales, however, the law of large numbers is clearly kicking in: for the second straight quarter, this marked Dexcom’s lowest YOY quarterly growth since 3Q12. The US business provided 90% of the quarter’s growth on sales of $131 million (~44% YOY growth) vs. sales of $18 million internationally (~29% YOY growth).

Figure 1: Dexcom Quarterly Worldwide, US, International Product Revenue (2Q12-3Q16)

2. The international business saw a slowdown in the quarter, with sales of ~$18 million flat sequentially and rising 29% YOY. This was the lowest growth we’ve seen in the international business in more than two years, and management attributed the attenuation to traction of FreeStyle Libre in countries without CGM reimbursement (i.e., most of Europe). Q&A shared a lot of optimism about the international business going into 2017, including the reimbursement win in Germany announced in June (pricing looks favorable at this stage) and some positive early signs in UK and France (no firm timing shared). Dexcom recently established an international headquarters and is beginning to ramp up its internal sales and support staff outside the US. German reimbursement still isn’t expected to impact revenue until 2017.

3. Management maintained full-year revenue guidance from 2Q16 – $550 million-$575 million (37%-43% YOY growth) – but expects revenue to fall in the “mid to upper end” of the range. On one hand, this sounds conservative: Dexcom is 72% of the way to the mid-point of the guidance already, and both 2014 and 2015 saw 33% of full-year sales in Q4. However, there are some current variables that Dexcom hasn’t seen in prior years:

  • The receiver recall has resulted in many returns free of charge; historically, these would have been revenue-generating replacement receivers, particularly in the fourth quarter as patients’ out of pocket expenses are lower than any other time of the year. While only 30% of the business is from hardware, each receiver is several hundred dollars, which definitely adds up.
  • FreeStyle Libre traction in Europe. The impact was not quantified, but given the strong momentum for this device (200,000+ users as of Abbott’s 3Q16 call), there is more competition for Dexcom.
  • FDA approval of Medtronic’s MiniMed 670G in September (launching in spring 2017). Said CEO Kevin Sayer, “The media blip surrounding what is being called the new artificial pancreas has been deafening. Based on what they've been told, many patients currently perceive that they will no longer need to manage their diabetes if they purchase this product. Therefore, many patients are willing to delay purchase decision until this system becomes commercially available, or they're being directed to purchase the current Medtronic offering, with the promise of being first in line for the 670G.” We agree about expectations …

4. Prepared remarks and Q&A extensively discussed the US competitive landscape following FDA approval of Medtronic’s MiniMed 670G and FreeStyle Libre Pro in September. Management referenced marketplace “confusion” and “noise,” expressed confidence in the near-term and long-term pipeline, and highlighted Dexcom’s advantages: greater cost-effectiveness of standalone CGM vs. the 670G; “compelling” data from the DIaMonD study (presented at ADA and close to publication); the “extremely complex” 670G and unknown real-world performance; the blinded nature of FreeStyle Libre Pro (not a direct competitor to G5); data from EASD suggesting G5 outperformed Libre (consumer) in patients with impaired hypoglycemia awareness; a pediatric study (n=56) presented at ISPAD suggesting Libre accuracy “is inconsistent and not reflective of previously published data” (the MARD was 16.7% vs. fingersticks); and the existence of alarms and Bluetooth connectivity in G5 vs. Libre.

  • The competitive nuances are important to note, but the “us vs. them” discussion does obscure a key point: CGM is still a tremendously underpenetrated market globally. With ~15%-20% penetration in US type 1s and far lower penetration in Europe – not to mention type 2 – there is substantial room for at least three companies to succeed globally. Analysts on the call seemed genuinely concerned about both competitors (for obvious revenue, stock price reasons), though we think Dexcom’s pipeline is well positioned to compete – assuming execution happens fairly quickly, particularly from the company’s historically slow-moving pump partners.
    • The payer landscape is difficult to predict at this stage, and Abbott could bring a significant pricing threat to Dexcom, depending on what Libre’s (consumer) go-to-market strategy is in the US. We expect Libre will be positioned as “CGM” to take advantage of existing reimbursement in the US, though we wonder if a cash-pay option will exist and if it will be similar to the European pricing (read more on that in Abbott 3Q16).
  • On the MiniMed 670G: “...there has been significant press surrounding this product, which has created considerable confusion in the marketplace. For example, we've recently attended a diabetes charity event where it was declared from the podium that type 1 diabetes is now being cured because of the FDA approved artificial pancreas. Clearly, this is not the case, patients and caregivers are showing signs of skepticism and frustration due to the over-hype of the promise of this technology. Our review of the currently available data suggests that although this product may be an incremental step in automated insulin delivery, it appears to be an extremely complex system and its real world performance remains to be seen.”
  • On FreeStyle Libre Pro: “This system is not a direct competitor to our core commercial product portfolio in the US, as the data generated by the sensor is blinded to the patient and only made available to the clinician retrospectively. Additionally, the first independent studies regarding Libre's performance as a consumer device are beginning to be published, comparing Libre to Dexcom G5 [see our EASD report]. Dexcom CGM outperformed on nearly every metric, including improved time in range and reduced time in hypoglycemia. Just last week, at the ISPAD Meeting in Spain, a pediatric study was published showing that Libre accuracy is inconsistent and not reflective of previously published data. Put simply, this form of technology does not achieve the same results in diabetes management as CGM and alerts and alarms and real-time communication are very critical in generating real clinical outcomes.” [Editor’s Note: The FreeStyle Libre pediatric study presented at ISPAD was relative to fingersticks and not YSI.]

5. Cash-based net income was $15 million in 3Q16, declining 25% YOY from 3Q15, though a 37% improvement sequentially from 2Q16. In addition to the receiver recall, management cited three drivers of lower-than-expected profitability: (i) building out customer ops and tech support infrastructure to address the customer service issues experienced in early 2016; (ii) increased direct-to-consumer marketing (strong conversion rates so far will drive further increases in Q4); and (iii) incremental compensation expense and some severance expense following new senior talent adds in 1H16. We see these as growing pains more than cracks in the underlying business model. Net loss in 3Q16 totaled $19 million, relatively flat with 2Q16 ($20 million) and a dramatic improvement from 3Q15’s $42.5 million loss (reflecting a one-time $36.5 million payment for the Google Life Sciences [Verily] collaboration).

  • Gross margin was 68%, down from 71% in 3Q15, but a substantial sequential improvement from 62% in 2Q16. Notably, gross margins reached the high end of the August expectation for “mid-to-upper 60% range” margins in 2H16. The receiver recall impacted gross margin by 1-2 points during 3Q16. As previously expected, warranty costs as a percentage of sales began to decline in Q3 and “should normalize” before the end of the year.
  • Operating expenses also included increased spend on Dexcom’s four strategic investments in 2016 (first outlined at JPM): the Verily relationship; building out its data analytics capabilities; international expansion; and a new manufacturing facility in Arizona. Dexcom is “now trending to the $40 million of incremental expense” outlined in January.

Pipeline Highlights

6. FDA approval of a non-adjunctive (insulin dosing) label claim for G5 is “more likely an early 2017 event.” Medicare coverage is still expected in 2018 (“a really big deal”), and CMS is already in discussions with Dexcom and the FDA. We hadn’t realized CMS was already involved, as we thought Dexcom couldn’t approach Medicare until post approval; this is great to hear. The FDA approval discussions are ongoing and taking a bit longer than expected following the July 2016 positive advisory committee meeting vote. As expected, talks remain focused on (i) the additional training and education that will be required for non-adjunctive sensor use (we assume the recent Edelman and Pettus DT&T publication in August will be helpful); and (ii) the final size and scope of a post-market study.

  • FDA is also working to establish standards for a non-adjunctive CGM, including sensor accuracy, manufacturing, and performance standards. Said management, “I think what the FDA is really focused on right now is what sort of standards do they want to set. What is going to be the bar for a dosing type sensor? Meaning, not just MARD, but we focused in on % 20/20 and outliers and things like that. The FDA is focused on manufacturing standards, manufacturing variability, things like that. So there is a whole bunch going into this, and very thoughtful, frankly, on the part of the FDA; and it's just taking a little bit longer, is all. I don't expect they come back and ask us to modify the product in any way.
  • These Dexcom-FDA discussions could have major implications for approval of Abbott’s FreeStyle Libre (consumer). The factory-calibrated product was submitted to the FDA in 3Q16, and we’re very eager to see what the labeling looks like and how long the review takes. Arguably, a no fingerstick calibration sensor is safer, as there is less risk of erroneous entries (e.g., from a false BGM high). However, a factory-calibrated sensor also means less of a safety check for the sensor. 

7. The G6 sensor “will more than likely be a 2018 launch.” The pivotal trial has started as expected, and to date, already 40%-50% of patients have completed the study. Initially, G6 will be 10-day wear and one calibration per day after startup, though management said “it will be the foundation of our no-calibration technology, both with our first Verily products and with our Gen 7 system.” We cannot ever recall G7 being mentioned by name, and this was the first time (in our memory) that Dexcom confirmed the Verily product will be factory calibrated (more below). We cannot find the G6 pivotal on ClinicalTrials.gov, but it is the largest pivotal study that Dexcom has ever executed: pediatric and adults in one combined study of ~300 participants. The 2018 launch timing sounds slightly conservative to us and we wonder if things might actually go faster.

  • “G6 is as important an advance in our technology as was the shift from Seven Plus to G4 Platinum in 2012.” That is a big statement, since G4 was a big step up over Seven Plus in terms of accuracy, reliability, and large outliers. As a reminder, G6 is a complete change in Dexcom’s sensor platform: a new algorithm, a 10-day-wear sensor with interferent blocking, and new features like predictive alerts. The sensor is expected to retain the accuracy and insulin-dosing claim of G5: “All we can tell you now is, we are not going backwards with respect to accuracy and reduction in and/or elimination of calibrations.”
  • Data from G6 pre-pivotal studies will be presented at DTM next weekread our preview of the meeting here. Dexcom’s Peter Simpson will present in the CGM industry update session on Saturday afternoon.

8. The Verily partnership will launch the first-gen product in the second half of 2018, with the smaller bandage-like sensor to be available “as early as 2020.” Management confirmed for the first time that the first-gen sensor will have no calibration, 14 day wear, Bluetooth (real-time), a smaller form factor than Libre, and non-adjunctive labeling (“in many respects it's a smaller, better performing version of the Libre”). This specific commentary was great to hear and cleared up many unknowns for the first time after the partnership was announced in August 2015 (factory calibration, real-time, Bluetooth, insulin dosing claim). The first-gen sensor will use the G6 sensor, and we’re particularly glad to hear it will be factory calibrated at launch – a critical competitive move to keep up with Abbott’s FreeStyle Libre consumer version (under FDA review). Assuming FreeStyle Libre is approved in the first half of 2017, it will be more than a year ahead of the first-gen Dexcom/Verily sensor in the US.

  • The timing for a 2H18 first-gen launch is more specific than the previous “2018” expectation, a good sign in our view– Dexcom and Verily probably have the path to market more firmly laid down at this point. The second-gen bandage-like sensor timing was the same shared in the August call; the first feasibility study of the second-gen sensor will take place this quarter.
  •  “We don't believe the first product that we launch in partnership with Verily is going to be a null product, and we've said that on multiple occasions. It will definitely be a smaller form factor than the Libre platform. In many ways, similar attributes, no calibration up to 14 days. I think the performance characteristics – and you guys will be pleased when you see the data at DTM next week – it incorporates the core Gen 6 sensor technology. So you'll see some preliminary data on that next week, and you'll understand the levels of performance that we'll be able to achieve with that product. So I think that'll be a real positive for us. But in many respects it's a smaller, better performing version of the Libre.
  • Management’s excitement remains high for this partnership’s ability to expand CGM in type 1 and drive entry into type 2: “...our collaboration to develop simple, low-cost, disposable sensor systems is one of the most aggressive innovative development opportunities we've ever engaged in.”

9. Dexcom finally filed an Android version of the G5 mobile app with the FDA in 3Q16 (PMA supplement); launch is expected in late 2016 or early 2017, depending on the FDA review. Notably, Android G5 launched “in several key markets internationally” earlier in 3Q16 – this was not otherwise mentioned and we cannot find it on the Google Play store. From a competitive perspective, Abbott’s LibreLink Android app is currently available in many European countries; Medtronic’s Guardian Connect standalone, Bluetooth-enabled CGM will launch in Europe in ~Nov 2016-Jan 2017, but it is only on Apple devices to start.

  • To our surprise, Android is actually the dominant smartphone both outside AND inside the US (more than two-thirds of the US and EU markets). This launch, therefore, will be key for expanding G5’s potential market. Kantar WorldPanel ComTech data is what many mainstream tech press articles quote, and this market share data was released on October 12:

 

Android Market Share

Apple iOS Market Share

US

65%

31%

EU (UK, Germany, France, Italy, Spain)

78%

17%

China (urban)

86%

14%

10. The next-gen one-button insertion system and 50% smaller transmitter were filed with the FDA in 3Q16 (PMA supplement). Dexcom is now calling this “G5x.” No approval timeline was shared. Previously this inserter was expected to launch by the end of 2016, meaning Dexcom could still hit the timing if things go rapidly at the FDA. This pipeline project requires a major manufacturing change, but is unquestionably needed to improve the patient experience from the current syringe-like insertion device – particularly as Abbott’s FreeStyle Libre consumer version and excellent insertion process comes to the US market.

11. Dexcom recently received FDA approval for a new version of the iOS G5 Mobile app and a transmitter firmware update, adding a mute override (when the phone is on silent/do not disturb, G5 will still audibly alarm), improved Bluetooth connectivity, an improved Apple Watch experience, and other features. The app update launched on Monday. These changes respond to critical user feedback, though the update currently has just 2/5 stars on the iTunes store – while many asked for the mute override, some reviews on the app store seem pretty frustrated they can’t silence it during meetings, etc. Dexcom also added a vibrate-only feature for alarms, though it’s clear some don’t know this. It’s hard to please everyone, and boy are the expectations high with apps these days.

  • We still see many patients at diabetes conferences using the G5 receiver, even though the app has at least three critical advantages: (i) less stuff to carry around; (ii) data automatically sent to the web-based Clarity platform; (iii) data sent to Apple Health, allowing other apps to use it; (iv) no need to replace broken/lost receiver. We wonder what percentage of current G5 users utilize the app vs. receiver, and whether this update will expand the former. From an HCP and Dexcom perspective, the G5 app is definitely preferred, given the auto-data upload. 

12. Dexcom’s next-gen color touchscreen receiver remains under FDA review (PMA supplement filed in 2Q16). No approval or launch timeline was shared. We might assume a fairly quick review, as the user interface will be in line with the already-approved G5 mobile app. We haven’t seen updated pictures on what this will look like, though presumably it will give the medical-device-looking receiver a welcome facelift.

Pipeline Summary

Sensor and Software Pipeline

Pipeline Product

Timeline

Non-adjunctive label claim (insulin dosing)

Approval expected in early 2017, Medicare coverage expected in 2018. FDA labeling discussions underway following advisory panel meeting on July 21.

Touchscreen receiver

Submitted to FDA in 2Q16.

Android G5 app

FDA submission in 3Q16, launch in late 2016 or early 2017 in US. Launched in several key international markets in 3Q16.

Next-gen one-button insertion system and 50% smaller transmitter (G5x)

FDA submission in 3Q16.

G5 app improvements

  • Mute override, improved Bluetooth
  • Ability to receive insulin dose data from other devices

 

  • Launched on October 31.
  • FDA submission previously expected before the end of 2016.

G6 sensor

New sensor and algorithm with 10-day wear, one calibration per day after startup, insulin dosing claim, acetaminophen blocking, predictive alerts

 

Launch expected in 2018. Pivotal study underway  (n~300; pediatrics and adults), with 40%-50% of patients completed.

Verily [Google Life Sciences] partnership

Simple, low-cost, disposable, 14 day, Bluetooth-enabled, factory calibrated sensor. First-gen smaller than FreeStyle Libre, second-gen bandage-like.

First gen to launch in second half of 2018.

Second-gen, bandage-like sensor “could be available as early as 2020.” First feasibility study in 4Q16.

Automated Insulin Delivery Partner Pipeline

See our latest competitive landscape here.

Animas

Hypoglycemia-Hyperglycemia Minimizer with Dexcom G5 CGM

 

As of July update, pivotal study was expected to begin enrolling in 4Q16, with a launch by November 2017-May 2018. We’re not sure if the study has begun yet.

Insulet

Next-gen OmniPod PDM with G5 app integration

Hybrid closed loop (Mode AGC algorithm)

 

Will “debut” at ADA 2017.

First in-clinic feasibility study started in September.

Tandem

t:slim X2 with Dexcom G5 integration

Predictive low glucose suspend

Hybrid closed loop (TypeZero algorithm)

 

 

FDA submission by end of 2016. Launch in mid-2017

Pivotal trial in 2016, launch in 2017

Pivotal trial in 2017, launch in 2018

Bigfoot Biomedical

Automated insulin delivery system with Dexcom CGM, Asante pump, and smartphone app

Began first feasibility study in July 2016. Pivotal in mid-2017, FDA submission in early 2018. Raised $35.5 million in October 2016.

Beta Bionics

Bionic Pancreas dual-chamber iLet with integrated Dexcom CGM

Insulin-only: Pivotal trial start in 2H17; PMA submission by mid-2018.

Bihormonal: Pivotal trial expected in 1H18; PMA submission timing to be discussed with the FDA. Received $12 million from NIH for bihormonal pivotal (October 2016).

Questions and Answers

Q: It sounds like the approval of 670G and all the discussion around it has basically taken up all the noise in the room and made things a little bit more difficult for the rest of the diabetes players. I don't know if you saw it, but Tandem reported relatively soft results and guided down for the year. It sounds like that's kind of occupying all the discussion right now in this space. Can you just characterize the tone of business pre and post that? Obviously, it's reflected in your guidance commentary for the fourth quarter, but any other color there would be great. And then, two, how do you change that dynamic? How do you make it, so that you're not spending the next six months trying to get people’s attention when Medtronic seems to have dominated the discussion over the last month?

A: Not only getting their attention, but talking about a product that's not commercially available yet that patients can't see, and the perception being created that it truly is an artificial pancreas. Where we're taking the discussion is back to the fact that CGM is the most important tool in intensively managing diabetes. And the outcomes we've achieved in the DIaMonD study, there are several other studies coming out in the not too distant future that are similar to that – real world studies – are great outcomes. The doctors have been very supportive. They're very much taking a wait-and-see attitude. We've done very well with clinics in getting our message across to them. Again, it's just this confusion or cloud that's hanging above us all.

We'll focus on CGM first, quite honestly, in the fourth quarter, in particular the investment in CGM if somebody's met their deductible is not tremendously large. We will offer programs to encourage new patients to sign up. You'll see those come out over the Internet over the next couple weeks to see if we can get more patients on.

I would tell you from a color perspective, our reorder patterns still remain very strong. We're not seeing reorders fall out of the queue, we're not seeing new patients. It's just combating the overall noise. And we went through not as much noise with 530G, but we did go through noise with 530G as well. The difference being that product had been out in Europe and we knew exactly what we were going after. There have been studies comparing our sensors to theirs and other things. So we could adopt a technical and a tactical approach. This time, there isn't one because again their product is not out. So we'll just continue to deploy, we'll continue to offer promotions, our DTC campaign has been very effective as far as generating the leads. But that's just where we are today.

Q: Talk a little bit about international being flat sequentially. Can you attribute some of that to an impact from Libre? How should we think about the impact of you guys going direct and getting reimbursement, and when that will begin to impact revenues at least in Germany?

A: There was no impact in Q3 with respect to reimbursement in Germany. And if anything, what we did is, we significantly increased our operating expenses by adding the operating expenses of our distributor to our own expense line.

We are seeing now, as reimbursement is starting to sporadically come in for some of the payers in Germany, that our opportunities of new patients have gone up quite significantly. I think, there will be a bit of a bump in Q4 for 2016 for Germany. But it really is more of a 2017 event than it is late 2016, probably more Q2 and beyond – again, as we have to get reimbursement and insurance contracts signed up.

We have the staff to go do that and we're calling on the people we need to meet with and talk to now. We've gone direct in the UK. UK was very small for us, it's still not very big, but the fact is our business has gone up at a nice rate since we hired a great team over there. We've hired people who can go drive reimbursement in there, as a percentage of our patients who are getting covered in local geographies there.

So I think German reimbursements can help significantly, UK growing reimbursement might help a tad, a bit, but it's not going to be a huge driving factor. As we look at other countries, we certainly have had meetings and discussions in France. We see reimbursement continuing to be talked about in Italy. Australia is now looking at reimbursement procedure as well. And we've got programs going on in Canada also.

So there are a lot of opportunities. And once we get reimbursement, we think we can grow nicely in those markets. But with respect to cash payers, I'll get back to what I said earlier, our product is very expensive, and there is a reason; it's really good. It provides a lot. And it is a large cash commitment for somebody. And when offered at much lower cash price upfront, parties have elected to try that [Libre]. I think over time when we get reimbursement, we can certainly work through that because of the features that our product set offers.

Q: As we think about kind of the transition here while we wait for the other products to show up in the market, with most likely 670G and then whatever label that Libre consumer might get next year, do we think about different growth rates before and after that happens? So if there is a transition where people are holding back a bit and we've come through with kind of maybe subpar revenue growth, but after that gets out there and you can message, you got the G5x, that you would expect to see patient adds accelerate perhaps in the second half of 2017?

A: We'll give formal guidance for next year in January; it's something we're still evaluating. I would tell you, we're starting to see new patient adds pick up again. I think it's too early to tell. I wouldn't even want to speculate on Abbott. The challenge we have with Medtronic, we went through this exact same thing with 530G, it was supposed to put us out of business, but we had the benefit of having an actual product and independent investigators had the benefit of having an actual product to test it. We've scoured their safety statement. It appears that patients are being recommended to take at least four fingersticks a day to keep the thing running. They also have to take fingersticks every time it drops out of automatic mode to keep it into automatic mode. The false alert rates seem pretty high; in fact, potentially, even worse than the 530G. And as we know in the 530G context, that’s the one thing that patients complain about most in the field is that the false alarms they receive from that system. So, look, we're fighting a ghost for the next six months to nine months and we won't really know what we're up against until we actually have physical product to get a test.

I think we need to see what the fourth quarter brings. We also have a non-adjunctive claim on the horizon, and I believe that could be a very big deal for marketing and positioning Dexcom relative to other products further even out there, because when there is a sensor that is labeled you can dose insulin off of this thing, I think that will be very compelling for physicians, caregivers and patients, and present a very good story for us.

We have G5x coming, which is the smaller transmitter and a much easier insertion system – much more practical and an easier experience for patients. So we have a number of positives, but we're not quite sure of the timing that goes into this entire discussion. We'll be a little more concrete in January.

Q: Is it then fair to assume hardware as a percent of revenues drops below 30%, just because of the warranty stuff you mentioned and then perhaps the patients starts and replacement of hardware for maybe the next one quarter to three quarters?

A: what we've said when we started the G5 transmitters is that it was two for one. We expected our gross transmitter hardware dollars to go up and our gross margin percentage to go down, because we were supplying two transmitters for the price of one. And that has happened. What has affected us again is the lack of receiver revenues, but there has been an increase in transmitter revenues. We think ultimately the percentage will go down and we'll be more sensor-driven than transmitters. We are selling a lot of transmitters. I don't have percentages to give you where that goes yet. We'll provide more color in January.

Q: Have you seen any change in the attrition rate with the launch of G5?

A: No. We think our patient retention metrics are still very good. I would tell you the thing that may have changed and we're studying this a little bit, and I can't give you complete color, because we study retention all the time. As we enter a broader market of patients, our early adopters were pretty intense on CGM, “I need CGM or I'm going to die” basically was often the statement that we'd hear from patients. As we're entering a group of patients who would say, “okay, I'll give this a try,” we see their use patterns may be different than our long-time, use it every day, all the time patients. We also see that, just given the reimbursement world, the continued use of our sensor for more than seven days remains prolific. And so that is something we factor into our revenue models and revenue analyses as well. But I don't see a huge increase, we don't see a huge decrease, we think it's remained relatively consistent.

Q: You mentioned some of the marketing initiatives that you have, but just given some of the stiffer headwinds, any thoughts on adding to your sales team going forward?

A: We are going through our 2017 planning efforts right now. More than likely, we will expand the team. We'll announce how much, and to what extent, in early 2017.

Q: On Germany, can you just give us any clarity on the level of reimbursement, or how ultimately price will shake out? Are the economics comparable to the US for Dexcom?

A: actually without giving you specifics, the preliminary numbers that I've seen look quite favorable, even in comparison to what we see here in the U.S. And that, I would tell you, I spoke to John Lister, who heads up our European operation yesterday, and he informed me that we're currently in discussions with approximately 40% of the covered lives, if you want to look at it that way, in Germany, and the pricing looks quite favorable.

Q: Just to touch on G5 and the potential dosing claim: any color you can add with the conversation with the FDA, as far as potential changes from a product perspective that they may be looking for, whether it be a required number of calibrations a day or sensor shut off, or something along those lines?

A: No, nothing in particular. I think what the FDA is really focused on right now is really, what sort of standards do they want to set. What bar is going to be the bar for a dosing type sensor, right? Meaning, not just MARD, but we focused in on % 20/20 and outliers and things like that. The FDA is focused on manufacturing standards, manufacturing variability, things like that. So there is a whole bunch going into this, and very thoughtful, frankly, on the part of the FDA; and it's just taking a little bit longer, is all. I don't expect they come back and ask us to modify the product in any way. And that's one of the reasons for the post-market study. And after the post-market study, if we see anything, then we would go back and discuss it with them. But for today, there's nothing significant.

Q: Any comments directionally on new patient additions and what the type of mix was there? I believe last quarter you talked about 50% of patients coming from MDI. Just any sort of color you can give us on this quarter?

A: Once again, more of our patient adds came from MDI than from the pump world, so we continue to make inroads there. With respect to numbers, just to get to $149 million in revenues we reported, the new patient number had to be significant. It certainly was within our target. And quite candidly, year-to-date, based on our internal targets, we are spot on at the end of Q3. So new patient growth remains very, very robust for us.

Q: Just starting with sensor utilization, you mentioned that reorder rates haven't changed; but then you also noted that the use patterns of the product may have changed, given the broader patient base you now have. Can you clarify these comments? And is average sensor utilization in Q3 and through the first nine months of the year maybe closer to the lower end of your historical range?

A: I think sensor usage for the first nine months of the year is really consistent with what we've had before. What I was trying to do in making my comments was indicate that, again, patients continue to use it long, particularly those where reimbursement is tough for them and they're not covered as well as others, and again, some of the new patients. But I don't think there has really been a significant change in sensor utilization. But as the whole population gets bigger, the portion or just the raw numbers of people who use their sensors longer is going to go up, and so it becomes more visible to us as we look at it, same with the use patterns.

We go through exercises where we pick samples of the new patients in the patient base and try and follow them all the way through, look at their purchasing patterns, look at their use, history, and everything. And every time we do it, we've learned something different and something new. And those were really the two things that were pointed out from the last group and the last population that we studied, that we still see increased sensor use and maybe the use patterns for some of these patients are not what they were before. And that's okay, they're still patients and they still derive great benefit from it.

Q: You talked about the impact Libre had internationally. Can you talk more about the features of your first-generation Verily product and how it stacks up versus Libre for Type 2 patients, or maybe Type 1 patients in the near-term who might desire a less comprehensive solution?

A: We don't believe the first product that we launch in partnership with Verily is going to be a null product, and we've said that on multiple occasions. It will definitely be a smaller form factor than the Libre platform. In many ways, similar attributes, no calibration up to 14 days. I think the performance characteristics – and you guys will be pleased when you see the data at DTM next week – it incorporates the core Gen 6 sensor technology. So you'll see some preliminary data on that next week, and you'll understand the levels of performance that we'll be able to achieve with that product. So I think that'll be a real positive for us. But in many respects it's a smaller, better performing version of the Libre.

One of the things that will dictate the software around that product and the feature set is going to be our non-adjunctive label. We are not going to give up our non-adjunctive label for anything. Once we get it, we're going to stay there. And we believe our technology is more than robust and that – again, as Steve said, as you'll see next week to maintain that level with reduced and with no calibrations over time.

I should add two key differentiating factors. Of course, that it does have Bluetooth technology and does have real-time alerts and alarms as part of the system.

Q: I wanted to go back to the topline guidance. You mentioned the number of times that you see the sustainable growth rate of this business, about 35% to 40% I think is what you've said historically. And the guidance range, as we look into the fourth quarter, I guess, implies just shy of that even at the top end. So should we think of the fourth quarter more as fully baked in for these competitive pressures, and that from that base we can start to recover back to more of a normalized rate? Or do you think that could potentially get larger impact sequentially for a couple of quarters as the launches continue competitively?

A: We don't have a crystal ball to tell you. There's initially substantial confusion in the marketplace. We're now starting to see some press to the contrary where they're trying to clear up the mistaking notion that this is actually an artificial pancreas coming. But we don't have a perfect crystal ball to tell you what this is going to look like going forward. We're being conservative, obviously, in our guidance. I want to take a step back, we stated 35% to 40% sustainable growth something like four or five years ago. Four to five years ago, looking back, we successfully sustained and exceeded that level of growth. At some point, we've said, even on recent calls, that the law of large numbers is such that on a percentage basis, growth is not going to continue to exceed 40%, 50% like it has in the past. So you guys are trying to read a whole lot more into this, I think.

We need some time. The flip side of that is that penetration in the Type 1 market for CGM is still not above 20%; there is plenty of runway there. We're not penetrated in someone using type 2s at all where reimbursement will come there. So I think it'll be a combination of when we can get our new technologies and products out and feature sets, we can expand to broader populations, combined with just market penetration in the current markets where we sit. So we'll just see it over time.

Q: A follow-up on gross margins. You mentioned last quarter they would jump back up to the upper 60%s; they certainly did. Sounds like there's still elevated levels of warranty expense. So normalized for that, is that – is it fair to say that we can get back in towards the 70% levels?

A: I think that's fair. Our warranty expense was about 3.5% of revenue vs. our historical averages of about 2%, so we have some opportunity there. The margins came up this quarter kind of in combination of lower warranty cost as a percentage of revenues. We also had better yields particularly on our G5 Mobile transmitter, and then of course we had higher sales volume.

Q: We focused mostly on German reimbursement. Can you give me a timeline of when we might get additional reimbursement? And then, my second question has to do with the non-adjunctive claim. Is there anything qualitatively that you can provide regarding the conversation that you're having with the FDA?

A: On the non-adjunctive claim, we're really not going to provide any more details than we have. We need to work through the requirements and standards and really have a clear position before we explain everything to everybody. The discussions have been very good and very focused and very regular. They've been paying significant attention to the process and we hope to get through it soon.

With respect to the other geographies, we really don't have timelines. One of the things we're learning as we go to other geographies, a lot of these places with national systems are very local and it's state-by-state or region-by-region.

We certainly have very good reimbursement in Sweden right now through the tender process. We have the German reimbursement decision that just came. There is reimbursement in Switzerland and Austria. And there's some other regions, like I said, in the UK where we're getting a little bit of reimbursement and a little bit is coming forth in Italy, but there has not been any national decisions. We'll just continue to work on region-by-region and we'll give you more color as time goes on.

Q: Am I right in sort of estimating that there is maybe about a $3 million impact from receivers on the revenue line in the quarter, that kind of ballpark?

A: We're not going to quantify it.

Q: In terms of CMS, it doesn't sound like these are sort of linear discussions where you're waiting for FDA approval of the non-adjunctive and then you'll go to CMS. It sounds like there is positive dialogue going on ahead of that, is that fair?

A: Yeah. That's a fair statement. In fact it's communication together, CMS and FDA and our regulatory folks. So, yeah, there are certainly discussions ongoing. Obviously, we will still need the non-adjunctive labeling before we can really become active with CMS. But I would say this is a – and it's highly visible not just with FDA, but within CMS as well. Our belief is that CMS is going to ask us to go direct with our national coverage decision as opposed to trying to go regional like we had discussed previously. From what we're hearing from our consultants and from our internal folks, that they don't think even if the non-adjunctive claim pushes into the early part of 2017, that that will impact our path to Medicare by some time in 2018.

Q: In terms of data that's going to come out next week, are these patients from the pivotal trial or is this another study that you have going on with...

A: With all of our sensors we run several pre-pivotal studies. The data that we’ll present next week is from relatively large pre-pivotal studies in our Gen 6 platform. It's not the pivotal data, because we wouldn't open that up until it's done and we submit it.

Q: Maybe just a little bit on CGM pricing broadly. You know you have 670G out there and then Libre and I think there is the potential for some other CGM clearances in 2017, if I'm not mistaken. How do you see pricing kind of shaking out with all the increased technologies on the marketplace at this point?

A: We've remained relatively consistent on the price front for quite some time. I think future prices will be determined by the outcomes one can generate with the technologies. Another big factor on pricing is going to be the length to wear the sensors, because we've talked about our Gen 6 sensor, is a 10-day product vs. 7-day product, which gives us an opportunity to either increase price or increase value to the patient and payer, and we'll probably end up with a combination of both with the Gen 6 sensor.

We prepare internally scenarios for every alternative; we prepare for higher prices, we prepare for lower prices, we prepare for everything you can imagine. And so we've modeled several scenarios, we'll adapt to where the market takes us, and being the leader we want to take the market obviously to the right place. But these technologies have to be affordable for patients and they have to provide outcomes that are worth more than the cost of the CGM system. And that's how we look at pricing in the future.

Q: Just a follow up on 670G, and when you look at the dynamics between pumpers and MDI, it seems to me that the initial low-hanging fruit would be pumpers. How much of a broad push are you expecting from 670G? Do you think it's sort of going to be broad based in those two categories and therefore potentially more competition in the MDI segment of the marketplace?

A: I think that's the big question, right? When Terry and Kevin sold MiniMed 15 years ago, pump penetration was at about 20% in the US; today it's at about 30%. So 15 years later, four commercial pump companies have been able to drive an additional 10 points of penetration. I mean, I personally don't believe that 670G is going to move the needle from 30% penetration to 40% or 50% penetration in any meaningful period of time.

I think, particularly, outside the US, MDI is going to remain the vast majority of patients' method for insulin delivery. And I think even in the US, I think, as we continue to improve our technology, improve the form factor, shrink the form factor on the technology, add patients with additional tools, additional decisions support tools with respect to their insulin on board, etc., I think the product offering, including an intelligent insulin pen, for example, will be a very compelling, both from an outcomes perspective and from a cost perspective.

Q: Have you seen any impact on your patient adds from the DIaMonD data that was released at ADA this year?

A: The DIaMonD study has not been officially published in a journal yet. So we remain relatively low key about that data. At this point of time, we know it is under review. Right now, we are going to be a little more aggressive with marketing it, because we think it's getting closer to publication. So we will really unleash the DIaMonD blitz over the next month of November here, and hopefully that will help increase patient adds – not a lot so far, because we really haven't pushed it.

Q: Can you provide more color on the timing of the rest of the DIaMonD data to be released this year?

A: There are a couple of other study arms. There is a type 2 patient study arm. There is also an arm of the study where some of the patients continue to use CGM for a longer period of time, another group of patients who continue to use CGM and go to an insulin pump in addition to CGM. I know the studies are winding down. I don't have timelines for you, and they'll be affected by publication timelines as well. But the study will be wrapped up by the end of the year. And so the data will come after that.

Q: One just on G6, it seems like the pivotal is moving very quickly there and FDA has also seemed to be moving fairly quickly in some of their approvals recently. So just wondering how solid you feel on that 2018 commercial launch time? What are the options to maybe pull that forward or see that move up a little bit from a potential launch standing?

A: We have a sequence of launches next year, particularly with the new applicator and transmitter of the Gen 5 system, but we need to get those manufacturing processes up and mature really quickly. And that's really important to us. That is a bigger endeavor than anything we've undertaken for a long time. So it's probably better timed the G6 come out a bit after that, because we fully intend to use the new transmitter and new applicator configuration, and we'd be much happier if we didn't have to start from scratch. If we could pull it in, that would be wonderful, and we work to pull it in, but there are a lot of boxes to check.

With respect to FDA timing and speeding things up, I would agree they appear amenable to that. We have been discussing non-adjunctive claim for a very long time, and it's not through. So when you get into complex issues, it can become real tough. I think they'll be very cooperative with Gen 6 and I think we can push it through quickly, provided the dataset is reflective of what you guys will see next week. I think it's a tremendous advance in sensor technology that we can get through. But we've got to remain diligent and keep going, but there is a lot to get ready before it comes.

Q: Just want to go back and clarify something you said on Germany. So it sounds like the reimbursement rates are coming in, you said pretty solid, especially even relative to U.S. rate. So just wondering, that's the kind of the sell-in price or the price you are going to recognize for sensors and transmitters and receivers is on par relative to U.S.?

A: That's exactly right. I mean, none of this is finalized, right? We still don't expect to see an impact until next year. But I would tell you, so far so good.

Q: And do have any idea the co-pay patients are going to have on that system in Germany? How does that stack up relative to self-pay for a Libre?

A: I don't, but maybe for the next call, we'll have a little more detail, or we'll have the contracts signed by that point.

Q: Relative to Europe, since Germany was just discussed: do you have any greater visibility on the other countries coming through? Our checks on Germany have been pretty positive over the last several weeks, just in terms of not only how reimbursement is shaping up, but potential demand. Do you think that Germany could be bigger in 2017 than maybe you're assuming?

A: Yeah. It certainly could be, particularly as I mentioned, we're in discussions with about 40% of the covered lives. So as we continue to accelerate that, if we can truly contract with something like 15-ish kind of critical payers to get the vast majority of covered lives, Germany could be a much bigger contributor next year.

I'm just giving you an example, in the UK, we're actually seeing reimbursement. Most of our patients are actually seeing reimbursement these days, but it's a one-off basis and they're having to apply individually, so it's not something that's scalable. So we really do need to reengage with NICE and push that issue, and really try to make it more formalized.

France the same thing, we're in discussions. There's not been any formal action taken in France. The signs are looking positive. So that could be a contributor, potentially as early as next year as well. So yeah, I think you're thinking about it right, Mike, that Europe as a collective, if we knock off a couple more of these big countries, it could be a big push for us next year.

Q: Obviously there is something to learn here from what's going on with Libre, right? I mean, Libre is taking off in Europe. There is this demand for diabetes technology in Europe that hasn't been satisfied. And so, when you think about the fact that there's now a couple of hundred thousand patients that potentially are on Libre or tried Libre, better put, at this point, how do you tap into that once you get reimbursement in a country like Germany? What's the launch strategy look like? So I have to think that, with the Libre experience and the enthusiasm that product has seen, that how you think about going to market has to be different than maybe you were thinking a year or two ago?

A: We are very much evaluating product configurations. I'll give you one example. We have CE Mark for the system without use of the receiver in Europe now. So if a patient wants to use the system, unlike in the US where we get a new patient we are required to sell them a receiver, we don't have to do that anymore. So we've put ourselves in a position where we can somewhat reduce the upfront cost for patients to get them to start.

And as far as enthusiasm, we are enthusiastic. In fact, our German team is going through budget processes and their request per sales heads is remarkable. So they obviously are very optimistic about their ability to go sell, and poor Jess has to be the bad guy down at the end of the table. But we're going to add significant feet on the street in Germany to go after that.

I think one of the other lessons learned is, people do want information. And if they're co-paying out of pocket and the reimbursement covers this, you know, by offering connectivity and the things that we offer from a feature set vs. this product, we think they're going to really lead to fast growth. We will configure future products as we look at it. We're going to offer a no calibration sensor over time. We're just going to offer one that's going to be accurate enough to meet our standards and to meet the non-adjective claim standards that we believe will be set up by the FDA as well. So there is a lot to learn here and we'll aggressively go after it.

Concluding remarks: We want to thank everybody for being on our call today, and everybody for their questions, they're very insightful. We continue to address, what we believe is, one of the most meaningful opportunities in all healthcare. The treatment of diabetes type 1 and type 2 is one of the leading cost drivers in healthcare all over the world. We're going to continue to focus on what we do best, developing, manufacturing, distributing the world's leading CGM technology. Our future platforms, our sensor platforms, Verily’s technology, our new algorithms and advanced data analytics are going to be able to change the conversation of diabetes management over the next several years.

Our growth opportunity, even in times like we described today, is still huge. The type 1 market is very underpenetrated. There is a completely un-penetrated type 2 intensively managed insulin market. The global opportunities we've discussed with Germany, the UK, the other places we're seeing reimbursement come up are huge, even in light of the increased competition that's coming. It's a great time to be here at Dexcom. We'll continue to deliver great products and great performance for our investors. Thank you.

-- by Adam Brown and Kelly Close