Memorandum

Dexcom 2Q15 – Record sales of $93 million; CGM as a pharmacy benefit through United and Anthem; G5 launch by end of year – August 5, 2015

Executive Highlights

  • Dexcom reported all-time high product revenue (again!) of $93 million in 2Q15, a 60% year-over-year (YOY) rise and a remarkable close to 30% sequential gain. The full-year 2015 revenue guidance was increased to $350-$375 million (36%-46% YOY growth) from $340 million-$360 million.
  • United Healthcare and Anthem will now process CGM as a pharmacy benefit. Other payers may follow, offering cost advantages to patients, Dexcom, and even payers!
  • Management continues to expect FDA approval and launch of the G5 mobile app and new Bluetooth-enabled transmitter later this year for both adults and pediatrics. We’ve seen folks who have it in trials and it seems very cool. We continue to think the emergence of Share will spawn a slew of interesting apps that make it much easier for patients to better understand their diabetes.

Dexcom held its 2Q15 financial update this afternoon in a call led by CEO Kevin Sayer, who did a very impressive job leading the call. See our top financial and pipeline highlights below, followed by a pipeline summary and Q&A.

Financial and Business Highlights

1. Dexcom reported all-time high revenue of $93 million in 2Q15, a 60% year-over-year (YOY) rise and a 28% sequential gain. This now marks nine consecutive quarters of 55%+ YOY growth. That the company can continue to do this from a higher and higher base is indeed remarkable. Management increased the full-year 2015 revenue guidance to $350-$375 million (36%-46% YOY growth) from the previous $340 million-$360 million (32%-40% YOY growth).

2. The business continues to see improved profitability: cash-based net income of $19.7 million (21% of revenue) more than doubled from 2Q14’s $8.2 million (14% of revenue) in 2Q14. Gross margin reached 71%, up from 68% in 2Q14.

3. United Healthcare and Anthem will now process CGM as a pharmacy benefit (effective July 1). This allows patients to pick up Dexcom sensors at retail outlets like CVS and Walgreens (!), offering lower out-of-pocket expense, faster time from the initial prescription to sensor pickup, cost savings for Dexcom, and often lower prices for payers. Other major payers may follow.

4. Dexcom began its first major CGM awareness campaign during the quarter, fueled by celebrity Nick Jonas. Dexcom’s website had more visitors than in any other day in its history following the campaign’s video launch. Awareness of CGM is one of the two biggest challenges as Dexcom keeps growing (the other is getting new products out quickly enough, and we would add perceived hassle factor and doctor inertia).

5. In a question on the outlook for CGM in five years, management envisions a miniaturized sensor (“more like a bandage than a medical device”) that is fully disposable and factory calibrated. That would be amazing, of course! CEO Kevin Sayer believes a minimum of 30% penetration in type 1 in the US is feasible in five years (similar to what pumps are at now).

R&D Pipeline Highlights

6. Management continues to expect FDA approval and launch of the G5 mobile app and new Bluetooth-enabled transmitter later this year (submitted in February). In new news, the launch will be for both pediatric and adults – this is a major win for patients. G5 will initially launch for Apple iOS, and move to Android in the future (no timing specifics).

7. Management expressed optimism that an insulin-dosing label claim could be attained in the US sometime next year. Dexcom is now “in sync” with the FDA regarding the necessary pre-market and post-market data. This would also apply to G5, since it uses the same sensor and algorithm as the G4. An insulin-dosing claim is most important from a Medicare coverage perspective, though it could also encourage more doctors to prescribe CGM. It’s an effective negative for strip companies who are already losing strip sales from those patients using CGM who had previously used a relatively higher number of strips than usual.

8. Dexcom expects to begin a pre-pivotal study of the 10-day-wear Gen 6 sensor later this year, with a pivotal study to shortly follow, an FDA submission in early 2016, and launch in early 2017. G6 will likely start with one required fingerstick calibration per day (down from the current two per day with G4), with a gradual reduction over time.

9. Dexcom expects to launch a “robust” new data platform developed by SweetSpot in 2H15. The company recently hired a Senior Data Executive with years of experience in Big Data in Silicon Valley. We don’t know who.

10. “Good progress” continues on a new insertion system, a new lower-cost, higher-quality receiver, and several generations of transmitters. Transmitter R&D is focused on reducing the size, expanding the battery life, and going disposable.

11. Management echoed 1Q15’s commitment to automated insulin delivery, specifically mentioning Tandem, Bigfoot Biomedical, and the Bionic Pancreas.

12. In a brief ADA recap, management highlighted T1D Exchange data comparing the A1c of MDI vs. pump users on CGM, suggesting little difference. It counters the perception (to Dexcom’s frustration) that CGM should only be added onto a pump. Anecdotally, we believe this is changing.

Financial and Business Highlights

1. Dexcom reported all-time high product revenue of $93 million in 2Q15, a 60% year-over-year (YOY) rise and a 28% sequential gain. The sales were Dexcom’s highest ever by far, beating the company’s blowout 4Q14 performance of $84 million. The YOY comparison was also quite challenging, given 64% growth in 2Q14. This now marks nine consecutive quarters of 55%+ YOY growth for Dexcom, and 2Q15 sales tied sales recorded for all of 2012. Given the upcoming launch of Gen 5, pharmacy distribution, and a very stacked pipeline, continued growth at this rate seems achievable.

  • Management increased the full-year 2015 revenue guidance to $350-$375 million (36%-46% YOY growth) from the previous $340 million-$360 million (32%-40% YOY growth). CEO Kevin Sayer was frank in calling the updated guidance “pretty ambitious,” noting that Dexcom needs to generate at least $200 million in sales in 2H15 to achieve the mid-to-upper half of the range – through 1H15, Dexcom is 46% of the way to the new guidance’s midpoint, which is actually ahead of this point in 2014 (when 1H14 sales accounted for 41% of full-year sales). Dexcom had a back-loaded year last year, though with the upcoming launch of G5, similar performance is possible this year.

Table 1: Worldwide Revenue

 

2Q14

3Q14

4Q14

1Q15

2Q15

Product Revenue (millions)

$58.2

$67.9

$84.3

$72.8

$92.9

Year-over-Year Growth

64%

60%

64%

56%

60%

Sequential Growth

25%

17%

24%

-13%

28%

 

Figure 1: Dexcom’s Quarterly Worldwide Product Revenue (2Q12-2Q15)

  • The international business generated sales of $12.2 million, rising 61% YOY and 31% sequentially (very consistent with the US business). International sales now represent 13% of total revenue, down slightly from 15% in 1Q15. The international business provided 13% of the quarter’s growth in 2Q15, down from 28% in 1Q15 (when the US business was seasonally weaker). As a reminder, Dexcom has just four international employees (as of the 1Q15 call), so there is lots of upside here. Still, we wonder if broader availability of FreeStyle Libre potentially tempered European uptake of G4 in 2Q15 – that is speculation on our part, though many early FreeStyle Libre users have been type 1, so it is at least possible.
  • Dexcom continues to see ~60% of its new patients on pumps and ~40% on MDI. The launch of the Animas Vibe in January has helped Dexcom on the pump side, though management did not quantify the impact. Further uptake of Tandem’s t:slim and Insulet’s OmniPod have also helped, as those patients migrate to Dexcom despite no firm integration at this time. 

2. Even as topline revenue grows at ~60%, the business continues to see increased profitability. Cash-based net income of $19.7 million (21% of revenue) more than doubled from 2Q14’s $8.2 million (14% of revenue) in 2Q14. Net loss of $3.7 million improved from $6 million in 2Q14 despite non-cash expenses of $23.4 million. Much of this reflects increased share-based compensation (Dexcom’s market cap is now at ~$6.7 billion ($86 per share), doubling from ~$2.9 billion and $44 one year ago). As the stock price continues to rise, shared-based compensation is going to be a significant expense going forward. Some contend that the valuation is too high, though the company’s growth has come from still very limited penetration of CGM, and there is unquestionably a lot of upside in both type 1 and type 2.

  • Gross margin reached 71%, a very solid improvement from 68% in 2Q14. Dexcom is now at the upper end of its gross margin targets, though further improvements could come from increased volumes, a further shift in sales mix to disposables (right now it’s still 70% disposable / 30% hardware), and cost savings in future product designs.
  • Average selling price per sensor was on the high end of the standard $70-75 range; despite analyst excitement, the bump really reflected quarter-to-quarter variability in direct vs. indirect business (Dexcom gets higher pricing when it sells directly vs. through distributors, which still reflect about 50% of its business). Management hopes pricing will stay roughly neutral as distribution shifts to the pharmacy channel. Patients are still using 2.5 - 3 sensors per month on average, meaning wear time is around 10 days for those using it 24/7.

3. Dexcom had two major pharmacy distribution wins during the quarter: United Healthcare and Anthem will now process CGM as a pharmacy benefit (effective July 1). This allows patients to pick up Dexcom sensors at retail outlets like CVS and Walgreens, offering lower out-of-pocket expense (similar to a drug co-pay), faster time from the initial prescription to sensor pickup, less expense for Dexcom (e.g., no need to have a huge call center for reorders like in the DME world), and often lower prices for payers (since DME distributors mark up the price). Dexcom will continue to expand efforts on this front, with the goal of the “vast majority” of its business running through the pharmacy channel over the next several years. Dexcom is building out its managed care team and field presence to support this move, which will more direct contact with pharmacists (who now need to know the product!).

  • The move to pharmacy distribution is expected to impact new patients “very favorably.” In many cases, a patient will only need a prescription and “maybe some limited prior authorization” to get on CGM. This is expected to shorten the time it takes a new patient to get on sensors. Management was honest in stating that a “few bugs” will need to be worked out, but the potential is high. In one anecdotal story, a mother with a newly diagnosed two-year-old picked up CGM at her local drug store within 24 hours! 
  • Notably, United and its DME distributor sent correspondence to all Dexcom patients notifying them of the change prior to July 1. We will be interested to hear how many patients go this route in the near-term. Management was not sure if Anthem did the same outreach.

4. Dexcom began its first major CGM awareness campaign during the quarter, fueled by celebrity Nick Jonas. Following the launch of the Nick Jonas/Dexcom promo video, Dexcom’s website had more visitors than in any other day in its history. The month-old video now has over 23,000 views on YouTube, the same number of views as the four-month-old Share receiver video. Nick appeared on 17 interviews with national media outlets, giving Dexcom a huge social media boost. Management called the campaign a “great success.”

  • Dexcom’s next big marketing and awareness push will come around the launch of Gen 5 later this year. Management was frank in stating that awareness of CGM is one of the two biggest challenges as Dexcom keeps growing (the other is getting new products out quickly enough). Of course, Gen 5 should help on both fronts.
  • As part of its evaluation of CGM awareness, Dexcom has conducted a series of focus groups of non-CGM users around the US. Management shared just one anecdote – an MDI user for 20+ years who burst into tears when she learned about Dexcom CGM. “We need to do more,” said CEO Kevin Sayer.

5. In a question on the outlook for CGM in five years, management envisions a miniaturized sensor (“more like a bandage than a medical device”) that is fully disposable and factory calibrated. Of course, this sounds a lot like what Abbott has done with FreeStyle Libre, albeit that device is not continuous. It will be interesting to see what product design tradeoffs Dexcom must make for a fully disposable, continuous transmitter that is still cost-effective to produce at scale. Perhaps it is just a matter of waiting until the component technology gets cheap enough. A major core design choice for Abbott was using NFC in FreeStyle Libre, which helped bring the sensor patch cost down and made it disposable, but sacrifices the continuous data transmission. It is also interesting to think about smartphones and watches with NFC built-in, which would eliminate the need for a separate medical device reader. We assume that would not be continuous, given the limited range of NFC, but it is intriguing to think about potential product profiles. Dexcom is certainly well positioned to look into that sort of design, given its tremendous progress on connectivity with Share and Gen 5.

  • CEO Kevin Sayer believes a minimum of 30% penetration in type 1 in the US is feasible in five years (similar to what pumps are at now). Exceeding that mark will depend on overcoming the biggest barriers: awareness, access, and cost. Dexcom is working on all three of those fronts through marketing, pharmacy distribution, clinical studies, and future product designs. Still, Kevin Sayer was frank in saying that getting to 70% penetration is probably not going to happen. To those three factors, we would add, “I do not want to wear something on my body,” something we hear all the time from non-CGM users. We wonder which of these four factors is holding back CGM the most, and where Dexcom can make the most meaningful progress.
  • Within five years, Dexcom also hopes to make a “very meaningful” penetration into intensive insulin-using type 2s, and potentially broader type 2s. Executive VP Steve Pacelli called type 2 an “ever increasing and important market segment for us.” Management believes a different product will be needed for type 2s – something that is ideally simpler to use, smaller on the body, disposable, and feasible as a diagnostic and behavior modification tool. Again, it sounds similar in approach to FreeStyle Libre, though we assume Dexcom has many ideas on how to build on what Abbott has commercialized. Steve Pacelli categorized this in the “advanced technology pipeline,” though admitted that five years out is a long time for Dexcom – indeed, back in 2010, the Seven Plus had just been commercialized, a reminder of how far things have advanced in half a decade.
  • Dexcom plans to gather more clinical data and support more investigator-initiated studies going forward. Notably, “several studies” will be related to use of CGM for some period of time immediately following diagnosis. It was not clear if these will be in type 1 or type 2.

R&D Pipeline Highlights

6. Management continues to expect FDA approval and launch of the G5 mobile app and new Bluetooth-enabled transmitter later this year (it was submitted in February). In new news, the launch will be for both pediatric and adults – a major win, given the typical lag time between those approvals. Identical to the 1Q15 call, CEO Kevin Sayer said the company has had “excellent dialogue” with the Agency. G5 will launch for Apple iOS to start, but move to Android in the future. There was no timing on an Android version of G5, though Dexcom will move it ahead in lieu of first launching an Android version of the Share app; this is the right call from our view, even if it results in a slight delay. As we understand it, the wider variety of Android phones means more software testing is required than iOS to gain approval.

  • Dexcom has planned the transition to G5 brilliantly from an operational perspective, and current G4 patients will only need a new transmitter, a remote software update of their Share receiver, and a download of the free iOS mobile app. We’re not sure if original G4 Platinum receiver (without Share) users will be forced to buy a new receiver, though we assume they will need to, since the G5 transmitter uses Bluetooth. As a reminder, we learned in 1Q15 that the FDA will likely require a receiver with Gen 5 to serve “as a backup.” In practice, we assume most patients will probably use their smartphone as the primary receiver, so the required receivers may simply sit in patients’ drawers (our wording!) ... From a business perspective, receiver upgrades are expected to decrease with G5.
  • We also learned in 1Q15 that the new G5 mobile transmitter will have a labeled useful life of three months, down from six months for the G4 Platinum and 12 months for Seven Plus. The reason is because the Bluetooth radio requires more power, and Dexcom wanted to keep the Gen 5 transmitter on-body footprint the same as the current G4 transmitter – a tough tradeoff for sure, but we think the company made the right call, all things considered. On the plus side, the shorter transmitter lifecycle will shorten patient upgrades to new platforms. For example, when Gen 6 is launched, the new smart transmitter and corresponding app will be available to all existing patients during their normal purchasing cycles within a few short months. Moving forward, management expects total transmitter revenue to possibly increase.

7. In a major update, Dexcom is now “in sync” with the FDA regarding the pre-market and post-market data required to obtain an insulin-dosing claim for CGM. Management is optimistic this could launch in the US sometime next year. Notably, the FDA has “very strongly indicated” that even without the interference blocking of Gen 6, the existing G4AP data (MARD: 9.0%) is strong enough to support an insulin-dosing claim. Though management previously forecasted a large clinical trial, today’s call was more vague on what will be required – CEO Kevin Sayer brushed off the question in Q&A, perhaps for competitive reasons as Abbott is presumably in the same conversations for FreeStyle Libre in the US.

  • Since Gen 5 uses the same sensor and algorithm as the current G4 Platinum sensor, Dexcom can immediately launch the insulin-dosing label claim for G5. By contrast, the G4 Platinum-integrated pumps (Animas Vibe, Tandem t:slim G4) will not have the dosing claim until they integrate with G5 (since they don’t have the updated G4AP/Software 505 algorithm). For current users, we think this label claim may not be that impactful, since many Dexcom users dose off the CGM readings anyways. However, it could persuade more patients to wear CGM, since the need for confirmatory fingersticks does dilute the value proposition of CGM somewhat.
  • An insulin-dosing claim is most important from a Medicare coverage perspective, though it could also encourage more doctors to prescribe CGM. As a reminder, CGM’s need for a confirmatory fingerstick has been a major sticking point for Medicare coverage. Once that is removed, perhaps the agency will entertain covering the devices for Medicare beneficiaries. Management also said (anecdotally speaking from the sales force) that some doctors might be much more willing to prescribe CGM if patients don’t have to take a confirmatory fingerstick. “It could be a bigger catalyst than we had anticipated in the past,” said management.
  • With artificial pancreas systems coming, an insulin-dosing claim is also important from a labeling perspective. Indeed, it would be odd to use CGM readings to automatically dose insulin (in an AP) if the standalone CGM was not also approved for doing so! Management does not believe an insulin-dosing claim accelerates the timing of AP systems, but it is “important.” Indeed, Steve Pacelli speculated this is “one reason why the FDA is pushing us.” That is clearly a testament to how forward thinking the Agency’s device side has become – they seem to be pushing Dexcom to get this through!

8. Dexcom expects to begin a pre-pivotal study of the 10-day-wear Gen 6 sensor later this year, with a pivotal study to shortly follow, an FDA submission in early 2016, and launch in early 2017. The timing is identical to that shared in 1Q15, though the pre-pivotal timing has slipped from “this summer” to “later this year” (summer technically ends on September 22, so Dexcom may still hit that previous timing).

  • Management gave the most granularity ever on the G6’s calibration scheme, which will likely start with one required fingerstick calibration per day at launch (down from the current two per day). It sounds like this primarily reflects a conservative regulatory approach to get through the Agency “efficiently.” The company will seek to “gradually” reduce the number of required calibrations over time – this step-by-step approach definitely worked well on the connectivity front (Share cradle in October 2014, Share receiver in March 2015, Gen 5 by end of 2015), and it will interesting to see at what pace it progresses on the calibration front. Management was firm in saying that calibration “will be completely eliminated in the future.” These remarks were consistent with the 1Q15 call, which said Gen 6 has potential for factory calibration, but some confirmatory fingersticks may be required for safety/regulatory purposes. 
  • The calibration scheme also has important implications vis-a-vis the insulin dosing claim. Management said that different labeled products may emerge over time as Dexcom operates within the labeling for insulin dosing off of CGM. The company is “breaking new ground” here and it still sounds early to know exactly what could come out in two years time. 

9. Continuing a company focus on software, Dexcom expects to launch a “robust” new data platform developed by SweetSpot in 2H15. Management believes it will “set a new standard” for visualization of CGM data. This is the most concrete timing on a SweetSpot software launch since the acquisition in February 2012. We’d note that Dexcom has already launched Portrait on the Web, which offers a very easy Internet browser upload of the G4 Platinum – for now, Portrait on the Web is very basic, offering modal day graphs, simple statistics, and basic high/low pattern recognition identical to the Dexcom Studio PC software that launched with the G4 Platinum.

  • Dexcom recently hired a Senior Data Executive from outside the medical technology field who has years of experience in Big Data in Silicon Valley. Dexcom’s leadership page has not been updated with any new hires that we can see. The move is definitely needed, given how much real-time data is already flowing with Share and how much more will come with Gen 5.
  • Management did not mention the partnerships with Diasend (launched), Glooko (launched last week), or Tidepool (still in beta), other than saying these are helpful for patients but not a driver of revenues. Still, we appreciated CEO Kevin Sayer’s modesty that if these partners have better answers than Dexcom, “that makes everybody happy.” Dexcom has been very forward thinking on an open architecture approach to its data, and we’ve been impressed to see the ecosystem already developing through data flowing seamlessly from the Share iPhone app to Apple Health – see the recent diaTribe test drive for more specifics on this through the app Meal Memory. We do think this will be a driver of revenue since better apps will make it more likely that patients want to use Dexcom. We’ve been saying CGM needs a “killer app” for awhile, and we definitely believe the field is getting closer to providing one.

10. In line with the 1Q15 call, Dexcom continues to make “good progress” on a new insertion system, a new lower cost, higher quality receiver, and several generations of transmitters. We caught glimpses of these at JPM and ATTD. In Q&A, management expanded on the transmitter comment, where R&D is focused on reducing the size, expanding the battery life, and going disposable. It’s a tough balancing act to jointly optimize the on-body form factor (getting it smaller but not too small, since that limits dexterity), battery life, and cost profile. This is especially critical as Dexcom thinks about potential competition from Abbott’s FreeStyle Libre and even longer-shot products from Google/Novartis (contact lens, “several years away”) and potentially from Sano Intelligence (a wellness-focused CGM patch slated for a 2016 launch).

11. Management echoed 1Q15’s commitment to automated insulin delivery and specifically mentioned Tandem, Bigfoot Biomedical, and the Bionic Pancreas. As a reminder, Tandem announced an expanded partnership last week to integrate G5 and G6; Bigfoot Biomedical announced a development agreement with Dexcom just prior to ADA; and Dr. Ed Damiano unveiled the dual-chamber Bionic Pancreas device at FFL last month. Animas was not mentioned on today’s call, other than a brief comment that J&J has seen “nice growth” in pump sales related to the Vibe launch; as of J&J’s 2Q15 call, a next-gen artificial pancreas system is still in development. Insulet was not mentioned, though we assume the plan to integrate the new PDM with Gen 5 is still on track (Insulet pre-released 2Q15 numbers last week, but has indefinitely postponed its call).

  • As a reminder, Dexcom is committing future technology (e.g., Gen 5, Gen 6) to pump partners only if they can commit to products that go beyond CGM display and offer some measure of insulin control. The 1Q15 call shared Dexcom’s strongest comments yet on the artificial pancreas: “...automated insulin delivery is going to come...We want a product in that space. We want a partner to be in that space.”
  • The big question is how quickly Tandem, Insulet, the Bionic Pancreas, and TypeZero will move on pivotal studies, considering that Medtronic has already started the pivotal study of the MiniMed 670G/Enlite 3. It also remains unclear how tightly the 670G will manage blood sugar, how good the Enlite 3 sensor will be, and how the integrated 670G will compare to Dexcom-integrated offerings from the aforementioned contenders. Lots of uncertainty, but unquestionably plenty of exciting movement and competitive offerings for patients!

12. In a brief ADA recap, management highlighted T1D Exchange data comparing the A1c of MDI vs. pump users on CGM – this was on display in Dexcom’s booth, suggesting very little difference in A1c between CGM+MDI vs. CGM+Pump users: 7.9% vs. 7.8% (<13 years), 8.0% vs. 8.1% (13-26 years), and 7.2% vs. 7.2% (>26 years). Certainly, the data counters the common perception (to Dexcom’s frustration) that CGM should only be added onto a pump (i.e., it’s valuable in MDI users too). We believe the company’s DiaMond study (n=338, 24 months, testing CGM in MDI users), if successful, may help put this to rest.

R&D Pipeline Summary

Pipeline Product

Timeline

Tandem t:slim insulin pump with G4 Platinum CGM integration

PMA filed in July 2014; launch expected in 2H15.

Gen 5 system

[iOS Mobile app, Bluetooth transmitter, Share Receiver, G4 Platinum sensor]

FDA submission in 1Q15; FDA approval and launch by end of 2015.

Insulin dosing claim

Sometime in 2016

Data Management

-Glooko

-SweetSpot

-Tidepool

 

- Launched last week

- Launch in 2H15

 - Launch in coming months

Gen 6 sensor

[New sensor with initial goal of one calibration/day, insulin dosing claim, new calibration algorithm, interferent blocking,]

 

Pre-pivotal trial later in 2015;FDA submission in early 2016; early 2017 launch

Insulet next-gen OmniPod with G5 transmitter/smartphone app integration

2016+ (following approval of next-gen PDM, which is slated for FDA submission by end of 2015)

Tandem next-gen t:slim with G5 and G6 transmitter/smartphone app integration

Unknown

Bigfoot Biomedical artificial pancreas system with G5 and Asante pump

Pivotal study by end of 2016

Bionic Pancreas dual-chamber iLet with integrated G5

Pivotal study in 1Q17-1Q18

Next-gen insertion system

Wi-Fi enabled receiver

New generations of transmitters

Could come before or after Gen 6 sensor

Questions and Answers

Q: Congrats on a great quarter guys. This is now something like three years in a row where every quarter has had 50-60% growth. It looks like every quarter, you keep finding new sources of patients and new areas of growth. Where are the new patients coming from? Is it pumps, is it MDI patients, and how sustainable is this? How important is G5 going to be in growing this patient base?

A: Let’s start with the pumps vs. MDI mix. Still a majority of our patients come from the insulin pump world, but we are seeing an increase in MDI patients. But we haven’t totally moved the needle from the 60/40% split that we talked about earlier. I think the Vibe traction has helped us. I also think the new pump starts on Tandem and Insulet help us, as those patients migrate to our sensor. So we’ve had good success there. They’re coming from across the board. With respect to patient groups, we cannot walk away from how good the pediatric launch has been for us. I was at a show in Florida, Children With Diabetes, and the guy giving a speech about all the therapies asked all the pediatrics and parents, who uses CGM. A bunch of hands went up. He then asked how many use Dexcom, and no hands went down. So we’ve done very well in the peds market, but it’s very much been across the board.

Everything we do helps sustain this growth. And I don’t think anybody can underestimate how difficult it is to grow this fast. To grow a business this fast you have to have new things after new things that enhance the patient experience to get deeper into the patient community. Gen 5 is that next step. We believe that having data going straight to a phone will enable patients to not have to carry an extra thing in their pocket. But we know the next step after that. We have to make it more convenient, we have to eliminate calibration, the hassle factors, and we have to maintain our excellent level of accuracy. All those things matter, and we address every one every time we iterate.

Q: Maybe just one follow up, a question we get a lot from investors. How is the shift to pharmacy going to impact your financials? I was surprised to see ASPs were actually at the high end of the range this quarter. What is the impact on sales and down throughout P&L. Is it going to be beneficial or neutral? How should this impact new patients getting through the door?

A: It should impact new patients getting through very favorably. In many cases, all they will need is a prescription and maybe some limited prior authorization on a phone call. So it will take the cycle of new patients going into system down significantly. It should also reduce their copays. In regards to the pricing of system in the pharmacy channels, there will be some situations where we lower prices to get pharmacies through the door and we should get in some situations where price increases. We hope over time to be price neutral. We have another pricing lever coming with the Gen 6, which has extended wear. So as we move to pharmacy, if we see movement down a little bit or up a little bit in pricing, in the pharmacy channel, we have something coming with Gen 6 system, where we can pick all, or some of that up with the extended wear sensor. We monitor it closely, and we have long discussions before we embark in pricing discussions every time we have a pharmacy contract. But so far so good, even as more of our business has shifted there, we’re doing quite well.

Q: Did you give a % of revenues from hardware this quarter?

A: 30%.

Q: Is that roughly split between new patients and replacement hardware? Because that would put you at an awfully big new patient number, something near 18,000 or even 20,000 new patients in a quarter.

A: We’re not going to break it out, just as we haven’t in past. We should assume that revenue levels that we’re at, and the growth rates we’re experiencing, we’re adding quite a few new patients, but we’re not going to be specific on the new patient numbers.

Q: So the Share receiver launch obviously plus pediatrics have to be two major things that really changed it. But was the momentum actually building though quarter? Since you’ve only had that product for a short time, and obviously there’s a little bit of backlog built through Q1 with seasonality. Was it actually getting even stronger as the quarter went on?

A: The month to month swings in our business are difficult to predict. Some years, all the months are equal in 2Q. Other years it starts fast and slows down. This quarter went as we planned it. We look at our monthly targets every quarter, and the monthly targets did well every month. I don’t think there was demand that was built up for the Share receiver. We fulfilled demand in Q1. That was a very effective and efficient launch by our team. There wasn’t anyone sitting on the sidelines waiting for a product that couldn’t get it. But I think the share receiver certainly helped, continued pediatric growth helped, and all those things continued to add to the business.

Q: You’re at 60% growth for now 10 quarters in a row-ish. You’ve talked about the challenges of sustaining that. What is the primary bottleneck right now? Is it awareness? At these kinds of growth rates, it does take a lot to feed the beast and sustain it.

A: I think awareness is the biggest challenge to keep growing at this rapid rate. Our operations team hasn’t had a backorder since we started. And we’ve gone through all the hassles that any manufacturing plant would go through and these guys are very, very good at fixing them. Another challenge in the future is getting new products out quickly enough. We’ve been iterating very quickly, with all the approvals we’ve had last year and all the approvals we’ve gotten this year. Things move pretty fast here and we need to keep innovating at a robust pace. We keep going.

Q: You mentioned you’ve been evaluating a series of different transmitters. Can you characterize some of the technologies, or the concepts you’re thinking about?

A: Sure. Battery life is one thing we look at and think about. The Gen 5 transmitter has a three-month life, vs. the six-month life of G4. Those of you who’ve been on this call for a long time will remember that the Seven Plus had a 1+ year life. We’ve looked at battery life and extending anywhere from 3-6 months, or going disposable, or looking at costs. We’ve looked at those transmitter options and certainly different electronics configurations. The biggest issue with the transmitter is simply size. The smaller we can make it, the better, to a certain degree. You can’t make it too small or it becomes unmanageable from a dexterity perspective, as we learned in our human factor studies. We’re looking at optimal size and shape and maximum economic benefit.

Q: Is it safe to say that the patient profile for Share is skewing more towards pediatrics and young adults? And you’re maybe actually seeing a turbo boost from that launch?

A: I think it does skew that way. But I would tell you that it is not the only place it goes. I’ve been around a number of adults who are using it. I was in a physician’s office just a week or so ago, where a physician showed me data from a 72 year-old patient, who had severe hypoglycemia unawareness and couldn’t manage himself. He’d moved to small town, to take care of his 90 year-old parents, and instead they were taking care of him, until he got on CGM. Well, Share data is very important for that guy. I think any place where people are experiencing high and low swings, and particularly with hypoglycemia, it’ll make a huge difference. It obviously will help more in pediatric patients. If not for the simple reason that seven-year-olds can now tell their parents they need an iPhone, and “I have to have this if you’re going to watch my data.” An interesting side note of all the Share data is that there’s a whole new culture of “How are we going to interact with CGM?” Parents are freaking out and calling all the time with the CGM. Then kids turn off parents from sharing. And with adults, you have spouses following each other. We’ve seen some very interesting text messages between followers and patients, so it’s across the board, but clearly the concept is being proven out dramatically.

Q: I wanted to ask a question on the various open source programs. You mentioned SweetSpot, but I know you’re also very familiar with Tidepool and Glooko. Can you talk in general about how those might impact business over next couple of years? It seems like there’s a broader concept of sharing information, but there are also seems to be a specific effort to allow Medtronic pumpers to interact with Dexcom devices.

A: Well that’s certainly being taken care of by those entities that are independent of Dexcom. That’s great for patients and health care professionals who want to see data put together. SweetSpot is our tech platform. We’ll ultimately upload pumps into there as well, particularly from our own pump partners.  And we’ll reciprocate pump partners in terms of sensor data being in their own devices. As far as other platforms being a huge driver of our business, it’ll be helpful, it’ll be a nice feature for our patients; we’re not banking on that to drive our revenues. We think it’s nice to have. If they have better answers than we do, that makes everybody happy. So, we’re fine.

Q: On G6, I think based on your comments, your pilot and pivotal will go this year, when do you expect to complete pivotal and submit?

A: The pivotal timing will depend a lot on of what we learn in the pre-pivotal. We expect to submit early next year, and I hope we can launch in early 2017.

Q: You believe with that, you’ll have the 10-day extended wear? Correct?

A: That’s correct.

Q: You’re working on that on G4, to get the dosing claim?

A: That’s correct. We don’t have to wait for Gen 6. The FDA in our discussions has very strongly indicated that even without interference blocking of Gen 6 membrane, our G4 data is strong enough to support an insulin dosing claim. So we’re working with the agency on what work we have to do and what labeling to add the feature to our product.

Q: If G5 is approved, do you still get the dosing claim off of G4? Do you turn around and then baby step to get that dosing claim on the G5? Or would it automatically come with it?

A: They’re the same sensor and algorithm, so more than likely, what we would do is launch G5 with the claim and not go backwards.

Q: So when you get the G4 dosing, it basically applies to G5, and then you have that locked up. On G6, you discussed reduced calibration, what exactly does that mean?

A: When we run studies, we calibrate it, and we have calibration schemes and our patients take calibrations. And we go simulate data in a number of ways. We simulate the data if they only took a calibration on the first day, or in the middle, or in any number of ways. More than likely, in the first pass with this, particularly in looking at timing with the Agency, it would be one calibration per day. But we developed an entirely new algorithm platform for this product, one that can adapt to very few calibrations. So more than likely, I would say our first pass with G6 will be a calibration every single day - not really as much for accuracy, as it is for reliability and safety of the system and looking to get through the Agency efficiently. Then we would look at a number of options after that with filings after the fact. One a day for G6 when we start.

Q: On UNH and Anthem, have those programs started? And if so, when did they start?

A: They were both effective July 1.

Q: What percent of the managed care guys do you think sign up? Or are they waiting for UNH and Anthem to have 12 months of data, and say yes, this is the way to go. Or is this fluid, and we will see more by the end of the year?

A: We could see more by end of the year. We’re in active discussions with virtually every major payer.

Q: And you talked about sensor pricing being in the upper end of the range. Can you explain why that has moved up a little bit? Or why you’ve talked about it at this time?

A: It’s a combination of factors. Any given quarter, given how much of our business is processed directly versus third party distribution, and the mix of which direct contracts vs. which third party distributors, the pricing varies, and that results in variability in sensor pricing. I think there’s been this weird concern on the Street that somehow our revenues would be severely impacted by pharmacy. What you have to keep in mind is that we have to give in some instances, maybe in a case where we have direct contract with payer today and in order to get them to move to pharmacy, we may have to do a slight discount. But remember, 50% historically of our business are through DME distributor. So where we’re looking to establish first time contracts with some of these payers, these guys were getting gouged by 3rd party distributors, so the pricing we can come in at can be quite favorable to Dexcom on an overall ASP basis, and the payer gets a significant rate reduction from what they were paying their DME supplier. What I want to caution though: we’ve never said the pharmacy channel will drive ASPs. Our goal again is to stay net neutral on pricing basis as we shift to the pharmacy.

Just to clarify on the distributor front, some distributors have gouged payers. Certainly not all of them. We’ve spent time a lot of time this year working on those relationships, making them more consistent and more seamless. We have a channel team working with those guys as well and those efforts have been very good.

Q: On the five-year outlook--when I look at CGM, what do you think the percentage penetration in type 1 is right now, where is it going, where do you see it 1 year now, 5 years from now?

A: That’s a great question. Pumps have reached 30% but haven’t had a tremendous amount of growth over the past 10 years from where they were before. I believe at a minimum standard, 30% for CGM penetration to start in US, and I think it should be higher. I think our biggest barriers are awareness, access, and cost. We are doing things on all those fronts to try and make a difference. So about 30%, I don’t think we’re ever getting to 70%. But I think CGM’s should be around 30. If we make these things small enough and easy to use, we can take away most every excuse that people have.

Five years from now, our goal would be to have a sensor that’s miniaturized, that looks more like bandage than a medical device and fully disposable – then you get it to the point where you overcome basically all of the objections any patient may have to wearing the device. It talks to your phone, doesn’t have to be calibrated, and you don’t take fingersticks anymore. There isn’t a reason you couldn’t get very significant penetration in type 1 market in the US, provided we can achieve access via the payer community and continue to drive awareness. So if you rehash our conference call script today, that’s kind of the message: we need to keep our foot on the pedal on growing our commercial enterprise and on the R&D front because we need to continue to iterate new and better technologies to capture a larger and larger share of type 1. I would also add that I think in five years, we should have very meaningful penetration into intensive insulin-using patients who are type 2s, but I would suggest even potentially broader into type 2s. I think that’s going to become an ever increasing, important market segment for us.

Q: On utilization trends in the quarter - did that change at all?

A: No. We don’t have perfect data. We’re starting to capture a lot more data as we move to Share. But we still think patients are using 2.5-3 sensors per month on average. They are still extending the wear.

Q: With UNH and Anthem, you’ve had month of experience with that, any anecdotal information or data that you’ve been able to obtain? When you think about current patients, will they now be able to go to a CVS or Walgreens or get their script refilled at pharmacy vs. going in and calling their distributor?

A: Theoretically that will happen. There will be a few bugs to work out as we get this started and going. Anecdotally, we got an email from a mother, who had a 2.5 year-old with type 1 diabetes, recently diagnosed and 24 hours later she had a CGM from the local drug store. Her email was very grateful. The ability to turn this stuff around very quickly is very important to our business. I think over time, we will be able to move over there, but there are bugs in this like everything else. This is going to take a little bit of work. But we’ll get there.

Q: Is the dosing claim only going to be applicable to the G4 with artificial pancreas algorithm?

A: That’s correct.

Q: So any of the integrated pumps will not have that dosing claim?

A: Not until they integrate with Gen 5.

Q: Share is currently only on the iPhone platform, when do you start thinking about the other platforms, Android and whatnot?

A: The Follow app is on Android. Followers can follow whoever they’re following on either platform. We probably will not move the Share app to the Android platform because we think Gen 5 will be here relatively quickly, and we’d rather put our resources towards that. But we’ll evaluate it as we go.

Q: Will Gen 5 be iPhone only?

A: It’ll be iOS to start, and then it’ll move to Android as well.

Q: On the dosing claim you hope to get next year, could you elucidate what you hope that does for you? CMS has wanted that before they would open up coverage so is that one aspect? When we start thinking of closing the loop, does this move you quicker down that path and when do you think we might see that?

A: You’re right. Initially, our belief was that the dosing claim was most important from a Medicare perspective. Medicare has indicated that until we get the adjunctive labeling removed, they would not entertain a coverage category for CGM. That opens up both type 1 and type 2 insulin using patients. It was really eye-opening this year, when I was at our national sales meeting, talking to the field force, who were much more excited than we were internally about the dosing claim because they tell stories of doctors who are truly unwilling today to prescribe as much as we would like. They’re not prescribing it as much because they have to tell their patient when they prescribe it, “Why don’t you purchase this thing. Oh, by the way, you have to take two fingersticks a day to calibrate it, and every time you eat or take an insulin injection you need to take a fingerstick.” All of this is anecdotal from our sales force, but these doctors would be much more willing if they don’t have to tell their patients that they have to take a confirmatory fingerstick to prescribe for it. I think the dosing claim could be a bigger catalyst than we had anticipated in the past.

Q: Does that precipitate the artificial pancreas - the holy grail here in terms of closing the loop?

A: Yes. The FDA is cognizant that with a fully automated system or even a semi-automated system, you’re going to have to use CGM information to dose insulin. It’s hard to have system that’s dosing insulin that isn’t labeled for dosing insulin, so I think there’s something there as well. I’m not sure it accelerates the timing to approval for artificial pancreas, but I think it’s important.

I think that’s one reason the FDA is pushing us. That’s for sure.

Q: As you start to eliminate calibrations, it’s counter to the idea of a dosing claim. Those two almost oppose each other. How do you marry those as you think about moving forward about different devices in the future?

A: Welcome to what I wake up to thinking about at 3 in the morning. It’s a very interested quandary because if you’re not going to take any fingersticks, there’s no safety check to make sure that everything is well. If you’re going to have a dosing claim, but take no fingersticks, that thing has to work every single time. That’s why I said our first pass with the Gen 6 system will be one fingerstick per day for calibration. You’re going to see us move into this area gradually and we have a number of system options we’re considering so we can stay in the dosing claim world, but eliminate calibrations. Maybe you’ll see different labeled products from us over time to offer patients more convenience and certainly glucose information, but maybe not give them the labeling to dose insulin off this thing. We’re considering a number of things and breaking some new ground here.

Q: One last thing on the DTC program. As you start to get more pharmacy benefit coverage, while Nick Jonas probably resonates with certain audiences, how do you think about expanding those programs in terms of appealing to a different audience?

A: That’s why we’re running the focus groups and really seeing what message resonates with this community. We’ll start launching those messages. The direct to consumer efforts in all reality, are about creating more awareness, and are probably going to coincide more with our Gen 5 launch, in terms of connectivity of the transmitter going straight to the phone. We’re working on this and planning this to be implemented later this year and into early next year. We’ll have more on this; we’re just not quite ready to share.

Q: Following on the type 2 market opportunity, I know you said in 5 years, you could really start to penetrate that market. What will it take to really get you there? Is it going to be product-driven, data-driven - showing clinical trials that type 2 patients really benefit?

A: It’s going to take all of the above. Those two factors, plus third being establishing a reimbursement category. We’re starting to see some reimbursement for intensive insulin using type 2. These are folks who are on basal/bolus, mealtime injections of insulin, and they mostly look and feel like type 1. You need to segment by intensive insulin users versus just a daily Lantus user versus those that are on oral, diet and exercise. I think it’ll take probably a different product than what we have today. Something that’s simpler to use, that’s smaller on body, and these are all the things that we’re exploring internally, but I think to appeal to a much broader patient population, I mentioned disposable, something you can put on as more of a diagnostic tool as a type 2 patient, use it as a behavior modification tool on a periodic basis, and then be able to throw it away when you’re done, that’s quite appealing. These are the kinds of things that are in our advanced technology pipeline. Five years out is a long time for us from a development perspective, so you’ll see some iterative steps along the way.

Q: In your prepared remarks, you make some comments about some of the increased spend going toward feet on the street. Maybe you could flesh that out a little bit for us. Are you talking about OB, ICU, those kinds of offices? I always thought that was more of a G6 timing when you start going into those offices.

A: I can give you a couple of examples. If we have pharmacy benefits, pharmacists are going to need to know about our product. We’ll look at that area. We’ll also look at some more internal medicine doctors and primary care doctors who see a lot of people with diabetes. We’ll pilot some programs in some geographies and see what that generates for us. We’re looking at a lot of things across the board, and just to make it easier for people. One of the stats that Terry often quoted was how many prescribers we had in a year and how many of them prescribe one unit. Last year, we had something like 13,000 health care professionals who prescribed CGM and over 8,000 of them did one. What do we have to do to get those people who prescribed one to do more? They’re not all endocrinologists. They’re across the board. We’ve also got programs targeted more for diabetes educators, physician assistant markets, because those people see a lot of patients and involved in patient care recommendations. We’re going to look at a number of different things outside of traditional endocrinology as to where we’re going to focus those dollars. As you talk about the ICU and OBGYN markets and such, those are places where we’ll start running clinical studies, particularly with our Gen 6 system with 1 calibration per day, when that comes out, it becomes more convenient for example in the hospital environment and more convenient for a patient who doesn’t have diabetes. It’s time to start running those studies and seeing what we can do in those markets. All of the technologies we’re developing for our core business will be applicable to those patient groups and enhancing their lives and improving their healthcare outcomes. So it’s time for us to go there.

Q: Just as UNH and Anthem theoretically ramp here during the third quarter, I know you guys don’t set sensor pricing, but would it be expected that that sensor pricing comes down off $75? Just to set expectations, should we expect to see something in the $70 or the lower end of range?

A: I think it will be somewhere in range of $70-75. I don’t see any change.

Q: Just as we think theoretically about some of the integrated pump stuff you’re doing with G5 and G6, with the Tandem announcement, is it better to scattershot and go with a few different partners and see which one works the best? Or do you control that from beginning because you want to have that spread across all different pumps?

A: We certainly control the way our data is displayed, which respect to how the FDA approved it. Those pump companies refer to our data set, with respect to our PMA filing, so it has to look very similar to ours. As long as it doesn’t alter the data, that’s fine. There will be different views among data as everybody adds their own touch. We’ve learned with our Gen 5 application, if you think you know how CGM data is displayed, it’s all about to change. The G5 system displays things much differently than before. Everyone has their own interpretation. A lot of that will depend on FDA process. It doesn’t make a lot of sense for a pump company to come along and change everything we do completely and drag themselves through an extended FDA review. On the other hand, if they can offer something more beneficial to their patients, that’s great. Remember that with Gen 5 and Gen 6, that transmitter can talk to two things at same time. What our patients are offered as the integrated pump moved to a Gen 5 or Gen 6 platform, is they can look at their data on their pump or their phone. If it’s on their phone, then it goes to their watch as well. I find the watch view very nice. In fact, I’m wearing an Apple watch and sensor today looking at it. There are going to be lots of different ways this is done, given our motto here is “Glucose for everybody anywhere they want.”

Q: On the insulin dosing claim, do you believe you need to run a trial, or do you think you can satisfy the FDA with existing data you have?

A: I’m not going to get into the specifics of what we’re doing. We’re getting pretty comfortable with the pre- and post-market work that we have to do and I’m going to leave it at that.

Q: In terms of the pharmacy channel, with United and Anthem, can you give us more detail on the rollout? Are these national, regional programs? Are the payers actively steering folks to the pharmacy, or is that your responsibility?

A: I can answer with respect to United, but can’t with respect to Anthem just yet. United actually sent correspondence to all of their patients, as did the DME distributor, just prior to July 1, saying that as of July 1, they’ll be able to purchase through their pharmacy. I’m not sure if Anthem doing same sort of outreach, I’d have to check.

Q: Is there any reason why someone would not go through the pharmacy?

A: If they’re used to going through DME – and there are certainly still DME arrangements with United – they might stay where they are. They might have an employer plan that gives them no deductible from DME, so they might make a decision to do it that way. When a patient calls us, we do run the numbers for patients, and we go through both the DME and the pharmacy to determine what is best for them, and give them both options.

Q: It’s fair to assume that going through the pharmacy is the most cost-effective solution for the user?

A: In almost every case it will be. It is certainly the most cost-effective for us. Once that drug store is online, patients won’t be calling us every time they need sensors, so that’s one of our motivations for doing this.

Closing: The era of connectivity is here. We’re learning all sorts of things about our patients, about our product, and the difference we can make in people’s lives through this connectivity. That has been a great, great thing this quarter and we’re very pleased with it. I do want to conclude with a couple of thoughts.

As you know, over the summer, there are many diabetes meetings we attend - AADE, ADA, and several others. I was attending a meeting in Colorado and in a diabetes symposium, an exceptional, Denver-based professional gave a talk and she asked a simple question at the start of her talk: who needs CGM? And her answer was just as simple as her question: everybody with diabetes. She proceeded to present case study after case study of type 2, type 1 patients, all in completely different circumstances, and completely different ages, all of whom experienced a significant reduction in A1c, and or hypoglycemia improvement through the use of personal or professional CGM. CGM tells a story nobody can get from fingersticks or any other medical device.

The second thought comes from a discussion I had with another physician at this conference, a very well-known endocrinologist. We were talking about business and he simply said this: nobody comes to my office anymore and asks for CGM. They come to my office and ask for Dexcom. And that’s a position we want to maintain. We’re not going to move away from our long-term goal of replacing fingersticks and we’re going to become first tool in treating diabetes.

 

-- by Adam Brown, Esther Wu, Tiffany Kha, and Kelly Close