Memorandum

Dexcom 1Q18 – G6 limited launches to big enthusiasm, full US launch in early June; sales accelerate to $184 million, +30% YOY; 2018 guidance raised – May 2, 2018

Executive Highlights

  • Dexcom’s worldwide 1Q18 sales grew 30% year-over-year (YOY) to $184 million, exceeding expectations. Sales declined 17% sequentially (seasonality), though this was still Dexcom’s third highest sales ever following two straight all-time records. The company’s Q4 momentum is clearly continuing. Gross margin fell slightly to 65% (YOY) from 66%, and net loss totaled $24 million, down from $42 million.

  • US sales of $145 million grew 25% YOY (“robust,” even with Libre competition), while international sales rose 49% YOY to a record $39 million. US sales declined 23% sequentially from the strong Q4, but still drove 69% of Dexcom’s 1Q18 growth (US sales still represent 80% of total sales, after all.) OUS sales are seeing continued momentum, exceeding 20% (just!) of total sales for the first time in Dexcom’s history – very notable, even though these sales contribute less to the bottom line.

  • As noted in our test drive on Monday, Dexcom’s factory calibrated G6 CGM has begun a limited US launch (<1,000 patients), with a larger launch starting in early June before ADA. The team is “sprinting” to get manufacturing running at full scale, building up inventory, working with payers (especially on pricing), and learning from early users. CEO Kevin Sayer said initial excitement for G6 has been “phenomenal” in the field. We used G6 for the first time on Monday, and our initial test drive shares our experience – it is as good as we expected, headlined by strong factory calibration accuracy with a short two-hour warmup (labeled MARD is 9.0% with the new applicator), extending to 10-day wear, and bringing a step-function improvement on the inserter. The new transmitter is slimmer than we expected, and the clearance for interoperability as an iCGM (510(k) pathway) seems gamechanging for future innovation.  Inputting a calibration code is a bit of a hassle compared to Libre but overall, not a huge negative.

  • Notably, management increased 2018 revenue guidance to $850-$860 million (~19% YOY growth), up on the bottom and top end from the prior $830-$850 million. This does imply a slowdown in the back half of 2018, reflecting some caution related to competition and pricing headwinds.

  • Dexcom is currently “putting plans in place” to file G6 with FDA so that the receiver becomes optional. G6 has already been filed with Medicare; a decision is expected this fall. CMS talks on remote monitoring/app use “hope” to resolve in 2Q18. 

  • Dexcom/Verily gen one is in “validation and verification,” and Dexcom will run a clinical trial in 2H18 testing 14-day sensor wear. This first disposable configuration sounds like a 2019 launch to us, though no specific timing was shared.

Dexcom announced 1Q18 financial results today in a call led by CEO Kevin Sayer, CFO Quentin Blackford, and EVP Steve Pacelli. Almost to the day of his seven-year anniversary participating in Dexcom calls, Mr. Sayer put it well: “This is the most anticipated, exciting, and complicated period we’ve ever been associated with.” Yes, it is! Read the G6 launch, financial, and pipeline highlights below.

Table of Contents 

G6 Highlights

1a. G6 begins limited US launch (<1,000 users so far), with full US launch starting in early June; big early enthusiasm, including in our test drive

  • In line with our report on Monday, management confirmed that the factory calibrated G6 CGM has begun shipping in “limited quantities,” with a larger scale US launch starting in early June before ADA. Technically, the G6 FAQ page says “June 4,” but this specific date was not mentioned today. The limited launch is <1,000 patients so far, and all patients are still on their first 10-day sensor wear. This impressively fast rollout comes right on par with “late Q2” launch expectations following FDA clearance on March 27. Dexcom is currently getting manufacturing running at full scale (this is the company’s biggest ever manufacturing change, given the new applicator), building up inventory, working with payers and distributors, and learning from early users in the limited launch. On the latter, Mr. Sayer said the team has already “learned a few things.” The entire company is “sprinting to get [G6] out the door,” and we expect a big presence at ADA for the no-calibration, 10-day wear system – especially to compete alongside FreeStyle Libre, which is also making its first ADA appearance as an FDA approved product.

  • CEO Kevin Sayer said initial excitement for G6 has been “phenomenal” in the field – one teenager in the limited launch summarized it as “beautiful, incredible, and unbelievable.” Joked Mr. Sayer, “Do teenagers say that about anything other than their phones?”

  • We inserted and used G6 for the first time on Monday, and our initial test drive shares extensive observations on every piece of the system. Overall, it is as good as we expected - bringing factory calibration with a short two-hour warmup, maintaining G5’s performance (G6’s MARD is 9.0% with the new applicator), 10-day wear, and bringing a serious step-function improvement on the inserter. The new transmitter is slimmer than we expected, and the clearance for interoperability as an iCGM (510(k) pathway) could be gamechanging for the field and for Dexcom - especially for accelerating pump, pen, and software partners’ efforts. G6 sustains Dexcom’s app leadership (iOS, Android, smartwatches), keeping it ahead of Abbott and Medtronic in the US. Never in its history has Dexcom had a product that so meaningfully updated every component of the system.

    • CEO Kevin Sayer: “The thing that I have heard the most about is the insertion system and its ease of use. Text messages I get do talk about no calibration too, and that is awesome, but the insertion of this system and ease of use is such a step up – it’s almost indescribable. This is also big for new patients that are foreign to Dexcom – the old inserter looks like another great big needle. We just saw a video of a child inserting his first G6, and he said, ‘I don’t feel anything – is it on? And it was.’ That experience from a new-to-CGM patient is going to be very, very important. Of course, we cannot ignore the other things. No calibrations with continued Dexcom performance puts us in a class no one else is in. That could lead to increased compliance and possibly increased utilization over time. We can truly eliminate fingersticks from the equation. We couldn’t be more excited about a product launch than this one.”

  • See our test drive for complete commentary on the no-calibration startup process (four-digit code must be entered/scanned for each sensor), massively improved applicator (it delivers big time), meaningfully slimmer transmitter (much closed to the G4 transmitter), and much more.

    • G6’s user guide shares the labeled accuracy for the first time: with the new applicator, MARD was 9.0%, with 93.5% of points within 20 mg/dl or 20%. An impressive 91% of values are within 20/20 on day 1. This is slightly buried in table 10, appendix F in the user guide in an n=62 trial testing the new applicator. In the larger pivotal study – which used manual insertion (as we understand it) – overall MARD was a similar 9.8%, with 91.7% of values within 20/20. It is surprising that G6 does not massively exceed the accuracy benchmarks set by FDA as part of the iCGM special controls. See an updated system-to-system comparison here.

    • What could be improved in G6, in our view? The heavy-duty touchscreen receiver is the weakest part of the system in our opinion, moving away from the sleeker G4/G5 first-gen receiver towards something more durable (the tradeoff is more clunkiness). Our G6 app was not able to scan the sensor calibration code via photo, which could have been user error or a bug; manual entry of the four-digit code was not hard, but could also be improved and is higher hassle than Libre. Obviously, moving to a disposable G6 transmitter and optional receiver purchase would be nice upgrades too (see the pipeline section below). The G6 app has meaningfully reduced the size of the CGM value/trend bubble vs. G5’s interface – we’re not sure why this change was made, but we preferred the larger display on G5. Overall, these issues are fairly small and the overall improvement with G6 vs. G5 is definitely bigger than we expected a year ago.

1b. Our G6 (no cal) Real-World Accuracy Over 48 Hours vs. G5 and Accu-Chek Guide BGM – Very Strong

  • G6 has had excellent accuracy in the first 48 hours of our first sensor wear without any calibration – we’ve seen very little discrepancy between fingersticks (Accu-Chek Guide), G5, and G6 readings, even after the very short two-hour G6 warmup: the first G6 reading was 127 mg/dl, while the meter read 130 mg/dl. G6 has also had smoother readings than a newly inserted G5, with less jumpiness. (Adam is wearing a sensor on each side of his abdomen.) See the pictures below for hours 1-24 (G6 vs. G5) and hours 24-48 (G6 vs G5) – they track very well together. The table further below the pictures shows the point comparison between both systems and an Accu-Chek Guide BGM. It’s clear to us, from this tiny sample (n=1), that Dexcom’s G6 has maintained the accuracy and reliability of G5, while eliminating fingersticks.

Hours 0-24 – G6 (no-calibration) vs. G5 (two calibrations/day) worn simultaneously:


Hours 24-48 – G6 (no-calibration) vs. G5 (two calibrations/day) worn simultaneously:

* Adam is running the G6, G5, and Loop apps all at once on an older iPhone SE. The data gap in the above happened overnight - the G6 app quit in the middle of the night, something Adam has also seen with the G5 app. We’re not sure if that relates to running so many apps together, Adam’s old phone, or something else.

 

 

G6

Accu-Chek Guide BGM

G5

Sensor start

11:40 am

-

12:07 pm

t=0 hours (first reading post two-hour warmup)

127 mg/dl

130 mg/dl*

117 mg/dl*
(fingerstick: 122 mg/dl)

t=1.5 hours

115 mg/dl

107 mg/dl

104 mg/dl

t=3 hours

124 mg/dl

100 mg/dl

121 mg/dl

t=4.5 hours

135 mg/dl

124 mg/dl

120 mg/dl

t=5 hours (post walking)

110 mg/dl

100 mg/dl

86 mg/dl

t=7 hours (post dinner)

225 mg/dl

197 mg/dl

181 mg/dl

t=9 hours

176 mg/dl

154 mg/dl

141 mg/dl

t=10 hours

135 mg/dl

136 mg/dl

114 mg/dl

t=18 hours

97 mg/dl

93 mg/dl

104 mg/dl

t=24 hours

118 mg/dl

128 mg/dl

129 mg/dl

t=28 hours (post bike ride)

71 mg/dl

95 mg/dl

74 mg/dl

t= 30 hours

75 mg/dl

89 mg/dl

79 mg/dl

t=36 hours

102 mg/dl

122 mg/dl

130 mg/dl

t=43 hours

76 mg/dl

80 mg/dl

101 mg/dl

t=46 hours

94 mg/dl

109 mg/dl

99 mg/dl

t=49 hours

170 mg/dl

179 mg/dl

164 mg/dl

 

* Adam started this G5 sensor 27 minutes after G6; for t=0, a calibration fingerstick of 122 mg/dl was entered twice to start G5 – this resulted in G5 displaying 117 mg/dl. Beyond t=0, all G5, G6, and BGM values were taken at the same point in time.

2. G6 sensor pricing discussions still “very, very early”; possible upside for expanded access, monthly bundling, pharmacy channel?

  • Dexcom is still “very, very early” in the pricing/payer discussions on the 10-day-wear G6 sensor. For now, management remains conservative and is internally assuming similar sensor pricing, utilization, and per-patient revenue as G5 – i.e., net pricing of ~$70 per sensor. Given that G6 extends wear to 10 days and eliminates fingersticks, this level of pricing would be a better deal for payers. If Dexcom keeps the same per-day net sensor pricing (~$10/day), it would see more revenue per sensor ($100/sensor) – something we would like to have seen for some time (we have encouraged a “shut-off” for many years) and very positive from our view. That said, things are early and management did not share a lot of specifics – despite many analyst questions. Management acknowledged, unsurprisingly, that the “environment” is competitive on pricing (i.e., FreeStyle Libre), and that Dexcom is prepared for lower pricing with future lower-cost products.

  • G6 does give Dexcom a chance to trade lower pricing for pharmacy access, Medicare monthly bundled pricing, or even broader access into type 2s: “There is some upside, if we maintain a per-day pricing similar to G5, we get a slight uptick in [per-sensor revenue]. We could use some of that upside, if we negotiate a lower price per day, for improved access in the pharmacy, for potential bundled solutions, for adding intensive type 2s to a bundle, etc. … There are lot of moving parts.”

3. G6 already filed with Medicare, decision expected this fall; Share/app discussions continue with CMS, with hope to resolve by end of Q2

  • Dexcom has moved quickly to file the G6 system with Medicare, with a decision expected this fall. The ~$250/month therapeutic CGM pricing won’t change, a huge win for Dexcom since it will only need to supply three sensors per month with G6. We’re not sure if Dexcom will continue to supply 60 Contour Next strips per month (two per day) and an Ascensia BGM. G6 does not need fingerstick calibration, though we wonder if patients will still use at least a couple strips per week – and whether Dexcom will offer this.

  • Discussions/advocacy continues with Medicare regarding Share (remote monitoring) with G5 – CEO Kevin Sayer hopes for an update by the end of this quarter (2Q18). “If we cannot get this resolved,” he said, “We’ll come up with a product configuration in our portfolio that does. We cannot let seniors go without Share. I’m really hopeful I have a positive announcement on the next call.” This is consistent with February’s call, which hoped to resolve this issue “in the next six months.”

    • Senate and House Diabetes Caucus leaders recently sent a letter to CMS urging Medicare to allow use of CGM smartphone apps – this should hopefully provide some political impetus to get this resolved. The two-page letter, signed by Susan Collins, Jeanne Shaheen, Tom Reed, and Diana DeGette, notes that the current Medicare restriction “does not make sense” and CMS should work “expeditiously” to support the full spectrum of CGM use – i.e., receiver alone or receiver/smartphone app. This has been a big issue ever since therapeutic CGM received coverage last March – if a beneficiary uses an app for viewing data, the CGM supplies are non-covered. Accordingly, Dexcom had to disable app communication for G5 Medicare users.

4. No G6 international timing update; presumably on track for 2H18 launch following CE Mark

  • No timing was shared on an international launch of G6. We assume the March expectation is still on track: an international launch in 2H18 following CE Mark. Management has continually referred to 2018 as a “global launch” for G6, though has not been more specific. We assume Germany will be the biggest focus for G6 initially, as Dexcom has had a lot of success there.

  • Getting the much-improved, factory calibrated G6 out in Europe will be a big priority to compete with Abbott’s accelerating FreeStyle Libre adoption – as of Abbott’s 1Q18 call, FreeStyle Libre had more than 650,000 users globally, with over 90% of those outside the US.

5. iCGM designation to accelerate interoperability; “rigorous bar” for accuracy; special controls a “significant barrier to entry”

  • Management commented enthusiastically on G6’s FDA clearance as the first “integrated CGM” (“iCGM”) within a lower-risk 510(k) pathway: the interoperability drives clarity into the product roadmap, while the special controls set a “rigorous” accuracy bar for others to meet.

    • EVP Steve Pacelli said that “certain product configurations” were previously ruled out because of the class III requirements around CGM – taking 9-12 months to get these out would not have been worth it under the old regulatory path. Now in the 510(k) iCGM pathway, some of these configurations –  particularly on the software side – “may be put back on the table.” No specifics were offered, but given the number of use cases for CGM (e.g., insulin dose decision support, pump and smart pen integrations, display on wearables, use within partner apps), we imagine many potential products will fall into this.

    • Dexcom has had conversations with its insulin delivery partners, and believes the iCGM designation “could accelerate interoperability.” Tandem was not called out specifically, though its 1Q18 call last week said it hopes to integrate and launch G6 within the Basal.IQ (PLGS) system by the end of this year – a remarkable turnaround given the Basal.IQ/G5 product was only filed in January and expected to launch this summer! (For comparison, the Tandem-Dexcom G4 t:slim took ~2 years to be updated and approved to incorporate G5!) Mr. Pacelli also alluded to last week’s JDRF/Helmsley Charitable Trust meeting on AID Interoperability – see our highlights coverage here. He noted that this area is “still early,” but Dexcom believes the “iCGM category is a major step in the right direction.” We wholeheartedly agree!

    • Dexcom aims to propel a “steady cadence of highly competitive (AID) systems.” Mr. Pacelli noted “potential to leap frog current products,” borrowing language from Tandem 1Q18 call in referring to the Control.IQ system with automatic correction boluses. We’re glad to see Dexcom focused on driving an ecosystem of insulin automation around its sensor – well ahead of other companies. For comparison, Abbott is only partnered with Bigfoot at this point, and Medtronic has everything in house. Senseonics does have an AID partnership with Roche and might be used in Beta Bionics system, though the latter is not official yet.

  • In line with the March call, CEO Kevin Sayer noted the competitive moat created by the iCGM pathway with special controls: “a rigorous bar for others that aim to use this regulatory path” and “a significant barrier to entry.” He even noted that some Dexcom employees wondered if next-gen Dexcom sensors (e.g., 14-day wear) will need to meet the accuracy standards – “Yes, yes we do.” Indeed, Dexcom will need to run clinical trials to ensure its 14-day version of G6 meets the bar for special controls (more in the pipeline section below). Based on our read of the data: (i) Senseonics appears to meet the bar, assuming PRECISE II and PRECISION are included; (ii) FreeStyle Libre (real-time) falls short of the hypoglycemia standard; and (iii) Medtronic’s Guardian Sensor 3 (two cals/day) slightly misses the outlier requirements for euglycemia and hyperglycemia. Since current CGMs’ labels don’t align with the thresholds for iCGMs, there is some subjectivity to this analysis, and it’s impossible to be certain. We’ll be interested to see if other companies go for the iCGM path and how quickly.

Financial Highlights

1. Sales of $184 million rise 30% YOY (better-than-expected); US grows 25% to $145 million, driving most of growth; OUS grows 49% to $39 million

Dexcom Quarterly Sales (1Q12-1Q18)


  • Dexcom’s worldwide 1Q18 sales totaled $184 million, rising 30% YOY and exceeding the prior expectation for ~$168 million in (max) 1Q18 sales. In line with the typical seasonality, global sales declined 17% sequentially. This was Dexcom’s third highest sales ever following two straight all-time records (Q3: $185 million; Q4: $221 million), an impressive feat in the first quarter of a year – especially with Abbott ramping FreeStyle Libre in the US. The YOY gain did come on an easier comparison for Dexcom to 23% growth in 1Q17. Still, the performance was clearly better-than-expected, attributed to strong sensor volumes, international sales, and new patient adds. Management also noted that 30% growth sits well above the 2018 sales guidance for ~19% growth for the year ($850-$860 million) – this provoked a lot of analyst questions, since it implies conservatism in the back half of 2018. More on that below.

  • US sales of $145 million grew 25% YOY for the second straight quarter, driving 69% of 1Q18 growth. US sales declined 23% sequentially, identical to the 1Q17 sequential decline a year ago. Management called the US business “robust,” and despite more competition, said there was “nothing out of the ordinary” in the quarter. Dexcom met expectations for sensor usage and goals for new patient adds. “We’ve not seen Dexcom losing patients to Libre,” said CEO Kevin Sayer. He added that with Medtronic (soon) launching Guardian Connect (May-July limited launch), Dexcom has “not heard excitement for it” in the field” – the twice-daily calibration and clunker on-body experience are obviously big drawback relative to G6 and FreeStyle Libre. Mr. Sayer added that where Dexcom has gone head-to-head in Australia, Dexcom holds ~70% market share – even with Medtronic priced lower.

  • Record-high international sales of $39 million rose 49% YOY and a robust 17% sequentially, driving 31% of Dexcom’s 1Q18 growth. OUS sales exceeded 20% of total sales for the first time in Dexcom’s history, up meaningfully from 15%-18% in 2017. This business remains small but has been on a nice steady growth trajectory – we’ll be interested to see if G6 brings a step-function change for the OUS business, since factory calibration will be critical for competing with FreeStyle Libre abroad.

    • In line with comments from JPM, Dexcom expects Japan approval “any time now,” with a hope to launch before the end of 2Q18. A launch in Korea is presumably still expected in 2H18. There was not update on reimbursement discussions in the UK or France.

2. 2018 Guidance raised to $850-$860 million (+19% YOY), up meaningfully from prior $830-$850 million

  • Dexcom meaningfully raised 2018 revenue guidance to $850-$860 million (~19% YOY growth), up on the bottom and top end from the prior $830-$850 million. We felt the previous guidance was on the conservative side (given the Q4 momentum), so this feels more reasonable – especially with a better-than-expected Q1 now in the books. The new guidance has a lot of puts and takes – including potential purchasing delays as G6 launches in early June, revenue deferral from the G6 upgrade program, and “other market conditions” (e.g., what happens with G6 sensor pricing and utilization, broader Abbott and Medtronic launches, potential Senseonics launch, etc.).

3. Net Loss of $24 million; gross margin of 65% down slightly from 66% in 1Q17; strong leverage as operating expenses grow 11% vs. 30% sales growth

  • Dexcom reported a GAAP net loss of $24 million in 1Q18, a significant improvement from 1Q17’s $42 million net loss. Net loss on a non-GAAP basis was a bit higher at $28 million, reflecting: (i) $7 million in income related to Dexcom’s equity investment in Tandem (see 4Q17 report – Dexcom purchased 2.5 million shares at $2/share); and (ii) a $3.3 million non-cash interest expense related to senior convertible notes.

  • Gross margin was 65% in 1Q18, down slightly from 66% in 1Q17 and down meaningfully from 69% on 4Q17’s record sales. Management attributed the margin decline to: (i) a charge for obsolete receiver inventory as Dexcom accelerates the move to the touchscreen receiver; (ii) more sales from international and Medicare, where Dexcom has lower margins; and (iii) deferred revenue related to the G6 clearance. Gross margins are expected to fall in the range of 65%-68% for the year, consistent with the prior guidance.

  • The business again showed positive operating leverage – operating expenses of $150 million grew 11% YOY in 1Q18, one third the rate of revenue growth (30%). For the year, operating expenses (excluding investments into non-intensive programs) are expected to grow 10% vs. 2017, about half the rate of expected revenue growth.

4. Where could CGM penetration in type 1 diabetes go? Kevin Sayer thinks “at least 80%” is not a stretch

  • In response to a question on expected CGM penetration over time, CEO Kevin Sayer said the following: “Type 1 penetration of CGM should be at least 80%. I don’t believe that’s a stretch. I don’t know how anyone could manage the disease without it. Over time, as we refine our portfolio, with G6 no calibration, connectivity, interoperability – we can lead that charge. I truly believe penetration intensive type 2 diabetes should be just as high. They similarly need data to manage this condition and their overall health.”

    • For context, the latest estimate from the T1D Exchange last fall suggested CGM penetration stands at ~24% in the US. We assume broad penetration in type 1 is lower when non-T1D Exchange centers are included. Certainly 2018 should see meaningful expansion with two factory calibrated sensors on the market (FreeStyle Libre, Dexcom G6), Medtronic launching its own standalone system (Guardian Connect; May-July limited launch), MiniMed 670G expanding its launch, and Tandem launching Basal.IQ (this summer). Automated insulin delivery – either pump- or smart-pen-based) should hopefully be a killer app for CGM.

    • Many people with type 1 can barely afford insulin and other co-pays, and many people have managed for years without CGM; still, we absolutely agreed that patients deserve CGM and will do much better on it. 

Other Pipeline Highlights

1. G6 receiver optional – “putting plans in place” for FDA filing

  • Dexcom is currently “putting plans in place” to file G6 with FDA so that the receiver becomes optional. Right now, it must be purchased, just like with G5. Medtronic received the first US approval for smartphone-only display with Guardian Connect in March, following a very lengthy FDA review. This application consists of “risk mitigation and human factors,” and considering that Dexcom has done all the hard work on standalone CGM apps for G5 and G6, this doesn’t seem like too much work. (Certainly, Dexcom has the strongest and most productive FDA relationship of any company in diabetes tech.) Dexcom already has this indication in Europe, and only ~40% of patients purchase the receiver, even where there is reimbursement. This will be a meaningful advantage for reducing the upfront cost of starting on CGM, since it will move a G6 start to transmitter, sensors, and the free app.

2. Verily Gen one in verification and validation; 14-day wear trial to be run in 2H18

  • In a call dominated by G6, there was no specific launch timing update on Verily gen one, though this sounds like a 2019 launch to us. The system is in “validation and verification.” Today’s call confirmed that Dexcom will run a clinical trial in 2H18 testing 14-day wear with the G6 sensor. This is the configuration that Dexcom plans to use in the Verily product, and we’d assume with the big change to a disposable transmitter, a launch this year would be a stretch. With the iCGM pathway and special controls, Dexcom is discussing internally how large the 14-day study needs to be – as a reminder, the special controls for iCGM accuracy focus on the lower-bound of the 95% confidence interval, meaning studies have to show a pretty high degree of confidence in sensor performance – i.e., larger study. For context, Verily gen one was last slated for a late 2018/early 2019 launch as of JPM in January. The more recent February call said timing depended on G6’s FDA path – the latter will of course inform the study design ahead.

  • CEO Kevin Sayer also brought up some business model implications – what happens if a 14-day-wear sensor falls off on day four? Will Dexcom send a new sensor? Will Dexcom send a new sensor if it falls off on day 12? “I don’t know how that works yet,” he said. We appreciated this candor! With every new innovation, Mr. Sayer looks at three pillars: (i) “performance – we will never compromise sensor performance”; (ii) “convenience – see what we’ve done with G6”; and (iii) “cost – G6 will cost less to manufacture than G5.” Dexcom will consider all those things as it goes forward on Verily and other products.

3. G6 lower-cost transmitter to launch in late 2018 or early 2019

  • Wasting no time, Dexcom already has a G6 lower cost transmitter that is “wrapping up engineering,” with a launch expected in late 2018 or early 2019. CEO Kevin Sayer hoped this filing might go faster in the 510(k) pathway. We assume this transmitter will have the same on-body profile as the current G6 transmitter, which already lies quite flat on the body and fits snugly in the sensor cradle. See pictures in our test drive.

4. Verily Gen 2 still slated for 2020 launch, will have a new insertion device

  • The Verily Gen 2 penny-sized CGM – with a far smaller and lower-cost disposable transmitter – is still slated for a 2020 launch. With all of the effort Dexcom put into G6’s new applicator, the company will do it again with this product – it will have a new insertion device and be as big of a manufacturing change as for G6. Mr. Sayer comments that there is a lot of research going on in this R&D process, and the team will have to be disciplined about what to include between new electronics, wires, algorithms, sensor membranes, etc. This product “contemplates many, many changes.” Mr. Sayer added that Dexcom R&D investment in this project is already “really big,” and of the two systems with Verily, this second one is the “biggest project, is the second verily product.

5. CGM pilot in type 2 with UHC: 1,000 patients enrolled, more updates closer to end of 2018/early 2019; hospital, prediabetes, gestational, obesity longer-term opportunities

  • Dexcom and UnitedHealthcare have now enrolled ~1,000 people with type 2 diabetes in their 10,000-person pilot testing CGM, connected devices (e.g., Fitbit), and coaching. Dexcom will not be publishing clinical data in a medical journal, but will share anecdotal experience closer to the end of 2018 or early 2019. Goals of this program include driving cost savings and defining Dexcom’s CGM business model for non-intensive users – just in time for the Verily gen one to launch, and certainly by the time Verily gen 2 is ready. This exciting initiative was first announced in November, received a lot of air time at JPM and CES, and is clearly moving ahead quite quickly.

  • Dexcom remains interested in pushing CGM into the hospital, prediabetes, gestational diabetes, and obesity. “CGM offers significant value beyond core intensive insulin users.” Mr. Pacelli was candid that G6 may not be the right product for these groups, but it lays down an important sensor platform for the future. We assume simplifying the hardware (disposable) and pushing cost down will drive much of this going forward. Of course, there is so much runway in intensive diabetes management alone that it may be 3-5 years before Dexcom meaningfully pushes ahead here (our speculation). Still, it was notable to hear these groups mentioned a couple times today.

 

 

-- by Adam Brown and Kelly Close