Memorandum

Medtronic 3Q18 (F2Q19) – CGM sales now larger than pump sales; Global sales rise 26% on easy comp; Advanced HCL launch by April 2020 – November 20, 2018

Executive Highlights

  • Worldwide Medtronic Diabetes sales of $583 million grew 26% YOY as reported and 28% operationally in 3Q18, propelled by continued 670G adoption (>135,000 active users) and 70% global CGM sales growth. Notably, the strong gain came on the second easiest YOY comparison in the past decade: sales were flat in 3Q17. Sequentially, global sales grew a modest 2% YOY from 2Q18 – we might have expected this to be higher. US sales of $334 million rose 30% YOY and 3% sequentially, also on an extremely easy YOY comparison (-5% in 3Q17). OUS sales of $249 million increased 22% YOY as reported. The US provided 63% of 3Q18 growth. Much softer high-single-digit growth is expected in Q4 on a tougher YOY comparison.

  • Remarkably, Medtronic’s revenue from CGM now exceeds pump revenue. This comment from CEO Omar Ishrak was the biggest financial highlight of the call in our view – we believe that although Ishrak did not reflect on it, it still signals a rapid shift to become, as some may see it, a CGM company that sells pumps, rather than the other way around. This implies Medtronic CGM sales are now at ~$200 million/quarter or perhaps more, though the company could not comment. We are not clear about the exact economics of what is counted how and where.

  • The MiniMed 670G now has >135,000 active/trained users globally, up a robust 39% from 2Q18 (>97,000 active US users) and nearly double 1Q18’s >70,000 US users. The hybrid closed loop has launched in “select” European countries, with more international launches (including Canada) to come between now-April 2019.

  • The next-gen “advanced hybrid closed-loop” with automated correction boluses and >80% time-in-range goal is now expected to launch in FY20, implying availability in April 2019-April 2020. This is a meaningful ~one-year acceleration in timing from the June 2018 Analyst Meeting, and comments implied it will bring connectivity via Bluetooth. We noticed the NIH-funded FLAIR trial – comparing advanced HCL to the 670G – has an updated start date of March 30, 2019. We believe this speedier timing reflects a more action-oriented company eager to respond to feedback.

  • By April 2019, Medtronic still expects to launch Sugar.IQ’s hypoglycemia prediction feature (up to four hours in advance). A Guardian Connect CGM app for Android is also expected outside the US in the next five months.

Medtronic reported 3Q18 (F2Q19) financial results this morning in a call led by CEO Omar Ishrak. Diabetes is slide #11, and the call’s diabetes highlights and Q&A are enclosed below.

Financial Highlights

1. Global sales of $583 million rise 26% on easy YOY comparison, mostly driven by the US; >135,000 active 670G users, up 39%; CGM sales rise 70%

Medtronic Diabetes Quarterly Sales (1Q13-3Q18), millions


Worldwide
  • Worldwide Medtronic Diabetes sales totaled $583 million in 3Q18, growing 26% YOY as reported and 28% operationally on continued 670G adoption (>135,000 active users) and 70% CGM sales growth. The strong YOY growth came on Medtronic’s second easiest YOY comparison in the past decade: sales were flat in 3Q17 on sensor capacity constraints. Sequentially, sales grew a modest 2% YOY from 2Q18. “Continued” demand for the MiniMed 670G drove the performance, which just launched in “select” European countries in addition to ~18 months of availability in the US. The system now has >135,000 active and trained users globally, up a robust 39% from 2Q18 (>97,000 active US users) and nearly double 1Q18’s >70,000 US users. CGM growth was also a major point of emphasis: CGM sales grew 70% globally and 90% in the US, driven by the US launch of the standalone Guardian Connect (June) and continued CGM uptake in 670G/640G users – see the highlight below for more on CGM sales, which are notably now bigger than pump sales. The US provided 63% of the quarter’s growth, in line with 2Q18 and well above the 0% share one year ago (3Q17). Sales did come on the upper end of guidance, which called for growth of ~20%, “plus or minus.”

US
  • US sales of $334 million rose 30% YOY and 3% sequentially, also on an extremely easy YOY comparison: US sales declined 5% in 3Q17 on sensor capacity constraints. Still, this was the second-highest level of reported US growth ever seen in our Medtronic model, only slightly exceeded by 2Q18’s 33% YOY US growth. Prepared remarks and the slides noted “sustained” and “continued” US market demand for the 670G, a good sign for Medtronic – both Tandem and Insulet reported record Q3 results. Tandem’s Basal-IQ/G6 is a particularly competitive product in the US, offering no fingersticks with G6, PLGS (the bottom-half of hybrid closed loop), Bluetooth and remote monitoring, and a software-updateable pump. The current quarter will reflect far more pressure in the US, as Medtronic’s comparison is significantly tougher (+15% US growth in 4Q17) and Basal-IQ will have a full quarter on the market.

OUS
  • Outside the US, sales of $249 million increased 22% YOY as reported, including 22% YOY operational growth in developed markets ($203 million) and impressive 40% operational growth in the smaller emerging markets ($46 million). The YOY comparison was tougher than in the US, as OUS sales grew 7% YOY in 3Q17. Sequentially, OUS sales were essentially flat (+0.4%) from 2Q18, an indication that 670G is rolling out slowly and in a limited, country-by-country fashion. The slide deck noted some OUS growth was driven by recent Guardian Connect launches in Israel and Korea. While OUS sales should see 670G-propelled strength in the coming quarters, like the US, the Q4 comparison will be much more difficult (+20% YOY).

Guidance
  • FY19 sales guidance was maintained: “low-to-mid-teens” diabetes growth. In the upcoming quarter (4Q18), management expects diabetes growth to decelerate to the “high single digits.” Indeed, the upcoming Q4 comparison to 17% growth will be the toughest Medtronic has faced in seven years, and the one after that is even harder (+26% YOY). Assuming Medtronic delivers 8% YOY growth, that would put global sales at $630 million, still behind 1Q18’s blowout sales growth.

2. CGM Sales Now Larger Than Pumps (!), A Major Milestone for Medtronic

“Revenue from CGM now exceeds our pump revenue and is establishing a consistent, long-term and dependable revenue stream for our diabetes group.” This comment (under-emphasized) from CEO Omar Ishrak was the biggest highlight of the call in our view – it signals a rapid shift as Medtronic has moved closer to a CGM company that happens to sell pumps, rather than the other way around. While this has been a strategic goal for some time now, the sales crossover is a big-time milestone. Medtronic could not share specifics with us, though it’s clear our last update from the company (mid-2017) – CGM at ~25% of global sales – is now very outdated. (That would back out to ~$146 million in CGM sales vs. ~$437 million for the rest of the business.) We assume CGM sales are now well-north of 30% and perhaps even over 40% of the global business now – it is never exactly clear what Medtronic is counting under what category (some time ago, part of each pump was valued as CGM). All in all, this move might put Medtronic CGM at a ~$200 million quarterly run rate (or perhaps more), with pump sales under that and the remaining diabetes revenue from other areas (e.g., Diabeter). Medtronic could not comment on these estimates, and we’re the first to note that the revenue recognition assumptions here can get confusing – e.g., when Medtronic sells a 670G system, what is “counted” in CGM vs. pump revenue? Still, the takeaway is clear: Medtronic has built a significant ~$800 million/year business in CGM and is seeing growth via both sensor-augmented pumps and the standalone Guardian Connect. For context, Abbott’s FreeStyle Libre seems poised to break $1 billion in 2018 sales (see its 3Q18) and Dexcom expects ~$975 million in sales in 2018 (see its 3Q18) – we also think there’s certainly some chance for that business to outperform even more, though the $975 million itself reflects a fantastic annual performance – or outperformance!

  • As noted above, worldwide CGM sales grew 70% in 3Q18, including robust 90% growth in the US. This represents an acceleration from 2Q18’s “nearly 50%” worldwide CGM sales growth. The press release added that CGM sales tied to pumps (e.g., 670G, 640G) grew in the mid-60% range operationally.

  • Medtronic’s Emerging Technologies division – which includes Guardian Connect and professional CGM – grew >100% YOY, driven by the ongoing US launch of Guardian Connect/Sugar.IQ. This represents a very notable acceleration from 2Q18, when this division saw growth of “>50%.” Sales were not broken out, but given overall Diabetes growth of 26% as reported roughly backs out to Emerging Technologies at ~8% of the business on sales of ~$47 million, while the AIM division is at ~92% of the business on sales of ~$536 million. We had Emerging Technologies at ~$30 million in 2Q18, implying the division might have seen ~50% sequential growth – a good sign for US uptake of Guardian Connect.

3. 670G Launched in “Several” European Countries; Real-World Time-in-Range Continues >70%, Mirroring Pivotal Trial

In line with the EASD update, Medtronic has launched the MiniMed 670G in “several” European countries, with more international launches to come in the next six months. The specific countries were not shared, though at EASD, initial launch countries were Belgium, Denmark, Finland, Italy, Netherlands, Slovenia, Spain, Sweden, Switzerland, and UK/Ireland. Management noted that “initial user feedback has been very positive” outside the US. It’s not clear how many of the >135,000 active users are from the US vs. OUS, though we imagine it’s still <10,000 users Europe at this point (e.g., as of August, the US had over 97,000 users alone). Last month’s Health Canada approval for 670G was not mentioned, though we assume the launch expectation for “fall 2018” (i.e., by mid-December) still stands. In both Europe and Canada, Medtronic is first-to-market with hybrid closed loop, with Diabeloop poised to be second-to-market starting in France in early 2019.

  • “We continue to generate strong real-word clinical outcomes with time-in-range range exceeding 70%.” We love hearing real-world data on time-in-range – especially when it bolsters pivotal/regulatory data – and wish more companies would share their global time-in-range. For now, achieving >70% time-in-range is the watermark for systems, and we’ll be excited to see where various next-gen products stack up: (i) adding automated correction boluses and lower targets to further boost time-in-range (e.g., see below; Tandem’s Control-IQ); (ii) combining CGM + smart pens + insulin dosing decision support to “close the loop” for injectors; and (iii) better forward-looking CGM prediction and pattern recognition via smarter algorithms and multi-sensor data streams (e.g., sleep, activity, habits).

Pipeline Highlights

1. Advanced Hybrid Closed Loop with Auto Boluses: Launch by April 2020 (moved up!) with Remote Monitoring, >80% Time-in-Range Goal

Notably, Medtronic’s next-gen “advanced hybrid closed-loop system” with automated correction boluses is now expected to launch in FY20, implying availability in April 2019-April 2020. This is a meaningful ~one-year acceleration in timing from the June 2018 Analyst Meeting, which called for a launch of this next-gen system “beyond” FY20 (i.e., after April 2020). The timing is consistent, however, with the FY20 launch plan for a Bluetooth-enabled pump and a mobile display app, which it sounds like this pump will add (see Q&A comment below). It’s unclear when in the FY20 window Medtronic will launch this product, which is critical from a competitive perspective – Tandem expects to bring its no-fingerstick hybrid closed loop with automatic bolusing (Control-IQ/G6) to the US in summer 2019, which is not very far away if the timing holds. We imagine Control-IQ will launch ahead of Medtronic’s Advanced HCL, as Tandem’s pivotal trial is fully enrolled (see 3Q18). In Q&A, Mr. Hakami said three feasibility studies of advanced HCL have been done. The NIH-funded FLAIR trial – comparing advanced HCL to the 670G – has an updated start date of March 30, 2019, with completion expected by December 31, 2019. The ClinicalTrials.gov page also lists FDA as a collaborator, which could imply FLAIR will serve as the Advanced HCL pivotal. Indeed, the crossover trial will enroll n=112 participants at seven big-name centers with three months on each system. Medtronic’s goal remains to deliver >80% time-in-range with this product, in line with Dr. Fran Kaufman’s commentary from Keystone. This will finally bring to market the DreaMed fuzzy logic algorithm combined with the 670G’s basal modulation.

  • Will Medtronic launch any 670G enhancements before this advanced hybrid closed loop? We had previously thought a Bluetooth-enabled 670G would come before the advanced hybrid closed loop, though today’s call positioned them as the same product. In early October, we learned that Medtronic had submitted a 670G enhancement to the FDA to reduce unnecessary, repeated requests for fingerstick BG checks; however, this is only a transmitter software update and there was no mention of Bluetooth.

  • Hooman Hakami in Q&A on Advanced HCL: “Our next generation algorithms will improve time in range to over 80% by automating insulin delivery following a snack or a meal. In addition, the system will reduce the burden of carb counting, enable remote monitoring, and automatic software downloads…We’ve done a lot of work, we've got three feasibility studies done…” We do note that of course the system can be designed to reach a certain time in range, but ultimately it is up to patients to see where time in range lands.

2. Sugar.IQ Hypoglycemia Prediction: On Track to Launch by April 2019

In line with the Analyst Meeting, the long-awaited Sugar.IQ hypoglycemia prediction feature is expected to launch by April 2019 (“in the back half of this fiscal year”). The four-hour prediction feature is designed for the companion app for Guardian Connect users: You are expected to experience a low within the next four hours.” This hypoglycemia prediction has long been a big – and cool! – selling pitch of Medtronic/IBM’s Sugar.IQ, but is running almost two years behind the 2016 Analyst Day hope to launch by “summer 2017.” It will be fascinating to see how much of a value-add this feature is for its standalone CGM – Medtronic has previously claimed ~90% hypoglycemia prediction accuracy, though notification fatigue and continued Sugar.IQ engagement will be key to watch. Of course, for those at high risk of severe hypoglycemia, this could make a very meaningful difference.

  • Sugar.IQ is only available in the US on Apple iOS and only for Guardian Connect standalone CGM users. Sugar.IQ has a better-than-average 3.5/5 stars (n=28 ratings) in early use though there’s obvious potential upside there. The closest CGM-data-focused comparators to Sugar.IQ are arguably: (i) Dexcom Clarity Mobile, which has 2.4/5.0 stars on iOS (n=110); and (ii) Glooko, which has 3.0/5.0 stars on iOS (n=52). We note that most CGM apps have historically had a negative bias (Dexcom G5/G6 included), though the just-launched US FreeStyle LibreLink on iOS appears to be an exception – it has 4.4/5.0 stars and n=110 reviews. Diabetes apps from mySugr (4.7/5.0, n=1,900) and One Drop (4.5/5.0, n=6,300) have set the high watermark for reviews – both in terms of positivity and quantity.

3. Guardian Connect Android OUS Availability Between Now-April 2019

Though not mentioned on the call, the diabetes slide said a Guardian Connect CGM app for Android will be available between now and April 2019 (“H2 FY19”). To date, the standalone CGM app has only been on Apple iOS, a competitive disadvantage relative to Android apps for FreeStyle Libre, G5/G6, and Eversense.

4. Next-Gen CGMs: No timing, but iCGM Standard, Non-Adjunctive Labeling, Less Calibration, Smaller, Longer Lasting in Pipeline

No timing was shared on Medtronic’s next-gen CGM sensors, though the company continues to move towards non-adjunctive labeling, iCGM labeling, reduced need for calibration, smaller and longer-lasting, and using IBM Watson to enhance personalized insights. From our view, Medtronic is ahead of the competition on analytics, but well behind on form factor and factory calibration. In Q&A, Mr. Hakami said the next-gen CGM timing is on schedule with the 2018 Analyst Meeting expectations:

  • Harmony CGM sensor with “adaptive calibrations” and “non-adjunctive” claim – launch by April 2020. At ADA, Medtronic expected a PMA filing for a non-adjunctive CGM label claim “this year” – it was unclear if this referred to calendar or fiscal year, though the given the current accuracy (MARD 9%-10% on two calibrations/day), we don’t imagine huge issues obtaining this. This claim will be important for marketing against G6 and FreeStyle Libre, as well to obtain Medicare therapeutic CGM reimbursement. Medtronic said in June that an iCGM designation is “absolutely” in its pipeline, though the accuracy needs to improve a bit. We saw some Medtronic self-calibrating CGM data at DTM earlier this month, though things still seem in the feasibility stage on this front.

  • Envision Pro (formerly iPro 3 professional CGM) outside the US – launch by April 2020. The ADA Briefing said that Envision Pro will be blinded, fully disposable, no-calibration, and Bluetooth-enabled (single-use data transfer, like Libre Pro).

  • Personalized guidance/intelligent therapy advice for T2D professional CGM – launch by April 2020. This could be Pattern Snapshot 2, which we saw at EASD and looks outstanding for giving clinical decision support to HCPs. Dr. Vigersky said it will launch to a small number of clinics in 4Q18.

  • Disposable, no-calibration, 10-14-day wear CGM – launch beyond April 2020. As of ADA, a clinical trial for a Medtronic no-calibration real-time CGM was expected to happen this year, with a slightly more aggressive planned launch within two years. We’re not sure how this is progressing, particularly the disposable part.

  • Combo CGM-insulin set with extended wear – launch beyond April 2020. As of ADA, talks were ongoing with the FDA to run a trial for a seven-day wear infusion set, with a goal to start before the end of this fiscal year (by April 2019). Unomedical’s coated Lantern catheter reported mixed seven-day wear results in a poster at ADA. Earlier this month at DTM, we learned a large multi-center study of coated Lantern is slated for 2019. An upcoming Stanford study will examine ten-day coated Lantern wear.

5. Not Mentioned: Inner Circle, 670G Guarantee, Mio Advance, discontinued Pro-Set, MDI Decision Support, Other Pipeline Products

  • Inner Circle CGM Gamification for Guardian Connect and 670G users – launched in 2Q18 – we’d love to hear more on this

  • MiniMed 670G Performance Guarantee for Payers – launched in 2Q18

  • 670G Enhancements to Reduce Unnecessary Calibrations – Submitted to FDA, per our October coverage

  • Unomedical Mio Advance inserter in the US – FDA cleared in March and available outside the US; when will it come to the US? Medtronic told us at EASD that it is seeing major demand for the set outside the US, and the lag time between clearance and US launch reflects the need to build sufficient supply. Will we see a launch this year?

  • “Sugar.IQ Dosing Assistant” and insulin pen dose capture – as of the Analyst Meeting, the launch was expected by April 2020

  • Pipeline products 2+ years out: “glucose prediction” for MDI users; a combo CGM-insulin set with extended wear; personalization through AI & Cloud; “real time coaching”; and a real time professional CGM (Envision/iPro 3) for type 2 diabetes.

  • MiniMed Pro-set with BD – discontinued as of BD 3Q18 call – we now look back on our writing about this as a bit naïve to really believe that Medtronic would have wanted to partner on this and share the upside

Questions and Answers

Q: There's been a lot of noise on two of your key product lines in the market this quarter both in diabetes from some competitors about the timing of your next-generation pump and then in this spinal cord stim market about what the true growth on a volume basis is there. You put up great results in both, so can you talk about the sustainability of the spinal cord stim market growth and then comment on reiteration of the fiscal year 2020 for your next-generation pump?

[…]

Hooman Hakami (EVP, Diabetes Group): On Diabetes, your question about our advanced hybrid closed loop systems, there's been no change to the timing of that product versus what we said at Analyst Day and also at ADA. We know who is communicating this information and it’s just incorrect. There's no changes from the timeline and it's just as simple as that. I'd say beyond those timelines we’re just really excited by what this technology and what this product promises to bring. We’ve done a lot of work, we've got three feasibility studies done, those feasibility studies indicate a time and range that is close to 80%. To put that in context, the person without diabetes has a time-in-range of 85%, so we're really excited about that. And we take a look and compare our performance against published data from what’s out there on competitive systems, our nearest competitor is going to have a time-in-range that's less than what we have with our 670G system and with a higher targeted glucose range. So timelines are holding. We're making great progress, and we’re excited about what this system's going to be able to do.

  • Editor’s Note #1: The new timeline is actually accelerated. Medtronic’s Analyst Day expected a launch of the Advanced Hybrid Closed Loop “Beyond” FY20 (after April 2020), while today’s call indicated a launch in FY 20 (April 2019-April 2020). We note that adding Bluetooth functionality was expected in FY20, though during the Analyst Day that was positioned as a separate pipeline product from the advanced HCL. Now it seems Medtronic has combined the two into a single launch in April 2019-April 2020.

  • Editor’s Note #2: The time-in-range point is an exaggeration based on recent data. Robust data from EASD using the G6 (Peters, Beck et al.,) showed people without diabetes have a time-in-range of 97%, and that used the much tighter target of 70-140 mg/dl. While 80% time in the wider 70-180 mg/dl range is phenomenal – especially relative to how people with type 1 diabetes do right now – this remains very far from 97% of the time spent in the tighter range of 70-140 mg/dl. We would note, however, that achieving 80%+ in 70-140 mg/dl is possible on AID now, but the biggest driver is food choices – not algorithm design.

Omar Ishrak (CEO): Let me also add that we are very excited about these two segments. We're going to own these segments, and I am personally engaged with both of these groups on a very regular basis. I'm interested in it, and I'm very confident that we will not only introduce these products, as we talked about, with the features that we talked about, but we will lead in these markets.

Q: It looks like the implied second-half guidance for Diabetes is mid-single digits. I know it’s high-single digits for the third quarter but a little lower in the fourth quarter just on an implied basis. Is that how we should think about Diabetes going forward? And is there any update on the Harmony Sensor timeline? And what's the regulatory pathway for Harmony and 690G? I heard the fiscal 2020 launch.

Mr. Hakami: As far as the back half goes, you heard that we’re reiterating from Karen the fact that we expect to grow for the full year low-to-mid-teens. This obviously translates to an applied deceleration of growth from what we've seen over the past four quarters. There's no doubt about that. But let me just remind you that, if you go back to last year, during the first half of last year, we had CGM capacity constraints that prevented us from not only meeting full CGM demand, but also pump and integrated CGM system demand. Then, we started to increase that capacity in Q3 of last year, and we’re at full capacity in Q4. And so we had a lot of pent-up demand. We had some backorders. And so that revenue catch-up that basically came in the back half of last year is impacting our comps for the back half of this year.

The other dynamic I'd point out is Animas. We are anniversary-ing two out of the three revenue components in Q3 of this year. So all of the consumables revenues, all of the out-of-warrantee conversions are really going to no longer provide a year-over-year benefit. Now, that's just commentary on comps, but if you take a look at it, there's going to be these quarterly fluctuations. We're really excited about the opportunity that we have here. We talked about our pipeline, Advanced Hybrid Closed Loop in FY 2020, which we're committed to delivering. This is a multiyear growth opportunity for us, for the company. And we intend to innovate and to lead.

Now very quickly from a CGM perspective, what you’re going to see from us are sensors that are going to come out with non-adjunctive labeling, iCGM standard. And then following that, what we're going to achieve are smaller sensors, longer-life sensors with continued accuracy that also feed in cognitive capability into it. And so as far as the sensor pipeline goes, there's really been no change versus what we've said at either the Analyst Day or ADA, and the teams are executing against the plan that we laid out.

 

--by Adam Brown and Kelly Close