Tandem 2018 Analyst Meeting – Control-IQ launch in summer 2019 with iPump designation; t:sport in 2H20; record Q3 sales of $42-$44 million rise 56%-63% YOY; first OUS sales – September 25, 2018

Tandem held its first-ever Institutional Invest and Analyst Meeting today, gathering ~80 people in San Diego to review the company’s automated insulin delivery plans, international strategy, and long-term growth strategies. Slides from the 3.5-hour webcast are posted here, many of which are enclosed below.

Overall, it was an extremely positive and persuasive presentation on Tandem’s near-term automated insulin delivery pipeline (strong initial Basal-IQ reception, iPump FDA filing plans, Control-IQ hybrid closed loop to launch in summer 2019), first international sales (~$2 million), next-gen pipeline after Control-IQ (miniaturized t:sport with app control in 2H20), and five-year patient base goal (triple from ~66,000 to 175,000). The company is well positioned to compete with Basal-IQ and Control-IQ near-term, with clear OUS upside and potentially more MDI appeal with t:sport.

Read the key highlights below from Tandem speakers – CEO Kim Blickenstaff (very short opening remarks), CFO Leigh Vosseller, COO John Sheridan, CCO Brian Hansen, Director of Connected Health Xavier de Anda – and guest speakers Dr. Patrick Keith-Hynes (TypeZero) and Laurel Messer (Barbara David Center).

Pipeline Highlights

1. Interoperable pump (iPump) FDA de novo submission to start in October for t:slim X2; separate PMA for Control-IQ algorithm; Control-IQ system launch still “summer 2019”

In a major change in regulatory strategy for the Control-IQ hybrid closed loop, Tandem has decided to submit the t:slim X2 pump to FDA as the first interoperable/integrated “iPump” (de novo application), followed by a separate PMA submission for the Control-IQ hybrid closed loop algorithm. There is no change to the previous “summer 2019” launch goal for the integrated hybrid closed loop system with Dexcom’s G6 iCGM and automatic correction boluses. Tandem has been in conversations with the FDA and will take the “formal steps” to submit the t:slim X2 as an iPump in “October” under a de novo application – following Dexcom’s lead with the G6 iCGM clearance. This submission will be followed by a separate PMA submission for the TypeZero Control-IQ hybrid closed loop algorithm (no submission timing given), which will include the clinical data from the ongoing iDCL pivotal trial (expected completion: April 2019, per CT.gov). Tandem will work with FDA to create the new device class of iPump that can be used in an AID system with interoperability in mind – this will separate the regulatory submission for the pump from the algorithm, moving towards a plug-and-play ecosystem that is simpler for FDA and Tandem moving forward (more on this below). Presumably the TypeZero algorithm would then be approved and labeled for use with an “iPump” (t:slim X2) and iCGM (Dexcom G6). Tandem believes it will be the first iPump, though we know that at least SOOIL hopes to do this too with its Dana pump (per ADA 2018). It did sound like FDA talks are pretty far along with Tandem on this front. For context, Tandem previously expected the entire Control-IQ submission – pump, embedded AID algorithm, CGM – to be a single PMA, with a “modular” (staggered, stepwise) PMA submission to begin in September. It’s impressive this new strategy, which will go into FDA one month later, is not expected to change the summer 2019 launch timing. Assuming Tandem hits summer 2019, it is highly likely to be second-to-market in the US with hybrid closed loop – Bigfoot, Lilly, Beta Bionics, and Insulet are converging on ~2020-2021 launches. (See our AID Landscape, updated last week.)

• Tandem and FDA will work to define special controls around an iPump, which will likely focus on “pumping accuracy” and “communication parameters” between the CGM and algorithm. COO John Sheridan clarified in Q&A that most pumps are “essentially equivalent”  in terms of pumping accuracy, so this is “not going to be the area that differentiates” Tandem’s t:slim X2. Obviously it will be in Tandem’s best interests to set a high bar for the iPump special controls – as we saw with the iCGM special controls for accuracy when Dexcom’s G6 was cleared. We look forward to watching how FDA thinks about this and if other pumps will choose to go this route. For commentary on iPump, see our coverage of JDRF/HCT’s Interoperability meeting in April.

• What’s the advantage of an iPump + iCGM + algorithm? One immediate benefit is a faster path to get t:sport with Control-IQ to market. Tandem will not have to do an additional PMA and new clinical study with t:sport. As long as t:sport meets the iPump standards (still to be determined) and it communicates with Control-IQ and iCGM as expected, Tandem will not need a clinical trial or a PMA – it can get t:sport cleared under the 510(k) pathway using the t:slim X2 as a predicate device. This has allowed Tandem to pull t:sport’s expected launch timing to “2H20” instead of the previous “2020-2021.” (More on this below.) Like Dexcom’s G6, Tandem will also be able to make changes to the pump component of its AID system without requiring a whole new PMA each time– a significant win for iterative innovation.

• COO John Sheridan highlighted the advantages of Basal-IQ’s iCGM compatibility: “It took us a year to go from G4 to G5; it took a couple of weeks to go from G5 to G6.” As a reminder, Tandem submitted Basal-IQ to FDA with G5, but once G6 was cleared with iCGM in March, Tandem worked with FDA to get Basal-IQ compatibility with an iCGM. Tandem had to take the iCGM specs and show that Basal-IQ would still work effectively – Mr. Sheridan said Tandem had to focus on corner cases and show that the algorithm continued to suspend/resume insulin in an appropriate manner with the G6 iCGM. Basal-IQ was technically approved in June with both G5 and iCGM compatibility; for simplicity, Tandem chose to go forward with the G6 iCGM only. As we noted at the time, this was a remarkable, forward-thinking move from FDA: “We made the decision to change from G5 to G6, and all we had to do is change the labeling in the manuals – it took a couple of weeks…iCGM has reduced the regulatory burden on FDA and companies like us.”

• Tandem has not decided on pricing to upgrade the t:slim X2 to add Control-IQ via remote software update. As a reminder, Basal-IQ was launched as a free software upgrade for t:slim X2 users – an industry first! However, there is clearly consideration of whether Tandem wants to charge some amount for the hybrid closed loop upgrade. We think it would show tremendous leadership not to charge anything, though even something like $99 could be attractive for Tandem – that alone would boost revenue by >$6.5 million if the entire installed base upgrades.

• Management was vague about integrating other iCGMs with the t:slim X2: “Whether or not we include other iCGMs is TBD…We’re working with Dexcom now, they have the best CGM on the market.” Noted Mr. Sheridan, “We have a great relationship with Dexcom/TypeZero, and our business agreement defines access to the technology. Dexcom is only company with an iCGM. If Dexcom goes from G6 to G7, and G7 is an iCGM, all we have to do is change the labeling to use the G7 – that’s going to take us a few weeks. That’s the benefit we have.” This came in response to an analyst question about the dynamics now that Dexcom owns TypeZero – would Tandem even be able to include another CGM or another algorithm in its system? Mr. Sheridan clarified, as we’ve heard before, that integrating another iCGM would definitely require some business relationship with the other company – even if no new clinical study would be needed.

• Under this regulatory paradigm – separating the iPump, iCGM, and algorithm – Tandem would likely need a new PMA and clinical study if it decides to significantly change its AID algorithm. For instance, if Tandem moves from Control-IQ to Control-IQ+ with some new fancy features, it would have to work with the FDA to define what is required. Obviously big changes would need new clinical studies. Said Mr. Sheridan, “The algorithm is the one thing that is still going to require clinical data.”

• “Interoperability is changing the regulatory landscape. There’s been a lot of progress, sponsored by the FDA.” Mr. Sheridan showed similar slides to what Dr. Courtney Lias has shown over the course of this year following G6’s clearance – the advantage of creating the iCGM category with special controls, which will help avoid duplicative FDA PMA submissions and lag time between sensor generations. The third slide below was new, showing the advantage of an iPump – again, no new PMA, only a labeling update.

Old PMA process for AID: CGM change requires a new PMA, new clinical data

New PMA Process for AID: iCGM change requires labeling update, no new PMA or clinical data

Potential PMA Process for AID with iCGM & iPump: iPump change requires a labeling update, no new PMA or clinical data

2. t:sport launch in 2H20: half the size of t:slim X2, no screen, smartphone control (or handheld), Control-IQ algorithm, iPump, bolus button

In the biggest update on t:sport ever, Tandem now plans to launch the miniaturized, screenless, wirelessly controlled durable tubed pump in the second half of 2020. t:sport will be submitted to FDA as an iPump with Control IQ algorithm and iCGM integration, meaning no new clinical data will be needed. Tandem expects t:sport to be controlled using a mobile app on a user’s own smartphone or via a separate touchscreen controller – “it’s a matter of preference of what you choose to use.” (This was refreshingly firm commentary – sounds like FDA is very open to this.) This design would put t:sport on par with plans from Bigfoot and Lilly, who also aim for screenless hybrid closed loop pumps controlled wirelessly from a phone. Tandem has mostly finalized t:sport’s hardware design (only small iterations left), and the team is starting to work on manufacturing to build the product. Notably, Tandem will switch to a syringe-driven pump design, a major departure from the current micro-delivery shuttle approach. That also means scaling the manufacturing will be very key to watch, since it is a new pumping mechanism for Tandem. Assuming Tandem can hit its timing, it will have three straight years with major product launches: Basal-IQ in 2018 (August), Control-IQ in summer 2019, and t:sport in the second half of 2020.

• t:sport has been in development for a remarkable ~9 years, and as noted in the feature set above, improves on the current pump in a number of ways: half the size of the current t:slim, smartphone control, wireless charging, water resistant (IPX8: submerged up to eight feet for two hours), and a quicker/easier fill process. t:sport will retain Bluetooth communication, iCGM integration, and remote software update capabilities. The screenless pump will include Control-IQ and G6 iCGM communication embedded in the pump, but add an on-device bolus button – this means users will remain in hybrid closed loop when the phone/controller is out of range, but retain the ability to take a bolus directly on the device. t:sport will move down to a 200-unit reservoir from the current 300-unit reservoir, a strategic size tradeoff since presumably most type 1s’ needs will be met with 200 units.

• Tandem will have “multiple infusion set options,” and the pictures implied both short and long infusion sets will be available – e.g., t:sport could be worn on the arm with a short infusion set (i.e., patch-like, similar to Cellnovo), or in the pocket with a longer infusion set (i.e., like a traditional tubed pump). The main goal with t:sport seems to be miniaturization and moving control to the phone, rather than being truly tubeless and trying to copy the Omnipod.

  • “Why not make this a true tubeless patch?” Slides noted a few interesting reasons why: (i) people have the freedom to disconnect with t:sport; (ii) t:sport preserves insulin if the infusion set fails (adhesive, occlusion); and (iii) t:sport preserves Tandem’s current durable pump reimbursement model. To this we’d add the design, profitability, and manufacturing challenges of making a truly tubeless, all-in-one patch like Insulet’s Omnipod – it’s far easier for Tandem to remain a tubed pump at this stage. Obviously t:sport is not a perfect competitor to Insulet’s Omnipod, though it will be interesting to see if Tandem sees stronger uptake in MDIs with t:sport. (Currently, ~50% of Tandem’s new patients are from MDI vs. ~80% for Omnipod.)

• Part of the move to t:sport will allow Tandem to add “consumer electronics features” that weren’t available when the original t:slim was developed and designed 7-8 years ago – we certainly like the move to wireless charging, something Beta Bionics is also planning with the iLet Gen 4.

• Said Mr. Sheridan, “This is going to be a product that really does differentiate us. This is a home run as far as I’m concerned.” Now that Basal-IQ is on the market and Control-IQ is in the clinical regulatory phase, “t:sport” is fully resourced internally and Tandem’s “top development priority…we’re putting our foot on the gas.”

• In a recent >50-person focus group in four cities, t:sport “did very well” when compared to expected competitive offerings in 2020-2021. It was not clear how Tandem framed competitors expected products, but presumably it was based on information in the public domain Mr. Sheridan said that relative to the competition, t:sport and the t:slim X2 “gained significant preference share.” Notably, t:sport was “extremely well received” in the MDI community, which would certainly be a win for Tandem as it looks to expand the market and compete with Insulet.

• Since t:sport pulls out the pump’s screen, the gross margin profile is expected to be better over time vs the t:slim X2. What’s unclear is how long it will take to scale manufacturing and see improved gross margins. Management did not give specific numbers or timing expectations for manufacturing at scale, but acknowledged that t:sport is “much simpler” and there will be “some cost savings.”

3. t:slim X2 first-gen mobile app (secondary display) beta launch by end of 2018, additional features throughout 2019; smartphone control by 2H20?

Tandem expects to launch its paired t:slim X2 mobile app in a beta launch by the end of this year, with additional features to launch in 2019. The timing is slightly back of the 4Q17 plan to launch in 2H18. As expected, the app will allow users to view real-time pump/CGM status on their own smartphone (i.e., secondary display without having to pull out the pump), enable remote monitoring by caregivers, and automatically upload data to the web-based t:connect platform. Tandem’s new Director of Connected Health called the app “a baseline for much more to come with this wonderful platform.”  Over time, Tandem hopes to add many features to the app: Apple Health app integration; biometric authentication (presumably a fingerprint or Face ID for a bolus); integration with diet, sleep, and exercise information; decision support; and smartwatch integration. 

• What about smartphone control? Management expects direct smartphone control will come for the t:slim X2 before t:sport launches, implying a launch of smartphone pump control could happen in 2H20 or even sooner. “We’ve had meetings with FDA in the last month or two and they wanted us to implement phone control. There has been ambiguity in past, but that is gone now. They are expecting we have the capabilities to do that.” Great news!

• Notably, Tandem has over 55,000 users on t:connect, a large majority of its users. This represents an impressive ~83% of its >66,000 users shipped a pump in the last four years and ~70% of the 78,000+ pumps since Tandem’s inception. (It’s true installed base is somewhere in between, as some users from 5-6 years ago are still on their Tandem pump and have not renewed at four years.)

4. Next-gen AID: “Advanced hybrid closed loop” with no carb counting, personalization; “Fully Closed Loop” w/ full bolus automation; no timing shared

Today offered a peek at longer-term AID plans beyond the Control-IQ hybrid closed loop algorithm: advanced hybrid closed loop (no carb counting for meals, personalized settings) and fully closed loop (bolus automation). No timing was given, though these seem like launches beyond 2020, given the plan to include Control-IQ in the 2H20 t:sport. Mr. Sheridan emphasized that “AID does not have to be hard to use or hard to train,” and the existing Control-IQ algorithm is already “best in class” with >300,000 hours of clinical testing and automatic correction boluses. When it came time to add it to the t:slim X2, Tandem didn’t want to change anything – the goal was to get through the FDA as quickly as possible. Now both Tandem and TypeZero are “just beginning to work” on next-gen algorithms. Optional carb counting, meal adaptation and personalization, and flexible glycemic targets would certainly be at the top of our list. Tandem expects to give more timing color once Control-IQ is on the market. Management was also clear that fully closed loop is “not going to get there in the near-term.”

• “We work closely with the DIY community, we have gotten feedback, and we know the things they value – the innovative features, the great user interface features. These are things we can implement on our pump as well. [In talking to] DIYers, they don’t want to have to do what they’re doing. At Tandem, we feel it’s our obligation to develop products that meet their needs – so they don’t have to do [AID] in a non-regulated manner.”

Financial and Business Highlights

1. Estimated Q3 sales of $42-$44 million, +56%-63% YOY and a record-high; first international sales of ~$2 million

Tandem reported record-high estimated 3Q18 sales of $42-$44 million (GAAP), an impressive 56%-63% YOY gain on a fairly tough comparison to 20% growth in 3Q17. This just edged out the previous record of $40 million in blowout sales in 4Q17, represents a robust 26% sequential increase from an already-strong 2Q18, and provides a great indication that Basal-IQ is seeing nice early uptake following the August US launch. On a non-GAAP basis – excluding Tandem’s technology upgrade program – 3Q18 sales grew an even more robust 81% YOY.

• 3Q18 saw Tandem’s first international revenue of ~$2 million following an under-the-radar launch in August (first disclosed today). Tandem has launched in Italy, Scandinavia, Spain, UK, South Africa, New Zealand, and Australia. This is actually quite strong uptake only ~1 month into the launch (~5% of quarterly sales), and is expected to be followed by ~$5-$6 million in international sales in 4Q18. The t:slim X2/G5 remains under Health Canada review, with a launched planned for 4Q18.

• Management fielded several questions about OUS channel stuffing in these early days. Tandem is currently seeing distributors buying at a similar cadence to the US – on an ~2-week cycle. Said CFO Leigh Vosseller, “Most of those distributors don’t have the ability or financial backing to do large stocking orders.” Added COO Brian Hansen, “Most distribution channels are fairly small in nature, and they’re not going to buy crazy amounts of stock.” Obviously with inventory issues common for Insulet/Ypsomed, many will be watching OUS sales closely.

2. 2018 Sales Guidance Increased $10 million on both ends: $150-$158 million, acceleration to +39%-47% YOY; $7-$8 million in OUS sales

Management significantly increased 2018 sales guidance to $150-$158 million (+39%-47% YOY), up $10 million on both ends of the range from the previous $140-$148 million (+30%-38% YOY). The enthusiasm stems from the “strong domestic sales” in 3Q18 and the encouraging OUS launch in select countries (Italy, Scandinavia, Spain, UK, South Africa, New Zealand, and Australia). Noted COO Brian Hansen, “Our pipeline is as large as it’s ever been.” The updated guidance includes a projected $7-$8 million of international sales for all of 2018, presumably baking in a planned 4Q18 launch in Canada (still under review). This major guidance raise follows an already-significant 2Q18 guidance raise of +$8 million from the original $132-$140 million expectation for 2018 (+26% YOY). Compared to 2017 sales (GAAP) of $108 million, which grew 28% YOY (the lowest full-year in Tandem’s history), the company has seen an impressive acceleration this year – see the chart below.  

• An initial slide from CEO Kim Blickenstaff flashed “$1 billion in sales,” though CFO Leigh Vosseller said this is not a five-year target. Management noted a “solid long-term growth trajectory” with four key drivers: (i) product pipeline; (ii) renewals; (iii) Animas conversions; and (iv) international. As she has on recent earnings calls, Ms. Vosseller noted that renewals will continue to be a big driver of growth, though it will take Tandem time to reach its goal of a 70% retention rate today – “We’re still actively renewing people from 2012.” In other words, people are holding onto their pumps more than two years after their four-year warranties expire! In 2012 Tandem shipped ~1,000 pumps, with ~6,500 pumps in 2013 almost 11,000 in 2014. Also in line with recent calls, Tandem has seen “higher conversion” from Animas patients than in the past. The company estimates there were ~45,000 Animas patients in the US and another ~45,000 OUS.

• Following the launch of Basal-IQ, Tandem plans to have a conversation with UnitedHealthcare. “This [product] is exactly what we were asked to do.” As a reminder, UHC announced in May 2016 that Medtronic would become the preferred supplier of pumps, hurting Tandem’s 2H16 and 1H17 sales. Tandem hopes to have that discussion in the next 90 days. (Insulet announced in-network coverage with UHC in March.) The current guidance assumes Tandem does not get the UHC contract.

• What have been the payer conversations related to the MiniMed 670G Performance Guarantee? Said COO Brian Hansen, “I can’t answer whether they are resonating or not. We haven’t been pressured to match it. We have a differentiated product. Payers recognize the need for choice and having multiple pumps on the market; having one would be a really bad idea. We haven’t been asked for it and continue to work on our strengths and show data.”

3. Goal of 175,000 Installed Customers by 2023 (Nearly Tripling from 66,000+ now), 50,000 From Outside of the US

Tandem aims to triple its installed base from 66,000+ at the end of 2Q18 to >175,000 customers at the end of five years, and for more than 50,000 of these to be from outside of the US. Such growth would entail adding ~22,000 customers each year for five years, which seems easily doable – Tandem is on pace to do that this year, and the coming five years will have the AID products and international runway. As it stands, Tandem operates in ~70 territories, with 150-160 people in the field (reps + clinical); a sales force expansion is not expected next year.  Notably, only 40% of current users integrate their pumps with CGM, far lower than we would have guessed given Tandem’s tech-savvy population. (We wonder if this is an underestimate – is Tandem capturing phone-only CGM users?) IN any case, there is lots of runway for Dexcom here. 90% of Tandem’s current customers are type 1, with a wide age distribution and an equal mix of males and females. With respect to customer acquisition, ~50% of new customers are from MDI (on par with previous updates, and a great sign for an expanding market), and dQ&A data suggests that Tandem is getting “a very fair share of Animas transition” pumpers. As noted in previous updates and above, Tandem is still well shy of its 70% renewal target, as these have a very long cycle that goes well beyond the four-year warranty.

• Mr. Hansen underscored a bold goal for industry: bringing overall US type 1 pump penetration from an estimated 28% (550,000 pumpers) to 50% (900,000 pumpers). No time frame was attached to this. Tandem estimates that 1.7 million people live with type 1 in the US, on top of 1.6 million insulin-dependent type 2s. (The former number sounds a bit too high to us, particularly given a recent BMJ article and CDC data ballparking US type 1s closer to roughly ~1.2-1.3 million.) According to Tandem, Medtronic owns the lion’s share of the domestic pump market (65%), while Insulet holds 15%, Tandem has 12%, and Animas/others have 8%. See our 1H18 Diabetes Technology Industry Roundup for more granular sales details.

4. Profit/Gross Margin: goal of positive EBITDA 4Q18, cash-flow breakeven goal in 2H19; annual EBITDA profit in 2020

• CFO Leigh Vosseller shared upcoming profitability goals: (i) the first positive EBITDA quarter in 4Q18; (ii) cash-flow breakeven in 2H19 (goal of 55% gross margin); (iii) annual EBITDA profit in 2020; and (iv) 25%+ operating margin long term. While the cash flow breakeven goal has been shared for several quarters now, the rest were new. Tandem has had a remarkable year relative to this time last year, with over $100 million in cash on hand and no debt following an incredibly successful fundraising year – cash stood at a dangerously low $23 million one year ago (3Q17). Now, Tandem has quadruple that in the bank, has paid off $89 million in debt (3Q18), and the stock is up to a remarkable ~$48 (market cap: ~$2.7 billion)

• Profitability improvements will come from better manufacturing, increased volume, and better product reliability. The latter has already seen impressive gains with the t:slim X2. In 2016, warranty reserves represented a low teens as % of revenue; now they have dropped to the single digits. Beyond the 55% gross margin, management hopes for “additional reimbursement” for AID algorithms (theoretical at this point), more direct contracts with payers, and launching new products like t:sport. Margins will be offset by international sales, which will put some downward pressure on gross margins in the near-term.

• Expenses are expected to grow at ~10%-15% over the long term, slower than the expected double-digit growth in revenue. Management expects 10% long-term average annual growth in R&D expenses and 15% growth in SG&A expenses.

• Moving to the pharmacy channel is “not a primary goal,” given the t:slim X2’s DME nature and Tandem’s current business model. While Insulet and Dexcom have talked about increasingly moving in this direction (and Abbott already has with FreeStyle Libre), Tandem plans to remain in the DME channel for now.

5. Manufacturing Capacity Can Support Installed Base of 100,000, Capacity for the Next 12-18 Months

News was very positive on the manufacturing front, where management confidently asserted that the current infrastructure can support a base of 100,000 users, and there is sufficient capacity for the next 12-18 months. A slide parsed the numbers: Two pump lines can currently generate 110,000 pumps per year, and the four cartridge lines can contribute 12 million units annually. Tandem estimates that the production of 12 million cartridges per year is sufficient to support an installed base of 100,000 (10 cartridges per patient per month). Further, in light of the transition to a larger facility with double the space (only minimal rent increase), there is reportedly room for two more cartridge lines, which suggests a boost in annual production to ~18 million units, enough to support a base of 150,000 users. Presumably these additional lines could be turned on as needed, though Tandem’s current active installed base of 66,000+ as of 2Q18 together with the goal of reaching a base of 175,000 users in five years suggests the current setup will more than suffice for the near term. Even with the increased sales guidance and greater expected demand, management is confident that manufacturing will not be an issue.

Basal-IQ and Control-IQ Data and Clinical Commentary

1. Real World Basal-IQ Data (n=6): +5 Hours/Day In-Range; dQ&A Data Shows t:slim X2 Compares Favorably to 670G

Management presented a first look at real-world Basal-IQ data (n=6; +5 hours/day in-range) as well as dQ&A user preference data vs. MiniMed 670G, and described the simplicity of the Basal-IQ user experience (just one additional screen!).

• We were intrigued to see the first batch of t:connect real-world data from use of Basal-IQ in an unnamed payer’s population, featuring a whopping 21%-point increase in time 70-180 mg/dl (+5 hours/day) after Basal-IQ vs. before Basal-IQ (n=6). Management said that the data are a mere one month old and were used in a payer conversation just “last week”; they also said they can run similar reports on a regional level and can delve into CGM usage and other parameters. This should really help in payer conversations! We love the focus on near real-time data, and the fact that payers are apparently so open to it. Medtronic has produced dashboards enabling clinics to compare the performance of their 670G patients against that of the national average, updated in real time as more patients upload to CareLink – this seems like a logical possible step for Tandem to build out for clinics, payers, and internal use through t:connect. We note the obvious limitations of this data: Small (selected?) sample size, undisclosed baseline therapy, selection bias from early users, undisclosed time period lengths, etc. However, a five-hour per day bump in time-in-range is remarkable with predictive low suspend alone.

• Data from diabetes market research firm dQ&A showed that t:slim X2 (n=168) compared favorably to Medtronic’s 670G hybrid closed loop system (n=112) with respect to ease of use and treatment satisfaction. Surveyed patients were asked to rank the degree to which they agree with statements on a 1-10 Likert scale; the data presented by Tandem showed that significantly greater percentages of t:slim X2 users agreed strongly (defined as a 9 or 10) with statements, “my pump is easy to use,” “helps me have good BG control,” and “helps me feel more in control of my diabetes” (see specific numbers in the slide below). The data as presented skew strongly in favor of t:slim X2, a sign it should compete quite well against the 670G’s additional automation. We note caveats below.

• This above comparison of users is fascinating, and we’d love to dig further into the nuances. First, we wonder about baseline characteristics and differences in users that responded to the survey. Perhaps some of the t:slim X2 users included patients on Basal-IQ, who would presumably be excited early adopters and a potentially different group than the 670G users. There may also be a difference in underlying expectations – arguably users would expect more out of the 670G than t:slim X2; some patients believe hybrid closed loop will drastically improve outcomes and reduce burden. The first-gen 670G was designed for safety, and thus demands more user interaction and is built with a pretty conservative design on hyperglycemia mitigation side. (Is this a potential advantage for those coming second and third and fourth to market with AID? Can Medtronic recover and make 670G more usable – and quickly?). We’d also love to see the rest of the distribution of responses (how many said “strongly disagree” to the questions?), the demographic breakdown of the respondents (including how many t:slim X2 users were on Basal-IQ), and which questions were in the survey but not included in the investor presentation.

• Remarkably, upon remote software upgrade to Basal-IQ, t:slim X2 users just had to familiarize themselves with a single new screen on their pumps. This may sound standard in the consumer sphere, but such ease of upgrade is nearly unheard of in diabetes technology! Management added that patients did not have to go through a laborious upgrade process, thanks to the 45-minute online training module. As we’ve reported previously, the default setting for alerts is OFF (insulin suspension and resume),  but patients have the option to turn them off should they desire. It’s still early days, but management claimed that Basal-IQ is “plastered all over social media” – we’ve certainly seen our fair share of positive accounts online!

2. Patient and Clinician Panel Emphasizes Usability of Basal-IQ; Proposed Deterrents to Pump Therapy and Wish Lists for Ideal Pump

Attendees were treated to an offline panel moderated by Tandem’s Dr. Steph Habif and composed of Ms. Laurel Messer (CDE at the Barbara Davis Center), Dr. Gaja Andzel (an endocrinologist with type 1 diabetes), Mr. Peter Nerothin (a patient with type 1 diabetes), and Mr. Clayton McCook (father to a 10-year-old with type 1 diabetes). After a morning packed with exciting pipeline, regulatory, and financial updates, the candid discussion served as an important reminder that ultimately even the most cutting-edge technology is rendered useless if it is not deemed valuable by the actual patients. All four panelists emphasized usability as a critical factor when it comes to technology adoption. Ms. Messer was especially adamant on this point, asserting that “Tandem is currently ahead of the market with their touchscreen, their ability to upgrade remotely, and the fact that their algorithm runs in the background without requiring users to interface.” Dr. Andzel agreed, describing the ability for remote upgrading as particularly compelling for patients. Both Mr. Nerothin and Mr. McCook praised Basal-IQ for its simplicity. Mr. Nerothin was happy to avoid fingersticks (and the prerequisite diabetes supplies) given the G6 integration, while Mr. McCook was grateful that both his 10-year-old daughter and 70-year-old mother could operate the pump with ease. They were also equally impressed with the glycemic results achieved with Basal-IQ. Mr. Nerothin, a self-described “situational pumper” likes to use Basal-IQ for improved nocturnal time-in-range; Mr. McCook explained that his daughter Lilly was previously on a DIY system and just 10 days into using Basal-IQ is seeing “very similar results.” Wow!

See below for some of our favorite quotes from the panel, followed by select questions and answers with the audience. We were especially eager to hear the panelists’ responses when invited to share what would make it onto their wish lists for the ideal pump – Dr. Andzel’s response sounded a lot like Beta Bionics’ iLet Bionic Pancreas system…

• “I left insulin pump therapy about seven to eight years ago. I came back to the t:slim X2 recently because I don’t see it as traditional pump. To me it’s a laboratory where I can experiment with all sorts of opportunities to improve my life with diabetes. I’m most excited about using X2 overnight. I’m a situational pumper – other times I’ll still use basal injections.” – Mr. Peter Nerothin

• “We’ve been using OpenAPS for the last two years with Lilly very successfully. I felt good about our DIY system. Lilly’s numbers were fantastic, her time-in-range was brilliant, her hypoglycemia was very limited. 10 days in with the t:slim X2, we’re seeing very similar results with just Basal-IQ. It’s about data for me. I want to know at all times how much insulin she has on board, what’s the trend, where she’s headed. Then I want to know that I can go to sleep at night. I’ve been amazed to see her results with Basal-IQ. To see that straight line all across and the number of times her pump was suspended and to have no alarms. Most importantly, this growing 10 year-old doesn’t have to wake up at 3:00 AM and drink orange juice. Quality of life for all of us has so dramatically improved. That’s what it’s all about.” – Mr. Clayton McCook

• “Zero of the people with diabetes I met on a retreat that were on Basal-IQ had bad things to say about it. Their time-in-range was pretty much flawless. Being out of range just wreaks havoc on your body. If I can check that box, even eight hours of the day while I’m sleeping, that’s a situation I’ll be pumping. Then, I’ll say: “I’m going to go surfing and leave my pump on the dresser and use basal injection for the next eight hours.” Not a lot of people do that but it’s my approach right now. I’m a patient that will ease into it.” – Mr. Nerothin

• “I prescribe pumps to make life with diabetes better, to make life easier and to empower patients. When? When they are ready for it. Basal-IQ changed everything. Even before it, though, I valued pumps especially for kids taking injections – during school hours it’s a big hassle – using a pump makes life more normal. Regarding when to start pump therapy, every patient is different. I start the conversation from day one. They are sent to me from urgent care. I want them to know their options and to be open with me and discuss. I tell them pumps can make your life easier but they are not necessary. I always give them something to read. I always try to set realistic expectations because a lot of newbies think pumps will be magic. I always say it will be in the future but not yet.” – Dr. Gaja Andzel

• “Lilly’s DIY system worked very well, but it was also very intense. The setup, the maintenance – there’s no 24-hour tech support besides a Facebook group. If your pump fails, too bad. You have to carry a lot of devices. We really got tired of the burden of constantly troubleshooting and updating. Lilly was getting tired of carrying a bag with all the necessary devices. I had been watching Tandem very closely and was very impressed by their emphasis on customer service. Looking at us as people is huge. It’s a tough balance. The emphasis on the human element is what brought me to Tandem. And then there’s the technology itself. I’ve seen people posting screenshots of their pump data. Lilly has wanted the t:slim X2 for two years now. She thinks it’s cool; she likes the touchscreen.” – Mr. McCook

• “Now there is a situation where you get a pump and if something is new, you upgrade it. It’s one of the big reasons why this pump is quite popular right now [panelists nodding]. Patients are not getting the t:slim X2 now but for the future.” – Dr. Andzel

• “When you increase functionality of a device and minimize the time people with diabetes have to think about their glucose levels I think you’re going to accelerate market share. The usability piece will be key. I’ve had a few individuals ask, “what about algorithms, they’ve got to be important?” Yes, but patients still have to use it. If you can eliminate cost barriers and if you find a system that is usable, patients will use it. My message to industry is to make a device that is easy to use – then we can academically squabble over algorithms but it won’t matter if people won’t use it. Tandem is currently ahead of the market with their touchscreen, their ability to upgrade remotely, and the fact that their algorithm runs in the background without requiring users to interface.” – Ms. Laurel Messer

• “An aesthetically pleasing insulin pump does matter. Lilly was attracted to that. She went to the Tandem booth and said, “that’s cool I want that.” I think it’s a crap deal that she got dealt diabetes, but she ought to have a cool pump that she can take out and show to her friends. Usability was huge for our family as we did research into pumps. Can I drop Lilly off at my 70-year-old mom’s house and not worry? Yeah, I can. Can my 10 year-old do everything she needs to do? Yes. Usability was right at the top of factors for our decision.” – Mr. McCook

• “The expectation piece is very important. If patients have unrealistic ideas they’ll get on technology but they won’t sustain it. It’s not about the initial prescription, it’s about maintaining.” – Ms. Messer

•  “People who use technology should be able to detach from technology. Digital health as a field has not done a very good job to allow for situational use.” – Ms. Steph Habif

• “The more we learn about technology adoption the better we will be as providers.” – Ms. Messer

• “Working in pediatrics, it’s every parent’s fear that glucose control in the long term will cause their children to lose eye function or kidney function. Any ability to give someone a tool to increase time-in-range and reduce glycemic variability is a huge selling point. It’s always on the mind of someone with diabetes.” – Ms. Messer

• “After Lilly was diagnosed we didn’t sleep. We didn’t have CGM for the first two years – we were completely blind. When I think back to that it’s horrifying. That’s a really scary thought. The real change for me was in terms of wanting access to knowledge of how my daughter is doing. I have one job on this earth and that’s to protect family. How can I not do that if don’t use all the tools at our disposal?” – Mr. Clayton McCook

Questions and Answers

Q: What are some of the biggest deterrents to pump adoption? Does Basal-IQ help to address them?

Dr. Andzel: Pumps are quite complicated for providers. Turns out they’re too complicated for a lot of patients, especially teens. A little too complex. I hope we are going towards something simple so you don’t need a PhD in carb counting.

Mr. Nerothin: This is what I call “full pocket syndrome” [dumps out bag of diabetes supplies] This is the life of a person with diabetes a few years ago and now it is feasible to have everything right here [holds up t:slim X2]. With the G6 you don’t even need to test your blood glucose.

Mr. McCook: From a parent’s perspective, pumps aren’t cheap. We do have a lot of people we work with that are struggling to afford insulin, so cost is definitely a factor. Another factor is body image. It’s the reason why we never went with Omnipod – Lilly said absolutely not, it’s very huge, obvious. She’s getting more self-conscious. I know parents whose kids do MDI because they don’t want to be attached to a device. Part of that is because they don’t want their friends to see them attached to device.

Q: From a user standpoint, what else are you looking for? Things like voice control?

Dr. Andzel: I like this question. I want a pump where to set it, I just put the patient’s weight in. Second, when it’s time to bolus it’s just bolusing a meal (small, medium, large) because carb counting doesn’t work.

Mr. Nerothin: I want all of this stuff to go away. I don’t want to sit down and toggle around with everything. I just want to not have diabetes. If you want to talk about space age, I think it would be really cool to have a stable glucagon chamber. Also, improvements to infusion sites.

Mr. McCook: I don’t think my daughter cares about voice control. My daughter wants to be 10 years-old. She doesn’t want me to ask her about her blood sugar, to stop and check, etc.

3. Control-IQ (Initial) Study Impressions, Granular Algorithm Details; >$40m Invested in Tandem/Dexcom/TypeZero Core Tech!; Successful 6-12 Year-Old Control-IQ Ski Study Hopefully Published Soon

BDC’s Ms. Laurel Messer and TypeZero’s Dr. Patrick Keith-Hynes offered a deep dive on Tandem’s AID portfolio, including clinical pearls, granular algorithmic details, and an update on clinical studies. We were absolutely blown away when we learned that the systems’ core technology is based on 10+ years of R&D, >$40 million in investment from NIH, JDRF, UVA, and Helmsley, 30 clinical studies, 450 patents, and 300,000+ hours of clinical data. Wow have TypeZero’s algorithms seen some insane investment across the spectrum!

• Control-IQ features and commentary:

  • What are Control-IQ study patients saying (so far)? “They are sleeping through the night for the first time since diagnosis. They’re not having to interrupt their life to test glucose levels and treat low/high glucose levels. They’re feeling safer driving, working, etc. They have better glycemic control with exercise. I’m dreading the end of the study – people saying I want it back. Control-IQ: ‘good news.’” – Ms. Messer

  • “What makes this easier to use, is the individual is able to have more control over what the system is doing. This is working as a pump. If you’re not in hyperglycemia or hypoglycemia, it’s delivering basal rates of insulin, just like a traditional pump. You can change the dose. There is a lot of customizability. That’s hugely important… there’s a nice ability to make it work even better individual by individual.” – Ms. Messer

  • Control-IQ has three algorithms: (i) basal attenuation module (reduces basal in response to predicted hypoglycemia, modulates slowly); (ii) basal increase module (manages basal rate in response to hyperglycemia prediction); and (iii) hyperglycemia correction module (delivers periodic automated correction boluses as needed, once per hour). We’d assume the auto boluses are given for blood sugars over 180 mg/dl or perhaps 200 mg/dl, and presumably they have some conservatism built in (i.e., not a full correction dose).

    • “We are the only algorithm that I am aware of that does gradual attenuation of basal insulin as hypoglycemia approaches – it’s a fundamentally different approach, a risk-based approach.” – Dr. Keith Hynes contrasting Control-IQ with algorithms that suspend insulin to avoid lows, but don’t attenuate it

    • The hyperglycemia correction module, will blunt the size of excursions – e.g., if the user misses a meal bolus. Dr. Keith-Hynes emphasized that the system works better with pre-meal bolusing, but there’s the automatic correction bolus safety net if the user doesn’t for whatever reason. New to us, Control-IQ is limited to providing a correction bolus once per hour.

    • “Control-IQ runs on a five minute cycle, when new CGM values are available. Every five minutes, it gets the last six hours of meals, CGM, insulin, and hypoglycemia treatments, and predicts forward: “Here’s where you are going to be in 15 and 30 minutes. We want a soft floor, and to make sure we never give too much insulin.” – Dr. Keith Hynes

• Basal-IQ features and commentary:

  • Ms. Messer described Basal-IQ as having “high ease of use” and exclaimed that the lack of fingersticks is “game-changing.” She also raved about the option to turn OFF alarms for suspension/resuming insulin delivery and the ability to upgrade software remotely – clear advantages over the 670G.

  • “Alarm fatigue is really important to ease of use and technology acceptance. The G6 has the least alarms – there’s so little troubleshooting involved…With Basal-IQ and also Control-IQ, there are no more alarms than if you’re using the G6.”

• Ms. Messer shared that data from a ski camp study (kids ages 6-12 years) will hopefully be published “in the next few months.” The skiing study consisted of three days and two nights at Breckenridge, and Ms. Messer gave the most important topline results orally: “everyone had fun, no one broke any bones, and the system performed very well.” She called the X2 pump/Control-IQ onboarding process in children “revolutionary.” If the system’s glycemic outcomes come close to resembling those in the 13-18 year-old ski study we first reported on at ATTD (+6 hours per day in range), it would be a major victory, particularly in such a young population! The IDCL Control-IQ pivotal study is only enrolling individuals down to age 14, though we’d hope to see this extend lower given Ms. Messer’s glowing review of the system in pediatrics.

International Launch Highlights

1. t:slim X2 Launches in Italy, Scandinavia, Spain, the UK, South Africa, New Zealand, and Australia; ~150,000 Addressable Pumpers in Tandem’s Near-Term Global Market

CCO Mr. Brian Hansen provided exciting updates on Tandem’s international launch, announcing that t:slim X2 has launched in Italy, Scandinavia, Spain, the UK, South Africa, New Zealand, and Australia. Mr. Hansen characterized the past year as a “whirlwind,” noting that conversations around entering these markets were only just beginning as of EASD in September 2017. As noted above, ~$2 million is expected in Q3 OUS sales. Mr. Hansen noted that the first international pump was placed on a patient in Stockholm just three weeks ago! Ms. Vosseller guided for $7-$8 million in 2018 international sales, a pretty strong uptick. We’re very impressed by how quickly Tandem has been able to mobilize its OUS efforts. According to Mr. Hansen, following the Animas shutdown news last October, Tandem was able to prepare for an international launch “in record time.” This process included a variety of critical developments, including translating the pump’s software into seven languages, adding mmol/l, and obtaining a CE Mark.

• Mr. Hansen shared that France, Germany and the Netherlands are the next international targets for Tandem. He explained that Tandem is preparing to begin discussions regarding entrance into these markets as the team heads to EASD next week. Apparently, enthusiasm for the t:slim X2 is quite high – Mr. Hansen noted that “we’re getting phone calls from you name it.” However, he views certain markets as less-viable options for Tandem than others, as they may not be able to be supported with the current reimbursement structure or lack a sufficiently active type 1 diabetes community.

• Tandem estimates internally that there are ~42.5 million people living outside the US with type 1 diabetes – this is way, way too high of an estimate in our view, as it implies 10% of the global diabetes population (both diagnosed/undiagnosed) has type 1. Based on recent CDC data and BMJ data, we’d guess the global type 1 number is more like 5% of the global diagnosed populations, or ~23 million people, with ~1.3 million of those in the US.

• Tandem estimates 150,000 pumpers are in its near-term target markets, with 3 million people in its long-term markets (3-5 years). Slides were not specific about the latter, but Mr. Hansen mentioned Japan, China, and India as phase three markets in the outer bubble – obviously these are many years away, and the initial focus will be on Europe and Canada and Australia/New Zealand. Being deliberate here will be key; Tandem could easily stretch itself too thin.

• Mr. Hansen noted that Canada is a “big piece” for Tandem’s near-term strategy, anticipating Tandem to have “very good penetration” and “very good reimbursement” in Canada. He noted that t:slim X2 is still within the Health Canada regulatory process, but that he is looking forward to approval “shortly.” The press release announcing sales estimated a Q4 launch in Canada, pending approval.

• Mr. Hansen acknowledged that the international market is “a little different” than the US pump landscape. In addition to Insulet and Medtronic, which both hold strong market share in the US and internationally, Mr. Hansen listed Ypsomed, Cellnovo, and SOOIL as competitors unique to the international space. (We’d add Roche, who was not mentioned.) SOOIL’s Dana pump does have the significant advantage of being controlled on a smartphone, which has opened it up for early adopters users in the DIY community (as a reminder, SOOIL plans to submit an open-protocol pump to the FDA, hoping to commercialize a device tailored for the DIY community by ADA 2019). Ypsomed is also jumping on the open-protocol bandwagon, announcing two-year JDRF funding in August to develop an open-protocol, fully interoperable mylife YpsoPump. Mr. Hansen remained confident, asserting that given its initial reception overseas, Tandem’s “premium product will have a place.” There is clearly more than enough runway globally for many companies to compete.   

2. OUS Reimbursement ($7,000-$8,000/Patient) is Roughly Equivalent to US ($8,000/Patient); Future OUS Targets Include France, Germany, and the Netherlands; “Unique Profitability” in “Filling the Animas Void”

Ms. Vosseller expects international pricing to be roughly equivalent to that in the US over a four-year reimbursement cycle. She explained that the total amount realized per patient in the US at the projected near-term distributor mix (80%) is ~$8,000 ($4,000 for pump, $4,000 for supplies), whereas she expects $7,000-$8,000 realized per international patient at a projected near-term distributor mix of 85-90%, given a much lower $2,000-$3,000 in pump costs and a higher $5,000 in supplies. These estimates assume one pump reimbursed every four years, and 120 cartridges and infusion sets used per patient annually.

• Tandem’s involvement to support distributors differs in the US, Canada, and in other OUS geographies. For example, about 80% of Tandem’s US business is through distributors. Per Mr. Hansen’s comments, Tandem plans to use a logistic distributor in Canada but will mainly rely on its US sales force, hiring only ~10 sales reps in Canada. Conversely, in other international geographies, Mr. Hansen explained that Tandem will “make the pump and ship it to [distributors.] They will do everything else.” Mr. Hansen characterized this model involving only a “skeleton crew” as presenting “unique profitability” potential. He described “phase one” of the international launch as focused on “filling the Animas void,” given the opportunity to plug into Animas distributors, as well as existing customer support groups, tech support, and payer contracts.

3. International Launch of Basal-IQ with Dexcom G6 Integration Expected in Early 2019; Control-IQ to Follow “with Similar Cadence” (~2020)

The t:slim X2 is currently launching OUS with Dexcom G5 integration, and in the first OUS timing update, an early 2019 OUS launch is expected for G6 integration with Basal-IQ. Mr. Hansen shared that Tandem is coordinating with Dexcom as it launches the G6 internationally, which obviously has to happen before Basal-IQ can launch. Meanwhile, Tandem is working on translating the Tandem Device Updater for foreign countries. Presumably Basal-IQ will also need a CE Mark. He explained that Control-IQ will follow “with similar cadence,” implying it will be about 6-9 months behind the US launch (~2020). Mr. Hansen emphasized that getting to this stage alone has required a “herculean effort” in a very short amount of time. We’re very impressed by the pace at which Tandem has managed to bring its product to market internationally and salute the company for aiming to provide its latest technology to a global patient population.

--by Adam Brown, Maeve Serino, Brian Levine, and Kelly Close