Insulet 1Q13 – Sales up 20% year-over-year; second-gen pod shipments strong; installed base conversion delayed; U-500 OmniPod in development – May 6, 2013

Executive Highlights

  • Sales in 1Q13 totaled $57 million, a 20% increase year-over-year (YOY) and 1% decline sequentially.
  • Since the launch of the next-gen pod, new patient referrals and initial shipments increased 40% compared to the same time period in 2012.
  • Shipments to new pod users have exceeded management’s expectations. Due to a manufacturing issue (now resolved), the installed base conversion will begin in the next few weeks (~90 day delay).
  • Insulet is developing a special version of the OmniPod for use with Lilly’s Humulin R U-500 insulin.

This afternoon, Insulet CEO Duane DeSisto led the company’s 1Q13 financial update. Revenues in 1Q13 totaled $57 million, up 20% from 1Q12 and down 1% from 4Q12. Sequentially, this was quite comparable to 1% in 1Q12 and 2% in 1Q11. Core OmniPod growth was 25% year-over-year (YOY). Since launch of the next-gen pod in February, the company has seen impressive gains in patient referrals and initial shipments (40% growth since launch vs. the same time period last year and 50% growth in the month of April compared to April 2012). However, Insulet continues to see reorder disruption from existing patients delaying or decreasing orders in anticipation of switching to the new pod (expected to continue in 2Q13). Notably, Insulet was operating cash flow profitable for its second consecutive quarter. The company ended 1Q13 with $148 million, of which ~$93 million was attributed to the company’s sale of almost five million shares. Excluding the public offering, Insulet had a cash burn of ~$2 million in the quarter. Looking forward, management guided for revenue of $59-62 million in 2Q13 (~16-22% YOY growth) and reiterated its forecast for full-year 2013 sales of $240-255 million (~14-21% YOY growth).

The new OmniPod started shipping to new patients in late February, and management emphasized that demand has exceeded expectations. CFO Brian Roberts was excited to report that over 10% of new shipments since launch have come from diabetes practices that had no shipments in 2012 – the big difference has been the smaller size and updated insulin-on-board calculation (especially for pediatric endocrinologists). Unfortunately, an “eyelash”-sized manufacturing problem (one component off specification by two ten thousandths) led to lower than expected second-gen OmniPod production in 1Q13. The result is that installed base conversion will now start in the next few weeks, representing a delay of ~90 days. On the bright side, Insulet quickly identified and remedied the manufacturing problem – great to hear that no pods left the building. The goal is still to convert the entire installed base by the end of 3Q13 (September), consistent with remarks in previous calls.

Much of the call focused on the new development agreement with Lilly to build a new OmniPod for use with U-500 insulin. Management estimates a target market of approximately two million type 2s, representing a potential doubling of the company’s current type 1 target market. No timelines were shared, though management gave nuance on the regulatory side (a 510(k) application from Insulet plus an updated insulin label from Lilly) and shared its plans to conduct a clinical study. Regarding the LifeScan-integrated PDM, management hopes it will be commercially available by year-end 2013; the company cannot file with the FDA until it receives some information from LifeScan. No new details were shared on Insulet’s new plans to integrate a CGM into the OmniPod patch – the goal is still to enter human trials in 2014.


  • Total revenue reached $57 million in 1Q13, up 20% from $48 million in 1Q12. First quarter sales fell in the middle of the company’s 1Q13 guidance for $56-59 million. The company noted higher initial shipments, offset by disruption in reorder patterns by existing OmniPod customers. Management explained that existing customers are either delaying shipments or reducing order quantity in anticipation of converting to the new pod. Insulet expects these delays to persist into 2Q13, and then normalize in 2H13 once the patient base has been transitioned to the new pod.

    • Core OmniPod growth was 25% from 1Q12. As such, we estimate that Insulet’s core OmniPod business brought in ~$42-44 million. Management remarked that Neighborhood revenue was similar to 4Q12, when sales were just under $13 million. We expect ~$1.3 million attributable to Abbott (as a reminder, Abbott provides Insulet with approximately $500 for each OmniPod PDM).














Total Revenue (millions)













Year-Over- Year Growth













Sequential Growth













  • Sequentially, sales were down 0.8% from $58 million in 4Q12. For historical context, sequential growth was 1% in 1Q12 and 2% in 1Q11. Given the delay in patient conversion and the seasonality of sales typical to medical device companies (i.e., stronger 4Qs and weaker 1Qs), the sequential decline was not unexpected.

  • Insulet was operating cash flow profitable for the second quarter in a row. As a reminder, September was the first month in the company’s history in which it achieved operating cash flow profitability. The company ended the quarter with $148.1 million, up from $57.3 million at the end of 4Q12. Excluding the company’s sale of 4.7 million shares of stock (resulting in~$92.8 million net proceeds), Insulet had a cash burn in the quarter of ~$2 million. In 4Q12,Insulet was striving to be operating profitable (earnings before interest and taxes [EBIT]) in 4Q13; management did not comment on this guidance during the call.

  • Since the launch of the new pod, new referrals and initial shipments have increased by over 40% compared to the same time period in 2012. Impressively, in the month of April, new referrals and initial shipments were up by over 50%. Said management, “I think it’s fair to say that it’s the largest first month of the quarter that we have had as a business.” When queried as to whether the 50% increase was attributable to “pent up demand,” management remarked that it believes the growth is reflective of an expanding prescribing base.

  • For 2Q13, Insulet expects total revenues of $59-62 million (~16-22% YOY growth). Management underscored that this guidance assumes that Insulet will no longer have a Medicare business for Neighborhood Diabetes after July 1 (when the competitive bidding program is slated to go into effect). However, management noted that there are lobbying efforts ongoing to delay the program’s implementation until December.

  • For full-year 2013, management reiterated its guidance for revenue to reach $240- 255 million, reflecting 14-21% YOY growth. Insulet expects international revenues to grow two- to three-fold. The full-year guidance, in light of the delayed conversion, inspires confidence in the company’s ability to attract both new prescribers and new patients. Insulet’s ability to penetrate the insulin-pump naïve market is certainly great – in the previous quarter, over 70% of its new customers have never used an insulin pump. Given the quarter’s major news (strong initial shipments and the delayed installed base conversion) we expect to see Insulet modify the full-year guidance up at some point, though today may have been a bit soon.

  • Insulet’s 1Q13 gross profit was $25.2 million (44% gross margin), up 24% from$20.3 million in 1Q12 (42.5% gross margin). On a sequential basis, gross profit was nearly flat (-0.6%) from $25.3 million (44% gross margin) in 4Q12. This was in line with expectations set forth in 4Q12 when management expected relatively flat grows margins in 1H13 compared to 4Q12. Management reaffirmed this expectation for 2Q13. At the time of the 4Q12 call, management also forecasted a 2-3% uptick in 3Q13 (and a 3%-plus uptick in 4Q13). However, given the delay in patient conversion, Insulet anticipates that it’s 3Q12 expectation may be ~1% lower depending on how quickly the company can transition its customer base.

    • For the OmniPod business specifically, management continues to expect 60%-plus margins in the US by year-end 2013. Longer term, the company maintains that it can reach 65%-plus gross margins, assuming an average selling price of$28.

  • Total operating expenses in 1Q13 reached $31.4 million, reflecting nearly flat (0.5%) growth from $31.2 million in 1Q12 and flat sequential growth.

    • Insulet spent $4.4 million on research and development in 1Q13, down 19% from $5.4 million in 1Q12. Sequentially, R&D fell 25% from $5.8 million in 4Q12.

    • First-quarter general and administrative costs of $13.1 million were nearly flat (0.6%) from $31.0 million in 1Q12. G&A costs were up 2% on a sequential basis from 4Q12 expenses of $12.8 million.

    • Sales and marketing costs reached $13.9 million in 1Q13, an increase of 9% from $12.7 million in 1Q12. Sequentially, sales and marketing expenses were also up 9% from 4Q12 expenditures of $12.7 million.

  • Operating loss totaled $6.2 million, a drop of 43% from $10.9 million in 1Q12 and up 2% from $6.1 million in 4Q12.


  • The second-gen OmniPod began shipping to new patients in late February and the launch has exceeded management’s expectations. CEO Duane DeSisto led off the call with positive anecdotal stories from the field, stating that management has been “thrilled” by the number of HCPs that have opened their doors to reps – e.g., one New England practice only averaged one new patient addition per month in 2012, compared to 20 new patients in the last 60 days. Management specifically cited the smaller pod size and modified insulin-on-board calculator, improvements that have gained traction among pediatric endocrinologists. CFO Brian Roberts was most enthusiastic that over 10% of new shipments since launch have come from practices that had no shipments in 2012.

    • Second-gen OmniPod shipments to existing Insulet customers will now start in the next few weeks, representing a delay of ~90 days. As a reminder, the 4Q12 call in late February called for the transition to start “in the coming weeks.” We were glad to hear that the target is still to convert the entire installed base over to the new pod by the end of 3Q13 (September). The delay will give Insulet time to build additional inventory to support the conversion. Said management in Q&A, “We feel good at the 200,000 per month with the first line. It’s like changing a tire on a car – want to do it while the car is stopped. Not while the car is going down the road at 30 mph.”Production of the new OmniPod was lower than expected in late 1Q13 due to an unexpected component issue. The problem was discovered in QA and no pods left the building – it was great to hear that “it was pretty obvious” something was wrong. The component was two ten-thousandths off spec (“an eyelash”). This occurred as Insulet began automating a couple stations on line two as production scaled up (line one is currently producing 200,000 pods per month, with an additional 400,000 expected on line two).

    • Insulet quickly identified and remedied the manufacturing problem. Management emphasized that these component issues come up “in the normal course of business as you ramp up production.” Management has worked extensively with suppliers to ensure the line’s consistency.

  • Management expressed strong confidence in manufacturing at this point. Second quarter reorder cycle lumpiness was characterized as a bigger worry than manufacturing (i.e., patients holding off on a supply reorder to perfectly time conversion to the new pod). Indeed, management is “confident enough” that they’re shutting down the old OmniPod manufacturing line in the next month.

  • The goal is still to have a third manufacturing line up and running by this summer (consistent with remarks in the 4Q12 call). The two current manufacturing lines can produce 600,000 second-generation OmniPods per month (20% excess over current demand), and the third line will give Insulet capacity to produce almost one million pods per month.

  • On the international front, sales in 1Q13 “tripled” from 1Q12 (when volume was “lighter”). Management said there was also an uptick compared to 4Q12 – for 2012, international sales were ~$5 million, roughly ~$1.25 million per quarter. Managementspecifically cited the Netherlands and Switzerland, where Ypsomed “continues to take significant share.”

    • Management believes international revenues in 2013 should double or triple compared to 2012. This was on par with comments in 4Q12, where the international business was expected to “more than double in 2013.” Management said Insulet will launch the OmniPod in “several additional countries” in 2013 (no further specifics). As a reminder, Insulet announced an amended and expanded agreement with Ypsomed in 1Q12. The amended agreement made Ypsomed the exclusive distributor of the OmniPod in a total of 22 countries (double the original eleven countries), including markets in Europe, the Middle East, and Asia.


  • Last week, Insulet announced plans with Lilly to develop a new version of the OmniPod for use with Lilly's Humulin R U-500 insulin. This would involve a label change on both the drug and device side (see below). Insulet has held an initial meeting with the Lilly team and is planning the clinical development program that will allow both companies to appropriately update the labeling. Management expects to provide an update as timelines become clearer.

    • Right now, the Lilly/Insulet collaboration is only a development agreement. Insulet will be selling the PDM and pods, while Lilly separately sell the drug. During the development phase, there will be clinical development costs (see below) that both companies will share. Management characterized these as “not material” and “little expense over the next couple of years.”

  • The new OmniPod for use with U-500 insulin will be a two-part regulatory filing, requiring a drug label update from Lilly and a 510(k) application from Insulet. U-500 is currently being used in the marketplace, though it is not approved for us in pumps.

  • Insulet plans to create a new PDM as part of the development program with Lilly; the pump will use the same second-gen pods and manufacturing lines. The type 2- targeted OmniPod will be a ~three-day, basal-bolus device. Said management, “It looks like a type 1 diabetes device but without the bells and whistles.” The PDM will need to be updated to accurately reflect the dosing and other criteria for type 2 diabetes patients. The other key question is whether type 2s need all the confirmation screens that the current PDM has. We will be interested to hear about FDA’s thinking on this front.

  • Management is “most excited about conducting a study to demonstrate positive clinical outcomes with the U-500 OmniPod. The expectation is that strong clinical data will help with reimbursement and “easily justify the current price in the marketplace.” Certainly, the higher price of the OmniPod might cause some payers to shudder at the thought of type 2 reimbursement, though we agree it will be justified if the outcomes are as strong as some groups have reported in trials:

    • Management noted that data in the literature has been quite positive in off- label studies of U-500 in insulin pumps. A quick literature search confirms this. In our search on PubMed, we found very few studies of U-500 in insulin pumps. Here are three particularly interesting abstracts we found: 

      • A compelling 2003 study (n=4) appearing in Endocrine Practice (Knee et al.) demonstrated an A1c decline from 10.8% to 7.6% in three months, with a further decline to 7.3% by six months. Meanwhile, the absolute volume of insulin infused per day decreased by at least four fold (savings of up to $2,600 per year for insulin and $3,400 for pump supplies!).

      • Another small study (n=6) appearing in Endocrine Practice in 2007 (Bulchandani et al.) demonstrated similar results – A1c declined from 9.1% to 6.9% after just six months of therapy.

      • A 2013 retrospective chart study appearing in Endocrine Practice (Leinung et al.) examined seven patients using U-500 insulin in a pump – A1c dropped by 1.1% (p<0.001) along with a 0.071 units/kg drop in basal insulin requirements (p <0.001).

  • Management estimates that the target market for a U-500 OmniPod is approximately two million patients with highly insulin resistant diabetes (>100 units per day). This was characterized as “up to 10% of people with type 2 diabetes.” CEO Duane DeSisto emphasized that this will “potentially more than double” the company’s current target market (~1.5 million type 1 patients in the US).

  • We see this move as a very smart way to expand the company’s target market without a significant capital investment. It’s encouraging to hear that only the PDM will need to be updated and the company will use the current second-generation OmniPod. That should offer a quicker regulatory path on Insulet’s side. Still, we expect it will take some time to develop the product and get the clinical study off the ground. Given management’s remarks about development costs over “the next couple years,” we don’t expect to see this on the market before 2015/2016..


  • Insulet hopes that the Verio-integrated PDM will be available by the end of the year. Management is waiting for information from LifeScan before it can file the new handheld with the FDA. The timing is still consistent with the 4Q12 call, where management called for a filing “in the coming months” and availability by year-end 2013. We would not be surprised if the timing slipped a bit, as LifeScan has an ongoing global recall of several Verio products. For more information, see our J&J 4Q12 report at

  • “Feasibility work continued” in 1Q13 with the unnamed CGM partner to integrate a sensor into the OmniPod patch. The goal is still to enter human trials in early 2014. “The single biggest issue” Insulet is dealing with is sterilization – as a reminder, this was cited in the 4Q12 call as a reason for not working with an established player like Dexcom. Management would not share accuracy details, only revealing that a combination of animal and human data suggests that accuracy is “comparable” to the products that are on the marketplace. CEO Duane DeSisto emphasized that this project is “on the drawing board.” We look forward to seeing data on the sensor next year. It’s a bold move not to work with an established player, in addition to relying on a three-day sensor. Still, we think the integration would be a boon for many patients, especially young type 1s with limited on-body space (or appetite) for devices.

  • Insulet management reiterated comments from JP Morgan 2013 concerning the artificial pancreas. CEO Duane DeSisto was quite negative, noting that “no one on this call” will live long enough to see a closed-loop system. With its integrated CGM sensor, Insulet plans to pursue what could be called a smart open-loop strategy. The company has an option to license software algorithms, which would ideally talk to the CGM and “keep the patient out of trouble.” In other words, lights on the handheld would appear green if thing are fine, while yellow would suggest a predicted problem in three to five hours, and red would notify the patient that he or she is “bumping against the guard rails. This is similar in user interface design to the University of Virginia’s Diabetes Assistant system, though that device also has closed-loop capabilities.

    • Management is very negative on the artificial pancreas. We think there is some misunderstanding on the terminology front; many believe that hybrid closed loop systems could be available within, for example, three to five years. Others believe strongly that a truly closed loop system will never be available. We also believe the “hybrid” closed loop will be how the closed loop still emerges – and although management is negative on the “closed loop,” we believe they are thinking of the closed loop where patients did absolutely nothing. We believe at this stage that it’s not a matter of “if” but “when” and “how” – what will the first product look like, when will it hit the market, and will reimbursement look like. We believe that over time, as the concept of the closed loop emerges and more people discuss hybrid models, that Insulet management may themselves be more balanced in their comments about the AP.

  • Insulet is looking at other drug opportunities besides insulin – management said the company has partners in fertility, oncology, and notably, obesity. Insulet “cannot identify” its oncology partner at this time, but the company has made “tremendous” clinical progress and anticipates a commercially available product in 2014. We were very interested to hear obesity and wonder what injectable drug this would be – we might speculate a GLP-1 agonist or pramlintide but note that it could be a range of peptides.


Q: Can you talk about the gross margin expansion as you transition the existing patient base? What’s the outlook for the cadence – is there a change at all given PDM upgrade? And for 2014 and beyond, can you talk directionally about gross margins getting beyond the 60% range?

A: In the short term, depending on the rate that we’re able to convert folks starting here in a few weeks, we will factor a little bit into where the Q3 gross margin comes out. You will recall from the last earnings call that we said Q1 and Q2 will remain flat with Q4. Q1 followed that. And then we would see 200-300 basis points of improvement in both Q3 and Q4. We may lose 100 basis points or so of that in Q3 depending on exactly how many people are converted in May and June vs. July, August, and September. That piece of it will have a little bit more color as we get to our next earnings call and figure exactly what percentage of the base has been moved.

Longer term, I can tell you certainly that Charlie and the team have been doing a fantastic job working with Flextronics and all of our suppliers to continue to move the manufacturing process in the short term and find additional efficiencies to drive the price down. As we look out longer term, using a US OmniPod $28 average selling price, we certainly still believe 65%-plus margins are very doable for the business – once we have fully transitioned everybody and eliminated some of the inefficiencies that would remain with the old line that need to be eliminated. We are not ready to go beyond that at this point, but I can tell you that the team is certainly working on a bunch of different projects that we hope will get it even further.

Q: On neighborhood diabetes, do you have thoughts strategically on that business, especially with competitive bidding into effect on July 1?

A: For competitive bidding, we’re following it closely. There is a significant amount of activity down in Washington. The guidance we gave you assumes we are not going forward with the Medicare piece of the business. If you read all the stuff going on in DC, there is a whole movement afoot to try to push that to December. How that all turns out, I guess only time will tell here. We've dialed in assuming it's all gone, come July 1. With regards to Neighborhood in general, all the reasons we want it all make sense. The Medicare business came along with it, but it’s a small piece of the business. We are cognizant that we have to continue to wring efficiencies out of that. We're going to make each individual sales worth more money, and we continue to work with that. Obviously, with the price that came out in competitive bidding – for us as a side player – we were kind of surprised how low it was going. But when you look at all the guys that had billions of dollars at stake here, they were surprised also. So stay tuned. The guidance we gave you assumes that business is going away.

Q: On the component supply issue, how confident are you that the issue is fixed? What impact did the component disruption have on the construction of the second quarter revenue guidance?

A: We're confident enough that we're shutting down the old line. That will be done here probably at the end of the month. So that's all going away. We feel good about it. Having been through this now a couple of times, it's kind of been in the normal course of business as you ramp this stuff up.

From the Q2 guidance standpoint, we're more worried about disruption. As we stated, some people now only take only want one box – say their reorder comes up in April, they’re hoping that we can convert them by the end of May. It is reflected in the numbers that we gave you that there will be some disruption. It’s not just they only ordered one box and then you can ship them two when you're ready. You’ve got to go through the third-party reimbursement and have to understand what their insurance plans are willing to pay for.

It’s reflected a little bit in the number. But we're not surprised. We've built in enough timeframe. I think what we told everybody was by the end of third quarter we'd have the installed base converted. We still think by the end of the third quarter, we'll have the installed base converted. It's taken us probably a little bit longer than we had hoped to get going. You want to start a machine up and you want to make 400,000 of these a month. We feel real good at the 200,000 level that's on line one, and this second machine we're going to squeeze it out. I guess the best analogy I can give you is if you're going to change a tire on the car to get it run better – you want to do it while the car is stopped. You probably don't want to be doing it while it's going down the road at 30 miles an hour. So we had the window, we took advantage of it to squeeze this efficiency out of it, and we feel pretty good about it.

Q: Great. That sounds like a little hiccup and a manageable one. But it’s fair to say that initial demand with the 40% number might be outpacing your expectations? On the annual guidance, you did not change it despite more conservatism on the second quarter. What gives you confidence to maintain the high end of the guidance?

A: Overall, we’re extremely encouraged. We had a very, very slow January and first half of February because everybody on the new shipment side was anticipating moving to the new pod. Since the launch, I think it's fair to say that it's exceeded our initial expectations. April was up 50% referrals and shipments year-over-year. I think when we look at the volume that we processed in April, I think it's fair to say that it's the largest first month of a quarter that we've had as a business.

A lot of these component issues have happened throughout the course of the last few years. It's just that as we have inventory on hand and as we got the production ramped and run, a lot of it happens behind the scenes and seamlessly. In this case, given that it was right at the start of this conversion, we just wanted to make sure and be doubly cautious that we were taking the right approach. We have to make sure we have enough inventory and that we’re be able to, once we convert people over, we never want to go back.

I think the field is very jazzed up about what we're seeing so far. Duane gave you a couple of different stories. We're getting tons of these different examples back about practices letting us come in. Again, if you think about a launch that's effectively only 60 days old, to have over 10% of our shipments coming from doctors who did not prescribe the product in 2012 is pretty telling. It means doctors are taking these visits, they're looking at the product, they're excited about it, and they're putting a few people on. We think that trend is going to continue. So that gives us a lot of confidence as we look at that revenue range throughout 2013.

Q: I just want to make completely sure – you are fully resolved on the component issue and you are building inventory. Do you have inventory levels today that will support that transition, or do you think you will have that by the end of the month?

A: We will have inventory levels here in the next few weeks where we will start converting people in the quarter. But as of today, we're still doing it [building inventory].

Q: And the component issue is fully resolved? So it's just a capacity issue?

A: We are talking about a component that was two-ten thousandths off the spec, which just gives you some indication of how precise we are in this design.

A few hundred thousand of these components are in manufacturing, so the long-winded answer to your question is yes.

Q: On Neighborhood Diabetes, is $13.5 million correct?

A: Not surprisingly Q1 always has seasonality in these types of businesses. The level of revenue in Q1 for Neighborhood was similar to Q4.

Q: For Ypsomed, we saw the announcement where they talked about their mylife pump and then they came out and talked about bringing out their own tubed pump next year. Is that just to supply the market with every type of pump there is? I am trying to figure out why they would come out with another pump.

A: Since we signed the deal with Ypsomed, they've talked about coming up with a very cheap traditional insulin pump because they believe there is a subset of this market that cannot afford our price point, a typical insulin pump price point. Not for me to speak for them, but to give some clarity on it. It's kind of a down and dirty pump for the person that probably could do a little bit better than shots but probably couldn't afford the higher care that comes with our products. They are trying to come up with a down and dirty way to do that and that's what this will be.

Q: As you look at this transition, does this impact your attrition rate or is it too early to tell on the new OmniPod what those attrition rates look like?

A: If we look at what's happening in Europe, attrition rates in Europe have been lower than what they have been here in the US. That's probably the best indication. It's too early to say anything in the US except for our overall attrition in the first quarter was down slightly from where we were in Q4. Although everybody wants to get their hands on the new pod and we certainly understand that, we are not seeing people leave.

Q: Where did you find the manufacturing and the component issue, in the field or internally in quality assurance (QA)?

A: We found it in QA. Line one was running pretty well. In line two, we’ve automated a couple of the stations from line one to help continue to drive the cost out and in that automation process is where it popped out. The product never left the building. It was pretty obvious, pretty quickly that something was different. We immediately went to a couple of these stations that we automated as the ones being different from the original line that was running. And we noticed – like I said – we noticed basically an eyelash of tolerance difference in the particular product.

Comment: So internally, that's the key.

A: Yes, it never saw the light of day.

Q: The 50% growth in shipments and referrals in April – can you tease out how much of that is pent-up demand versus something more durable?

A: It’s hard for me to say. What I keep going back to is the fact that over 10% of the shipments over the last 60 days have come from brand new doctors, defined as doctors who haven't prescribed the product in at least the last 15 or 16 months. So when we look at that metric and you're seeing over 10%, that to me is not pent-up demand. That is the expansion of the prescribing base, which I'd say is what we're most excited about.

Q: On CGM, you talked about the path to human trials, or at least human trials in 2014. Can you give us any update on what sort of data you have in humans so far and where that path takes you in 2014?

A: I think it's too early to go there. I would tell you obviously the single biggest issue for us with integrating – and if you talk to any of the guys that are in the CGM business – the single biggest thing that we are wrestling to the ground as we speak is we've done the testing on CGM, we kind of like what we see, but the next big step before we invest any kind of money in this whole process here is really to wrestle down the whole sterilization cycle for the fluid path as well as for the sensor. That's kind of where we're at at the moment.

We'll keep you updated, but I think it's too early to tell in terms of accuracy versus other products that are out there or any of that now. I think the real test for us is we really want to clear the sterilization cycle issue. And then once we get there, then I think we have a couple of really good ideas on the drawing board, but I would re- emphasize they're ideas on the drawing board. We really want to get that thing pinned down before the end of the year.

Q: Can you say anything more about what human data there is for just the sensor itself though? What either is published or what you guys saw as part of your diligence?

A: The sensor data that we had were based off of the animal studies we've done and then some of the sensor data that they had done, a combination of human and animal. Again, what we came away from was that we thought that the accuracy of the sensor was comparable to what we're seeing out in the marketplace today. That's probably as specific as we can be at this point.

Q: In my opinion, Lilly putting this project in your hands is a nice endorsement. I don't know much about Humulin U-500. Is it a different consistency? Are there different properties? Is there a major technical change that has to be made to the pump?


A: The typical type-1 diabetes patient uses U-100. U-500, as the name depicts, is five times the strength. The typical U-100 is fast acting. U-500 is not in that same formulation. U-500 is being used in the marketplace today and what we are going for, from a drug side is obviously a labeling change, so it could be used in the pump. From the pump side, we're trying to build a device that accurately reflects dosing and the other criteria that are needed for a patient with type 2 diabetes. And then the other question, as part of this study is, does the type 2 patient need all the things that a type 1 patient needs in terms of the various screen shots?

That's what we're going through. What we're talking about doing is, in combination with the drug, doing a clinical study with some hopefully very positive clinical outcomes. When you get this all approved, not only are you going to the market with a product and a drug specifically designed for one another, but also clinical outcomes that should reflect, hopefully, an improvement in the patient's life.

Q: Is the mix changing significantly between brand-new-to-pump customers as opposed to people switching pumps?

A: I haven't seen the change in the mix. What we are seeing, though, which is in my opinion and my opinion only at the moment, is that there seems to be more pediatric-type centers that are really, really starting to pay attention to this because of the size. Prior to this new product, we had the smallest pod, period, in the marketplace. This size reduction for a lot of these pediatric centers is meaningful. And the insulin on-board calculation, the methodology for that, they are much, much more comfortable with. It's still too early to tell, obviously, but we are seeing an uptick across a bunch of these pediatric endocrinology centers.

Q: We saw on the blogs that some users were upgraded to new pods in April. Were those patients using specific distributors versus direct ship? It seems like some of your current users were hearing about those single folks getting the upgrade and were getting pretty nervous and/or upset.

A: To be clear we have not commenced the conversion yet of the existing customer base to the new OmniPod. If someone wound up receiving new pods, it was probably somebody who was out of warranty who maybe went through a distributor partner and was able to obtain it looking more like a new shipment. Those are onesies and twosies, but not something that we've really seen at this point.

Q: At ADA, are you having any special sessions for the OmniPod?

A: I don't think we are planning any special sessions. Certainly in the booth we will have a lot of information about the new OmniPod. We will continue to do what we think has worked well for us, which is to be able to gather key opinion leaders in different settings and talk through it with them as compared to something that's formal.

I think that will be our approach with all of the conferences this summer. We had folks at the AACE conference last week. I can tell you that we were thrilled by the amount of people that attended, like I said in an informal session, where it was literally standing-room-only. We presented a couple of slides and we had people standing behind me, basically standing behind the screen because we couldn't fit all of them in the room.

Q: Regarding your type 2 pump with Lilly, do you think it is generally a three-day wear device, one-day wear basal/bolus combine, and will it have a controller?

A: To make it clear, it's our device using Eli Lilly's U-500. We're the device, they're the drug. And it will have a handheld. We believe for these highly insulin-resistant type-2 patients with U-500, one of the unique features of this is that it should be a three-day device for a lot of them. There will be a handheld. It


will be basal and bolus because there have to be meals. It will look a lot like a type 1 device but without some of the bells and whistles that are unique to patients with type 1 diabetes.

And it will leverage the new OmniPods.

Q: It won't be a completely separate manufacturing line down the road then?

A: No new manufacturing line.

Q: Could you give us some sense of international sales growth in the quarter either year- on-year or relative to the fourth quarter of 2012?

A: Compared to Q1 of 2012, which was probably much lighter in volume, it was probably more than triple what that number was back then and it was an uptick over Q4 of 2012. Ypsomed has done a great job of accelerating demand. Like here in the US, the first quarter is historically kind of the weakest of quarters and I think they were very pleased with their results as well, especially in markets like the Netherlands and Switzerland where they continue to take significant share. We've talked about that our international revenues should be probably somewhere between two times and three times what it was in 2012.

Q: On the efforts with Lilly, what is the filing for this device by Insulet? When the device is ready for market, to what extent will Lilly be engaged in the marketing and what can you say relative to the economics between Insulet and Lilly?

A: It's dual-path filing. Lilly's filing for a labeling change on the drug and we're going to file a 510(k) for use of the product. Right now, the only agreement we have with Eli Lilly is the development of this. We will be selling the pump and the pods and they will be selling the drug.

Q: Is there some sort of a transfer price from Lilly to Insulet?

A: No. During the development phase, there is some certain cost around the clinical and other things that both companies will share. That's a little bit of an expense for us over the course of the next couple of years while this is ongoing – not material. And then going forward, we will sell pods and we'll sell the system and they'll sell the drug. There is no cost share, if you will, for a patient once they are on product.

Q: I heard you say about highly resistant type 2 patients. Obviously, that's a qualitative characterization. Care to put some numbers on it in terms of patients in the US that you see? The reason I ask is, we are obviously seeing a recent push of insulin pens in type 2 patients. And there's a lot of buzz about betatrophin, although these are far out. I'm just trying to get a handle on the market size eventually when this product comes on line?

A: Let me take a step back and define what we call highly resistant. This is our definition – people in need over a 100 units of insulin a day. It's a pretty large dosage. I think if you take a look at this and these are our numbers, we are not speaking for Lilly, but we've been looking at this for a while. We've talked to a lot of our physicians and the belief is this market is going to be as big as the type 1 market if not bigger over time.

If you would look at what's going on in the type 2 space, there are more and more oral medications that are having bigger and bigger problems. There are a lot of different panel, FDA panels to look at these various oral medications as a potential type 2 [therapy] as to what the side effects are and there is more and more studies coming out that say the best way to treat – especially adolescent young kids that I would define as large body mass index kids for a lack of a better term year here – is to treat early on with insulin and try to reform the diet and maybe give them a little exercise. There is a trend that we believe that will make this market as big if not bigger than the type 1 market in the US, now that's our opinion, but it's a substantial market.

Q: Your internal market analysis must have indicated a certain price elasticity of demand. Can you give us some color on what your analysis is suggesting?

A: I think it's simpler than that. If you go back and you look at the various studies that have been performed using U-500 – all kinds of off-label use by various doctors over the last three, four years – there have been multiple studies. All of these are showing better outcomes in terms of improved day one sees for these patients. The single biggest issue for many of these patients is compliance, which it is with about any drug. If you do a full-blown study with clinical outcomes, we think we can easily justify the current pricing that we have in the marketplace for the type 1 product.

Q: On closed-loop systems, everyone is talking about the fact that you are not pared with Dexcom. It's too obvious to miss right now. You have made a strategic bet with this so far unknown player with a certain technology. Help us understand. Accuracy is one factor, but what features does your CGM partner have that can counter...?

A: I think despite what everyone's telling you, I think the idea of a closed loop system in which the patient is not involved, no one on this phone call will live long enough to see that, that's my humble opinion. Having said that, I think the best tool for patients is one they're going to use. And what we know is 90%- plus of people with type 1 diabetes are willing to live with one thing on their body if they need insulin. So our theory behind this is to integrate a continuous sensor in it. We have an option to license a couple of different software algorithms that are out there. Our version of the right product isn't CGM to replace a finger-stick, it's CGM to talk to a predictive algorithm – taking all these data points to keep the patient out of trouble, to only buzz, wake, kick, jump, whatever you have to make the handheld do, to keep the patient out of trouble. It could be in its simplest form, it could be a turn on the handheld and if the screen's green, then everything's fine. If the screen is yellow, let's say that says three to five hours based on the parameters you put in and based on what we are seeing here with the continuous sensing, you are going to run into a problem and then if it's a red screen, based on the parameters you put in, you are now bumping up against the guardrails.

We are not going into the CGM business to sell CGM. We believe the combination of CGM with one thing easy to use on the body is going to result in better clinical outcomes, reduced hospital stays and a much, much better like for people with type 1 diabetes.

Q: I wanted to confirm – are both gen-2 lines up and running again? What's the new timeline or what is the timeline for getting the third line up and running?

A: Both lines are up and running. And the third line, as we talked about in the last call, we're still targeting towards sometime this summer.

Q: Looking at the plan to transition to the LifeScan PDM, what's the timeline on that and is there any chance you can get that into the mix during the conversion process?

A: No, it won't be during the conversion process. At the moment, we are waiting for some information back from LifeScan. Although it does not impact our meter, our product with them, they do have a recall that they are dealing with that has probably distracted them slightly. We are working with them to be able to just get the rest of the information needed to be able to file with the agency.

Q: Is that potentially still a 2013 development or could that get pushed a little bit?

A: We were hoping it could be the end of the year that we'll have it.

Q: Then lastly, did you say 25% core OmniPod business growth year-over- year?

A: Yes.

-- by Adam Brown, Kira Maker, and Kelly Close