Memorandum

Insulet 3Q15 – Revenue hits a record $87 million as US OmniPod grows 12%; next-gen PDM filing pushed to mid-2016; committed to AP – November 6, 2015

Executive Highlights

  • Total revenue reached a record $87 million, increasing 16% year-over-year (YOY) and 16% sequentially. US OmniPod sales rebounded from a weak 1H15, growing 12% YOY and 12% sequentially to a record-high $50 million. International OmniPod sales grew 9% YOY and 75% sequentially to $13.5 million.
  • Insulet now plans to submit its next-gen Bluetooth-enabled PDM to the FDA in mid-2016; launch is expected by the end of 2016. The company has identified artificial pancreas algorithm partners and mapped out an early clinical development pathway. Lilly has begun a 416-patient study testing the U500 OmniPod. End date is Dec 2016.

Insulet reported 3Q15 results yesterday afternoon in a call led by CEO Patrick Sullivan. Sales beat analysts’ and management’s expectations, as the US OmniPod business rebounded and drug delivery had another strong showing. We enclose our top 10 highlights below.

Financial and Business Highlights

1. Total revenue reached a record high $87 million, increasing 16% year-over-year (YOY) and 16% sequentially. The past two quarters (2Q15, 3Q15) have both set all-time records for total revenue. Management raised the low-end of its full-year revenue guidance (+$5 million) to $310-$320 million (9% YOY growth at the mid-point) from the previous $305-$320 million.

2. US OmniPod sales had a solid rebound from a weaker 1H15, growing 12% YOY and 12% sequentially to a record-high $50 million. This was the highest YOY growth seen in the past year and recovers from three straight quarters of sub-1% YOY growth. US new patients starts increased almost 25% YOY and 10% sequentially, the highest in the company’s history.

3. International OmniPod sales grew 9% YOY and 75% sequentially to $13.5 million. Ypsomed’s ordering patterns have returned to normal. These results “reflect true end-user demand.”

4. Management’s enthusiasm for the drug delivery business remains sky high. Sales of $7.1 million more than tripled YOY and were down 5% sequentially. As of Amgen’s 3Q15 call, 19% of all Neulasta doses in the US are now delivered via OmniPod. Management called the trajectory “terrific.” 

5. CEO Patrick Sullivan was “very unhappy” about the OmniPod field safety notification issued earlier this week. Insulet has corrected the manufacturing issue and product quality remains a serious priority. The P&L took a $7.7 million hit, a key driver in the quarter’s $15.7 million net loss.

6. Insulet’s key growth initiatives are aimed at retaining new patients and growing demand. The company is gathering better data to prove the OmniPod’s value to clinicians and payers. Studies are underway and one has already been submitted for publication.

R&D Pipeline

7. Insulet now plans to submit its next-gen, touchscreen, Bluetooth-enabled PDM to the FDA in mid-2016; launch is expected by the end of 2016. The new submission timing marks the second delay in a row (2Q15: “delayed a few months” from the previous “end of 2015”). Insulet is “on track” and working with Dexcom on G5 integration, though there were no product specifics or timing.

8. In Q&A, management called the artificial pancreas program “very active” and said it is “committed” to a product – Insulet has identified algorithm partners and mapped out an early clinical development pathway. Oddly, this was not mentioned in prepared remarks.

9. At a drug delivery conference last week in Vienna, Insulet debuted a larger volume OmniPod (for delivering biologics) and a Bluetooth-enabled OmniPod for transmitting dose information. The Bluetooth OmniPod could hypothetically send insulin delivery data directly to a phone, and perhaps allow a phone to control the pump in the future.

10. Lilly enrolled the first patient this month in a clinical trial of the U500 insulin OmniPod. According to the ClinicalTrials.gov posting, the 416-patient study is expected to complete in December 2016. Primary outcome is change in A1c at 26 weeks, and the study will compare U500 insulin via OmniPod vs. U500 insulin via MDI.

Financial and Business Highlights

1. Total revenue in 3Q15 reached a record high $87 million, increasing 16% year-over-year (YOY) and 16% sequentially. For the second straight quarter, the performance came in ahead of management’s guidance ($82-$85 million), which likely reflects some conservatism, stronger-than-expected drug delivery sales (it’s still early for that business), and the new US commercial team coming up to speed. Remarks highlighted that all four business segments reported YOY growth, though US OmniPod and Drug Delivery provided the vast majority of the quarter’s growth (44% and 39%, respectively). The encouraging results came on a tough YOY comparison to 23% growth in 3Q14, and now mark two consecutive quarters of double-digit sequential gains. The past two quarters (2Q15, 3Q15) have both set all-time records for total revenue – the business seems to be gaining some momentum. More on the US OmniPod, International OmniPod, and Drug Delivery segments is below.

  • Management raised the low-end of its full-year revenue guidance (+$5 million) to $310-$320 million (9% YOY growth at the mid-point) from the previous $305-$320 million. Only when prompted in Q&A did management share Q4 forecasts for $50-$52 million in US OmniPod sales (~17% YOY growth), $14-$16 million in international OmniPod sales (~13% YOY growth), and ~$16 million in Neighborhood Diabetes sales (~8% YOY growth). Prepared remarks called out the ~$2-$3 million rise in the full-year drug delivery guidance (now $25-$30 million from the previous $23-$27 million; ~$6-$10 million in Q4) – there continues to be lots of enthusiasm there.   

Insulet Worldwide Revenue (2012-3Q15)

  • Operating loss in 3Q15 increased substantially to $15.7 million from $2.9 million in 3Q14. Management highlighted two key drivers: (i) non-recurring charges of $7.7 million associated with certain product that did not meet the company's new higher quality expectations (see #5 below); and (ii) increased investment in commercial and development initiatives, in-line with previous expectations.

2. US OmniPod sales had a solid rebound from a weaker 1H15, growing 12% YOY and 12% sequentially to a record-high $50 million. This was the highest YOY growth seen in the past year and recovers from three straight quarters of sub-1% YOY growth. It was a relatively easy YOY comparison (for Insulet) to 9% YOY growth in 3Q14. Still, 3Q15 marked two straight double-digit sequential increases, a good sign that the new US commercial team is indeed coming up to speed. Remarks did note that the US results were buoyed by a sales force expansion, so it remains to be seen if the growth is sustainable going forward. Q4 US OmniPod sales are estimated at $50-$52 million, a very slight sequential increase and slightly better ~17% YOY growth.

  • US new patients starts increased almost 25% YOY and 10% sequentially, the highest in the company’s history. Pediatric new patient starts had even better metrics: ~30% YOY growth and a 15% sequential increase. The OmniPod has always been strong in pediatrics, though the competition is also getting much stiffer, particularly Medtronic (MiniMed Connect for remote monitoring, 640G and 670G coming in the next couple of years). Will the tubeless differentiator lose its advantage once competing automated insulin delivery devices come to market first? Management maintained that 70% of new patients come from MDI.

Insulet US OmniPod Revenue (2013-2015)

  • Can Insulet return to the 25-30% YOY US OmniPod growth it saw for years? New patient starts seem to be growing at this rate, but given the larger installed base and greater base of sales, this level of growth gets tougher every quarter. And of course, competition is only going in one direction.

3. International OmniPod sales grew 9% YOY and 75% sequentially to $13.5 million. Ypsomed’s inventory and ordering patterns have returned to normal. These results “reflect true end-user demand,” said management – as a reminder, 1H15 results were very negatively impacted by inventory de-stocking (indeed, this quarter’s $13.5 million in sales eclipsed cumulative 1H15 sales of $11.4 million). Ypsomed’s F1H16 update last week suggested 19% growth in the OmniPod patient base in the first half of its fiscal year, which we estimate added a couple thousand new patients in Europe. Q4 international OmniPod sales are estimated at $14-$16 million, a slight sequential increase and ~13% YOY growth.

  • The big question is how sales will change once Ypsomed launches its own durable YpsoPump in early calendar 2016. More speculation on that here. Insulet management did not comment on this, though both companies seem outwardly committed to the OmniPod partnership at this point in time.

Insulet International OmniPod Revenue (2013-2015)

4. Management’s enthusiasm for the drug delivery business remains sky high. Sales of $7.1 million more than tripled YOY and declined 5% sequentially. The Amgen product is seeing very strong uptake – as of Amgen’s 3Q15 call, 19% of all Neulasta doses in the US are now delivered via OmniPod. This product only launched earlier this year, and Insulet hopes to secure 100% of all Neulasta doses (~1 million per year). Management called the trajectory “terrific,”  and one it sees continuing in the future. Full-year guidance for drug delivery was increased for the second straight quarter, now at $25-$30 million (from the previous $23-$27 million and $15-$20 million). 

  • Insulet has signed six additional drug delivery agreements, three with top 15 big pharma partners. No further details were given, though these won’t be revenue drivers in the near-term. We wonder if any of these partnerships concern pumpable diabetes medications, such as GLP-1 or amylin.
  • As a reminder, the Amgen Neulasta delivery kit is a single use, customized OmniPod for oncology medication pegfilgrastim. The pod automatically delivers a dose of Neulasta ~27 hours after it is applied (the same day as chemotherapy). The Neulasta device automatically shuts down and alarms the patient to remove it. The kit offers major patient convenience and adherence advantages.

5. CEO Patrick Sullivan was “very unhappy” about the OmniPod field safety notification issued earlier this week. Insulet has corrected the manufacturing issue and product quality remains a serious priority. The P&L took a $7.7 million hit, a key driver in the quarter’s $15.7 million net loss. Insulet noticed a slight increase in reported cases of OmniPod needle deployment failure or delay (a low 1-2% incidence) in 15 lot numbers distributed in the US. Once the issue was recognized in the field (10 days after shipment), Insulet voluntarily recalled the product and added additional inspection steps in its manufacturing process. To us, the low absolute risk here is more of a signal of management’s commitment to quality than a serious safety issue. Like any notification, however, it is not reassuring to users.

  • Why did it happen? Insulet made changes in its manufacturing line in July to improve product performance. Product manufactured in July passed all existing testing and release criteria, but performance in the field identified a failure mechanism that the company had not seen in production. That issue is now tested for with improved inspection procedures in the manufacturing process.
  • The notification was voluntarily issued with the knowledge of the FDA, possibly as an act of caution given the warning letter Insulet received from the Agency in June. As a reminder, Insulet imposed a higher quality bar for acceptable OmniPods in 2Q15, which resulted in a one-time $2.5 million charge from incremental scrap and warranty costs.
  • CEO Patrick Sullivan reminded listeners that newly hired VP, Regulatory, Quality and Clinical Affairs Michael Spears brings significant experience from Medtronic and Covidien. Mr. Spears reports directly to CEO Patrick Sullivan, who repeatedly emphasized how seriously he takes product quality.

6. Chief Commercial Officer Shacey Petrovic outlined key growth initiatives aimed at retaining new patients and growing patient and clinician demand. The clear focus is on getting better data to prove the OmniPod’s value to clinicians and payers – we see this as very key, given the higher costs of pumps and modest ~0.5% (or less) benefit on A1c. Of course, there are other benefits to pumps too (quality of life, less hypoglycemia), and we hope robust studies can tease out the value for payers.

  • Insulet has several clinical studies in process to demonstrate the “clear clinical differentiation of OmniPod.” The first of several papers was submitted to an journal this quarter (not specified), and Insulet expects at least two more submissions this year. We are curious to see the study designs. The company has also developed a validated healtheconomic model to demonstrate the OmniPod’s value to payers.
    • Ms. Petrovic highlighted the observational study published in BMJ this year (and shown at EASD), suggesting reduced morbidity and mortality in 2,441 patients using pump therapy in Sweden. Patients in the study were followed from baseline in 2005-2007 to 2012, with a mean follow-up of seven years. Overall, pump use was associated with an impressive 44% reduction in fatal CVD (p=0.003) and a 29% reduction in total mortality (p=0.003) compared to those using MDI. The researchers used propensity scoring to account for the baseline differences between the pump and MDI groups; A1c’s were not different at baseline, though things like age, use of lipid drugs, smoking, GFR, and previous CVD all favored the pump group. The results drew many critical questions at EASD around correlation vs. causation in Q&A. Observational studies are always hypothesis generating, but this does not seem like the data payers are going to jump at.
  • The sales force expansion is complete, which Ms. Petrovic said is “clearly driving growth.” She pointed to strong record-high revenue and new patient starts this quarter, in part due to more effective targeting. The plan as of the 4Q14 call was to hire 15 new reps, bringing the total sale force to ~150.
  • Insulet has hired a larger in-house field clinician team to handle new patient training. Previously, Insulet-employed clinicians trained less than half of new patients. So far, the team has seen a 50% increase in the number of new patients trained by Insulet employees – presumably the base is low, though the call did not specify how large this team is or many patients this included. The big advantage is more standardized training.
  • Optimized internal support to improve the new customer experience. Ms. Petrovic said service levels have improved “dramatically,” particularly in giving new patients faster research and delivery of benefits. This has resulted in a higher conversion from leads to new patients starts, and improved time to conversion.

R&D Pipeline

7. Insulet now plans to submit its next-gen, touchscreen, Bluetooth-enabled PDM to the FDA in mid-2016. Launch is expected by the end of 2016. This important pipeline project was skipped in prepared remarks and only came up in Q&A. The new submission timing marks the second delay in a row – the 2Q15 update pushed submission back from “end of 2015” to roughly early 2016 (“delayed a few months”). Management did not characterize today’s update as a delay.

  • Insulet is “on track” and working with Dexcom on G5 integration, though there were no product specifics or timing. We would be surprised if the new handheld launches with Dexcom G5 integration on the PDM itself – that is presumably a PMA application, and a sub-six-month FDA review seems unlikely. Much more likely is some kind of app integration, whereby the Insulet handheld communicates with the Dexcom G5 app, or perhaps Dexcom’s G5 communicates with an Insulet app. Management is evaluating ways of working with Dexcom to accelerate CGM integration, “maybe using mobile.” We assume an Insulet app would only passively display data at first and not control the pump.
  • “Our market research demonstrates patients want to see information on the phone.” It was encouraging to hear management’s appreciation of this fact, especially because the criticism of the PDM is that it looks like a clunky medical device. Medtronic’s MiniMed Connect has also launched to good reviews, particularly in parents who value the remote monitoring of insulin and CGM data. Insulet is definitely playing catch-up here. 
  • Management called out the “somewhat obsolete” Tandem and Animas G4 integrations. As a reminder, Dexcom is still supporting these companies with G4, though the goal is to eventually move them over to G5. Tandem has announced plans to do that officially with an automated insulin delivery product (IDE filing by end of year to begin a feasibility study); Animas has not commented on G5 integration. The challenge is the regulatory review cycle for these pumps – 13 months for the Tandem t:slim G4 and 20 months for the Animas Vibe vs. a remarkable six-month PMA review for Dexcom’s G5.
  • As a reminder, this partnership was rekindled year ago at ADA 2014, allowing Insulet/Dexcom patients to view their pump and CGM data on both their smartphone (within a modified version of the base Dexcom Gen 5 app) and right on the Insulet PDM screen itself (the Gen 5 transmitter will pair with two Bluetooth devices).

8. In Q&A, management called the artificial pancreas program “very active” and said it is “committed” to a product – Insulet has identified algorithm partners and mapped out an early clinical development pathway. Like the next-gen PDM, however, this critical pipeline project received no airtime in prepared remarks, presumably because it is so early. Still, management was adamant that it is coming: “We are committed to an artificial pancreas product offering. I believe we have the best product for artificial pancreas use. We are committed to get there.”

9. At a drug delivery conference last week in Vienna, Insulet debuted a larger volume OmniPod device and a Bluetooth-enabled OmniPod for transmitting dose information (“remote patient manager”). Management said feedback has been “strong” and the developments stemmed from market feedback from pharma companies. These comments were part of the overall drug delivery business update in prepared remarks – the diabetes implications were only probed during Q&A.

  • The Bluetooth-enabled OmniPod could theoretically send insulin delivery data directly to a phone, and perhaps allow a phone to control the pump. Management was mildly optimistic on the possibility of this, and the manufacturing capabilities exist to do this now. The regulatory path is the big stumbling block, and we imagine it is several years away. Eliminating the PDM would be a dramatic product improvement for Insulet, particularly if an artificial pancreas algorithm could sit in the phone too.
  • The larger capacity OmniPod will not be applied to diabetes – Insulet is instead pursuing the concentrated U500 insulin OmniPod with Lilly (see below). Management believes those who need more insulin capacity would prefer the smaller OmniPod with concentrated U500 insulin vs. a larger OmniPod with standard U100 insulin.
    • There is potential value in both a U500 OmniPod and a larger reservoir U100 version. The current OmniPod only holds 200 units of U100 insulin, putting it well behind Tandem (480-unit t:flex), Medtronic (300 units), and Roche (315 units). Is there a middle ground to expand the reservoir size for patients needing ~100 units of insulin per day? We assume that patient population wouldn’t opt for U500 insulin, and the current OmniPod wouldn’t last two days. The upside to a larger reservoir U100 OmniPod is potential faster time to market, as the U500 road is longer and more uncertain (it relies on Lilly, an update to the U500 insulin label, massive human factors overhaul, lots of new education, etc.). Of course, a larger reservoir OmniPod has the big downside of requiring new manufacturing, not easily done.

10. Lilly enrolled the first patient this month in a clinical trial of the U500 insulin OmniPod. According to the ClinicalTrials.gov posting, the 416-patient study is expected to complete in December 2016. Primary outcome is change in A1c at 26 weeks, and the study will compare U500 insulin via OmniPod vs. U500 insulin via MDI. Lilly is running the study and recruiting at an impressive 64 locations. It is terrific to see the project finally getting off the ground – this partnership was first announced in May 2013, so it has been almost three years in the making just to get to this study.

Pipeline Summary

Pipeline Product

Timeline

Next-gen Bluetooth-enabled PDM

FDA 510(k) filing in mid-2016, with launch by end of 2016

PDM integration with Dexcom Gen 5 app and transmitter

Timing unknown (i.e., either with the new PDM launch or after it)

Updated OmniPod PDM for use with Lilly’s Humulin U500 insulin

Study underway, expected to wrap up in December 2016.

Artificial Pancreas

Identified algorithm partners and mapped out an early clinical development pathway

OmniPod with integrated CGM

[single on-body device]

We assume no longer in development. No updates since 4Q14, when development was ongoing and the first in-human trial was expected in early/mid-2015.

Bluetooth-integrated OmniPod for sending dose information

Unknown

Larger Reservoir OmniPod for drug delivery

Unknown

Drug Delivery

- Amgen Neulasta Delivery Kit (oncology)

- Ferring (infertility)

- Capricor (heart failure)

- Other partnerships (obesity, pulmonary hypertension, oncology, and Parkinson’s medications)

 

- Launched

- Launched

- Supplying for phase 2 trial

- No specifics shared

-- by Adam Brown and Kelly Close