Memorandum

Abbott 2Q16 – Global sales up 3% operationally; FDA approval of FreeStyle Libre (consumer version) as early as 1Q17? “More than 125,000 patients” using Libre in Europe – July 20, 2016

Executive Highlights

  • Global Diabetes Care sales totaled $283 million in 2Q16, rising 2% as reported and 3% operationally year-over-year (YOY). The international business drove the performance with revenue of $210 million growing a strong 14% as reported and 15% operationally on the strength of FreeStyle Libre. US sales, on the other hand, declined 22% YOY, to mark the second-lowest revenue ever recorded in our Abbott model ($73 million).
  • Pooled global revenue for Abbott and J&J (who reported yesterday) totaled $754 million, falling 2% YOY relative to pooled revenue in 1Q15 (~$772 million). Pooled declines were driven by US weakness where combined sales of ~$250 million fell ~19% YOY against an easy comparison to pooled sales in 2Q15 (sales fell 4%).
  • CEO Mr. Miles White pushed back the US timeline for the FDA approval of FreeStyle Libre (consumer version), which is now expected in the “first quarter” of 2017. The device still has not been filed with the Agency, though this is planned for the near future. The timing is ambitious, as the lower risk, blinded FreeStyle Libre Pro has been under FDA review for more than a year at this point. On the plus side, management confirmed with us that approval of FreeStyle Libre Pro is expected imminently.
  • Mr. White shared that there are presently “more than 125,000 patients” using FreeStyle Libre in Europe – the first public estimate we’ve gotten on this front. For comparison, this is ~85% of Dexcom’s global installed base (140,000-150,000 patients) in less than two years on the market and in a direct consumer pay market. EU Libre sales back out to ~$45 million if all patients are using two sensors per month (~120 euros per month).

Early this morning, Abbott CEO Mr. Miles White led the company’s 2Q16 financial update. Below, we bring you our top financial and pipeline highlights from the call following by Q&A.

Financial Highlights

1. Global Diabetes Care revenue totaled $283 million in 2Q16, climbing 2% as reported and 3% operationally year-over-year (YOY). Guidance for 3Q16 called for mid-single digit sales growth, driven by sales of FreeStyle Libre – that seems reasonable considering relatively easy comparison against 2Q15’s 8% decline. Sequentially, sales grew 17% from a very low base.

2. US Diabetes Care sales declined 22% YOY, totaling $73 million in 2Q16. This marked the second-lowest revenue ever recorded in our Abbott model after last quarter’s $69 million. Notably, 13 of the past 14 quarters have now seen negative growth – yikes! There seems to be no end in sight to the challenges in the US, and given further CMS pricing cuts, the declines will likely continue until either FreeStyle Libre system is approved and launched in the US.

3. International Diabetes Care had another strong quarter, with revenue of $210 million growing 14% as reported and 15% operationally. This marks six consecutive quarters of operational growth. Management noted that FreeStyle Libre is driving the performance in a major way, though actual sales figures were not shared. Back of the envelope: assuming the 125,000 EU users are using Libre full time (two sensors per month, ~120 euros), revenue was ~$45 million from Libre. Revenue would be just over $20 million if users wear one Libre sensor per month on average. The former would be impressive, equating to over 20% of OUS sales.  

Pooled Financial Highlights

4. Pooled global revenue for Abbott and J&J totaled $754 million, falling 2% YOY relative to pooled revenue in 1Q15 (~$772 million). Pooled declines were driven by US weakness where combined sales of ~$250 million fell ~19% YOY against an easy comparison to pooled sales in 2Q15 (sales fell 4%). We are very struck that the US portion of total revenue doesn’t even reach 33% of the worldwide total - that’s a striking testament to how devastating the impacts of competitive bidding have been!

Pipeline Highlights

5. In the biggest news on today’s call, CEO Mr. Miles White pushed back the timeline for the FDA approval of FreeStyle Libre (consumer version), which is now expected in the “first quarter” of 2017. The device has not been filed with the Agency, though this is coming “as soon as possible.” The previous timeline “optimistically” called for a US launch by late 2016,” and given that it is not at FDA, the new timeline may be a bit too optimistic. This is certainly not for a lack of effort, however; said CEO Mr. Miles White, “When you've got a product that's as good as [Libre] is and that is trending like it is, you're in a big hurry to get it approved in the US yesterday.” We wonder if the submission is hung up on an insulin dosing claim, manufacturing, or other factors.

6. Mr. White estimated that there are presently “more than 125,000 patients” using FreeStyle Libre in Europe – WOW! This is the first figure Abbott has shared publically on this front and, for context, it is ~85% of Dexcom’s global installed base (140,000-150,000 patients) in less than two years on the market and in a direct consumer pay market. This is a testament to the composite profile of this product: fingerstick elimination, strong accuracy, low cost, a terrific on-body form factor, no-prescription ordering, etc. Libre is one of the rare devices that subtracts burden (fingersticks) at the same time it adds value (better glucose data).  

7. Despite some confusion on the call itself (see Q&A), we received word from management that FreeStyle Libre Pro is slated for a near-term US launch (submitted in 2Q15, over a year ago). Specific timing was not shared, though we’re interpreting the news as a slight delay considering the previous guidance for a mid-2016 launch (it’s already July, after all!).

8. Abbott’s six-month IMPACT study in 225 well-controlled type 1s (baseline A1c: 6.7 %) was not mentioned on the call, despite the long-awaited results presentation at ADA 2016. As a reminder, the study met its primary endpoint at six months – relative to patients on SMBG, patients using FreeStyle Libre spent a striking ~74 minutes fewer per day <70 mg/dl (a 38% reduction; p<0.001). For detailed thoughts on IMPACT, please see our full coverage from ADA.

9. There were no new details on LibreLink – the free Android app for scanning FreeStyle Libre sensors without the need for a separate reader – that has now officially launched in Sweden, the Netherlands, Germany, Italy and the UK. Abbott has said that more countries in Europe will roll out the app in the coming weeks.

10. A number of FreeStyle Libre studies have been added or updated on ClinicalTrials.gov: (i) a diabetes management study in pediatrics (SELFY) is currently recruiting participants (great name!); (ii) an accuracy study of the FreeStyle Libre consumer version in type 1 patients should wrap up data collection this month; and (iii) a recently initiated study investigating FreeStyle Libre in pregnant women has completed recruiting.

Honorable Mention: There were no additional pipeline updates on FreeStyle Libre or other Abbott meters.

Table of Contents 

Financial Highlights

1. Global Diabetes Care revenue totaled $283 million in 2Q16, growing 2% as reported and 3% operationally year-over-year (YOY). The performance came against an easy comparison to a reported sales decline of 5% YOY in 2Q15 though the operational comparison was a bit tougher (5% growth). The performance breaks a streak of ten consecutive quarters of reported declines (since 4Q14), but also represents a sixth consecutive quarter of sales below $300 million. Overall, we think the financials reflect the challenging environment in the US, though there is lots of momentum and upside with FreeStyle Libre outside the US (see below). We assume the business is still profitable and cash-generating, even at the lower prices.

  • On a sequential basis, global sales grew 17% from a very low base (sales of $243 million in 1Q16 were a record low).
  • Management does expect to carry this quarter’s momentum into 3Q16, sharing guidance of mid-single digit sales growth driven by international Libre sales. It is great to see this optimism after an extended rough patch. We wonder how much of the selling woes in the US have to do with: (i) CMS’s competitive bidding program that recently forced mail order blood glucose strips down an additional ~20% (see below); (ii) greater use of CGM in high frequency testers; (iii) less use of fingersticks with non-hypoglycemia-causing type 2 therapies; (iv) more high deductible health plans, causing users to buy less strips; or other factors.

Figure 1: Global, US, and International Quarterly Sales (1Q12-2Q16)

  • As Figure 1 demonstrates, the US business has driven global sales declines over the past three years, whereas OUS sales have remained relatively steady, with a solid uptick this quarter. Two-thirds of Abbott’s business now comes from outside the US, which has helped buoy the pooled global number – this used to be about 50%. Given the strengthening of the US dollar, Abbott’s underlying OUS business has actually grown more than the chart above would suggest.

2. US Diabetes Care sales declined 22% YOY, totaling just $73 million in 2Q16. This came on an easy comparison to 2Q15 when sales fell 5% YOY. This marks 13 of the past 14 quarters with negative growth (1Q15 was the lone exception, and it was against a very easy comparison) and four of the past five quarters when stateside sales have come in under $100 million. Sales did rise 6% sequentially, but that was against the smallest base we have on record ($69 million in 1Q16).

  • The US remains an incredibly challenging region for all BGM companies, and this quarter saw yet another particularly striking drop. It was only a little over two years ago (1Q14) that US sales dropped sub-$100 million for the first time since the acquisition of FreeStyle manufacturer TheraSense in 2004. At that time, the inability to hit the $100-million mark was seen as an ominous watermark, but we’re now at a “new normal” – Abbott’s US sales have come in under $100 million in six of the past ten quarters. Now, the even starker reality is that lower milestones ($60 million? $50 million?) are very much within reach.
  • We wonder how much of the steep US decline can be attributed to negative impacts from the second round of CMS’s competitive bidding program vs. other factors. The program has established lower new payment amounts for mail order and retail blood glucose strips – $8.32 per 50-count box ($0.17 per strip), down 20% from the current allowable payment of $10.41 ($0.21 per strip), and a striking 76% reduction from the pre-competitive bidding price (~$35).

3. International Diabetes Care revenue totaled $21o million, growing 14% as reported and 15% operationally YOY. For comparison, 2Q15 sales fell 6% as reported (easy comparison), but grew 10% operationally (tough comparison). The 2Q16 performance marks the second consecutive quarter of reported growth for the segment and the fifth consecutive quarter of operational improvement. The operational performance is particularly impressive on the heels of double-digit 10% growth in 2Q15, and it was no surprise to hear that FreeStyle Libre drove the success.

  • FreeStyle Libre sales could be estimated as ~$45 million in Europe, assuming the 125,000 EU users are using it full time (two sensors per month, ~120 euros). Revenue would be less than that if many use Libre intermittently; in a far more conservative case of one 14-day sensor per month, revenue would be ~$23 million. If the former is true, this is a striking 21% of international sales already, not including sales in other countries outside of Europe. Abbott did not report specific sales for FreeStyle Libre – this is our guess – but the user number does give a good broad stroke on the larger-than-expected size of the business.
  • Management alluded to tremendous FreeStyle Libre upside moving forward, and there is plenty of momentum to carry the product in new geographies: a recent reimbursement victory in Belgium and recent expansions into Argentina (forthcoming), Australia (June 2016), Israel and Brazil (July 2016).
  • Sales grew 21% sequentially, though the seasonality of Abbott’s business means that we always see positive sequential 2Q growth. However, 21% is still the single highest quarter-to-quarter growth that we have ever seen in our model, internationally or otherwise, hinting that this business is indeed trending in the right direction.

Pooled Financial Highlights

4. Pooled global revenue for J&J and Abbott reached ~$754 million, falling 2% YOY relative to pooled revenue in 2Q15 (~$772 million). This comes against an easy comparison as combined revenue declined 9% YOY a year ago. While some of the weakness comes from currency headwinds, it cannot explain the entire drop as the US market is driving the declines.

  • US weakness drove the pooled declines in 2Q16, with combined sales of ~$250 million down a striking ~19% YOY against a relatively easy comparison to pooled sales in 2Q15 (sales fell 4%). The performance essentially ties 1Q16’s all-time low pooled J&J and Abbott quarterly revenue (~$249 million).
  • Pooled international sales of $504 million rose 8% YOY in 1Q16 against an easy comparison to pooled revenue in 2Q15 (down 13% YOY). Depressed revenues as a result of foreign exchange had characterized these businesses for several quarters, and it seems like the market has bottomed out, and both companies now have room to show growth. This represents the first positive pooled YOY growth since 2Q14. Pooled sales rose 19% sequentially.
  • As a reminder, direct comparisons between Abbott and J&J are difficult because each company’s Diabetes Care business includes a fraction of non-BGM revenue. J&J has global insulin delivery, and Abbott has continuous glucose monitoring (Navigator II) outside of the US.

Pipeline Highlights

5. On the pipeline front, CEO Mr. Miles White unsurprisingly pushed back the timeline for the FDA approval of FreeStyle Libre (consumer version), now expected in the “first quarter” of 2017. This is a delay relative to Mr. White’s 4Q15 commentary that had suggested the consumer version could “optimistically” make it to the US market by late 2016. Consistent with our ADA coverage, we received confirmation from management that the consumer version has still not been filed with the Agency. The slow US timing remains a disappointment for a product that has been out for over two years in Europe. (As a reminder, the US pivotal study wrapped up in March 2015.) It is more surprising that the blinded “Pro” version hasn’t yet been approved, and we think that is on the way soon.  

  • The new timeline for FreeStyle Libre (consumer) feels a bit optimistic, given the ongoing 12+ month FDA review of the blinded “Pro” version. We’re surprised that the low-risk, retrospective-only Libre “Pro” sensor has not been approved, and one might guess the real-time consumer version will take as long … if not longer. The consumer version of FreeStyle Libre will presumably need an insulin-dosing claim to keep Abbott’s major marketing message (“No fingersticks!”). D`espite a rocky regulatory history, Abbott has some advantages: (i) a factory calibrated system with no noise from fingerstick calibrations; (ii) two large outcomes studies in IMAPCT and REPLACE showing non-adjunctive use in RCTs; and (iii) 125,000+ users on FreeStyle Libre using it non-adjunctively in the EU. On the other hand, Libre does not have alarms, and has slightly lower point accuracy than G5. The Dexcom CGM meeting could provide some clues on what FDA is thinking in this area, and there will undoubtedly be implications for Abbott. Dexcom management has said, “We will undoubtedly be the first company to have this [insulin dosing] label. And we want to set the bar high. You’ve got to have a sensor that performs as well as ours does.” We interpreted that as a competitive barrier to entry, though it’s possible that the FDA will respond in the other way: opening the door to an insulin-dosing claim and clarifying the regulatory path.
  • Despite the delay, sky-high expectations for FreeStyle Libre continue to be striking – said Mr. White: “When you've got a product that's as good as [Libre] is and that is trending like it is, you're in a big hurry to get it approved in the US yesterday.” This enthusiasm isn’t new but it’s remarkable to see just how much excitement the franchise continues to generate. The product is on the company’s radar in a big way, and judging from the way it was discussed, the device is clearly a source of pride for the company (as it should be, given the tremendous achievements on factory calibration, accuracy, cost, and on-body form factor).

6. Mr. White estimated that there are presently “more than 125,000 patients” using FreeStyle Libre in Europe – WOW! This is the first figure Abbott has shared publically on this front and though management has raved about Libre over the past year, we did not expect this level of penetration. This is already ~85% of Dexcom’s global installed base (140,000-150,000 patients) in less than two years on the market AND in a direct consumer pay market. Of course, CGM has historically been poorly reimbursed in the EU – a factor that has made it a no-brainer for EU patients to pay out-of-pocket for Abbott’s relatively cheaper product vs. Dexcom’s or Medtronic’s – though given the recent, positive reimbursement decision for CGM coverage in Germany, it stands to reason that Libre may face less of a pricing advantage (though getting CGM reimbursed may still be a hassle). Still, the current trajectory bodes very well for the future, and Abbott certainly has momentum of its own with its recent reimbursement victory in Belgium and recent expansions into Argentina (forthcoming), Australia (June 2016), Israel and Brazil (July 2016).

7. Despite some confusion on the call itself (see Q&A), we received confirmation from management that FreeStyle Libre Pro (blinded, retrospective) is still slated for a near-term US launch (submitted in 2Q15). Although an approval would have to come very soon to meet this timeline exactly, we assume even a slight delay isn’t really a big deal, especially since CEO Miles White did temper his forecast last year with “optimistically …”. We are very excited to see the Libre Pro approved in the US since it is so easy to use and we believe it will contribute to real therapeutic change (that is easy!) for so many patients and HCPs. The company’s US pilot study of the Pro – which began recruiting type 2 participants in April 2015 (n=132) – wrapped up last August, though results still have not yet been posted. It is possible that Abbott is waiting for the lower-risk Pro version to be approved before submitting the consumer version, but as far as we know, both submissions can be in at the same time.

  • During Q&A, Mr. Miles commented broadly on the challenge of hitting product launch timelines and chose to use FreeStyle Libre as an example. We’re not sure if he was referring to the US Pro or consumer regulatory process (or the EU process) but the remarks are striking all the same: “There are always speed bumps. Something slows you down somewhere, whether it's reimbursement or the timing of an approval or whatever. And in all my years in the industry, I have never seen, on a particularly consistent basis, any company including my own deliver its new products on time, on plan, on schedule without some kind of delay. Even with Libre, our ambitions, we're – in terms of our own internal goals, we're probably a couple of months behind where we wanted to be – months, okay, couple months. And yet, we have a pretty stiff goal in our own minds about what we're trying to do here. And if you look at the product, it's screaming north at a rocket-like pace, which is a good thing. And all it takes is some minister somewhere to slow down his decision a few weeks and you're off by couple weeks.”

8. Abbott’s six-month IMPACT study in 225 well-controlled type 1s (baseline A1c: 6.7 %) was not mentioned on the call, despite the long-awaited results presentation at ADA 2016. The study met its primary endpoint at six months – relative to patients on SMBG, patients using FreeStyle Libre spent a striking ~74 minutes fewer per day <70 mg/dl (a 38% reduction; p<0.001). Pre-specified secondary endpoints were particularly compelling – patients using Libre spent ~49 minutes fewer per day <55 mg/dl (a 50% reduction; p<0.0001) and ~33 minutes fewer per day <45 mg/dl (a 60% reduction; p<0.0001). Measures of nocturnal hypoglycemia were also significantly lower with FreeStyle Libre as patients spent ~28 minutes fewer per night (a 40% reduction; p<0.0001) in hypoglycemia. Patients using Libre spent ~22 minutes fewer per day in extreme hyperglycemia > 240 mg/dl (p=0.02) and spent ~60 minutes greater/day between 70-180 mg/dl (p=0.0006). There was not a significant difference in A1c between the groups, and both saw a marginal ~0.15% increase by the end of the study – a positive given that these patients were spending three hours per day in hypoglycemia at baseline! In other words, FreeStyle Libre prompted a higher quality A1c, with one hour less hypoglycemia per day.

  • IMPACT highlights the truly scary amount of hypoglycemia type 1s on insulin therapy are experiencing every day, and the tremendous challenges of dosing insulin as A1c approaches goal – all patients at baseline were spending ~200 minutes per day <70 mg/dl!!! The hypoglycemia data discussed above is very clinically meaningful (-74 minutes per day), and there is still room to improve therapeutic approaches: patients on Libre were still spending two hours <70 mg/dl per day at six months. Was the residual hypoglycemia driven by over-treating hyperglycemia while on FreeStyle Libre (so-called “hyper avoidance” in intensively managed patients)? Could that be managed with better education?
  • For detailed thoughts on IMPACT, please see our full coverage from ADA.

9. There were no new details on LibreLink – the free Android app for scanning FreeStyle Libre sensors without the need for a separate reader – that has now officially launched in Sweden, the Netherlands, Germany, Italy and the UK. Abbott has said that more countries in Europe will roll out the app in the coming weeks.

  • At ADA, we learned a number of new details on LibreLink: (i) that a new integration agreement with Diasend enables data from the LibreLink app to automatically populate a patient’s diasend account wirelessly and passively; (ii) that the LibreLink Apple platform expansion is “coming soon” though may require an iPhone adapter to power the NFC-reading capability; and (iii) that LibreLink will soon sync with LibreView, Abbott’s new cloud-based data management software. On the latter, we learned that the software allows patients to download data from both Abbott and other proprietary BGMs using an appropriate USB cable and can give healthcare professionals access to glucose data. Of course, it was only a matter of time before this web platform was made available, as both Dexcom (Clarity) and Medtronic (CareLink) already offer such platforms. These cloud data management systems may help with regulatory approvals, algorithm and device improvements, and even drive reimbursement in the future.
  • For more details on all the above, please see our detailed coverage from ADA.

 10. A number of FreeStyle Libre studies have been added or updated on ClinicalTrials.gov: (i) a diabetes management study in pediatrics (SELFY) is currently recruiting participants; (ii) an accuracy study of the FreeStyle Libre consumer version in type 1 patients should wrap up data collection this month; and (iii) a recently initiated study investigating FreeStyle Libre in pregnant women has completed recruiting.

  • SELFY will compare the efficacy of FreeStyle Libre vs. SMBG in self-management of diabetes. The multicenter trial hopes to enroll 70 type 1 patients between 4-17 years old, who will wear FreeStyle Libre (real time) for eight weeks. Primary endpoints are time in range and change in time in range, and the study’s estimated completion is November 2016.  We applaud Abbott for sponsoring a study in a population that has historically struggled to derive the same benefits with diabetes devices due to adherence issues, and we’re hopeful the study will confirm what many have shared anecdotally about the convenience of the device.
  • Abbott has updated an accuracy study of FreeStyle Libre in type 1 patients in the UK (Clinicaltrials.gov Identifier: NCT02734745) that is not yet open for recruitment. The single-center trial is aiming to enroll 24 patients, who will wear Libre for 14 days at home. During the study period, patients will perform at least seven fingersticks per day and will return to the clinical center on two separate occasions: (i) to assess accuracy vs. YSI during a meal test; and (ii) to assess accuracy vs. YSI during induced moderate hyperglycemia and subsequent hypoglycemia. The primary endpoint will be accuracy at 14 days; the study’s estimated completion is listed as May 2016. We wonder if this is testing a new algorithm or changes to the sensor.
  • A study investigating FreeStyle Libre in pregnant women with type 1, type 2, or gestational diabetes is now listed as “ongoing” and is no longer recruiting participants (ClinicalTrials.gov Identifier: NCT02665455). The trial enrolled 83 patients to wear a masked FreeStyle Libre for 14 days at home. The primary endpoint is point accuracy judged using the Clarke Error Grid, and study completion is slated for September 2016. We appreciate Abbott’s strong push to emphasize the utility of Libre across various age groups (pediatrics, adults), diabetes types (type 1, type 2, gestational), and CGM varieties (consumer, professional). Expanding the footprint in alternative populations is critical as Abbott faces increasing sensor competition from Dexcom and Medtronic.

Honorable Mention: There were no additional pipeline updates on FreeStyle Libre or other Abbott meters. Abbott’s newest BGM (the FreeStyle Precision Neo) launched in April 2015. Management has alluded in the past to expanding the Flash Glucose Monitoring category to other areas, and the Pro version is the first of these. We wonder if a continuous version of FreeStyle Libre will be developed though this would incorporate different principles that went into the original Libre (i.e. no alarms) – we assume this would be very possible. A tough call, and obviously, more SKUs add manufacturing complexity and could reduce margins.

Questions and Answers

Q: FreeStyle Libre is up and running in Europe. You said you're working to bring it to the US. What's needed to make that happen? What kind of timing should we expect for FreeStyle Libre in the US?

Mr. Miles White: I'm not so sure how specific to be about that. We are going to submit [the consumer version] as soon as possible and I would estimate it’ll be approved in the US in the first quarter. That's what I would guess. There are two stages to this – there's a Libre Pro and a Libre consumer. Our goal is to get this approved within the next six to eight months, something like that. But, we will see.

There are a number of steps that are going to happen in the interim. Also, the product is doing exceptionally well in its rollout in Europe where Libre's success thus far has been driven by consumer pay. We are achieving our goals of getting key countries to reimburse the product, which should frankly make a big difference in the performance in the market, and it's already trending at a pretty good clip.

So, I'm pretty enthusiastic about Libre and, of course, when you've got a product that's as good as this is and that is trending like this is, you're in a big hurry to get it approved in the US yesterday. So that's where we are.

Q: One Libre clarification question – when you mentioned approval in the “first quarter 2017,” is that the professional version or the consumer version? As far as I know, you haven't yet disclosed when you're filing the consumer version in the US. And second, on the Dexcom panel tomorrow, what do you think the implications are for Libre?

Mr. Scott Leinenweber (VP, Investor Relations): I'll just touch base on the professional version very quickly. As you know, we filed for that device in the middle of last year. We're working with the FDA on that and that is the approval we would expect in the next six to nine months. [Editor’s Note: Abbott management confirmed after the call that FreeStyle Libre Pro approval is expected in the near term rather than in six-to-nine months. FreeStyle Libre (consumer version) approval is expected in six to nine months (1Q17).]

As it relates to the Dexcom panel tomorrow, we're really excited about the US market. We think there's lots of opportunity there. We think Libre is a unique technology. It's a discussion we'll be watching closely to see how the FDA is thinking about these novel technologies and how to apply them for diabetic patients' needs in today's market.

 

-- by Brian Levine, Varun Iyengar, Adam Brown, and Kelly Close