Memorandum

AZ 1Q18 – Portfolio grows 6% YOY on strength of Farxiga, +44% YOY to $299M; GLP-1 class climbs 30% YOY to $1.9B; SGLT-2 class grows 41% YOY to $1B – May 18, 2018

Executive Highlights

  • AZ Diabetes started off the year with a positive performance, driven entirely by SGLT-2 inhibitor Farxiga. Whole portfolio sales grew 6% YOY to $607 million, up from $574 million in 1Q17 (-16% sequentially, from $725 million in 4Q17). Farxiga sales of $299 million grew 44% YOY, and fell 10% sequentially, from $207 million in 1Q17 and $332 million in 4Q17. Expanded indications could spur additional growth in the Farxiga business – for example, dapagliflozin has been submitted to EMA for a type 1 indication (FDA filing anticipated in 2H18). Management was enthusiastic about Bydureon BCise (the new GLP-1 autoinjector), but in 1Q18, the franchise fell 9% YOY and 5% sequentially to $139 million. AZ emphasized that ~half of BCise patients are new to GLP-1 therapy, so there is potential here for the autoinjector to grow the class, expand GLP-1 use to more type 2s, and achieve commercial success. We agree.

  • On a pooled basis, GLP-1 agonist revenue rose 30% YOY in 1Q18 to $1.9 billion, vs. $1.4 billion in 1Q17. The market dipped a modest 1% sequentially. Novo Nordisk’s Victoza and Lilly’s Trulicity drove almost all of this YOY growth, although new once-weekly Ozempic (semaglutide) also contributed $11 million revenue after launching to US pharmacies in February. Coming up, all eyes are on REWIND to see if Trulicity (dulaglutide) demonstrates cardioprotection.

  • The SGLT-2 class grew 41% YOY to $1 billion, up from $715 million in 1Q17. The market dropped 2% sequentially. The strong YOY performance is a larger YOY jump than seen from GLP-1 agonists, which makes sense because the SGLT-2 market is rising from a lower base, but it’s also notable because GLP-1s were growing faster from a higher base for much of 2017. Lilly/BI’s Jardiance led the class by value, with $456 million revenue reflecting 45% of pooled sales. AZ’s Farxiga captured 30% of the market by value with $299 million sales, and J&J’s Invokana held 25% with $248 million. Merck/Pfizer’s Steglatro posted ~$8 million in its first (partial) quarter on the market, launched at the end of January.

  • The DPP-4 market climbed 9% YOY to $2.4 billion, from $2.2 billion in 1Q17. Pooled sales fell 6% sequentially, as this market continues to fluctuate. Nonetheless, the DPP-4 inhibitor market was second only to the basal insulin market in total 1Q18 revenue (see the first table in our report). Merck’s Januvia is unchallenged as the frontrunner – the sitagliptin franchise posted $1.4 billion in 1Q18 and captured 60% of the market by value.

  • Including insulins, pooled branded diabetes drugs grew nearly 10% YOY to just under $10 billion – they fell 3% sequentially. Basal insulin and DPP-4 inhibitors continue neck and neck at $2.4 billion each in pooled sales.  

    AZ provided its 1Q18 update this morning in a call led by CEO Mr. Pascal Soriot. Click for the press release, presentation slides, and clinical trials appendix - we love that AZ provides this level of detail on R&D and we’d absolutely love to see other companies follow suit.

    We begin this report with three pooled market analyses on the GLP-1 class, SGLT-2 class, and DPP-4 class. We’ve also summarized results on all major diabetes drug classes in the table below. Then, you’ll find financial highlights covering AZ’s major diabetes products, followed by pipeline highlights relevant to diabetes, obesity, and NASH. There were no diabetes-specific questions during Q&A.

    Pooled Markets Summary for 1Q18

    Therapy Class

    Pooled Sales 1Q18 (billions)

    YOY Growth

    Sequential Growth

    Basal insulin

    $2.4

    +2%

    -9%

    DPP-4 inhibitors

    $2.4

    +9%

    -6%

    GLP-1 agonists

    $1.9

    +30%

    -1%

    Rapid-acting insulin

    $1.7

    +8%

    +4%

    SGLT-2 inhibitors

    $1.0

    +41%

    -2%

    Basal insulin/GLP-1 fixed-ratio combinations

    $0.07

    +252%

    +24%

    Sum of branded diabetes drugs (from therapy classes above)

    $9.5

    +11%

    -3%

    1Q18 Financial Results for AZ’s Major Diabetes Products

    Product

    1Q18 Revenue (millions)

    Year-Over-Year Reported (Operational) Growth

    Sequential Reported Growth

    Share of Growth

    Farxiga

    $299

    +44% (+39%)

    -10%

    100%

    Bydureon

    $139

    -9% (-11%)

    -5%

    0%

    Onglyza

    $129

    -16% (-19%)

    -28%

    0%

    Byetta

    $31

    -33% (-35%)

    -35%

    0%

    Symlin

    $9

    -36% (-36%)

    -31%

    0%

    Total Diabetes

    $607

    +6%

    -16%

    --

    Pooled Market Highlights

    1. GLP-1 Agonists Post Impressive YOY Growth (+30%) Despite 1% Sequential Fall; Pooled Revenue Totals $1.9 Billion; Victoza Claims 53% Value Share, Trulicity Up to 36%

    The GLP-1 agonist class grew 30% YOY to $1.9 billion in 1Q18 (vs. $1.4 billion in 1Q17). We see this as another very strong performance from the advanced injectable agents, despite a small 1% sequential dip from >$1.9 billion in 4Q17. Save the addition of Novo Nordisk’s Ozempic, which posted $11 million in its first partial quarter (<two months) on the market, class growth in 1Q18 was driven by Lilly’s Trulicity (+82% YOY) and Novo Nordisk’s Victoza (+4% YOY), which were the only GLP-1s to experience YOY growth. Moreover, Trulicity was the only GLP-1 agonist to rise sequentially in 1Q18, and even that was mild at +5% – the product’s lowest sequential jump since launch. These muted sequential performances from individual GLP-1s certainly explains the 1% QOQ drop in the market, although we could just as easily consider this a flat sequential change. It’s important to note that the GLP-1 market is now growing from a much higher base. Short-term fluctuation is not uncommon, and the class also experienced a 1% sequential loss in 3Q17 before posting record sales of >$1 billion in 4Q17 and growing 32% YOY to $6.5 billion in 2017 overall. Speaking of a higher base, Novo Nordisk highlighted in its 1Q18 press release that GLP-1 agonists now comprise 12% of the total global diabetes market by value, an increase from 10% one year ago; in the US, GLP-1s make up 15% of total diabetes revenue. At the same time, we continue to see plenty of room for growth. GLP-1s make up only ~7% of second-line diabetes prescriptions in the US and ~30% of basal insulin prescriptions worldwide, despite impressive glucose-lowering efficacy, potentially lower injection burden, weight loss, and little to no hypoglycemia risk. We expect class growth to be further spurred by the Ozempic launch and, in a few years, the entry of oral semaglutide and Intarcia’s ITCA 650, but we also recognize that much more work needs to be done on access, cost, and provider education. For all the thought leader enthusiasm around GLP-1s, perhaps Dr. Ralph DeFronzo put it most aptly in arguing that the best GLP-1 agonist is the one a patient can access (i.e. what will his/her insurance cover?): “All of these drugs are fantastic, and all you need to do is make sure your patient is on one of them,” he said at AACE 2018.

    • By value, Novo Nordisk’s Victoza retained market leader status in 1Q18 with 53% share and $989 million in revenue. Next up, Lilly’s Trulicity claimed 36% of pooled revenue with $678 million sales. Stunningly, these two products left only 11% of the market for five other agents. That share was split between AZ’s Bydureon (7%, $139 million), AZ’s Byetta (2%, $31 million), GSK’s discontinued Tanzeum (1%, $18 million), Novo Nordisk’s Ozempic (1%, $11 million), and Sanofi’s Lyxumia (<1%, $6 million). This reflects a slight shift in value shares from 3Q17 and 4Q17, when the breakdown didn’t change at all among Victoza, Trulicity, and Bydureon. From 4Q17, Victoza’s value share fell slightly from 54% to 53% and Bydureon’s from 8% to 7%, while Trulicity’s climbed from 34% to 36%; Byetta also fell from 3% to 2% value share, and Ozempic newly claimed 1% of the market in 1Q18. We expect further shifting over 2018 and beyond as formulary access is established for Ozempic, as Trulicity likely keeps growing, and as familiarity with GLP-1 continues to build.

      • Market share by volume tells a slightly different story, at least for the US. Stateside, Victoza held 44% of all GLP-1 agonist prescriptions (TRx) in 1Q18. Trulicity held 37% TRx, followed by 15% TRx for Bydureon and Byetta combined. Per our understanding, this most likely reflects a higher realized price for Victoza (i.e. from lower rebates paid to PBMs/payers – the reason Novo Nordisk’s value share is higher) and a lower realized price for Bydureon (the reason volume share is higher than value share), with the caveat that there is also likely some discrepancy between US and global volume. Still, it’s interesting to note that AZ mentioned pricing pressure around Bydureon in the US, while Novo Nordisk said the opposite – that pricing pressure around Victoza is relatively light. Novo Nordisk management attributed lower pricing pressure in the GLP-1 category to greater within-class differentiation; where payers largely consider all mealtime insulins the same, they recognize the key differences between GLP-1 agonists. So, there is also the possibility that AZ is up against more pricing pressure in the GLP-1 category because Bydureon is considered an inferior product.

    • Cardioprotection is a key point of conversation with respect to GLP-1 agonists, and Novo Nordisk has an edge for now, as Victoza is the only GLP-1 indicated for CV risk reduction. All eyes are on REWIND, the CVOT for Trulicity expected to complete in July 2018, with topline results anticipated in 4Q18. Lilly is aiming to present full results at ADA 2019. Dr. Ralph DeFronzo has forecast that REWIND results will be positive. Novo Nordisk CSO Dr. Mads Thomsen has cast doubt on the ability of Trulicity to demonstrate cardioprotection due to a larger primary prevention cohort in the trial, but in response, Lilly management has elaborated that REWIND compensates for its larger primary prevention cohort with a longer duration of follow-up (median 6.5 years vs. 3.5 years in LEADER). We’ve heard nothing but confidence from Lilly about what REWIND will show, and management is acutely aware that positive CVOT results are a near imperative for Trulicity’s long-term commercial success. Lilly’s Head of Diabetes and SVP Mr. Enrique Conterno has acknowledged as much, pointing to the rising bar for all diabetes drugs: Patients and providers alike are learning that they can get CV and renal protection from their glucose-lowering medicine. Victoza received a CV indication last August. Ozempic has also demonstrated CV benefit in the smaller and shorter (pre-market) SUSTAIN 6 trial, but Novo Nordisk will conduct the post-market SOUL CVOT before pursuing a formal CV indication. Our sense is that a fair number of thought leaders already consider semaglutide to be cardioprotective, while also recognizing the need for a definitive CVOT.

      • In addition to what it means for Lilly’s GLP-1 agonist business, REWIND will have huge implications on the diabetes field. The trial should shed valuable light on whether or not cardioprotection is a GLP-1 class effect. It will also offer insight into the role of GLP-1s in primary prevention of CV disease. REWIND enrolled a 69% primary prevention cohort, compared to 19% in LEADER, 17% in SUSTAIN 6, and 27% in EXSCEL, making up for the lower event rate with a longer median follow-up of 6.5 years, compared to 3.5 years in LEADER, ~2 years in SUSTAIN 6, and 3.2 years in EXSCEL. Average duration of diabetes is 10 years in REWIND vs. 13 years in LEADER and 14 years in SUSTAIN 6, so in general, Lilly’s enrolled population is “less sick.”

    Pooled GLP-1 Agonist Sales (1Q06-1Q18)

    2. SGLT-2 Inhibitor Class Grows 41% YOY, Falls 2% Sequentially to $1 Billion; Jardiance Leads with 45% Total Sales; Farxiga Leads with 41% Global TRx; Invokana Continues Decline; Steglatro Enters Market

    The SGLT-2 inhibitor market climbed 41% YOY to just over $1 billion in 1Q18, up from $715 million in 1Q17. This is strong YOY growth – the strongest, in fact, that the class has seen since 4Q16 (pooled revenue rose 18% YOY in 1Q17, +20% YOY in 2Q17, +35% YOY in 3Q17, and +23% YOY in 4Q17). That said, AZ management highlighted on today’s call that SGLT-2 growth decelerated in 1Q18, and indeed, pooled sales fell 2% sequentially from an all-time high in 4Q17 (>$1 billion). The decline was driven entirely by J&J’s Invokana (canagliflozin), which dipped 13% YOY to $248 million. AZ’s Farxiga (dapagliflozin) revenue grew 44% YOY as reported (+39% YOY operationally) to $299 million, and we elaborate on this first quarter performance under financial highlights below. Lilly/BI’s Jardiance (empagliflozin) shows the steepest climb, as reflected on the graph below; 1Q18 sales of $456 million more than doubled YOY (albeit, from a lower base of $224 million in 1Q17, but this is still in the ballpark of Invokana and Farxiga sales now). For the purposes of this analysis, we’ve estimated total Jardiance franchise sales based on the assumption that Lilly collects 33% of the global total (only Lilly’s revenue from the product, and not BI’s, is publically disclosed). Notably, this pooled analysis includes Merck/Pfizer’s SGLT-2 inhibitor Steglatro (ertugliflozin) for the first time, and we’ve calculated that 1Q18 revenue totaled ~$8 million; this was not reported explicitly, but Merck’s 1Q18 materials included a figure for total diabetes sales and a slightly lower figure for Januvia/Janumet sales. Jardiance surpassed Invokana and Farxiga to become the market leader by value in mid-2017, and in 1Q18, captured 45% of pooled revenue by our calculations (vs. 42% in 4Q17 and only 31% in 1Q17). Farxiga captured 30% of the market by value, Invokana captured 25%, and Steglatro captured <1% (we didn’t expect ertugliflozin sales to be very high, as these products were only just launched in 1Q18 and reimbursement will be low until 2019). AZ included volume estimates in its 1Q18 earnings presentation: Farxiga remains the market leader by volume with 41% of total global SGLT-2 prescriptions (TRx), while Jardiance held 36% TRx in 1Q18 and Invokana held 16%. From the graph on slide 17, we notice that Jardiance’s volume share has been climbing steadily as Invokana’s falls, and all the while, Farxiga has maintained its ~40% volume share. This matches our sense of sales trends for the SGLT-2 class, in that Jardiance is “stealing share” from Invokana as Farxiga grows at pace with the overall market. That said, AZ management emphasized that Farxiga’s growth in 1Q18 exceeded whole class growth (44% YOY vs. 41% YOY).

    • We suspect that AZ’s commentary on slowing SGLT-2 growth applies primarily to the US market, where pricing pressure abounds. Farxiga sales rose 32% YOY in the US (to $127 million) vs. 55% YOY OUS (to $172 million). J&J’s Invokana franchise is facing several headwinds, including a black box warning for lower-extremity amputations – but we don’t want to underestimate the adverse impact of pricing pressure on Invokana sales, especially since J&J’s SGLT-2 business is much bigger in the US vs. OUS (in 1Q18, $204 million vs. $44 million, respectively). In recent quarters, J&J management has cited high rebating (Invokana is preferred on the CVS Health formulary in 2018, and we imagine the company agreed to pay a larger average rebate to secure this positioning), segment mix (more prescriptions going to patients on Medicaid), and generous patient discounts to explain Invokana’s sluggish financial performance. We’d add that Janssen’s Patient Assistance Program (PAP) for Invokana and Invokamet was ranked highest (alongside Merck’s PAP for Januvia) in diaTribe’s recent rundown of PAPs for diabetes drugs. Meanwhile, we imagine the Farxiga business has benefited from the fact that it generates a majority of its revenue (58%) OUS.

    Pooled SGLT-2 Inhibitor Sales (1Q13-1Q18)

    • The SGLT-2 inhibitor class is affected by multiple opposing forces, a series of pushes and pulls. We list some of the key factors here, alongside key upcoming events for the class.

      • Cardioprotection: The evidence points to CV benefit as a class effect of SGLT-2 inhibitors, and most thought leaders have espoused this view, although there are notable exceptions (e.g. Dr. Sanjay Kaul at CMHC West). By evidence, we mean EMPA-REG OUTCOME for Jardiance, CANVAS for Invokana, and AZ’s CVD-REAL + CVD-REAL 2 (as well as EASEL, sponsored by J&J). In December 2016, FDA updated the Jardiance product label to include a CV indication (for the reduction of CV death in patients with type 2 diabetes/established CV disease). Jardiance has certainly benefited from being the only SGLT-2 inhibitor approved for CV risk reduction (again, product sales more than doubled YOY in 1Q18), but we hope Invokana will be joining these ranks soon – an FDA decision on Invokana’s CV indication is expected by year-end. Of note, Dr. Kaul was skeptical at CMHC West that FDA will grant J&J’s sNDA request for a CV indication; he presented a Bayesian statistical analysis in which CANVAS data on three-point MACE was no longer significant. In our view, it would be good for patients, for providers, and for public health to add CV benefit to another SGLT-2 label, because this would spread awareness about CV risk in diabetes and about cardioprotective therapy options. Of course, FDA shouldn’t add a CV indication to any medicine unless there’s sound evidence pointing to CV efficacy, but our sense is that a majority of diabetes thought leaders believe in a cardioprotective class effect for SGLT-2s. The DECLARE trial for AZ’s Farxiga will report topline results in 2H18, and VERTIS CV for Merck/Pfizer’s Steglatro is expected to complete in October 2019.

      • Amputations/safety: Counteracting the positive commercial impact of a CV indication for Jardiance is the black box warning on Invokana’s label (added in May 2017). We were worried that after FDA issued this update, and after EMA extended an amputation warning to the entire class, that patients/providers might be more reluctant to start or continue SGLT-2 therapy. We’re glad to see that so far, the negative commercial impact of amputation concerns has been mostly restricted to Invokana; it hasn’t held back impressive sales growth for Jardiance or Farxiga. We’ll be very interested to see if/how amputation concerns affect Steglatro uptake, because the VERTIS pivotal program did find an imbalance in lower limb amputations (11 on ertugliflozin vs. 1 on placebo) and this is included in the label, but FDA has not suggested causation and hasn’t required an official warning. We continue to believe that amputation risk should be highly manageable in the real world with strong patient education, careful patient selection, and diligent monitoring of the feet. To be sure, we see an unmet need for this better education around foot care in diabetes. We’d love to see SGLT-2 manufacturers team up to promote proper foot care.

      • Expanded indications: There is exciting clinical development happening with SGLT-2 inhibitors. These agents are being investigated in type 1 diabetes, in heart failure, and in chronic kidney disease (CKD). AZ has already filed Farxiga for a type 1 indication to EMA (based on positive DEPICT data) and is targeting 2H18 for an FDA submission. Sanofi/Lexicon could enter the class with their SGLT-1/2 dual inhibitor sotagliflozin, submitted to FDA and EMA for type 1 and currently under phase 3 investigation for type 2. After EMPA-REG and CANVAS found significant heart failure benefit with empagliflozin and canagliflozin, respectively, manufacturers launched dedicated studies of SGLT-2s in heart failure (with or without diabetes): This includes Lilly/BI’s EMPEROR HF program with Jardiance and AZ’s Dapa-HF trial with Farxiga. There’s growing recognition in the diabetes field of the immense overlap between diabetes/heart failure, and you can bet we’re enthusiastic about this possible application of SGLT-2s. Lastly, we’re looking forward to readouts on SGLT-2 inhibitors in kidney disease. CREDENCE (Invokana in DKD) is expected to complete in June 2019, Dapa-CKD (Farxiga in CKD, participants with or without diabetes) is expected to complete in November 2020, and EMPA-KIDNEY (Jardiance in CKD, with or without diabetes) is slated to start later this year. Expanded indications would also mean an expanded market for SGLT-2 inhibitors. In our view, getting these advanced products into more patient hands would be a victory, given that they offer glucose-lowering, weight loss, blood pressure reduction, and possible CV/renal protection all in a convenient oral tablet.

    3. DPP-4 Inhibitor Sales Rise 9% YOY, Fall 6% Sequentially to $2.4 Billion; Another Quarter of Fluctuation in the Face of Pricing Pressure

    The DPP-4 inhibitor market grew 9% YOY but fell 6% sequentially to $2.4 billion (vs. $2.2 billion in 1Q17 and $2.5 billion in 4Q17). This sort of fluctuation has come to characterize the class over the past few years. YOY growth each quarter in 2017 was -3%, -4%, +2%, and +7%, consecutively, while sequential growth was -8%, +13%, +1%, and +1%. It’s unusual for a company besides Merck to comment on its DPP-4 inhibitor. For their part, though, Merck management has cited pricing pressure in the US as a key driver of these fluctuations, and in our view, increased competition from GLP-1 agonists and SGLT-2 inhibitors isn’t helping. Interestingly, Merck shared on its 1Q18 call that Januvia + Janumet prescription volume is actually rising (though management didn’t specify by how much), but US sales are pushed down by lower net price, indicating high rebates and patient discounts. Indeed, this dynamic isn’t unexpected in a five-member class with little differentiation between agents (giving PBMs more leverage in designing exclusive contracts), and Merck has a high-quality patient assistance program. Merck’s Januvia remains a dominant force in the market nevertheless: By our calculations, the Januvia franchise captured 60% of the market by value in 1Q18 with $1.4 billion in sales, followed distantly by Lilly/BI’s Tradjenta at 16% value share with an estimated $392 million in revenue and Novartis’ Galvus with 13% value share and $318 million in sales. The remaining 10% of the market was split between AZ’s Onglyza ($129 million) and Takeda’s Nesina ($115 million), each with 5% value share. This breakdown is very in line with past quarters; as with the DPP-4 market overall, small fluctuations in value share are also commonplace. Notably, this analysis assumes that Lilly collects ~36% of Tradjenta franchise revenue (only Lilly’s share of sales, and not BI’s, is reported publically).

    • While we’re certain that DPP-4 inhibitors will continue to play an integral role in diabetes care for years to come – perhaps an even larger role as they go generic over the next few years – our thinking on them also continues to evolve. We’ve heard recent pointed criticism of DPP-4 inhibitors: At ENDO 2018, Dr. Steven Nissen contended that DPP-4s are among the least effective glucose-lowering agents but still are on the high end of the cost spectrum (though they may be quite affordable for many patients – Merck, for instance, offers a $5 copay card). We don’t think DPP-4s alone can match the glycemic efficacy of an SGLT-2 inhibitor, a GLP-1 agonist, or another more advanced therapy, especially for patients with more progressed diabetes (in this sense, AZ’s strategy to deprioritize Onglyza in favor of Farxiga seems highly logical). We’d also point to their neutrality on weight and CV outcomes as a further point of inferiority to GLP-1s and SGLT-2s. That said, DPP-4s are also considered among the most tolerable, if not the most tolerable, glucose-lowering agents, providers are very familiar with them, and they also have a well-established role as treatment for those with renal impairment or who are early in the course of disease. Moreover, we’re keen on fixed-dose combinations of DPP-4 inhibitors with SGLT-2 inhibitors (AZ’s Qtern, Merck/Pfizer’s Steglujan, and Lilly/BI’s Glyxambi), which offer greater efficacy vs. either monotherapy but with similarly mild side-effects, and we’d especially love to see this oral combination used more often and earlier in diabetes.

    Pooled DPP-4 Inhibitor Sales (1Q07-1Q18)

    Financial Highlights

    4. AZ Diabetes Rises 6% YOY to $607 Million; SGLT-2 Farxiga Drives 100% of Growth and Dominates Diabetes Discussion

    On the strength of SGLT-2 inhibitor Farxiga, AZ Diabetes climbed a solid 6% YOY to $607 million in 1Q18, rising from $574 million in 1Q17. One hundred percent of this growth came from Farxiga (+44% YOY), which more than compensated for losses in DPP-4 Onglyza (-16% YOY), GLP-1s Bydureon + Byetta (-15% YOY), and Symlin (-36% YOY). In fact, looking at these sluggish performances throughout AZ’s diabetes portfolio, we wouldn’t be likely to guess that the business grew overall, which speaks to Farxiga’s very impressive 1Q18. The portfolio fell 16% sequentially from $725 million in 4Q17, but we note that this was a challenging comparison; AZ’s total diabetes sales increased 23% sequentially in 4Q driven by Farxiga and by an uncharacteristically positive quarter for Onglyza. The lion’s share of diabetes discussion on the call focused on Farxiga. Management emphasized that Farxiga’s growth exceeded SGLT-2 class growth at +44% YOY; by our calculations, the SGLT-2 inhibitor market climbed 41% YOY worldwide in 1Q18 (see above). We’re certainly glad to see such a strong performance in this segment of AZ’s diabetes business (and a consistently strong one at that, since Farxiga is always a bright spot within this portfolio). On the promise of continued growth for Farxiga (with DECLARE results coming up in 2H18, plus potential for new indications in type 1 diabetes, heart failure, and kidney disease), AZ management continues to be fairly positive about its diabetes business.

    • Farxiga’s $299 million in sales reflected 49% of AZ Diabetes in 1Q18, an increase from 46% in 4Q17 and 36% in 1Q17 – this is telling, though it’s also important to keep in mind that AZ’s GLP-1 and DPP-4 sales have been flat, falling, or fluctuating in recent quarters. Onglyza’s $129 million in 1Q18 revenue reflected 21% of total portfolio revenue, down from 25% in 4Q17 and 27% in 1Q17. Once-weekly GLP-1 agonist Bydureon pulled in $139 million in 1Q18, comprising 23% of portfolio revenue, a decrease from 27% in 1Q17 but an increase from 20% in 4Q17. Twice-daily Byetta continued to decline, posting only $31 million in 1Q18 to make up 5% of profile revenue, below the 8% and 7% it accounted for in 1Q17 and 4Q17, respectively. With $9 million, Symlin held steady at ~2% of profile revenue. On a final note, AZ didn’t break out sales for fixed-dose SGLT-2/DPP-4 inhibitor Qtern (dapagliflozin/saxagliptin), as the company did for the first time in 4Q17 when the product sold $5 million. In our view, this signals uncertainty across the industry about how and when to focus on combination products, and in all likelihood, Qtern sales were very low in 1Q. You can read more on each product in subsequent highlights.

    Total Sales for AZ’s Diabetes Portfolio (4Q12-1Q18)

    5. Farxiga Sales Rise 44% YOY to $299 Million, Falling 10% Sequentially; Management Shares Very Positive Outlook on DECLARE & CV Indication, DEPICT & Type 1 Indication

    Farxiga (dapagliflozin) sales of $299 million grew 44% YOY as reported (+39% YOY operationally) from $207 million in 1Q17, marking another impressive performance for this now-blockbuster drug ($1.1 billion in 2017). Revenue fell 10% sequentially from $332 million in 4Q17, and in this way, Farxiga’s 1Q18 performances matches that of the SGLT-2 inhibitor class overall (+41% YOY, -2% sequentially – see above). By geography, US sales rose 32% YOY and dropped 15% sequentially to $127 million, while OUS sales jumped 55% YOY and only declined 6% sequentially to $172 million. Management alluded to pricing pressure in the US, and we’re not surprised to see a somewhat stronger performance OUS, as dapagliflozin was the first-to-market SGLT-2 in Europe and dominates there. Still, AZ highlighted double-digit growth in the Farxiga franchise across all geographies, including the US. In emerging markets, revenue climbed 62% YOY to $69 million, and in Europe, climbed 30% YOY to $74 million. Management explained that they missed the cutoff for Farxiga to be included on China’s National Drug Reimbursement List, but underscored good progress in securing reimbursement in the provinces. Farxiga was a major talking point on the call. We note that it was the lone bright spot of AZ’s diabetes portfolio in 1Q18, driving 100% of growth; all other products experienced YOY loss, so it speaks to remarkable strength in Farxiga that the portfolio grew at all, rising 6% YOY. Management positioned the SGLT-2 inhibitor as a very important product and growth driver for AZ’s overall pharmaceutical business. We have no doubt that the company is exceptionally committed to SGLT-2. If we needed any more reassurance, we heard significant discussion from management today on dapagliflozin clinical development – much more detail than we’re accustomed to getting during a major pharma company’s earnings call, which was amazing!

    Farxiga Franchise Sales (1Q13-1Q18)

    • We sensed a lot of confidence from management in upcoming DECLARE results, and they confirmed that a topline release will be issued in 2H18. We were hoping for full results, and it seems like we may have to wait until early 2019, though management hasn’t been all that specific about this timing. Dr. Itamar Raz, a DECLARE investigator, shared at CODHy that he hopes the paper is published by end of 2018. Management discussed the CVOT’s co-primary endpoints – one is the classic three-point MACE (non-fatal MI, non-fatal stroke, or CV death) and the other is a composite of hospitalization for heart failure/CV death – touching on the potential of Farxiga to protect against heart failure. Indeed, thought leaders have expressed distinct enthusiasm for SGLT-2s in heart failure; Dr. Subodh Verma has referred to these drugs as a “sweet spot” in heart failure management due to their multifaceted effects on glucose, blood pressure, and body weight (not to mention possible renal protection). According to AZ’s presentation (slide 31), the company plans to submit DECLARE data to FDA for inclusion on the Farxiga label in 2019.

    • Management reviewed CVD-REAL 2 results, presented by Dr. Mikhail Kosiborod at ACC 2018. They called attention to the fact that 75% of SGLT-2 exposure in this real-world study was to dapagliflozin (fewer patients were on empagliflozin, canagliflozin, ipragliflozin, tofogliflozin, and luseogliflozin). The implication here is that Farxiga may have driven the positive findings, and that there’s no reason to believe dapagliflozin won’t be as cardioprotective as empagliflozin and canagliflozin – in other words, this bodes well for DECLARE. Management also noted that treatment with an SGLT-2 inhibitor in CVD-REAL 2 lowered risk for stroke (HR=0.68, 95% CI: 0.55-0.84, p<0.001) and MI (HR=0.81, 95% CI: 0.74-0.88, p<0.001), in addition to all-cause mortality and hospitalization for heart failure. These atherosclerotic CV events weren’t analyzed in the first CVD-REAL, also presented by Dr. Kosiborod, at ACC 2017. We heard much discussion of CVD-REAL 2 at CODHy Asia Pacific in Shanghai last week, because the study included Asian populations from Japan, South Korea, and Singapore normally under-represented in RCTs and real-world trials, and because it showed a beneficial effect of SGLT-2s on stroke events, which trended in the wrong direction in EMPA-REG OUTCOME (favoring placebo over Jardiance, but not reaching statistical significance).

    • AZ mentioned several times that Forxiga has been submitted to EMA for a type 1 diabetes indication. The company is planning a parallel FDA filing for 2H18, and both submissions are based on DEPICT results. A decision for the EU market is expected in 1Q19, and if all proceeds according to plan, FDA should issue its decision by end of 2019. By 2020, we could have SGLT inhibitors on the market for type 1 diabetes, and we see this as a major win for patients. Many type 1s are already taking SGLT-2 inhibitors off-label for the extra glucose-lowering, weight loss, insulin dose reduction, and most notably, time-in-range benefit. We’re eager for this practice to become safer, because we do not deny that DKA is a significant and costly risk, devastating for quality of life and expensive for the health system when patients are hospitalized. AZ management didn’t comment on DKA risk; Lexicon has perhaps been the most vocal about this, with regard to its Sanofi-partnered SGLT-1/2 dual inhibitor sotagliflozin, also submitted to FDA and EMA for a type 1 indication. Ultimately, we think these manufacturers should be working together (to the extent possible) to promote education so that real-world patients/providers are well-versed in minimizing, recognizing, and managing DKA. We’ll be watching SGLTs for type 1 closely, especially now that they’ve reached regulatory desks. Lilly has also promised phase 3 data on Jardiance in type 1 (EASE-2 and EASE-3) sometime this year.

    • While not mentioned on the call or in any of AZ’s prepared materials, we note that the DERIVE study of Farxiga in type 2s with stage 3A CKD read out at ENDO 2018. Results showed that dapagliflozin was safe and effective in patients with eGFR as low as 45 ml/min/1.73m2, and we wonder if AZ will seek a label update, since Farxiga is currently not recommended for anyone with eGFR <60 ml/min/1.73m2. The company has also launched the Dapa-CKD trial probing for Farxiga’s renal protective effects in patients with chronic kidney disease, with or without diabetes. Dapa-CKD is expected to complete in November 2020. See the table below for more detail on ongoing clinical trials of dapagliflozin.

    Ongoing Farxiga Clinical Trials

    Trial/Indication

    Estimated Enrollment

    Timeline

    DECLARE (cardiovascular outcomes)

    17,276

    Expected to complete July 2018; Data anticipated in 2H18

    Dapa-HF (heart failure with reduced injection fraction)

    4,500

    Initiated February 2017 and expected to complete December 2019

    Dapa-CKD (chronic kidney disease)

    4,000

    Initiated February 2017 and expected to complete November 2020

    DEPICT 2 (type 1 diabetes)

    768

    Topline results available in house as of 4Q17 (along with 52-week DEPICT 1 results); Study completed in April 2018; DEPICT 1 (n=833) reported positive 24-week data at EASD 2017

    DELIGHT (Farxiga vs. Qtern in people with diabetes/DKD)

    460

    Completed April 2018

    DEFINE-HF (heart failure biomarkers in people with comorbid diabetes/heart failure)

    250

    Expected to complete September 2018

    PRESERVED-HF (heart failure biomarkers in people with prediabetes or diabetes alongside chronic heart failure)

    320

    Expected to complete March 2019

    DAPASALT (evaluating 24-hour sodium excretion following dapagliflozin treatment)

    51

    Expected to complete August 2018

    6. Bydureon Falls 9% YOY to $139 Million, but Management Shares Optimism in BCise Launch (Leads to Uptick in New-to-Brand Bydureon Prescriptions); EXSCEL Data Submitted to EMA for Label Inclusion

    Sales of Bydureon (exenatide once-weekly) dropped 9% YOY as reported (-11% YOY operationally) to $139 million, compared to $153 million in 1Q17. Sequentially, revenue fell 5% from $147 million in 4Q17. While we’re disappointed that the new BCise autoinjector has yet to translate to an improved commercial performance for AZ’s long-acting GLP-1, management maintained that the launch has been encouraging to-date. A graph in AZ’s presentation (slide 10) does illustrate a jump in new-to-brand prescriptions for Bydureon, driven by BCise, at the start of 2018. According to AZ, roughly half of patients on BCise are new GLP-1 agonist patients, indicating that the product is possibly even contributing to class growth. About one-third of patients now on BCise switched from the Bydureon dual chamber pen – this is in line with what we’d expect. It’s still early for BCise, and we expect the autoinjector to make more meaningful waves in future quarters. On the whole, Bydureon sales have been stagnant for the past two years, falling 1% YOY in 2017 and <1% YOY in 2016 after 27% YOY growth in 2015, which pales in comparison to the exceptionally strong underlying class growth for GLP-1s (see above for our pooled analysis of the class). In many ways, we think AZ’s GLP-1 business was in-need of a more patient-friendly device/better overall product, and we hope to see steeper franchise growth as BCise gains traction. An EMA decision on the Bydureon autoinjector is expected in 2H18.

    • By geography, US sales of Bydureon fell 13% YOY to $111 million, from $127 million in 1Q17. Sequentially, US revenue dropped 4% from $115 million in 4Q17. OUS revenue partially offset this loss: Sales rose 8% YOY to $28 million from $26 million in 1Q17, and management cited growth in Europe in particular. This OUS revenue does represent a 13% sequential loss, though this was a challenging comparison as sales rose 14% sequentially in 4Q17 to $32 million.

    Bydureon Sales (3Q12-1Q18)

    • During prepared remarks, management highlighted the April FDA approval of Bydureon as an add-on to basal insulin. In DURATION-7 (n=461), the addition of Bydureon to insulin glargine therapy gave a mean A1c reduction of 1.0% at 28 weeks, vs. 0.3% with placebo (baseline A1c 8.5%, p<0.01). Participants on Bydureon were 25% more likely to reach an A1c <7%, and 20% more likely to achieve A1c <7% without weight gain or severe hypoglycemia. Prior to this approval, the Bydureon label recommended against adding the GLP-1 on top of basal insulin, and the update will allow AZ reps to promote Bydureon for patients already on basal insulin. It also brings Bydureon more in line with other leading GLP-1 agonists: Lilly’s Trulicity received a similar update in February 2017, and Novo Nordisk’s Victoza and Ozempic are both approved for use with basal insulin. The EMA-approved label for Bydureon already reflects safety with basal insulin.

    • AZ’s press release indicates that the company has submitted EXSCEL CVOT results to the EMA (page 45), and that a similar FDA filing is slated for 2Q18. As a reminder, full EXSCEL results presented at EASD 2017 established CV safety but not efficacy for Bydureon, missing the threshold for superiority by a razor-thin margin (p=0.06). These submissions are almost definitely to include CV safety data on the EU and US labels for Bydureon, but we’ll still be curious to see precisely what language is approved by EMA and FDA. Will it resemble the language describing SUSTAIN 6 as a safety trial on the Ozempic label?

      • We continue to hear KOL commentary suggesting that EXSCEL was a positive study that appears neutral because of elements of pragmatic study design. Earlier this week at AACE 2018, Dr. Ralph DeFronzo argued that the use of reconstitution kits rather than pens or the BCise autoinjector likely led to poor adherence and dropout rates. We’ve also heard remarks focused on the trial’s larger primary prevention population and real-world design from leaders including Dr. Rury Holman and Dr. Francesco Giorgino. In the end, though, it’s hard not to imagine these results hindering Bydureon’s commercial success, no matter how convinced we are in CV benefit as a GLP-1 class effect. This class is becoming increasingly competitive and differentiated based on cardioprotection, and AZ is limited with Bydureon on that front, although the company does have patient-friendliness in tow now that BCise is available on the market. Still, we imagine the technically neutral EXSCEL results may shift AZ’s focus/investment even more toward SGLT-2 inhibitor Farxiga (see above).

    • Sales of Byetta (exenatide twice-daily) dropped 33% YOY and 35% sequentially to $31 million, from $46 million in 1Q17 and $48 million in 4Q17. This decline was driven entirely by a 50% YOY fall in US revenue, from $30 million in 1Q17 to $15 million in 1Q18. Meanwhile, OUS revenue was stagnant at $16 million. We expect Byetta is losing sales to both Bydureon and other GLP-1 agonists on the market (Victoza, Trulicity, and now Ozempic). Revenue for AZ’s exenatide franchise (Bydureon + Byetta) totaled $170 million in 1Q18, representing a 15% YOY decline from $199 million in 1Q17.

    Byetta Sales (3Q12-1Q18)

    7. Onglyza Revenue Dives 16% YOY, 28% Sequentially to $129 Million as AZ Prioritizes Farxiga, Cites GLP-1 and SGLT-2 Competition

    Sales of DPP-4 inhibitor Onglyza (saxagliptin) fell 16% YOY and 28% sequentially to $129 million, dropping from $154 million in 1Q17 and $180 million in 4Q17. This lackluster performance follows an uncharacteristically strong 4Q17, when Onglyza broke a six-quarter streak of YOY declines; that said, 21% YOY growth in 4Q17 came against an easy comparison of 22% YOY decline in 4Q16. On an annual basis, Onglyza sales have been falling since 2014, and AZ management has been explicit since as least 2Q16 that they are deprioritizing Onglyza while increasing support for SGLT-2 Farxiga. In fact, management has gone so far as to say that they’re willing to let Farxiga cannibalize some Onglyza sales, because they’re interested in positioning their SGLT-2 inhibitor as initial therapy and as the preferred choice for any patient switching over from a DPP-4 inhibitor. AZ’s 1Q18 announcement cites several sources of adverse pressure on DPP-4s, including (i) pricing pressure and (ii) increased movement of patients toward classes with proven CV benefit, i.e. SGLT-2 inhibitors and GLP-1 agonists. These market dynamics are not entirely dissimilar to what we’ve seen with other DPP-4 inhibitors, but where class-wide revenue is fluctuating, Onglyza seems to be on a relatively consistent decline. Unlike other DPP-4s, saxagliptin showed a worrisome signal for increased heart failure hospitalizations in the SAVOR-TIMI CVOT, and we think concern lingers among many HCPs. Moreover, while some thought leaders maintain that DPP-4 inhibitors have a role to play in those with renal impairment or earlier in the course of disease, the majority opinion, in our perception, is that their inferior glucose-lowering efficacy combined with a lack of CV or weight loss benefit make them a subpar option vs. GLP-1s and SGLT-2s. For a deeper dive on DPP-4 inhibitors overall, scroll up to our pooled market highlights.

    • By geography, US Onglyza revenue plummeted 40% YOY and 52% sequentially to $49 million in 1Q18, falling from $81 million in 1Q17 and $103 million in 4Q17. Outside the US, Onglyza fared better, rising 10% YOY and 4% sequentially to $80 million, up from $73 million in 1Q17 and $77 million in 4Q17. However, sales only actually grew in emerging markets, where revenue climbed 33% YOY to $40 million, a performance attributed to Onglyza’s entry onto China’s National Reimbursement Drug List in 3Q17. Sales in China more than doubled YOY to $16 million in 1Q18, compared to a 15% YOY decline in Europe to $23 million.

    Onglyza Sales (4Q12-1Q18)

    8. No Qtern Sales Reported, Implying Continued Sluggish Uptake of Fixed-Dose Combination (Farxiga + Onglyza)

    Somewhat strangely, AZ broke out Qtern sales for the first time in 4Q17 but did not continue this in 1Q18. Global revenue from the fixed-dose combination (SGLT-2 dapagliflozin + DPP-4 saxagliptin) totaled $5 million in 4Q17, with the majority of sales ($4 million) coming from the US market. This was an underwhelming financial performance, even for a very new product. We expect that Qtern uptake is still extremely low, which would explain why AZ hasn’t reported the product’s revenue separately in its 1Q18 materials. We have high hopes for fixed-dose and fixed-ratio combination products, because of the convenience they offer, not to mention superior efficacy, milder side-effects, and greater adherence compared to co-administration of the component monotherapies. And yet, we’ve been continuously disappointed by sluggish uptake of Qtern and Glyxambi (Lilly/BI’s empagliflozin + linagliptin), of Novo Nordisk’s Xultophy (insulin degludec + liraglutide) and Sanofi’s Soliqua (insulin glargine + lixisenatide). To be sure, pricing and reimbursement are a major barrier, in that prices are too high and current reimbursement prospects are poor. For example, we learned at AACE this week that patients can get Merck’s SGLT-2 inhibitor Steglatro (ertugliflozin) and DPP-4 inhibitor Januvia (sitagliptin) together for ~$5/month with copay cards, but Steglujan (ertugliflozin/sitagliptin fixed-dose) would cost the same patients ≥$60/month out of pocket. In other words, co-administration is simply a more affordable option than fixed-dose combination. We’re curious to know if similar factors are at play with AZ’s Qtern and with Lilly/BI’s Glyxambi, because none of these products have really taken off commercially. We expect reimbursement for combination products to improve over time (along with attitudes), and meanwhile, we think medical education efforts (e.g. Sanofi’s peer-to-peer education campaign for Soliqua) will be key, to make sure HCPs are comfortable with the concept of fixed-dose/fixed-ratio combination therapy.

    Pipeline Highlights

    9. AZ Adds New Type 2 Diabetes Candidate to Phase 1 – Undisclosed Mechanism of Action

    In 1Q18, AZ added a new large molecule to its phase 1 pipeline for type 2 diabetes. Very little information has been disclosed on MEDI7219, but we found a phase 1 study (n=126 healthy volunteers) on ClinicalTrials.gov that’s expected to complete in September. From the name, we’re guessing this candidate was advanced into clinical-stage development from MedImmune, AZ’s biologics research arm. The company’s 1Q18 call offered no other updates on the diabetes pipeline (we discuss news on Farxiga clinical trials above). See our pipeline summary table below.

    AZ Diabetes Pipeline Summary and Major Ongoing Clinical Trials

    The table below reflects the latest updates, as far as we are aware, on AZ’s diabetes pipeline candidates. Rows highlighted in yellow indicate notable changes to the pipeline in 1Q18.

    Product

    Details

    Status

    Timeline

    Bydureon BCise (exenatide)

    Autoinjector that eliminates need for reconstitution

    Launched in the US; Under review with EMA

    FDA-approved in 3Q17 with US launch in 1Q18; Submitted to EMA in 3Q17 with decision expected 2H18

    Saxagliptin + dapagliflozin + metformin

    Fixed-dose combination for type 2 diabetes

    Phase 3

    FDA/EMA filing planned for 1H18

    Farxiga (dapagliflozin)

    Indication for type 1 diabetes

    Phase 3

    Submitted to EMA, decision expected 1Q19; FDA filling planned for 2H18; 52-week DEPICT 1 data and 24-week DEPICT 2 data reported internally in 4Q17; Positive 24-week DEPICT 1 results presented at EASD 2017; DEPICT 2 completed April 2018

    Farxiga (dapagliflozin)

    Cardiovascular outcomes data

    Phase 3

    DECLARE CVOT expected to report data 2H18

    Farxiga (dapagliflozin)

    Outcomes data in patients with chronic heart failure with reduced ejection fraction

    Phase 3

    Dapa-HF initiated in February 2017, expected to complete December 2019; PRESERVED-HF evaluating heart failure biomarker (not outcomes) expected to complete March 2019

    Farxiga (dapagliflozin)

    Outcomes data in patients with CKD

    Phase 3

    Dapa-CKD initiated in February 2017, expected to complete November 2020; DERIVE results presented at ENDO 2018 showing clinically-meaningful and statistically significant A1c reduction in people with type 2 diabetes/CKD stage 3A (eGFR down to 45 ml/min/1.73m2)

    Brilinta (ticagrelor)

    P2Y12 receptor agonist antiplatelet agent

    Phase 3

    THEMIS outcomes study in patients with type 2 diabetes and coronary artery disease expected to complete November 2018; Data expected 2019

    roxadustat

    2-OG inhibitor for anemia in chronic kidney disease or end-stage renal disease

    Phase 3

    FDA filing planned for 2H18; Approved in China

    MEDI0382

    GLP-1/glucagon dual agonist

    Phase 2

    Phase 2 dose-ranging study in people with type 2 diabetes and overweight/obesity expected to complete March 2020

    AZD4076

    Anti-miR103/107 oligonucleotide for NASH

    Phase 2

    Partnered with Regulus Therapeutics; Phase 1/2 trial in type 2 diabetes/NASH expected to complete November 2018

    verinurad

    Uric acid reabsorption inhibitor (URAT-1) for CKD in type 2 diabetes

    Phase 2

    Phase 2 trial expected to complete August 2018

    MEDI7219

    Undisclosed

    Phase 1

    Added to pipeline in 1Q18; Phase 1 study expected to complete September 2018

     

    -- by Ann Carracher, Payal Marathe, and Kelly Close