AZ 4Q17 – SGLT-2 class grows 24% YOY to $3.5 billion in 2017; DPP-4 market flat at $9.7 billion in 2017; Diabetes portfolio +3% YOY in 2017 and +21% YOY in 4Q17; Farxiga hits blockbuster status; Muted Bydureon performance – February 6, 2018

Executive Highlights

  • AZ’s overall diabetes portfolio grew 21% YOY in 4Q17 and 3% YOY in 2017, reaching $725 million (up from $598 million in 4Q16) and $2.5 billion (up from $2.4 billion in 2016), respectively. After falling 1%, 7%, and 3% YOY in the first three quarters of 2017, it was shaping up to be a soft year for the diabetes franchise – until 4Q17. SGLT-2 inhibitor Farxiga was the primary growth driver in both the quarter and the year: Product sales rose 39% YOY to $332 million in 4Q17, reflecting 46% of total portfolio revenue and accounting for a 69% share of portfolio growth. In 2017, product sales rose 29% YOY to $1.1 billion, hitting blockbuster (!) status – this reflected 43% of total revenue and accounted for a whopping 95% share of growth.
  • On a pooled basis, SGLT-2 inhibitor sales grew 24% YOY in 2017 to $3.5 billion (compared to $2.9 billion in 2016). Pooled revenue in 4Q17 broke $1 billion, rising 23% YOY and 11% sequentially (from $841 million in 4Q16 and $935 million in 3Q17). Lilly/BI’s Jardiance was the clear market leader by value in 4Q17, with $436 million sales reflecting 42% of pooled revenue. With $339 million sales, Farxiga held a 32% market share by value, and with $267 million sales, J&J’s Invokana held 26%. As a sign of how these dynamics have changed over the past year, market share by value in 4Q16 was 44% for Invokana in first place, 28% for Farxiga in second place, and 27% for Jardiance in third place – this has essentially flipped in 12 months. While Jardiance had a stellar 2017, revenue more than doubling YOY for an estimated $1.4 billion, Invokana served as its foil, falling 21% YOY to $1.1 billion – J&J management acknowledged share loss to Lilly/BI.
  • On a pooled basis, DPP-4 inhibitor sales grew 7% YOY in 4Q17 to $2.5 billion (vs. $2.4 billion in 4Q17). The market was flat sequentially. Full year 2017 pooled sales were flat YOY at $9.7 billion. By our estimates, pooled revenue from all SGLT-2s and GLP-1s will sum to ~$2.9 billion in 4Q17, which is greater than the $2.5 billion in total sales from all major DPP-4 products. Annualized, the two new classes could post >$10 billion, which also surpasses the $9.7 billion from DPP-4s in 2017. We include this analysis to show that while there’s no doubt in our minds that DPP-4 inhibitors will continue to serve an important purpose in diabetes care into the foreseeable future, competition from advanced therapy classes is also impossible to ignore.
  • On the pipeline front, several Farxiga outcomes studies will report in 2018, including the DECLARE type 2 diabetes CVOT (in 2H18), DERIVE investigating dapagliflozin in comorbid diabetes/kidney disease, and DEPICT 2 in type 1 diabetes (not to mention 52-week results from DEPICT 1, which was originally presented at EASD 2017). AZ plans to submit Farxiga to FDA for a type 1 diabetes indication in 2H18. The company is also developing a triple combination of dapagliflozin + saxagliptin + metformin, which plans to file with FDA in 1H18.

AZ announced its 4Q17 and 2017 financial results early Friday morning in a call led by CEO Mr. Pascal Soriot. SGLT-2 inhibitor Farxiga gave a strong performance to close out 2017, and reached blockbuster status for the year. We were also excited to see the first Qtern sales (however small!) broken out in financial tables. To get a first-hand take, check out AZ’s press release, slide deck, and clinical trials appendix.

This full report is expanded from the First Look we published on Friday. It starts with a section of pooled market highlights covering the SGLT-2 class and the DPP-4 class. We provide much more detail on each of AZ’s major diabetes products (the company has multiple exciting clinical trials ongoing for Farxiga). We also include illustrative graphs to show sales trends, and we end this piece with a comprehensive pipeline summary table and relevant Q&A from the call.

Now’s as good a time as ever to download the Closer Look app if you haven’t already! See our earnings reports as they’re posted to get all the essentials.

Table 1. Full Year 2017 Financial Results for AZ’s Major Diabetes Products


2017 Revenue (millions)

Year-Over-Year Reported (Operational) Growth

Share of Growth



29% (28%)








-1% (-1%)




-31% (-30%)




-15% (-16%)




20% (20%)


Total Diabetes




Table 2. 4Q17 Financial Results for AZ’s Major Diabetes Products


4Q17 Revenue (millions)

Year-Over-Year Reported (Operational) Growth

Sequential Reported Growth

Share of Growth



39% (37%)










4% (2%)





-13% (-13%)





21% (19%)








Total Diabetes





Table of Contents 

Pooled Market Highlights

1. SGLT-2 Inhibitor Sales Rise ~24% YOY to $3.5 Billion Total in 2017, >$1 Billion in 4Q17; CV Benefit Balanced by Amputation Concerns; Jardiance “Stealing Share” from Invokana; Class Will Soon Become Four-Product Market with Merck/Pfizer’s Steglatro

Pooled SGLT-2 inhibitor sales from Lilly/BI, J&J, and AZ totaled $3.5 billion in 2017, rising 24% YOY from $2.9 billion in 2016 (which is in line with our most recent forecast). Notably, pooled sales in 4Q17 broke $1 billion, rising 23% YOY and 11% sequentially (from $841 million in 4Q16 and $935 million in 3Q17). While this is an impressive showing for the quarter and year, we do point out that SGLT-2 market growth in 2017 has consistently been in the low double-digits (+18% YOY in 1Q17, +20% YOY in 2Q17, +35% YOY in 3Q17), muted in comparison to YOY growth in the high 40s for most of 2016 (+47% in 1Q16, +49% in 2Q16, +33% in 3Q16, +48% in 4Q16, +44% in 2016 overall). Following groundbreaking results from EMPA-REG OUTCOME and the new CV indication for Lilly/BI’s Jardiance (empagliflozin), we expected a steeper climb from this advanced therapy class. But there have been headwinds as well, namely the boxed warning for lower limb amputations added to the Invokana (J&J’s canagliflozin) label.

  • In 4Q17, Jardiance was the clear market leader by value with $436 million sales reflecting 42% of pooled revenue (note that this analysis is based on our estimate for total global sales from the Jardiance franchise, since only Lilly’s portion and not BI’s is reported to the public). With 39% YOY growth to $339 million in 4Q17, AZ’s Farxiga (dapagliflozin) held a 32% market share by value, and with $267 million sales, J&J’s Invokana held 26%. As a sign of how these class-wide dynamics have changed over the past year, market share by value in 4Q16 was 44% for Invokana in first place, 28% for Farxiga in second place, and 27% for Jardiance in third place – this has essentially flipped in only 12 months. While Jardiance had a stellar year, revenue more than doubling YOY for an estimated $1.4 billion in 2017, Invokana served as its foil, falling 21% YOY to $1.1 billion – J&J management acknowledged share loss to Lilly/BI. Farxiga had its first blockbuster year in 2017, growing 29% YOY to $1.1 billion, and keeping pace with overall class growth. As has been the trend for SGLT-2s, Jardiance was the primary growth driver in 4Q17 and 2017, while Farxiga contributed positively to growth and Invokana detracted. Notably, Farxiga leads in volume, with 41% of total prescriptions globally, followed by 35% for Jardiance and 17% for Invokana – a marked shift from the 40%/30%/22% split in 2Q17, which once again sends the message that Jardiance is “stealing share” from Invokana while Farxiga holds its ground and grows steadily.
  • Adding to the pushes and pulls on the SGLT-2 inhibitor market, Merck/Pfizer will launch the Steglatro (ertugliflozin) franchise in 1Q18 – we’re looking forward to writing about a four-product class in three months’ time. The two companies have developed a trio of products – standalone Steglatro, fixed-dose combination Steglujan (ertugliflozin/sitagliptin), and fixed-dose combo Segluromet (ertugliflozin/metformin) – that were all FDA-approved in December 2017. These drugs also received a positive CHMP opinion in January, and final marketing authorization for Europe is expected in 1H18. Merck/Pfizer have opted for a lower list price on Steglatro, Steglujan, and Segluromet vs. competitors. In the US, the companies announced a list price for Steglatro at $8.94/day, considerably lower than the average list price for other SGLT-2 inhibitors (~$17/day). For Steglujan, the companies have listed a price of $17.45/day vs. ~$22/day for Lilly/BI’s Glyxambi (empagliflozin/linagliptin) and ~$24/day for AZ’s Qtern (dapagliflozin/saxagliptin). While list prices only mean so much without accounting for patient discount programs and reimbursement, we imagine the lower cost of Merck/Pfizer’s SGLT-2 products will be attractive to patients, HCPs, and payers alike. We’re optimistic that Steglatro can grow the entire SGLT-2 inhibitor class, especially because there’s ample head room. According to a recent Diabetes Care article, a mere ~7% of second-line diabetes prescriptions in the US go to an SGLT-2 inhibitor (a majority still go to a sulfonylurea) – the market is so far from saturated when it comes to this advanced class of oral agents. Moreover, Lilly management suggested on the company’s recent 4Q17 earnings call that it will benefit all SGLT-2 manufacturers to have another two players promoting the class (to patients, HCPs, and payers). We certainly agree, and we’d add that Merck has extensive experience in diabetes with DPP-4 inhibitor Januvia (sitagliptin), and could be very successful commercializing Steglatro and its sister products.
  • AZ’s DECLARE study will report results in 2H18, becoming the third complete SGLT-2 inhibitor CVOT. This will be an important milestone for the class, with potential to demonstrate a very compelling cardioprotective class effect. What’s more is that DECLARE enrolls a larger primary prevention cohort than either EMPA-REG OUTCOME (for Jardiance) or CANVAS (for Invokana). As AZ management alluded to during Q&A, positive DECLARE results could expand the “breadth” of the market for Farxiga, as these findings could support a CV indication on the label that applies to a broader spectrum of real-world patients. In contrast, Jardiance is FDA-approved for the reduction of CV death only in type 2 diabetes patients with established CV disease. Some thought leaders have also questioned the generalizability of CVOT results given high-risk study populations, and DECLARE could (start to) address this issue. We continue to push for more standardization of CVOT design overall, so that the collection of studies can actually inform personalized treatment decisions in real-world clinical practice. That said, we’re excited to see if Farxiga and SGLT-2s show promise in primary CV prevention through DECLARE. AZ’s CVOT also includes heart failure in a co-primary endpoint. Can SGLT-2 inhibitors show CV benefit, particularly in heart failure risk reduction, in a lower-risk population? A post-hoc analysis of CANVAS found significant heart failure benefit in the primary prevention cohort, even though the benefit on three-point MACE did not reach statistical significance in this subgroup, which has led some experts (e.g. Dr. Itamar Raz at WCTD 2017) to predict that DECLARE will show significant findings on the primary composite endpoint that includes heart failure, even with a much larger primary prevention cohort.
  • SGLT-2 inhibitors are currently being investigated for a handful of additional indications outside diabetes as well, which could stimulate further class growth. In 2018, Lilly/BI will report phase 3 EASE results on Jardiance in type 1 diabetes. AZ has already presented DEPICT 1 results on Farxiga in type 1 (at EASD 2017) and will follow-up this year with findings from the DEPICT 1 trial extension out to 52 weeks and from DEPICT 2. According to 4Q17 presentation materials, AZ will file Farxiga for a type 1 diabetes indication in 2H18. The markets for heart failure and chronic kidney disease (CKD) are even bigger than the market for type 1, and to this end, studies are ongoing to assess SGLT-2 agents in these patient populations (in people with or without diabetes). The Dapa-HF trial for Farxiga is expected to complete in December 2019, while the EMPEROR HF-Preserved and EMPEROR HF-Reduced trials for Jardiance are expected to complete in June 2020. J&J’s CREDENCE study investigating Invokana in people with type 2 diabetes and comorbid kidney disease should wrap up in June 2019, while the Dapa-CKD study for Farxiga is scheduled to complete in November 2020, and Lilly/BI’s dedicated outcomes study for Jardiance in CKD (participants with or without diabetes) is scheduled to begin sometime in 2018. While we find it somewhat excruciating to wait for these results, we’re equally excited for the impact they could have: Treating patients with renal impairment remains a challenge in the real world, and thought leaders have suggested that SGLT-2 inhibitors could be the next wave of CKD therapy (for example, see our coverage of Dr. David Fitchett’s talk at ESC 2017). SGLT-2 inhibitors are getting a lot of attention at diabetes and cardiology conferences alike for their impact on heart failure outcomes, and Drs. Subodh Verma, John McMurray, and David Cherney published a JAMA Viewpoint in September 2017 positioning SGLT-2s as the “sweet spot” for heart failure management.  

Figure 1. Pooled SGLT-2 inhibitor Sales (1Q13-4Q17)


2. DPP-4 Inhibitor Market Remains Relatively Flat at $9.7 Billion in 2017, $2.5 Billion in 4Q17; Merck’s Januvia Maintains Frontrunner Status with ~60% Market Share by Value

On a pooled basis, the DPP-4 inhibitor class grew 7% YOY in 4Q17, posting $2.5 billion in revenue (vs. $2.4 billion in 4Q17). The market was flat sequentially. Full year 2017 pooled sales were flat YOY at $9.7 billion. We’ve come to expect small fluctuations in the DPP-4 market (-3% YOY in 1Q17, -4% YOY in 2Q17, +2% YOY in 3Q17) as these agents face pricing pressure and increasing competition from SGLT-2 inhibitors and GLP-1 agonists. By our calculations, Merck captured 60% of this market by value with $1.5 billion from Januvia and Janumet; this market share is on par with previous quarters (61% in 1Q17-3Q17, 62%-63% throughout 2016). Lilly/BI’s Tradjenta (linagliptin) was in a distant second place, reflecting 14% of pooled sales ($361 million). With $327 million in revenue, Novartis’ Galvus (vildagliptin) reflected 13% of pooled sales. With $180 million in revenue, AZ’s Onglyza (saxagliptin) reflected 7% of pooled sales. Lastly, with $137 million in 4Q17 revenue, Takeda’s Nesina (alogliptin) reflected 5% of pooled sales. In general, these market dynamics aren’t changing all that dramatically or quickly (e.g. Tradjenta’s market share by value has been between 13%-17% throughout 2017).

  • There’s no doubt in our minds that DPP-4 inhibitors will continue to serve an important purpose in diabetes care into the foreseeable future. That said, the competition from advanced therapy classes is impossible to ignore. By our estimates, pooled revenue from all SGLT-2s and GLP-1s will sum to ~$2.9 billion in 4Q17, which is greater than the $2.5 billion in total sales from all major DPP-4 products. Annualized, the two new classes could post >$10 billion, which also surpasses the $9.7 billion from DPP-4s in 2017. To be sure, DPP-4 inhibitors still comprise a substantial market within diabetes care. These agents are highly-familiar within the diabetes community, and they offer patients decent A1c-lowering alongside strong safety/tolerability. Some thought leaders, including Dr. Robert Ratner, have defended this class for the elderly and for patients with renal impairment. On the other hand, where DPP-4 inhibitors are weight neutral, SGLT-2s and GLP-1s produce meaningful weight loss; where DPP-4 inhibitors are CV safe, SGLT-2s and GLP-1s are cardioprotective. Given this, other experts including Drs. Jay Skyler and Steven Nissen have advocated for the de-prioritization of DPP-4s in treatment guidelines, in favor of SGLT-2 inhibitors and GLP-1 agonists. We imagine this is the direction the field is heading, but for the time being, we hope to see DPP-4 inhibitors used earlier in the course of diabetes development and widely-used among patients who could stand to benefit, especially considering they are somewhat more accessible/affordable than the newer classes.

Figure 2. Pooled DPP-4 Inhibitor Sales (1Q07-4Q17)

Financial Highlights

3. AZ’s Overall Diabetes Portfolio Grows 21% YOY in 4Q17 to $725 Million, Primarily Driven by SGLT-2 Farxiga; Full Year Sales Rise 3% to $2.5 Billion

A strong 4Q17 and a particularly impressive performance from SGLT-2 inhibitor Farxiga buoyed an otherwise lackluster year for AZ Diabetes. The portfolio grew 3% YOY in 2017 to $2.5 billion, up from $2.4 billion in 2016. After falling 1%, 7%, and 3% YOY in the first three quarters of 2017, it was shaping up to be a soft year for the diabetes franchise – until 4Q17 brought 21% YOY growth for $725 million in total sales. By our calculations, Farxiga sales of $332 million reflected 46% of total diabetes revenue and drove 69% of portfolio growth (the SGLT-2’s share of growth was even more remarkable in 2017 overall, at 95% vs. 43% of revenue). DPP-4 inhibitor Onglyza posted $180 million in 4Q17, reflecting 25% of total revenue and driving 23% of growth. Prior to the last quarter of 2017, it was a year of rather sluggish sales for Onglyza, so we see this performance as particularly positive and important in allowing AZ’s diabetes business to return to double-digit YOY growth. Fixed-dose combination Qtern (4%) and once-weekly GLP-1 agonist Bydureon (4%) drove the remaining portfolio growth in 4Q17, reflecting <1% and 20% of total diabetes revenue, respectively. AZ broke out Qtern (dapagliflozin/saxagliptin) for the first time in 4Q17. US launch was very low-key – this was the first indication, for us, that the combo was available in US pharmacies – and the product pulled in $5 million worldwide ($4 million in the US). Historically, AZ has been quite positive about the combo; management’s extended silence made us skeptical of their commitment to Qtern, although we imagine there could be more commentary from here on out as there’s real “news” on covered lives, total prescription volume, sales growth, etc. On the year, only Symlin (+20% to $48 million total) and Farxiga (+29% to $1.1 billion total) grew, balancing a 15% YOY drop in Onglyza, a 30% drop in Byetta, and a 1% drop in Bydureon. AZ management was fairly optimistic about its overall business, stating a focus on “returning to growth” and pointing to momentum in product sales as well as a strong flow through the pipeline.

Figure 3. Total Sales for AZ’s Diabetes Portfolio (4Q12-4Q17)

4. SGLT-2 Farxiga Hits Blockbuster Status with $1.1 Billion in 2017 (+29% YOY); Revenue in 4Q Rises 39% YOY to All-Time-High $332 Million; DECLARE (CV Outcomes), DERIVE (DKD), DEPICT 2 (Type 1) Results Upcoming

As we anticipated, Farxiga (dapagliflozin) reached blockbuster status in 2017: Sales of AZ’s SGLT-2 inhibitor rose 29% YOY to hit $1.1 billion (vs. $835 million in 2016). In 4Q17, Farxiga revenue rose 39% YOY and 17% sequentially for a record $332 million in quarterly sales (this was from a base of $239 million in 4Q16 and $285 million in 3Q17). The product accounted for 46% of AZ’s diabetes portfolio revenue and drove 69% of portfolio growth in 4Q17, by our calculations, accounting for 43% of revenue and driving a whopping 95% of growth in 2017. The Farxiga franchise has proven to be remarkably stable for AZ, with only one quarter of sequential decline in the past three years (-14% in 1Q17). Growth throughout 2017 has been quite positive (+26% YOY in 1Q17, +19% YOY in 2Q17, +30% YOY in 3Q17); this is muted from the 70% overall growth in 2016, but that’s to be expected for a product now several years into its launch cycle (dapagliflozin was the first-to-market SGLT-2 therapy OUS, launching in Europe in early 2013). Farxiga’s OUS performance continues to drive sales. In 2017, OUS revenue rose 55% YOY to reach $585 million, comprising 54% of total global sales. US revenue, in contrast, rose only 7% YOY to $489 million (capturing the remaining 46% of total global sales). In 4Q17, OUS sales climbed 67% YOY and 20% sequentially to $182 million, while US sales grew 15% YOY and 13% sequentially to $150 million. Notably, Farxiga was second-to-market in the US after J&J’s Invokana (canagliflozin), which is one likely reason for the slower growth stateside. We also attribute the smaller upward slope in the US to persistent pricing pressure around diabetes drugs, which has come up on almost all earnings calls for diabetes pharma companies throughout 2017. AZ management emphasized that Farxiga is still experiencing growth in all global markets, and that this impressive financial performance is occurring “in spite of a very competitive marketplace.” Indeed, with the flurry of discussion surrounding Invokana and CANVAS results this past year, as well as the impressive uptick in Jardiance (Lilly/BI’s empagliflozin) sales following the drug’s new CV indication, it’s been easy to forget that Farxiga leads the market by volume with 41% of total global prescriptions, followed by Jardiance with 35% and Invokana with 17% (more on this above). The US list price for Farxiga is nearly identical to that of Jardiance, which suggests to us that the lower total revenue for Farxiga is due to some combination of (i) lower realized prices in the US (ie. due to PBM/payer rebates) and/or (ii) lower OUS pricing.

  • Management confirmed that the DECLARE trial (a “substantial opportunity”) is on track to read out in 2H18. According to, the CVOT will complete in July 2018. We’re hoping for results at EASD 2018 in Berlin. DECLARE will go a long way in answering lingering questions about the SGLT-2 inhibitor class. Is cardioprotection a class effect? All signs thus far (EMPA-REG, CANVAS) point to yes, but AZ’s study could build an even more compelling case for HCPs, regulators, guideline-writers, and payers (indeed, the most exciting implication might be more leverage in payer negotiations for all products in the class). The VERTIS CV CVOT for Merck/Pfizer’s ertugliflozin (now branded Steglatro) is expected to complete in October 2019. Is amputation risk a class effect? Some have expressed concerns following a signal for lower limb amputations in CANVAS, but Lilly/BI have released data underscoring no imbalance in this complication between empagliflozin vs. placebo. There are key differences in trial design across EMPA-REG and CANVAS that make it impossible to compare directly, but in any case, we believe amputations should be highly-manageable in the real world with stronger patient education and diligent monitoring of the feet. It would be a shame for amputation worry to stand in the way of so many people with diabetes benefiting from SGLT-2 treatment (considering the base rate for amputations, even in a diabetes patient population, is low) – DECLARE data could help mitigate these concerns.
    • With >17,000 participants, DECLARE will surpass EXSCEL (also an AZ trial) as the largest diabetes CVOT. Approximately 7,000 DECLARE participants have established CV disease at baseline, while ~10,000 have multiple CV risk factors but no prior CV events. Notably, this is a much larger primary prevention cohort than seen in either CANVAS or EMPA-REG OUTCOME: The former enrolled ~34% in a primary prevention subgroup, while >99% of participants in the latter had established CV disease at baseline. Heart failure is included as a co-primary endpoint in DECLARE. Can SGLT-2 inhibitors show CV benefit, particularly in heart failure risk reduction, in a lower-risk population? A post-hoc analysis of CANVAS found significant heart failure benefit in the primary prevention cohort, even though the benefit on three-point MACE did not reach statistical significance in this subgroup, which has led some experts (e.g. Dr. Itamar Raz at WCTD 2017) to predict that DECLARE will show significant findings on the primary composite endpoint that includes heart failure, even with a much larger primary prevention cohort.
  • AZ shared that topline results from the DERIVE trial (n=323) of dapagliflozin in people with type 2 diabetes/CKD stage 3A became available internally in 4Q17. The early data show that Farxiga is both safe and efficacious as a glucose-lowering agent in patients with moderate renal impairment. This 24-week trial included endpoints on A1c, weight, fasting plasma glucose, and systolic blood pressure to evaluate both the glycemic efficacy and renal safety of Farxiga. Currently, the product is not recommended for people with eGFR between 30-60 mL/min/1.73 m2, and it’s contraindicated for eGFR <30 mL/min/1.73 m2. All SGLT-2 labels actually feature similar contraindications, but as Dr. David Fitchett suggested at ESC 2017, these designations are due to a presumed loss of efficacy given a mechanism of action that relies on kidney function – they’re not meant to imply any specific or major safety concerns. Dr. Fitchett implied that patients with renal impairment may actually stand to benefit the most from SGLT-2 therapy, and DERIVE could be the first hard evidence toward this answer. We’ll be curious to see whether AZ pursues a label update for Farxiga based on this data. Our understanding is that treating patients with renal impairment is particularly challenging, and it’s great to see evidence supporting SGLT-2 inhibitor use in this population. The company is also conducting Dapa-CKD in patients with CKD (eGFR between 25-75 ml/min/1.73m2) with or without diabetes. We can’t wait for DERIVE results to be reported publically, and we hope the study will be presented at an upcoming scientific meeting in 2018.
  • Following positive DEPICT 1 results at EASD 2017, AZ announced in its 4Q17 press release that Farxiga will be filed with FDA for a type 1 diabetes indication in 2H18. The company’s presentation materials also indicate that 52-week DEPICT 1 results and 24-week DEPICT 2 results were internally read out in 4Q17, and we look forward to public presentation of both sets of data. The topline results showed significant reductions in A1c, body weight, and total daily insulin requirements with both the 5 mg and 10 mg daily doses. Compared to evidence in people with type 2 diabetes, the safety profile was similar, save a higher DKA rate with Farxiga vs. placebo (no further details were provided). For context, Sanofi/Lexicon plan to file SGLT-1/2 dual inhibitor sotagliflozin for type 1 diabetes with FDA in 1H18. The companies have indicated that they hope to submit as early as possible – and we think they should appreciate the company of AZ, as navigating the regulatory landscape for an oral adjunct type 1 diabetes drug could be an uphill battle. Lexicon has been extremely positive to-date about its interactions with FDA, but we’re a bit uncertain of how the agency is going to approach this therapy class and its associated DKA risk in type 1 diabetes (we continue to believe this is manageable, but we wouldn’t be surprised if FDA exercised caution). We’re so excited about both dapagliflozin and sotagliflozin for type 1 diabetes, especially in terms of improving time-in-range (and in turn, enhancing quality of life). All things considered, we’re confident that AZ has the experience to bring this new indication to market, and we’re grateful for the time and resources the company has put into helping people with type 1 diabetes; the same goes for Sanofi/Lexicon. Lilly/BI’s EASE trials of Jardiance in type 1 will also read out this year, so we could be looking at a more robust commercial landscape for adjunct oral treatments two-three years down the line.
  • Farxiga is clearly a top priority for AZ, within the diabetes business and for the overall pharmaceutical business, as reflected in the robust program of ongoing clinical trials (see table 3 below). It’s noteworthy that the company is continuing to invest in commercial and clinical development for the SGLT-2. During AZ’s 2Q16 update, management announced that they would deliberately shift focus away from DPP-4 (Onglyza) toward SGLT-2, positioning Farxiga as the preferred option for patients switching off a DPP-4 inhibitor onto an SGLT-2 inhibitor. AZ seems in tune to the notion that DPP-4s are advantageous early in the course of diabetes, but that they can’t necessarily compete with SGLT-2s in terms of glucose-lowering potency or beyond-A1c benefits as the disease progresses. Certainly, the SGLT-2 class is posing competition for DPP-4s inhibitors on the market (see our pooled market highlights above). AZ’s strategic shift has been fruitful and smart, from our view.

Figure 4. Farxiga Franchise Sales (1Q13-4Q17)

Table 3. Ongoing Farxiga Clinical Trials


Estimated Enrollment


DECLARE (cardiovascular outcomes)


Data anticipated in 2H18 (moved up from April 2019 as of 2Q17 update)

Dapa-HF (heart failure with reduced injection fraction)


Initiated February 2017 and expected to complete December 2019

Dapa-CKD (chronic kidney disease)


Initiated February 2017 and expected to complete November 2020

DERIVE (glucose-lowering in people with comorbid diabetes/DKD)


Completed and topline data read out internally in 4Q17, showing significant glucose-lowering vs. placebo

DEPICT 2 (type 1 diabetes)


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

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


Expected to complete April 2018

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


Expected to complete September 2018

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


Expected to complete March 2019

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


Expected to complete August 2018

5. AZ’s GLP-1 Products Bydureon and Byetta Show Flat Sales at $197 Million in 4Q17; Bydureon Revenue Falls 1% YOY to $574 Million in 2017 – In-Class Competition Continues; Potential Tailwind in 2018 with New Bydureon BCise Autoinjector

Sales of GLP-1 agonist Bydureon fell 1% YOY to $574 million in 2017, from $578 million in 2016. In 4Q17, sales grew 4% YOY and 15% sequentially to $147 million (from $142 million and $128 million, respectively). This was an easy QOQ comparison for Bydureon, as quarterly sales fell 14% sequentially in 3Q17 (and for that matter, also dropped 9% sequentially in 2Q17). A net decline on yearly sales for Bydureon is disappointing in the context of a rapidly growing overall GLP-1 agonist class, and within-class competition continues to be a challenge. In contrast, sales of market leader Novo Nordisk’s Victoza climbed 16% YOY to $3.6 billion in 2017, and sales of Lilly’s Trulicity (a fellow once-weekly option) more than doubled YOY to hit $2 billion. That said, AZ’s GLP-1 business could experience a tailwind in 2018 with the new Bydureon BCise autoinjector (US launch expected in 1Q18, EMA approval expected in 1H18). Management positioned BCise as a significant opportunity for the franchise, offering a more patient-friendly alternative to dosing compared to single-dose Bydureon reconstitution kits or even the dual chamber pen (which still required a lengthy tapping and mixing process). We hope for more granularity in future earnings updates on prescriptions for the autoinjector vs. dual chamber pen vs. single-dose reconstitution kits, with TRx share ideally shifting toward the more advanced Bydureon BCise.

  • Bydureon (exenatide once-weekly) and Byetta (exenatide twice-daily) together posted $195 million in 4Q17, which is essentially flat from $197 million of combined sales in 4Q16. Bydureon continues to cannibalize some Byetta sales (which is to be expected, given the much lower injection burden), although we imagine Byetta is also losing share to other GLP-1 products on the market (namely, market leaders Victoza and Trulicity).
  • We estimate that the GLP-1 class will reach $6.5 billion in 2017 (pooled sales neared $5 billion for the full year 2016). YTD, pooled sales from all major GLP-1 agonists are up 28% YOY ($4.6 billion in 1Q17 + 2Q17 + 3Q17 vs. $3.6 billion in 1Q16 + 2Q16 + 3Q16). We believe new market entries will only spur further growth moving forward: This includes Novo Nordisk’s highly-anticipated once-weekly Ozempic (semaglutide), which will be launched in the US in 1Q18, as well as Intarcia’s ITCA 650 down the line. Stay tuned for a much deeper dive on this market for 4Q17 and 2017 as a whole. Sanofi reports this Wednesday, February 7.
  • We’re also continuing to watch the impact of neutral EXSCEL results on the Bydureon business. Even though we believe this study actually lends evidence in support of a cardioprotective class effect around GLP-1 agonists, the results won’t support a label update for Bydureon the way LEADER did for Victoza. And now that Novo Nordisk’s liraglutide is indicated for CV risk reduction, we imagine GLP-1 agonists without demonstrated CV benefit may have difficulty competing commercially over the long term. It’s too early to tell exactly how this will play out, and we think Bydureon could still see a boost in sales if the EXSCEL data is added to the label – after all, these findings offer very compelling evidence for CV safety, which is still meaningful in reassuring patients/HCPs in the diabetes community.

Figure 5. Bydureon Sales (3Q12-4Q17)

Figure 6. Byetta Sales (3Q12-4Q17)

6. DPP-4 Onglyza Sales Rebound in 4Q17, Growing 21% YOY to $180 Million Against Easy Comparison; Full Year Sales Drop 15% to $611 Million; Company Continues to Shift Focus to SGLT-2 Over DPP-4 Business

Sales of DPP-4 inhibitor Onglyza (saxagliptin) grew 21% YOY in 4Q17 to $180 million (from a base of $149 million in 4Q16), breaking a five-quarter streak of double-digit YOY decline. This is positive news for AZ, but it also means that the growth in 4Q17 came against an easy comparison of 22% YOY decline in 4Q16. Full year 2017 revenue for the product totaled $611 million, dropping 15% from $720 million in 2016. Though Onglyza received no air time on the call, AZ’s press release described “adverse pressures” on the DPP-4 inhibitor class as a whole, as prescribing habits move in favor of newer SGLT-2 inhibitors and GLP-1 agonists. Management has been saying since 2Q16 that the company will prioritize resources around SGLT-2 Farxiga over DPP-4 Onglyza, suggesting that AZ is prepared to let dapagliflozin cannibalize some saxagliptin sales – this position is reiterated in 4Q17 presentation materials. All in all, we think this is a smart strategy, one that has certainly paid off for AZ given Farxiga’s climb to blockbuster status. Unlike other DPP-4 inhibitors, Onglyza showed a worrisome signal for heart failure hospitalization in its CVOT, and while some have suggested that DPP-4 agents are most effective in recently-diagnosed patients, the majority opinion among diabetes thought leaders is that this class simply can’t compete with the potency of SGLT-2 inhibitors or GLP-1 agonists later in the course of disease (on A1c, body weight, or CV benefit). Onglyza’s new status on China’s National Reimbursement Drug List (NRDL) as of 3Q17 could potentially boost sales in this large market. In line with this, product revenue in emerging markets grew 16% YOY as reported (13% in constant currencies) in 4Q17 to reach $37 million. For a deep dive on the DPP-4 inhibitor class as a whole, scroll up to our pooled market highlights.

Figure 7. Onglyza Sales (4Q12-4Q17)

7. Fixed-Dose Combo Product Qtern Posts $5 Million in First Quarter of Reported Sales, Including $4 Million from US Market

This was the first time AZ broke out sales for fixed-dose combination Qtern (dapagliflozin/saxagliptin), reporting $5 million in product revenue for 4Q17. Qtern received no mention during the call, and it’s unclear exactly when the product was launched in the US (following FDA approval in March 2017). At $4 million, US sales accounted for the lion’s share of Qtern revenue, compared to only $1 million ex-US (though Qtern has been approved in the EU since 2Q16). We note that $5 million quarterly sales are rather low for a product that was highly-anticipated (AZ has expressed a great deal of optimism around Qtern in the past, and the drug actually received a Complete Response Letter from FDA the first time it was submitted). We suspect that reimbursement is a current challenge, and the company will have to focus on payer negotiations to improve uptake. As we understand it, HCPs in the US are also particularly hesitant to prescribe fixed-dose (or fixed-ratio) combination treatments because this is still an unfamiliar concept. To this end, we hope to see concerted HCP education efforts from AZ as well. We do think it’s early and we believe AZ is well-positioned to carry forward commercial success for this combination product, particularly given the company’s enthusiasm around Farxiga (Qtern’s SGLT-2 component).

  • As for the SGLT-2/DPP-4 class, it’s unclear how $5 million compares to revenue from Lilly/BI’s Glyxambi (empagliflozin/linagliptin), as Lilly does not break out Glyxambi sales. Merck/Pfizer’s Steglujan (ertugliflozin/sitagliptin) is poised to enter the market in 1Q18, and notably, will be priced lower than its competitors. At our local pharmacies, the list price for Glyxambi is ~$22/day and the list price for Qtern is ~$24/day, while Merck issued a statement setting list price for Steglujan at $17.45/day. List price only means so much without considering payer coverage and patient discounts, but it seems like reimbursement is poor across all these fixed-dose combos, so Steglujan’s lower price could make it a more attractive option. The average list price for standalone SGLT-2 inhibitors right now is ~$17/day, and for just a little more, patients would also get the DPP-4 component, plus superior efficacy and a milder side-effect profile overall. This is only our speculation for now, but we’ll be watching these market dynamics with keen interest throughout 2018, and we hope to see volume/sales growth for these advanced combination products.

Pipeline Highlights

8. Triple Fixed-Dose Oral Combination Dapa + Saxa + Metformin Slated for 1H18 FDA/EMA Filing

AZ recently added a new combination therapy to its pipeline page – fixed-dose oral tablets of SGLT-2 inhibitor Farxiga (dapagliflozin), DPP-4 inhibitor Onglyza (saxagliptin), and metformin. The company is targeting 1H18 for FDA and EMA submission. We hadn’t realized this triple combination was in development at AZ, though it’s certainly a natural follow-up to Qtern (dapagliflozin/saxagliptin), which launched last year in the US (after FDA approval in 1Q17). Lilly also announced during its 2018 financial guidance call that the company will file a triple fixed-dose combo of SGLT-2 Jardiance (empagliflozin), DPP-4 Tradjenta (linagliptin), and metformin with FDA in 2018. These two products are neck-and-neck to be first-to-market in this specific combination therapy class, although we anticipate AZ and Lilly/BI are investing in these new drugs more so as a way to stimulate their overall SGLT-2 inhibitor business (in other words, management may not be expecting either triple combo to generate substantial revenue as an individual product early on). Real-world providers/patients have been slow to uptake fixed-dose SGLT-2/DPP-4 combos thus far (Qtern posted only $5 million in 4Q17, its first quarter of broken out sales, while Lilly management has cited HCP hesitation to prescribe fixed-dose medications as a reason for sluggish Glyxambi sales). The metformin component of these triple combinations could boost uptake, given how familiar the generic agent is within the diabetes community. We do wish more patients were reaping the benefits of these fixed-dose combination products, which offer superior A1c-lowering and weight loss alongside a milder side-effect profile and lower pill burden, plus one co-pay instead of two, simplifying medication regimen. Building payer coverage will be key. Neither Qtern nor Glyxambi are very well reimbursed right now, as far as we can tell, and payer negotiations will be crucial for these fixed-dose combos as well as the new triple combos, pending regulatory approval. See the table below for a comprehensive summary of AZ’s diabetes-related pipeline. We discuss some of the 4Q17 updates to dapagliflozin clinical trials in our highlight on Farxiga, above.

Table 4. 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 4Q17.





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 in 3Q18 or 4Q18

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

FDA/EMA filing planned for 2H18; 52-week DEPICT 1 data and 24-week DEPICT 2 data reported internally in 4Q17; Positive DEPICT 1 results presented at EASD 2017; DEPICT 2 expected to complete 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 with expected completion in December 2019

Farxiga (dapagliflozin)

Outcomes data in patients with CKD

Phase 3

Dapa-CKD initiated in February 2017 with expected completion in November 2020; Topline results from DERIVE reported internally in 4Q17, showing clinically-meaningful and statistically significant A1c reduction in people with type 2 diabetes/CKD stage 3A

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


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

Phase 3

FDA filing planned for 2H18


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


Anti-miR103/107 oligonucleotide for NASH

Phase 2

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


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

Phase 2

Phase 2 trial expected to complete August 2018

Select Questions and Answers

Q: Could we talk about Farxiga, which is obviously a $1 billion product now? Pending data from DECLARE, this could be a substantially under-valued asset. Humor us: If DECLARE hits for both primary and secondary prevention, and is further validated by heart failure trials, how large a product can this actually be prior to patent expiry? The interim analysis for Dapa-HF – we’re expecting that this year. Would I be right in thinking that’s a reasonable assumption?

Mr. Sean Bohen (CMO, AstraZeneca): I’ll be very happy to forecast what I think Farxiga is worth. The Dapa-HF trial investigates Farxiga for the treatment of heart failure in people with diabetes as well as people without diabetes. Positive results there would enable the use of Farxiga outside of diabetes as well. There are interim analyses coming. We don’t get into the details of how they’re powered or when they’re done. Once we have them, if they’re positive, we announce them. Otherwise our assumption is that we go ahead into the final analysis, so I cannot provide more detail than that.

Mr. Mark Mallon (EVP, Global Medical Affairs, AstraZeneca): I’m cautioned to say that we don’t provide guidance, but what I can say is that we see substantial further opportunity for Farxiga in the class. Of course, it’s a $1 billion-plus brand already, growing in the mid-20s. We’ve expanded access in the US. There’s still tremendous room in terms of changing clinical practice. There are only a couple of countries in the world where SGLT-2s are ahead of DPP-4s in guidelines, even though we know have outstanding evidence showing that this really, as a class, has a great CV benefit. So, we still have a lot of work to do on the education. DECLARE will have a very big impact, if it’s positive, because of the breadth of the patient risk profiles in the study. And then of course, that’s just in the diabetes area. If we’re successful in the heart failure program, that’s further upside. This could be a very substantial product. We’re absolutely committed to it across the globe. In Brazil, Farxiga is the number one innovative oral anti-diabetes product, and that’s a major market; it’s bigger than Galvus, bigger than Januvia. That’s where we’re aiming to take this medicine.

Mr. Pascal Soriot (CEO, AstraZeneca): Unfortunately, the countries where SGLT-2s are recommended first instead of DPP-4s are small countries; one of them is Singapore. But hopefully, with a lot of good work, we’ll be able to modify the guidelines. If the class becomes first-line after metformin, the potential is enormous. We could also have added CKD, because we also have a program in kidney disease.


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