Memorandum

AstraZeneca 3Q15 – Diabetes sales up 10% to $577 million, boosted by Farxiga and Bydureon pen; CRL for saxa/dapa not due to heart failure or DKA – November 5, 2015

Executive Highlights

  • AZ’s diabetes portfolio posted total sales of $577 million in 3Q15, up 10% year over year (YOY) as reported (17% in constant currencies) and 1% sequentially.
  • Farxiga/Forxiga (dapagliflozin) sales reached $135 million in 3Q15, up 5% sequentially from $129 million in 2Q15. Onglyza (saxagliptin) continued to decline, falling 8% YOY as reported and 2% sequentially to $203 million. Uptake of the new Bydureon (once weekly exenatide) pen remained strong, driving an impressive 30% YOY growth for the product to $162 million.
  • AZ shared that the recent FDA Complete Response Letter for its saxagliptin/dapagliflozin fixed-dose combination was not due to concerns about heart failure or DKA.

AstraZeneca (AZ) provided its 3Q15 update Thursday morning in a call led by CEO Mr. Pascal Soriot. Below we include our top ten highlights from the call, followed by relevant Q&A.

Financial Highlights

1. AZ’s diabetes portfolio posted total sales of $577 million in 3Q15, up 10% year over year (YOY) as reported (17% in constant currencies) and 1% sequentially.

2. Farxiga/Forxiga (dapagliflozin) sales reached $135 million in 3Q15, up 5% sequentially from $129 million in 2Q15. The product has faced some challenges in the US due to formulary changes and concerns about DKA, but AZ expects its performance to improve in 2016.

3. Sales of the Onglyza (saxagliptin) franchise fell 8% YOY as reported (were flat in constant currencies) and 2% sequentially to $203 million. AZ is still awaiting a potential label update based on the heart failure results from SAVOR.

4. Bydureon (once weekly exenatide) sales rose an impressive 30% YOY as reported (34% in constant currencies) and 16% sequentially to $162 million, driven by continued uptake of the new pen and overall class growth.

5. Byetta (twice daily exenatide) sales declined 22% YOY as reported (17% in constant currencies) and 12% sequentially to $72 million.

6. During Q&A, management acknowledged that AZ has revised its medium-term expectations for US diabetes sales downward, though the expectations outside the US remain high.

Pipeline Highlights

7. Management shared that the recent FDA Complete Response Letter for its saxagliptin/dapagliflozin fixed-dose combination (FDC) was not due to concerns about heart failure or DKA.

8. AZ confirmed that the DECLARE CVOT for dapagliflozin is now fully enrolled.

9. The company is conducting two phase 3 trials of dapagliflozin in type 1 diabetes: DEPICT 1 and DEPICT 2, scheduled to complete in 2017.

10. AZ advanced a new candidate for diabetic kidney disease into phase 1.

Financial Highlights

1. AZ’s diabetes portfolio posted total sales of $577 million in 3Q15, up 10% year over year (YOY) as reported (17% in constant currencies) and 1% sequentially. Management stated that this growth was driven by the progression of several strong launches (particularly Farxiga/Forxiga [dapagliflozin] and the Bydureon [once weekly exenatide] pen) and AZ’s “global footprint.” Management also suggested that the company may boast the fastest-growing diabetes franchise in the field at the moment. While the portfolio grew in all geographies, growth was strongest in emerging markets (73% for the first nine months of the year), while the US market posed the most challenges due to competition and formulary losses.

Figure 1: Total Sales for AZ’s Diabetes Portfolio (3Q12-3Q15)

2. Farxiga/Forxiga (dapagliflozin) sales reached $135 million in 3Q15, up 5% sequentially from $129 million in 2Q15. Unlike in 2Q15, sales were fairly evenly split between the US ($69 million) and outside the US ($66 million). This reflects contrasting sequential growth trends, as US sales declined 12% sequentially in 3Q15 while ex-US sales rose 29%. The split in ex-US sales was $36 million in Europe, $19 million in emerging markets, and $11 million in other established markets.

Figure 2: Farxiga/Forxiga Sales (1Q13-3Q15)

  • During Q&A, management largely attributed the US sequential decline for Farxiga to an overall class slowdown prompted by concerns about DKA. Farxiga remains a driver of growth in the US overall, underpinned by greater promotional activity. However, the company stated that a class decline occurred following the burst of concern about DKA with SGLT-2 inhibitors in May and June and particularly the FDA's Drug Safety Communication. Interestingly, management commented that the lost volume was picked up by GLP-1 agonists (including Bydureon) and DPP-4 inhibitors (not including Onglyza). AZ suggested that this slowdown was a temporary setback and that it expects the class to rebound, especially following the positive EMPA-REG OUTCOME results for Lilly/BI’s Jardiance (empagliflozin). Lilly referred to an SGLT-2 inhibitor class slowdown in its 2Q15 update to explain the sequential decline in Jardiance sales that quarter, though the company did not explicitly state that the DKA controversy was responsible. There was no mention of a class slowdown in 3Q15, when Jardiance experienced a post-EMPA-REG bump in sales.
    • Pooled revenue results do show a bit of a plateau in growth for the SGLT-2 inhibitor class in 3Q15. Total sales for the class rose 7% sequentially in 3Q15, compared to sequential growth rates well into the double digits in recent quarters. However, it is difficult to know how much of this is due to a specific event like the DKA controversy vs. a natural plateauing of growth as the class becomes more established. Either way, we agree that the positive EMPA-REG OUTCOME results will likely outweigh any concerns about DKA going forward, particularly following AACE/ACE’s recent announcement advising no changes to current recommendations for SGLT-2 inhibitor use due to DKA risk.

Figure 3: SGLT-2 Inhibitor Sales (1Q13-3Q15)

  • AZ expects Farxiga’s US performance to improve in 2016 due to better formulary access and changes to patient access programs. The company has alluded to US formulary challenges for Farxiga throughout 2015, but it appears that it had better luck in the 2016 negotiation cycle. One positive change for AZ was the news that CVS Caremark has excluded J&J’s market-leading Invokana (canagliflozin) from its 2016 formulary for a portion of covered lives, leaving both Farxiga and Jardiance as preferred options. We imagine that this could help competitors cut into Invokana’s dominance within the class, although J&J stressed in its 3Q15 update that it expects continued broad access for Invokana in 2016.  The EMPA-REG OUTCOME results could also help Jardiance make some inroads against its competitors if they lead to improved formulary positioning. In addition, AZ stated that the patient support program for Farxiga has been “tightened up.” The company has been hinting at potential changes to the savings card program for several quarters; we assume this may involve greater restrictions on eligibility or on the length of time patients can receive the drug for free.
  • Growth for Forxiga in ex-US markets remains strong. AZ highlighted the product as a driver of growth in the EU and emerging markets and noted that additional launches are ongoing, though the company did not provide details on penetration in specific markets as it has in the past. In contrast to the situation in the US, Forxiga has benefited from its first-in-class status in Europe in particular and typically holds much greater market share in ex-US markets.
  • AZ appears to be operating under the assumption that the EMPA-REG OUTCOME results will be interpreted as a class effect. There was perhaps surprisingly little discussion of the results during the call – management mentioned them only as a factor likely to provide a boost to the entire SGLT-2 inhibitor class. The consensus in the scientific community seems to be that a class effect is the most likely explanation, but that it would be prudent to wait for results from the remaining SGLT-2 inhibitor trials before jumping to broad conclusions. We expect that for the most part, guidelines committees and payers will want to wait until the question of a class effect is resolved before implementing any dramatic changes. However, we do expect the results to have some effect on prescribing patterns that could benefit Jardiance at the expense of its competitors. Indeed, several speakers at the recent Cardiometabolic Health Congress said they would support switching patients from other SGLT-2 inhibitors to Jardiance based on the results.

3. Sales of the Onglyza (saxagliptin) franchise fell 8% YOY as reported and were flat in constant currencies at $203 million. Sequentially, sales declined 2% against a difficult comparison (14% sequential growth in 2Q15). US sales totaled $111 million, down 15% YOY as reported and down 2% sequentially. Outside the US, sales rose 2% YOY as reported and fell 3% sequentially to $92 million. These figures include both standalone Onglyza and Kombiglyze (saxagliptin/metformin). AZ acknowledged the continuing challenges for Onglyza in the US, including decreased market share within its class, a shift in promotional resources to the Bydureon pen and Farxiga, and a lower net price. As with Forxiga, the situation is more positive ex-US, as AZ highlighted Onglyza as a driver of growth in Europe and emerging markets.

  • AZ is still awaiting a potential update to Onglyza’s label based on the heart failure results from SAVOR. We would certainly expect that update to arrive soon, as it has now been almost seven months since the FDA hosted an Advisory Committee meeting to discuss the results from SAVOR and EXAMINE (the CVOT for Takeda’s Nesina [alogliptin]). AZ characterized that meeting as a success during its 1Q15 update, as the panel voted 13-1-1 in favor of an acceptable cardiovascular risk profile for Onglyza. However, the panel also voted 14-1 in favor of adding new safety data from SAVOR to Onglyza’s label, and we assume at least some additions are on their way. We assume that the DPP-4 inhibitor/heart failure controversy is at least one factor that has contributed to Onglyza’s recent decline. It has also likely helped cement Merck’s Januvia (sitagliptin) as the clear leader of the DPP-4 inhibitor class, as it is so far the only agent to emerge from a CVOT completely untarnished by the heart failure question. That said, it is not clear whether an official label update would do significant additional damage to Onglyza’s prospects, as we assume awareness of the issue is already relatively high.

4. Bydureon (once weekly exenatide) sales rose an impressive 30% YOY as reported (34% in constant currencies) and 16% sequentially to $162 million. US sales were $138 million, up 29% YOY as reported and 19% sequentially. Ex-US sales rose 33% YOY as reported and were flat sequentially at $24 million. Management continued to characterize the Bydureon pen as an important driver of growth. Conversion from the old vial-syringe system is now up to 60%. The company also noted that US growth has been driven by growth of the overall GLP-1 agonist class and by higher net prices.

Figure 4: Bydureon Sales (3Q12-3Q15)

  • Based on data from Novo Nordisk’s 3Q15 update, Bydureon has lost some share to new competitors, but overall class growth has limited the impact on sales. IMS Health data presented in Novo Nordisk’s 3Q15 update showed both the exenatide franchise (Bydureon and Byetta) and Victoza losing some share (measured by total prescriptions) throughout the year to newcomers Lilly’s Trulicity (dulaglutide) and GSK’s Tanzeum (albiglutide). As of August, the exenatide franchise held 27% market share by volume compared to 59% for Victoza and 7% each for Trulicity and Tanzeum. By value, Bydureon accounted for 16% of total GLP-1 agonist sales in 3Q15, and the total exenatide franchise accounted for 23%. Data in the Novo Nordisk presentation showed US volume growth for the GLP-1 agonist class rising to close to 20%, up from ~8% at the beginning of this year. The graph below of pooled sales over time tells the same story: class growth has rebounded this year following challenges in 2014.

Figure 5: GLP-1 Agonist Sales (3Q13-3Q15)

  • According to the company’s pipeline, AZ still intends to file a Bydureon suspension formulation in the US and EU in 4Q15. Such a product would presumably enable an auto-injector device and would likely help AZ compete with products like Trulicity and Sanofi’s Lyxumia (lixisenatide) that come in ready-to-use devices.  

5. Byetta (twice daily exenatide) sales declined 22% YOY as reported (17% in constant currencies) and 12% sequentially to $72 million. US sales declined 18% YOY and 15% sequentially to $45 million. Ex-US sales declined 27% YOY and 7% sequentially to $27 million. Byetta sales have generally been on a downward trajectory as AZ has deemphasized the product in favor of Bydureon. Long term, we do not expect Byetta to remain a strong competitor in a more crowded and patient-friendly GLP-1 agonist class, though its sales have remained relatively strong for some time, perhaps due to reimbursement advantages.

Figure 6: Byetta Sales (3Q12-3Q15)

6. During Q&A, management acknowledged that AZ has revised its medium-term expectations for US diabetes sales downward, though the expectations outside the US remain high. These comments came in response to a question on whether AZ thought its targets for diabetes sales for 2023 were still achievable given the controversies over side effects and regulatory hurdles. Management responded that the company’s overall expectations for 2023 sales remained unchanged but that the current plan reflected lower diabetes sales and higher oncology sales than originally anticipated. The unexpected challenges in diabetes are due entirely to pressure in the US. Management did not comment on specific contributing factors, but the company has referred to increased pricing pressure from US payers several times in quarterly updates over the past year. By contrast, management emphasized that growth remains quite strong in Europe and emerging markets and that AZ remains competitive in diabetes in the US – the performance is just not quite where the company had originally expected it would be. The expectation for increased pricing pressure in the US has been an almost universal one in pharmaceutical company updates this quarter. In some ways this is a reflection of the fact that the US has contributed disproportionately to growth in recent years relative to cost-conscious European markets, suggesting that this could be more of a transition to a new equilibrium rather than a long-term negative trend.

Pipeline Highlights

7. AZ shared that the recent FDA Complete Response Letter for its saxagliptin/dapagliflozin fixed-dose combination (FDC) was not due to concerns about heart failure or DKA. This is very reassuring to hear, as we had assumed that the heart failure question with saxagliptin was the most likely reason for the CRL. Management did not disclose exactly what the reasons were but stated that the FDA requested more clinical data “to support various dosing regimens.” While it is impossible to know for sure, this suggests to us that some minor additional work testing particular dose combinations is all that will be required, which would hopefully mean a relatively short delay before approval. AZ has not yet met with the FDA to discuss the required next steps and the anticipated timeline, but management promised more specific guidance as soon as the meeting takes place. We are glad to hear that the CRL was not due to any significant safety concerns, though it is still frustrating if approval was delayed by a relatively minor issue that could likely have been avoided with better communication.

  • “Saxa/dapa” is still on track for an EU regulatory decision in 2H16. Management emphasized that the CRL in the US should have no impact on saxa/dapa’s timeline in the EU.
  • We wonder whether any heart failure additions to saxagliptin’s label will affect saxa/dapa. AZ indicated in its 2Q15 update that it was considering mechanistic studies to evaluate the risk of heart failure with the combination, but it has not yet initiated any to our knowledge. We assume it is likely that some language about heart failure will be included in the label, but it will probably not be as definitive given that there is no evidence of increased risk with the combination itself. AZ has even suggested that dapagliflozin’s small diuretic effect could potentially counterbalance any increased heart failure risk with saxagliptin. Overall, we assume the impressive efficacy and clean safety profile for saxa/dapa in clinical trials would preclude any major concerns on this front among payers and providers.
  • It remains to be seen whether Lilly/BI’s Glyxambi (empagliflozin/linagliptin) will gain a substantial advantage from its now-extended period as the only SGLT-2/DPP-4 inhibitor combination on the market. AZ suggested in 2Q15 that there could be some advantages in being second to market, as the company can learn from Lilly/BI’s experience with payers and take cues from them on pricing. Lilly has offered little commentary on Glyxambi since its launch in March, and the product does not appear to have contributed significantly to sales in the past two quarters. We are not sure where reimbursement stands, but Glyxambi will certainly require robust coverage to be successful since its out-of-pocket cost ($581 for a monthly supply) will be out of reach for most patients. AZ has suggested in the past that it plans to price saxa/dapa at a similar level to Glyxambi, but we wonder if there is any chance the company would agree to a lower price point in exchange for increased access.   

8. AZ confirmed that the DECLARE CVOT for dapagliflozin is now fully enrolled. The company’s supporting materials in 2Q15 indicated that the trial had completed enrollment, but this was the first confirmation we had heard during prepared remarks. The study aimed to enroll 17,276 patients and still has an estimated completion date of April 2019. Dapagliflozin will be the last of the three currently approved SGLT-2 inhibitors to report CV outcomes results. By the time those results are presented, there should either be a fairly strong consensus around an SGLT-2 inhibitor class benefit if the CANVAS trial for Invokana (completion expected in 2017) is also positive, or a great deal of confusion and debate if it is not.

  • In the opening session of CODHy today, Dr. Stefano Del Prato (University of Pisa, Italy) noted that an interim analysis of DECLARE will take place next September and could provide hints about the question of a class effect. He raised the possibility that the trial could be stopped early for superiority, which would provide evidence in favor of a class effect sooner than anticipated. He suggested that if the trial is not stopped at that point, one can assume that a class effect is less likely. We assume the trial would not be stopped early unless the interim analysis shows an enormous benefit (or harm) – EMPA-REG OUTCOME was allowed to run to completion even though the curves for mortality began to separate almost immediately after the trial began.

9. AZ is conducting two phase 3 trials of dapagliflozin in type 1 diabetes. The first, DEPICT 1, is evaluating two doses of dapagliflozin (5 mg and 10 mg) vs. placebo as an adjunct to insulin in 768 patients with type 1 diabetes. The primary endpoint is change in A1c at 24 weeks. Secondary endpoints include percent change in daily insulin dose, percent change in body weight, several parameters obtained from CGM data, and the proportion of participants achieving an A1c reduction ≥0.5% with no severe hypoglycemia. Safety parameters including hypoglycemia and DKA are also being prospectively adjudicated. The trial is expected to complete in May 2017. The second, DEPICT 2, has an almost identical design, though it does not list hypoglycemia and DKA as key outcome measures on ClinicalTrials.gov. It is expected to complete in December 2017. Interestingly, BMS is listed as the lead sponsor on both trials with AZ as a collaborator, though we would be very surprised if the company is actively involved in these studies given its exit from diabetes in 2013. We are very glad to see that AZ has decided to continue investigating dapagliflozin in type 1 diabetes, and these robust trials should help paint a more complete picture of how the benefits of SGLT-2 inhibitors in this population stack up against the risk of DKA.

  • AZ is conducting several other intriguing trials that could support expanded indications for dapagliflozin. A phase 3 study (n=660) of dapagliflozin in combination with exenatide is ongoing and has an expected primary completion date of May 2016 (moved up by one month since the 2Q15 update). We expect that this combination has the potential to produce very impressive effects on A1c and body weight. A phase 2 study (n=80) investigating the effect of dapagliflozin in combination with Epanova (omega-3 carboxylic acids) on liver fat in patients with type 2 diabetes is expected to complete next month.

10. AZ advanced a new candidate for diabetic kidney disease into phase 1. The compound, AZD9977, is a selective mineralocorticoid receptor modulator – a variation on the mineralocorticoid receptor antagonist class that appears several times in the competitive landscape for chronic kidney disease. The company is conducting three phase 1 trials of the candidate: (i) a study (n=88) assessing safety, tolerability, and pharmacokinetics of single ascending doses in healthy males; (ii) a study (n=12) comparing the pharmacokinetics of various extended release formulations of the drug and the impact of food in healthy males; and (iii) a study (n=24) assessing the pharmacodynamics of single doses in healthy males. The trials have expected completion dates between November 2015 and January 2016. We are glad to see AZ jumping back into the diabetic nephropathy arena after the company discontinued its partnership with Ardelyx on tenapanor following negative results from a phase 2a trial. AZ’s early-stage pipeline also includes a phase 1 GLP-1/glucagon dual agonist (MEDI0382) for diabetes/obesity. The company has also announced research collaborations over the past year with the Montreal Heart Institute, the Joslin Diabetes Center, and the Harvard Stem Cell Institute focused on cardiometabolic disease.

Questions and Answers

Q: Last year, you’d given a target for 2023 of reaching $8 billion for diabetes. There’s been a lot going on in diabetes in terms of side effect issues and regulatory. Do you still see that as potentially achievable – that would mean that the business has to triple between now and then – or could that now be challenging?

A: We remain committed to our long-term goals. In fact, we just finished our long-range planning process and we presented it to the board this week as we do every year. Our plan is very much in line with previous planning and reflects that our 2023 goals are achievable from our perspective. Having said that, as you would imagine, in every plan you have ups and downs. So what our planning reflects now is lower diabetes sales but higher oncology sales. If I talk about diabetes, we are doing better outside the US than we had expected originally. And in every emerging market I go to, when I ask people about their growth story, their growth story is very much about diabetes. So we’re very delighted to have diabetes as part of our portfolio.

We are a global company and even though every geography behaves a bit differently and it’s clear that diabetes is a bit of a competitive field today in the US, outside the US there are enormous opportunities. We do very well in Europe. And in emerging markets, comparatively we do very well. You’ve got to realize in those countries cancer, of course, is an issue, but the big issues they are dealing with today are cancer, hypertension, diabetes, COPD, asthma, and there are enormous opportunities in all of those diseases to help patients. So we do very well outside the US. In the US, we are doing okay, but certainly not as well as we were expecting originally. So if you take the totality of global sales, we see less diabetes sales and more oncology sales.

Q: The Farxiga sales in the US were actually lower in the third quarter than they were in the second quarter. Is there some competitive aspect that’s important there? Is there any shift in patients away from commercial plans toward government plans or is this just settling in of new promotional programs and we should start to see a strong acceleration?

A: There was an overall impact on the class with the signal we’re seeing with DKA in May and June, and the subsequent letter from regulators, and you can really see that effect on the overall class. We obviously were impacted by that with Farxiga. If you look at the share trends, that volume was picked up by GLP-1s, which we benefited from. And also by DPP-4s, which we didn’t. I think as we look further out, clearly there are some positive trends for SGLT-2s in terms of the EMPA-REG data. Also in October, we tightened up some of our access programs and affordability programs particularly in terms of cash. There was not a huge amount of movement between government and other programs. And again, in terms of actual sales, it was around $9 million. Again, we remain confident about the class.

It’s important to keep in mind that the SGLT-2 class is a new class and it’s an innovative new class, and as with many new classes you expect a few bumps in the road as you get started. If you go back to the statins, if you go back to the launch of Crestor, we suddenly experienced quite a number of eventful developments as we launched it many years ago. If you look at the SGLT-2 class in Japan, we (meaning the whole class) had this issue of dehydration of patients that has really slowed down the class, but we see a pickup now. It’s a good class, and we believe it helps patients and it will grow, and we see a pickup in Japan. In the US, the DKA has slowed it down. I believe that it’s going to pick up again, especially if this cardiovascular risk reduction is real.

Q: Have you met with the FDA regarding the specifics of the Complete Response Letter for saxa/dapa?

A: The Complete Response Letter stated that we would need more clinical data to support the application. We are committed to the saxa/dapa combo approval. We have not yet met with the FDA, and as a result of that, we don’t know exactly what more clinical data will be required. And that leads us to not being able to provide any clearer guidance with regard to how long the delay might be. But we’re hopeful to meet with them soon and then we will have clarity around that. I think it’s important to recognize that we do not believe this particular Complete Response Letter will affect filing outside the United States.

Q: On saxa/dapa, it’s been pretty clear which concerns FDA did not raise in their Complete Response Letter, i.e., increased risk of hospitalization for heart failure and diabetic ketoacidosis. Any chance you could be a bit more specific about which concerns it did raise? Is it a problem related to manufacturing, efficacy, safety, or dosing?

A: There’s no real, serious concern attached to the questions. They’re more to do with producing clinical data supporting the combination of the various dosage regimens. We can’t be a lot more specific than this because we haven’t met with the FDA and we need to understand exactly what they expect of us. But it has nothing to do with any safety or adverse events like DKA or heart failure and it has nothing to do with CMC. It’s purely producing the clinical data that they need to see to uphold the fixed-dose combination across dosage regimens. So we need to talk to them before we can answer specifically because the important part is to understand how long it’s going to take to produce those data.

--by Emily Regier and Kelly Close