Memorandum

Novo Nordisk 3Q14 – Diabetes Care up 11% to ~$3.1 billion, powered by Levemir and Victoza; New strategy for possible Tresiba NDA resubmission developed with FDA input – October 31, 2014

Executive Highlights

  • Diabetes Care sales rose 11% as reported (12% in constant currencies) to $3.1 billion following four quarters of low-single-digit growth.
  • Levemir (insulin detemir) drove over half of whole-company growth, with sales rising 27% to $660 million and capturing ~25% of the US modern basal insulin market.
  • Victoza (liraglutide) returned to double-digit growth, rising 21% as reported to $610 million and defending its leading share in the midst of a continued class slowdown.
  • A US Tresiba resubmission in 2015 based on interim data is no longer a given; rather, a firewalled group will review the data (with FDA input) and will decide whether to file or to wait for full trial results.

Novo Nordisk reported 3Q14 financial results yesterday in a call led by CEO Mr. Lars Sørensen. Total Diabetes Care sales rose 11% year-over-year (YOY) as reported (12% in constant currencies) to DKK 17.7 billion (~$3.1 billion) in 3Q14, representing a rebound from the slower beginning of the year (the portfolio grew 1% in 1Q14 and 3% in 2Q14). Remarkably, nearly half of diabetes growth driven by Levemir (insulin detemir), with a strong contribution from Victoza (liraglutide) as well. Levemir’s 27% growth as reported to DKK 3.7 billion (~$660 million) and continued strong performance in North America (up 41% to DKK 2.5 billion [~$450 million]) stood in contrast to the trend we saw for Lantus in Sanofi’s recent 3Q14 update, which had a fairly slow quarter, especially in the US. Victoza returned to double-digit growth after several sluggish quarters, rising 21% to DKK 3.4 billion (~$610 million) in 3Q14 amidst a continued class slowdown. Management highlighted Tresiba’s (insulin degludec) 24% basal insulin value share in Japan, a market where the new product is reimbursed on a comparable level to Lantus.

Keeping with the past few updates, the call included a major update on Tresiba’s DEVOTE cardiovascular outcomes trials (CVOT) and progress towards US regulatory resubmission. Following interactions with the FDA and an August FDA hearing on interim data disclosure from CVOTs, it is no longer a given (though it is still likely, in our view) that Novo Nordisk will resubmit the NDA for Tresiba based on interim data from DEVOTE. Instead, a firewalled group within the company, with input from the FDA, will examine the data and make a decision on whether to resubmit the NDA or to wait until the full trial is completed. Management will remain blinded to the interim results, which will also not be disclosed publicly in any way to maintain trial integrity. This is a change to the way interim data disclosures and NDA submissions are done, and although the agency clearly has a high bar for preserving interim data confidentiality, it was heartening to hear that the company and the FDA are completely on the same page, and that the agency might be willing to provide insight on what a comforting point estimate for MACE might be before the time of submission.

For plenty more pipeline updates and the full range of financial results, read on below.

Table: Financial results for Novo Nordisk’s major diabetes products in 3Q14

 

3Q14 Revenue (billions)

Reported/Operational Growth from 3Q13

Reported Growth from 2Q14

Total Diabetes Care

DKK 17.7 (~$3.1)

11%/12%

4%

Total Modern Insulins

DKK 10.6 (~$1.9)

13%/15%

3%

   NovoRapid/NovoLog

DKK 4.5 (~$0.79)

9%/11%

5%

   NovoMix

DKK 2.4 (~$0.43)

4%/5%

-2%

   Levemir

DKK 3.7 (~$0.66)

27%/28%

3%

Human Insulin

DKK 2.5 (~$0.44)

-4%/2%

0%

Tresiba

DKK 0.18 (~$0.03)

300%/NA

24%

Victoza

DKK 3.4 (~$0.61)

21%/21%

13%

Financial Highlights

  • Total Diabetes Care sales rose 11% as reported (12% in constant currencies) to DKK 17.7 billion (~$3.1 billion) in 3Q14. Sequentially, sales rose 4% as reported from 2Q14. As in past quarters, the main drivers of growth were Levemir (insulin detemir) and Victoza (liraglutide), particularly in North America. It is encouraging to see a return to double-digit growth for the portfolio following several quarters of more sluggish performance (growth was 3% in 2Q14, 1% in 1Q14, and 4% in 4Q13).
    • By geography, North America remains the main driver of growth, accounting for 65% of the company’s growth in the first nine months of 2014 (figures were not provided for 3Q14 alone). International Operations and China also made significant contributions to growth, accounting for 23% and 16%, respectively, of growth in the same time period.
    • Management noted that sales continue to be hurt by the difficult reimbursement and contract environment in the US. Specifically, 3Q14 growth was negatively impacted by the partial loss of reimbursement from Express Scripts, expanded Medicare Part D utilization, and adjustments to rebate provisions in 2013. If not for Levemir’s continued upward momentum, the pricing pressure in the US that we learned about through Sanofi’s recent 3Q14 update would have had a much larger impact on the diabetes portfolio’s bottom line.
  • Given Sanofi’s gloomy 2015 outlook for its diabetes portfolio in its recent 3Q14 update, we were interested to get a look at Novo Nordisk’s early guidance for 2015. Preliminary plans for 2015 in local currencies include sales growth in the high single digits, based on a forecast for “continued robust performance” for the modern insulins, Tresiba, and Victoza, as well as a modest contribution from Xultophy. The preliminary guidance mentions that these factors will be tempered by the impact from increased rebate levels in the US and intensifying competition – these are the factors that led to Sanofi’s forecast, but overall a forecast for growth in high single digits (driven by diabetes) is a much more positive outlook.

Insulins

Table: Insulin market share

 

Share of Total Insulin Market

Share of Modern and New-Generation Insulin Market

 

August 2014

August 2013

August 2014

August 2013

Global Total

47%

48%

46%

46%

US

37%

38%

38%

38%

Europe

48%

49%

48%

49%

International Operations

55%

55%

52%

53%

China

58%

59%

64%

64%

Japan

52%

53%

49%

49%

  • Novo Nordisk’s modern insulin portfolio (NovoLog/NovoRapid, NovoMix, and Levemir) collectively grew 13% as reported (15% in constant currencies) to DKK 10.6 billion (~$1.9 billion) in 3Q14, driven largely by Levemir. Sequentially, the modern insulins grew 3%, slightly lower than the 10% sequential growth in 2Q14. Novo Nordisk’s share of the global insulin market was 47% as of August 2014, compared with 48% share in August 2013. The company’s share of the modern insulin and new-generation insulin market (which now accounts for 80% of Novo Nordisk’s insulin sales) held steady at 46%. Growth was driven primarily by North America (which accounted for 65% of growth), interestingly driven by price (which overall is under a lot of pressure right now) along with share gains for Levemir, although the loss of reimbursement (due to the lost Express Scripts contracts) and expanded Medicare Part D utilization was a drag on sales. International Operations and Region China also made substantial contributions. By contrast, sales have declined in Europe, at least in part due to strict pricing requirements by several European governments, and in Japan due to declining volume and increased competition.

Levemir (insulin detemir)

  • Levemir continued its stellar 2014, growing 27% YOY as reported (28% in constant currencies) to DKK 3.7 billion (~$660 million) in 3Q14. While sequential growth was less robust at 3%, this is the second straight quarter of >25% YOY growth for the product. More than any other single product, Levemir was responsible for a greater share of the diabetes portfolio’s growth, at 42%.
    • Much of Levemir’s strong performance resulted from very robust growth in North America. Sales rose 41% YOY in the region to DKK 2.5 billion (~$450 million) in 3Q14, and the product has gained more than two percentage points in the US basal insulin market over the past year to reach ~25% market share. Levemir also performed well in China (albeit from a smaller base), where it grew 34% YOY to DKK 83 million (~$15 million) in 3Q14. Performance was weaker in other markets; sales rose 2% in Europe to DKK 735 million (~$131 million) and declined 25% to DKK 51 million (~$9 million) in Japan and Korea.
      • This geographic breakdown was almost the opposite of what we saw for Sanofi’s Lantus in 3Q14. US sales growth slowed to 6%, whereas Western Europe performed uncharacteristically well, growing 10%.
    • Management did not comment specifically on pricing and rebate negotiations for Levemir with US payers – for background, Sanofi made waves during its recent 3Q14 update when it disclosed that immense payer pressure and resulting high rebate levels could completely stifle growth into 2015. As a reminder, Sanofi revealed that it was compelled to accept a major increase in rebates for Lantus in 2015 due to competitive pressure, code (ostensibly) for aggressive discounting by Novo Nordisk during negotiations with payers. We were somewhat surprised to hear only one question on the topic during Q&A, as the subject completely dominated discussion during Sanofi’s update and even contributed to the firing of Sanofi’s CEO the day after the call. Novo Nordisk management did not provide specific details about the pricing of Levemir compared to Lantus but noted that both the list price and the absolute rebate level for Levemir have both increased over the last few years.
    • Management attributed much of Levemir’s gain to the launch of the FlexTouch pen in June. The device is the only pre-filled insulin pen with a push-button that does not telescope out when patients dial up the dose, which the company believes will be an important convenience factor for patients. The FlexTouch was on show in a big way at the Exhibit Hall at ADA this year.
    • Novo Nordisk does not believe that upcoming changes in the basal insulin market will necessitate an increased US sales force for Levemir. During Q&A, one analyst inquired whether a sales force increase would be necessary if Sanofi “gets their act together” and increases the sales force for Lantus. Novo Nordisk management indicated that the current sales force for Levemir should be adequate for the foreseeable future and that the FlexTouch pen should allow continued gains in market share. We also suspect that the US launch of Tresiba (which will likely take place in early 2016) will divert at least some resources away from Levemir. Levemir’s growth has been so exciting that during Novo Nordisk’s 2Q14 update, some investors seemed worried that a launch for Tresiba (definitely a more exciting product in terms of clinical characteristics) might get in the way of Levemir’s momentum – that fear ignores the fact that it was the initial CRL for Tresiba, and Novo Nordisk’s resultant reallocation of marketing resources to Levemir, that contributed to Levemir’s strong growth in the first place.

Tresiba (insulin degludec)

  • Sales of Tresiba (insulin degludec) reached DKK 175 million (~$31 million) in 3Q14, up 24% from DKK 141 million (~$26 million) in 2Q14. The uptake of Tresiba in the countries where it is launched remains highly dependent on the level of reimbursement. The product has gained significant basal insulin market share in countries like Japan (24%), Switzerland (22%), Mexico (16%), and India (12%) where it is reimbursed at similar levels to Lantus. Market share is much lower in markets like the United Kingdom (2%), Denmark (2%), Germany (4%), and Sweden (6%) where access is more restricted. Tresiba has now been launched in 22 countries, most recently in Brazil, Slovakia, Chile, and Russia.
    • Management used Japan as a case study to illustrate why the company believes launching Tresiba will likely increase Novo Nordisk’s overall share of the basal insulin market. Tresiba now accounts for 24% of the basal insulin market in Japan (by value) 20 months after its launch, and its strong penetration has raised Novo Nordisk’s total share of the basal insulin market from ~30% before the launch to 39% in August 2014, with most of the gains appearing to be coming at Lantus’ expense. In the company’s 2Q14 call, there was some concern that the Tresiba launch might stunt Levemir’s momentum in the US, but management has repeatedly said that Tresiba appears to be drawing patients primarily from market leader Lantus.
    • Management also highlighted the recent launch of Ryzodeg (insulin degludec/insulin aspart) in Mexico in September, saying that early feedback from that launch has been encouraging. We saw full phase 3 data on Ryzodeg for the first time at EASD, where the product was also featured prominently in Novo Nordisk’s exhibit hall booth; the product barely missed achieving statistical non-inferiority vs. basal-bolus therapy with Tresiba and NovoLog, although the A1c reductions were numerically similar between the groups. We expect that the main initial advantage of Ryzodeg from a patient perspective will be the reduced injection burden, which will hopefully facilitate better adherence to treatment; further on, we expect it may be used in combination and that it would have an advantage there.

Other Insulins

  • NovoLog (NovoRapid; insulin aspart) sales rose 9% as reported and 11% in constant currencies to DKK 4.5 billion (~$790 million); sequential growth was 5% as reported. NovoRapid appears to have recovered to some extent after a difficult past few quarters (sales declined 1% YOY in 2Q14 and 3% in 1Q14), but sales likely continue to be hurt at least to some extent by the loss of the Express Scripts formulary contract to Lilly’s Humalog (insulin lispro) last year.
    • By region, NovoLog was strongest in China (up 33% to DKK 153 million [~$27 million]) and International Operations (up 31% to DKK 471 million [~$84 million]). Growth was also positive in Europe (up 7% to DKK 1.0 billion [~$181 million]) and North America (up 7% to DKK 2.6 billion [~$464 million]), but sales declined by 25% in Japan and Korea to DKK 213 million (~$38 million).
  • NovoMix sales rose 4% as reported and 5% in constant currencies to DKK 2.4 billion (~$430 million); sequentially, sales declined 2% as reported. Not surprisingly, sales declined 8% to DKK 2.4 billion (~$107 million) in North America, where the market for premixed insulin has generally been shrinking; sales also declined 13% to DKK 164 million (~$29 million) in Japan/Korea and dropped 4% to DKK 577 million (~$103 million) in Europe. However, growth was positive in markets like China (up 12% to DKK 563 million [~$100 million]) and International Operations (up 31% to DKK 471 million [~$95 million]) where premixed insulin is more popular – we expect that Ryzodeg sales might show a similar geographic variation once it is launched.
  • Sales of Novo Nordisk’s human insulins fell 4% as reported and rose 2% in constant currencies to DKK 2.5 billion (~$440 million); sales were flat sequentially. Sales declined in all markets except International Operations, where growth was 2%. In North America, sales fell 3% to DKK 438 million (~$78 million), following an unusual bump of positive growth (5%) in 2Q14.

Victoza (liraglutide)

  • Victoza (liraglutide) returned to double-digit growth amidst a continued slowdown of the class as a whole. Total global Victoza sales rose 21% as reported in 3Q14 to DKK 3.4 billion (~$610 million) following several quarters of more sluggish performance. Most of Victoza’s growth is driven by North America, where sales rose 29% to DKK 2.3 billion (~$416 million) due to favorable pricing and the overall growth of the class, despite headwinds from the partial loss of reimbursement from Express Scripts. Growth was also positive in China (up 46% to DKK 41 million [~$7 million]) and Europe (up 10% to DKK 821 million [~$146 million]), but sales declined by 1% to DKK 172 million (~$31 million) in International Operations and by 17% to DKK 65 million (~$12 million) in Japan/Korea. Despite the slower volume growth of the GLP-1 agonist class in recent quarters, it now accounts for 6.9% of the total diabetes care market, up from 6.7% in 2013.
    • The presentation slide deck featured IMS Health data showing that the GLP-1 agonist class slowdown is continuing. Compared to a volume growth rate of 20%-25% in early 2013, the rate appeared to be closer to 8% in August 2014. The introduction of SGLT-2 inhibitors and remnants of the incretin-pancreatitis controversy are likely two of the biggest contributors to this trend.
    • Management argued that the increasing competition within the GLP-1 agonist class should not challenge Victoza’s position as the clear market leader. Victoza’s share of the GLP-1 agonist market is currently 72%, up from 70% in 2013. Management downplayed analyst concerns about a loss of market share to AstraZeneca’s Bydureon (exenatide once weekly) and the recent launch of the new, more user-friendly pen, saying that while the new Bydureon pen may lead to some short-term turbulence in new-to-brand (NBRx) share, Victoza’s compelling clinical profile will likely be the deciding factor for most patients. Management also argued that the main effect of the upcoming launch of Lilly’s Trulicity (dulaglutide) will be to expand the overall GLP-1 agonist market rather than drawing market share directly from Victoza – Lilly’s management has voiced similar sentiments about class expansion vs. competition for share of the existing market. We will be curious to see if this forecast proves correct, as we believe Trulicity could represent the most meaningful competition Victoza has seen since it reached the market, given its very user-friendly delivery device and the convenience of once-weekly administration; it seems that we will find out sooner rather than later, as we heard during Lilly’s recent 3Q14 update that the product has already begun shipping.

Table 3: GLP-1 Agonist Market Share

 

GLP-1 Agonist Share of Total Diabetes Care Market (by value)

Victoza Share of GLP-1 Agonist Market (by value)

 

August 2014

August 2013

August 2014

August 2013

Global Total

6.9%

6.7%

72%

70%

US

8.4%

8.4%

69%

66%

Europe

7.9%

7.4%

78%

78%

International Operations

2.4%

2.6%

76%

76%

China

0.7%

0.5%

60%

69%

Japan

2.1%

2.2%

61%

73%

Other Diabetes Products

  • Sales of oral antidiabetic products largely Prandin (repaglinide) fell 2% as reported, largely due to generic competition. The impact of the product’s genericization is annualizing, and should be less of a detractor from whole-portfolio growth moving forward.

Pipeline Update

Tresiba (insulin degludec)

  • Novo Nordisk provided excellent clarity on its approach to the review and possible submission of interim data from the DEVOTE CVOT to the FDA. As Chief Scientific Officer Dr. Mads Thomsen noted, the plan was designed in response to an August 11th FDA hearing on the confidentiality of interim CVOT data as well as specific guidance derived from direct discussions with the agency. There was a lot valuable learning in this section of the presentation, both in terms of what we can expect for Tresiba’s resubmission and, more broadly, the FDA’s thoughts on interim data confidentiality.
  • The take-home message for Tresiba is that a 2015 resubmission is no longer a given, although in our view it is still quite likely. Dr. Thomsen disclosed that interim data will not be shared with management, and will instead be restricted to a small unblinded team within the company that, with guidance from the FDA, will decide whether to resubmit the NDA based on the interim data – this is one of the main ideas we heard voiced at the FDA’s hearing. The agency involvement in providing guidance to this group on whether or not to submit is key, and quite interesting in our view – it suggests that a decision to submit could bring with it slightly more confidence in the chances for approval, as the agency would have already had some input in the process.
  • One key element of the plan, included to rigorously defend the integrity of the ongoing trial, is that no data will be publicly disclosed from the interim analysis. The firewalled group and the FDA will have access to this data, and no other groups will, regardless of whether the group decides to re-file next year or to wait for the full trial to complete. During Q&A, Dr. Thomsen expressed confidence that Novo Nordisk is doing the “ultimate” in terms of what can be done to preserve DEVOTE’s integrity – this will be critical, for as we learned with the FDA’s recent approval of Orexigen/Takeda’s obesity medication Contrave (which came with a requirement for a completely new CVOT), any hint of damage to the integrity of an ongoing CVOT can have drastic repercussions.
  • Despite the complexity of this approach, during Q&A Dr. Thomsen defended the practice of using interim trial data to support regulatory filings. He suggested that the August 11th hearing revealed that proper protocols for firewalling data and maintaining trial blinding can preserve trial integrity. This was not a surprising point, as the alternative to using interim data to support earlier submission would be waiting until CVOTs were complete, which would add years on to the timeline of new type 2 diabetes drugs. Dr. Thomsen noted that the FDA appeared eager to have Novo Nordisk and other companies help communicate to the broader industry that submissions based on interim data are possible if done properly, and do not necessarily invalidate ongoing trials.
  • All in all, we see this new paradigm as a potential plus for industry. One of the more confusing elements of the FDA’s 2008 CV Guidance was the vague requirement of a “reassuring point estimate” for MACE to accompany an upper confidence interval bound that met the strict guidelines. The FDA now appears willing to work with manufacturers to discuss interim data and provide input on whether an application is worth submitting based on interim data, or if waiting for the full trial results would be more prudent. As long as the agency’s approach to these conversations is not overly cautious, it could mean that the cardiovascular safety evaluation process for new drugs is less of a black box.
  • During the call, Novo Nordisk expressed confidence in Tresiba’s cardiovascular safety and the ability of the full DEVOTE trial to prove it, although management did acknowledge that interim analyses involve a higher degree of uncertainty. Dr. Thomsen alerted call participants to the fact that, in some cases, interim analyses from trials do not support approval but the final analysis of the full trial does. As a result, he reminded participants that a decision not to resubmit in 2015 should not be seen as a signal that there is definite cardiovascular risk with the drug. Although this candor may have spooked some investors, it is important to keep in mind that management is blinded from the study results at the moment. As a result, we interpret this statement as a wise word of warning and clarity rather than a hint of worries about how the trial is actually going. 
  • In light of the strict thresholds put in place by the FDA’s 2008 CV guidance, management expressed hope that the agency’s examination of the interim data from DEVOTE in light of the surrounding dataset. That surrounding data set likely refers to the PK/PD stability and hypoglycemia benefits, among other factors. We would go even beyond that, and suggest that the agency consider the fate of the potentially transformative GLP-1 agonist/basal insulin combination Xultophy (see below) in its decision making process for Tresiba.
  • The timeline of the DEVOTE trial itself continues to progress ahead of schedule. At this point, the majority of the 7,500 planned patients have been recruited. For background, Novo Nordisk management advanced the target date for the interim analysis twice over the past six months, first from late 2015/2016 up to mid-2015 during the company’s 1Q14 update, then again from mid-2015 up to early 2015 during the 2Q14 update. As Dr. Thomsen reiterated during the call, the event rate seen in the trial has been much higher than the ~2% rate initially expected. Although he did not quantify the event rate exactly, he suggested that the event rate seen is closer to (and might even be exceeding) the rate seen in the LEADER CVOT for Victoza, in which the event rate is above 3%. The company stuck to its guidance that the full trial is likely to end between late 2016 and late 2017, for a total study duration of three to four years.

Xultophy (IDegLira – insulin degludec/liraglutide fixed-ratio combination)

  • Management expressed excitement about Xultophy’s recent European approval and the product’s strong label. The strength of the label largely draws from the results of the phase 3 DUAL program – DUAL I found that IDegLira was able to deliver strong a strong mean A1c reduction of 1.9% at 26 weeks and 1.8% at a full year, compared to -1.4% with Tresiba and -1.2% with Victoza at a full year from an average baseline of 8.3%. The data, as well as Novo Nordisk’s 3Q14 presentation slide on Xultophy, also highlighted Xultophy’s superior weight loss and lower hypoglycemia vs. Tresiba and lower rate of nausea vs. Victoza.
  • Novo Nordisk expects to launch Xultophy in the first European countries in 1H15. We have not yet had the chance to put our hands on the Xultophy device, but from an image it looks like it will be a pen or pen-like device. Xultophy was approved in Switzerland (which has a separate regulatory pathway from the EU) in September 2014.
  • In the US, Xultophy cannot be filed until Tresiba is re-filed. Novo Nordisk has not commented specifically on whether it is looking to file Xultophy soon after Tresiba, or if it may wait until a final Tresiba regulatory decision before filing Xultophy.

Obesity Pipeline

  • In late September, Novo Nordisk announced the establishment of a new obesity research center in Seattle. Although this news item was not discussed much during the call, it did lay the groundwork for some of the exciting obesity pipeline updates that the call featured.
  • Novo Nordisk listed the positive 14-to-1 FDA Advisory Committee vote for Saxenda (liraglutide 3.0 mg for obesity) as one of the highlights of 3Q14. Interestingly, the PDUFA date for the drug – October 20th – has passed, although the FDA occasionally takes slightly longer to finalize a regulatory decision. It is too early to speculate that the delay is due to challenges in finalizing labeling, but one tricky issue that arose during the AdComm was how to approve Saxenda as a first-line agent for obesity when Victoza is not approved as a first-line agent for type 2 diabetes. The labels would overlap, meaning patients with type 2 diabetes that could not receive Victoza (liraglutide up to 1.8 mg) as a first-line agent to treat their diabetes would be able to get a higher dose of liraglutide for obesity. Most panelists appeared willing to accept a degree of inconsistency between labels in order to preserve a first-line indication for Saxenda in obesity, but it was hard to gauge the FDA opinion on the matter.
  • Novo Nordisk disclosed that it has moved a new glucagon analog (G530L/NN9030) into phase 1, with the initial target being combination use with liraglutide in obesity. The trial will investigate single doses of the analog alone and in combination with liraglutide in ~160 overweight/obese but otherwise healthy male subjects, with an estimated primary completion date of October 2015. 
    • Dr. Thomsen shared Novo Nordisk’s thinking on the development and investigational indication of the glucagon analog during Q&A. He began the story with oxyntomodulin, an endogenous GLP-1/glucagon agonist that has demonstrated greater weight lowering effects than GLP-1. He raised the possibility that co-activating GLP-1 and glucagon receptors could yield double-digit weight loss, as opposed to the ~8% weight loss seen with Saxenda (liraglutide 3.0 mg for obesity).
    • Rather than pursuing a single compound that activates both GLP-1 and glucagon receptors, Novo Nordisk’s approach is to coformulate a glucagon analog with its GLP-1 agonist liraglutide. The rationale, as Dr. Thomsen explained, is that animal studies have not indicated exactly what ratio of GLP-1 vs. glucagon action is necessary. Working on a coformulation rather than a single molecule with dual activity is that a dose ratio can be optimized in early human testing.
    • Companies working on GLP-1/glucagon co-agonists include: Lilly/Transition Therapeutics (TT401, phase 2), Lilly alone (oxyntomodulin, phase 2), Xenetic Biosciences (phase 1), Hanmi Pharmaceuticals (phase 1), Zealand/BI (Zealand now has sole rights to phase 1 candidate ZP2929; BI selecting new lead candidate for partnership), and Prolor/OKPO (preclinical).
    • We are very curious if G530L would be suitable for use in applications beyond obesity, perhaps as a better hypoglycemia rescue or even in the closed loop. Zealand has expressed remarkable excitement in its aqueous glucagon analog ZP-GA-1, which is currently in phase 1 and could reach regulatory submission as a hypoglycemia rescue as early as 2017.
  • Dr. Thomsen disclosed that Novo Nordisk will begin looking into a possible obesity program for its once-weekly GLP-1 agonist semaglutide (currently in phase 3 for type 2 diabetes; see below), with the aim of possibly beginning phase 2 dose-ranging trials in 2015. During Q&A, Dr. Thomsen discussed brain scan findings that support management’s earlier claims that semaglutide has the possibility for even greater weight loss relative to Victoza. The brain scans focused on the arcuate nucleus, a center of appetite regulation, and found greater accumulation of drug following semaglutide. Weight is an extremely important factor in patients’ mind, and modest weight loss could have the dual benefits of reducing patients’ risk of complications while also improving adherence. Weight loss appears to be one of Victoza’s biggest strengths – even Lilly’s Trulicity (dulaglutide), which was the only once-weekly GLP-1 agonist to match Victoza on A1c lowering, fell slightly short of Victoza on weight loss – and an improvement over Victoza would stand to be a powerful differentiating factor in an increasingly crowded once-weekly GLP-1 agonist market.

Semaglutide

  • In line with previous guidance, the final of the phase 3 trials for the once-weekly GLP-1 agonist semaglutide began in October. The trial is comparing semaglutide to Merck’s Januvia (sitagliptin) in ~300 Japanese type 2 diabetes patients. Although there are six numbered sustained trials, we noticed a seventh phase 3 trial on ClinicalTrials.gov testing semaglutide in monotherapy or in combination with oral type 2 diabetes drugs in Japanese patients. See our table below for an overview of the SUSTAIN program.
  • See the previous section on obesity for Dr. Thomsen’s discussion of a possible obesity program for semaglutide.

Table: Overview of SUSTAIN phase 3 trials for semaglutide

Trial (ClinicalTrials ID)

Estimated Enrollment

Comparator

Estimated Primary Completion Date

SUSTAIN 1 (NCT02054897)

390

Placebo

May 2015

SUSTAIN 2 (NCT01930188)

1200

Januvia (sitagliptin)

October 2015

SUSTAIN 3 (NCT01885208)

798

Bydureon (exenatide once weekly)

July 2015

SUSTAIN 4 (NCT02128932)

1047

Lantus (insulin glargine)

September 2015

SUSTAIN 5 (NCT02254291)

306

Januvia (sitagliptin)

February 2016

SUSTAIN 6 (CVOT) (NCT01720446)

3297

Placebo

January 2016

SUSTAIN [No #] (NCT02207374)

595

One unspecified OAD

May 2016

Oral Insulin and GLP-1 Agonists

  • Phase 1 testing has been completed with two oral formulations of liraglutide, OG987GT and OG987SC – each uses a different carrier technology to protect the liraglutide molecule through the harsh stomach environment. Management did not provide guidance on if or when those molecules might move into phase 2, but did mention that both candidates were associated with statistically significantly greater weight loss than placebo.
  • Novo Nordisk’s most advanced oral GLP-1 agonist candidate is the phase 2 OG217SC, an oral formulation of semaglutide. The compound was the first oral GLP-1 agonist to reach phase 2, and Novo Nordisk has in the past suggested that it could take some small shortcuts with the compound’s development because it is already conducting an extensive phase 3 program for injectable semaglutide. The compound uses the Eligen carrier technology from Emisphere to protect the peptide as it passes through the harshly acidic stomach, allowing it be absorbed in the gut. Another oral formulation of semaglutide (OG217GT/NN9928) remains in phase 1, and was not mentioned during the call.
    • All in all, Novo Nordisk has kept four discrete oral GLP-1 agonists candidates in its pipeline. This represents quite a commitment, but should yield in excellent choice in terms of eventual compound selection for phase 3.
  • According to Novo Nordisk’s 1Q14 update, the company plans to advance its primary oral insulin candidate OI338GT (NN1953) into phase 2a in 2015. Novo Nordisk has completed three phase 1 trials on the candidate. Earlier this year, Novo Nordisk discontinued development of its two other oral insulin candidates, OI287GT and OI362GT. OI338GT is a novel long-acting basal insulin analog that uses Merrion’s GIPET carrier system to carry it from the GI tract into circulation. OI338GT’s phase 1 testing included three clinical pharmacology studies in a total of 118 healthy volunteers and people with type 2 diabetes, supporting its safety and dose-dependent glucodynamic effects. According to the company’s 1Q14 update, Novo Nordisk is planning for OI338GT’s phase 2a trial to be a proof-of-principle trial investigating glucose lowering and safety, including rates of hypoglycemia, after individual titration of OI338GT in people with type 2 diabetes. Based on results of this phase 2a trial, larger phase 2b proof-of-concept studies may be initiated. We haven’t heard much on the oral insulin program since that 1Q14 update.
    • Other companies developing oral insulins include Biocon (phase 2), Oramed (phase 2), and Aphios (preclinical), among others. Novo Nordisk has the advantage of having the most experience in insulin and a large amount of resources to support this project – in 2013, management estimated that $3 billion might be required to fully develop oral insulin and oral GLP-1 agents. Although not quite an insulin tablet, MannKind’s inhalable insulin Afrezza was recently licensed by Sanofi.

Long-Acting Insulin

  • Management shared during the call that a new long-acting insulin analog (LAI338/NN1438) has entered clinical testing. The first phase 1 trial, currently recruiting, is expected to include around 70 healthy volunteers and type 1 diabetes patients. It has an estimated primary completion date of April 2015. Novo Nordisk has not shared what the target dosing for this insulin is, but we imagine once-weekly is a definite possibility.
    • Novo Nordisk has another long-acting insulin (LAI287/NN1436) in phase 1, having began a second phase 1 study testing once-weekly dosing in 2Q14. The study (ClinicalTrials.gov Identifier: NCT02148861) will enroll around 48 type 2 diabetes patients, comparing LAI287 to Tresiba. It is currently recruiting, and has an estimated primary completion date of March 2015.
    • We have not heard much from management about this candidate, and in general the once-weekly insulin competitive landscape is not yet too crowded. In June of last year, PhaseBio announced that it had initiated phase 1 testing of its once-weekly basal insulin Insumera (PE0139), while AntriaBio is advancing its PEGylated basal insulin AB101 into clinical studies.

Other R&D Highlights

  • The day following the 3Q14 update, Novo Nordisk announced a ~$130 investment in new diabetes research laboratories in Denmark. The new facilities will be located in the company’s existing Måløv R&D campus, which currently accommodates around 2,300 employees. The expanded facility, expected to come online in early 2016, will provide room for around 350 additional employees. 

Questions and Answers

Big Picture

Q: Do you see a change in the overall design of contracts – in terms of duration, in terms of price protection, whether we’re seeing more exclusivity?

A: It’s not a question that’s easy to answer briefly because for each segment in which we operate, the contracting environment is slightly different. There’s a big difference whether you talk fast-acting insulin, basal insulin, GLP-1, etc. I assume the question mostly relates to modern insulins, and in the modern insulin space, we’ve always been a believer in having free choice for the doctor and the patients. We support access to more than one product in each category and that’s what we’ve been striving for. It’s quite clear that from the pharmacy benefit manager side, there has been a desire to have more exclusive contracts and increased rebates. We try to maneuver in this landscape the best we can and have done so for many, many years, and of course we want to have good longstanding relationships with both pharmacy benefit managers and managed care organizations. And we want to remain competitive, and as a consequence of this, you have seen that some contracts in segments where there are very similar products (such as the fast-acting insulin segment), there is a tendency to have longer contracts, higher rebate levels, and more price protection. In other areas where the products are more uniquely differentiated, that’s less so. Getting more specific than that would take it too far.

Levemir

Q: What net US insulin pricing is reflected in the guidance for 2015? That question is in light of Sanofi’s commentary this week – they believe a price war is beginning there. Can you give any color on the net price differentials, post-rebate, between Levemir and Lantus in the US? Clearly there was some differential at launch; does that still exist and has it narrowed at all?

A: The assumption for US pricing is a modestly positive to flat impact from prices in the US. We’ve noted in the nine-month results that we’ve had a positive pricing impact in 2014. And hence in 2015, we’re anticipating a slight positive to neutral impact from prices on our gross margin in 2015.

This is a complex topic with many moving parts; at the end of the day, it is the net ARP per unit of insulin that really drives the business. And that of course is a combination of the product mix, the effect of pricing, the effect of all the different rebates in the US. Some of these rebates are related to the government (e.g., Medicaid) and some are related to managed care, and with the Affordable Care Act, there have been changes in some of those rebates. And of course, there’s the whole negotiation with pharmacy benefit managers, where individual customers have individual contracts, and that can also affect the net pricing. So getting into a very detailed discussion about the specifics of every moving part would be very complex, and would not necessarily lead to a better overview than the one we just provided. I can’t really comment with regard to the specific question on Levemir and Lantus. I can only say that with regard to Levemir, we have over the last years seen increases in the list price, we have also seen increases in the net ARP, but we have also seen increases in absolute rebates.

Q: Sanofi recently highlighted that they think they had sales force deficiencies in the US. So going into 2015, with your competitor potentially putting those right, do you think you might need to increase the spend behind the basal insulin sales force in the US to head off further competition there?

A: It’s probably most polite not to comment on the performance of their sales force, but I can comment on the performance of our own sales force this year. I’m very happy with the performance of our sales force in the space of basal insulin. We are very happy with the market share gains we have been seeing consistently on Levemir over the last couple of years. This continues as we speak. We have upgraded the device from the FlexPen to the FlexTouch. As I mentioned before, FlexTouch is a unique pre-filled device, which is based on a spring mechanism with high-quality springs that ensure a self-injection once you touch the push-button. This is new in the space of pre-filled devices. It is more convenient and we think it is further strengthening the attractiveness of Levemir in the US marketplace. So I don’t have any plans to dramatically change the size and the way our sales force works. We’re very happy with the way it’s going right now.

Tresiba

Q: Should we expect Novo to announce that they’ve decided to file Tresiba? Or will that communication to the investor base happen only after the FDA’s decision on whether to accept the filing?

A: On the day when we hear from the unblinded team that has discussed the situation with the FDA at a pre-NDA meeting, once we are informed by them – not about any data or any other outcomes but on whether they will submit or not – this is the communication that will be given to the market. And it will likely be supplemented with information if and when the FDA has acknowledged the receipt of the dossier and that they will review it with a six-month time span. We would expect to communicate that as well.

Q: We all know the FDA’s caution regarding integrity and interim data, but has this been driven solely by that or also by an increase in concern at the FDA about the results of interim analyses? Do you see a smaller likelihood of Tresiba being submitted based on an interim analysis?

A: We can put to rest the notion that interims are dangerous to do and you would not be able to submit interim data because you would taint the rest of the study. I think the August 11 meeting revealed that if you do your work diligently, with the right firewalls, with the right blinding, which is the case for DEVOTE, and you negotiate all of that with the regulator, you will actually be in good shape regarding preservation of the integrity of the remainder of the trial. And this is totally reflected in the communication between the Agency and Novo Nordisk. The Agency is also suggesting that companies like us should actually contribute to informing the market that indeed one can do interims without this negating or invalidating the rest of the trial. We would also like to state that when you do an interim, it doesn’t mean that the interim necessarily will be robust enough to actually reassure the agency about either the confidence interval (which is easier to do because that’s very binary) or what constitutes a reassuring point estimate.

Is there now a smaller likelihood of approval? I would say that an aggregate risk has now been split into two. If you recall the theoretical scenario that Novo Nordisk put forward one quarter ago, where we basically said that we intended to submit the interim analysis as part of a Class 2 resubmission of the NDA to the Agency in the first half of next year. That incurred, by definition, a risk that the point estimate would not be reassuring, even to Novo Nordisk. But we would have at that point decided to do so, to get it through the Agency, and then postpone the risk of what could be a biasing of the residual trial until the very end of the FDA action period. Now the FDA didn’t quite agree with that notion. They’re basically saying, let’s have a discussion. Let’s look at your data. And if we feel you should submit this data, that is when you submit and that’s why the unblinded team is meeting with the FDA. And that basically also means that now you have the risk that we will not submit based on the interim, but when we submit, it will have been discussed with the Agency and the Agency at that point has actually been in concordance with the notion that it should be resubmitted. So the overall risk, short answer, is the same.

Q: You said you received some specific guidance from the FDA on the DEVOTE interim analysis. Could you provide any color on that, and specifically, do you have any view of what the point estimate and upper end of the confidence interval for that interim data should be?

A: Many of us heard the August 11 FDA meeting about the general use of interim analysis in cardiovascular outcomes trials, and we have had more discussion with the Agency that specifically relates to the DEVOTE trial. And quite frankly, I think when you do things the way that we have now decided and designed the study to address the firewall between the very small unblinded team and the rest of the company, etc. that we have done the ultimate that can be done to preserve the integrity of the entire trial, regardless of the conduct of the interim analysis. I think it’s safe to say that the Agency is in full agreement with this. So that means we have a path forward of exactly how to handle this, and exactly which data are to be submitted. As you know, it’s not only a small dataset compared to the full trial, it’s also only a small subset of the full trial data that the FDA would desire to see at this point.

About the numbers, a number is never just a number. The number should be seen in the context of its surroundings. And that means that inasmuch as the 1.8 upper bound of the 95% confidence interval as per the December 2008 guidelines is kind of carved in stone, the point estimate that will be considered reassuring should be seen in the light of the surrounding dataset. They will also be receiving serious adverse events and other things they might have asked for, and hence I can’t really respond at this point. We’re simply looking forward to what will happen in the first half of next year.

Q: What are you seeing in terms of the CV event rate for patients on Tresiba in Europe and Japan?

A: The lines of evidence as to cardiovascular safety that go beyond just what we saw in the BEGIN program, at least threefold. One is, as you allude to, the Japanese and European performance, i.e., the pharmacovigilance systems. What is in our system? I can say it’s actually not much. It’s very difficult to pick up anything unless you have very dramatic signals from such databases. But indeed, there is no such signal at all. We’ve also done the first Periodic Safety Update Report, the PSUR, with Europe without giving rise to any new cause for concern. As regards the overall program, we have today more than 22 studies involved in the phase 3b and phase 4 program, and when we put all those together, I can say that there’s no either new or changed cause for concern from those studies at all. In Japan we’ve also committed to do, as you always do for new molecules, a kind of safety monitoring study in a few thousand patients for a couple of years after approval, and that is still ongoing, so I have no findings from that one.

Q: What are your expectations for the event rate in the DEVOTE study itself?

A: The original planning of the study included a 2%-plus annual MACE event rate, and here we’re talking MI, stroke, and death. That is now obsolete, and what we’re witnessing is what I would call a more LEADER-like situation, and LEADER is 3%-plus in MACE annual event rate. I cannot be more specific than that because it’s still early days. But this is of course underlying the basis for the upgraded guidance as to the duration of the study that we’ve given you over the last six months or so. These patients in general have baseline demographics that are slightly more sick than those in the LEADER trial. So the best estimate I can give you is LEADER or something maybe slightly exceeding LEADER. That’s what we would expect. We have no data or the data we have at this point is still uncertain.

Q: I just want to be absolutely sure that you say in the first half of next year, you will make the decision to resubmit. When you make the decision, are you then ready to push the button right away so you also make the submission in the first half and have a launch in early 2016? Is that still your base case scenario?

A: Yes, we will be in a position to make a decision to file, and we will file in the first half of 2015 if the data are convincing.

Victoza

Q: I think we’re seeing the beginnings of some share loss to Bydureon’s new pen. Is there anything you can do there to head that off? And what do you expect on Victoza going forward? Clearly with dulaglutide coming, this suggests that Victoza is going to be under some pressure. What are you expecting in 2015?

A: In the GLP-1 space, Victoza is the gold standard. I don’t think there’s going to be any change in the perception about the relative strength of the products since it has been proven over and over again that Victoza has a superior clinical profile. However, every time there is a launch of a new product, you will see NBRx data react quite violently due to the nature of new launches and sampling. With the launch of the new Bydureon pen, I don’t think we should expect any major changes in market dynamics. With the launch of Trulicity, hopefully we’ll see an increase in the total volume of GLP-1 use in the US. GLP-1 is a very attractive therapy for type 2 diabetes and if our competitor is able to bring more patients into the GLP-1 segment, that would only be good for patients. So we expect that Trulicity will expand the market more than directly take market away from Victoza.

Saxenda

Q: What’s the current situation with Saxenda? You’ve obviously had the discussion with the FDA and the PDUFA date has passed.

A: Since the 14-1 AdComm on September 11, we have been in a close and constructive dialogue with the Agency. That’s basically what we can say and we hope we’ll be able to get back to you sooner rather than later with the FDA action.

Other Products

Q: Can you give us a bit more detail on the new glucagon analog? What are its advantages and why isn’t it being developed first in diabetes?

A: Glucagon has an extremely short half-life, a matter of a few minutes. What we’ve managed to come up with is a physically, chemically, thermally stable analog with a very prolonged half-life. Why would you do that? Because it has been demonstrated from the experiment of nature, oxyntomodulin, where cells in the blood actually produce a precursor stage of GLP-1 that binds both to the glucagon and the GLP-1 receptor, albeit with different affinities. And this oxyntomodulin has documented a greater weight-lowering effect than native GLP-1. This has given rise to the speculation that if you tweak the dose of glucagon (in the low end) with a therapeutic dose of GLP-1, you’ll be in a very good situation and have a synergistic effect giving greater weight loss, maybe double-digit weight loss as compared to 8% with Saxenda. So the reason we’re not creating Saxenda/oxyntomodulin is that we as a company believe that the data we have seen in animals have given rise to a lot of confusion with regard to the ratio you need. So our approach is making a ratio where we define the ratio after human experiments. Right now the human being is being used as the ultimate experimental animal, to define the exact ratio of the fixed-ratio combination, which we’ll update you on if and when the trial pans out successfully.

Q: You alluded to the idea of semaglutide’s possible development in obesity. Can you update us on that?

A: We have data that may explain the enhanced weight loss that semaglutide has demonstrated both in animals and humans. And the data basically suggests that at the level of the appetite control center in the arcuate nucleus, we seem to have greater accumulation of semaglutide than Saxenda. And this would explain a more sustained and pronounced efficacy. We’ll look into that later this year, with the aim of potentially starting phase 2 dose range-finding trials next year.

 

-- by Emily Regier, Manu Venkat, and Kelly Close