Novo Nordisk 4Q14 – Diabetes care up 13% to $3.2 billion in 4Q14, up 7% to $12 billion for full year, riding Levemir and Victoza’s rebound; DEVOTE interim analysis beginning – January 30, 2015

Novo Nordisk reported its 4Q14 and full-year 2014 update in a call led by CEO Mr. Lars Sørensen. Total diabetes care sales grew 13% as reported to DKK 19.3 billion (~$3.2 billion) in 4Q14 and grew 7% as reported to DKK 70 billion (~$12 billion) for the full year 2014. Levemir (basal insulin detemir) was a strong performer throughout the year in the US, benefitting from the reallocation of promotional resources intended for Tresiba (next gen insulin degludec that should have been approved at FDA in 2013); sales grew 17% as reported in 4Q14 and 23% as reported for the full year for totals of DKK 3.8 billion (~$620 million) and DKK 14.2 billion (~$2.5 billion) respectively. Although Victoza (liraglutide) was also a driver of whole-company growth for much of the year, it sputtered during 1H14; growth for the full year was 15% (to DKK 13.4 billion, ~$2.4 billion) vs. 23% in 2013. However, Victoza appears to be picking up steam, posting 4Q14 growth of 24% as reported; management suggested that the end of the incretin-pancreatitis issue and new product approvals should help the overall GLP-1 agonist class rebound. Although we expect Victoza to retain the lion’s share of the once-daily GLP-1 agonist market, we expect that segment of the market to be squeezed by new once-weekly products such as Lilly’s Trulicity (dulaglutide) in particular (potentially GSK’s Tanzeum [albiglutide] as well). Looking ahead, Novo Nordisk forecast sales growth of 6%-9% in local currencies (18%-21% as reported) driven again by modern insulins and Victoza but with greater contributions from Tresiba and some help from launches for Saxenda (liraglutide 3 mg for obesity) and Xultophy (insulin degludec/liraglutide).

Turning to the pipeline, we received the exciting and long-awaited news that the DEVOTE cardiovascular outcomes trials (CVOT) for Tresiba has completed enrollment and that enough events have accrued for the interim cardiovascular safety analysis. As was disclosed during the company’s 3Q14 update, the plan now is for a small firewalled group within the company (with FDA input and guidance) to review the data and decide whether to resubmit the NDA or wait for the full trial to finish – the forecast for that milestone was moved up slightly to 2H16. To whet our appetites for the day when Xultophy makes it to the US, management shared topline data from the DUAL V trial showing greater A1c reductions (-1.8%) from baseline (8.4%) in type 2 diabetes patients who switched from Lantus to Xultophy vs. patients who stayed on Lantus (-1.1% from a baseline of 8.2%).  Pretty remarkable! We continue to wonder if more patients should be moving directly from metformin to this compound and we are wishing this and SGLT-2s were included in GRADE. These improvements also occurred in the context of less hypoglycemia and weight loss rather than weight gain. We also received updates on Saxenda’s launch, a new phase 1 amylin agonist for obesity, new topline data on faster-acting insulin aspart (FIAsp), and a discontinuation of one candidate in the oral GLP-1 agonist program, among other R&D highlights.

Editor's Note: We will return to this report soon to add highlights from a Tuesday investor presentation in London.

Financial results summary for Novo Nordisk’s major diabetes products in 4Q14

Financial Highlights

• Novo Nordisk’s diabetes portfolio posted sales of DKK 19.3 billion (~$3.2 billion) in 4Q14, representing year-over-year growth of 13% as reported and 9% in constant currencies. As reported, this makes 4Q14 Novo Nordisk’s strongest quarter of 2014 in terms of diabetes sales growth, although 3Q14 was not far behind at 11%. This is due largely to an easier comparison for 4Q14: reported year-over-year growth was 4% in 4Q13 compared to an average of 12% in the first half of 2013.
• For the full year 2014, diabetes care sales grew 7% as reported and 9% in local currencies to DKK 70 billion (~$12.4 billion). This was squarely in line with guidance from the beginning of 2014 for reported growth of 4.5%-7.5% (8%-11% in local currencies). The first half of the year was much slower in terms of reported growth (~2%) than the second half of the year (~12%). Performance for the full year 2014 was roughly comparable with performance for the full year 2013 (growth of 8% as reported vs. 7% in 2014), although in 2013 growth was much stronger in 1H13 (12%) than 2H13 (3%).
  
  • By product, Levemir (~55%) and Victoza (27%) comprised the largest share of whole-company growth for the full year 2014. Tresiba comprised an 8% share of whole-company growth; despite somewhat slow uptake in some markets and a non-presence on the US market, it is encouraging to already see Tresiba contribute meaningfully to whole-company sales growth and this bodes well for Tresiba’s contribution in 2015.
  • By geography: North America was the primary driver of growth for the full year, comprising a 61% share of growth based on 11% growth as reported. In North America, insulin growth was driven by a positive contribution from price (we imagine this will be increasingly rare moving forward given increased competition and price pressure) and also driven by continued market share gains for Levemir. These drivers worked against a headwind created by expanded Medicaid and Medicare Part D utilization. Japan/Korea and Europe posted essentially no growth (surprising given Tresiba’s strength in Japan), while China and International Operations were in between (see table below).

Whole-company sales and growth by geography for the full year 2014

• A paragraph in the press release on sales and distribution costs in 2014 (which were roughly flat) mentioned sales force investments in the US, China, and selected International Operations countries but also lower promotional spend in the US and Europe. The reasons behind these particular dynamics are not entirely clear, but movement into the new field of obesity could have been a cause for a ramp-up in sales representatives.
• R&D costs rose 17% as reported in 2014 to DKK 13.8 billion (~$2.3 billion). This was largely due to the phase 3 SUSTAIN program for semaglutide, clinical trials for Tresiba (the DEVOTE CVOT was specifically included), the phase 3a program for faster-acting insulin aspart (FIAsp), and a phase 2 trial for oral semaglutide.
• For 2015, Novo Nordisk forecasts whole-company sales growth of 6%-9% in local currencies (18%-21% as reported). The financial summary suggests that growth will mainly be driven by modern insulins, Victoza, and Tresiba as well as modest contributions from the launches of Saxenda and Xultophy; that list includes the majority of Novo Nordisk’s diabetes portfolio and is not altogether surprising. These drivers will be countered by increased rebate levels in the US that are part of the continuing battle for access to narrowing formularies (prices have just been highest in the US so there is more to lose there), intensifying competition in diabetes (largely SGLT-2s as well as other more advanced GLP-1s – we think this will ultimately be offset by patients living longer, which should help insulin manufacturers over the long run), and macroeconomic conditions in certain markets in the International Operations segment.
  
  • While the new 6-9% 2015 forecast may have once seemed somewhat conservative of a forecast, it compares very favorably to Sanofi’s forecast for flat sales for its diabetes portfolio in 2015. Lilly in its 4Q14 update the same morning did not provide a forecast specifically for diabetes. Much of the pressure coming into 2015 appears to be from the price pressure in basal insulin: Levemir’s share of Novo Nordisk’s whole-company growth slowed from ~80% in 2Q14 to ~25% in 4Q14.

Modern Insulin Summary

• Modern insulins (NovoLog, NovoMix, and Levemir) collectively grew 10% as reported and 6% in constant currencies in 4Q14 to DKK 19.3 billion (~$3.2 billion). For the full year 2014, modern insulins grew 9% as reported to DKK 41.5 billion (~$8.2 billion). The modern insulin category does not include Novo Nordisk’s newest insulin, Tresiba (insulin degludec) – they are calling this “new-generation insulin,” partly in an effort to differentiate it. See below for a detailed breakdown of the 4Q14 and full-year 2014 results for each of Novo Nordisk’s modern insulins.
• During Q&A, management acknowledged that price increases benefitted all three modern insulins in 2014 and forecast single-digit price increases in 2015. We were a bit surprised that any price increases (especially in aggregate) are expected at all. Importantly, Novo Nordisk shared that it does not forecast any major contract changes on the level of the loss of the Express Scripts rapid-acting insulin and GLP-1 agonist contracts in 2014. We aren’t sure that management could have forecast the Express Scripts loss, so we aren’t sure what to make of this, though presumably, management has worked hard to get to a point where it can say that major surprises are not expected.
• During Q&A a caller suggested that Sanofi might be trying to increase its promotional investments coming into 2015. Novo Nordisk management responded that the company has not noticed any significant changes in the promotional landscape. Given the immense pressure on Sanofi, especially following its recent forecast for flat diabetes sales growth in 2015, we have a hard time imagining that the company will not try to find ways to stem the slow bleed in share from Lantus to Levemir. That said, in recent presentations Sanofi management has highlighted efforts to increase sales force targeting and productivity with Lantus rather than efforts to increase sales force sizes or promotional investments. Additionally, much of the change in share is likely pricing-oriented – while we know Novo Nordisk has been spending more on Levemir, we would be surprised if the majority of Lantus losses stemmed from this. We’re thinking it’s more likely that at least some losses are stemming from contracting changes and pricing reductions that Novo Nordisk has accepted on behalf of Levemir – that’s speculation on our part. However, the focus will likely shift away from the Lantus-Levemir dynamic as both Sanofi and Novo Nordisk launch their next-generation basal insulins over the coming year or two and as Lilly enters the basal market – that last will be very important to watch. Can they expand the market? We would certainly think so, though it will be important (and fascinating) to watch how average pricing declines.

Levemir (insulin detemir)

• Levemir was once again a star performer in 4Q14, growing 17% as reported and 13% in constant currencies to DKK 3.8 billion (~$620 million). This was against a fairly standard comparison, as 4Q13 growth as 18% as reported (same as for the full year 2013) and just a few million DKK shy of Levemir’s all-time sales peak in 3Q14. Levemir’s share of the US modern basal insulin market stands at just under 25% as of November 2014, up from under 20% in November 2011. Device penetration for Levemir is holding steady at ~70%, where it has been for roughly two years, although the new FlexTouch pen (unveiled at ADA 2014) may be improving mix within the device-using segment.
  
  • For the full year, Levemir sales rose 23% as reported and 25% in constant currencies to DKK 14.2 billion (~$2.5 billion). Truly an exceptional 2014 showing for a product that has always played in Lantus’ shadows; while Sanofi’s Lantus still dominates  the basal insulin’s massive shadow over Levemir is now slightly smaller. That said, we are not sure at what cost, i.e., we don’t know how much more Novo Nordisk is spending on marketing or how much lower the aggregate price of Levemir is – we are assuming there has been significant downward movement on both fronts but again this is speculation. Levemir’s success was certainly helped along by the FDA’s CRL for Tresiba in 2013 (which of course had a net very negative effect on the company in aggregate), the resulting re-allocation of Tresiba sales representatives to Levemir, the launch of a new FlexTouch pen, and (perhaps biggest in the news recently) a willingness to be more aggressive on rebating and to cut prices overall in order to gain more volume. The latter put pressure on Lantus, but Sanofi made what we think was a wise move to prioritize access over price. For both Sanofi and Novo Nordisk, the room for increases in price from this point forward is far more limited than in days past, especially with their more innovative new basal insulins (Tresiba and Toujeo [insulin glargine U300]) on or near the market. As another silver lining, once a manufacturer has claimed a single-source formulary their leverage once again increases because payers are hesitant to switch suppliers and inconvenience patients. On the positive side, there is volume grown coming from more patients being diagnosed (going on basal insulin will come downstream for those that go on basal insulin), and more patients staying older longer and going on basal insulin – on the other side, more patients are trying more orals and more oral combinations (and, some oral/GLP-1 combinations, which could help Novo Nordisk).

NovoLog/NovoRapid (insulin aspart)

• NovoLog sales grew 8% as reported and 4% in constant currencies in 4Q14 to DKK 4.8 billion (~$810 million). This came against a very easy comparison as sales grew only 4% in 4Q13. Sequentially, sales grew 8% as reported in 4Q14. Despite the somewhat slow growth, 4Q14 represented a new sales peak for NovoLog coming off a previous peak in 3Q14 of DKK 4.5 billion (~$800 million). Even so, the results were not particularly strong for Novo Nordisk’s former flagship and as a result we were not surprised to see very little mention of NovoLog during the presentation. Lilly’s Humalog also had a challenging quarter in 4Q14, falling 1% to $730 million.
  
  • For the full year, NovoLog grew 4% as reported and 5% in constant currencies to DKK 17.5 billion (~$3.1 billion). As with 4Q14, this figure represented an all-time peak, and compares to year-over-year reported growth of 7% and total sales of DKK 16.8 billion (~$3.0 billion) in 2013. .
• We have been seeing a number of newer drug classes positioning themselves as safer alternatives to the intensification of basal insulin therapy than rapid-acting insulin. GLP-1 agonists in particular are being positioned as a “smarter” (more glucose responsive) and safer (less hypoglycemia) way to intensify basal insulin treatment, with additional weight advantages. SGLT-2 inhibitors also deliver strong postprandial glucose efficacy with the advantage of oral administration, low hypoglycemia, and weight gain. These competitors will be counteracted by growth in the patient population that is older and/or has longer-standing diabetes that exhausts non-insulin treatment options, and applications like the closed loop could also breathe renewed life into the rapid-acting insulin segment.

NovoMix

• NovoMix continued its lukewarm trajectory, growing 3% as reported and remaining flat in constant currencies in 4Q14 to achieve sales of DKK 2.6 billion (~$440 million). For the full year 2014, NovoMix grew 1% as reported and 4% in constant currencies to DKK 9.9 billion (~$1.8 billion). Nearly half of NovoMix sales come from International Operations and China, a much higher percentage than for NovoLog and Levemir. The premixed insulin market has grown more slowly than the rapid-acting and long-acting segments (CAGR of 3.5% for premix vs. 6.0% for rapid-acting and 7.1% for long-acting according to IMS Health data in Novo Nordisk’s supplemental slides). However Novo Nordisk should also benefit from the introduction of Ryzodeg (insulin degludec/insulin aspart, aka Tresiba/NovoLog) starting now in markets ex-US.  

Tresiba (insulin degludec)

• The new ultra-long-acting basal insulin Tresiba posted global (but ex-US only as the FDA gave it a complete response letter in early 2013) sales of DKK 262 million (~$44 million) in 4Q14, up 50% sequentially from DKK 175 million (~$30 million) in 3Q14. Tresiba sales for every quarter from the beginning of 2013 could be found in the full financial information document. The growth trajectory for Tresiba appears promising; although some would argue that revenue should be higher two years following its introduction, we point out that the countries in which it is approved are not known to be major funders of diabetes therapies. Part of the reason may also be that Novo Nordisk has not shied away from pricing Tresiba at a premium that it believes is proportional to the product’s clinical benefit – CEO Mr. Lars Sørensen has previously spoken emphatically on the need for markets other than the US to support the price of innovation – we agree though are not optimistic on this front. 
  
  • Management displayed an updated version of a now-familiar chart showing Tresiba’s launch trajectory in different markets. Tresiba has now been launched in 23 countries, most recently Brazil, Russia, Italy, Chile, and Slovakia. The continuing theme is that uptake is quite good in markets where Tresiba is reimbursed similarly to Lantus. In Japan, for example, roughly 20 months from launch, Tresiba holds a 26% value share of the basal insulin market (25% in Switzerland 20 months after launch, 15% in Mexico 20 months after launch). However in markets without reimbursement, uptake has been minimal (2% in Denmark and the UK more than 20 months post-launch). We see the major value of Tresiba in the combination with Victoza and believe this combination drug will do incredibly well commercially once approved in more places. That this drug has effectively been delayed multiple years in the US is very disappointing from our view.  
• Ryzodeg, a premixed insulin including insulin degludec and insulin aspart (essentially Tresiba/NovoLog), has launched in India and Mexico. We saw data presented at EASD from a phase 3 trial in which Ryzodeg barely missed achieving statistical non-inferiority vs. basal-bolus therapy with Tresiba and NovoLog, although the A1c reductions were numerically similar. Even though one could make the case for Ryzodeg from a convenience perspective and perhaps from a hypoglycemia perspective, as noted, in our view, the most exciting combination that Novo Nordisk has to offer is Xultophy, which combines Tresiba with Victoza rather than NovoLog. 
• There was big news from Chief Scientific Officer Dr. Mads Thomsen on the DEVOTE CVOT for Tresiba, which is gating for US resubmission. See the pipeline highlights section below for more details.

Victoza (liraglutide)

• Victoza showed signs of a continued rebound following a bit of a sluggish 1H14: sales in 4Q14 grew 24% as reported and 19% in constant currencies to DKK 4.0 billion (~$680 million). This compares to 9%, 6%, and 21% year-over-year growth as reported in 1Q14, 2Q14, and 3Q14 respectively. For the full year 2014, Victoza grew 15% as reported to DKK 13.4 billion (~$2.4 billion) – if sales hold at that level, this makes Victoza nearly a $3 billion product. As for 2014 weakness, Novo Nordisk attributed the full-year results to lower GLP-1 agonist volume growth as well as the Express Scripts contract loss.
  
  • IMS Health data included in the slide deck painted an improving – or at least stabilizing – picture for the GLP-1 agonist class. Following a period of 20%-25% growth in early 2013, volume growth for the class (moving average total) plummeted to a low of roughly 8% earlier this year. Since then, it has stabilized and appears to potentially be picking up: as of November 2014 volume growth was at 10%. While it is hard to interpret much from such a modest upswing, during Q&A management described some concrete reasons why a rebound should be on its way: (i) the launch of new drugs in the class, especially those broadening the list of once-weekly options (although these represent competition for Novo Nordisk, we do expect them to expand the market); and (ii) the diminution of the incretin/pancreatitis controversy that peaked in 2013.
  • The same IMS Health slide showed that Victoza appears to be maintaining its share of the class (for now) despite recent launches of new agents in the class. As of November 2014, Victoza was holding steady at a 66% share of total prescriptions in the class, while AZ’s exenatide franchise (Byetta and Bydureon) was at 33%. GSK’s Tanzeum/Eperzan (albiglutide) and Lilly’s Trulicity (dulaglutide) were only just emerging on the chart at that point. Management noted during Q&A that new once-weekly agents will not entirely swallow the once-daily class, and there should be a “strong” part of the market that will prefer once-daily options.  Once-daily administration may indeed be better for some patients, especially those using a combination product like Xultophy or those needing more postprandial glucose control. However in our view, over the coming year it is likely that Victoza will lose share to the emerging players while the class expands – we would be surprised if Victoza didn’t continue to grow, but perhaps at a slightly slower pace as monotherapy. Then again! Depending on where and when Xultophy (IDegLira) is launched – we expect that product to be supremely successful and this will obviously contribute to Victoza revenue (we’d love to learn more about exactly how this will happen). Regarding the competitive landscape for GLP-1, we learned during Lilly’s 4Q14 update that Trulicity is being priced roughly on par with Victoza and has even gained positioning on Express Scripts’ formulary (which Novo Nordisk cannot claim for Victoza). Notably, Novo Nordisk management continues to reiterate the fact that Victoza achieved statistically significantly greater weight loss than Trulicity – we think this will be one won more through the formulary wars than anything in the US and won’t actually be as much subject to patient or provider preferences. 
  • The conversation during Q&A also covered GSK’s Tanzeum, or at least a hypothetical Tanzeum-like product with relatively low efficacy, lower cost, and low side effects. Management argued that efficacy is the prime consideration for patients and providers, both in terms of glucose control as well as parameters such as weight. Eschewing the use of Tanzeum’s name, management suggested that there is unlikely to be appetite on the market for a low-efficacy discount product. We expect to gain more clarity on Tanzeum’s early performance during GSK’s 4Q14 update on Wednesday, February 4^th. As noted, we believe success of where a product lands on formulary will play a major role.    

Pipeline Highlights

Tresiba (insulin degludec)

• Excitingly, we learned that the DEVOTE CVOT has completed enrollment and has accumulated enough events for an interim analysis that could enable FDA resubmission. Enrollment has ended at 7,644 type 2 diabetes patients. The availability of interim data occurred well within the company’s previously stated timeline, and enables the possibility of an FDA class 2 resubmission in the next five months (1H15), opening the possibility for possible US approval by the end of the year. Novo Nordisk now expects the full DEVOTE trial to end in 2H16 as opposed to previous guidance for October 2016 – October 2017. We have consistently been impressed over the past year with the movement in the timeline for DEVOTE. The event rate appears to have been higher than expected (~3% or higher rather than the ~2% rate initially estimated). Though obviously negative from a patient perspective, this did help accelerate key trial milestones. We do wonder how “generalizable” these results are for the “average” person with diabetes, particularly if the higher event rates represent patients who are “sicker” than the average patient.
• As a reminder, it is no longer a given that Novo Nordisk will decide to resubmit Tresiba to the FDA based on interim results from DEVOTE. As Chief Scientific Officer Dr. Mads Thomsen reiterated during prepared remarks, access to the interim data has been restricted to a small firewalled team within the company that (with FDA input) will examine the interim data and decide whether to resubmit the NDA for Tresiba or to wait for the full DEVOTE trial to be completed. This new plan was largely the outcome of an FDA hearing in August on the confidentiality of interim results from CVOTs, as well as direct discussions between Novo Nordisk and the agency. As part of the new paradigm, interim data will not be publicly disclosed even if the NDA is resubmitted – the only news we will receive is whether the resubmission occurs or not. That is positive in that patients won’t have this data to decide whether or not to stay in the trial; that said, if interim data is submitted and the drug were approved, we would assume that would affect the sanctity of the trial. While some patients may stay in the trial due to getting free drugs, etc., we have to assume some patients would leave the trial and try to get on the approved drug.  
  
  • While this new paradigm appears to be a plus for industry, as it better preserves the integrity of CVOTs while still allowing for approvals based on interim data, it is unclear the impact on trial integrity. The fact that this compromise was reached also indicates a reassuring willingness to reconsider the implementation of certain parts of the 2008 FDA Guidance on the FDA’s part. However, we note that the new plan to preserve confidentiality of interim CVOT data addresses only one of many potential issues with the CV Guidance, and we hope the conversation continues beyond what we heard at the August hearing. Specifically, we hope that FDA is able to review the CVOT requirements in 2015 to be able to better advise companies on what is in the best interests of patients. Currently, we believe that significant resources are being spent that may be spent better elsewhere and that represent barriers for companies and researchers to address diabetes.
• There is a chance, of course, that the FDA may take a more conservative approach when evaluating the interim DEVOTE data (if submitted) because the full trial will be completed one to two years from now. This possibility came up in a question during the Q&A session, but Dr. Thomsen noted that the FDA is a science-based agency that at least theoretically should only consider the submitted data without taking the trial time course into account. One slight advantage for Novo Nordisk with the revised resubmission plan is that the firewalled team will work closely with the FDA to determine whether to resubmit the NDA. Therefore if the FDA does not see the interim data package as approvable, it may advise the firewalled team to not resubmit the package, which would at least let Novo Nordisk avoid the consequences of a second CRL. Ideally, of course, the DEVOTE data will be fully reassuring and pose no barrier to approval. However, the 4Q14 document does acknowledge that interim analyses are inherently riskier and more unpredictable than full trial results, and that interim data may not support resubmission even if final trial data does.
  
  • Even with this new plan, there may still be risks in terms of preserving DEVOTE’s integrity if the NDA is resubmitted. In DEVOTE, patients are randomized to Tresiba or placebo and are therefore only have a ~50% chance of receiving the study drug. If the NDA is resubmitted and the drug is approved, patients may drop out of the trial to directly pursue the drug. Dropout rates may be higher in the placebo group assuming there is a sufficient efficacy difference, which could skew trial results.
• Reassuringly, management commented that long-term follow up from Tresiba’s phase 3b program has, if anything, shown a regression towards the mean rather than a preserved or heightened deviation in the hazard ratio for CV events. We assume the phase 3b data refers to trials that were not included in the initial NDA or longer-term follow-up data from studies that were. Early pharmacovigilance data from markets where the drug is available (largely Europe and Japan) also does not show any worrying safety signals so far, although the size of the database is still fairly small.
• Tresiba is an exciting product in its own right, as noted, but we are most excited for its US approval because it will allow Novo Nordisk to file Xultophy (insulin degludec/liraglutide). Read on below for a few of the reasons why this product could be so transformational.

Xultophy (insulin degludec/liraglutide)

• Novo Nordisk shared new compelling topline results from the 26-week DUAL V phase 3 trial on the conference call: the primary finding was a statistically superior A1c reduction (-1.8%) from baseline (8.4%) in type 2 diabetes patients on background metformin who switched from Lantus to Xultophy vs. those who stayed on Lantus (-1.1% from a baseline of 8.2%) – that was a pretty big difference in our view. A full 72% of patients in the Xultophy arm reached a final A1c target of 7% or below vs. 47% in the Lantus arm. The improvement in glycemic control with Xultophy occurred in the context of reduced confirmed and nocturnal hypoglycemia vs. the Lantus arm (specific values not disclosed) and 1.4 kg (~3 lbs) mean weight loss vs 1.8 kg (~4 lbs) mean weight gain with Lantus. This data was particularly meaningful because it was some of the first data we have seen (to our knowledge) comparing Xultophy with a non-Novo Nordisk product; DUAL I and II used Victoza and Tresiba monotherapy as comparators.
  
  • These topline results are largely in line with what we would expect based on data from DUAL I. That trial found a mean A1c reduction from a similar baseline (~8.3%) of 1.9% after 26 weeks vs. 1.4% with Tresiba (insulin degludec), an incremental benefit of 0.5%. The difference in the A1c changes between the two study arms in DUAL V was 0.7% but the 0.2% difference in mean baseline A1c between groups means that the incremental benefit may more appropriately be characterized as 0.5%-0.7%. The difference in weight at 26 weeks in DUAL I was 2 kg (~4 lbs); the greater difference seen in DUAL V may be due to the fact that all patients were on insulin at baseline and could have been heavier on average.
  • As a side finding: the fact that investigators were able to achieve an A1c reduction of 1.1% in the Lantus group while only adjusting dose shows how much room for improvement there is in basal insulin titration and how sub-optimally managed so many patients are who are on insulin.

Table: Topline results from DUAL V

* = Xultophy statistically significantly better than insulin glargine

• During Q&A, management shared that Novo Nordisk is looking to target Xultophy to the millions of patients worldwide not at control on basal insulin alone. For these patients, Novo Nordisk believes Xultophy could be a safer and more weight-friendly way to intensify therapy, ostensibly compared to rapid-acting insulin. We have heard it suggested at a few past conferences that Xultophy could one day be a preferred option as patients’ first injectable (see item #4 of our IDF 2013 Day #1 Report), but this would represent more of a paradigm shift than using it as an way to intensify injectable therapy. Long-term we do think Xultophy could be an initiator injectable, in large part because of the very gradual up-titration process that avoids much of the initial nausea patients experience when initiating Victoza.
• Xultophy’s closest competitor is Sanofi’s LixiLan (insulin glargine/ lixisenatide), which is currently in phase 3 and could be filed in 2H15. Phase 2 results for LixiLan included an A1c reduction of 1.8% with LixiLan and 1.6% with Lantus from a baseline of 8% - the difference was modest (perhaps due to a stronger-than-expected performance in the Lantus group) but still statistically significant. The LixiLan group also saw significantly less weight gain.
• Xultophy was approved in Europe in September and had its first launch recently in Switzerland. We do not have any information yet about reimbursement.

Obesity

• Saxenda (liraglutide 3.0 mg for obesity) is still on track to be launched in the US in 1H15; it very recently received a positive CHMP opinion in the EU. The CHMP opinion recommends Saxenda’s approval for weight management as an adjunct to a reduced-calorie diet and increased physical activity for adults with a BMI of at least 30 kg/m² or with a BMI of at least 27 kg/m² and at least one weight-related comorbidity. The company announced that it expects to receive marketing authorization from the European Commission within the next two to three months and would begin launching Saxenda in several European markets in 2015. The indications listed in this opinion closely match those in Saxenda’s US label, although we were excited to see the addition of prediabetes as a weight-related comorbidity.
  
  • During Q&A, management confirmed that it is targeting a sales force of approximately 500 representatives for Saxenda. This group will target a relatively narrow group of specialists already working in weight management, a wise move given that an injectable therapy is somewhat more complex than a weight loss pill might be. For comparison, from updates we heard over the course of last year, Eisai has a recently reduced sales force of ~450 representatives for Belviq (lorcaserin), Orexigen has a planned sales force of ~900 representatives for Contrave (naltrexone/bupropion), and Vivus has a sales force of roughly 150 for Qsymia (phentermine/topiramate) – we note that these numbers may have since changed as the products’ launches have progressed.
  • Management shared that conducting studies aimed at expanding Saxenda’s label to include patients on insulin  is “not highest on the list of priorities.” There is little established reason why the use of Saxenda along with insulin would pose a safety risk, but the phase 3 SCALE program did not collect data to fully support clinical safety and efficacy in patients taking insulin. Dr. Mads Thomsen suggested that the decision to conduct such trials would be a management decision. We can understand why Novo Nordisk’s focus is on the launch rather than on expanding Saxenda’s label, but we hope an insulin study remains on the company’s long-term radar. Insulin is a cause of weight gain and there are certainly many obese patients on insulin that could benefit from Saxenda.
• Excitingly, we learned today that Novo Nordisk has a long-acting amylin analog (NN9838) in phase 1 for obesity. Other than AZ’s Symlin (pramlintide) for diabetes, we have not heard much about amylin analogues in clinical development recently. Novo Nordisk’s major recent move into obesity has already led to a few new obesity candidates popping up early in the pipeline, including the glucagon analog G530L (NN9030), which was unveiled during Novo Nordisk’s 3Q14 update.
• Towards the end of Q&A, management briefly confirmed that Novo Nordisk still plans to explore an obesity indication for the phase 3 once-weekly GLP-1 agonist semaglutide. We have heard earlier that imaging data suggests that semaglutide could have even greater weight effects than liraglutide. During Novo Nordisk’s 3Q14 update, management suggested that phase 2 dose-ranging trials in obesity could begin this year.

Semaglutide

• SUSTAIN 5, the final of six global phase 3 trials for the once-weekly GLP-1 agonist semaglutide, began in October. This news was included in Novo Nordisk’s 3Q14 update as well. See a table below for an overview of the global SUSTAIN program – the only change we noticed was a slight acceleration of the estimated completion date for SUSTAIN 5 from February 2016 to December 2015 (notable given that SUSTAIN 5 was initially the last non-CVOT scheduled to be completed).

Oral GLP-1 Agonists and Insulin

• Management disclosed that the oral GLP-1 agonist OG217GT has been discontinued in phase 1. Management shared during Q&A that OG217GT was discontinued because it achieved insufficient drug exposure in healthy volunteers. We do not see this as disappointing news as the company has a more advanced oral GLP-1 agonist (OG217SC) in phase 2, with data expected in 1H15. However, it does underscore how many obstacles there are in consistently and safely delivering a polypeptide orally. Dr. Mads Thomsen noted during Q&A that an oral formulation of a peptide drug must show a normal efficacy distribution with no unexpected safety findings, requiring a more complicated set of analyses than a clinical development program for an injectable GLP-1 agonist would. Ultimately, he stated that the end goal is to match injectable GLP-1 agonists in clinical safety and efficacy in a commercially viable manner – quite a tall order. He also shared that the phase 3 go/no-go decision for OG217SC will likely take place in the second half of this year. Both OG217GT and OG217SC use semaglutide as the active ingredient.
  
  • Dr. Thomsen delved into the scientific basis of the differences between the two oral peptide carrier technologies. Merrion’s GIPET carrier uses a para-cellular mechanism to loosen the tight junctions between cells lining the GI tract. Interestingly, Dr. Thomsen said that GIPET works well for insulin but appears to work less well for GLP-1 agonists, although the cause of this difference is unclear. OG217SC (still under development in phase 2) uses Emisphere’s SNAC enhancer that uses a trans-cellular carrier mechanism.
• There were no updates on the long-acting insulins LAI287 and LAI338, which remain in phase 1. See our Novo Nordisk 3Q14 report for more on these candidates. In the competitive landscape for basal insulins with once-weekly potential, 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 program into clinical studies.

Faster-Acting Insulin Aspart (FIAsp)

• The presentation included new topline phase 3 results showing that the addition of FIAsp to basal insulin yielded a 0.9% greater mean A1c reduction vs. continuation of basal insulin alone. The overall change in A1c in the FIAsp group was from 7.9% at baseline to 6.8% after 18 weeks; assuming a similar baseline in the comparator group, patients continuing basal insulin alone ended at a mean A1c around 7.7% for a mean A1c reduction of ~0.2%. The trial, Onset 3, enrolled 323 type 2 diabetes patients on basal insulin and metformin at baseline. FIAsp was taken thrice daily. Other findings from the trial were an overall improvement in postprandial glucose parameters (no specifics provided). Weight gain and the rate of hypoglycemia were both increased in the FIAsp group compared with basal insulin alone; while the slide correctly pointed out that this is not unexpected, the degree of the difference will be important to ascertain when full results are presented (we imagine ADA is a possibility). Overall, the safety and tolerability profile was in line with what has been seen from FIAsp in previous studies.
  
  • These results were the second topline phase 3 results we have seen out of the four current Onset studies. The first trial to complete was Onset 4, a smaller six-week double-blind parallel-group pump trial. In that trial, both NovoLog (standard insulin aspart) and FIAsp demonstrated pump compatibility and there was no apparent difference in the number of microscopically confirmed infusion set occlusions. We noticed on ClinicalTrials.gov that a third Onset study, Onset 2, was completed earlier this month; the trial compares FIAsp to NovoLog on background insulin glargine and metformin in patients with type 2 diabetes. Onset 1 is still ongoing and has an estimated primary completion date of June 2015. The trial compares FIAsp and NovoLog on background Levemir (insulin detemir) in 1,095 patients with type 1 diabetes.

Questions and Answers

Insulins

Q: One question on Levemir and the price protection that we see kicking in, in 4Q14. Could you explain how that's working on a quarterly basis going forward? Can you help us understand what we should expect from price caps you've built into contracts and how that should evolve?

A: Price protection is a mechanism in contracting in the US whereby you – either on a one-year basis or on a several-year basis corresponding to the contract you might have with a PBM – go in and agree that you will not have a net price increase over and above a certain percentage threshold. So, if we take an example of 8%, for instance, that means that if prices increase in a calendar year – because it's a one-year price protection – you only get the 8% net in pricing effect. Then, if you have a contract over several years, it might be that you have a cumulative price protection of 8% per year over a period. That means that if your list price increases are above that level, you get the same situation over a number of years.

When you want to analyze the detailed effects of this, it's quite complex because you need to compare to what was the price protection situation a year ago vs. how is it now, and at what time of the calendar year were the price increases actually executed. It gets very detailed when you want to get to the granularity of how to analyze this. Overall, what we can say is that longer contracts with price protection will have a tendency to lead to lower net price increases, but also more stability, because often, when you renegotiate contracts, your gross rebate gets pushed up. In a way, there's a trade-off between having longer contracts with price protection to having very short contracts maybe without price protection but where the rebate demand might go up year-by-year. In the future, we expect to see both types of contracts – short, long, with, and without price protection; it's all a part of, you could say, the wheeling and dealing about net pricing.

A: It's basically a reflection of the competitive environment in the given segment that you're contracting on.

A: Yes. Then there can be tactical considerations for the PBMs, which mean that they are more in favor of a certain setup. Of course, at the end of the day, what matters is the net price.

A: Also, the length of these price protection contracts also is influenced by the advent of new products into the marketplace in the coming years. It's kind of complicated.

Q: On my estimates, NovoLog is still seeing low-single digit net price increases in the US. NovoMix has roughly flat net price, while Levemir is seeing net price rises well into the teens. Is it fair to assume that US net price may actually be negative for NovoLog and NovoMix for this year? Then on Levemir, you still seem to expect a net positive price impact? If we strip out the R&D charge from Q3 operating profit growth for this year – around 8% - will pricing get better in 2016 to help get you back towards your target of 15% operating profit growth?

A: Here, we have to be careful whether we talk about 2014 and 2015, because there are certainly changes expected from 2014 to 2015. There was also a comment on whether or not we're anticipating the price picture changing in 2016 as a way of getting back to double-digit growth.

A: It is, of course, a multi-faceted, complicated topic. However, if you look at 2014, it's quite clear that we have seen net price of all the three of the modern insulin analogs that are in the US increase. If you look at 2015, we also expect net price increases. We do not expect double-digit price increases, but we expect single-digit price increases on all three modern insulins. The details about how it will play out for each of them is down to details about contracts seeing these price increases, timing and so on. We don't expect to see any dramatic change in the pattern compared to what we have seen in the past. We don't have any major contract changes the way we had last year. So, we expect a more even and calm development, you could say, from a contracting and pricing and rebating point of view. In regard to 2016, we don't have anticipation at this point about exactly how it will look. We will have to comment on that once we get closer to 2016.

A: If any thing, 2016 looks like a time where there will be more products on the market, if we are to project new entrants. So, everything else being equal, that would bode for more competition in the field.

A: In terms of the long-term financial targets, I think the comment from our side on the longer-term horizon is that a pre-condition for longer-term achieving the 15% operating profit growth target will be that we get our top-line growth up to the double-digit level we historically operated on. A pre-condition for achieving that will clearly be having access in the US market with Tresiba and the portfolio of degludec products that goes with Tresiba.

Q: One question on the cost lines for sales and distribution. That seems to stick out as being relatively low, and you talk about promotional cost being low in the US and Europe. Is this now sort of the real world of being big pharma and being more cost conscious going forward, or is this a phasing effect? Could you give me some examples of why or how you're reducing costs in the quarter, or at least keeping costs stable?

A: I think there is a significant element in Novo Nordisk having established a global footprint in terms of our sales force. We feel that we can comfortably operate in the ballpark of 26% as an S&D ratio out of total sales. The savings in 2014 was partly related to the promotional activities in the European and the US markets. We would anticipate that we would slightly ramp up our direct-to-consumer marketing in the US in 2015, but, with the growth in our business, we will feel comfortable operating at a similar ratio. We don't have any current plans for expansion of the US sales force, which is, of course, a key cost driver. So, I think a fair assumption for 2015 will be around 26%. As the S&D, costs are distributed with sales, and we don't expect any significant currency impact on the ratio.

Q: One question on the promotion in the long-acting insulin seen in the US. Apparently, Sanofi is trying to catch up in terms of fueling some more promotion. What will be your position in 2015? Will you increase your promotion again, which has been very successful last year?

A: We're not seeing any significant changes in the promotional landscape. Sanofi is obviously promoting Lantus heavily as they've been doing for the past several years. We are promoting Levemir heavily in the marketplace, and we intend to continue to do so.

Tresiba and DEVOTE

Q: On DEVOTE, is there's any real-life CV event data from launched territories that you could potentially use to supplement the DEVOTE interim data?

A: As you're well aware, the situation is that there are two components to a potential NDA class 2 resubmission. One component is the DEVOTE data per se, the interim analysis that includes the MACE events and a few other end points, so to speak, agreed upon with the Agency. Then, there's the rest – the rest is a completely standard update, including an analysis with also a MACE analysis of all the existing phase 3b trials. That's more than 20 that have been entertained by the company, and also, investigations on pharmacovigilance systems that are in place for the drug.

With that being said, even though we have more than 100,000 patients in Japan, and this year, may have 100,000 patients also in Europe, it is still so that in the big broad context, the pharmacovigilance real world evidence database, so to speak, is still relatively loosely developed because it is early days. We are not seeing any signs or signals, neither in the pharmacovigilance systems nor in the 3b analysis that change our clear view that Tresiba is as safe as other insulins. However, it is the DEVOTE data that ultimately, as you know, will be gating for what will happen later this half.

Q: Turning back to DEVOTE and your comments on the 3b analysis. Could you give us a little bit of granularity on the hazard ratio? Is it an improvement over the analysis is the FDA did back in 2012, which I think got to about 1.3 on the of their analyses, and whether the hazard ratio you’re seeing is improving?

A: In terms of what is happening in the 3b program, if anything, I would say we are seeing a regression towards the mean, so to speak – i.e., not any stronger deviations from a hazard ratio of 1.0. So, there's no signals of alert of any kind in what we are seeing in the ongoing 3b program at all, if that is what you're aiming at. Now, do bear in mind that neither did we see that in the pre-specified analysis in the BEGIN trials because, as you recall, we have events rates of 1.44 and 1.48. It is only when you start massaging the data that you can see outlier findings. I think it's all in the DEVOTE data, and we'll be much wiser either this year or next year.

Q: In the DEVOTE study, given that you see it enrolling or both accruing and completing faster than expected, can you elaborate a bit on the risk for the FDA – saying that ‘you're just a year-and-a-half from completion, let's just see the final data?’ Is that a risk with the FDA, or do you see that as just speculation?

A: I will start by quoting my boss. That is pure speculation. Nobody can have any guesses on that. I have to remind everybody, however, that the FDA is a science-based agency. That implies that they should be totally able to disconnect any time course relationship between these things. They should look at the data, and I would expect that they will look at the data, in the discussion with this small un-blinded team, and that will form the basis for a potential resubmission and the action thereupon. Therefore, I would not expect any connection between these two things. But again, it's pure speculation.

Q: How do you think about the number of events that is driving a hazard ratio vs. duration of exposure? One of the concerns in the market seemed to be that even though you may have a hazard ratio, one could argue that the duration on average isn't long enough to really allow you to have a view at that point in time.

A: Obviously, this question as to which way do you put together the amount of MACE events: either a low event rate and a long exposure, or high event rate and the low exposure, or somewhere in between. This has obviously been discussed with the Agency. The way we conduct this trial, including the conduct up to the interim analysis in the very near future, is completely compatible with what we have discussed with the FDA. There's no concern as to so-and-so many have to be exposed for so and so long. However, you’re right in stating it's typically a phenomenon that will be associated with the final outcome of the trial. For instance, in the LEADER trial, for safety reasons, the European agency actually demanded a minimum of 3.5 years of exposure in each individual cardiovascular patient in that trial. However, for the interim analysis, there is no such requested amount, and it has been discussed.

Victoza and Injectable GLP-1 Agonists

Q: One question on the GLP-1 contracting market. Trulicity apparently has position on the Express Scripts' national formulary, which I believe was one of the contracts you were excluded from. Does that positioning give you any color on how aggressive Lilly is going to be with their rebating strategy?

A: I think we should be careful with how we interpret what competition does, but we can, of course, conclude that Trulicity has been launched and has open access under several plans, and has a list price that is very comparable to the list price of Victoza, and has a slightly less effect on weight than we see with Victoza. But that being said, I would rather comment on how we see the market. We think that there should be free choice among doctors and patients in the GLP-1 space. So, our contracting strategy is to try not to go for exclusives, but to have several products listed so that there is choice for the patient and the healthcare professionals.

Q: Can you discuss the marketing positioning of Trulicity? If you see their marketing material, are they arguing for Trulicity being the first product for everybody, or is it more of a distinction between daily and weekly options? Essentially, are they looking for a co-position here and position themselves against Bydureon, or are they going after Victoza? And GSK captured a couple of market share points in the US with somewhat of a lower efficacy product. I was wondering if you can discuss what patients can prefer? Are you seeing primarily a match up between you and Lilly?

A: If we start with Trulicity, of course, the details about that product's positioning should be asked of our competitor and not us. However, the way we see it, there will be a strong part of the market that will prefer once daily, and there will, for sure, also be a segment of the market that will prefer once weekly. It is unlikely that one of the two regimens will sort of take the whole market, which can actually be seen from analyzing the split that has been seen with the first GLP-1 that hit the market or the GLP-1 look-alike Byetta and follow on Bydureon. So, I think it's fair to say that it will not be one regimen taking it all.

If we then look at what seems to matter in our research in the market, it's quite clear that in any diabetes medication –injectable diabetes medication – it is efficacy that matters. It is efficacy that's really number one, two and three. Then it's, of course, efficacy both on glucose, and it could be efficacy on other parameters, for instance, weight.

Therefore, I find it very unlikely that there's appetite for sort of a low-efficacy discount product. However, with companies coming out, the products that have maybe a slight edge on one part of the efficacy vs. other parts of the efficacy; we do hope to see an expansion of the total market; if you look at the patient numbers, the GLP-1 market in the US is very small in patient numbers considering the superior efficacy that GLP-1 has.

Q: Regarding the GLP-1 market, we've seen a small rebound. Is that something you see as sustainable? Do you see this coming from many more players promoting GLP-1, or is it a lack of noise around side effects?

A: I think it is sustainable. I think the two components are the ones you mentioned. One is the reduced focus on this not-real safety risk related to pancreatic cancer. That has been refuted by the relevant authorities. Then there is also the fact that more players are entering the GLP-1 field. It's like mixing hot and cold water. You can't really say whether it's one or the other of the effects that is leading to this slightly improved growth level in the GLP-1 total market. I think it's sustainable because I think the newcomers are there to stay; they will keep on promoting. I also think we have sort of refuted the pancreatic cancer safety story. So, hopefully, we will see a level of growth in the GLP-1 market this year that's better than last year. However, we won't get back to, of course, the percentage growth levels that we had in the beginning.

Xultophy

Q: One launch question on Xultophy, now that you have the approval in Switzerland. Could you offer an update on how you're thinking about positioning the product within the space?

A: Xultophy, of course, combines the effects of Victoza with the strong effect of Tresiba. The way we are positioning it globally is for all those millions of patients worldwide that are currently on a basal-only regimen but that have inadequate control of their glucose. Basically, we have millions of people who are taking basal insulin analogues, but their A1c numbers are failing to reach target. That is the target we going for. It's safe, it's convenient and effective.

Saxenda

Q: On Saxenda, given that the US label doesn't allow it to be used by patients on insulin, does that impact your target population? Is that something you believe you can change in the future or is that just fundamentally how it is?

A: It's the same situation as when we were developing Victoza, both in Europe and the US. When you don't have a statistically reliable database of safety and efficacy when combining two agents together, then obviously, you will not get that as an indication. Our SCALE program has obviously not investigated, to any significant extent, co-use of 3 mg of liraglutide together with insulin, and that is why it's not indicated. It is indicated for use in people with diabetes, but not to those who are in insulin therapy. That would be a management call whether we want to investigate that at some point in time. However, it would not be highest on the list of priorities.

Q: Back to Saxenda, I think there were some news reports of 500 US salespeople. Are these incremental employees or have they come from the Tresiba pool that was previously recruited, and what products are they marketing in the meantime?

A: Yes. It's great that for Saxenda, we expect to have roughly 500 sales people targeting a relatively narrow target group of specialists that are already involved in weight management in the US. The majority of these people will be coming from our existing sales force, and we'll have those people trained and ready once we launch within the next few months.

Earlier-Stage Pipeline

Q: On the oral GLP-1, could you talk about your thoughts on the product 9924 in phase 2 given the discontinuation of the 28 product? Obviously, they are different carrier mechanisms and different oral GLP-1 included. So, just wondering whether you think this result's more related to the carrier mechanism or the GLP-1 in the discontinuation.

A: The OG217GT, that was, as you correctly stated, with the GIPET carrier. And the GIPET carrier is noted for its predominantly paracellular optic mechanism by reversible loosening of the tight junctions between the enterocytes, to be very specific. Now, that works well for insulin, but it seems to work less with GLP-1, and if you ask me why, I cannot give you an answer. However, the short effect is that exposure for this particular molecule, semaglutide, is a distinctly better in the case of the SNAC enhancer, which predominantly uses a transcellular optic mechanism. That also means that when we see data from Phase 2, we have to look at many things. We have to look at how good are the data, at which dose were the data obtained, was the population distributed in a normal distribution manner in terms of efficacy, and what were the results in terms of the expected safety findings such as the GI and so on. Overall, basically, it all boils down to one thing: Can we find a way where we can claim that all semaglutide will actually match a injectable counterpart in terms of clinical safety and efficacy in a way that is commercially viable for the company to develop in Phase 3? If you can say yes to all of that, then the company can, in the 2H15, look into Phase 3 stop-go decisions. That is also why this is truly pioneering territory. There are many more analysis to be done in such a program than in a straight-forward injectable program. It'll be very exciting.

Q: Do you have any comments on semaglutide for obesity?

A: I can only confirm your notion that, based on what we have seen, i.e., potential for more efficacy with semaglutide compared to liraglutide at higher doses, the management does intend to stick to exploring an indication for semaglutide.

-- by Manu Venkat and Kelly Close