Novo Nordisk 2Q12 – Diabetes Care up 24%; Victoza sales increase 15% sequentially; phase 3 results for IDegLira reported – August 14, 2012

Novo Nordisk reported financial results for 2Q12 in a call led by CEO Lars Sorenson last Thursday. Diabetes Care revenue totaled a record DKK 15.0 billion ($2.59 billion), increasing a spectacular 24% year-on-year (reported) with growth driven by Victoza and modern insulins. Novo Nordisk remains the leader in the global diabetes care market with a 25% share, up from 23% in 2009 and 2010 and 24% in 2011. Sales of modern insulins reached DKK 8.6 billion ($1.5 billion), rising a reported 23.5% year-on- year and 9.5% sequentially. Management attributed the growth to strong sales in North America, China, and International Operations. Notably, there were major differences in performance for N. America and Europe, with the former up 29% year over year and the latter up just 2%. N. America represented 42% of sales in 2Q12 and 64% of total growth for Novo Nordisk, while Europe represented 26% of sales but just 6% of total growth – more affirmation of cost pressure in Europe. All three modern insulins posted their highest revenue to date, with particularly strong sales of NovoRapid (DKK 3.8 billion [$661 million]) and Levemir (DKK 2.4 billion [$423 million]). Based on IMS moving annual total (MAT) data from May, Novo Nordisk possesses a 50% share of the global total insulin market by volume (down from 51% in 2010) and a 46% share of the modern insulin market (consistent with recent quarters). Sales of Victoza totaled DKK 2.29 billion ($396 million) in 2Q12, up 83% from DKK 1.3 billion ($241 million) in 2Q11 and up 15% sequentially. GLP-1 agonists now account for 5.1% of the global diabetes market by value and Victoza leads the class with a 65% share, up from 62% in 1Q12 despite the launch of Bydureon – we believe that they are also benefiting from increased GLP-1 education to HCPs. Notably, management increased the expected growth rate for 2012 slightly, to 9-12% in local currencies compared to the 8-11% growth forecasted in 1Q12, and to 7% from 4% on a reported basis.

On the pipeline front, very few details were provided on why the FDA has called for an advisory committee meeting for degludec and degludecPlus on November 8, 2012. Management maintained that they have not yet been informed about what the focus of the meeting will be and that more detail might be available during the 3Q12 call on October 31, 2012. As we understand it, the FDA is requiring, by default, advisory committee meetings going forward for all new molecular entities unless specific reasons exist for not holding one. Overall, the advisory committee for degludec and degludecPlus came as a bit of a surprise because it was announced so late in the review process – still, with the current FDA, nothing that would prompt a delay would be very surprising. Regulatory action for these drugs in the EU is expected during 2H12, as Novo Nordisk has received a CHMP “Day 180” report (CHMP recommendation is typically expected by day 210). New data from the DUAL I phase 3a trial for IDegLira (fixed dose degludec and liraglutide) were reported, demonstrating IDegLira’s superior glucose compared to each of its individual components and intermediate weight benefit (superior weight benefit compared to degludec but not to liraglutide, as expected). In the development of a once- weekly GLP-1 candidate, Novo Nordisk chose semaglutide over once-weekly liraglutide to enter phase 3 trials in 1H13. The phase 2 data demonstrating semaglutide A1c superiority over liraglutide were briefly reviewed and will be presented at EASD 2012 in October. Although developing once-weekly liraglutide could have been an easier choice from a regulatory perspective, presumably semaglutide has superior efficacy and/or other notable advantages. The phase 3a program will enroll >8,000 patients, with the first trial expected to be a cardiovascular outcomes trial (CVOT). No updates were provided on Victoza’s CVOT, LEADER, but management did disclose during Q&A that currently, there is no requirement for a CVOT for Victoza’s proposed obesity indication. Finally, a phase 1 trial for a novel oral basal insulin analog (NN1954/OI362GT) was initiated and is expected to complete in 1H13. We are very eager to see how this moves along – there would be tremendous interest from our view in any kind of oral insulin that works. We will be testing this in the upcoming dQ&A surveys. In the roadshow presentation, there were two oral GLP-1s in development in phase 1 as well. Finally, the NovoPen 5, the first new durable insulin pen for adults since 2005, was launched in June in Denmark. This pen has a memory function that displays information related to the last dose administered. Novo Nordisk expects to launch the pen in additional European countries in 2H12, although no details were provided regarding launch in the US.

FINANCIAL AND PRODUCT NEWS

• Diabetes Care revenue totaled DKK 15.0 billion ($2.6 billion) in 2Q12, rising an impressive 24% on a reported basis from DKK 12.o billion ($2.3 billion) in 2Q11. While this was a relatively easy comparison (the 5% year-on-year growth in 2Q11 was the lowest reported growth since 4Q09, albeit compared to a challenging base of 21% growth in 2Q10), Diabetes Care sales achieved the largest year-on-year growth in 2Q12 since 4Q10 (when growth was 28%). Sequentially, sales grew 7.5% from DKK 13.9 billion ($2.4 billion) in 1Q12. As in previous quarters, growth was driven mainly by sales of Victoza (36% of total growth in 2Q) and modern insulins (particularly NovoRapid [26% of total growth in 2Q] and Levemir [20% of total growth in 2Q]; see table below). According to IMS moving annual total (MAT) data from May 2012, Novo Nordisk currently holds a 25% share of the global diabetes care market by value, slightly higher than its 24% share in 2Q11 and 23% share in 2Q10. IMS MAT data from May 2012 indicates that Sanofi trails Novo Nordisk with a ~17% share by value, followed by Merck (~15%), Lilly (~13%), Takeda (~11%). We would anticipate, of course, that Takeda’s share will drop precipitously with the move of TZDs to generic, and it will be interesting to see what happens with the rest of the rivals – the slope of Merck’s revenue trajectory would suggest it may gain global market share the fastest.

Table 1: 2Q12 revenue from modern insulins

* The above assumes an exchange rage of 1 DKK to 0.1728 USD

• Sales of modern insulins reached DKK 8.6 billion ($1.5 billion) in 2Q12, increasing 23.5% from DKK 7.0 billion ($1.4 billion) in 2Q11. Again, this was a relatively easy comparison given the low reported growth of 3% in 2Q11 (albeit from a challenging base of 25.5% growth in 2Q10) – that said, Novo Nordisk has certainly never achieved insulin sales exceeding8.0 billion DKK, so to achieve such a record quarter was quite an accomplishment. The high growth stemmed from strong sales in North America (which accounted for more than half the growth), Region China, and International Operations – we point out that Europe is the main geographic region absent from this list – note that Novo Nordisk does not give insulin sales by region. Sequentially, revenue from modern insulins rose 9.5% from DKK 7.9 billion ($1.4 billion) in 1Q12. Sales of modern insulins now account for 75% of Novo Nordisk’s total insulin’s sales, compared to 72% in 2Q11, ~70% in 2Q10, 64% in 2Q09, 57% in 2Q08, 53% in 2007, 45% in 2006,35% in 2005, 26% in 2004, and 16% in 2003! In 2Q12, Novo Nordisk posted its highest sales for NovoRapid, NovoMix, and Levemir to date each – quite an accomplishment. Specifically, according to our model, sales of NovoRapid rose an impressive 24% year-on-year and 8.8% sequentially on a reported basis to DKK 3.8 billion ($661 million). For comparison, 2Q12 sales of Lilly’s Humalog rose 4.5% year-on-year to $613 million following four straight quarters of double- digit growth and sales of Sanofi’s Apidra grew 5.7% year-on-year to $72 million - we don’t see Sanofi investing much in this product and Apidra has never had more than 5% of the rapid acting analog market. Lilly introduced Humalog, the first rapid acting analog, in 1996 in the US – Novo Nordisk followed with Novolog in 1998. Novo Nordisk has had a larger share of this market with Novolog and NovoRapid since 2010, but each quarter of late, the companies are very close, as in 2Q12 ($647 million vs. $613 million).

• According to our model, market shares for the fast-acting insulin analogs (NovoRapid, NovoMix, Humalog, and Apidra) have changed little between 2008 (the year Apidra was launched) and 2Q12 – NovoRapid’s market share increased from 34% to 38%, Humalog’s share decreased from 38% to 35%, NovoMix’s share declined slightly from 25% to 23% and Apidra’s share moved up from 3% to 4%. Sales of NovoMix totaled DKK 2.3 billion ($404 million), rising a reported 15% year-on-year and 8.6% sequentially. Revenue from Levemir totaled DKK 2.4 billion ($423 million), up 31% year-on-year on a reported basis and 11% sequentially. For comparison, 2Q sales of Lantus also grew substantially (27%), exceeding €1 billion (€1.23 billion; $1.58 billion) for the third straight quarter and surpassing $1.5 billion in quarterly sales for the first time.

• During the call, management noted that in June, Novo Nordisk launched the NovoPen 5, its first new insulin pen for adults since 2005 (see below for details).

Table 2: Sales of fast-acting and basal insulin analogs by product

* The above assumes an exchange rage of 1 DKK to 0.1728 USD

• In May 2012, Novo Nordisk announced that the FDA had approved Levemir for use in children aged two to five with type 1 diabetes. The approval was based upon data from a subpopulation analysis (n=82) showing that Levemir exhibited similar effects on A1c compared to NPH after one year (8.1% for Levemir and 8.3% for NPH, from baseline A1c levels of 8.2% and 8.1%, respectively), with lower incidences of overall and severe hypoglycemia (Thalange et al., Ped Diabetes 2011). Levemir’s expanded indication represents its third US label update in 2012 – as a reminder, in April, Novo Nordisk updated Victoza’s US label to reflect superior efficacy to Januvia and the FDA’s recent approval for use of Victoza with basal insulins (for details, please see our April 9, 2012 Closer Look at http://www.closeconcerns.com/knowledgebase/r/0326a3cd). The FDA’s decision follows the EMA’s Committee for Medicinal Products for Human Use’s (CHMP) September 2011 positive recommendation to extend the use of Levemir to the same pediatric population. For comparison, in May 2012, the European Commission authorized extended use of Lantus for children with type 1 diabetes aged two to five years old. Sanofi plans to request six-month Pediatric Exclusivity for its existing Supplementary Protection Certificates (SPC) for Lantus in Europe, which if granted, would make most SPCs in Europe expire in May 2015. Sanofi has already received pediatric extension for Lantus in the US, giving Lantus patent protection until February 2015. However, from our understanding, Lantus is not currently approved for pediatric use in the US – this is a little confusing from our view.

• Sales of human insulins and protein-related products (such as pens) achieved positive year-on-year growths (5.3% and 18% on a reported basis, respectively), with nearly flat sequential growths. Sales in both categories represent a vast improvement over the large negative year-on-year growths observed in 2Q11 (decreases of 15% for human insulins and 10% for protein related products). Notably, sales of human insulins have increased in the past two quarters (2.4% in 1Q12 and 5.3% in 2Q12) following four consecutive quarters of negative growth (-4% to -14%) in 2011 – we suspect this has to do with increased cost pressures globally, particularly in Europe. Revenue from oral antidiabetic products (NovoNorm, Prandin, and PrandiMet) was constant with 2Q11 (compared to a 7% year-on-year decrease in 2Q11) and declined 8.8% from DKK 720 million ($130 million) 1Q12.

Table 3: 2Q12 revenue from human insulins, protein-related products, and oral antidiabetic products

* The above assumes an exchange rage of 1 DKK to 0.1728 USDManagement offered data on the global insulin market and provided a detailed characterization by region.

• Global – Novo Nordisk’s Growth: IMS MAT data from May 2012 indicates that Novo Nordisk remains the leader in the global insulin market with a 50% share of the totalinsulin market by volume (remaining approximately constant since 2006) and a 46% share of the modern insulin market (consistent with May 2011). For comparison, the remainder of the total insulin market and the modern insulin market is split mainly between Sanofi (~25% share and ~34% share, respectively) and Lilly (~21% share and~20% share, respectively). MAT quarterly growth rate for Novo Nordisk’s total global insulin sales was 8%, having climbed steadily from a low of ~5% in 4Q11 (the growth rate dropped continuously from ~11% in 4Q10 to ~5% in 4Q11 but increased to 6% in 1Q12). Consistent with the 1Q12 earnings update, management attributed the slow growth in 2011 to healthcare reforms across most major markets (including US, Europe, China) and to competitive pressures in North America, Europe, and Japan. How they presented the challenges was interesting – we assume this means that Japan represents one market without big cost pressures and that in China, management doesn’t currently perceive competitive pressures – though on the latter, we note that both Sanofi and Lilly seem to be making big investments in China. Management cited the increasing prevalence of diabetes and continued improvements in insulin therapies as factors that will help sustain the growth of insulin sales – it will be interesting to see how degludec fares in the global market and how differentiated it is perceived. Management’s decision on pricing will certainly be critical. We would anticipate that over time, Novo Nordisk would stop making Levemir, though there has been no official word on this.

• Global – Insulin Market Growth: Insulin therapy is currently used by ~26% of diabetes patients and accounts for nearly 50% of sales from diabetes treatments. According to IMS MAT data from May 2012, the value of the total insulin market totals DKK 98 billion ($17 billion), split between DKK 82 billion ($14 billion) for modern insulins and DKK 16 billion ($2.8 billion) for human insulins. The global insulin market consists of 37% long-acting insulins, 33% fast-acting insulins, and 30% pre-mixed insulins by volume (for comparison, the insulin market was more evenly distributed in May 2007 – 34% long-acting, 34% pre-mixed, and 32% fast-acting insulins). Europe has the largest volume share of the global insulin market (38%), followed by North America (35%), International Operations (17%), China (6%), and Japan and Korea (4%). During Q&A, management stated that the global insulin market is growing at about 5% by volume, driven by strong growth in China and International Operations (both ~18%), offset by slowing growth in North America (3-4%) and continued low growth in the EU (2-3%) and Japan (<2%). It is quite something from our view that Novo Nordisk’s insulin growth in DKK increased 18.5% during the same period (analogs and human insulin combined). Management commented that despite increased competition from newer antidiabetic medications (in particular DPP-4 inhibitors), Novo Nordisk anticipates thatthe increasing prevalence of diabetes and the longer lifespan of diabetes patients will help drive the use of insulin and sustain the global volume growth at approximately 5%. We have been trying to ascertain for some time how to measure lifespan of people with type 2 diabetes as we would also believe it must be increasing (more aggressive therapy earlier in disease progression, faster moves to better therapy as disease progresses) but have been unable to find data showing this.

• North America: Sales of modern insulins, human insulins, and protein-related products grew 27% on a reported basis in the first six months of 2012 (for comparison, growth in 1Q12 was 21%). Sales were driven by the increased market penetration of modern insulins partly offset by a continued decline in human insulin sales. IMS MAT data from 2012 indicates that the market value of insulin therapies in North Americaamounts to DKK 57.9 billion ($10 billion; up 14% from 2Q11) and that the market consists of 48% basal insulins, 37% fast-acting insulins, and 15% pre-mixed insulins by volume. As noted above, management cited the slowing volume growth of the US insulin market (currently ~3-4%) as one of the primary reasons for the low growth of the global insulin market and speculated that the slow US growth is due at least in part to the increased use of DPP-4 inhibitors (more details below). IMS MAT data from 2012 indicates that in the US, Novo Nordisk has a 41% share by volume of the total insulin market and a 37% share of the modern insulin market (both consistent with 1Q12 and 2Q11).Notably, Novo Nordisk trails Sanofi in the US modern insulin market (Sanofi has a 44% share), though both companies have a strong lead over Lilly (with a 19% share). Since May 2007, Lilly’s share has decreased roughly eight percentage points while Novo Nordisk’s share has increased roughly the same amount. Roughly 48% of Novo Nordisk’s modern insulin volume in the US is sold via the FlexPen, compared to ~47% in 1Q12 and ~44% in 2Q11,~41% in 2Q10, ~39% in 2Q09, and >37% in 4Q08, >35% in 4Q07, >33% in 4Q06,suggesting that penetration of pens certainly continues to increase over syringes – notably, however, in Europe and China, the percentage of modern insulin sold through pens is far higher.

• Europe: Sales of modern insulins, human insulins, and protein-related products in the first six months of 2012 remained unchanged with those observed in the first half of 2011, driven by revenue from NovoRapid and Levemir that were partly countered by declining human insulin sales – we were surprised by the latter since we had assumed cost pressures would prompt more growth in human insulin. IMS MAT data from 2012 indicates that insulin therapies have a market value of DKK 24.9 billion ($4.3 billion; up 3% from 2Q11) in Europe and that the EU market consists of 36% basal insulins, 39%fast-acting insulins, and 25% pre-mixed insulins by volume. Management noted, however, that sales in Europe have been negatively impacted by the continued low volume growth of insulins (<3%), gradual market share losses (this was surprising), and healthcare reforms in several European markets. When asked to comment on the low growth of the European insulin market, management speculated that insulin sales have been negatively impacted by the greater penetration of DPP-4 inhibitors in the EU and the US, noting the patients are increasingly using DPP-4 inhibitors before GLP-1 agonists, GLP-1 agonist/insulin combination therapy, and insulin alone. Presumably, patients are also using GLP-1 before insulin – Novo Nordisk, of course, also benefits from this as they have the top-selling GLP-1. Management noted that they expect the uptake of DPP-4 inhibitors to impact insulin sales for the next few years, similar to how the introduction of insulin sensitizers initially dampened the growth of the insulin market for three to four years – the introduction of that class also lengthened the lifespan of diabetes patients, we believe. Novo Nordisk currently possesses 51% of the total European insulin market by volume (flat with 1Q12; down from 52% in 2Q11) and 50% of the modern insulin market (flat with 1Q12 and 2Q11). The remainder of the Europe’s modern insulin market is split mainly between Sanofi (30% share) and Lilly (18% share). Regarding devices, the use of the NovoPen and FlexPen in Europe remains high, accounting for 96% of Novo Nordisk’s insulin volume.

• International Operations: Sales of modern insulins, human insulins, and protein- related products grew by 19% on a reported basis in the first half of 2012 – given that growth in 1Q12 was 24%, growth in 2Q12 appears to have been quite low, although the graph on the roadshow presentation on this geographic region overall is quite instructive– this region now has 17% of insulin volume, up from about 10% in 2002. Growth was driven by modern insulins with a “solid contribution” from human insulins. MAT data from 2012 indicates that the market value of insulin therapies in International Operations totals to DKK 6.8 billion ($1.2 billion; up 10% from 2Q11) and that the market segment consists of 54% pre-mixed insulins, 27% basal insulins, and 19% fast-acting insulins by volume. As noted above, during Q&A management highlighted the strong growth of the insulin market in International Operations (~18% by volume). Novo Nordisk holds a 59% share of the total insulin market by volume in “International Operations” and a 56% share of the modern insulin market (both consistent with 1Q12 and 2Q11). The remainder of the total insulin market is split mainly between Lilly (18%), Sanofi (14%), and Wockhardt (2%). Use of devices account for roughly 58% of Novo Nordisk’s insulin volume in major private markets (not defined).

• Region China: In 1H12, sales of modern insulins, human insulins, and protein related products increased 27% on a reported basis (14% operationally) – growth in 2Q12 appears to have been extremely high given that sales grew only 17% on a reported basis in 1Q12 (8% operationally). Notably, growth was again driven by sales of modern insulins, while sales of human insulin “only grew modestly” (for comparison, sales of human insulins were flat in 1Q12). During Q&A, management noted that the growth in 2Q12 was relatively high at around 18% operationally, reflecting favorable price changes. Management continued to guide for operational growth in 2012 in the ballpark of 15%, but qualified that growth would depend on possible pricing reforms for NovoMet (metformin hydrochloride) in 2H12. MAT data from 2012 indicates that insulin therapies have a market value in China of DKK 3.2 billion ($550 million; up 27% from 2Q11) and that the Chinese market consists of 65% pre-mixed insulins, 22% fast-acting insulins, and 13% basal insulins by volume. Also during Q&A, management stated that the insulin market in China is growing at an impressive ~18% by volume. Novo Nordisk possesses 62% of China’s total insulin market by volume (flat with 1Q12; down from 63% in 2Q11) and 66% of the modern insulin market (also flat with 1Q12; down from 68% in 2Q11). Lilly is the second-leading insulin distributor in China with a 14% share of the total insulin market. As a reminder, in the 1Q12 update, management stated that Novo Nordisk had lost some market share to Lilly due to Lilly’s expansion of provincial reimbursement drug listings and forecasted a continued strong battle with Lilly in upcoming quarters; no updates on Novo Nordisk’s competition with Lilly were provided in the 2Q12 earnings call. Regarding device use, roughly 97% of the insulin volume in China is attributed to sales of Penfill, for use in the NovoPen.

• Japan and Korea: Sales of modern insulins, human insulins, and protein-related products increased by 3% on a reported basis (compared to a 9% in 1Q12). Management attributed the low growth to the lack of volume growth in the Japanese market (<2%) and to a continuously challenging competitive environment. MAT data from 2012 indicates that the market value of insulin therapies in Japan and Korea amounts to DKK 5.4 billion ($930 million; up 5% from 2Q11) and that the market segment consists of 36% fast-acting insulins, 35% pre-mixed insulins, and 29% basal insulins by volume. In Japan, Novo Nordisk remains the leader in both the total insulin market (57% share by volume; down from 58% in 1Q12 and 61% in 2Q11) and the modern insulin market (52% share by volume; down from 53% in 1Q12 and 55% in 2Q11). For comparison, both Lily and Sanofi hold a 24% share of Japan’s modern insulin market by volume (since May 2007, Lilly’s share has increased 14 percentage points while Novo Nordisk’s share has decreasednearly the same amount). Currently, more than 98% of the Novo Nordisk’s modern insulin volume in Japan is sold through devices, primarily the FlexPen (for comparison, the NovoPen was also listed as a significant drive of device use in 1Q12).

• As the table below shows, Novo Nordisk has a very strong share of the total insulin market, but directionally, the company lost a little share in Europe, Japan, and China. Similarly, although overall share was the same at 46% for Novo Nordisk’s share of the modern insulin market, its share did decline in both Japan and China.

Table 4: Novo Nordisk’s value share of the total and modern insulin markets

Data based on IMS Health, IMS Midas Quantum data, May 2012

* Data for 12 selected markets representing approximately 60% of diabetes sales in the region

^#Data for mainland China, excluding Hong Kong and TaiwanSales of Victoza totaled DKK 2.29 billion ($396 million) in 2Q12, up 83% from DKK

• Sales of Victoza totaled DKK 2.29 billion ($396 million) in 2Q12, up 83% from DKK 1.3 billion ($241) million in 2Q11. Sequentially, sales increased 15% from DKK 1.99 billion ($350 million) in 1Q12. The sequential comparison is fairly easy, as 1Q12 posted the only negative sequential growth (-5%) since Victoza’s launch in 1Q10 (management had attributed 1Q12’s negative growth to rebate reversals in the US and UK in 4Q11 and to normal seasonal trends in which first quarter sales are typically lower than previous quarters –the latter reason is difficult to verify as 1Q11 had a positive (15%) sequential growth, though it was lower that previous quarters (40%-70%) and reflects initial growth during Victoza’s roll-out), As a comparison, worldwide 2Q12 sales of Byetta totaled $169 million (~$115 million in domestic sales to Amylin and ~$53 million in international sales to Lilly) and worldwide Bydureon revenue amounted to $35 million (~$28 million in domestic sales to Amylin and ~$7 million in international sales recognized by Lilly). According to IMS MAT data from May 2012, GLP-1 agonists account for 5.1% of the global diabetes market by value (up from 4.8% in 1Q12 and 3.8% in 2Q11) and Victoza continues to lead the GLP-1 class with a 65% share by value (up from 62% in 1Q12 and 47% in 2Q11).

• Management provided an overview of Victoza’s global progress and provided information by region. In 2Q12, sales of Victoza were drive mainly by revenue in North America (63% of total Victoza revenue), followed by Europe (27%), International Operations (4%), Japan and Korea (5%), and China (1%). At the end of June 2012, Victoza had launched in 53 countries, with ten more countries (details not provided) preparing to launch the drug in 2H12.

  • North America: Sales of Victoza in 2Q12 reached DKK 1.4 billion ($250 million), up 19% from DKK 1.2 billion ($214 million) in 1Q12. The growth in 2Q12 and 1H12 reflects continued expansion of the GLP-1 class, which now represents 6.3% of the US diabetes care market (up from 6.0% in 1Q12 and 5% in 2Q11). Victoza continues to lead the US GLP-1 agonist market with a 58% share by value (up from 56% in 1Q12 and 41% in 2Q11). During the 2Q12 earnings call, management noted that growth of Victoza in the US continues to be “encouraging” despite the launch of Bydureon. Victoza holds more than a 57% share of total GLP-1 agonist prescriptions in the US (on par with 56% in 1Q12) and a 1.2% share of total prescriptions in the US diabetes care market and more than two-thirds of prescriptions are for the three-pen pack (on par with 1Q12). Victoza is capturing roughly 55% of all new GLP-1 agonist users (up from 53% in 1Q12 but down from 62% in 4Q11 and 60% in 311, 2Q11, and 1Q11). Of course, we’ll be eager to track how these metrics evolve in the coming quarters as Bydureon continues its penetration into the US market and BMS/AZ leverage their combined sales force to market Bydureon and Byetta. Of note, despite the likely delay in the approval of degludec (Tresiba) and degludecPlus (Ryzodeg) in 2013 (at least in the US) as discussed below, Novo Nordisk continues to plan to expand its sales force in 2012 in preparation for the launch of those products. However, until their approval, the company plans to commit this additional sales force to Victoza and the company's insulin portfolio, which we would expect would translate to even stronger sales for these products in the coming quarters.

  • Europe: Sales grew 24% sequentially to DKK 616 million ($106 million) in 2Q12 reflecting the continued rollout across Europe and driven by strong growth in France, Italy, the UK, and Spain (all countries but Spain were cited as revenue drivers in the 1Q12 earnings update). According to IMS MAT data from May 2012, Victoza has a market value of DKK 11 billion ($1.9 billion) in Germany (up 10% from DKK 10 billion [$1.7 billion] in 1Q12), DKK 6 billion ($1.0 billion) the UK (up 20% from DKK 5 billion [$0.87 billion] in 1Q12) and DKK 7 billion ($1.2 billion) in France (flat with 1Q12). GLP-1 agonists account for 5.8% of the European diabetes care market (up from 5.4% in 1Q12 and 4.3% in 2Q11) and Victoza is by far the market leader with a 73% share by value (rising from 59% in 2Q11). This reflects Lilly’s relative lack of attention to building Bydureon overseas – it will be interesting to see how this changes and how fast, with BMS/AZ taking over the responsibility.

  • International Operations: Victoza revenue amounted to DKK 98 million ($17 million), declining 39% from DKK 159 million ($28 million) in 1Q12. However, revenue increased nearly 300% on a reported basis in 1H12, albeit from a low base in 2011. Sales in 1H12 were again driven by strong growth in Brazil and certain Middle Eastern countries (not specified). Management indicated that the GLP-1 agonist class is experiencing “significant expansion” in these markets and currently accounts for 2.5% of the total International Operations diabetes care market (compared to 0.4% in 2Q11), with growth driven by strong uptake in Brazil. Victoza leads the GLP-1 market with a 79% share by value (compared to 18% in 2Q11).

  • Region China: Sales rose 15% sequentially to DKK 23 million ($4 million) in 2Q12. Following Victoza’s launch in China in 4Q11, management noted that early market feedback has been positive and that hospital listings are developing satisfactorily. No details were provided regarding reimbursement, which was previously cited as a barrier to uptake – presumably it is far better, relatively speaking, in Beijing and Shanghai. GLP-1 agonists account for a small portion of the total diabetes care market (0.4% in 2Q12 compared to 0.2% in 2Q11) with Victoza holding a minority share of the GLP-1 agonist market by value (23%).

  • Japan and Korea: Victoza sales reached DKK 115 million ($20 million), up 15% from DKK 100 million ($18 million) in 1Q12. In Japan, the GLP-1 market represents 2.1% of the total diabetes care market (up from 0.8% in 2Q11) and Victoza leads the GLP-1 market with an 82% market share by value (down from 92% in 2Q11).

• Management revised portions of its financial guidance for 2012. Specifically, sales growth for 2012 is now guided to be 9-12% in local currencies compared to the 8-11% growth forecasted in 1Q12 and to be up 7% rather than 4% on a reported basis. Notably, operating profit growth is anticipated to achieve 15% in local currencies, up from the 10% growth expected as of 1Q12. The expectation for a higher level of operating profit growth reflects the increased expectations for sales growth and a deferral of significant costs related to the US launch of Tresiba to 2013, which previously were expected to be incurred in 2012 – although the costs related to the expansion of the sales force in preparation for the launch of Tresiba and Ryzodeg (as discussed above) will still be recorded in 2012. Management noted in the roadshow presentation that “mix” of products is improving – this statement probably reflects Victoza’s high margin. Finally, management anticipates a net financial expense of around DKK 1.95 billion ($337 million) compared to the previous forecast of DKK 800 million ($141 million).

REGULATORY AND PIPELINE UPDATES

• Management did not provide details on why the FDA has scheduled an advisory committee meeting for degludec and degludecPlus. As a reminder, the candidates were submitted to the FDA on September 29, 2011. The PDUFA date for degludec (Tresiba) and degludecPlus (insulin degludec/aspart; Ryzodeg) was initially set for July 29, 2012, but in early June, the PDUFA date was delayed to October 29, 2012 following Novo Nordisk’s submission of a “substantial amount of new data” upon request by the FDA. The FDA’s decision to hold an advisory committee meeting (tentatively scheduled for November 8) came as a surprise since the agency had not requested a committee meeting prior to the initially scheduled July 29 PDUFA date. Management maintained during the call that they have not been informed yet what the focus of the advisory committee meeting will be, but said they hope to provide greater insight on this front during their 3Q12 call on October 31, 2012. We speculate that the likely topics of discussion at the meeting will be the strength of the data that would allow Novo Nordisk to claim superiority regarding hypoglycemia or dosing flexibility, as management affirmed during Q&A that they had not identified any preclinical or clinical safety signals. The FDA has also not yet informed Novo Nordisk of the new PDUFA dates for the two candidates. During Q&A, when asked about how pricing strategies for degludec might be impacted by not receiving the hypoglycemia superiority claim, management declined to consider such a situation, simply stating that degludec would be priced competitively. Also during Q&A, management confirmed that the costs for expanding the sales force in preparation for Tresiba and Ryzodeg’s launches would still be recorded in 2012 and not deferred despite the likely delay of the candidates’ approval (at least in the US) until 2013. In the US, Novo Nordisk will use the additional sales force to help market Victoza and its insulin portfolio.

  • As a reminder, both drugs were submitted to the EMA on September 26, 2011, and a CHMP “Day 180” report was received during 2Q12. An opinion fromthe CHMP (typically expected by day 210) should presumably be expected very soon. The European commission typically provides a final decision on approval within three months of the CHMP’s recommendation, and management mentioned during the call that EU regulatory action for Tresiba and Ryzodeg is expected during 2H12. Novo Nordisk has also submitted NDAs for Tresiba and Ryzodeg in Japan (regulatory action also expected in 2H12), Switzerland, Canada, South Africa, India, Australia, Brazil, Mexico, and Russia.

• Data from the phase 3a DUAL I trial for IDegLira (a fixed-ratio degludec and liraglutide combination product) demonstrate superior glycemic control with IDegLira over each of its individual components. In the recently completed trial, individuals (n = 1,663) with type 2 diabetes inadequately controlled on metformin pioglitazone were randomized in a 2:1:1 ratio to receive IDegLira, Victoza, or degludec, all in a pre-filled pen. As we understand it, participants were allowed to take the three treatments at any time during the day. Notably, at 26 weeks, IDegLira provided significantly greater A1c reductions (1.9%) than degludec (1.4%) or Victoza (1.3%) from a baseline A1c of 8.3% in all three groups. The percentage of participants achieving A1cs of 7% and 6.5% were greater for IDegLira (81% and 70%, respectively) vs. degludec (65% and 48%) or liraglutide (61% and 41%). At the end of the trial, fasting glucose levels were similar between the IDegLira and degludec arms while the liraglutide component of IDegLira provided significantly greater reductions in postprandial glucose (PPG) equally across breakfast, lunch, and dinner (based on CGM and nine-point BGM) – this result likely helped drive the greater A1c reduction observed with IDegLira over degludec. As expected, IDegLira provided significantly better improvements in weight compared to degludec (-0.5 kg [1.1 lbs] vs. +1.5 kg [3.3 lbs]), but not liraglutide (-3.0 kg [6.6 lbs]). Hypoglycemia for IDegLira was characterized as “low,” and management reported that it was statistically significantly lower than hypoglycemia for degludec. As expected, the most frequent side effect of IDegLira was nausea, which occurred in <10% of patients (nausea rates were not reported for either liraglutide or degludec. At the end of the trial, the mean insulin dose was significantly lower in the IDegLira arm than in the degludec arm, while the mean liraglutide dose was within the approved dose range for Victoza (not specified).

  • Management indicated that a number of possibilities existed for positioning IDegLira going forward: 1) as a means to intensify treatment for Victoza patients that are not at their target A1c; 2) as a means to intensify oral treatment that are not resulting in optimal diabetes management; 3) for newly diagnosed patients with a very high A1c; and 4) as a means to strengthen insulin therapy by adding a GLP-1 agonist. We were pleased to see management position GLP-1 as an early option in the course of diabetes treatment. The “unspoken” was that IDegLira would almost certainly prompt a delay in use of rapid acting insulin. From an HCP and patient perspective, we would characterize this as positive since presumably IDegLira should be far easier to teach and take due to minimal hypoglycemia and easy dosing.

  • The second phase 3 trial in the DUAL program (DUAL II; ClinicalTrials.gov identifier: NCT01392573) is expected to complete towards the end of 2012, According to management, regulatory dossiers for the product will then be compiled in the first half of 2013 – presumably, this should allow for regulatory submission by early 2013, as previously guided. This timeline puts IDegLira well ahead of Sanofi’s lixisenatide/Lantus combination product submission expectations. Phase 3 trials for this product are expected to initiate in early 2013. As a reminder, while IDegLira is a fixed-ratio combination of GLP-1/basal insulin, Sanofi’s device will reportedly allow for insulin dose individualization while delivering a 10 mg or 20 mg dose of lixisenatide.

• Progress continues on semaglutide, Novo Nordisk’s once-weekly GLP-1. The first study in semaglutide’s phase 3 program is expected to be a cardiovascular outcomes trial starting in 1H13. As a reminder, in June, semaglutide was selected over once-weekly liraglutide as Novo Nordisk’s once-weekly GLP-1 candidate to advance into phase 3 development. The phase 3 SUSTAIN program will start in the first half of next year, and its phase 3a program alone will enroll >8,000 patients with type 2 diabetes. Management did not disclose the doses to be examined in phase 3, but did say that doses would be selected with the goals of 1) showing superiority regarding A1c reduction compared to the existing once-weekly GLP-1 product on the market (Bydureon), and 2) offering the flexibility of choosing between a fully glycemic efficacious dose or one that may provide additional weight loss benefits, similar to what is available with Victoza in the 1.2 mg and 1.8 mg doses. Management commented that the phase 3 program would be initiated with the cardiovascular outcomes study to enable filing as soon as possible. We are curious to see whether Novo Nordisk will also eventually pursue a weight loss or prediabetes indication for semaglutide, although we anticipate such development programs would be initiated (if at all) only after the drug receives approval for the treatment of type 2 diabetes. The decision to pursue an obesity indication will also likely depend upon the success of the liraglutide obesity development program and the commercial success of that product as an obesity agent once it is marketed. BMS/AZ have a once-monthly GLP-1, courtesy of Amylin, that will be entering phase 3 shortly as well – it will be interesting to see how timing to submission goes for each compound.

• We look forward to hearing more about semaglutide at EASD. In a phase 2 study, semaglutide lowered A1c and weight in a dose-dependent manner (doses ranged from 0.1 mg to 1.6 mg), and, notably, increased the likelihood of achieving an A1c<7% compared to liraglutide. Novo Nordisk will present data from this study at EASD in October (data are currently available in abstract form on EASD’s website here: http://www.abstractsonline.com/Plan/ViewAbstract.aspx?sKey=8e40db00-2d48-40da-891e-e4c9db8d9378&cKey=c653f1a2-0976-4db0-bba6-668fe7af5cb6&mKey={2DBFCAF7-1539-42D5-8DDA-0A94ABB089E8}). In the 12-week trial, 411 patients were randomized to one of five weekly semaglutide doses (0.1 mg - 1.6 mg), open label liraglutide (1.2 mg or 1.8 mg), or placebo (n=43- 50 per group). Semaglutide 1.6 mg titrated by 0.4 mg weekly (the highest dose tested) produced a mean 1.2% placebo-adjusted A1c reduction from a baseline of 8.1%, which was statistically significantly greater than liraglutide 1.8 mg (0.8%; baseline of 8.1%). No other dose of semaglutide was shown to provide superior A1c reductions vs. liraglutide (placebo-adjusted A1c reductions observed were 0.9% for 0.8 mg titrated by 0.4 mg weekly, 1.0% for 0.8 mg untitrated, 0.6% for 0.4 mg untitrated, 0.4% for 0.2 mg untitrated, 0.1% for 0.1 mg untitrated). More patients achieved an A1c of < 7% when treated with semaglutide >0.8 mg (69% of patients on 0.8 mg titrated by 0.4 mg/week, 73% of patients on 0.8 mg untitrated, and 81% of patients on 1.6 mg titrated by 0.4 mg/week) than with liraglutide 1.8 mg (57%). Management noted during Q&A that semaglutide is the only GLP-1 agonist that has shown greater A1c benefits than Victoza. However, withdrawal rates for semaglutide 1.6 mg titrated (27.7%), 0.8 mg titrated (18.6%), and 0.8 mg untitrated (14.3%) due to gastrointestinal (GI) adverse events were higher than those for liraglutide 1.8 mg (10%) or 1.2 mg (2.2%). Overall GI adverse events were reported to decrease upon titration and were indicated to be mild to moderate in severity (overall rates of GI adverse events aside from the causes of withdrawal were not provided).

  • Novo Nordisk announced that it has selected semaglutide over liraglutide as its lead once-weekly GLP-1 candidate; although liraglutide once-weekly could have been less cumbersome and costly to develop since once-daily liraglutide (Victoza) is already approved. As a reminder, management noted during the 4Q11 earnings call thatonce-weekly liraglutide might have been able to skip phase 2 studies using pre-existing Victoza data. However, presumably efficacy would not have been better with once-weekly lira. Various other once-weekly GLP-1 agonists are in development ahead of semaglutide (phase 3 trials for Eli Lilly’s dulaglutide are expected to complete later this year; GSK/Human Genome Science’s albiglutide expects to submit in early 2013). PhaseBio’s Glymera is another once-weekly GLP-1 in development with phase 2b initiation anticipated at the end of 2Q12 (no news of initiation yet). Given that semaglutide could potentially be third to market (and follow the release of the Bydureon pen in 2013), management must have confidence in semaglutide’s improved efficacy (A1c and weight loss) over Victoza, and more importantly, Bydureon. We are curious whether the candidate’s potentially improved A1c control and weight loss benefits will be enough to offset the higher associated nausea rates. Management did note in Q&A that this was the first time a GLP-1 agonist has demonstrated superior A1c reductions to Victoza. We are also curious how Novo Nordisk will position semaglutide in relation to Victoza. Even with Victoza’s continuing strong growth, we assume that Novo Nordisk would be glad to cannibalize Victoza sales with a GLP-1 agent that is both more efficacious and more convenient, particularly since the company could probably price semaglutide at a slight premium to Victoza – though we would assume that in this environment of extensive global cost pressures, pricing at parity might well prompt significantly higher volume.

• During Q&A, management commented that an additional CVOT for Victoza is not currently required to receive an obesity indication approval. Management noted that if Victoza is not approved for obesity before the FDA implements additional cardiovascular guidelines for obesity drugs (which may be published as early as later this year although we imagine it will take longer), they may have to engage in further discussions with the agency, but that as the guidelines currently stand, Victoza’s LEADER trial (its CVOT for diabetes) will provide sufficient CV data for Victoza. We imagine the company has already begun to speak to FDA about this. As a reminder, Victoza’s efficacy and safety as an anti-obesity medication is being evaluated in the phase 3 SCALE program, which includes two ongoing trials. The first trial (ClinicalTrials.gov identifier: NCT01272219) is examining the ability of liraglutide 3 mg to induce or maintain weight loss in obese participants without diabetes over 56 weeks. The study is also investigating whether liraglutide can delay the onset of diabetes (after 160 weeks) in people with pre-diabetes at baseline. The second trial (ClinicalTrials.gov identifier: NCT01272232) is evaluating the ability of 1.8 mg or 3 mg liraglutide to induce or maintain weight loss in obese patients with type 2 diabetes over 53 weeks. The estimated primary completion dates for the two trials are March 2013 and January 2013, respectively so we would assume a second half of 2013 submission for obesity. For our coverage of the FDA’’s advisory committee meeting on obesity CV guidance, see our March 30, 2012 Closer Look at http://www.closeconcerns.com/knowledgebase/r/63ce77cd.

  • No updates were provided on Victoza’s cardiovascular outcomes trial LEADER. For details on this study and other CVOTs being conducted in the GLP-1 class, please see our Novo Nordisk 1Q12 report at http://www.closeconcerns.com/knowledgebase/r/edc801c9.

• Notably, Novo Nordisk has moved a novel oral basal insulin analog (NN1954/OI1362GT) into phase 1 and has initiated a trial for the compound in ~100 healthy adults. The study is expected to complete in 1H13. As of our 4Q11 report, NN1953, a different oral basal insulin analog, had completed a phase 1 study, but no updates have been provided on this candidate since. As a reminder, other companies developing oral insulins include Biocon (IN-105, phase 3; see our Biocon F1Q13 report at http://www.closeconcerns.com/knowledgebase/r/db8093e8), Oramed (ORMD- 0801, phase 2), and Generex (Oral-lyn, phase 3 with submission in India slated for August 2012). This area has been fraught with failure to date. Additionally, MidaSol is developing a buccal insulin product (Midaform Insulin PharmFilm, phase 1; more information is available at http://www.closeconcerns.com/knowledgebase/r/85175809).

• Other product updates included: the approval of NovoMet, the fixed dose combination of repaglinide and metformin (marketed as PrandiMet in the US) in China with launch expected at end of 2013 and the launch of NovoPen 5 (the first new durable insulin pen for adults since 2005) in June in Denmark. This pen has a memory function that displays information related to the last dose administered. Novo Nordisk expects to launch the pen in additional European countries in 2H12, although no details were provided regarding launch in the US.

• In late July, Novo Nordisk launched the Changing Diabetes Barometer (www.changingdiabetesbarometer.com/), which bills itself as a tool for measuring progress in the fight against diabetes. We hope the program will help policy makers grasp the need for much greater resources dedicated toward diabetes prevention, treatment, and education. Projections for 2025 estimate that US diabetes prevalence will be 53 million people, costing over $500 billion (this is comparable to the total budget for Medicare in 2010) compared to 26 million people costing over $180 billion today. For more of our commentary on these daunting statistics, see the July 27, 2012 Closer Look at http://www.closeconcerns.com/knowledgebase/r/de008900.

• In early June, Novo Nordisk and JDRF announced the establishment of an R&D collaboration for the discovery and development of novel immunotherapies for type 1 diabetes. The collaboration will be based out of Novo Nordisk’s new type 1 diabetes research center in Seattle, WA, which is led by the highly regarded Dr. Matthias von Herrath (UC San Diego, La Jolla, CA). For more details on this center, see our January 27, 2012 Closer Look at https://closeconcerns.box.com/s/b07d362785d41b1d2769.

• No updates were provided on NN1218, the company’s ultra-fast-acting formulation of insulin aspart, which completed an initial phase 1 trial in 1Q12. On ClinicalTrials.gov, the compound is listed as recruiting for a phase 1 study examining its pharmacodynamics properties in patients with type 1 diabetes (identifier: NCT01618188). The company planned, as of 1Q12, to select a candidate to advance into phase 3 studies by the end of 2012. Additionally, no updates were provided on the trial initiated in 1Q12 for liraglutide as an adjunct to insulin therapy in people with type 1 diabetes, which is estimated on ClinicalTrials.gov (identifier: NCT01536665) to complete in August 2012, one month earlier than initial estimates of a September 2012 completion. Finally, no updates were provided on the company’s oral long- acting GLP-1 candidates (NN9924, NN9926; both in phase 1).

Close Concerns Questions and Answers

• What was driving lower sales of human insulin sales?

• Is 2014 for semaglutide submission in the ballpark?

• 24% modern insulin growth seemed remarkable, even though it was an easy comparison – was there anything unusual driving this quarter?

• Is there anything documented about people with type 2 diabetes living longer? This seems very logical with more aggressive treatment earlier and more optimized treatment on an ongoing basis but we have not seen yet any published data.

• In China, in 1H12, sales of modern insulins, human insulins, and protein related products increased 27% on a reported basis (14% operationally), compared to a 17% reported growth (8% operationally) in 1Q12. Can you point to what caused the sharp increase here in 2Q?

• Sales of modern insulins, human insulins, and protein-related products grew by 19% on a reported basis in the first half of 2012 in “international operations” – since growth in 1Q12 was 24%, can you point to what caused the relatively large downturn here, which countries?

• Overall, how is competition going in rapid acting analogs? Will Novo Nordisk ever reach 50% market share or more? How does ASP compare between the “Big 3” rapid acting insulins?

• Is any further detail about the launch of the new pen available? When will the US launch be and what “order” will EU countries get it? How much will it cost? Could you combine CGM with it?

• How much is Novo Nordisk benefiting from more global education on GLP-1? What dose on average is NN selling? Is there anything new on the GLP-1 type 1 trial?

• In the semaglutide trial, 0.8 mg dose prompted a 1.5% A1c drop untitrated and 1.4% A1c drop titrated, but just a 1.2% A1c drop for 1.6 mg titrated - that seems odd to have a lower drop for a higher dose?

• Lira for obesity – is there a name for the drug yet? Is 2H13 reasonable guess for submission? Have you heard of any IIT for lira and Qsymia?

• How is NN1954 different from NN1953? How did the phase 1 study go for NN 1953?

• Is there anything you can say about the ongoing generic lawsuit situation with Caraco?

• Is NN using CGM in all its studies now at least part-time?

Questions and Answers

Q: Are there any issues being discussed with the FDA regarding degludec that are not commonly being discussed in the investment community? Is there anything outside of open label, Lantus timing, strength of hypoglycemia data, or flexible dosing that we should be aware of? The background to the question is just that with Victoza, the thyroid issue came up quite late, so is there anything similar happening here? And then, related, I think you commented that you’d have further color during the third quarter results. Just what will you have then that you don’t have now, from your perspective? Finally, thinking about a potential degludec label in the US and what you may be able to promote, I wonder if you could describe, how you are able to market the superior hypoglycemia profile for Levemir versus human insulin while the label for Levemir doesn't have that data explicitly claimed?

A: We have not been told what the specific focus of the advisory committee will be at this point. And of course we are hoping by the end of the third quarter, i.e., at the company announcement in late October, that we will be able to put more color on that because we will know more at that point. In terms of the scope and focus, what I can tell you is that in the case of liraglutide, we had the preclinical findings that are shared by all GLP-1 agonists, i.e., the c-cell findings that occur in rats and mice but not in monkeys and humans. And those were clearly there, as you'll recall, they were heavily debated during that 2nd of April meeting. In the case of degludec, we do not have any preclinical findings suggestive of such a surprise. But we will get back to you hopefully over the next quarter on that one.

I think the other one was about Levemir, where we have got a new label recently, by the way, in the US. What I can tell you is that you can actually promote certain data that are not found directly in your label for a product, if that data set is a) consistent with the label and b) if the data are sufficient to meet the agency's, what we call, evidentiary standards for those particular claims. This is also reflected by such material having to be approved by the FDA's Office for Prescriptions Drug Promotion known as the OPDP and formally known as DDMAC.

Q: Even if you don’t know what the advisory committee might be related to, based on the discussions to date, is there anything outside of the investment community being discussed between yourself and the regulators?

A: My understanding is that the investment community discusses everything, but I’ll have to investigate further and read all those reports.

Q: Have you seen any receivables come back from Southern Europe, and if you have, have you been paid for those? We've seen that with a number of companies this quarter. Relatedly, could you let us know whether any of those debts had been previously written off, which would have positively impacted sales this quarter?

Do you see any future impact from healthcare reforms in 2H12? Clearly, the insulin growth rates are looking very good, but do you see any further reforms coming? We've heard from some companies about potential price cuts in Italy; your perspective there would be great.

A: In terms of receivables in Southern Europe, the way that we actually record the provisions is so that when we make provisions for our receivables, they would actually be included in the selling and distribution line, so it's a selling and distribution cost. The sales would have been recorded. So there's no effect on the sales from a change in the provision levels for our Southern European debt. And if we look at the sales and distribution cost line for the second quarter, and also if we look at it for the first half-year, there's no significant impact from changes in provisioning levels occurring as for the first six months of the year. But changes in provision levels would typically go through as a reversal on the selling and distribution line, but that's not the case for the first six months.

Again, we do make ongoing provisions for the sales we are having if they become more than 90 days overdue, and we are still seeing slow payments from the public part of the system, whereas the traditional distribution system for our insulin that flows through the pharmacies, we are seeing a continued payment scheme and no significant change in our days of sales outstanding in Southern Europe, so I'd say no specific concern there.

In terms of the expectations we had given for local-currency sales growth of 9% to 12%, that includes a continuation of expected healthcare reforms in Europe. It is not so that we are hit in all countries at once; it will be country by country. Currently, we are feeling impact from, especially, a reform in Poland. Overall, I would estimate the annual effect in Europe on European sales to be on the magnitude of 3% to 4% of European sales, and that would be equivalent to close to 1% of global sales, and that is the going rate that we have assumed for the full year.

Q: On IDegLira, you mentioned a significant improvement in the mealtime glucose for all meals. Was there a differential effect on each separate meal? Was there an effect that was higher the first meal after dosing, or was it relatively stable across all meals?

And finally, were all the products were dosed at the same time in the trial?

A: Regarding the mealtime effects of IDegLira, I would like to remind everyone of the paper by Professor Louis Monnier from Montpellier that was published back in 2003 that actually defined the relative contributions of postprandial glucose and fasting glucose to hemoglobin A1c. To summarize: at high A1c levels, 70% of A1c is governed by fasting glucose and 30% by post-prandial glucose – and that's when you have 9% to 10% A1c – while at low A1c levels, around 7% and 6.5%, it is the reverse. That is, 70% of A1c is governed by post-prandial glucose and only 30% by the fasting plasma glucose. And that means that for a patient with 7% A1c, you have to improve the post-prandial glucose control at all three meals to be able to take that patient down to below 6.5%. That is what the Romans call a sine qua non.

Patients were allowed to take IDegLira any point in time, and that means in the morning, in the evening, during the day. And we thought there would be a split between morning and evening users, but early evidence shows that it doesn't make any difference because of the long action profiles of these compounds.

Now, importantly, we not only monitored blood glucose by nine-point self-measured plasma glucoses, but also by CGM. What we found is that there are equal and highly significant reductions mediated by the liraglutide component in IDegLira component across each of the three meals of the day. And that is also reflected by the fact that fasting glucose levels in IDegLira were similar to those in the Tresiba arm, suggesting and proving that it is indeed the post-prandial glucose excursions that are significantly improved toward the levels observed in non-diabetics.

Q: A follow-up on IDegLira, could you perhaps shed some light on your strategy? Now that you've shown superiority in this study to each of the components, how should we see this positioned in the market going forward? You're filing within the next nine or 10 months, so it's rather close to the market, so it would be interesting to learn a bit more about that and any pricing information.

A: In terms of the positioning of IDegLira, the results are not entirely surprising in the sense that we've basically been able to show that when we put the two products together, you get the effect that you would have anticipated if you modeled both. And then because of the glucose-dependent effect of Victoza, you get the ability to drive A1c further down without any untoward effect when making the combo. These are all things we hoped for and anticipated.

Now, the key question here is, of course, how to position this going forward. There are a number of options. One would be to intensify treatment for Victoza patients that are failing or are finding it difficult to retain their A1c. Another opportunity would be to start patients on the combo if they have a very, very high A1c when they are failing on orals or if they were never on orals and are diagnosed with a very high A1c. And finally, there is, of course, also the opportunity of strengthening insulin therapy by adding GLP-1 to insulin therapy, which is in fact already done today and practiced by physicians and patients.

I would also add that this is obviously a very competitive area, and we find no interest whatsoever in disclosing our positioning statement for this drug until we have concluded further studies and also get closer to the actual launch, because of the very clear competitive situation that we would be facing, especially with our dear competitors from Sanofi.

In terms of pricing, there are our certain limits in how we will price this, because if we overprice this, then people would just start using the individual components. And so the pricing is likely to be relatively close to one-plus-one, basically taking the cost of Victoza and adding the cost of Degludec to the combo in the given ratios, and then perhaps a little fee for the nuisance of having had to make it.

Q: Which doses of semaglutide have you selected for phase 3 development?

A: Regarding dosing for semaglutide in phase 3 – first of all the phase 3 program is very big. We will enroll 8,000 patients in the SUSTAIN program and that is of course driven by the cardiovascular guidelines of the FDA and soon to be, actually this autumn, also the European authorities. And that is also why the first trial we're going to start is going to be a cardiovascular trial, where we're going to measure for a couple of years, in high-risk patients, outcomes such that this will enable a new drug application as soon as possible. Now, the phase 2 data that you may have seen on the EASD website are very powerful in terms of the A1c reductions. In fact, it's the first GLP-1 agonist that has shown higher A1c reductions than liraglutide. It normally is the other way around.

We've selected a couple of doses of semaglutide, and let me just say that they are selected in such a way that a) we should be able to beat the existing – or sorry, we're going to set out to beat the existing once- weekly GLP-1 on the market, based on efficacy, simply, and b) we will have a competitive offering where people can either have a fully glycemic efficacious dose or one that may add the benefit of some more weight loss, as is much the case for Victoza today.

Q: Can you remind us how much the 10,000-patient BEGIN/BOOST program for degludec cost to run, and maybe give a ballpark figure for the cost relating to the launch of degludec that's being pushed now into 2013?

A: Yes, on the BEGIN/BOOST trial that we ran for Tresiba and Ryzodeg, I think as a rule of thumb, it's fair to say that it cost us in the ballpark of $100,000 per patient, so you'd get to a number very close to DKK 1 billion or so for this trial, or around US $200 million.

In terms of the Tresiba launch, we have to note that the costs that have been deferred to 2013 are solely the launch marketing costs, whereas the cost of expanding the sales force are still be on the 2012 statement, and of course there will then be a higher focus on the marketing of Victoza in the second half of 2012. So those costs will still be incurred. So I think if you could give a ballpark figure, it would be something on the magnitude of half a percentage point on the S&D ratio that has been impacted.

But again, the precise effect of that will have to come down to what the outcome of the advisory committee meeting is, and then how quickly can we ramp up for a potential launch in the US. But in the ballpark of half a percentage point, I think that's a reasonable estimate.

And then on the US, of course it is interesting, ironically enough, that nothing is disappointing that is not good for something, and the fact that we are now have to plan a launch of degludec or Tresiba in the beginning of 2013 means that we have full marketing power on Victoza and our modern insulins. And they are actually the products that are driving the predominant part of the growth in the first half, and so this will strengthen our situation in the US, where we have a good momentum behind our insulin business.

Q: US pricing trends across the interim portfolio for the industry continue to surprise on the upside, so just wondering what your thoughts are there. You've been pretty relaxed about the ability to put pricing through in the US. I'm wondering if you've become more confident, or has anything changed as you get the crystal ball out and look out a few years.

A: Yes, it is perhaps surprising that we have been able to follow the competition in the sense of increasing prices, in general, on insulin products to the tune of some 6%, 8%, 9%, on that order of magnitude. You've got to bear in mind, though, that these are list price adjustments, and they're often associated with very significant rebates, so the flow-through on our net sales is much, much less than the increases that you hear about.

In terms of going forward, I think our picture is still the same that we would be expecting that we'd see relatively modest price increases going forward. But it's all determined by the competitive situation, and so far, it has predominantly been Lilly and ourselves that have determined the price level of insulins in the US., with of course Sanofi driving the basal price levels, given they're the leaders in this segment in the US. But there's still caution about the outlook going forward.

Q: Coming back to the obvious question about the advisory committee meeting, I'm still puzzled by the FDA decision to convene this meeting so late in the file. I realize it's their prerogative to do so and that you're not going into a lot of detail with respect to the ongoing discussions. But just to be absolutely clear, I know the question's already been asked earlier in the call, but you talked about no pre-clinical signal. Just I want to be absolutely clear. In the past, you've been on record as saying there's no safety signal in the data package. First, I wanted to make sure that's the case.

And then secondly, if that is the case, can we assume that it's all focused on the strength of the data allowing you to make claims on hypos and dosing flexibility?

A: First I’d like to remind everyone that things have happened since the former generation of insulins, such as Lantus and Levemir, at least along three lines. One is that in 2007, the FDA issued their Amendment Act, according to which new molecular entities, as a default position, would undergo an advisory committee unless there were specific reasons why they need not do so. And since then, we've also seen the advent of two sets of FDA-mandated guidelines within diabetes, a general development guideline and a CV guideline.

Specifically regarding degludec: first is that degludec has been proven and has been submitted in the dossier to work as insulin and only as insulin, period. It doesn't bind to any level of receptor. It has an on and off rate constant that is compatible with that of human insulin, and so on and so forth. So degludec is going to be an insulin with an ultra-long action profile. And by the way, the way it's degraded by the body, whether it's animals or humans, is into natural ingredients such as carbon dioxide, amino acids that can reenter the pool of substrates in the body, and water.

Finally, we have not had any statistically significant findings at all pre-clinically or clinically in the statistical analysis plans that have been incorporated in the various protocols and studies, aside of the benefits we have seen on hypoglycemia, which we've discussed many times.

Q: I realize this is fairly early in the year for you to give us any specifics around next year, but just as we think about some of the pushes and pulls on the margins, given the cost- phasing into next year, how should we think about the broader operating margin picture for next year?

A: Yes, we are a bit early in terms of discussing 2013. But you could say, in a way, that we anticipated that we would have cost pressures in 2012 because of the launch of Tresiba and Ryzodeg. These launch costs are primarily being deferred now to 2013, of course with some potential for impacting Europe and Japan from launches late in the year, but that will be so marginal that it really won’t have any impact. So we anticipate that we can look at a sales growth in the area of around double-digits, and then we will have a cost pressure next year, which would put pressure on our ability to grow operating margins.

Our long-term operating margin target has been 15%. This is where we are guiding right now. So I would anticipate that next year, that number will come down as a result of the launch of these products globally.

Q: What would you do in terms of pricing for degludec/degludecPlus if you don't get your [hypoglycemia] superiority claim?

Could you update us as to any formulary negotiations you may have been able to make for next year? This seems to the beginning of the formulary season. How should we look at that for next year?

A: I'd like to get an approval first before we start to speculate on that. I don't want to speculate on a hypothetical scenario of not being able to get the hypo claims. We have strong hopes and optimism that we will get some kind of hypo data, or if not claims, that we can promote these products on. And we hope, in fact, that we will be able to use the advisory committee meeting to demonstrate to the agency and the experts how important hypoglycemia still is. And so the advisory committee meeting functions, of course, to review risk, but we also have to highlight the benefits, and we're certainly going to do that. So I don't what to speculate on a negative hypothetical scenario because I don't think it's going to happen.

Comments on the formulary – we reacted to quite aggressive pricing strategies from our competitors about a year ago. You can say, by doing so, we sent a signal that we want to stay competitive, and so our action on formulary negotiations will be dealt with according to the competitive scenario for each individual managed care organization, the upside with exclusives, and when it's non-exclusive, where our current market share is with the particular healthcare organization, whether they can pull patients and switch patients from a competitive product to ours. It's a very complex picture, and you can rest assured we will stay competitive also going forward.

Additionally, there is very significant price competition occurring in the declining human insulin segment in the US.

Q: Regarding liraglutide in obesity – I know, of course, a lot of discussion is whether or not you need to do additional CV studies ahead of approval. Can you tell us, in your discussions with FDA, when can you actually gain clarity on this?

A: Well, first of all, as you're probably also aware, as would be typical for a company like Novo Nordisk, we have already discussed with the agency and came to the understanding under the circumstances that prevailed at that point when we had the discussion. We did indeed ask these questions about cardiovascular profiling, including the LEADER program done for liraglutide in diabetes, and all of the overall studies that have been done at the doses that are relevant in diabetes, and then what we intended to do in the Phase 3 SCALE program. And at that point when we had this discussion, it was deemed to be sufficient evidence for the cardiovascular safety of liraglutide without further pre-approval of 3.0 mg studies related to cardiovascular risk in the obesity setting.

Of course, we are also cognizant of the notion that the FDA is coming up with obesity CV guidelines, potentially later this year, and if and when they are published, it is of course highly relevant to discuss with the agency how to handle those products that have been caught in the inter-regulum period between one and the other set of guidelines. But at this point there are no specific plans to do so because we've had the discussion previously.

Q: With respect to China, we see some volatility in your growth rates in China, with 2Q12 being quite good. Are there any external factors affecting this? How will this play out long- term? Are we still at the 15% insulin growth level?

A: The China business for Novo-Nordisk, more or less, is only a diabetes care business, and the guidance that we've provided historically would be expected to grow in the ballpark of 15%, and I think that's still a very meaningful guidance for 2012, and the numbers I say here are of course in local currencies.

If you look at it year-to-date, we have been looking at a growth rate in China in the ballpark of 13%, but I do note that the comparisons going into the third quarter will be easier. And actually, the second quarter did see a higher level of growth. I think local-currency growth in China was about 18% in second quarter, and that was reflecting the impact from the changed price occurring in China, where there was some insulin restriction they already imposed in Q2 last year, and that impacted the growth in Q2 alone. And Q3 last year we had the change in pricing for the human insulins in China occurring, and hence, when we get to the full year, I'm pretty confident we'll be in the 15% ballpark.

There is an uncertainty factor surrounding potential price reform for impacting the modern insulins, and especially for Novo Nordisk, NovoMix in China being a very significant product. And we saw, when the price reform was impacted last time, there was an immediate significant stocking effect, and hence, whether or not there will be a pricing reform impacting NovoMix in China in the second half of the year could impact overall growth numbers, but 15% would be the ballpark.

Q: Regarding insulin markets – despite the obesity/diabetes pandemic in the developed world and increased life expectancy, insulin volumes don't seem to be growing in Europe or Japan. I know you've spoken about this before, but I was wondering if you could offer any updated thoughts on the dynamics occurring in these markets.

A: If we look at the overall insulin market globally, we're talking about a volume growth of around 5%. This has been coming down somewhat due to a slowdown in the volume growth in North America, and also, as you already alluded to, a not-recent slowdown in Europe and in Japan. However, we do see the trend from growing wealth and urbanization in China and in International Operations, where we have growth rates in China and IO at around 18% in both regions. But, as we were talking about, the very low growth rates in Europe of around 2% to 3%, and very low growth rates in Japan, even less than the 2%. United States is currently trailing at around 3% to 4%.

The reason behind that seeming slowdown, you can speculate, our current speculations are that the very expanded use of DPP-4 inhibitors, which we see typically in Europe and the United States, are impacting the number of patients with a new treatment option, which they can take advantage of for a certain period of time before they need to do something else, and that something else being going on to GLP-1, which is a more powerful treatment, or a combo product of GLP-1 and insulin, or an insulin. So it’s a question of getting to a point of equilibrium, where the DPP-4 inhibitors can penetrate the market to the extent that we get the same flow-in and flow-out of patients that are identified in therapy.

This usually takes two to three years. We saw it when saw the introduction of insulin sensitizers as a new class of drugs that was added to the armament of treatment of type 2 diabetes, and that led to stagnating growth rates for, if I'm not mistaken, three to four years, and then we later saw growth. The underlying dynamics are not going to change. More people are getting diabetes, people with diabetes are living longer, and so we still uphold the overall estimate that the volume dimension should grow approximately 5% per year.

-- by Jessica Dong, Nina Ran, Ben Kozak, Lisa Rotenstein, and Kelly Close