Oral agents are the fastest-growing drug class for diabetes treatment. Approximately 6.8 million (out of 13.1 million diagnosed and treated) Type II diabetes patients are treated with oral medications in the U.S. Combination pharmacotherapy, in conjunction with dietary and exercise intervention, is the standard of care for Type II diabetics. Combinations of oral antidiabetic agents are routinely being used to help patients reach aggressive HbA1C targets, such as the 6.5% level recommended by the American College of Endocrinology. Nevertheless, recent studies show that only about 7% of patients with diabetes achieve their target goals for blood sugar, cholesterol, and blood pressure. These patients represent a significantly under-treated population that will drive future utilization growth for diabetes medications.

There are a plethora of drugs in the antidiabetics pipeline. Of the major drugs in development, those in the GLP-1 agonist and DPP-IV inhibitor classes show the most promise for reaching the blockbuster status of the thiazolidediones. Amylin’s Exenatide is a potent agonist for the GLP-1 receptor. Exenatide is an injectable drug whose primary advantage lies in the fact that unlike most other oral antidiabetic drugs which work by a single mechanism, it works by several mechanisms: it stimulates insulin secretion, slows emptying of the stomach and inhibits production of glucose by the liver. It also appears to suppress appetite and helps weight loss. Novartis’ LAF237 and Merck’s MK-0431 are two DPP-IV inhibitors in Phase III clinical development that also act via GLP-1 by preventing its degradation by the DPP-IV enzyme. The DPP-IV inhibitors offer similar benefits to Exenatide but are orally bioavailable.

We have updated and added the 16 drug revenue models for Type II diabetes listed at the end of this report. As highlighted in our 2005 outlook, we expect Amylin to have a good 2005 with catalysts from Exenatide and Symlin. While both Exenatide and Symlin will not be as big as LAF237 and MK-0431, Exenatide LAR could be. Data on Exenatide LAR in 2005 could also help Amylin, but it is still early in development.

BACKGROUND

Diabetes affects an estimated 194 million adults worldwide, including over 18 million in the United States. Within the United States, there are 13.1 million people diagnosed and 5.1 million people remaining undiagnosed. The World Health Organization projects that the number of people with diabetes will double to 366 million by 2030. Last year alone, more than 3.2 million deaths were attributed to diabetes or diabetes-related causes which represented the 6th leading cause of death in the US.

Insulin regulates the uptake of glucose, amino acids, and fatty acids and the production of protein, glycogen, and triglycerides. The failure of this hormone to exert those effects contributes to diabetes. The two main types of diabetes are Type I and Type II. A minority (10%) of patients suffer from Type I diabetes mellitus, which is caused by pancreatic B-cell failure and leads to absolute loss of insulin. This form develops most frequently in children and adolescents, however, it is increasingly presenting itself later in life. About 90% of patients with diabetes have Type II diabetes, also known as non-insulin-dependent diabetes mellitus (NIDDM), which is characterized by the failure of sensitive tissues to respond to insulin (i.e. insulin resistance).

The oral antidiabetic agents represent the largest drug class for diabetes treatment, with sales in 2003 of $8.3 billion and a 19% compound growth rate between 1988 and 2002. Since entering the market in 1999, the two thiazolidediones agents, Actos (Takeda) and Avandia (GlaxoSmithKline), have created a blockbuster class of oral diabetes agents. Worldwide sales in 2003 for Actos and Avandia were $1.8 billion and $1.4 billion, respectively. Both Actos and Avandia are known to cause serious side effects. Labels for both drugs carry a warning of the risk of cardiac failure and other cardiac effects due to edema, and both are known to cause weight gain and decreased hematocrit. Consequently, there are a number of biotech and pharma companies racing to bring out the next generation of novel oral antidiabetic agents.

Most of the agents have been combined by physician recommendations, but manufacturers are increasingly interested in the development of fixed-dose combinations of oral agents to help prolong their drug franchises and improve patient compliance with therapy. Patent expirations have prompted the introduction of combination pills such as Glucovance (Bristol-Myers Squibb), Avandamet (GlaxoSmithKline) and Metaglib (Bristol-Myers Squibb). Combination pills are only available as branded products and are more expensive. The benefits of an easier routine may make the added cost worthwhile, if the combination is able to provide better control of diabetes. There are also fewer pills to swallow and only a single insurance copayment for combination products.

DRUGS IN DEVELOPMENT - Near Term Outlook

Combination pills such as Avandaryl (GlaxoSmithKline) and Actoplus Met (Takeda) are on the near-term horizon, but there are three novel antidiabetic agents that are also close to FDA approval: Exenatide (Amylin/Eli Lilly), Symlin (Amylin), and Muraglitazar (Merck & Co.).

Amylin’s Exenatide and Exenatide LAR (NDA stage)
On September 1, 2004, the FDA formally accepted the New Drug Application (NDA) requesting approval of Amylin Pharmaceuticals’ and Eli Lilly’s Exenatide (synthetic exendin-4) for the treatment of Type II diabetes. With positive clinical results on both average glucose level reduction and body weight reduction with minimal hypoglycemic side effects, Exenatide is almost certain to be approved by the FDA in early 2005. The likely 2005 approval of Exenatide will invigorate Amylin’s anti-diabetic drug pipeline, given the difficulties it has had with obtaining approval for Symlin (pramlintide), a drug for insulin-using diabetics who are unable to achieve adequate glycemic control.

Exenatide is an injectable drug for Type II diabetics, many of who are currently taking oral antidiabetic drugs for glucose control. Clinical trials have demonstrated Exenatide’s efficacy for glucose level and body weight reduction with minimal hypoglycemic side effects and positions the drug to complement and compete with market-leading oral therapeutics such as Actos and Avandia in the lucrative Type II diabetic market which comprises 90% of all diabetics. Exenatide helped patients with glucose control in 3 distinct late-stage clinical trials, each of which tested Exenatide in combination with different diabetes drugs, and it also contributed to patient weight loss. The weight loss is significant since nearly all patients achieving tight control with existing medication gain weight in response to the glucose being pushed into cells by the therapy.

Exenatide is a potent agonist for the mammalian (glucagon-like peptide 1) GLP-1 receptor. Exenatide’s advantage lies in the fact that unlike most other oral antidiabetic drugs which work by a single mechanism, it works by several mechanisms: it stimulates insulin secretion, slows emptying of the stomach and inhibits production of glucose by the liver. It also appears to suppress appetite and helps weight loss. The multiple glucoregulatory actions of GLP-1 opens the door to a new class of antidiabetic drugs.

The main drawback of Exenatide, and other potential GLP-1 agonists, is that it has to be injected. Exenatide is injected subcutaneously twice a day at breakfast and dinner. Exenatide LAR (Long Acting Release) is being developed and may allow once-a-week to once-a-month administration. It is expected to be approved by the FDA in 2007.

Amylin’s Symlin (NDA stage)
On September 20, 2004, Amylin submitted a response of the FDA’s second approvable letter for Symlin. A response from the FDA is expected by March of 2005. Symlin is an injectable drug for insulin-using diabetics who are unable to achieve adequate glycemic control. For the Type I patient who has been taking insulin all along, Symlin will be used in addition to the insulin as an adjunct therapy. The second group of patients that Smylin will target is that comprising of Type II patients who are using insulin. Among adults with diagnosed diabetes, about 12% take both insulin and oral medications, 19% take insulin only, 53% take oral medications only, and 15% do not take either insulin or oral medications. It is estimated that there are approximately 4.5 million people in the United States who use insulin each day to control their diabetes.

Symlin’s advantage is in helping to control glucose levels without increasing the risk for hypoglycemia or increasing weight when added to insulin regimens. The main drawback is that patients must inject Symlin 3 or 4 times a day, on top of any insulin injections. Consequently, it is not likely to be used by more than 10-15% of insulin-using diabetics.

Bristol-Myers Squibb Muraglitazar (NDA stage)
On December 23, 2004, Bristol-Myers Squibb Co. filed an NDA requesting approval of Bristol-Myers Squibb and Merck & Co.’s Muraglitazar for the treatment of Type II diabetes. Muraglitazar is a dual peroxisome proliferation-activated receptor (PPAR) alpha and gamma agonist. The thiazolidinedione drugs Actos and Avandia bind with PPAR-gamma to enhance insulin-mediated transport of glucose into adipose tissue (fat cells) and skeletal muscle. Three PPARs have been identified – PPAR-alpha, PPAR-gamma, and PPAR-delta. Studies have suggested that co-activating both PPAR-alpha and PPAR-gamma works better than activating just one of the two. Modulating PPAR-alpha has been shown to lower LDL, lower triglycerides, and raise HDL. Therefore, dual PPAR agonists are a novel class of antidiabetic drugs that have the potential to treat blood glucose and lipid abnormalities in patients.

Despite the promising aspects of dual PPAR agonists, their clinical programs face a significant safety challenge. In animal studies, there seems to be a trend towards unusual cancers in rodents exposed to these compounds. In 2003, Merck & Co. and Kyorin Pharmaceutical Co. Ltd stopped Phase III development of their MK-767 dual PPAR agonist, citing cases of a rare form of malignant tumor in long-term safety trials in mice. In October 2002, Pharmacia and Japan Tobacco’s reglitazar was dropped following completion of Phase II trials due to safety concerns. Whether the observed carcinogenicity is exclusive to murine biology is not known, but it is unlikely that the FDA will approve any dual PPAR agonist until it is satisfied that the cancer risk is exclusive to mice. We do not know if Murglitazar has shown any murine cancer issues.

DRUGS IN DEVELOPMENT – Mid Term Outlook

Despite its attractiveness as a target, the potential of GLP-1 is limited primarily by its rapid degradation by the ubiquitous enzyme dipeptidly peptidase-IV (DPP-IV). The rapid inactivation of GLP-1, in addition to its renal clearance, contributes to a short half-life of less than two minutes. Consequently, a novel class of antidiabetic agents is targeting the inhibition of DPP-IV as a way to improve the duration of endogenous GLP-1 activity. The primary advantage of the DPP-IV inhibitors over the GLP-1 agonists is their oral bioavailability. Novartis’ LAF237 and Merck’s MK-0431 are two DPP-IV inhibitors in Phase III clinical development. In addition to the two DPP-IV inhibitors, there are two thiazolidindione agents that are also 2-3 years away from FDA approval: Galida (AstraZeneca) and R483 (Roche).

Novartis’ LAF237 (Phase III stage)
On June 6, 2004, Novartis presented results at the annual scientific meeting of the American Diabetes Association demonstrating that glucose reductions in response to combined treatment with LAF237 and metformin were sustained for one year. Moreover, the patients did not gain weight. There was another advantage in treating patients with LAF237 in that patients did not experience nausea. Based on the strength of those data and other findings from Phase II studies, Novartis launched a full Phase III clinical trial program in the second half of 2004. DPP-IV is an enzyme that is expressed in many tissues and acts on a number of substrates, so the question is whether it will cause any unexpected adverse events. There have been no reports of unusual safety issues. However, pursuing DPP-IV inhibition has some potential risks. The DPP-IV enzyme has multiple functions. It inactivates some substrates, such as GLP-1, but also activates others, such as some neuropeptides and molecules important in activating white blood cells of the immune system.

Merck & Co.’s MK-0431 (Phase III stage)
Merck is keeping pace with Novartis’s LAF237 with its DPP-IV inhibitor, MK-O431. Although minimal data has been available on this compound, it has been reported that in one study in patients with early-stage Type II diabetes there was a two-fold increase in active endogenous GLP-1 concentrations. In an oral glucose tolerance test, MK-0431 administration was stated to reduce plasma glucagons area under the curve and increase plasma insulin area under the curve. Furthermore, MK-0431 was generally well tolerated.

AstraZeneca’s Galida (Phase III stage)
Like Bristol-Myers Squibb’s Muraglitazar, Galida is a dual PPAR alpha and gamma agonist. AstraZeneca is developing Galida to treat both Type II diabetes and metabolic syndrome. Minimal data has been available on this compound, but on October 6, 2004, AstraZeneca announced during its annual business review meeting in London that the Galida Phase III program is progressing well. It has also been reported that its Phase IIb program included more than 1,000 patients and that 2-year carcinogenicity studies have been completed for Galida.

Roche’s R483 (Phase III stage)
R483 is a thiazolidindione (PPAR-gamma agonist) that has been reported to improve insulin sensitivity and inhibit glucose production in addition to offering superior efficacy. In May 2004, Roche announced during its R&D Day that it had made a decision to progress R483 into Phase III trials. The company reported that it was preparing for 5 large Phase III trials enrolling approximately 5,000 patients. In July 2004, Roche announced that, following new guidance by the FDA on the thiazolidinedione class of agents, it had decided to revise its Phase III development plans for the compound and to wait for the results of ongoing long-term toxicity studies. The toxicity studies will be completed in 2005.

GLP-1 Agonists Versus DPP-IV Inhibitors

The GLP-1 agonists and DPP-IV inhibitors may eventually inhabit different niches of the diabetes market. The DPP-IV inhibitors, LAF237 and MK-0431, are expected to compete with existing oral drugs such as Actos and Avandia. They could be recommended for early-stage diabetics. The DPP-IV inhibitors may also be used in combination with other existing antidiabetic drugs. The GLP-1 agonist Exenatide, because it must be injected, will likely compete with other injected drugs used in later-stage diabetes. It could become another option for avoiding last-resort insulin therapy, and may even replace insulin entirely in some Type II patients.

Near to Mid Term Pipeline for Type II Diabetes Drugs

The near to mid term pipeline for Type II diabetes drugs looks as follows:

REVENUE MODEL UPDATES

We have updated and added 16 drug revenue models for Type II diabetes:

Actos-Takeda-Approved
While Actos’ sales momentum is slowing, the drug should continue to perform well in the coming years. We project peak U.S. revenues for Actos of $1.99 billion in 2006, with peak worldwide revenue of $2.66 billion in the same year.

Amaryl-Sanofi-Aventis-Approved
Amaryl’s peak sales days are behind it and will decline going forward. We project U.S. revenues for Amaryl of $195.9 million in 2005, with worldwide sales of $510.4 million in the same year.

Avandamet-GlaxoSmithKline-Approved
We predict Avandamet will continue to grow revenues in the coming year, primarily due to the overall growth in diabetes. We project peak U.S. revenues for Avandamet of $683.3 million in 2011 and peak worldwide revenues of $1.27 billion in the same year.

Avandia-GlaxoSmithKline-Approved
Avandia should enjoy a relatively large part of the market in the coming years. We project peak U.S. revenues for Avandia of $1.84 billion in 2008, with peak worldwide sales of $2.37 billion in 2006.

Avandaryl-GlaxoSmithKline-NDA-81% LOA (Average)
Avandaryl should see nice growth over the next several years, though will generate a smaller amount of revenue than some of the upcoming drugs. We project peak U.S. revenues for Avandaryl of $708.7 million in 2012, with peak worldwide sales of $1.48 billion in the same year.

Muraglitazar-Bristol-Myers Squibb-NDA-81% LOA (Average)
Should Muraglitazar be approved it we feel it has less revenue potential than many of the other diabetes drugs. We project peak U.S. revenues for Muraglitazar of $753.1 million in 2012, with peak worldwide sales of $1.57 billion the same year.

Exenatide-Amylin-NDA-96% LOA (15% Above Average)
While the market for Exenatide will be smaller, its expected approval should bode well for Amylin in 2005. We project peak U.S. revenues for Exenatide of $363.2 million in 2008, with peak worldwide revenue of $706 million in the same year. Currently we feel Amylin to be undervalued based on its 5 and 10 year drug pipeline values and upcoming catalysts that could propel the stock.

Symlin-Amylin-NDA-87% LOA (6% Above Average)
A Symlin approval in diabetes would make for a nice addition to Amylin’s future portfolio and could provide additional benefit to the drug’s chances in the obesity setting. We project peak U.S. revenues for Symlin for Type II diabetes of $367.9 million in 2012, with peak worldwide revenue of $779.6 million in 2013.

Galida-AstraZeneca-Phase III-67% LOA (Average)
If approved, Galida would be a much welcome revenue generator to AZN. We project peak U.S. revenues for Galida of $813.3 million in 2013, with peak worldwide revenue of $1.71 billion in the same year.

R483-Roche-Phase III-67% LOA (Average)
Should R483 gain approval, it should perform in a similar fashion to Galida. We project peak U.S. revenues for R483 of $813.3 million in 2013, with peak worldwide revenue of $1.65 billion in the same year.

LAF237-Novartis-Phase III-72% LOA (5% Above Average)
We feel LAF237’s revenue potential is among the largest in the diabetes market. We project peak U.S. revenues for LAF237 of $2.02 billion in 2013, with peak worldwide revenue of $4.27 billion in 2014.

MK-0431-Merck & Co.-Phase III-69% LOA (2% Above Average)
Similar to LAF237, MK-0431 should generate a large amount of revenue. We project peak U.S. revenues for MK-0431 of $2.02 billion in 2013, with peak worldwide sales of $4.34 billion in 2014.

Exenatide LAR-Amylin-Phase II-52% LOA (22% Above Average)
Good data on Exenatide LAR in 2005 could further boost Amylin’s stock. We feel its market potential is quite a bit greater than the short release version of Exenatide and as such would be a great addition for Amylin. We project peak U.S. revenues for Exenatide LAR of $1.72 billion in 2015, with peak worldwide revenue of $3.54 billion in the same year.

BVT.3498-Biovitrum-Phase II-36% LOA (6% Above Average)
While its potential approval is further down the road, BVT.3498 has the potential to be a multi-billion dollar drug worldwide. We project peak U.S. revenues for BVT.3498 of $990.2 million in 2015, with peak worldwide revenue of $2.08 billion in the same year.

T131-Amgen-Phase II-30% LOA (Average)
T131 should be on the smaller side of the diabetes’ revenue generators. We project peak U.S. revenues for T131 of $348.4 million in 2015, with peak worldwide revenue of $731.1 million in the same year.

677954-GlaxoSmithKline-Phase II-30% LOA (Average)
We project peak annual U.S. revenues for 677954 of $495.1 million in 2015, with peak worldwide revenue of $1.44 billion in the same year.