The San Diego Company That Got There First

In 1998, a San Diego biotechnology company watched its last big partner give notice, saw its share price fall to under a dollar, and cut roughly three quarters of its staff 6. Seven years later the same company launched two first-in-class peptide medicines within six weeks of each other, one of which was the first of the GLP-1 drugs to reach the market. Seven years after that it was sold for $5.3 billion, to buyers who were paying for its future as much as for its present 5. The company was Amylin Pharmaceuticals, and its story answers a question that the obesity-drug boom has made urgent: why is being first in a field so often not enough?

This is a business history. The pharmacology of the two peptides at its centre is described in other pieces, and nothing here is guidance on any compound. The prescription medicines mentioned are named for their place in a corporate and regulatory timeline.

Oxford, 1987: a peptide in the amyloid of diabetic pancreases

Pathologists had known for decades that the islets of the pancreas, the clusters of cells that make insulin, acquire deposits of a glassy, insoluble material called amyloid in many people with type 2 diabetes. Nobody knew what it was made of. In 1987 a group at Oxford, led by Garth Cooper, purified the main component from the amyloid-rich pancreases of diabetic patients and found it was a peptide of 37 amino acids 1. Independent work in Sweden reached the same molecule, which was called islet amyloid polypeptide. The Oxford group called it amylin.

What made the finding more than a pathology curiosity was that the peptide was a hormone. It is made by the same beta cells that make insulin and released with it. A molecule that forms plaque in disease was also an ordinary part of normal physiology, and a peptide of that kind was an obvious thing for a drug company to pursue 4.

A Hybritech veteran starts a company around it

Amylin Pharmaceuticals was incorporated in San Diego in 1987, led by an alumnus of Hybritech, the local monoclonal-antibody company whose success had made San Diego a centre of biotechnology 3. Its founding idea was the hormone itself. Natural human amylin, however, was a poor drug candidate: the same property that makes it form plaque in disease makes it clump in a vial. The company's chemists solved the problem by borrowing from the rat, whose version of amylin does not aggregate, and substituting three residues in the human sequence with the rat's prolines. The result, pramlintide, was a stable analogue of a peptide that was famous for being unstable 4.

The 1990s: partners come and go

Developing a diabetes drug takes more money than a young company has. Amylin raised capital in the public markets and signed a collaboration with the pharmaceutical division of Johnson & Johnson to develop pramlintide, and for several years the project seemed well funded. But the clinical programme was slow, and the data were not clean enough to carry the bet. The pattern of the decade was familiar: partners arrived with large cheques and left when the numbers disappointed.

1998: under a dollar a share

Matters came to a head in 1998. After two phase III studies failed to reach statistical significance, the share price collapsed, and the partner gave notice that it would end the collaboration; the company's own filings tie the restructuring that followed to that notice 6. Trade press at the time reported that Amylin cut its workforce by about three quarters, to a few dozen people, and its shares traded for well under a dollar by the end of the year.

Companies in that position usually do not survive. Amylin did, with a skeleton staff, a lead product that still had to be proved, and a sceptical market. Whether it would have lasted another year without a second product is a question the record does not answer.

The lizard peptide licence that changed the company

The second product came from an unlikely direction. In the early 1990s John Eng, a researcher at a Veterans Affairs hospital in the Bronx, had isolated a peptide called exendin-4 from the venom of the Gila monster 2. It resembled the human incretin hormone GLP-1 but lasted far longer in the body, because the lizard's version resists the enzyme that destroys GLP-1 within minutes. Amylin acquired rights to the peptide in the mid-1990s, and the compound that would become exenatide waited for its turn while pramlintide absorbed the company's attention.

In 2002 Amylin entered a collaboration with Eli Lilly to develop and market exenatide, bringing in the large partner and funding it had lacked since 1998. The other history of exenatide, from the lizard to the clinic, is told on this site in the piece about the Gila monster and the GLP-1 drugs. What matters here is that a licence signed when the company was in trouble turned into its most valuable asset.

2005: two approvals in six weeks

In March 2005 the US Food and Drug Administration approved pramlintide, sold as Symlin, as an add-on for people with diabetes who were already using mealtime insulin. On 28 April it approved exenatide, sold as Byetta, the first GLP-1 receptor agonist to reach the market for type 2 diabetes, and it went on sale that June 3. A company that had been close to extinction seven years earlier had two first-in-class peptide medicines in the span of six weeks, and within a year the company had roughly doubled in size.

The long-acting version, the Lilly divorce and the sale

The original exenatide had to be injected twice a day. A once-weekly formulation, using microspheres that release the peptide slowly, was the obvious next step and was approved in early 2012. But by then the commercial situation had turned. Lilly's partnership, the engine of the 2002 rebirth, had ended, and the terms of the unwinding left a payment obligation to Lilly. Competitors were launching rival GLP-1 products: one from Novo Nordisk that was injected once daily, and weekly products from other companies that would follow.

In the summer of 2012 Bristol-Myers Squibb agreed to buy Amylin for $31 a share in cash, a 10 per cent premium to the previous close. The headline equity value was $5.3 billion. Including roughly $1.7 billion of Amylin's debt and the obligation to Lilly, the total value was about $7 billion. AstraZeneca then joined Bristol-Myers Squibb in a partnership for Amylin's diabetes portfolio, paying $3.4 billion to do so 5. Bristol-Myers Squibb later closed the San Diego operations, and at the end of 2013 AstraZeneca agreed to buy out its partner's share of the diabetes business.

Why the first mover did not become the winner

There is no single explanation, and a careful account should not pretend there is. Amylin had a small company's resources and a large company's problem: it had to teach prescribers an unfamiliar drug class, sell it through a sales force it was still building, and fund the next product from the profits of the first. Its partner relationship with Lilly, valuable at the start, became a constraint at the end. And the next generation of competitors was backed by companies with decades of experience in diabetes and the capital to sustain long development programmes.

What the record does show is a pattern common in drug development. The company that proves a mechanism is not necessarily the company that profits most from it. Amylin demonstrated that peptides mimicking an incretin hormone could be developed, approved and sold, and the lessons it paid for were available to everyone who came after.

Where the GLP-1 articles take over

Amylin's own molecule, the hormone from the pancreases of diabetic patients, outlasted the company. Analogues of amylin remain an area of research, and the pharmacology of the amylin receptor is described in the technical pieces on the sister sites in this network. For the story of what happened to the GLP-1 class once the large companies took it up, the pieces on the year the obesity drugs worked and on the chemist who gave GLP-1 a week carry the narrative forward. The small company in San Diego is where the commercial story began.