As Other Biotechs Sink,
Amgen, Genentech Soar

SOUTH SAN FRANCISCO, Calif. -- A few dozen miles north of the emptying office parks of hard-hit Silicon Valley, a different sort of high-tech corporate campus is still humming with activity.

Construction work proceeds at three huge new buildings here that will double laboratory space at Genentech Inc., whose revenue and profits are growing at roughly 20% a year. Instead of facing layoffs, employees at the biotechnology company still get bounties to recruit new scientists. Parking lots remain filled with BMWs and Mercedes-Benzes.

Four hundred miles south, rival Amgen Inc. also is thriving. Fresh off the largest biotech acquisition ever, its $11 billion purchase of Immunex Corp. this year, Amgen is putting up buildings and hiring more employees. The company is one of the most profitable in the world: Setting aside accounting charges for the acquisition, Amgen's net profit margin of 32% in the first nine months of 2002 was slightly wider than Microsoft Corp.'s.

Much of the economy may be in the tank, but a very few companies continue to mint money in biotech, a field in which companies generally seek to produce drugs through genetic science. Genentech and Amgen are the prime examples. They are a stark contrast to the traditional pharmaceutical industry, whose giants are having increasing difficulty turning research into new medications and face tough competition from generic drugs as well.

They also stand apart from the travails of the broader biotech industry. A quarter-century after the industry's birth, most companies are still unprofitable ventures that can take a decade or more to win regulatory approval for experimental drugs. Overall, stocks in the sector have fallen 41% this year as the 2000 biotech-stock bubble continued to deflate, forcing many cash-burning start-ups to search for fresh investment.

INDUSTRY PULSE   For more health coverage, visit the Online Journal's Health Industry Edition at wsj.com/health1 and receive daily health e-mails2.

Genentech and Amgen haven't been immune to the stock slump: Genentech's stock is down 35% this year and Amgen's is down 15%, dragged lower by pessimism hanging over the sector. So far, though, the two are doing surprisingly well financially for a simple reason: They hold franchises on a few specialty drugs that have shown themselves to be recession-proof.

How they got to this point tells a lot about how biotech works. Unlike other high-tech companies, drug manufacturers generally succeed or fail based on science bets made years earlier. The risks are especially high in biotech, where the science is newer and the companies are smaller, and therefore more vulnerable if a badly needed product fails in testing. Genentech and Amgen saw their bets pay off big thanks to a rare combination of research savvy, sharp legal maneuvering and no small amount of blind luck.

Genentech's current success traces back to 1995, when the company decided to proceed with development of two cancer drugs that were on the verge of being abandoned. 

One of those drugs, Herceptin, is now a standard treatment for a type of metastatic breast cancer. It was one of the first drugs to target a genetic defect that contributes to cancer and works by gumming up a cellular protein called Her2 that regulates growth in some fast-growing breast cancers. Doctors like it because it doesn't have the toxic side effects of chemotherapy and can be used along with traditional treatments.

"Herceptin has clearly changed the treatment of breast cancer," says Charles Loprinzi, chairman of the oncology department at the Mayo Clinic in Rochester, Minn.

No Sure Thing

In 1994, the experimental drug wasn't at all a sure thing. Genentech scientists knew that 25% to 30% of aggressive breast cancers appeared to express abundant amounts of the Her2 protein. But while Herceptin, a genetically engineered antibody designed to target Her2, had shown promise in early tests, a midstage study had narrowly failed to show that the drug made a difference in cancer patients. Genentech's product-development committee recommended that the company license out the drug rather than pursue it directly.

DRUG BENEFITS   A handful of successful drugs are propelling solid growth at Genentech and Amgen, despite the biotechnology sector's doldrums   GENENTECH   DRUG NAME CONDITION SALES (1) (IN MILLIONS) Herceptin Breast cancer $278.6 Rituxan Non-Hodgkin's lymphoma 816.3(2)     AMGEN   DRUG NAME CONDITION SALES Epogen (EPO) + Aranesp Anemia 1,849 Neupogen (G-CSF) + Neulasta Chemotherapy-related infections 1,302     1-For the first nine months of 2002 2-Revenue as shared with Idec Pharmaceuticals   Sources: the companies

But Arthur Levinson, Genentech's former research chief and new CEO, convened his top executives to take a second look at the drug. Her2 had been under study since the mid-1980s, and Dr. Levinson was a believer in targeting the protein, in part because it was one of the first linked specifically to tumor growth. The latest data hadn't met strict criteria for statistical significance, but it had been close enough that Dr. Levinson was willing to "roll the dice" and overrule the development committee.

"The question was, do you want to sink $500 million into a phase three trial that might fail?" Dr. Levinson says. "Some people thought we shouldn't do it. Others said the biology is just too compelling."

Herceptin's survival still wasn't assured until late 1995, when Genentech's board voted a large, across-the-board increase in research spending. That allowed a then-marginal product such as Herceptin to be funded. "Herceptin might not have survived an ordinary budget year," Dr. Levinson says.

The board's move was a gutsy one, considering the pressures Genentech was under. Its CEO had just been ousted for personal financial improprieties. And the company was under federal investigation for aggressive sales and marketing practices. (Genentech later paid $50 million to settle with the government, without admitting wrongdoing.)

Continued testing of Herceptin showed greater benefits than did the early trials, and the Food and Drug Administration approved it for late-stage metastatic breast cancer in 1998. Last year the drug racked up sales of $347 million, out of total Genetech sales of $2.2 billion.

"We slogged and plowed our way through that process, and at every stage, it was slightly more positive, and slightly more positive," Dr. Levinson says. "There was never a stage where we looked and said, 'This will be a killer drug.' "

Genentech had made big bets before. The company, founded in 1976, was the first to commercialize the new technology of gene-splicing. That technique, which gave birth to biotechnology, made it possible to churn out mass quantities of medical proteins by inserting human genes into bacteria and other cells. The company made its name with a string of accomplishments -- among them, ways of producing human insulin and human growth factor -- and quickly became a leader in the business of producing protein-based medications.

Genentech's other hit cancer drug, Rituxan, is now the best-selling branded cancer drug in the U.S., with $819 million in sales last year. (Many older chemotherapy agents are now generic drugs.) Like Herceptin, it has relatively few side effects because it is an antibody aimed at a specific type of cell, the white blood cells that can cause non-Hodgkin's lymphoma.

In the mid-1990s, Dr. Levinson had taken an interest in using antibodies to treat non-Hodgkin's lymphoma after a senior executive's wife died from the disease. Genentech found a struggling San Diego biotech, Idec Pharmaceuticals Corp., with an antilymphoma antibody no one wanted. Antibodies had a long record of failure as drugs, and non-Hodgkin's lymphoma, which affects roughly 50,000 people a year, seemed too small a niche. "Doors slammed everywhere we went," recalls William Rastetter, Idec's chief executive.

But Genentech had been working with antibodies for years. Before long, the company agreed to pony up as much as $57 million to help Idec develop the drug.

The results were exceptional. In patients with low-grade lymphoma -- a slow-growing version of the cancer that is virtually incurable with chemotherapy -- Rituxan proved it could frequently batter the disease into quiescence, at least temporarily.

Approved by the FDA in 1997, Rituxan's sales have almost doubled every year since its introduction. Although sales growth has started to slow, Genentech hopes to expand use of the drug to more aggressive cases of non-Hodgkin's lymphoma as well as to autoimmune conditions such as rheumatoid arthritis.

Like Genentech, Amgen relies on a few successful products. Its chief moneymakers, two bioengineered proteins, have become the best-selling biotechnology drugs in history. Last year, Amgen raked in $3.5 billion from sales of the two drugs, out of total sales of $4 billion. Erythropoietin, or EPO, fights anemia by promoting the growth of red blood cells, and granulocyte colony stimulating factor, or G-CSF, does the same for white blood cells and is useful in treating infections.

When introduced in 1989, EPO revolutionized the treatment of anemia. Patients with kidney failure, for instance, don't make enough red blood cells and are constantly anemic, which at the time required frequent blood transfusions. EPO proved more effective than transfusions and additionally spared dialysis patients the risk of blood-borne disease such as hepatitis or AIDS.

Today, EPO has vastly improved survival rates. "It's been huge," says Scott Liggett, the medical director of a dialysis center in Hastings, Neb. "It used to be a rarity to see patients in their 80s on dialysis. Now it's a common occurrence."

Among the tactics to which Amgen owes today's profits: It began development of second-generation versions of both drugs in the mid-1990s, bringing them to market just as growth of the original drugs was beginning to slow. Amgen can also credit a series of unexpected legal victories that preserved its lucrative franchise in the face of steep odds and even a major marketing misstep.

Capitalists' Creation

Founded in 1980 as Applied Molecular Genetics, the Thousands Oaks, Calif., company was largely a creation of venture capitalists looking to cash in on the early biotech craze. Its initial product candidates ranged from a hepatitis B vaccine and biologically produced dye for blue jeans to animal-growth hormones.

Then an Amgen scientist discovered the EPO gene in 1983. In an early study, EPO completely eliminated the need for blood transfusions in all the kidney-dialysis patients who received it, and it quickly became clear that it would become one of biotechnology's first true wonder drugs.

Hungry for cash, Amgen sold EPO marketing rights to drug giant Johnson & Johnson in 1985. In that hastily drafted agreement, Amgen reserved only the U.S. kidney-dialysis market for itself. J&J moved to lock up European and other non-U.S. markets, and pioneered a lucrative business treating anemia-related fatigue in cancer patients in the U.S. and abroad. It has fought a series of battles with Amgen over the U.S. dialysis market, with each side claiming the other has encroached on its territory.

Amgen's biggest strategic move of the last decade came in the mid-1990s, when executives realized that an idea bubbling out of its research labs might offer a way of freeing it from the hated marketing agreement with J&J.

The company's researchers had long pondered ways of making both EPO and G-CSF more effective, for instance by making them last longer, since both required frequent shots. By changing just six amino-acid components of EPO, the researchers found they could alter its structure to make it more difficult for the body to excrete the drug through urine, doubling or tripling the time it remained active.

Amgen executives soon realized the modified drug might not be bound by the 1985 marketing agreement. J&J immediately objected, and the matter moved to arbitration. Observers figured Amgen might, at best, win more favorable royalty terms from J&J. But in a surprising move, the arbitrators in 1998 handed down a binding ruling that freed Amgen to compete head-to-head against J&J.

Approved last year and now named Aranesp, the new version of EPO is off to a relatively strong start. Analysts predict the new drug, as well as the second-generation version of G-CSF, known as Neulasta, will eventually pull in more than $1 billion each in annual sales.

The big question for Genentech and Amgen these days is whether the decisions they're making now will serve them equally well a decade down the line. Genentech's pipeline of new drugs hit several setbacks this year, due to regulatory delays and unexpected failures. Next year, the company could potentially release two new drugs for allergies and psoriasis -- if the FDA is accommodating.

Amgen's challenge is to convince doubters it is more than a two-trick pony dependent on EPO, G-CSF and their derivatives. Chief Executive Kevin Sharer, who took over two years ago, says he isn't disappointed with the company's drug pipeline but adds that he is committed to improving it. He has installed a new team of drug-development executives, several from big pharmaceutical companies such as Merck & Co. The team now boasts that it has moved as many drug candidates into human tests over the past two years as the company had in the previous 10.

Amgen also has to successfully digest Immunex, which it acquired earlier this year, and prove that it can turn that company's rheumatoid-arthritis drug, Enbrel, into a blockbuster. Hamstrung by a shortage of production capacity, Enbrel has steadily lost ground over the past year to a competing treatment produced by J&J.

No one knows better than these executives how uncertain the future can be in biotechnology, where a company's prospects are only as good as its next hit drug. Reflecting on decisions he's made over the past year, Genentech's Dr. Levinson suggests asking him again in five years if they turned out to be good ones. "Maybe I'll say, 'Gosh, I was an idiot,' " he says. "I just don't know that yet."

Write to David P. Hamilton at david.hamilton@wsj.com³.

Updated November 18, 2002 8:59 a.m. EST