A long-struggling biotech startup is finally ready for its day in the sun. Its groundbreaking drug lands before the Food & Drug Administration after wowing doctors at various medical meetings with great clinical data. Then, seemingly out of the blue, the FDA rejects the drug, saying it wants more tests to prove its safety. The stock collapses, losing 60% of its value in just a few days, and the huge pharmaceutical company that has been helping to develop the drug pulls out of the deal.
No, this is not another recap of the troubles besetting ImClone Systems Inc. (IMCL) and its Erbitux cancer drug. The events above happened in 1999 to Scios Inc. (SCIO) and its Natrecor drug for acute congestive heart failure: proof that, in biotech, everything new is old again. One after another, biotech companies have been running into delays ranging from months to years after bringing their drugs to the FDA. Or their drugs have washed out in late-stage clinical trials.
These setbacks are occurring just when biotech is reaching critical mass, with close to 400 drugs now in various stages of human testing. Although many of these drugs will never pan out, those that do could have a huge impact on the treatment of disease. But first, biotech firms and FDA officials must figure out how to clear up the many hazards that have dragged out the process of testing and evaluating these novel treatments--to the detriment of patients.
Scios is a classic example. The FDA's rejection of Natrecor came in April, 1999, when the agency overruled its own outside advisory panel, which had recommended the drug a month earlier by a 5-to-3 vote. The FDA said the 500-patient trial used to support the application wasn't big enough to prove safety. That caused development partner Bayer to pull out of the deal, leaving tiny Scios to finance another year of clinical trials on its own. Natrecor was finally approved by the FDA in July, 2001, the first new treatment for deadly congestive heart failure in 14 years.
Scios and ImClone are not anomalies. In the past several months, a dozen or more biotech drugs have run into problems (table). The FDA review of Corixa Corp.'s (CRXA) cancer treatment, Bexxar, has been delayed twice because the agency needs more time to review the application filing. Chiron Corp. (CHIR) has run into a similar problem with its new HIV blood test, Procleix. That application was still pending a year after filing, before finally being approved on Feb. 28. Amylin Pharmaceuticals Corp. (AMLN) won a likely-to-approve letter from the FDA in October for its diabetes drug--contingent on the company's running more safety trials that will keep the drug off the market for at least another year. Even Idec Pharmaceuticals Corp.'s (IDPH) innovative lymphoma drug, Zevalin, was delayed for six weeks after receiving a likely-to-approve letter because of manufacturing concerns. (It won clearance in February.)
ImClone may still recover from its own bad stumble. After a meeting with the FDA on Feb. 26, the company said it may be able to resubmit its Erbitux application this year with existing data from U.S. and European trials, rather than having to run new trials.
Biotech executives, doctors, and FDA officials all see some common problems with drug delays: underfunded clinical trials; overenthusiastic CEOs who refuse to accept that their clinical trials, or the drugs themselves, may be deficient; an understaffed and leaderless FDA that may also be overcautious; and extremely complicated manufacturing processes. But what really has the industry spooked is that the FDA seems to be tightening up its requirements for approval of cancer drugs--the largest class of biotech agents under development.
Because cancer is so often fatal, the agency has often in the past accepted trials that proved only that a tumor responds to a drug. But increasingly, for cancers where chemotherapy or radiation is already doing some good, the FDA wants evidence that a new drug actually extends life. This "survival" data can take years to accumulate. "I think the FDA has raised the bar too high," says a biotech exec who asked not to be named.
Few biotech executives want to go on record criticizing the FDA. As Vaughn M. Kailian, vice-chairperson of Millennium Pharmaceuticals Inc. (MLNM), notes, "the main asset that most of us have is our relationship with the FDA." But off the record, many say the FDA has grown too conservative.
The agency, though, may simply be responding to the flip side of the problem: sloppy or incomplete drug applications. "I call it founder's syndrome," says one executive. "They discovered the drug, they are convinced it works, and they don't have anyone on their board who will question that assumption. So when the FDA tells them to slow down, they ignore them." This attitude really irks Tamar D. Howson, senior vice-president for corporate development at Bristol-Myers Squibb Co. (BMY), which owns 20% of ImClone. "I wish someone would invent a pill that would shrink the size of a biotech CEO's ego," she said at a recent industry forum.
These conflicts have come to a head in the cancer arena. Many oncology drugs have been approved on the basis of early stage clinical trials that involved only 100 or so patients and no control group, in which patients receive a placebo. In such trials, the patients must be "refractory," meaning they have failed all standard therapies. Rather than demonstrate that the drug significantly extends life--a tough proof in such patients--the trials needed only to show that the tumor shrank in size.
There are two problems with such trials. The company must have a rigid protocol for determining if, indeed, the patients failed standard treatment, and an equally rigid standard for measuring tumor shrinkage. All of this becomes that much murkier with many of the new drugs, such as ImClone's Erbitux, that are meant to be used in combination with existing chemotherapy.
But many biotech executives insist there is nothing murky about the goal. "It is really, really hard to shrink a solid tumor," says Dr. Susan Desmond-Hellman, executive vice-president for product development at Genentech Corp. (DNA) "Sure, survival is a wonderful endpoint, but those studies are very long and take lots of patients."
Companies have another major concern: cost. Getting through all three phases of clinical trials typically required for a new drug approval can run anywhere from $300 million to $800 million. "Everybody underestimates the cost of drug development," warns Idec Pharmaceuticals CEO William H. Rastetter.
The FDA feels that companies rushing their trials bring many of the problems on themselves. Kathryn C. Zoon, director of the agency's Center for Biologics Evaluation & Research (CBER), which reviews most biotech drugs, concedes that the FDA often works closely with companies to design pivotal clinical trials--only to ask for more data later. But, she says, "we try to caution that those agreements are pending only if there are no untoward adverse events." No matter how comfortable the FDA feels about a trial initially, surprises can change its view. "We don't usually pull these things out randomly," says Zoon.
And the agency can move boldly. Last spring, it approved Novartis' (NVS) Gleevec drug for a particularly deadly form of leukemia in a record 3 1/2 months, based on only a very small, early stage trial, with no control arm, and tumor response as the endpoint. In 1998, the agency stunned COR Therapeutics Inc. (CORR) when it approved Integrilin, for angina, after an outside review panel had raised concerns about the drug. The FDA sometimes overrules its outside panel's approval recommendation (witness Natrecor) but rarely ignores a negative review. COR, recently acquired by Millennium Pharmaceuticals, had no idea the FDA would vote in its favor, says Kaillian, then CEO of COR: "A company's relationship with the FDA is always extremely fluid."
The industry would like less wobble. Carl B. Feldbaum, president of the Biotech Industry Organization, says "biotech companies don't want to decrease the rigor of the process, but they want to make it a little more predictable. There are many complaints of inconsistency throughout the agency." He would like to see more definitive guidelines for clinical trials, so companies are clearer on what is required. In general, the FDA feels no compulsion to respond to such criticism. But a top official bristles at this argument. "What's happening is that companies are giving less complete data," he insists.
Many of these issues might be more easily resolved if the FDA had a commissioner. President George W. Bush has yet to name a permanent head for the huge agency, although on Feb. 25 he did appoint a deputy commissioner--a veterinarian, Lester M. Crawford--to act as senior official until the real thing comes along. Staffers are likely to be extremely cautious without any guidance from the top, say executives. That caution may well be warranted, but while the biotech industry and the FDA point fingers, patients continue to suffer.
Corrections and Clarifications
"New drugs: Why so many delays?" (Science & Technology, Mar. 11) said Chiron Corp.'s HIV blood test, Procleix, was pending Food & Drug Administration approval. The drug was approved on Feb. 28, a year after the application was filed and a day after BusinessWeek's issue went to press.
By Catherine Arnst in New York, with John Carey in Washington and Arlene Weintraub in Los Angeles