The Food and Drug Administration's back-to-back withdrawal of two drugs in three days hints at a deeper crisis in the agency.

   In March, the FDA withdrew Rezulin after it was implicated in 63 reported deaths, and 27 cases of life-threatening disability caused by liver damage.

   Two days later, the agency withdrew the heartburn drug Propulsid after 80 reported deaths from cardiac arrest and 347 life-threatening cardiac emergencies.

   These facts, taken directly from FDA press statements, might suggest a drug safety system operating reasonably well. Some would argue that despite the unfortunate death toll - the rough equivalent of, say, an airplane crash- millions of people have benefited from such innovative new drugs.

   But a closer look reveals an FDA in crisis, clinging to its historical reputation for vigilance even as it exposes an unprecedented number of consumers to new risks, and stifles concerns raised among its own ranks.

   The casualty toll from the two FDA-approved drugs recently withdrawn is far higher, perhaps 100 times higher, than that acknowledged by the agency and the two manufacturers involved (who privately negotiate with the FDA and then give out the same numbers).

   Because the system relies on voluntary reports, it captures only a small percentage - anywhere from 1 out of 10 to 1 in 10,000 - of drug adverse events, including deaths. That small number is reduced even further to arrive at official totals, because many additional reports are excluded on technicalities - such as insufficient information on the form, or because a patient has another problem complicating the picture. With Rezulin, for example, at the point the manufacturer, Warner-Lambert, and the FDA had publicly disclosed only 33 deaths or liver transplant cases linked to the drug, voluntary reports indicated at least 155 deaths overall, and hundreds more serious injuries.

   It is reasonable to estimate that Propulsid and Rezulin each caused thousands of serious injuries and hundreds of preventable deaths. And they are but two of the nine drugs, each with its own casualty toll,withdrawn for safety reasons since September 1997.

   The pharmaceutical industry and the FDA have gambled with the public's health at a time when mistakes can be deadlier than ever. In recent years, as a flood of new drugs has entered the market, the industry has perfected aggressive marketing techniques that often result in literally millions of people taking a new drug within months of its approval. If a dangerous new drug does reach the market, the consequences are far more devastating than a decade ago, when new drugs made a far slower entry into the marketplace, affecting far fewer people.

   The response of the FDA's senior drug regulators to these rising risks has been troubling. Since 1994, the FDA's drug approval unit has been led by Janet Woodcock, a rheumatologist who has been with the agency since 1986. Many key decisions have been publicly attributed to her senior deputy, Murray Lumpkin, another career FDA physician. So far, these two appear to have shaped current drug approval policy more than Jane Henney, the commissioner who has been in office less than 18 months.

   In medical journal articles, special reports, and press statements, Woodcock and others have repeatedly claimed that safety standards have not been lowered. Whether true or not, the agency could hardly say otherwise. But with each additional drug withdrawal, that claim became steadily less credible.

   The next line of defense has been the kind of spin-control usually employed by embattled politicians. For example, when the antibiotic Raxar was Withdrawn last October for safety reasons, the agency made no public statement whatever, and as a result got minimal media attention.

   The FDA timed its announcement on Rezulin and Propulsid so late in the day that the story was not included on network television evening news, and print reporters' ability to get any perspective other than the FDA's was limited. Meanwhile, safety concerns have been rising among FDA staff. When the Public Citizen Health Research Group anonymously polled FDA medical officers who perform the hands-on review of new drug applications, 34 said the pressure to approve new drugs had increased since 1995, and 19 officers named specific drugs they believed should not have been approved.

   By early this year, the staff concern had boiled over into rebellion. Staff members leaked agency e-mails documenting that a growing number of FDA specialists were urging the immediate withdrawal of Rezulin. But Warner-Lambert strongly resisted, and Woodcock and Lumpkin sided with the company.

   The agency's response to the controversy was to investigate - and threaten with dismissal - Robert Misbin, the medical officer who had initially backed approval of Rezulin, but later sought its immediate removal from the market.

   Investigators wanted to know whether Misbin had leaked agency e-mails voicing concerns about Rezulin safety. And, in what may be the low point in this agency's history, heavy-handed FDA investigators showed up in the office of Leo Lutwak, a medical officer and supervisor in the office that reviewed Rezulin, according to several press accounts. They threatened Lutwak, who is 72, with five years in jail unless he confessed that he was the one leaking e-mails from FDA specialists concerned about Rezulin. He told them they were wrong.

   This intimidation is a radical change from the agency's long and Honorable tradition of openness, in which its experts were free to discuss concerns with the press and public.

   Two pivotal moments illustrate how the FDA's new tilt toward industry has put consumers at risk:

   The first involves Duract, a painkiller similar to ibuprofen, aspirin, and Aleve. Early testing revealed that Duract could damage the liver, and this did not escape the attention of veteran FDA medical reviewer Rudolph Widmark. He noted that some other drugs in this family had also shown liver toxicity, but concluded Duract was the worst he had seen. It looked like an easy call.

    Does the world need an alternative to aspirin and ibuprofen noted mainly for its higher liver toxicity? But would the agency say no to the manufacturer, Wyeth-Ayerst, which had invested millions of dollars to develop what must once have seemed a promising new drug? The issue was appealed to Lumpkin, who tried to compromise. Duract was approved, but with a warning to doctors about its liver toxicity, and a recommendation limiting its use to only 10 days. But millions of people with long-term pain routinely take similar drugs for weeks or months. It was like unveiling a new car with a stern warning not to drive it over 35 miles per hour.

   And, then, could Wyeth-Ayerst actually sell an expensive alternative to aspirin with a liver toxicity warning that it could only be used for 10 days? So effective is the industry at marketing, and so uncritical are prescribing physicians, that more than 2.5 million prescriptions were written in the 10 months before Duract was withdrawn because of the expected, predictable, and entirely preventable liver deaths.

   Companies can now rack up sales with extraordinary speed if the FDA opens a window of opportunity. Before its withdrawal, Propulsid had reached nearly $1 billion in sales each year. Warner-Lambert captured about $1.8 billion in additional sales from the time Rezulin was withdrawn in Britain in late 1997 until its withdrawal here this month.

    Rezulin is at the center of the second example illustrating the FDA's tilt toward industry, according to an account first published in the Los Angeles Times.

    If it had been up to the original medical officer, John Gueriguian, Rezulin would never have reached the market in the first place. He believed that its benefits were minimal and poorly documented, and noted its potential toxicity both to the liver and the heart. He had written his preliminary findings in a draft report that was circulating among colleagues at the FDA and may have reached the company.

   But before he could finalize his review, Gueriguian was summarily removed from the Rezulin case. In a meeting with Mary Taylor, a senior Warner-Lambert executive, Gueriguian had used vulgar language to describe his negative view of the drug. After Warner-Lambert complained, Lumpkin reprimanded Gueriguian and banned him from further work on Rezulin.

   More mysteriously, his point-by-point critique was never filed and made part of the official Rezulin record. With the most experienced reviewer out of the way, and Rezulin still rolling forward on the fast track to approval, the liver toxicity of Rezulin was largely overlooked for many months - until the death reports started flooding in. Even then, it was 2 1/2 more years and thousands of additional liver cases later before the FDA finally took Rezulin off the market.

   Whether one looks at the big picture - nine drugs withdrawn in four years -or inspects individual episodes, the story is largely the same: The agency has accommodated industry by accepting risky drugs, and when these drugs triggered concerns, simply issued warnings to doctors, rather than promptly removing the drugs from the market.

    Industry marketing has portrayed each of these risky new drugs as a breakthrough with unique mechanism of action. In fact, the withdrawn drugs were largely "me-too" drugs, ones that offered consumers few new benefits, compared to safer alternatives already on the market. Rezulin was the 11th drug to treat adult-onset diabetes. There are at least five safer alternatives to Propulsid for heartburn. Duract was one among 18 similar painkillers; Posicor was the 104th drug for high blood pressure, and Raxar was the 13th fluoroquinolone antibiotic to reach the market.

    The pharmaceutical industry insists the current system is working well, and from its perspective, it may well be. Profits have reached recordlevels. But, in a push for quick approval for new drugs, the tilt to industry has gone too far.

   The nation needs a strong and independent FDA. Just as we have the nonpartisan National Transportation Safety Board as a watchdog over airline safety and the Federal Aviation Administration, we also need an independent panel to monitor the FDA and prevent further deterioration of the vital drug safety system. Until then, perhaps we should heed the suggestion of US Representative Henry Waxman, a California Democrat, who has called for an outside investigation into what is happening at this troubled agency.