Anti-medicine man.

IN 1991, when David Kessler became Commissioner of the Food and Drug Administration, he promised dramatic reforms to change the slow and ponderous ways of the agency, which regulates products accounting for 25 cents of every consumer dollar. He "would teach the elephant to dance," he said. So he has -- to a tune called by politicians.

Early in his administration, Kessler chose a high-profile case calculated to demonstrate his toughness on industry on the evening news. He didn't go after defective heart valves, or a death-inducing drug. He went after orange juice. Not because it was spoiled, contaminated, or unwholesome, but because it was labeled "fresh" even though it was made from concentrate. For that reason alone, the FDA sent federal marshals to confiscate 15,000 gallons of juice. On 60 Minutes, Kessler explained that it wasn't very hard to tell the difference: "My grandmother could have told you. I mean, it wasn't fresh." So consumers could easily tell by the taste that the juice came from concentrate, but they could not be allowed to decide whether they liked it enough to buy it.

Kessler followed this decision with an even higher-profile and more costly ban on silicone breast implants, although there was little evidence of health risks from them (see p. 50).

The agency's politicization accelerated before my eyes during the Clinton Administration. Direction on science policy (and sometimes even on civil-service personnel matters) was now provided by a political commissar, Jerold Mande, a former Senate staffer for Al Gore who conveyed the White House's instructions to Kessler. Linda Suydam, acting deputy commissioner for operations of the FDA, blandly described his portfolio as "areas that are of high visibility and/or of interest to the White House."

A case in point is the FDA's 1993 review of a new female contraceptive. This contraceptive had a 26 per cent failure rate, much higher than alternatives; so the agency was reluctant to approve it. But someone in the office of Donna Shalala, the Secretary of Health and Human Services, called the FDA to make it clear that HHS wanted this product approved. Approval followed immediately.

Biotechnology provides perhaps the most illuminating example of politics at the agency. Prior to the Clinton Ad-ministration, the FDA had a consistent and generally positive 15-year track record of regulating biotech. But some people in the Clinton Administration, influenced by the kind of primitivist environmental views propounded by Vice President Al Gore in Earth in the Balance, oppose biotechnology on principle. The Administration therefore set out to undermine the scientific paradigm that had been the basis for biotech policymaking.

Signaling this reversal, the FDA announced last year that it would soon require manufacturers to notify the agency before marketing foods manufactured with new biotech techniques, like the use of high-precision recombinant DNA (or gene-splicing), regardless of the risks posed. Foods manufactured using older, cruder techniques like hybridization were exempted from the rule. Confronted by a Republican Congress intent on the use of scientific risk analysis by regulators, the FDA shelved the biotech notification proposal, at least temporarily. But this volte-face itself reveals the unscientific basis of the FDA's approach.

Another example of the new approach was the absurdly prolonged review of the long-shelf-life tomato. This took four years, compared to five and eleven months respectively for the first two new biotech therapeutics in the early 1980s -- human insulin and human growth hormone.

Or take the review of the milk-stimulating drug, bST (bovine somatotropin). I was head of the FDA's Office of Biotechnology at the time, and knew that Jerold Mande had relayed instructions from the White House to Kessler, who then directed a vigorous FDA-wide search for reasons not to approve bST. Two completely superfluous meetings of advisory committees were held during the final stages of the evaluation process, as a kind of fishing expedition for arguments against approval. bST was reluctantly approved nonetheless. Now, however, the FDA is damaging its marketing potential by taking no action against stores that have signs proclaiming "NO BST." Since cows' milk naturally contains bST, these labels are inaccurate, a violation of federal law; yet the FDA has done nothing.

Kessler has even eliminated the two policy offices having extensive involvement with the biotechnology industry (the Office of Biotechnology and the Office of Small Business, Scientific, and Trade Affairs). Meanwhile, he has moved dozens of FDA employees from all over the agency to work on the anti-tobacco initiative. The result has been many critical jobs understaffed or undone, new drugs and other products not approved and therefore not available to patients, and inspections and other regulatory actions delayed.

The FDA's press releases, to be sure, tell a different story. In January, the FDA proclaimed with great fanfare that 1994 had been a "record-setting" year for drug approvals. It maintained, for example, that approval times for biologics (such as vaccines and biotechnology and blood products) had dropped precipitously, and they noted that 15 "major biologics actions" were taken. But the FDA's real success was only in manipulating statistics.

Some of this distortion is simple enough: reporting medians instead of averages, for instance, makes the numbers look better. Some of it involves a "gaming" of the drug evaluation process to produce better numbers: since the marketing approval of a drug is the last step of a lengthy process, the FDA can lift its approval rate by making the earlier steps so difficult that few manufacturers reach the stage of requesting it. FDA evaluators would then have more time to speed through the few applications that came in. It's like a quarterback going after a record for completions by repeatedly throwing short passes, though the plays lose yardage.

This appears to be what has occurred. During the past few years, the FDA has introduced various impediments to drug development, including extraordinarily stringent rules on the reporting of drugs' side effects; restrictions on "promotional activities" (which prevent companies from even convening focus groups to determine how to make packaging and labeling more user-friendly); and new controls over manufacturing that can cause interruptions in drug production during clinical trials for even the most trivial change in procedure.

As a result, the number of applications submitted to the FDA for marketing approval, in particular of biologics, has declined: during FDA fiscal year 1993, ten were submitted, compared to eight in FY 1994. The absolute number of products approved is declining as well. According to the Pharmaceutical Research and Manufacturers of America, in 1994 the FDA approved 22 drugs, down from 25 in 1993. In short, products are being further slowed in an already sluggish testing pipeline. Bear in mind that these products could help treat diseases such as arthritis, schizophrenia, multiple sclerosis, and AIDS.

As a consequence of the FDA's hyper-regulatory, "zero-risk" mindset, it takes 12 to 15 years and more than a third of a billion dollars to develop a new drug in the U.S. -- a price tag that applies whether the product ultimately serves one hundred patients or one hundred million. The FDA is blocking the development of new biotechnology that has the potential to provide farmers with innovative tools that can reduce the use of agricultural chemicals, and to offer consumers tastier and more nutritious food choices, such as leaner meats and more healthful cooking oils. These new policies afford no conceivable public-health advantage. Indeed, the reduction in numbers of new products may actually harm public health. At a time when the idea of regulation in the public interest is being questioned, the FDA's politicization can only make regulation seem less legitimate.

The agency's response to the criticism that these kinds of decisions have understandably provoked has been troubling. Agency officials have unequivocally warned drug and biotechnology companies: if you're tempted to criticize us publicly, don't forget who controls when -- and whether -- your products are approved. After these threats, several companies and the drug industry's trade association withdrew from a January 1995 public meeting to discuss reform of the FDA's regulation of biologics.

During my 15 years of observing the FDA -- years spanning the Administrations of four Presidents -- different management styles and regulatory philosophies have governed the agency. But raw political influence has never been permitted to dictate FDA decisions as much as in the Clinton Administration. It sometimes seems that the issue at stake in an FDA decision is which interest group currently deserves a reward. Science and law are now subordinate to politics and public relations at the FDA.