Sam Broder isn't used to the limelight. As director of the clinical oncology program at the National Cancer Institute, he has spent much of his career fighting cancer and, most recently, AIDS.

Lately, however, his work has come under increasing and at times bitter scrutiny, as he fights to halt the nation's most insidious public health threat in its deadly tracks.

As thousands die each year from a disease without a cure, critics say that useful drugs are trapped in years worth of needless red tape before they can hit the shelves.

"AIDS has really captivated people's imaginations, and they are paying attention to this process in a way they never have before," said Broder. "We have a sophisticated system of drug approval, and it takes a long time. That's because time and testing are the only things that can prove drugs work and are safe to use."

But time is something many people just do not have.

Every prescription drug sold in the United States must be approved by the Food and Drug Administration. Of the more than 2,000

drugs under study for commercial use in the United States, fewer

than 20 percent will make it to market.

It routinely takes years from the time a chemical compound is first tested to the day it goes on sale.

American drugs must follow a strict, slow path that requires a progressive series of experiments in test tubes, animals and humans.

Studies must be carried out to verify that drugs do what they are supposed to do, and that their risks are minimal. The larger the study, the better the results.

But large studies take time, and patients who are dying or in pain often focus on positive poten- tial rather than negative side effects.

AIDS activists have led growing charges that the lag time is just too long. California has moved to set up its own shadow FDA to test drugs in clinical trials.

AIDS research is not alone in sparking anger. For example, when an FDA advisory committee refused to approve a new heart drug called TPA earlier this year, many doctors were startled.

When injected immediately after heart attacks, TPA appears to be remarkably effective in dissolving blood clots, which kill hundreds of thousands of Americans every year. Unfortunately, when the committee met, test results suggested that TPA was also responsible for a dangerously high level of strokes among patients who received the drug.

"The public just doesn't know what it takes to approve a drug," said FDA Chairman Frank E. Young.

"They want drugs to be safe and effective," he said. "But you don't approve them after one study. The lack of understanding here is a real issue."

The process of bringing a drug from discovery to approval for public use can take from seven to

10 years, according to the Phar- maceutical Manufacturers Association.

Young has stated repeatedly that he wants to reduce that time whenever possible. AZT, the only drug now approved for the treatment of AIDS, took only 18 months to gain approval, faster than any other drug in FDA history.

"When you get a drug that looks promising, particularly for a deadly disease, there is a delicate weighing process for approval," said Robert J. Temple, director of drug research and review at the FDA.

"If you don't build from a reasonable scientific base," Temple said, "you will never have any idea where you are or if what you are doing is good. That would represent a tremendous loss to society.

"It is extremely difficult to balance those ethical needs," he said. "Take AZT. We know there was a desperation to get it out there. But the reason it is known to be useful is principally because the first

big study was well designed. If not, how would we have been able to judge other drugs? We wouldn't have the linchpin of information we needed."

Even most who accuse the FDA of reacting slowly to the AIDS crisis agree that AZT was approved with remarkable speed. But at least one significant group of gay activists has sued the agency, saying it is dragging its feet on testing new drugs to combat AIDS.

"We feel that the FDA is conducting business as usual when they should be waging war," said Ben Schatz, director of the AIDS Civil Rights Project of the National Gay Rights Advocates.

"There is a lot of ego on the line in the scientific establishment and in the AIDS industry," Schatz said. "The decisions are not being made strictly based on human compassion."

The basic drug-approval process arose 25 years ago, after thalidomide proved to be a harmful drug that ruined thousands of lives in Europe. The drug was never marketed in the United States because it failed extensive FDA tests.

Since then, caution has guided drug approval. FDA officials say they would rather be slow and right than fast and wrong.

Pharmaceutical companies can control the pace of development to some degree, but much depends on the type of drug involved. Most companies follow similar procedures.

First, chemists experiment with molecules, trying to create a compound that has certain specific properties they think might be useful in fighting a disease.

As soon as they develop one, they file a patent application.

These applications, which guarantee a company's right to develop a particular chemical compound, are important in an intensely competitive industry.

The company then screens the compound to learn as much as possible about how it works and what it does.

The FDA requires extensive testing of drugs with animals before any use with humans. Normally, a drug is tested first on rats or mice to determine toxicity levels. If a drug meets standard criteria here, tests on higher animals follow. At this stage, more than half of all drugs must be discarded.

If a drug shows promise without dangerous toxicity, the company files an application with the FDA for human tests. This is the Investigational New Drug (IND) application.

For the IND study, the FDA requires that the drug company file reports on the drug's chemical composition, the results of all animal tests, plans for the proposed human study, and qualifications of the researchers who will perform the study.

Companies must also make certain that they seek permission to test the compound for an exact purpose, and they have to specify the disease for which the drug is intended.

"It's a long, involved process, and if you mess up at any stage, you are back to square one," said Al Alberts, a senior scientist at Merck, Sharpe and Dohme who directed the scientific development of Mevacor, Merck's new cholesterol-reducing drug.

"You can bang your head about the process, but it does work. Maybe it could be just a bit quicker, but unfortunately you cannot take shortcuts when you are making drugs that are available to millions of Americans," Alberts said.

Clinical testing begins once the FDA approves a company's IND application. At this point, animal studies continue, but the focus shifts to human tests.

Drugs can be tested only on informed, consenting volunteers, and the FDA must be told at once of negative effects.

Clinical trials have three phases. In Phase I, doctors give the experimental drug, initially in low doses, to up to 100 healthy volunteers. This stage is designed to determine the drug's safety in humans. Researchers seek to establish only whether the drug has toxic side effects.

This stage is often misunderstood. For instance, DDC, a new AIDS drug that many experts felt might help stop the spread of the AIDS virus, has been shown to be harmful to many subjects in its first dose.

This result has widely been interpreted by many AIDS researchers to mean the drug has no value.

"I cannot understand why anyone would declare that DDC is a washout," said Broder. "We have data that shows it may be effective. That doesn't mean it will be, but it's way too early to write it off.

"We faced this problem with AZT, too," he said. "There is an attempt to draw conclusions about the value of this drug before the study has given us all the facts."

If Phase I trials show promise, then Phase II attempts to determine whether the drug is effective. At this point, scientists look main- ly at the dose of a drug to determine how different groups react to it.

In Phase III studies, researchers try to see just how well the drug works. Here, they normally use a double-blind, placebo-controlled study.

This means that doctors at medical centers across the country give the drug to one group of patients and a placebo, which has no effect, to another group that is equal in number.

Double-blind trials can have from 150 to 3,000 participants. Obviously, the more there are, the better the information generated.

After this process, if the test

results are promising, the company can make its New Drug Application.

Complications and questions often follow drugs to the day they are approved. When their promise is great, delays cause distress to many.

TPA, for instance, might save hundreds of thousands of lives in the United States each year, as heart disease is the nation's most potent killer. But because the drug could be so widely used, it could also harm many people if its side effects are dangerous.

"It doesn't help to clean out your arteries if you blow out your brains," Young said.