The long and difficult history of an experimental anticancer drug known as Indicine N-Oxide begins in 1965 with two brothers, one working for a large and modern pharmaceutical laboratory in the United States, the other traipsing into the fields of rural India to pursue a 2,000-year-old legend of folk medicine.

Venkateswara Rao, the botanist in India, had collected plants that he had read about in historical medical literature and sent them to his brother Koppaka Rao, a chemist at Pfizer Incorporated, the drug company in Maywood, N.J. One day, Venkateswara Rao read about a plant called Heliotropium indicum.

According to the lore of Indian folk medicine, the plant had been placed in tea many centuries ago to fight cancer. The only other thing known about the plant was that it was poisonous -- all over the world it had been known to cause severe liver damage in livestock.

At the time, Pfizer was under contract with the National Cancer Institute to help develop anticancer drugs. Such drugs have been derived from all sorts of toxic chemicals and plants -- from herbicides, pesticides, dyes, the barks of trees -- so it was not all that unusual that one could be derived from this low-growing weedy plant that poisons livestock.

At his brother's urging, Venkateswara Rao traveled to a field in the state of Andhra, Pradesh, India, collected about 10 pounds of the plant, and shipped the package to Pfizer. Then, over the course of a few years, the drug company extracted the substance and purified it so that it could be injected into mice with leukemia.

Initially, Pfizer researchers found that the mice receiving the plant substance lived about 160 percent longer than leukemic mice that were given no drug, Rao said. They also found that the compound showed activity against several other types of cancer in mice. There was a contract problem between the drug company and the NCI that delayed the process for two years. Finally, in 1969, the drug was sent off to the NCI for further testing.

"The drug looked active in the early tests," said Koppaka Rao. "I can't say it looked any better or worse than other drugs. But it did appear to have a good range of safety at the time. It looked less toxic than other drugs."

The story of what happened to Indicine N-Oxide from its harvesting in India to its formulation in New Jersey to its experimental use today is one of scientific hope and promise, pain and death, and bureaucratic arguments between regulators and the regulated. It exemplifies the difficult questions presented by many of the more than 150 experimental drugs tested in cancer patients by the National Cancer Institute in recent years.

The drug has shown some promise in fighting leukemia, but it has also caused severe side effects. Earlier this year, the NCI reported 17 cases of liver damage believed to have been caused by Indicine N-Oxide. According to the Food and Drug Administration, eight of those toxic reactions led to death.

What is known about the drug and what was known before it went to human testing? Should the human testing continue? Are the risks worth the possible benefits? Has the testing been properly monitored by the federal agencies charged with that responsibility? The people involved every step of the way in the evolution of Indicine N-Oxide address those questions.

Keith Jensen, director of cancer research for Pfizer from 1967 to 1980, ran the laboratory at which Indicine N-Oxide was purified and received its initial tests in animals. "I wasn't that excited about it," he said in a recent interview. "I knew this would not reach that level of great promise. It wasn't that exciting."

Jensen noted that it has been 11 years since his company was directly involved with the drug and that the NCI has still not made a determination on its effectiveness. "I'm surprised that the NCI continues to take an interest in it . . . . I hadn't paid much attention," he said. "I had seen abstracts that they were in Phase I the first phase of testing an anticancer drug on patients to determine its toxicity and I thought, 'Heaven's sake! They haven't gotten any further after all this time.' I was surprised that they were still messing around with it."

He added: "It just seems to me the same patients could have been used to test other drugs with more promise or been given drugs already at hand . . . . Their the NCI's job as they see it is to test all seven or eight different types of tumors . . . . I think it took too long. It seems to me they should have made a decision about the thing five years ago."

After Indicine N-Oxide moved from Pfizer to the NCI and showed some activity against cancer in animals, the drug went through another round of animal tests, this time to study its toxicology, or, in other words, to determine its side effects. These tests were conducted for the NCI by the South Shore Analytical and Research Laboratory in Islip, N.Y. Researchers at that lab injected Indicine N-Oxide into dogs, monkeys and mice.

The tests found that the drug caused blood and bone marrow damage and injury to the gastrointestinal system of the animals. They also showed that a single dose of the drug did not produce noticeable liver damage, but that after five days of doses the liver damage became evident.

In January 1975, the research lab sent more than 350 pages of data on the drug to the NCI, where it was reviewed by Anthony Guarino, then the chief toxicologist for the federal agency. Guarino ordered a special test to study liver toxicity because the plant from which the drug was derived was known for that poisonous effect. Those tests, he said, did not predict severe liver damage as a side effect.

Guarino prepared his data for submission to the Food and Drug Administration, which must review and approve the findings before a drug can go to human testing. The FDA received the data in November 1977. As part of the application, NCI submitted chemical analysis of the drug, results of animal testing for efficacy and toxicity and detailed plans for how the drug would be tested in humans -- what the doses would be, how often they were to be given and what the patient would be told before entering the experiment.

" We knew from the beginning that this caused toxicity in animals and the most prominent of these was bone marrow suppression and liver toxicity," said Dr. John Johnson, one of the FDA officials who reviews experimental cancer drugs. "None of this is very unusual for an investigational drug . . . . Almost all investigational cancer drugs are highly toxic."

Guarino eventually left the NCI -- "as a matter of conscience" -- after a dispute with the institute's director, Vincent DeVita. Guarino insisted that more detailed animal studies should be done before drugs are given to humans. DeVita said that fewer animal studies were needed to do the job more efficiently and economically. Guarino claimed that his findings in animal studies were often overlooked when a drug went into the clinic trials. Asked about the purpose of the animal toxicology studies, Guarino said: "I always told DeVita I feel it's a toxicologist's job to define toxicity -- and a clinician's prerogative to ignore it."

By early 1978, 13 years after the Indian botanist picked the first heliotrope plant for Pfizer, Indicine N-Oxide was ready to begin its first human tests. But there was one more consideration.

In January of that year, the FDA called NCI to criticize a proposed informed consent form -- the document a patient signs before entering the experiment. The FDA said that the form, prepared at the Mayo Clinic, failed to spell out the possible liver toxicity predicted by the animal data.

On Jan. 30, Andrew J. Turrisi III, the NCI monitor for testing of the drug, wrote to the Mayo Clinic: "We have reviewed the toxicology data and conclude that the hepatoxicity is probably too minor to be concerned with."

Turrisi told the Mayo Clinic's Charles Moertel in that letter that if he chose not to amend his consent form in line with FDA's comments "we will inform FDA that their intervention is probably overreaching." But he cautioned, that " . . . believe me the FDA does not consider these matters as trivial as they may appear to you."

Moertel chose to change the consent form, rather than "do battle with Rockville headquarters of the FDA ." The FDA then approved the drug.

The FDA's John Johnson monitored the progress of the drug through the course of human experiments which began early in 1978 at four medical centers around the country. His job is to protect patients from unwarranted risk and to ensure that NCI is closely watching the drug for both its usefulness and its possible hazards. But the job is not always clear-cut. Sometimes, as in the case of Indicine N-Oxide, it can take a philosophical or political twist.

In June 1980, Johnson said the NCI reported to him of two possible deaths related to liver toxicity. "That," said Johnson, "was the first time I was aware of any reports of serious liver toxicity."

Then, on May 5, 1981, Johnson said, an NCI official contacted the FDA and reported three possible drug-related deaths from liver toxicity in children. Johnson said the information was "pretty sketchy," so he called the NCI and asked for more information. He talked with NCI's John Penta. " He said there had been two children's studies in progress and that they had been put on hold as a result of these three deaths," recalled Johnson. "He didn't have many details. He said he would send full reports."

Johnson told the rest of the story this way:

"The next thing we heard was not the receipt of the full reports we were waiting for. It was another emergency telephone call from NCI . . . that they had reinstituted the studies without informing the FDA and without consulting the FDA. The first patient they gave it to was a 2-year-old child who died of liver toxicity. The next thing I have from them is dated July 13 in a letter to me . . . that reports to us for the first time 17 cases of liver toxicity." Johnson said he believes that eight deaths were probably related to the drug. "These should have been reported to us sooner and the NCI itself should have taken action on it.

Johnson concluded: "There have been quite a few things of this nature happen . . . . I am telling you this is not an isolated event. This is in my opinion a pattern. It's my impression that NCI never wanted to be regulated by the FDA. They have always resisted this . . . . If we don't know about these 17 cases of liver toxicity, we can't take regulatory action. It's my impression they are reluctant to report these adverse reactions to us because they believe we will take some regulatory action they don't like."

Dr. Daniel Hoth, head of the NCI's experimental drug program, takes issue with Johnson's claims. Their dispute over Indicine N-Oxide reflects a struggle between the two federal agencies that has persisted for years, with outward demonstrations of courtesy but sharp disagreements in private.

Hoth said he is proud of the way the NCI promptly responded to the question of liver toxicity in Indicine N-Oxide. The NCI, not the FDA, voluntarily halted testing of the drug when it became evident it could cause liver toxicity, he pointed out. He said he perhaps should have told FDA that the hold on testing had been lifted, but since it had been their decision to shut it down, they had the right to resume it when they felt the studies could be modified to avoid some of the risks to patients' livers.

"It would obviously have been ideal to inform them at the same time," said Hoth. "It does not represent a casual attitude on our part. It represented a crushing work load. In the ebb and flow of daily business, that's one detail. Our position is that we took the lead voluntarily in shutting down the drug. We took the lead."

He said there was no effort to conceal the 17 cases of liver toxicity. "I can list a half-dozen documents that indicated the hepatoxicity which we sent to them over the course of the year," he said. "The point is, the information was there if they wanted it." He said Johnson's impression that the NCI is trying to avoid FDA regulation is groundless.

Sometimes the dispute with FDA takes a more philosophical turn. There are many doctors who would argue that protecting the dying cancer patient from the risk -- no matter how serious -- of an experimental drug is simply stripping the patient of the right to choose how he will make his stand against the disease. Thus the difference of opinion between the Mayo Clinic's Dr. Charles Moertel and the FDA.

Although the first Mayo Clinic tests of Indicine N-Oxide showed little promise in fighting colon and rectal cancers, later tests showed good results in leukemia. Of 22 patients tested, five responded, three of those with complete remissions.

"Lo and behold, we got a complete remission in the disease," said Moertel, who is known as a cautious researcher. He was excited about pursuing the drug tests, but then came the federal decision to halt it temporarily.

"This is terribly frustrating to me," he said. "We have youngsters who are dying from cancer. There has been so much panic about this drug. Now they are all dying because of some bureaucratic hangup. They say 'You might kill somebody with this drug.' Hell, if you don't treat them, they're all going to be dead."

Dr. James Miser of Children's Hospital in Columbus, Ohio, has also had promising results with early tests of Indicine N-Oxide. In a group of 24 leukemia patients, he has recorded three complete responses and two partial responses. One of his success stories appears to be Joshua Waples of Galloway, Ohio. Joshua is just under 4 years old. His parents were informed last June that they boy appeared to have only two months to live because of a tumor between his ventricles, the two lower chambers of the heart.

The child then began taking Indicine N-Oxide. He has shown improvement since starting the drug, and he has not suffered side effects. Said his mother, Edeltraud Waples: "Right now, we feel it's doing some good."

But then, there is the case of Angela Frash. Last March 16 was Angela's seventh birthday. Her mother, Elizabeth, had baked her a birthday cake, but Angela didn't feel like eating it. Instead, she just lay on the sofa most of the day. At six that night, she took a dramatic turn for the worse. She started throwing up blood. Her mother drove her to the hospital in Columbus from their rural home.

She became incoherent on the way. "If you even touched her, she would holler in pain," said her mother. Two days later, on March 18, Angela died, ending a three-year siege with leukemia. At the end, remembered her mother, the 7-year-old was bald and looked pregnant, her belly was so distended from a swollen liver.

Elizabeth Frash remembers deciding whether to let her daughter take the experimental drug, Indicine N-Oxide. Angela was with her as she and her husband Eugene had a conference with Dr. Miser. "We knew it would either bring her back or take her from us completely," recalled Elizabeth Frash. "It was more or less our last hope. It was either that or another drug that wasn't as well researched as that one."

She said she never asked what her daughter died of. "We never asked, but I think we more or less knew," she said. "She'd been struggling for two to three years. We thought it was the cancer, because the pressure seemed to be more than she could bear . . . . "

Could it have been the drug?

"It could have been . . . . It's really too late to speculate on that, and we never had an autopsy . . . . It said on the death certificate leukemia, but she just wore out . . . heart failure. She just couldn't take any more."

In a review of the experiment filed with the NCI, James Miser concluded, "In this pilot patient, there is clear evidence of severe drug toxicity that resulted in her death." Miser said he would have filled out the death certificate differently. "I would have put down liver failure presumably due to Indicine N-Oxide."

Hoth of the NCI has hopes for the future of Indicine N-Oxide, but he is very aware of the difficulties of predicting a drug's possible usefulness. "The dilemma we face today," he said, "is that treatments are so much better than they were 20 years ago that by the time a patient gets to a Phase I study, he has already tried proven drugs. How likely is it that an investigational drug will be active?" He said that since patients on experimental drugs have already failed to respond to proven drugs, any response at all to an experimental one could be an important hint that the drug is worthwhile.

In the case of Indicine N-Oxide, Hoth points to studies of 46 patients with leukemia in which six have shown complete responses -- that is, all signs of their disease were gone for at least one month. In contrast, many experimental drugs show only one complete response in a group of 100 or more patients. "Based on these findings of CRs complete responses ," said Hoth, "I'd have to say that the drug may be promising."

But because of the reported cases of liver toxicity, Hoth said, future tests with the drug will be conducted "cautiously, with careful monitoring" in hopes that a safe but effective dose level can be established. He concluded: "We feel that it will be possible to use this drug safely."

Dr. Andrew Turrisi, who monitored Indicine N-Oxide for the NCI, left the institute in June 1980, just as the first reports of liver toxicity were coming in. He said he has not kept up with the drug since then. When told in a recent interview of the drug-related deaths, he said: "Amazing . . . there was nothing in there to set off any alarms . . . . It's interesting how your words come back to haunt you. It happens in that job because you leave a paper trail."

Turrisi said he was never really impressed with Indicine N-Oxide. "It just didn't look like a winner to me," he said. "I remember it was not a very promising looking drug from its activity in animals , but you still want to try it to see. Clearly, a lot of better minds than mine have looked at it, so we went with it . . . . At times I wonder. There's no one who had the whole picture, the type of view to know where we were going. We got caught up in the system we understood. There wasn't good enough reason not to go on with it."

After leaving NCI, Turrisi went into radiotherapy. He expresses some disenchantment with drug development: "I was disappointed by the effect those drugs had on patients . . . . I knew that pushing drugs on these people with precious little rationale was hard for me to do . . . . I think something should be done and there are times in a cancer patient's existence where they experimental drugs should be considered, but where a lot of b.s. comes is where you are telling a patient you are going to help him . . . . They the drugs rarely if ever make anyone live even one minute longer and even if they do, they don't add any quality to the patient's life."

Dr. Miser in Columbus has hope for Indicine N-Oxide. And he says the participation of families with cancer patients has a value beyond research. "The family will have the feeling they gave it everything they had," said Miser. "They can feel like they didn't leave any stone unturned."

Cancer patients must be given hope, Miser said, but the hope must be reasonable. That is one of the values of the experimental drug program, he said. "If we run out of reasonably based scientific things, they'll the patients turn to to crazy things -- like coffee enemas. I've had patients do that. We can't let them fall into the hands of charlatans.