On the Fourth of July in Peoria, Ill., an 8-year-old cancer patient named Sheri Beck was rushed to Methodist Hospital. She died there the next day.

Cancer did not kill Sheri Beck. Her treatment for cancer did. She died of congestive heart failure brought on by Mitoxantrone, an experimental drug derived from a dye used in ballpoint pen ink. She was one of hundreds of cancer patients who have been given the experimental drug and one of a growing number who suffered heart failure as a result.

Those given Mitoxantrone are not alone. A one-year study by The Washington Post has documented 620 cases in which experimental drugs have been implicated in the deaths of cancer patients. Each of the drug-related deaths was verified by doctors and recorded in government accounts of the experiments. And they amounted to merely a fraction of the thousands of people who in recent years have died or suffered terribly from cancer experiments conducted in the nation's hospitals.

Over the last decade, more than 150 experimental drugs have been given to tens of thousands of cancer patients under the sponsorship of the federal government's National Cancer Institute. Many of these drugs have been derived from a list of highly toxic industrial chemicals including pesticides, herbicides and dyes. They have all been tested by the NCI and hospitals around the country in the hope that they would prove to be lifesavers against a disease that has ravaged America and frustrated its medical and scientific community.

The official rationale for this experimentation, despite the fatalities, is that these patients were already terminally ill and that scientific progress required the risks of subjecting functioning human beings to highly toxic drugs. In this desperate fight, some hospitals have turned into laboratories and their patients have become subjects. Vincent Bono, former chief of the NCI's experimental drug branch, likened many of the studies to "donating someone's body to science while they are still alive."

"Sometimes there is little regard for people's lives," added Robert Young, of the Food and Drug Administration's drug review section. "In Boston, a hospital tested a new NCI drug . . . on children. Their kidneys were lost within days. This was no big deal, because NCI new drugs are routinely given out with literally no safeguards for people who will receive them."

"First do no harm." That is the oath that every physician swears to honor. Has that oath been upheld in the use of experimental drugs? Over the next four days, The Washington Post will report that:

* While all anticancer drugs can cause side effects among some of those who take them, the experimental drugs -- in addition to leading to hundreds of deaths -- have elicited a nightmarish list of serious adverse reactions, including kidney failure, liver failure, heart failure, respiratory distress, destruction of bone marrow so the body can no longer make blood, brain damage, paralysis, seizure, coma, and visual hallucinations.

* So little is known about many of these chemicals that doctors have found these ironic results: In some cases the experimental drug actually stimulated tumor growth rather than stopped the cancer; and in other tests, doctors and researchers found that the experimental drugs themselves caused cancer.

* In many cases, the experimental drugs have been given to patients even years after studies failed to show that they were of use in the fight against cancer. The first phase of the experimental drug program, in fact, is primarily designed not to combat cancer, but to find out how toxic the drugs are.

* Physicians at some of the nation's leading hospitals and cancer centers have inadvertently administered high or miscalculated doses of these potent toxins with fatal results.

* Records of some drug experiments have been so poorly kept that dates of patients' deaths are off by months, lab reports are lost or never done, and responses to drugs are misrepresented. In addition, some hospitals and the NCI have in written reports exaggerated the therapeutic performance of the drugs.

* Drugs and treatments not approved by the Food and Drug Administration for human use -- and not authorized by the National Cancer Institute -- have been administered to dozens of patients.

The litany of death and suffering from experimental drugs has become an accepted part of life at some hospitals around the nation. These human experiments have gone largely unchallenged and unquestioned by Congress, the medical profession and the scientific community at large. How and why did it reach this point? To understand, it is necessary to enter the world of the cancer patient and the cancer doctor.

For four of every 10 cancer patients, a cure is possible through surgery, radiation therapy or drugs. But for the rest -- for the majority of cancer patients -- death is imminent and inevitable, agony is common.

In this world, doctors trained and dedicated to healing are rendered helpless. They are profoundly frustrated and burdened by the knowledge -- painfully gained by experience -- that their patients will die of cancer. And the patients are in most cases frantically grasping for something and willing to try almost anything in the hope of defeating the cancer or, at least, delaying their deaths.

"The worst thing that can happen is they tell you, 'There's nothing that can be done,' " said Mel Wilkins, whose son died last July in Baltimore after taking an experimental cancer drug. "You're grasping for possibilities . . . You're asking for a miracle in a religious sense and in a technological sense . . . . If it's a one in a million shot, then you've got to go for it. There's little other choice."

Within such a bleak and agonizing environment, most cancer doctors work with what Daniel Hoth, head of the National Cancer Institute's drug testing program, calls the "risk/benefit ratio." They argue that the risks of these experimental drugs -- the serious side effects and even death -- are outweighed by the possible benefits. Dr. Emil Freireich, who works at one of the nation's leading cancer hospitals, M.D. Anderson in Houston, put it this way: "In circumstances in which people have what is clearly a hopeless prognosis, the willingness of physician and patient to accept a much higher level of risk . . . is of course obvious."

But what have been the long-term benefits of the experiments? Vincent DeVita, director of the National Cancer Institute, acknowledges that there are no statistics on the cure rates with drugs. He estimates that as many as 46,000 people each year -- or about 6 percent of all cancer patients -- can be cured by chemotherapy. To DeVita, that statistic makes the risks worthwhile. "Say there have been 1,000 drug-related deaths ," he said. "We're curing some 40,000 or 36,000 per year with drugs . . . We couldn't have gotten to that point without going through this."

The cure rates of which DeVita speaks are for the most part from a half-dozen commonly used drugs that are on the market and have been shown to be effective in treating several forms of cancer, among them some childhood leukemias, Hodgkin's disease and testicular cancer. Cure rates for these diseases -- all relatively rare -- have gone up dramatically since the advent of drugs. But in the vast majority of drug experiments, it is not uncommon for none or only one or two out of hundreds of patients to benefit from the drug.

Over the years, thousands of cancer patients have participated in the initial testing phase of the experimental drug program, known as Phase I studies. The primary purpose of these Phase I studies is to observe the side effects of the drugs and determine proper dosage, not to cure the patients. "I would say there's no potential for a cure on a Phase I study," said Stephen Carter, head of the NCI experimental program from 1970 to 1975.

Doctors know of the history and dangers of the experimental drugs and the remote chance of any benefit. Many times, patients do not. The patients are required to sign "informed consent" forms indicating that they have been informed about the risks of the experiment. Many doctors say the patients often don't read the forms thoroughly and that the patients usually are not told all the details. "I tell a patient we have a new drug that shows some promise in animals," says Gordon Zubrod, a pioneer in NCI's drug development program. He added: "I don't think I'd ever say to the patients the primary purpose is to determine toxicity."

Most doctors say they tell their patients that an experimental drug shows "promise," but they readily acknowledge that their patients -- who have to be expected to live for at least another two or three months to qualify for the program -- often put more faith in that "promise" than is warranted.

While Zubrod, like most cancer doctors, supports the experimental drug program, he acknowledges that progress has come only in inches. "We've been on a plateau for about 10 years," he said. "There've been virtually no breakthroughs. They've been inching ahead. It's almost like building a pyramid."

Charles Moertel, director of the cancer center at the Mayo Clinic, used a different analogy to make the same point. "We have hit on some areas more or less by luck. We have accomplished a lot, but it's been like swatting at flies with a sledgehammer," said Moertel. "It's been rather blind . . . . I suppose as far as end products are concerned it's been a bust, but then so is the overall treatment of cancer."

The FDA's Young takes a harsh view. "There are so many analogies between the War on Cancer and the War in Vietnam, it's scary," he said. "You've got the generals, the NCI. And you have this attitude among the generals: 'We've got to burn the village to save it.' The generals keep saying, 'We've invested so much; we can see the light at the end of the tunnel.' But you've got to keep asking yourself -- what's the gain? I mean, what about the patient?"

No one questions the need to test drugs in human beings. What is questioned is at what point a drug is promising enough to warrant human testing, and at what point an experimental drug should be considered ineffective or too toxic and withdrawn from human experiments.

Said scientist Linus Pauling, a two-time Nobel laureate: "Why make the patient miserable for his last few days? I don't think the terminal cancer patients should be used as guinea pigs. You have to carry out human experiments to make progress, but you must be careful that you don't sacrifice human beings."

The final judgment on the drug experiments of the last decade cannot yet be rendered, said NCI's DeVita. "If chemotherapy improves. . . if it becomes an easy treatment, they will look on it as one of the greatest eras in the history of medicine . . . .If we go no further than we are now, then we will look at it as a blind alley -- that we went up a blind alley."

What follows today and tomorrow are case studies of 20 experimental drugs tested in the last decade. They were not chosen as "good" cases or "bad" cases, but rather as ones that were representative of the more than 150 drug studies undertaken since the War on Cancer was declared in 1971. Eight of these drugs were listed by DeVita in 1979 as "high-priority drugs" -- the most promising experimental drugs in the country.

What doctors look for in the anticancer drug experiments are "complete" or "partial" responses from the patients. A "complete response" means that all signs of the disease have disappeared for a specified period of time, usually in excess of a month. A "partial response" means that at least 50 percent of the cancerous cells have disappeared for the same period of time.

A patient's life may or may not be extended by a response. In most cases, the response lasts for a matter of weeks or months, then the disease reoccurs and the patient dies. What the response tells a doctor is that the experimental drug may have activity in fighting cancer.

Risk/Benefit Ratio

As in an eerie nightmare, emergency phone calls and letters keep coming in to the experimental drug unit of the National Cancer Institute.

Congestive heart failure.

Unexpected cardiac arrests.

Children dying.

The cases have come from hospitals in St. Louis, Memphis, Houston, Baltimore and other cities. The doctors are reporting patient reactions to a new experimental drug called Mitoxantrone, which had been derived from a dye used in ballpoint pen ink.

The drug was developed specifically to avoid cardiac complications. So far this year, 22 people taking the experimental drug have had heart failure or other serious cardiac problems.

More than 1,000 people across the country are receiving experimental doses of Mitoxantrone, according to the NCI's drug branch. The toxicities to the heart have usually been delayed -- occurring weeks or months after injection of the experimental drug, although there also have been cases of sudden heart failure within hours of the drug infusion.

But NCI officials are not stopping the Mitoxantrone tests. They are accustomed to heart problems associated with some anticancer drugs. One of the most commonly used anticancer drugs on the market today, Adriamycin, has been implicated in scores of documented cardiac deaths.

"Even though a drug has a toxicity like that doesn't mean you have to throw it out," explained Hoth, head of the NCI's drug testing program. "There's what we call a risk/benefit ratio." But in regard to the benefit side of the ratio, the final results are not yet in. Of the first 568 people tested in a study in hospitals in the Southwest, there has been one "complete response" and five "partial responses."

Throughout the summer and fall, Hoth and others at the NCI were constantly reminded of the risks:

June 10: William Harp, 65, had lung cancer. On May 26, doctors at the Veterans Administration Hospital in Lexington, Ky., gave Harp his first dose of Mitoxantrone. A week later, on June 3, he went to the hospital for his second administration of the experimental drug.

Say the hospital records: "Shortly after receiving his second infusion of Mitoxantrone within 1 or 2 hours , the patient developed an episode of acute weakness, dyspnea difficulty breathing , diaphoresis sweating , and midsubsternal chest pain." He was taken to the coronary care unit. In the next five days his respiratory difficulties increased and he developed severe congestive heart failure. He died on June 10.

Dr. Philip DeSimone informed the NCI.

July 5: On Saturday, July 4, Virgil and Vivian Beck ran from their house in Henry, Ill., jumped in their car and "drove 80 miles an hour" to Methodist Hospital in Peoria.

Mrs. Beck recalls that she held her daughter, Sheri, in her arms "every second, until we were in the hospital." The 8-year-old girl, who had rhabdomyosarcoma (a skeletal muscle cancer), had completed treatments with Mitoxantrone several weeks before. She now had congestive heart failure, which doctors found was drug-induced.

She died on July 5.

July 6: Christopher Wilkins, 15, had osteogenic sarcoma, a form of bone cancer. On April 22, doctors at the Johns Hopkins Oncology Center began giving him Mitoxantrone. On July 4, he felt extremely ill, constantly vomiting. Two days later, the hospital admitted him with "severe intractable congestive heart failure." He was declared dead at 3:10 p.m., July 6.

Aug. 25: Theodore Rechtin, 64, began receiving Mitoxantrone at Washington University in St. Louis on June 2. On Aug. 25, according to hospital records, Rechtin was home alone; his wife was at the theater. When she returned that evening, Rechtin "was found slumped at the kitchen table. Paramedics were called and apparently he was in ventricular fibrillation (irregular contractions of the heart) and was unable to be resuscitated." He was pronounced dead at Lutheran Medical Center that night, with the cause of death listed as "probable cardiac respiratory arrest."

Those four cases and 18 others are now listed in a file at the NCI under the title: "Mitoxantrone Cardiotoxicity."

Doctors are not certain whether the heart failures were caused only by Mitoxantrone or whether that drug, combined with earlier drug treatments, caused the deaths.

In most of the cases, such as with 15-year-old Wilkins, the patients had earlier received Adriamycin. Many had experienced heart difficulties with the earlier treatment. However, in some cases, such as with 65-year-old Harp, the patient had not received any anticancer drugs before.

Most doctors agree that drug treatments caused the heart failures. Autopsies in some cases show that cells in the heart muscle were destroyed by the drugs. Doctors are uncertain, at this point, though, about exactly which combinations of drugs may be the culprits.

But the heart complications did not come as a surprise to everyone.

In fact, on Nov. 26, 1979, the director of the FDA's cancer drug division wrote a letter to an administrator of the NCI's experimental drug program warning of possible heart reactions to Mitoxantrone. He also warned about the increased danger of heart problems for patients who had been treated earlier with Adriamycin.

Dr. William Gyarfas of the FDA wrote to Dr. John Penta of the NCI:

"In view of the possible cardiotoxicity of this drug Mitoxantrone , we recommend that EKG heart monitoring tests be performed more frequently . . . and that consideration should be given to excluding patients with prior Adriamycin treatment and patients with heart disease."

Despite the warning, the NCI decided to test Mitoxantrone in all patients -- including those with past Adriamycin treatments and heart problems. In addition, the NCI circulated reports to doctors in June 1980 and in June 1981 that said specifically that there were no heart toxicities with Mitoxantrone.

In June 1980, a document circulated by the NCI reported to doctors across the country: "No evidence of cardiotoxicity was noted" in early studies.

In June 1981, a document circulated by the NCI to doctors reported: "No evidence of cardiotoxicity was observed for any patient" on Mitoxantrone.

In the same month as the latter publication, the first reports of heart failure in patients on Mitoxantrone began to come in to the NCI.

Throughout the summer, NCI officials sent letters to doctors across the country warning of the heart toxicities with Mitoxantrone and limiting use of the drug with patients who previously received Adriamycin. Still the drug is currently among the most widely used drugs in human experiments.

Mitoxantrone was derived from a dye used in ballpoint pen ink, according to Kenneth Paull of the NCI. Scientists modified the chemical structure of the compound to make it into a drug. In studies, doctors found patients' veins turning blue, their urine and serum (which is usually clear) turning green. Before receiving the experimental drug, doctors were given this advance notice by the NCI: "Warning -- Personnel handling and administering this product should take all precautions necessary to ensure that it will not come in contact with human tissue except during intravenous infusion."

Kidney Damage

Early tests in dogs and monkeys showed that an experimental drug called MeCCNU produced severe kidney damage in the animals. Those findings were published by the National Cancer Institute in 1974 and distributed to doctors before human experimentation with the drug began.

Over the next four years, children in clinics at New York University and Children's Hospital in Boston were given the experimental drug. At least 20 of those children suffered kidney damage -- in some cases irreversible and fatal.

In addition, doctors recently have been reporting another serious side effect of MeCCNU: In some cases the drug causes acute leukemia, according to NCI records. One of the children in Boston who developed kidney damage also contracted leukemia from the drug. He died of leukemia. Federal officials are now gathering data on the reported cases of leukemia from MeCCNU from across the country.

Thousands of people have received and still are receiving MeCCNU at numerous hospitals. Despite its effects on the kidney and other severe reactions that have been implicated in at least 24 documented deaths, some doctors use MeCCNU because they feel it may be more active against cancer than many other experimental drugs. In fact, MeCCNU is one of a group of drugs -- called nitrosoureas -- which NCI officials say are among the most important drugs to have come out of their experimental program.

Still, while MeCCNU is considered to be an "active" drug, its exact usefulness has not yet been determined. "In preclinical studies, MeCCNU was significantly superior to other analogues" in animal studies, a doctor wrote in an NCI report. "It has not been demonstrated to be more effective than the other nitrosoureas in the treatment of human tumors, however."

No General Rule

Experimental drugs generally are brought into use because they show some sign of anticancer activity in mice. There is no general rule to determine when human experimentation with a drug should be stopped.

In the 1950s, doctors began using an experimental drug from a substance called piperazine. Various derivatives of piperazine were tested over the next 25 years.

In 1975, the National Cancer Institute began full-scale experimentation with Piperazinedione, a piperazine derivative. From the start, there were problems. Early tests in the Southwest showed severe poisoning of the bone marrow -- causing dramatic drops in the blood counts of patients. Increasing the problem was the fact that the blood toxicity was "fairly unpredictable," according to doctors.

In 1978, NCI reported to the FDA that three patients taking the drug died from destruction of the bone marrow, and that "life threatening" blood toxicities and bone marrow toxicity were "relatively common" in one group of 17 patients.

The NCI concluded: "Piperazinedione has been well evaluated in some tumors; however, it has not been established as a very effective agent drug , and its toxicity can be formidable." In fact, the statistics to that point showed that 325 people across the country had taken the drug, with one complete response reported. There were 15 partial responses.

Despite these findings -- and after more than two decades of sporadic testing of piperazine derivatives -- the NCI decided to continue testing the drug.

A year later, the NCI reported to the FDA again, this time indicating that 87 more people were given the experimental drug; there was one "response." In the report, the NCI documented three more drug-related deaths.

The NCI concluded this time: "Since our last report, no new good evidence for effectiveness has emerged in the Phase II studies we have supervised, and toxicity has been prominent."

Still, the federal agency decided again to continue testing the drug in humans. "We still have to see a study in leukemia, a disease in which the most promise for this agent seems to rest," NCI wrote.

A year later, in October 1980, NCI wrote another report to the FDA. The results: 104 more people had been given the drug, with no complete responses. There were three partial responses. Of the 25 leukemia patients in the experiment, there was no therapeutic response to the drug at all.

The drug was implicated in the deaths of six more patients.

By late 1980, 516 people had been given Piperazinedione, and one complete response had been recorded. The NCI, in its October 1980 report to the FDA, said it planned no further studies as an anticancer agent. However, they said that because of the effect the drug had on the bone marrow, it might be given to patients being prepared for bone marrow transplants.

The NCI concluded: "The IND (permit) for Piperazinedione remains open to continue accruing patients for autologous bone marrow transplantation, where the drug seems to have its potential usefulness. There are no other studies planned."

Effectiveness

In 1951, researchers investigated possible use of a chemical that is now the experimental drug Hexamethylmelamine (HMM).

Some doctors consider it a valuable drug in treating ovarian cancer -- in many cases in combination with other drugs. It has been tested in well over 1,000 people.

The drug has numerous toxicities to the central nervous system, including, according to the NCI, "confusion, agitation, depression. Some patients have hallucinations and petit mal-like spells."

Also cited as drug reactions were insomnia, somnolence, vertigo and "tremors", and depression, "accompanied by suicidal ideation. One patient became frankly psychotic 3-4 weeks after one course of HMM," according to an NCI report published in November 1974.

In a meeting with the FDA in April 1977, Vincent DeVita, then director of cancer treatment for the NCI, said of HMM: "You might wonder why, with the marginal effectiveness of this particular compound, we're interested in this . . . . Actually our scientific program is going in the development of analogue a related drug . We are very excited about what may be the first metabolite of this compound, Pentamethylmelamine . . . so we have Pentamethylmelamine under a rush, trying to get it through the system in a hurry."

'Agitation'

Using the Hexamethylmelamine chemical structure, researchers designed a second drug, similar in structure, but more water soluble so that it could be given to patients intravenously. The drug derivative was called Pentamethylmelamine.

Patients at M.D. Anderson Hospital in Houston began receiving the experimental drug in 1978. Records show that some patients immediately began to suffer from nervous system disorders such as imbalance, weakness, somnolence, skin irritations such as itching and burning, hearing loss, slurred speech, personality change, amnesia, confusion and coma.

Memorial Sloane-Kettering Cancer Center in New York, the Mayo Clinic in Minnesota and at the National Cancer Institute in Bethesda also began studies of the drug. Patients receiving higher doses of it experienced severe nausea and vomiting. Some continued vomiting for as long as 36 hours.

The Mayo Clinic reported that one patient with severe nausea and vomiting and severe nervous system toxicity required hospitalization and was in a state of confusion and "agitation for 24 hours," according to an April 1981 NCI report.

Another Mayo Clinic patient who was hospitalized because of the drug "became unconscious after 1/2 of dose, drug stopped, awoke 1 1/2 hours later and was confused and disoriented for 24 hours." He also suffered from severe nausea and vomiting and a drop in blood pressure.

In Baltimore, doctors reported that one patient on Pentamethylmelamine "had PMM (drug) induced visual hallucinations."

In a report to the FDA last April, the National Cancer Institute wrote that "no significant antitumor effect was seen" with Pentamethylmelamine.

Four hospitals -- Mayo Clinic, M.D. Anderson, Memorial Sloane-Kettering and the University of Vermont -- had reported that the toxicities from the drug were so severe that they felt NCI should stop its studies with Pentamethylmelamine. Those same institutions also said that Pentamethylmelamine showed no advantage over Hexamethylmelamine.

Despite these recommendations, the National Cancer Institute reported in April that it had decided to continue its human experimentation with the drug, which was one of the "high priority" drugs listed by DeVita in 1979. However, no hospitals in America showed interest in using the drug. "The only Phase II trial so far is under the International Agreement with Egypt," the National Cancer Institute reported to the FDA. "The trial is in progress, but results are unavailable."

Tumor Stimulation

It is common for cancer patients to receive experimental drugs that do not kill the cancer. In the case of a drug called Estradiol Mustard, the drug did something for at least one person: It apparently stimulated the growth of his tumor.

Estradiol Mustard was a short-lived experimental drug, but for 69 people, the drug was held out as their last hope.

Thirty women with advanced breast cancer were given the drug. For 27, there was no benefit. The three others showed "objective" responses. The drug caused nausea and severe vomiting for some women, a drop in blood counts, diarrhea and "profound weakness." For one woman, the drug led to vaginal bleeding. Another became quadraplegic, according to a March 1977 NCI report.

In another study, Estradiol Mustard caused the same stomach and blood problems. In terms of treatment, the recorded results were:

Of 39 people, 37 received no benefit, one showed a partial response, and one "was believed to have had tumor stimulation due to the drug."

There are no studies using the drug today.

The Next Step

Cancer doctors across the country agree that one of the most successful experimental drugs introduced in the last five years is AMSA, which has shown considerable activity in fighting leukemia in adults. About 21 percent of the leukemia patients receiving AMSA have experienced complete responses, according to the NCI.

The success of this drug has caused excitement. But the AMSA story also shows vividly how tenacious a disease cancer is and how a genuine sense of excitement in drug experimentation may still be a long way from a cure.

While AMSA has displayed an impressive "response rate" in patients with acute myelocytic leukemia, most of those patients died from the disease several weeks or months later, according to a June 1981 NCI report. Some died from the effects of the drug while still showing a response to it. Most leukemias -- especially in adults -- are still incurable.

As well as AMSA works, to date it does no better than the drugs that were in use against leukemia before it was developed. "The next step is to show that it AMSA adds something" to existing treatments, explained Daniel Hoth, head of the NCI experimental program.

One of the real advantages first seen with AMSA was that it did not appear to cause heart failure. In an enthusiastic press release dated May 27, 1980, Memorial Sloane-Kettering Cancer Center wrote: "AMSA showed itself to be just as effective as anthracyclines drugs already in use , but it does not cause heart damage."

At about the time that press release was written, several doctors across the country began reporting sudden cardiac problems and deaths associated with AMSA. To date, 15 such cases have been reported.

Doctors at M.D. Anderson Hospital in Houston reported three cardiac toxicities that occurred while the patients were being infused with the drug, according to NCI records. One 40-year-old man "became dizzy and felt as though he was going to black out" while he was being given the drug, according to a hospital report. Minutes later the man went into seizures, his heart beating rapidly, and he had a cardian arrest. He died of body-wide infections a week later.

In addition, the drug can be extremely harmful to the blood and bone marrow. In one study in several northern California hospitals, the NCI reported that every patient in the experiment required blood transfusions because of the effects of the drug.

In all, AMSA has been implicated in the deaths of 18 cancer patients. But because many more than that have shown responses to the drug, the NCI is expanding AMSA's use, planning to test it in combination with other drugs to see if they can more effectively treat leukemia. NCI officials hold AMSA up as an example of the "risk/benefit ratio" -- the reason doctors are willing to accept severe side effects in human experiments.

Footnote: AMSA is so toxic that it has been reported to cause side effects in hospital workers who were in the room when a vial of the drug was opened. In the summer of 1979, six of eight staff members at the University of California experienced nausea, headaches, abdominal pains, nasal and eye irritations and some difficulty in breathing within minutes of opening an ampule of the drug, according to an NCI report. These symptoms generally ended within a few hours.

Medical Ethics

Although some doctors have seen their patients become seriously ill or debilitated by experimental drugs, they rarely write about the human suffering or raise questions of medical ethics in their reports to the National Cancer Institute.

However, Dr. Rupert Nitschke at the University of Oklahoma chose to raise some of these issues in a report earlier this year on a drug called F3TDR.

The drug -- in combination with three others -- was being tested in children with neuroblastoma, a usually incurable form of nerve cell cancer. Earlier reports found that the drug showed activity against the disease, but that it was severely toxic to the children.

"The investigators' conclusions were unfavorable," said a 1980 report to the National Cancer Institute, "because of the heavy toxicity of the regimen. The children spent 40 to 46 percent of their time on-study in the hospital because of complications related to drug therapy or for the administration of chemotherapy."

In his 1981 report, Dr. Nitschke noted that children on these experiments in Oklahoma, Texas, Arkansas and Alabama were suffering severe blood reactions, with four children dying from infection believed to be related to the drug. Nitschke seemed to be torn between the suffering caused by the drugs and his own frustration at not being able to do anything else for the children. In a wrenching statement, which seems to reach conflicting conclusions, he wrote:

"Metastatic neuroblastoma is a deadly disease. The treatment must be improved. For that reason, an intensive chemotherapy program like the one reported is ethically permissible. Unfortunately, the results of this study were not as good as expected. In view of the therapeutic results, the observed severe complications caused by the therapy, the financial burden due to frequent hospitalization, and the undue interference in the life of the child and his family, the 4-drug combination cannot be recommended in newly diagnosed patients. In spite of severe side effects, it is, however, a promising induction therapy in children who have disease recurrence."

After receiving that report from Nitschke, the NCI concluded in a report last March "that there is no further interest in the drug because of its toxicity. Because of this and the drug's lack of activity in other tumor systems, consideration is being given to closing" experiments with the drug.

'It Doesn't Work'

On Aug. 3, 1976, the first of 47 patients at Mount Sinai School of Medicine in New York was entered into a Phase I study of an experimental drug called Neocarzinostatin.

Within months, doctors from Mount Sinai reported to the National Cancer Institute that three of the patients on the study "developed immediate reactions to to the drug characterized by shaking chills, fever up to 40C, rigor, respiratory distress, choking sensation, hypertension high blood pressure of 210/100, tachycardia rapid heart beat of 180/minute and, in two, mental confusion."

They also reported that three other patients died because their bone marrow was so severely destroyed, apparently by the drug, that they could no longer make blood.

Studies with Neocarzinostatin also were conducted at the Baltimore Cancer Research Center, St. Jude Children's Research Hospital in Memphis, Children's Hospital in Los Angeles and Sidney Farber Cancer Center in Boston.

By late 1978, doctors from those hospitals reported a long list of side effects to the National Cancer Institute and reported one complete response and 19 partial responses from Neocarzinostatin among 260 patients. The NCI decided to test the drug in several hundred more people in 1979 and 1980.

In February 1980, doctors at NCI wrote in a report to the FDA that Neocarzinostatin showed "encouraging results" in treating some forms of leukemia. They concluded: "In the past year, more evidence has been accumulated which confirms Zinostatin's activity in leukemia."

The evidence they referred to was spelled out in data charts in the report. Of 287 patients given the drug, there were two complete responses and 16 partial responses. The two complete responses were in leukemia.

Studies were continued into 1981. In a report last April to the FDA, the National Cancer Institute wrote that it still felt the experimental drug might have some therapeutic value in treating leukemia. It also reported three "life threatening" cases of respiratory arrest which occurred during or just after infusion of the drug and one "drug-related death in a patient with cardiac arrest ten minutes after the sixth dose of the drug."

In an interview last month, Hoth, head of the NCI's clinical drug experimentation program, gave this status report on Neocarzinostatin: "It's dead. It doesn't work much. It has some nasty toxicities, too. There may be some studies trickling along, but I think the drug is heading for the box."

According to National Cancer Institute records, experimentation continues with Neocarzinostatin, which was one of the "high priority" drugs listed by DeVita in 1979. Dr. J.J. Lokich of Sidney Farber, one of several doctors who concluded the drug would not be of value in treating cancer, said:

"There's a lot of brouhaha when a drug comes out. But every drug follows the same pattern. They come out with a lot of excitement, a lot of laboratory rationale from animal tests. But it's given to 20 people and there's nothing. That's the same scenario for all of those 150 in the last 10 years -- except for four or five."