Altman, Lawrence K. "Fatal Drug Trial Raises Questions About ‘Informed Consent,'" New
York Times, 5 October 1993, p. C3.
The deaths of 5 of 15 participants in a drug trial at the National Institutes of Health over the last
few months have focused attention on the process by which patients come to participate in
studies testing the safety and efficiency of new therapies.
What seemed, after an uneventful pilot study, to be a safe and promising drug for hepatitis B
turned out to have fatal complications. In addition to those who died, two participants survived
only after having liver transplants. The remaining eight are unaffected so far. The deaths made
the trial the most catastrophic human experiment in many years. Now it may become the subject
of a Congressional hearing.
The experimental drug was Fialuridine, or FIAU. A four week pilot study of 24 hepatitis B
patients convinced Dr, Jay H. Hufnagle, the principal investigator, that FIAU was the most potent
drug ever tried against the hepatitis B virus. So he began a new study scheduled to last 24 weeks.
This time the doses were slightly larger, in an effort to evaluate the drug's safety and
effectiveness over a longer period.
None of the 24 patients in the pilot study suffered any serious ill effects, but in the new study
several patients developed sudden, severe and unrelenting liver failure. The study was stopped
immediately and all participants were taken off the drug. But then, in an extraordinary turn of
events, other patients who received the higher dose began developing deadly complications
weeks after they had stopped taking FIAU.
The researchers are still not sure what went wrong, but they theorize that the larger doses given
in the second trial damaged energy-producing units called mitochondria within the liver cells.
A review by Dr. Hufnagle's team of all previous studies of FIAU showed that its dangers could
not have been predicted from animal experiments and that monitoring tests had not hinted at
impending disaster in any patient, Dr. Hoofnagle told a scientific meeting at the Food and Drug
Administration on Sept. 21.
ow attention has focused on how well prepared the participants were for the risk of something
going wrong. Dr. Hoofnagle and officials of the National Institutes of Health in Bethesda, Md.,
said that because the participants had signed a six- page consent form before they entered the
study - as they would at any medical center conducting similar research - they knew the risks.
N.I.H. officials called the form a "model."
But Dr. Judith P. Swazey, a nationally known expert on the ethics of human experimentation at
the Academic Institute in Bar Harbor, Me., said if it was a "model," there may be problems with
other consent forms.
She is among a variety of experts in drug research and clinical trials at other medical centers and
drug companies who have reviewed the document and question whether it told the volunteers
enough. Some said it was flawed and should never have been approved. And an examination of
the form in their case raises the question of whether any consent in a study involving desperately
ill patients can be truly "informed."
The FIAU disaster comes at a time of increased pressure on researchers to find cures and of
increased scrutiny of both consent forms and the effectiveness of the 1,200 Institutional Review
Board (I.R.B.'s) at hospitals and other research centers in the United States. Before an
experiment can begin on any patient at any center that receives Federal funds, its I.R.B. must
approve the study and the consent form.
The F.D.A. does not approve specific consent forms, but it can delay approval of drug research if
it objects to a form;s wording. No objects were raised to the form in the FIAU study.
The forms are intended to explain the nature of the experiment and to spell out the potential
benefits and risks of the trial. But the researchers who write them face to major challenges:
determining how much detail to include and choosing words that patients can clearly understand.
Also, potential participants are often desperately ill and may grasp at any straw - even signing a
document without reading it. For this reason, many say there is no such thing as informed
consent, only consent.
"It's really an art to write a consent form so that a rational person would sign it," said Dr. Henry
Greenberg, a cardiologist at Roosevelt-St. Luke';s Hospital who is president-elect of the New
York Academy of Sciences.
George J. Annas, a professor of health law at Boston University, said: "Researchers tend to think
that they do good, that they don;t do bad. And patients feel the same way; they tend to minimize
or totally downplay the risks."
Researchers and their institutions may have vested interests in persuading patients to participate
in studies. Success can lead to larger budgets, academic promotions and, occasionally, to fame.
"Even thought doctors may not distinguish experimentation from treatment when they are doing
a study, they may shift ground once it is over," Mr. Annas said. "If it did well, it was treatment;
and if it did poorly or something went wrong, it was an experiment."
Many experts also say the emphasis among the review boards has shifted from protecting
individual to promoting participation in research, largely due to pressures from activists for
speedier treatment for AIDS.
While researchers generally want to write short summaries that will not scare off potential
volunteers, review boards are charged with protecting participants and trying to avoid false
Dr. Michael Grodin, who has long been chairman of a review board for the Department of Health
and Hospitals in Boston, said: "Research today has become positive, meaning that patients may
demand to get into it and we given them the opportunity to do so. That is a big shift from 25
years ago when I.R.B's were designed to protect people from being abused."
At a news conference, Dr. Hoofnagle said the public must understand that "drugs can do harm
just as much as they do good." He also said that despite all the laboratory and animal experiments
that precede trials on humans, the safety and effectiveness of a drug can be determined only when
it is tested in people.
He added that in his opinion, the FIAU consent form had included broad warnings.
The form spelled out six specific potential risks, including fatigue, nausea, skin rashes,
suppression of the bone marrow, seizures and pain and numbness in the arms and legs. But it
contained only a single sentence about other, unknown hazards: "FIAU is a new medication, and
its side effects have not be completely described."
Dr. Swazey, the Maine ethicist, said the term "new medication" was misleading. She said Dr.
Hoofnagle should have used the term "experimental anti-viral compound" on the form "and
underlined it." Dr. Swazey is co-director of a major study of review boards and consent forms for
Dr. Hoofnagle said he believed none of the patients had thought they were receiving a standard
treatment. And as for warning that FIAU might worsen the disease and even cause, Dr.
Hoofnagle said, "We did not imagine that would happen."
Dr. Grodin suggested omitting any mention of benefits because "probably the easiest, clearest
and most honest thing is to let patients realize that they are involved in research, where the goal
is get generalizable knowledge. If a study benefits a participant, then that is great. But that is not
the primary goal of research."
Dr. Hoofnagle said the policy of the National Institute of Diabetes and Digestive and Kidney
Diseases , where he works at the N.I.H., requires that benefits be listed in the consent form.
N.I.H. officials said at the meeting that they had considered the written document only a part of
the consent process, and Dr. Hoofnagle said he had tried to limit his consent form to six pages
because most patients would not read a longer document. He said other information, including
warnings of potential risks, was covered in discussions with patients.
Among those was the risk of giving up a standard treatment, alpha interferon, a licensed drug for
hepatitis B. All the participants said they either had not benefited from it or could not tolerate its
side effects, Dr. Hoofnagle said.
Critics also said the form was vague about the number of people who had taken FIAU in the pilot
study. Knowing that data would have allowed participants to weigh the relative risks and benefits
of taking part in the study. For example, there was a small risk of death or other complications
from a liver biopsy, a standard test required for the study. Dr. Hoofnagle said omission of the
data was an oversight.
In addition, critics faulted the document for not stating that the study was being funded in part by
Eli Lilly & Company of Indianapolis, which had bought marketing rights for FIAU from
Ocxlassen Pharmaceuticals of San Rafael, Calif. "You assume if you go to the N.I.H. it is a
Government-sponsored project," Mr. Annas said.
Although it is policy at many medical centers to include such information, Dr. Hoofnagle said he
had never included the names of drug companies in any consent forms.
When participants die after a drug study has ended, researchers often say the death was "off
study," not linked to the drug, a panelist at the F.D.A. scientific meeting said. One lesson from
the FIAU study may be that researchers need to be less cavalier and to extend follow-up periods
for many experimental drugs, to be more certain that deaths and other complications are not