July 27, 2001
By JAMES GLANZ
BALTIMORE, July 26 — In the nebulous language of eulogy, Ellen Roche, a bright, healthy, 24- year-old volunteer in an asthma study at Johns Hopkins University, died for the sake of science. In clinical terms, Ms. Roche died because a chemical she inhaled in the study led to the progressive failure of her lungs and kidneys.
But in the plain speech that has been missing from official reports on the disastrous experiment, medical experts say Ms. Roche's death on June 2 was almost certainly caused by the failure of scientists and administrators charged with her well-being to recognize that the experiment was unsafe.
Still, while they agree that clues about the risky nature of the experiment are clear in retrospect, scientists and other experts around the country remain divided about whether the evidence was obvious enough at the time to assign culpability to one researcher, to Hopkins, or to the overall system for assessing research involving humans in the United States.
An examination of the case by The New York Times, based on official documents and interviews with university officials, asthma specialists and experts on clinical trials, suggests that the researcher who conducted the experiment and the ethics panel that approved it overlooked numerous clues about the dangers of the chemical, hexamethonium, given to Ms. Roche to inhale.
While Hopkins has formally accepted what its officials have called "institutional responsibility" for Ms. Roche's death, many university officials have rejected criticisms of their procedure for evaluating the safety of experiments involving humans.
Much of that resistance remains, even though last week an agency investigating the death briefly suspended all federally financed research involving human subjects at Hopkins, saying the university's methods of reviewing such research did not meet federal standards.
The agency, the Office for Human Research Protections, eased the suspension this week after Hopkins officials agreed to change the way its internal panels assess proposals for human research. Those panels, called institutional review boards, have been ordered to review again more than 2,000 research projects before the work can resume.
Johns Hopkins has suspended all the research projects of Dr. Alkis Togias, the associate professor of medicine who designed and carried out the asthma study.
"In every investigation that has taken place we've cooperated fully," Dr. Togias said in a brief telephone interview. "The idea is to really find out what happened and if there are any issues to correct."
Despite the government's actions against Hopkins, Dr. N. Franklin Adkinson Jr., an allergist and immunologist, professor of medicine at the university and member of the review board that approved the fatal experiment, said that the death could not have been foreseen.
"Well-meaning and sophisticated individuals can come to different conclusions as to whether this was a grave departure from our responsibility," Dr. Adkinson said. He said the required changes were no more than busywork.
"I don't feel it's adding one iota of safety in human research protocols," he said.
Many scientists and investigating officials disagree. William Hall, a spokesman for the research protection agency, likened the fatal experiment to a plane crash and said the changes were essential for ensuring that other volunteers at Hopkins were not put at risk.
George Annas, chairman of the health law department at the Boston University School of Public Health, said, "Somebody didn't do their job." He said the resistance at Hopkins was understandable, in human terms, "because if the chemical you introduced killed them, then you killed them, and that's almost impossible to acknowledge."
Dr. Cynthia Dunn, director of the Clinical Research Institute at the University of Rochester, who oversees review boards there, said the Hopkins board had committed a "serious oversight" in its quick approval of the use of hexamethonium, which is unapproved by the Food and Drug Administration and has not been used clinically since 1972.
Some scientists, like Dr. Warren M. Gold, a pulmonologist who is a professor of medicine at the University of California at San Francisco, said the experiment had important flaws in its design. In hindsight, he said, evidence of the chemical's dangers could easily have been found in the published literature as well as from problems that cropped up during the experiment.
But he said that, tragic as the death was, some researchers might question whether that evidence would have been so clear to a typical scientist in the middle of an experiment.
"Anybody doing clinical research dealing with human beings knows that there are risks," Dr. Gold said. "Accidents do happen."
The purpose of the hexamethonium experiment was to test a theory held by Dr. Togias and a few colleagues about what happens during an asthma attack.
The airways of all people constrict when exposed to certain irritants like pollen, pollutants or some drugs. People without asthma can usually make the airways relax again quickly just by taking deep breaths. People with asthma cannot do that.
According to Dr. Togias's theory, in healthy people, the expansion of the airways during deep breaths may send a signal along nerve pathways to the brain, which sends back a return signal that relaxes the airways. But in people with asthma, the signal is somehow blocked, and the airways remain constricted.
The Hopkins study was intended to test this theory by giving healthy, nonasthmatic volunteers a drug that constricted their airways in a way that mimicked asthma. Some of those volunteers would also be given hexamethonium, a so-called ganglionic blocker that would prevent signals from returning to the lungs — and, if the theory was right, stop the airways from reopening.
Hexamethonium was once used orally to treat high blood pressure, but the F.D.A. rescinded its approval for the chemical in 1972 after deciding that it was ineffective. In the 1950's, research had also found that hexamethonium could be toxic to lungs.
But when Dr. Togias applied to the Hopkins review board for approval to perform the experiment, his literature search apparently failed to turn up those papers, even though investigators later found references to them on sites available through standard Internet search engines like Yahoo! and Google.
Instead, the application used several much more recent papers to argue that inhaling large doses of hexamethonium, up to a gram, would be safe for the volunteers.
"Obviously the dose should be defended," said Dr. Harold Nelson, an allergist at the National Jewish Medical and Research Center in Denver who described himself as a friend of Dr. Togias's. He and other researchers said Dr. Nelson had a good professional reputation.
"You have to say why you're using a particular dose, what the background science is for the selection of the dose," Dr. Nelson said.
Although typical doses of approved drugs used by inhalation, like the asthma drug albuterol, are thousandths the size of that given Ms. Roche, Dr. Nelson said there was no reason that inhaling a gram of an active substance should be inherently dangerous.
But the Hopkins review board was criticized by the federal oversight agency for not questioning the use of hexamethonium, for not finding the earlier literature after Dr. Togias had missed it, and for allowing the chemical to be called a "medication" in the consent document signed by the volunteers.
"This isn't a medication," said Leonard Glantz, a professor of health law at the Boston University School of Public Health and a member of a review board there. He said the experiment, when it came to the language of the consent form, was no different from one in which subjects might inhale gasoline or another nontherapeutic chemical.
"One of the things that makes this so disconcerting is that the best in the world didn't know this stuff," he said, referring to the problems with the literature search. "It's a little creepy."
Hopkins officials have protested such assessments, saying that the 1950's literature may not have made any difference to the approval of the experiment, since the earlier findings of lung damage from hexamethonium involved people with "malignant hypertension," or serious blood pressure problems.
"They were treated for many months before these pulmonary complications occurred," Dr. Adkinson said, arguing that the lung problems could have been unrelated to the hexamethonium.
The investigation by the oversight agency found, however, that the Hopkins review boards were overburdened, had little institutional support and were not versed in the federal regulations governing experiments involving humans.
Instead of discussing each proposed project in full meetings of the review board, for example, as required by federal law, the board circulated proposals individually to its members through a subcommittee that collected questions and criticisms from them and communicated with the researcher.
"Our belief is that this is an adequate review," said Dr. David Cornblath, a Hopkins neurologist who is a review board member. But he said that the number of review boards has been increased and the procedures changed to suit federal regulations, which he characterized as "checklist-driven" red tape.
Beyond the 1950's papers, a possibly more glaring indication of the chemical's danger to humans was known only to a small group of researchers in San Francisco.
The strongest piece of evidence offered by Dr. Togias that inhaled hexamethonium in gram amounts was safe was a 1978 asthma study by a group of researchers at the University of California at San Francisco (they did not include Dr. Gold). No adverse reactions to the chemical were reported in the published description of the experiment, but this week the university announced that two of the volunteers had become sick, one with lung problems.
Because doctors outside the study diagnosed the latter problems as viral pneumonia unrelated to the experiment, they were never reported. But after the death of Ms. Roche, the researchers say they are reconsidering that conclusion, and the university has opened a new review of the experiment. Daniel Kracov, Dr. Togias's lawyer, said this week that Dr. Togias would not have proceeded if he had known of those events.
In any case, Dr. Togias did receive the go-ahead from the Hopkins institutional review board. The letter informing him of that fact reminded him that any changes to the experiment had to be approved in advance by the board and that "adverse events should be reported to the I.R.B. promptly."
He followed neither admonition, but once again, the degree to which his actions might make him culpable for Ellen Roche's death is a matter of dispute.
The first subject in the study developed a cough and shortness of breath after receiving hexamethonium via a nebulizer, or compressor- driven spray, through a standard mouthpiece. Dr. Togias's lawyer later argued in a public letter that the origin of the cough was "either a cold or the acidity of the original hexamethonium solution."
Without informing the review board, Dr. Togias changed the solution in which the chemical was delivered, making it less acidic. Later, he also attempted to speed up the hours- long sessions, which had led to complaints among the subjects, by using a nebulizer that delivered the chemical more quickly.
Those specific changes, some scientists have argued, would probably have been approved by a review board, since they were meant to make the subjects more comfortable.
Dr. Nelson, Dr. Togias's friend, said that "technically, yes," the change in the formulation "should have been sent through" to the review board. The report would presumably have included the reason for the change in the acidity, the first volunteer's cough.
On May 4, the hexamethonium was administered to Ms. Roche, a technician from a nearby lab in the same institution who had volunteered for the study. A day later, she began coughing, and soon had a fever, runny nose and fatigue. By May 12, a CAT scan revealed that her lungs had the appearance of ground glass, indicating severe damage, and two days later she was put on a ventilator.
"Given her worsening condition and unsupportable oxygenation," reads a Hopkins report on the incident, "her family elected to withdraw support and Ms. Roche died on June 2, 2001."