NIH to reinvestigate disputed paper
By David P. Hamilton
Less than a week before a two-day congressional hearing is scheduled to review allegations of scientific fraud at an MIT biomedical laboratory, the National Institutes of Health have reopened an inquiry into the validity of the research under question.
The focus of both investigations is a scientific paper published in the journal Cell in mid-1986 by Nobel laureate and Whitehead Institute director David Baltimore '61 in collaboration with then-MIT researchers Thereza Imanishi-Kari and David Weaver. Three other scientists also contributed to the paper, which reported new results in immunological work.
New questions raised by Margot O'Toole, a former postdoctoral fellow at MIT and one of the paper's original challengers, led the NIH to reconvene a scientific panel for a second examination of all data associated with the paper, according to NIH deputy director William Raub, who was quoted in The New York Times.
NIH has been unable to find additional data O'Toole had requested from laboratory notebooks, and as a result it intends to conduct a detailed audit of all information supporting every detail in the paper.
"There is no presumption of right or wrong at this stage," Raub told the Times.
The panel's recently completed investigation of the research had found "significant errors" in the paper, but "no evidence of fraud, conscious misrepresentations, or manipulation of data" by the authors. An internal MIT review, led by Professor Herman N. Eisen, reached similar conclusions. The authors have written two letters of correction to Cell, although the second was written under protest at the behest of the NIH panel.
Forensics experts from the Secret Service were brought into the investigation by Rep. John Dingell (D-MI), chairman of the subcommittee on Oversight and Investigation, according to a committee aide. These experts have reportedly discovered evidence in Imanishi-Kari's laboratory notebooks that suggests several dates were changed and that some data was entered long after the experiments took place.
Alfred Kildow, a spokesman for the Whitehead Institute, disputed the conclusions drawn by the Secret Service, saying that the forensics work "doesn't relate at all to the data in the Cell paper."
Baltimore, Imanishi-Kari, and Weaver all intend to appear at the hearing later this week, Kildow said. He refused to discuss any preparations the scientists were making for the hearing.
An incriminating letter?
Another point of dispute lies in a letter Baltimore sent to Eisen
in September of 1986 in which Baltimore said that a particular antibody used in the experiments and known as Bet-1 did not work the way Imanishi-Kari had described it.
"The evidence that the Bet-1 antibody doesn't do as described in the paper is clear. Thereza's statement to you that she knew it all the time is a remarkable admission of guilt," Baltimore's letter stated.
Kildow said that Baltimore's letter was based on a misunderstanding conveyed by Eisen, who had a "chance meeting" with Imanishi-Kari. Because English is Imanishi-Kari's fifth language, Eisen misunderstood a comment she made about the Bet-1 antibody and repeated it to Baltimore, who responded with the letter a day or so later, Kildow said. Eisen confirmed Kildow's version of events.
Neither Baltimore nor any of the other participants ever made a written correction to Baltimore's letter. Kildow explained that the participants were all aware of the error and that no written correction was necessary.
The Cell paper reported experiments on laboratory mice that seemed to indicate that the introduction of foreign genes into an animal could lead to the expression of related genes within the animal, a topic which the biological community is still debating. The finding could have implications for immunological study and gene transplant work.
The congressional hearing and the NIH investigation are culminating nearly three years of controversy over the research. Among the serious issues raised have been questions about the efficacy of research institutions' attempts to respond to charges of error, fraud, and scientific misconduct, the motives of the politicians investigating scientific work, and the freedom of junior scientists to contest the work of their seniors without fear of reprisal.
Noting that members of Congress have proposed legislation to reinforce institutional safeguards against fraud, a subcommittee aide said, "I'm not sure we're in favor of legislative solutions to the problem. It's in the institutions' best interest to police themselves before the federal government steps in."
"This is a very instructive case," he continued, adding that the reluctance of the researchers -- Imanishi-Kari, in particular -- to cooperate with the investigators has complicated matters unnecessarily.
The opinion of the scientists involved, however, is that any government intrusion is too much. "The more this goes on, the clearer it is that this is not the way to adjudicate scientific issues," Eisen said. He worried openly that the hearing might be a "launching pad" for "the regimentation of science," with strict regulations for conducting experiments, taking notes, and verifying results that would "take all the fun out of science."
Eisen said that while the institutional safeguards against fraud are not perfect, he could not imagine any way in which federal intervention could improve them. "This whole matter is the most discouraging thing I've seen in 30 years," he said.