Misconduct in Science: BIOMEDICAL SCIENCE CASES

Misconduct cases have been more prominent in the biomedical sciences than in the physical and social sciences. This may be because there are more people working in biomedical research than in physical or social science research or because misconduct in biomedical research is more likely to have direct, harmful effects on human beings. Several have been high-profile cases, attracting the attention of the scientific community, independent watchdogs, governmental agencies, and the public at large. The following four cases are some of the best-known instances of alleged misconduct and depict a variety of the ethical issues related to misconduct in biomedical research.

The Sloan-Kettering Affair

In 1974 William Summerlin was at the Sloan-Kettering Institute for Cancer Research, continuing work on a project that he and his supervisor, Robert Good, had begun while the two were at the University of Minnesota. Preliminary data from experiments there had suggested that some tissues, when incubated for several weeks in culture, cease to produce an immune response. If supported, that finding would have dramatic implications for transplantation science, allowing transplants between any two individuals without the risk of rejection.

Summerlin and his coworkers at Sloan-Kettering were having difficulty replicating the results of those initial studies, and as a result, Summerlin had little to show Good in a progress meeting in March 1974. In the elevator on his way to the meeting Summerlin used a marker to draw what appeared to be successful skin grafts on two of the laboratory mice and represented them to Good as successful transplants. Good failed to notice the fraud, but a laboratory assistant caring for the mice discovered the black spots later that day. When he was able to wash the spots away with alcohol, the assistant reported Summerlin. A review committee was established to look into the case.

In the investigations of the affair it was found that Summerlin's data from another transplant experiment conducted in the same period had been falsified. Summerlin had begun a study with two ophthalmologists that was designed to test the same hypothesis: that incubated tissues would not produce an immune response. The protocol required the ophthalmologists to transplant a fresh human cornea onto a rabbit's left eye and then transplant the donor's other cornea into the rabbit's right eye after it had been in tissue culture for several weeks. When Summerlin observed the rabbits, he saw unsuccessful transplants in the rabbits' left eyes and what looked to be successful transplants in their right eyes. He disseminated those remarkable results at several scientific meetings with confidence. In fact, however, the two ophthalmologists had not done the second transplant on any of the rabbits; therefore, what Summerlin interpreted as successful corneal transplants were actually the rabbits' own corneas. Summerlin later claimed that he was unaware that his coinvestigators had not completed the second half of the protocol.

The institute determined that Summerlin had misrepresented data in both cases. The review committee further concluded that Summerlin had been experiencing emotional problems and placed him on medical leave for a year rather than imposing official sanctions.

Subsequent testing of the only available mouse from the Minnesota laboratory that had undergone a successful skin graft and that had formed the basis for Summerlin's work at Sloan-Kettering revealed that the mouse was a genetic hybrid rather than a purebred mouse, as had been recorded. Because the purebred mouse would have been expected to reject the skin graft but the hybrid mouse would not have, this explains the success of the graft in that case. It is not known whether the hybrid mouse was selected deliberately or accidentally.

The Darsee Case

John Darsee was a prolific and well-liked postdoctoral fellow at the Brigham and Women's Hospital, an affiliate of Harvard Medical School. In May 1981 Darsee's coworkers observed him fabricating data by recording data gathered over several hours so that the data appeared to have been collected over a two-week period. Caught in the act, Darsee apologized and claimed that it had been an isolated incident.

An internal investigation led by Eugene Braunwald, Darsee's supervisor, was conducted, but the incident was not disclosed publicly until several months later, when the investigators uncovered data that Darsee had generated for a multicenter study funded by the National Institutes of Health (NIH). Inexplicable discrepancies between Darsee's data and the results from other participating institutions were found, precipitating an independent investigation and notification of the NIH.

The NIH then launched its own review of Darsee's research. The review committee found problems in five of the papers that Darsee had published and on which Braunwald had been a coauthor and recommended that Darsee be barred from eligibility to receive NIH funding for ten years. The panel condemned Braunwald's supervision of Darsee, stating that his hands-off approach had inhibited the discovery of Darsee's fabrication. In response Braunwald argued that he had followed standard laboratory practices.

Further investigation into work done previously by Darsee at Emory University and Notre Dame uncovered instances of data fabrication and falsification in at least twelve of Darsee's papers that were based on research he had conducted at those institutions.

Harvard Medical School was criticized for its handling of the case and subsequently revised its policies for responding to charges of misconduct. In particular the review committee claimed that the NIH had had a right to know that Darsee, who had continued work on NIH-sponsored research for six months after the incident, had been caught fabricating data.

The Gallo Probe

A well-known controversy involving the isolation of the acquired immune deficiency syndrome (AIDS) virus illustrates a third form of scientific misconduct: plagiarism. In May 1984 a series of four papers appeared in the journal Science written by Robert Gallo and his team at the National Cancer Institute, stating that they had identified the virus that causes AIDS and proposing a process for developing a blood test for the virus. Mikulas Popovic, working in Gallo's laboratory, had been able to grow the retrovirus in cells that could survive infection with the virus, a cell line that he called H9. It later was revealed that the H9 cell line had not been developed by Popovic but instead had been cloned from a cell line called HUT78 that had been given to the Gallo laboratory by John Minna's team at the Veterans Administration. Minna's group was not credited in the Science papers for that significant contribution.

A second and more high-profile dispute accompanied the Gallo group's accomplishment. In July and again in September 1983 Luc Montagnier's team at France's Pasteur Institute had sent a sample of a viral isolate called LAV to Gallo's laboratory. In spring 1984 Popovic used the H9/HUT78 cell line to grow an AIDS retrovirus, which Montagnier's laboratory had been unable to do because it did not have a cell line that could survive infection with the virus. Gallo was able to produce sufficient quantities of the virus, which he named HTLV-III, to develop a method for testing for the presence of the virus in blood. It was discovered later that HTLV-III and LAV were the same virus, although Gallo had not acknowledged the contribution of the Pasteur Institute. Gallo claimed that the use of LAV was unintentional and must have contaminated his cultures accidentally.

The NIH's Office of Scientific Integrity (OSI) conducted an investigation and found Popovic guilty of four counts of misconduct but held Gallo responsible only for exhibiting a lack of collegiality. In a later investigation by the OSI's successor, the Office of Research Integrity (ORI) at the Department of Health and Human Services (DHHS), Gallo was found guilty of intention to deceive the scientific community about the origin of the materials used to isolate and replicate the AIDS virus. In 1993, however, a federal appeals board cleared Popovic and therefore Gallo of the misconduct charges, citing a lack of evidence that the virus had been stolen.

The Gallo case was significant not only because of the recognition and prestige associated with receiving credit for a discovery of this magnitude. The patent on the blood test for AIDS virus antibodies was lucrative, producing millions of dollars in royalties. It eventually was agreed that those royalties would be split evenly between the United States and France.

The Baltimore Case

Perhaps the most infamous instance of alleged misconduct in the biomedical sciences was the affair that would come to be known as the Baltimore case, even though David Baltimore, for whom the case is named, was not accused of fraud. Baltimore did, however, staunchly defend Thereza Imanishi-Kari against claims that she had fabricated data in a paper on which he was a coauthor. When the accusations were made, Baltimore was a professor of biology at the Massachusetts Institute of Technology (MIT) and the director of the Whitehead Institute. He had been awarded a Nobel Prize in 1975 for his work in virology. Imanishi-Kari was working with Baltimore on a complex project investigating the mechanisms behind the immune response.

Margot O'Toole took a postdoctoral fellowship in Imanishi-Kari's laboratory in 1985, and the two clashed from the beginning. O'Toole was having difficulty getting results consistent with Imanishi-Kari's data and had some problems with the experimental method. When she approached Imanishi-Kari with her concerns, she was dismissed and told that the discrepancies were due to incompetence. While attempting to understand the discrepancies between Imanishi-Kari's results and her own, O'Toole came upon evidence that she believed showed that data in a 1986 Cell paper coauthored by David Baltimore had been misrepresented.

O'Toole brought her concerns to senior scientists at both MIT and Tufts University, where Imanishi-Kari had taken a position. Informal investigations were conducted at both institutions. Errors were found in the paper, but the investigators believed that O'Toole's problems with the paper were scientific disagreements and did not demonstrate misconduct. O'Toole and a former graduate student of Imanishi-Kari's continued to push the issue, notifying NIH scientist's Walter Stewart and Ned Feder. In doing so, they sparked parallel investigations by the NIH and by the congressional subcommittee on oversight and investigation with jurisdiction over the NIH that continued for the next six years.

In 1994 a report by the ORI found Imanishi-Kari guilty of numerous counts of fabricating and falsifying data and banned her from receiving NIH funding for a period of ten years. However, two years later the DHHS's Research Integrity Adjudications Panel exonerated Imanishi-Kari of fraud. The panel made note of the many errors in the Cell paper as well as the sloppiness of Imanishi-Kari's bookkeeping but stated that solid evidence of intentional misrepresentation was lacking. That was the second ruling by the ORI that had been overturned by an expert panel (the first had been the Gallo ruling), shedding doubt on the office's ability to police scientific misconduct.

The Baltimore case also raised questions about the treatment and protection of whistle-blowers. Throughout the ten years of the ordeal O'Toole was alternately ostracized and praised for her actions and was unable to find work in science. Her experience and the similar experiences of others sparked a movement that resulted in improved protections for whistle-blowers.

Results and Changes

These four cases illustrate a variety of the difficult issues related to scientific misconduct. They raise questions about the high expectations placed on researchers and about authorship requirements, supervision of laboratory work, appropriate attribution of credit, collegiality, transparency of data recording, and treatment of whistle-blowers. These cases also demonstrate that the distinction between honest errors or omissions and intentional fraud is not an obvious one. Significant improvements in the process used to negotiate the murky waters of scientific misconduct have come out of these experiences.

In some cases, a rapid and transparent response to revelations of misconduct may minimize the damage done by those revelations. In 1996, Francis Collins, head of the human genome project at NIH, became aware that data had been falsified by one of his graduate students in five papers that he had coauthored. Collins promptly confronted the student, informed researchers for whom the information would be relevant, retracted two of the papers and corrected sections of three others. Although this case differs from those above in that the researcher accused of misconduct did not deny the allegations, it may illustrate the advantages of dealing with instances of misconduct quickly and openly.

JANET MALEK

SEE ALSO Bioethics;Medical Ethics.

BIBLIOGRAPHY

Anderson, Christopher. (1993). "Popovic Is Cleared on All Charges; Gallo Case in Doubt." Science 262(5136): 981–983.

Cohen, Jon, and Eliot Marshall. (1994). "NIH-Pasteur: A Final Rapprochement?" Science 265(5170): 313.

Culliton, Barbara J. (1974). "The Sloan-Kettering Affair: A Story without a Hero." Science 184(4137): 644–650. First part of a review of the Sloan-Kettering affair.

Culliton, Barbara J. (1974). "The Sloan-Kettering Affair (II): An Uneasy Resolution." Science 184(4142): 1154–1157. Second part of a review of the Sloan-Kettering affair.

Culliton, Barbara J. (1983). "Coping with Fraud: The Darsee Case." Science 220(4592): 31–35. Review of the Darsee case and its fallout.

Culliton, Barbara J. (1990). "Inside the Gallo Probe." Science 248(4962): 1494–1498. First part of a review of the Gallo probe.

Hamilton, David P. (1991). "NIH Finds Fraud in Cell Paper." Science 251(5001): 1552–1554.

Kaiser, Jocelyn, and Eliot Marshall. (1996). "Imanishi-Kari Ruling Slams ORI." Science 272(5270): 1864–1865.

Kevles, Daniel J. (1998). The Baltimore Case. New York: W. W. Norton. Detailed history and analysis of the Baltimore case.

Marshall, Eliot. (1996). "Fraud Strikes Top Genome Lab." Science 274(5289): 908.

Palca, Joseph. (1992). "'Verdicts' Are in on the Gallo Probe." Science 256(5058): 735–738.

Rubenstein, Ellis. (1990). "The Untold Story of HUT78." Science 248(4962): 1499–1507. Second part of a review of the Gallo probe.