Medicine: Immune Cells May See Strain of HIV But Be Blind to Viral Cousin Public Health: Harvard Reports Efforts Against AIDS in Africa Genetics: Formin Gene May Explain a Common Cause of Female Infertility International Health: HMI Trains Trainers on HIV in India Biosecurity: Conference Fosters Dialog on Biosecurity   Crystal Structure Sheds Light on Angiogenesis Small Molecules Confound Lipid-transferring Ability of 'Good' Cholesterol HMS Lends Hand to Landmark Mouse Genome Study   The HMS Faculty Council Senator Kennedy Honored with Richmond Award Harvard President Calls for Support of Scholars at Risk Grillo Surgery Professorship Announced Memoir Tells of Women's Heart Attack Survival, MGH Doctors Who Helped AAMC Honors Korsmeyer for Distinguished Biomedical Research Beals Endow Associate Professorship at HSPH In Memoriam: Ruthanne Simmons Honors and Advances   Scientific Sloppiness is Bad News for Translational Research   Front Page

FORUM Scientific Sloppiness is Bad News for Translational Research Jan Schmollinger Photo by Jeff Cleary About five years ago a scandal involving falsified data shook up several German universities and research centers. Due to the involvement of prominent biomedical researchers and the need to withdraw several articles in high-ranking journals, the case created interest in other countries and even outside the scientific community. Many institutions and funding agencies in Europe and the U.S. responded by establishing clear guidelines for researchers and by creating mechanisms to investigate scientific fraud. Despite all the newly implemented measures, however, two very recent examples of falsification of primary data at Bell Labs and the Lawrence Berkeley National Laboratory indicate how difficult it is to root out scientific misconduct. The suspicion of fabricated data is bad enough, but it is especially troubling in any study that involves patients or has a direct impact on patient treatment. In mid-November the University of Göttingen in Germany released the report of a misconduct panel that had investigated the clinical trial of an experimental cancer therapy. Two years ago scientists from Göttingen, along with their collaborators, had claimed a breakthrough with a vaccine that consisted of fused renal cancer cells and dendritic cells. While some scientists heralded the study as a significant success in the treatment of renal carcinoma, others remained skeptical. Subsequently, signs of sloppy science, such as using images without proper citation of the source, slowly started to surface, and an investigation was launched. The investigation did not turn up any evidence of scientific fraud in this case, but several unsettling findings emerged, leading the misconduct panel to conclude that principles of good scientific practice had been ignored. The press release issued by the university states that among other points, the ethics committee had not been informed and patients who had not fulfilled the study criteria had been included in the published evaluation. Furthermore, the steps that led to the production of the vaccine as well as the immunological results of the vaccination had not been properly documented. According to a recent news article in the journal Nature, at least 400 patients have received this type of vaccine. Betting on Science The bad news could get worse if the report of the misconduct panel results in problems for translational research in the future. When patients volunteer to take part in an experimental study, they accept some kind of wager. Cancer patients who are given the opportunity to take part in clinical trials often do not have much to lose. They are in a late stage of their disease and there is not much hope that conventional forms of treatment will improve their situation. Given this predicament, many patients are willing to try new therapeutic procedures. They accept the risk of certain side effects in exchange for the chance that the new treatment will turn out to have an effect on their disease. Clinical science relies on these volunteers to test new concepts and refine existing regimens. Experiments in model organisms like mice and primates sometimes have a very limited capacity to predict outcomes in humans, and toxic side effects can often only be estimated. The Fear of Fraud Reports like the one from northern Germany are disconcerting for bench-to-bedside research in general. Even if there was no fraud and even if none of the treated patients was in any danger at all, as the commission claims, many of the clinical study's participants are likely to feel misled, and other patients may now think twice if asked to participate in an experimental trial. Whether a patient seeks treatment in Europe, the U.S., or elsewhere, any news about a lack of quality control and bypassed ethics committees in a trial is going to discourage trust. A second problem that may arise is the introduction of unnecessary hurdles for new treatments. Following the proper guidelines for quality control, scientific evaluation and, above all, patient care is of utmost importance. The type of described violations, however, may cause overcautious officials elsewhere to be much more reluctant to accept new treatment protocols even if there is no scientific basis for this decision. Given that the well-being of patients is at stake, this hesitation is fully understandable. Yet most coverage of the investigations at the two physics labs and the flawed science behind the renal cancer trial suggests a less burdensome remedial step. All three cases, despite their significant differences, have one thing in common: too few of the study authors seem to have been involved in examining the primary data that represented the very core of the findings. Perhaps rectifying this problem--the apparent lack of collective ownership and responsibility on the part of authors--would be a good first step toward shoring up trust in clinical trials and the process of science in general. --Jan Schmollinger, an HMS research associate in medicine at the Dana-Farber Cancer Institute