Munich

The fallout from Germany's most notorious case of alleged systematic scientific fraud is proving even more damaging than expected to the reputation of the country's clinical research. Attention had previously focused on a cancer researcher who had worked at the universities of Freiburg, Berlin and Ulm, but it is now suspected that other top clinical professors may have been involved in possible scientific misconduct.

Such suspicions were raised in a report published earlier this week by a task force that has been analysing the publications of Friedhelm Herrmann, who was first accused of fraud in 1997, and who now works as a private practitioner.

The Deutsche Forschungsgemeinschaft (DFG), Germany's research council, has followed up suspicions raised by the task force and has launched a formal investigation into a paper published in the journal Blood in 1994, of which Herrmann was not an author.

Rapp: has analysed over 300 papers in the past two years.

The task force, a small team of scientists headed by Ulf Rapp, a cell biologist from the University of Würzburg, has over the past two years been analysing the data in 347 papers published by Herrmann between 1985 and 1996. The researchers used computer programs to look for evidence of copying or manipulation of computer-stored autoradiograms, as well as photographs of suspected proteins and other molecular biological data.

It also interviewed some of the papers' co-authors in a bid to reconstruct the history of each experiment. In its report, published this week, the task force says that 94 papers include data that it thinks have been either definitely or ‘highly probably’ manipulated.

Fifty-three of these were published jointly by Herrmann with his former colleague Marion Brach, who has already admitted fabricating data (see Nature 387, 750; 1997). Fifty-nine — including some involving Brach — were co-authored by Roland Mertelsmann, chair of the department of haematology and oncology at the University of Freiburg, in whose department Herrmann had worked.

Mertelsmann, who is well known for being the first scientist to conduct a gene-therapy trial in Germany, has accepted responsibility for the suspect publications. But he says that he was only an honorary author, and was therefore unaware of the details of experiments described in them.

Mertelsmann: data in one of his papers are under investigation. Credit: DPA/ROLF HAID

Only 132 papers have been given a clean bill of health by the task force. The remaining 121 were placed in a ‘grey category’; a lack of cooperation from some authors, says Rapp, limited his team's access to original data, and the manipulation of data could therefore be neither proved nor disproved. Three of the five most frequent co-authors provided no information at all to the task force.

But, in an effort to gauge the extent of the problem, the task force also looked at five randomly selected papers on which Mertelsmann — but not Herrmann — was listed as a co-author. It asked the authors for original patient data to allow them to analyse tables and graphs.

According to Rapp, sufficient original data were provided to fully analyse only one paper, published in Blood in 1994 (volume 84, 1421). This paper investigated a technique to help cancer patients recover from the bone-marrow damage that reduces blood-cell counts following high-dose chemotherapy.

The paper looked at the rate of recovery of blood cells when the patients' own peripheral-blood progenitor cells, taken before chemotherapy, were transplanted back, either unseparated or purified. The paper concluded that recovery was rapid and complete in either case.

But the task force's report says the paper contained many “irregularities and indications that data had been improperly handled”. Rapp told Nature that the published graphs describing blood-cell recovery omitted data on a significant proportion of the patients. “If you mis-portray clinical data in important journals you may encourage others to adopt a practice that could put patients at risk,” he says.

In response to the task force's findings, the DFG has launched a formal investigation of Mertelsmann and of Lothar Kanz and Wolfram Brugger, respectively professor and senior researcher at the University of Tübingen, and co-authors of the Blood paper.

The University of Freiburg has also opened an internal investigation into the allegations. Kanz declined this week through his lawyers to comment on the investigation; the two other authors did not respond to several requests from Nature to discuss the issues raised by the task force in its report.