A group of cancer researchers formerly centered at Johns Hopkins have retracted two more studies. The previous two retracted papers — one of which was the focus of a lawsuit — were about prostate cancer, while the new retractions are of papers about colon cancer.
Here’s the notice for one paper:
The authors wish to retract the article entitled “Evaluation of Colon Cancer-Specific Antigen 2 as a Potential Serum Marker for Colorectal Cancer,” which was published in the March 1, 2008, issue of Clinical Cancer Research (1).
As the authors re-reviewed the data presented in the article, they identified differences between some of the plate reader values and those that were reported in the article. For some of the data, the values were indeed from the plate reader data, whereas the source of other values was not apparent. Therefore, the authors believe that the article may contain data falsification. Furthermore, the values were not blanked. The authors feel that the inconsistencies in the data collection and recordation warrant retraction of the article.
Five of the 6 authors of the original article agreed to this Retraction. Attempts on the part of the journal office to contact the first author, Eddy S. Leman, were unsuccessful.
Robert E. Schoen
Division of Gastroenterology, Hepatology, and Nutrition and Department of Epidemiology, University of Pittsburgh and the University of Pittsburgh Cancer Institute, Pittsburgh, Pennsylvania
Ahmed Magheli
The Brady Urological Institute, Johns Hopkins University School of Medicine, Baltimore, Maryland
Lori J. Sokoll, Daniel W. Chan, and Robert H. Getzenberg
The Brady Urological Institute and Department of Pathology, Johns Hopkins University School of Medicine, Baltimore, Maryland
The paper, which has been cited 21 times, according to Thomson Scientific’s Web of Knowledge, had been criticized by the University of Toronto’s Eleftherios Diamandis. In a letter cataloging his issues with the technique, Diamandis wrote:
…the biomarkers proposed by this group of investigators for prostate and colorectal cancer suffer from methodological and interpretative shortcoming. These papers give the impression that newer and improved circulating cancer biomarkers have been discovered and are on their way to the clinic. I believe that it is premature to draw any conclusion on the validity of these series of biomarkers until they are validated independently and blindly by agencies like the Early Detection Research Network of the National Cancer Institute (http://edrn.nci.nih.gov/).
The study was funded by National Cancer Institute grant CA084968, Onconome, Inc. — which sued the researchers, and settled — and the Patana Fund of Hopkins’ Brady Urological Institute.
Here’s the notice for the other paper:
The authors wish to retract the article entitled “Initial Analyses of Colon Cancer–Specific Antigen (CCSA)-3 and CCSA-4 as Colorectal Cancer–Associated Serum Markers,” which was published in the June 15, 2007, issue of Cancer Research (1).
As the authors re-reviewed the data presented in the article, they identified differences between some of the plate reader values and those that were reported in the article. For some of the data, the values were indeed from the plate reader data, whereas the source of other values was not apparent. Therefore, the authors believe that the article may contain data falsification. Furthermore, the values were not blanked. The authors feel that the inconsistencies in the data collection and recordation warrant retraction of the article.
Six of the 7 authors agreed to this Retraction. Attempts on the part of the journal office to contact the first author, Eddy S. Leman, were unsuccessful.
Robert E. Schoen
Division of Gastroenterology, Hepatology, and Nutrition and Department of Epidemiology, University of Pittsburgh and the University of Pittsburgh Cancer Institute, Pittsburgh, Pennsylvania
Joel L. Weissfeld
Department of Epidemiology, University of Pittsburgh and the University of Pittsburgh Cancer Institute, Pittsburgh, Pennsylvania
Lori J. Sokoll, Daniel W. Chan
The Brady Urological Institute and Department of Pathology, Johns Hopkins University School of Medicine, Baltimore, Maryland
Grant W. Cannon, Robert H. Getzenberg
The Brady Urological Institute, Johns Hopkins University School of Medicine, Baltimore, Maryland
The second paper has been cited 31 times, and was also criticized by Diamandis. It notes that:
Dr. R.H. Getzenberg holds a patent for the technology described in this article. This patent is owned by the University of Pittsburgh and has been licensed to Onconome, Inc. Dr. R.H. Getzenberg is a consultant to the company.
The text of the retraction notices is basically the same as that of a Prostate paper the group retracted earlier this year, with the addition of the comment about Leman. Leman appears to have been at Stanford most recently, but we can’t find him, either. And Getzenberg left Hopkins in July, as we’ve reported. He appears to be an employee of GTx, a Memphis-based biotechnology company.
I wonder if retraction notices are copyrighted. Retracting a retraction notice due to plagiarism would make an interesting legal case.
BTW, “ad was also” should probably read “and was also”.
Fixed, thanks.
In fact, the retraction notice by the “Brazilian Forensic Entomologist” figuring in this very blog plagiarized another notice from a previous authors, as the author himself (who doesnt know English awfully well) wrote the text.
“Furthermore, the values were not blanked.”
So aside from suspected efforts to manipulate the numbers in order to produce the desired outcome, all of the values are simply wrong because they include the background value of the buffer. And this has happened more than once. Somebody needs a refresher on how to use the instrument.
So sad if the failure to blank turns out to be an error by an undergraduate worker who was running the plate reader. If it was a more advanced lab worker … . Education is a slow and inefficient process. It should not be subjected to high-throughput methods.
This addition might be their attempt to assuage other fears about data manipulation; namely, “Oh, well, the experimental results were inaccurate to begin with because the values weren’t blanked, so this means that we shouldn’t trust the data anyhow–manipulated or not.
No idea why they included that after the fact that the values had been changed because it’s pretty much moot, in my opinion. But I know a lot of post-docs and grad students who don’t run blank buffers in their plate readings. It’s a fundamental control, and I have no idea why it’s not done.
I wonder if they intended to do a thorough job of dismantling the reputation of the first author. Not only were the results faked, but the researcher had poor lab skills. And clearly the first author is the one who is getting the blame, because the other authors were responsible enough to sign the retraction request and to explain what went wrong. Only the first author is MIA.
Ideas why it’s not done:
1) the researcher has not been taught properly, and nobody else in the lab knows enough to spot the improper technique,
2) the researcher thinks only the relative values are important, not the actual values,
3) the researcher is hoping for high values, and failing to subtract the background is one way to get high values.
Everybody is guilty.
My admiration for Diamandis is very great. He previously criticized the EPCA-2 paper. Few are so brave.
It leads me to ask, should reviewers demand that some effort be placed on characterizing whatever it is that is being detected? (Am I right in thinking that it is essentially impossible for anyone else to confirm the findings regarding EPCA-2 or these CCSA markers cause we have no idea what they are?)
-Rork Kuick
“It leads me to ask, should reviewers demand that some effort be placed on characterizing whatever it is that is being detected?”
Several PhD projects.
It would become pathological.