A team of researchers in Ireland has retracted two papers from Cancer Letters after concerns were apparently raised about some of the studies’ figures.
Denise Egan, of the Institute of Technology Tallaght in Dublin, and colleagues published “In vitro anti-tumour and cyto-selective effects of coumarin-3-carboxylic acid and three of its hydroxylated derivatives, along with their silver-based complexes, using human epithelial carcinoma cell lines” and “A study of the role of apoptotic cell death and cell cycle events mediating the mechanism of action of 6-hydroxycoumarin-3-carboxylatosilver in human malignant hepatic cells” in 2007.
The two notices say the same thing:
This article has been retracted: please see Elsevier Policy on Article Withdrawal (http://www.elsevier.com/locate/withdrawalpolicy).
This article has been retracted at the request of the Authors.
The Authors are unable to guarantee the accuracy of some of the figures in the paper, specifically in the presentation of some of the cellular morphologies and electrophoretic data.
The coumarin-3-carboxylic acid paper has been cited 53 times, according to Thomson Scientific’s Web of Knowledge, while the 6-hydroxycoumarin-3-carboxylatosilver paper has been cited 14.
We’ve contacted Egan and the editor of the journal, and will update with anything we learn.
“The Authors are unable to guarantee the accuracy of some of the figures in the paper” – it must have taken hours to come up with this euphemistic statement. Come on, the authors must know the figures are not legit. Just throw one of the co-authors under the bus and be done with it.
Understood, but bottom line is that they retracted it, readers are no longer exposed to an inaccurate piece of work. It’s up to the institution to sort out if there was misconduct.
Egan and Thati (the first author of the two retracted Cancer Letters articles) have seven published paper where they are coauthors. Let’s have look (in chronological order)!
The two Cancer Letters do not seem to have any obvious issues (but they have images in common with other papers, see below). It is peculiar that they are the ones that got retracted, since they were the first published studies with the original images that were subsequently copied and pasted to subsequent papers.
The next paper (Toxicol In Vitro. 2007 Aug;21(5):801-8. Epub 2007 Feb 3. PMID: 17350222) doesn’t have any obvious issues.
The paper published in Chem Biol Interact. (2007 Jun 30;168(2):143-58. Epub 2007 Apr 19. PMID: 17512508) share some images with one of the retracted papers (Cancer Lett. 2007 May 18;250(1):128-39. Epub 2006 Nov 28.). Figure 8 of the Chem Biol Interact paper have several images that are identical to Figure 4 of the Cancer Lett paper (e.g. 8b 24h is identical to 4c 24h, slightly zoomed in version). The control plot of Figure 12a of the Chem Biol Interact paper is identical the control plot of Figure 8a of the Cancer Lett paper.
J Inorg Biochem. 2007 Aug;101(8):1108-19. Epub 2007 May 4. PMID: 17555821 doesn’t have any obvious issues.
Eur J Pharmacol. 2007 Aug 13;569(1-2):16-28. Epub 2007 May 21. PMID: 17585902 does hare some images with other papers. Figure 5 of the Eur J Pharmacol paper share the five first lanes with Figure 6 of the Chem Biol Interact paper. Figure 6a 24h (Eur J Pharmacol) is identical to Figure 7a 4h (Chem Biol Interact). Figure 7 of the Eur J Pharamacol paper share images with Figure 8 of the Chem Biol Interact paper and Figure 4 of the Cancer Letter paper (2007 May 18;250(1):128). Lane 2 of Figure 8 (Eur J Pharmacol) is identical to lane 3 of Figure 9 (Chem Biol Interact). Actin blot of Figure 10 (Eur J Pharmacol) is identical to Actin blot of Figure 11 (Chem Biol Interact). Lane 2 and control of the cleaved Parp blot of Figure 10 (Eur J Pharmacol) is identical to lane 1 and lane 2 of the cleaved Parp blot of Figure 11 (Chem Biol Interact).
All lanes of Figure 6 of the second Eur J Pharmacol paper (Eur J Pharmacol. 2009 Jan 14;602(2-3):203-14. doi: 10.1016/j.ejphar.2008.11.020. Epub 2008 Nov 19. PMID: 19041861) can also be found in Figure 7 of the first Cancer Lett paper (Cancer Lett. 2007 Apr 18;248(2):321-31). Figure 8 (Eur J Pharamcol 2009) share images with Figure 7 of the first Eur J Pharmacol paper (2007), Figure 8 of the Chem Biol Interact paper and Figure 4 of the Cancer Letter paper (2007 May 18;250(1):128). Look at for example at A 24h. Figure 9 (Eur J Pharmacol 2009) share some lane with Figure 5 of the second Cancer Lett (2008 May18;250(1):128) paper. Figure 11 (Eur J Pharmacol 2009) and Figure 7 (Cancer Lett (2008 May18;250(1):128)) are identical. Figure 7 (Eur J Pharmacol (2009)) share images with Figure 3 (Cancer Lett (2008 May18;250(1):128)).
Junk S – is this reason for further investigation or just a few unfortunate errors?
There is a reason for further investigation of all papers in my opinion. I had a quick look at other papers where the first author were not involved and interestingly the reuse issues are still there. E.g. Figure 6 (Chem Biol Interact. 2006 Dec 1;164(1-2):115-25. Epub 2006 Sep 16. PMID: 17030029) is identical to Figure 5 (Cancer Lett. 2007 Mar 18;247(2):224-33. Epub 2006 Jun 5. PMID: 16740357). Please look at lane 4 in both figures, which have had the white dots removed from one of the identical gel images to make it less obvious. Clerical error, not even close.
you meant to say that it is not only the first author issue – laboratory practise?
Right you are, JunkScience. Thanks for reporting this.
Curiously, the 2006 paper (PMID: 17030029 ) cites the 2007 paper (PMID: 16740357) as a previous study (2006 paper pages 119-120: “Phen and its copper complexes have previously been shown to exert a range of biological activities, such as anti-tumour [31–33], . . .” Ref 33 is the 2007 paper)
Perhaps a savvy biochemist can chime in here – the figures 5 and 6 are indeed the same gel, but the names of the compounds involved in lanes 6 – 8 are described differently, though the differences will no doubt only be appreciated by specialists in this field.
2006 paper: Fig. 6: . . . lanes 6–8 (pGEM-3Z DNA and [Cu(phendione)3](ClO4)2*4H2O at 1, 10, and 200 uM)
2007 paper: Fig. 5: . . . lanes 6–8 [Cu(phen)2(mal)]*2H2O (pGEM-3Z DNA and [Cu(phen)2(mal)]*2H2O at 1, 10, and 200 uM)
If the compounds associated with lanes 6 – 8 as stated in the papers are identical, then this is image and data re-use in two papers, and though the 2006 paper cites the 2007 paper, the image is reused where the 2006 paper should have just referred to figure 5 of the 2007 paper.
If the compounds as stated are different, then this is image re-use with different experimental conditions claimed.
Either way this is not proper presentation of results and evidence. JunkScience is right, this should all be investigated further.
I also got confused. Everything looks so alike, I believed for some time that all papers were reviews of just one important paper and they could therefore freely reuse images.
In reply to Junk Science April 27, 2013 at 2:54 am
You have done a lot of work bringing reuse of images to light. I am impressed. Going though quite a few papers makes you tired, which adds to the confusion.
fernando: you are very active on this board as well, so keep up the good work!
Figure 6 in this paper:
Finn GJ, Creaven BS, Egan DA. Cancer Lett. 2004 Mar 8;205(1):69-79. PMID: 15036663
is identical to Figure 5 in this paper:
Finn GJ, Creaven BS, Egan DA. Biochem Pharmacol. 2004 May 1;67(9):1779-88. PMID: 15081877