One of the issues we’ve touched on at Retraction Watch is what happens once papers are retracted. A few studies have found that other authors continue to cite those studies anyway, without noting their withdrawal from the literature. A more recent paper found that retractions are linked to a dramatic decline in citations (see last half of post). And we’ve reported on one case in which the authors of a retracted study decided not to cite it at all when they republished their findings elsewhere.
But it seems unusual for an author to cite his or her own retracted work without noting it had been retracted. That’s what happened in a recently published PLoS ONE paper, “Loss of Secreted Frizzled-Related Protein 4 Correlates with an Aggressive Phenotype and Predicts Poor Outcome in Ovarian Cancer Patients.” The second to last paragraph of that paper ends:
We and others have recently shown that Foxy-5 peptide, a six amino acid fragment based on Wnt-5a, with subsequent modifications has anti-metastatic properties [46].
Reference 46 reads:
Ford CE, Ekstrom EJ, Andersson T (2009) Wnt-5a signaling restores tamoxifen sensitivity in estrogen receptor-negative breast cancer cells. Proc Natl Acad Sci U S A 106: 3919–3924.
That PNAS paper, it turns out, is one that was retracted in late 2010. When we covered that retraction, we noted that the notice said that “the first author of the paper takes full responsibility for these inaccuracies.”
That lead author was Caroline Ford, who is now at the University of New South Wales, Sydney, Australia, and one of the co-authors of the new PLoS ONE paper. We pointed the unusual citation out to the PLoS ONE corresponding author, and she directed us to Ford. Ford told us late last week:
Yes, this reference is incorrect, and thank you for drawing it to our attention.
The correct reference should be
Säfholm A, Tuomela J, Rosenkvist J, Dejmek J, Härkönen P, Andersson T.
The Wnt-5a-derived hexapeptide Foxy-5 inhibits breast cancer metastasis in vivo by targeting cell motility. Clin Cancer Res. 2008 Oct 15;14(20):6556-63.
We will contact PLoS One immediately to amend the reference list.
The replacement reference includes Ford’s former boss at Lund University in Sweden, Tommy Andersson, who is now chief scientific officer of WntResearch, a company founded to commercialize the findings. As we reported, that company’s initial public offering (IPO) was delayed because of the retraction, news of which we thought they handled in a very transparent way. In fact, “the writer who took full responsibility for these mistakes” gave back $240,000 worth of shares, although the company wouldn’t confirm that the writer in question was Ford.
Meanwhile, WntResearch seems to have done quite well. According to a recent company announcement, the Foxy-5 project “has now been earmarked for funding from the [€400 million] Eurostars program.” The company didn’t know at the time of that announcement exactly how much it would receive, but the decision means the project will receive support from now through the first phase 2 study. You can follow WntResearch’s stock price here.
There’s a plausible enough explanation. I’d guess that in the original draft manuscript, that sentence “We and others…” had two references – one for the “We” (Ford et al 2009) and one for the “others” (Säfholm et al 2008). When the Ford et al paper was retracted, the follow-up MS was altered accordingly – but they mistakenly deleted the wrong reference from the bibliography.
If I’m right, it implies that the first draft of the paper was written before the retraction. The timeline is a little bit odd because the paper was submitted to PLoS One in August 2011 and the retraction was several months earlier in late 2010, but it’s not inconceivable, particularly if the latest paper was submitted somewhere else first and PLoS One was the second-choice journal for publication.
@peter ellis: it could be as you suggest but, all you’ve really pointed out is that in fact not only did the authors of this paper perhaps remove the wrong reference in error but that simply replacing it with the Safholm reference with still leave an error in the article, as the Safholm paper surly does not include a ‘we’ -there are no shared authors between that paper and this PlosOne paper!
Perhaps more important than the plausability of the mis-citation (I don’t agree with Dr Ellis)
RE “We will contact PLoS One immediately to amend the reference list.”
No correction, remains cited – this should have been followed up
RE the wider wnt story – well their approach is interesting but they face huge challenges in delivery