Retraction Watch

Tracking retractions as a window into the scientific process

Cell line switch sinks PLoS ONE cancer paper

with 4 comments

plosWe’ve written before about how common cell line mix ups are in cancer research; according to a 2012 Wall Street Journal article (paywalled), between a fifth and a third of cancer cell lines tested by suspicious researchers turned out to be misidentified.

Obviously, mistakenly studying the wrong kind of cancer is a waste of precious resources, both time and money. And it’s clear the problem hasn’t gone away. PLoS ONE just retracted a cancer paper originally published in December 2012 for studying two cell lines that had been contaminated by other cell types.

Here’s the notice for “Epithelial Mesenchymal Transition Is Required for Acquisition of Anoikis Resistance and Metastatic Potential in Adenoid Cystic Carcinoma”:

The authors retract this publication due to concerns about the cell lines employed in the study.

The study reports experiments looking at pathways involved in anoikis resistance in human adenoid cystic carcinoma; the experiments involved the use of the cell lines ACCM and ACC2. After the publication of the article, a reader raised concerns that these two cell lines may not originate from adenoid cystic carcinoma. The authors have completed a short tandem repeat analysis on the cell lines and this has revealed cross-contamination with other human cells, which compromises the relevance of the work to human adenoid cystic carcinoma, and thus the conclusions of the study.

In the light of this, the authors retract this publication.

PLOS ONE deputy editor Iratxe Puebla tells Retraction Watch:

The concerns about the article were raised by a reader who posted a comment noting that the ACC2 and ACCM cell lines employed in the study had been determined to be HeLa cells, the comment is available here: We followed up with the authors to request information regarding the characterization of the cell lines employed in the study. The authors submitted the cells for STR analysis, which revealed cross contamination.

The paper has been cited seven times, according to Thomson Scientific’s Web of Knowledge.

Hat tip: Rolf Degen

  • Fred Levine October 10, 2014 at 12:36 pm

    You state that “Obviously, mistakenly studying the wrong kind of cancer is a waste of precious resources, both time and money.” However, this is not so obvious in some respects. The fundamental biology that is being studied may not be specific to a particular tumor type. It is increasingly recognized that the old way of classifying tumors by site of origin is obsolete and that we ought to be examining the underlying drivers of oncogenesis, which are more often than not common between cancers arising from different tissues. Thus, the data in the paper may be completely correct and relevant to cancer biology, except for what may be a minor mistake of the cell of origin. As a regular reader of your blog, it is clear that many completely fraudulent papers get by with “corrections”, while this paper, which may report an important phenomenon that is relevant to many tumor types, is being retracted.

    • Erin October 13, 2014 at 2:23 pm

      Fred – I do respect your opinion but I think you are missing the point. The use of misidentified cell lines is pervasive and rampant in the scientific community and even IF, as you point out, the use of this particular type of cancer (ACC) may not be relevant to the overall meaning of this paper (and I would argue against that point if I had time), the use of these misidentified cell lines (ACC2 and ACCM) perpetuates the problem for other researchers whose conclusions ARE based on the use of actual ACC samples because as you know, the data in these papers gets cited many times over by other researchers. I respect and admire these authors for doing the right thing – unfortunately, they are few among MANY.

  • DocMartyn October 10, 2014 at 4:59 pm

    I was talking to a senior person at the ATCC and they are going through their bank and checking on the origin of their deposits. A list of the known ‘wrongun’s’ is here
    So two human GBM’s and an astrocytoma were rat cells.
    A ‘nameless’ group has published for decades examining the effects of hormones on a cervical cell line that appears to be a Melanoma.

    • PJTV October 11, 2014 at 2:07 am

      An impressive list of misidentifications. The number of undiscovered ones could be many more (as indicated by the reference to the Wall Street journal). Does this mean that the scientific world is structurally deficient in weeding out its errors? A disturbing thought.

  • Post a comment

    Threaded commenting powered by interconnect/it code.