Retraction Watch

Tracking retractions as a window into the scientific process

Journal to retract paper called “anti-vaccine pseudoscience”

with 5 comments

A journal is planning to retract a paper that purported to link a component of vaccines to autism in mice.

The paper, about the effects of aluminum adjuvants in vaccines on the immune response in the brains of mice, is the second retraction for co-authors Christopher Shaw and Lucija Tomljenovic, of the University of British Columbia. The journal’s editor told us he and the authors are jointly retracting the paper. 

Just over a month old, the paper has already received plenty of criticism. Numerous commenters on PubPeer have allegedly identified image duplications and other problems with the paper. One commenter described “clear and deliberate” removal of control results in the paper, while others suggested gel bands were duplicated within the paper, and appear similar to those from another paper published in 2014 by Shaw and Tomljenovic. In a blog post, David Gorski, a professor and surgeon at Wayne State University, called the paper “antivaccine pseudoscience.”

Shaw, the paper’s last author, told us that his lab became aware of the PubPeer discussion a few weeks after publication: 

We immediately did our own analysis. Indeed, some images have been altered. How that happened, we don’t know.

Subcutaneous injections of aluminum at vaccine adjuvant levels activate innate immune genes in mouse brain that are homologous with biomarkers of autism” was originally published Sept. 5 in the Journal of Inorganic Biochemistry. 

The journal’s editor, John Dawson of the University of South Carolina, told Retraction Watch:

The paper by Shaw and co-workers is being retracted jointly by the authors and the editor.

He noted there will be a “statement accompanying the retraction of the paper.”

Shaw told us that his lab began investigating the issues raised on PubPeer “within a day” and reported its findings to both UBC and the journal soon after. He said:

Our own analysis showed some figures had been altered. We requested a retraction because we could not understand how that had happened. We felt the data had been compromised.

Shaw said that the problems mostly lie with data showing no change in gene or protein expression levels after aluminum injections — but also with some data showing changes in expression, which the paper attributed to the injections. 

In February 2016, the journal Vaccine temporarily removed — and then retracted — a paper from Shaw and Tomljenovic which also raised concerns about the side effects of vaccines. The paper was then re-published in July 2016 in Immunologic Research.

About the Journal of Inorganic Biochemistry paper, Gorski wrote:

given Shaw and Tomljenovic’s history, it is not unreasonable to be suspicious of this study as well…

At best, what we have here are researchers with little or no expertise in very basic molecular biology techniques using old methodology that isn’t very accurate overinterpreting the differences in gene and protein levels that they found. At worst, what we have are antivaccine “researchers” who are not out for scientific accuracy but who actually want to promote the idea that vaccines cause autism….If this were a first offense, I’d give Shaw and Tomljenovic the benefit of the doubt, but this is far from their first offense.

Shaw said that first author Dan Li, a former postdoc who performed the molecular biology and gene expression analysis for the study, has agreed to the retraction but not yet offered an official explanation about the data. Shaw told us:

She denied that anything had been manipulated, or that anything was amiss.

He added that when Li left the lab in 2015, she took the original data with her:

UBC policy is that original data never leave the lab. We’ve asked for them to be returned to us. 

Shaw said he thinks the core data are “probably correct,” but said he plans to have the experiments re-done:

It is what it is. We’ve done everything we can on our end. We’re still having conversations with Li on where the data are and how we get them back. That’s as much as we can do at this point.

Like Retraction Watch? Consider making a tax-deductible contribution to support our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up on our homepage for an email every time there’s a new post, or subscribe to our daily digest. Click here to review our Comments Policy. For a sneak peek at what we’re working on, click here.

Comments
  • herr doktor bimler October 9, 2017 at 2:29 pm

    appear similar to those from another paper published in 2014 by Shaw and Tomljenovic

    Two other papers in 2014. Fig. 1 also appears in Shaw, Li & Tomljenovic, Immunotherapy.

    when [Dan] Li left the lab in 2015, she took the original data with her

    Then who assembled the diagrams? How can you submit a paper when you no longer have the data?
    It may be that the present paper is a version of “Gene-toxin synergy in the brain of autistic mouse model”, a 2015 presentation by Li, Li and Shaw to the 11th Keele Meeting on Aluminum (page 41 in the Abstract Book).

  • herr doktor bimler October 9, 2017 at 4:09 pm

    Our own analysis showed some figures had been altered.

    Shaw signed off on those compromised Figures, and presumably looked at them, before submitting the paper in Jan. 2017.
    But the issues weren’t purely data-related. Pubpeer contributors note that the first installment of data from these long-deceased mice came out in 2013 —
    https://pubpeer.com/publications/4AEB7C8F30015079E2611157CF8983#24

    but the 2013 explanation of the mouse pups’ treatment is different from the 2017 account.
    https://pubpeer.com/publications/4AEB7C8F30015079E2611157CF8983#29

    Were the mice euthanased at age 16 weeks or 34 weeks?
    https://pubpeer.com/publications/4AEB7C8F30015079E2611157CF8983#30

  • herr doktor bimler October 10, 2017 at 7:17 pm

    Some of the discussion at Pubpeer addressed the concern that one band in Figure 1C, for the level of protein TNF-A in the third Control Mouse (i.e. the 5th band in that slice of gel), looked suspiciously blank — a completely empty rectangle:
    https://pubpeer.com/storage/image-1505914692638.jpg

    Either it was blanked with image-processing tools (not good), or the rectangular appearance is an artifact of a high level of image compression, and the band was naturally blank.
    But that’s not good either, as the statistics were calculated on “Treated / Control pair” ratios, and in this case, with Control = 0, the ratio is necessarily infinity. So the authors calculated the mean and standard error for a distribution of three ratios, one of those ratios being infinite, and somehow ended up with mean = 3 for TNF-A, std.error = 1.
    https://pubpeer.com/publications/0546BC914E697A4DFB579214F68B4D#5

  • Ian October 16, 2017 at 10:46 am
  • Brian Deer October 17, 2017 at 6:27 am

    I notice from the cbc report that the pair of them both make statements inconsistent with authorship.

    The sequence appears to be that they publish papers, these papers are then found to be defective at face value, and they then deny any role in those papers that would entitle them to have claimed authorship in the first place.

    Clearly a time for institutional investigation into these individuals, irrespective of where the data is claimed to be, or who was responsible.

    http://www.cbc.ca/news/canada/british-columbia/ubc-autsism-vaccine-paper-retraction-chris-shaw-1.4351855

  • Post a comment

    Threaded commenting powered by interconnect/it code.