When journals learn papers are problematic, how long does it take them to act?
We recently had a chance to find out as part of our continuing coverage of the case of Anil Jaiswal at the University of Maryland, who’s retracted 15 papers (including two new ones we recently identified), and has transitioned out of cancer research. Here’s what happened.
As part of a public records request related to the investigation, we received letters that the University of Maryland sent to 11 journals regarding 26 “compromised” papers co-authored by Jaiswal, four of which had been retracted by the time of the letter. The letters were dated between August and September 2016 (and one in February) — although, in some cases, the journals told us they received the letter later. Since that date, three journals have retracted nine papers and corrected another, waiting between four and six months to take action. One journal published an editorial note of concern within approximately two months after the university letter.
And six journals have not taken any public action.
Many of the most recent retractions and corrections appear to be a direct result of the university’s letters, which explain that the university conducted an investigation in which it uncovered a “compromised” paper or papers on which Jaiswal is last author. The letters then outline why the article or articles in question deserve a retraction or correction, with reasons ranging from figure duplication and manipulation to unreliable data.
In most cases, the university’s chief accountability officer Roger Ward asked the journal to retract an article. But for four papers, he told the journal to use its judgment about whether to issue a retraction or erratum. (In one case, the university sent a letter to a journal providing an update to an existing retraction, but did not request any action. This journal is not included below.)
Here are more details from each journal.
Journal of Cell Science
On August 24, the university dated a letter to Journal of Cell Science, asking the journal to retract two papers. In mid-February, the journal retracted both articles, along with a third paper by Jaiswal that was not mentioned in the letter. Time from date of letter to retraction: Almost six months.
In a letter dated September 29 (which revised an initial letter dated in August 24), the university asked Oncogene to retract two papers. On March 6, the journal did. We recently spotted these notices, upping Jaiswal’s previous count of 13 to 15.
Here is the notice for “Stress-induced NQO1 controls stability of C/EBPα against 20S proteasomal degradation to regulate p63 expression with implications in protection against chemical-induced skin cancer,” published in 2012 and cited eight times, according to Clarivate Analytics’ Web of Science, formerly part of Thomson Reuters:
Following an internal investigation by the University of Maryland, Baltimore, the Editors and Publisher have agreed to retract this paper owing to the results of the investigation concluding that Figure 2 does not support the claim of presenting new data from keratinocytes.
And here’s the second notice, which cites band duplication:
Following an internal investigation by the University of Maryland, Baltimore, the Editors and Publisher have agreed to retract this paper due to duplication of bands used to depict varying NQO1 interactions in Figure 4.
“Disruption of NAD(P)H:quinone oxidoreductase 1 gene in mice leads to 20S proteasomal degradation of p63 resulting in thinning of epithelium and chemical-induced skin cancer,” published in 2011, has been cited 17 times.
Time to retraction: Just over five months.
Journal of Biological Chemistry
The university sent a letter dated September 26 to The Journal of Biological Chemistry (JBC), requesting retractions of four papers and retractions or corrections to two others. Three of these papers had already been flagged with an expression of concern. The letter also provided updates to two papers that JBC had retracted in April 2014 and in August 2014, but did not ask the journal to take further action. JBC issued retractions to five of the papers and an erratum to one; according to a spokesperson, the notices were issued in
early February January, but are stamped early February.
Time to take action:
Just over four Three months.
In a letter dated August 24, the university requested retractions of five papers in Cancer Research. The university also provided an update to a paper that the journal had retracted in 2015. The five papers are:
- “Inactivation of the quinone oxidoreductases NQO1 and NQO2 strongly elevates the incidence and multiplicity of chemically induced skin tumors,” cited 27 times.
- “Disruption of NAD(P)H:quinone oxidoreductase 1 gene in mice leads to radiation-induced myeloproliferative disease,” cited 12 times.
- “Low and high dose UVB regulation of transcription factor NF-E2-related factor 2,” cited 34 times.
- “Lower induction of p53 and decreased apoptosis in NQO1-null mice lead to increased sensitivity to chemical-induced skin carcinogenesis,” cited 56 times.
- “Disruption of the NAD(P)H:quinone oxidoreductase 1 (NQO1) gene in mice causes myelogenous hyperplasia,” cited 99 times.
Time to act: Eight months and counting.
Biochemical and Biophysical Research Communications
In a letter dated August 24, the university requested a retraction of one paper in BBRC. That paper is:
- “Overlapping signal sequences control nuclear localization and endoplasmic reticulum retention of GRP58,” cited six times.
Time to retraction: Eight months and counting.
In a letter dated August 24, the university requested Chemico-Biological Interactions retract the 2006 paper, “Si RNA inhibition of GRP58 associated with decrease in mitomycin C-induced DNA cross-linking and cytotoxicity,” which has been cited six times. Editor-in-Chief Daniel Dietrich told us he received the request in September, and the initial letter was “not sufficient to automatically invoke a retraction,” so he contacted the university requesting more information. He asked for:
… an explanation [of] how the “malpractice” or [“]misrepresentation/adulteration” was determined in a fair and transparent way had to be sent to us, prior to CBI initiating retraction procedures.
Dietrich did not specify when he asked the journal for more information. But on March 6, he got what he needed from the University of Maryland at Baltimore’s research integrity officer, and sent it to the publisher on March 8, 2017:
I immediately forwarded this information to Elsevier, who then submitted this information to their Panel. I have also notified the authors of the impending retraction, should they not be able to dispel the accusations brought forth by the University of Maryland. The authors usually have 48 hrs to respond. Once the response is in (failure to respond is taken as admission to the accusations of the University of Maryland), this will be discussed. If the authors can provide evidence that the accusations brought forth have no basis, we will send the response back to the University of Maryland.
If the journal and publisher found the authors’ explanation unconvincing, they would move forward with a retraction, he added.
Dietrich acknowledged that this process may seem slow, but he said it’s all in an effort to ensure that decisions are not made “in haste”:
Although we may appear as slow or reluctant to act, we must ensure that we do not take decisions in haste and unfairly to either the author(s) in question nor to the scientific audience at large.
Time to action: Seven months and counting.
Molecular Cancer Therapeutics
In a letter dated August 24, the university requested a retraction of one paper in Molecular Cancer Therapeutics. That paper is:
- “Aromatase Inhibitor-Mediated Downregulation of INrf2 (Keap1) Leads to Increased Nrf2 and Resistance in Breast Cancer,” cited five times.
Time to retraction: Eight months and counting.
Free Radical Biology & Medicine
In a letter dated August 24, the university asked Free Radical Biology & Medicine to issue a retraction or erratum to one paper, and described problems with a second paper without specifically asking the journal to act. That second paper, “NAD(P)H:quinone oxidoreductase 1 protects bladder epithelium against painful bladder syndrome in mice,” had already received two corrections, which the letter did not acknowledge. The initial erratum for the 2012 paper occurred in 2013 for a figure-related error, and the second was published in 2014, citing an “unintentional” figure error. The letter mentions issues with one of the figures called out in the 2014 erratum, stating that “there is no valid explanation for the absence of heavy or light chain signal in the Western blots.”
The first paper is “Nrf2-induced antiapoptotic Bcl-xL protein enhances cell survival and drug resistance,” cited 49 times.
Time to issue editorial notice on remaining paper: Eight months and counting.
Clinical Cancer Research
In a letter dated February 21, the university asked Clinical Cancer Research to issue an erratum or a retraction to one paper.
- “NRH:quinone oxidoreductase 2-deficient mice are highly susceptible to radiation-induced B-cell lymphomas,” cited eight times.
Time to action: Two months and counting.
The journal responded to the university’s August 24 letter with an editorial notice on November 5, instead of a retraction or correction. The paper in question, “Human Antioxidant-Response-Element-Mediated Regulation of Type 1 NAD(P)H:quinone Oxidoreductase Gene Expression,” was published in the European Journal of Biochemistry (now The FEBS Journal) more than two decades ago (in 1994), and has been cited 75 times. The notice explains:
This Editorial Note is intended to alert our readers to the fact that questions have been raised regarding a figure in the above-mentioned manuscript . It is unfortunate that we cannot state with complete certainty that the data in Figure 5 are accurate representations of the described experiments. However, we are satisfied that the key conclusions of the manuscript would not be affected by a potential error in this figure and thus will not be pursuing further action at this time.
We asked editor-in-chief Seamus Martin why the journal opted not to issue an erratum or retraction:
We issued an Editorial Note, instead of retracting the paper, because the data under question were limited to one figure (Figure 5), which appeared to contain a duplication of a CAT assay slot blot in both panels of that figure. As members of [Committee on Publication Ethics], we followed their guidelines on retraction vs. correction; notably, that an instance in which “a small portion of an otherwise reliable publication proves to be misleading (especially because of honest error)” warrants a correction.
I would note that the overall conclusions of the manuscript are not in dispute. Duplication of the same piece of data within the same figure (i.e. panels A and B of Figure 5 appear to be largely identical) looks, on the face of it, to be more likely due to sloppy data handling or poor figure preparation rather than wil[l]ful data manipulation. Thus, in our view, the COPE guidelines did not support a retraction and we proceeded accordingly.
We are aware of the ongoing investigation into Dr Jaiwsal’s work and, as mentioned in the Editorial Note, regret that we cannot state that the data in Figure 5 are reliable. If additional evidence comes to light that casts doubt on other figures in this paper, or the overall conclusions of the study, we are likely to take further action.
We’ve contacted the journals who have yet to take action to ask if and when they are planning to do so.
Sarah N. Archibald, research integrity officer at University of Maryland, told us she does not believe the university has held up the process:
The University completed its investigation process and then sent letters to journals. To the best of my knowledge, the University has responded to all requests for information made by the journals.
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.