Neurology has partially retracted a 2016 paper, replacing a figure and removing the author who contributed it after he was found guilty of misconduct.
The journal has replaced the figure with a new one that confirmed the findings of the original, and swapped the name of Andrew Cullinane with the scientist who constructed the new figure using a new dataset. Last year, the U.S. Office of Research Integrity declared that Cullinane had falsified data in this paper and one other while working as a postdoctoral fellow in the Medical Genetics Branch at the National Human Genome Research Institute (NHGRI).
Cullinane appears to be at Howard University in Washington D.C., according to his LinkedIn page. He is listed as an assistant professor in the Basic Sciences/Anatomy department of the university’s College of Medicine.
Here’s the partial retraction notice from the journal:
The Editors of Neurology® have investigated the article “Neurologic involvement in patients with atypical Chediak-Higashi disease” by W.J. Introne et al.1 following publication on September 28, 2016, of an Expression of Concern.2 The Editors’ investigation resulted from a notification by the US Office of Research Integrity on September 1, 2016, that an investigation from that office had found that figure 2A, contributed by coauthor Andrew R. Cullinane, had been manipulated and was fraudulent. The Editors were concerned about the implications of the fraudulent figure for the overall data and conclusions of the paper. They requested an explanation from the authors. Based on information provided, the Editors retract figure 2A and remove Dr. Cullinane from the author byline.
This one is a bit confusing, as the journal also issued a correction notice for the paper on the same day as the partial retraction. The correction reads:
In the article “Neurologic involvement in patients with atypical Chediak-Higashi disease” by W.J. Introne et al.,1 figure 2A was found to have fraudulent content and has been retracted,2 and Andrew R. Cullinane, the coauthor responsible for the figure, has been removed from the byline. A corrected figure, generated from use of a different antibody using data from prior experiments, is below, and confirms the findings of figure 2A, although it shows much lower expression of CHD-5 (consistent with cDNA expression). May Christine V. Malicdan, MD, PhD, who was instrumental in preparing the new figure, has been added as an author. An Expression of Concern was published on September 28, 2016,3 and the article has been corrected and republished.4 The authors regret the errors.
We don’t see many partial retractions — in fact, MEDLINE recently stopped identifying partial retractions, and will instead refer to them as errata. And the Committee on Publication Ethics (COPE) has said:
Partial retractions are not helpful because they make it difficult for readers to determine the status of the article and which parts may be relied upon.
Patricia Baskin, the executive editor of the Neurology journals, told us the paper has been republished with the replacement figure. She explained why the journal decided to publish both a correction and partial retraction:
The correction was meant to detail the correction and the partial retraction to indicate on our own website (where it will still be called partial retraction even if PubMed changes it to a correction) to indicate to our readers that fraud was involved.
Baskin explained why the journal needed to also call the change a “partial retraction:”
We considered retracting the paper; however, the figure was the only component of the paper that was questionable and the results and conclusions of the paper remained unchanged. We therefore asked the authors for a corrected version of the paper with a replacement figure. We have now published the corrected article with the new figure and replaced Dr. Cullinane in the byline with the new author who constructed the new figure. We also decided to publish a partial retraction for the Neurology® website to explain why the correction was made and that fraud had occurred.
She added that the journal has taken steps to address the concerns raised by COPE — namely, that readers won’t know which parts of the paper were problematic:
We are aware that PubMed no longer recognizes partial retractions, but we believe it is important to post the explanatory information on our own website. Since the paper has now been corrected and republished (the appropriate classification in PubMed), readers will be pointed to the corrected version.
What’s more, she added, the new figure confirms what the original said:
We not only removed the figure, but replaced it with another figure provided by a new author and constructed with a new data set. The conclusions of the paper were not affected as the new figure confirmed the findings in the original figure and both figures (old and new) confirmed the clinical findings in the paper.
She also explained the journal’s decision to remove Cullinane’s name from the paper:
An ORI investigation found that Dr. Cullinane had fabricated data. We decided to remove his name and the figure to ensure the integrity of the corrected version of the paper…
We contacted Cullinane and Howard University; we will update the post if they respond.
The other paper flagged by the ORI’s investigation was published by the American Journal of Human Genetics (AJHG) in 2011. “A BLOC-1 Mutation Screen Reveals that PLDN Is Mutated in Hermansky-Pudlak Syndrome Type 9,” lists Cullinane as the first and corresponding author. We contacted the editor of the journal, David Nelson from Baylor College of Medicine, who told us:
This issue is still in process.
Hat tip: Rolf Degen
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.