Retract… or correct… or retract…

(June 10th, 2015, updated June 12th, 2015)

 An Oncogene paper was found to contain manipulated data. Four years later, the different players are still fighting about the appropriate way to handle it. Coincidentally, one of them is a former postdoc of ETH Zürich plant biologist Olivier Voinnet – and thus also deeply involved in the current misconduct investigations of his papers.


The affair surrounding potential image manipulations in a number of articles by ETH Zürich plant biologist Olivier Voinnet (Lab Times online reported here and here) has started drawing even wider circles. Charles-Henri Lecellier, a former postdoc of Voinnet, has been associated with misconduct and data manipulation in his own publications. Lecellier is second author of a harshly criticised and freshly retracted Voinnet publication in The Plant Cell (Dunoyer et al., The Plant Cell 16 (5): 1235-50) as well as first author on a Science paper (Vol. 308: 557-60), which is also accused of containing potential image manipulations on the publication discussion platform PubPeer.

About ten years ago, after finishing his postdoc in Voinnet’s lab, then in Strasbourg, Lecellier moved as a staff scientist to the Institut de Génétique Moléculaire de Montpellier, which is part of the French Centre national de la recherche scientifique (CNRS) network. Five years later, data published in a 2010 Oncogene-paper by Guillaume Vetter et al. (Vol. 29, 4436-48) came under suspicion of being manipulated. The paper described the role of a certain microRNA in breast cancer progression and had originated from a collaborative research project (funded by the Luxembourg public body, Fonds National de la Recherche, FNR), in which Lecellier participated. Lecellier himself signed the paper as second-to-last author; first author was his close associate Guillaume Vetter, then at the University of Luxembourg. An official investigation into the case, initiated by the University of Luxemburg, finally concluded that Vetter had deliberately manipulated data for the publication. Therefore, a retraction request was sent to the journal, which Oncogene also initially accepted. Up until today, however, nothing has happened in that direction, because Lecellier and the CNRS leadership have, so far, opposed any retraction or correction of the paper, which would directly mention the involvement of intentional misconduct.

How could this happen?

At the time the paper originated, Guillaume Vetter worked as a postdoc in the lab of cell biologist Evelyne Friederich at the University of Luxembourg. Friederich told Lab Times that shortly after the Oncogene paper appeared, she became suspicious of the reliability of certain data published therein and had confronted Vetter with this matter. Vetter claimed to have lost his lab book, where the experiments described in the Oncogene paper were documented, but promised to eventually present the original raw data. Instead, he suddenly disappeared from Luxembourg (without even notifying Friederich) and soon after joined his research partner Lecellier in Montpellier as head of a laboratory of the diagnostics company Theradiag. Vetter, however, left behind some data files on the lab server as well as a CD-ROM with Oncogene-material (some of it later proven to have been manipulated or fabricated) – and the supposedly "lost" lab book that was eventually discovered in the Luxembourg lab.

Friederich, subsequently, requested an extraordinary meeting on good research practice of the Life Science Research Unit (LSRU) at the University of Luxembourg, which took place in October 2011. According to the meeting protocol, LSRU decided to follow the FNR guidelines on suspected misconduct, despite strong opposition from certain senior University academics. The case was thus passed on to Lucienne Blessing, Vice-Rector for Research, who at the time was setting up an Ethics Review Panel (ERP) on behalf of the University’s Board of Governors.

Right after the LSRU meeting, Friederich contacted Oncogene with her concerns and requested a retraction of the problematic Vetter et al paper. The Editor-in-Chief, Douglas Green, agreed to issue a retraction, as was stated by the journal’s editorial office. Therefore, the case was clearly on its way to being closed – but then for whatever reason a decision was made to await the results of the institutional investigation in Luxembourg.

This investigation started with the internal inquiry by the newly-formed ERP, which “viewed the available data and spoke to various people in the group”. Despite several requests, the ERP was never able to speak directly to Guillaume Vetter. Charles-Henri Lecellier, on the other hand, quickly distanced himself and his wife Anna Saumet (second author on the Oncogene publication) from taking any responsibility, by stating in an email from November 2011 that they both had “not done any of the experiments”.

Nevertheless, the ERP found that Vetter had systematically “modified” the results of two RT-PCR gene expression analyses, together with “possibly fabricated data” elsewhere. Given that clear evidence of “manipulation and fabrication”, the ERP suggested “an official investigation including external reviewers”. And this is where things started to become complicated.

After only one visit to Luxembourg in April 2012, the invited Investigation Panel, consisting of three external experts, had already made a decision. Two months later it concluded: “[…] the paperwork and the interviews leave very little if any doubt that the manipulation of the data […] was done by Dr. Vetter without the involvement of others. In fact he avoided answering a number of critical questions about the data from his collaborators”. The Investigation Panel therefore recommended that “the article published in Oncogene has to be retracted…” but, surprisingly, added “…or corrected”. According to the Panel, the latter should be conducted by “removing the manipulated data and adding the new data”. Furthermore, “it should be made clear why the corrections are necessary and that it is an isolated case of data manipulation by Dr. Vetter”.

At this point, it is noteworthy that this kind of recommendation clearly is in conflict with the guidelines of the Committee on Publication Ethics (COPE), which advise corrections only as appropriate in small cases of misleading sections and honest error. In cases of clear evidence of misconduct, however, COPE refers to a retraction as the only option.

In his email, however, Lecellier refused to “support a correction of the paper” – at least a correction along the recommendations of the external commission. His views were shared by his wife Anna Saumet as well as by another co-author and Montpellier team leader, Charles Theillet. On June 12th 2013, Catherine Jessus, Head of the Institut des sciences biologiques (INSB) in Paris (and currently in charge of the CNRS investigation into the “Voinnet case”), expressed towards Friederich and the new Vice-Rector for Research at the University of Luxembourg, Ludwig Neyses, her “proactive approach” by insisting that “a compromise formulation may be obtained on the basis of a correction of the article, and not its retraction”. Jessus moreover described the communications between the University of Luxembourg and Lecellier as “reactive” and offered “mediation in the event an agreement cannot be achieved”.

Lecellier was, apparently, less cooperative than Jessus expected. He demanded changes in the designated correction notice, which, ultimately, would have completely removed every indication of Vetter’s data manipulations. Moreover, Lecellier wanted the sentence to be added: “Overall the conclusions of the paper are valid but the authors feel appropriate to provide these new results for sake of fairness and objectivity”.

In October 2014, Oncogene received another request of retraction, signed by Vice-Rector Neyses as well as by Rolf Tarrach (then President of the University Luxembourg), Paul Heuschling (Dean of Faculty of Sciences) and lris Behrmann (Head of LSRU). Possibly, the CNRS has still been proactively mediating on the matter, as nothing has happened since. Finally, on June 8th 2015, Friederich herself sent another email request to retract the paper to the Editor-in-Chief of Oncogene.

The fact that data in the paper was manipulated has been multiply confirmed for years now. How long does Oncogene want to continue fooling its readers with the unchanged paper?

Leonid Schneider

(A couple of documents underlying this article can be viewed at Figshare.)


UPDATE (June 12th, 2015)

Lab Times has received the following statement on the matter from Lucinda Haines, Publishing Manager of Oncogene at Nature Publishing Group (dated June 12th, 2015):

"Following the request from Dr. Friederich the retraction will be published asap."

Last Changes: 07.23.2015