Fireworks: Belgian dispute over ovarian transplant findings includes claims of theft, arson

There’s a story brewing in Belgium that is, as one local newspaper put it, worthy of a TV drama.

Here’s our attempt at a summary: Jacques Donnez, chair of Catholic University of Louvain’s (UCL) gynecology department, and colleagues published two studies in Human Reproduction in 2010. One study claimed to show that a woman had given birth after undergoing chemotherapy for severe sickle cell disease and then getting an ovarian transplant from her sister.

The other of those studies, the authors noted, confirmed “data published earlier as a case report” in 2007. That case report of a woman with another type of anemia — the pregnancy only went as far as an embryo, which did not survive — garnered a good deal of attention, because, as New Scientist reported:

Crucially, the woman received the graft from a sibling who was not her identical twin, the first time an embryo has been produced from such a match.

But as Le Soir reported last week,  there have been some questions about whether the woman with sickle cell disease who successfully gave birth was really unable to produce eggs before the procedure — suggesting Donnez’s group’s feat wasn’t all it was cracked up to be. It turns out there had been a UCL investigation into the work, and a committee recommended that Donnez be fired.

He wasn’t. UCL ignored the advice of its committee, and in fact appears to have said Donnez was in the clear. The papers were corrected, but not retracted. And now Donnez is threatening to sue the whistleblower who brought the case to UCL’s attention.

According to La Libre, Donnez’s attorney said last week at a press conference that one of his clients’ computers had been stolen — a fact that the lawyer said was relevant because data on the computer was used to file a complaint about Donnez’s research ethics, among other things. But things got worse, according to the lawyer (Google Translated from French):

On 26 November 2010, the laboratory of Professor Donnez, located near the Cliniques Saint-Luc, was burned. The expert concluded that the prosecution arson, the origin being clearly voluntary. Authors were never identified. Even if we have suspicions, the investigation ended with an acquittal. With this fire, this is not only its research lab was destroyed, but also years of work have been wiped out.

Fire and theft would, of course, make it difficult to come up with data to corroborate what was in the papers that were published earlier that year.

It turns out that there was an exchange of letters about the 2010 and 2011 papers in Human Reproduction in May, in an eye-opening example of what happens when academic intrigue is translated into academic jargon. The letters between Donnez’s team, his now ex-co-authors Yvon Englert and colleagues,, and the journal’s editor, Andre Van Steirteghem, are more subtle, but tell basically the same story as the Belgian papers — sans fire and theft.

Donnez et al:

We wish to draw to the attention of the journal’s readership, two points that have recently come to light.

On point one, the evidence for activity in the woman’s allegedly “dead” ovary, they write:

This presence of ovarian follicles does not change the clinical diagnosis, the indications for surgery or the cause of the restoration of ovarian activity. Since the birth, the sister donor has been identified as one genetic parent of the child.

In response to the concerns about IRB approval, they say they thought they had it:

Secondly, in both papers we stated ‘Our protocol was authorized by the Ethics Committee of the Catholic University of Louvain which, back in 1995, had approved such research protocols, including reimplantation of ovarian tissue to preserve or restore fertility in women treated with high doses of chemotherapy, which could induceovarian failure’.

We considered that this approval back in 1995 covered both auto- and allografting. However, after recent discussions within the University and to avoid any potential misinterpretation in the future, our group has agreed that from here on, we will ensure that specific agreement of the Ethical Committee is sought and obtained, as required by the national and international regulations.

Donnez’s explanation didn’t convince three of his co-authors, who asked to be taken off the paper to which they contributed. In a letter requesting that removal, Englert et al raise their eyebrows at the apparent reversal by UCL:

In May 2011, we were informed by the Vice-rector of the Catholic University of Louvain that an internal procedure against our colleague Professor Jacques Donnez had driven the University to conclude that an article (‘Livebirth after allografting of ovarian cortex between genetically non-identical sisters’ by Donnez et al., 2011), for which I and two members of my research staff are co-authors, was considered to raise significant problems because ‘it contains statements that are not supported by documented evidence’. We understand that the letter containing this statement has subsequently been withdrawn by Universite´ Catholique de Louvain (UCL), considering that Professor Jacques Donnez acted ‘in good faith’ and that ‘there is no proven intention of scientific fraud’.

Since then, many actions appear to have taken place inside the Catholic University of Louvain, as well as between Jacques Donnez and the journal.

Many actions indeed.

Donnez’s letter didn’t impress his co-authors:

As external co-authors of the paper, we would like to stress the fact:

(i) that we could not verify or even imagine that aspects of the case were omitted by our distinguished colleague and that, in any case, our participation to the case was to perform IVF for tubal infertility in a laboratory specific for high viral risk as a standard procedure in that case;

(ii) that we did not take part in the internal inquiry performed by the UCL authorities and are not able to build a personal opinion on the facts themselves.

The letter proposed by Jacques Donnez was unable to convince us that the situation was entirely clarified. Therefore, we maintain our request from May 2011 to be withdrawn as co-authors of the paper.

The editor sounds somewhere between weary and exasperated in his own letter. You can almost hearing him sighing “Children!”:

As Editor-in-Chief, I received, in May 2011, a letter from the Vice-Rector of the Universite´ Catholique de Louvain (UCL) informing the journal that both articles contained statements that could not be fully supported by experimental evidence. The articles mentioned that no ovarian follicles were present in ovarian biopsy from the patient who had a child after allografting ovarian cortex. A laboratory report on a sample of the same ovary revealed the presence of ovarian follicles: an unexpected finding that was not mentioned in either article. A second concern, regarding the appropriateness of the ethical approval granted for these studies, was also raised.

The UCL authors (Donnez, Squifflet, Pirard, Jadoul, Dolmans and Cheron) now recognize their errors and, in a letter to the Editor in this issue, agree that, firstly the finding of the presence of a number of ovarian follicles should have been mentioned and commented upon, and secondly, that in future specific and appropriate ethics committee approval will be sought and obtained. They go on to elaborate why the presence of ovarian follicles is unlikely to change the clinical diagnosis, the indications for surgery or the cause of restoration of ovarian activity. In an accompanying Letter to the Editor, three co-authors of the second article (Englert, Delbaere and Armenio) explain their reasons for wishing to withdraw their names from the article.

As Human Reproduction is a member of the Committee on Publication Ethics (COPE), the advice of the committee was sought prior to any decisions being made by the journal. As Editor, I must accept the correction of the literature by both sets of authors, but I reiterate the importance that reports must mention all findings—even if they are unexpected and do not compromise the meaning of a study. I also cannot stress highly enough the necessity to obtain appropriate ethical approval.

After much discussion and careful consideration, the journal has decided not to retract either article based on the premise that the data are sound.

We reached Englert, who didn’t have anything to add. We tried to reach Donnez and Van Steirteghem, and will update with anything we hear back.

Please see an update on this post, involving another Donnez et al paper in The Lancet.