The scandal surgeon Paolo Macchiarini, now professor at Federal University of Kazan in Russia, is preparing to experiment on monkeys using plastic oesophagus. He proclaims previous success in rats, yet today’s decision of his former employer, the Karolinska Institutet (KI) in Stockholm declared that very paper as fraudulent and found Macchiarini and his acolyte Philipp Jungebluth guilty of research misconduct. An earlier decision of the Swedish Central Ethical Review Board (CEPN, read decision here) came on September 6^th 2016 to a similar conclusion, with one minor difference. Back then, all senior authors were found “guilty of scientific misconduct”, thus also including Macchiarini’s American collaborator Doris Taylor (who in 2014 almost made him a professor at her University of Texas). Now, all authors except of Macchiarini and Jungebluth have been virtually absolved, while the junior co-authors, the graduate student Sebastian Sjöqvist and the assistant professor Mei Ling Lim received an admonition.

KI now demands the retraction of this 2014 Nature Communications paper:

Sebastian Sjöqvist, Philipp Jungebluth, Mei Ling Lim, Johannes C. Haag, Ylva Gustafsson, Greg Lemon, Silvia Baiguera, Miguel Angel Burguillos, Costantino Del Gaudio, Antonio Beltrán Rodríguez, Alexander Sotnichenko, Karolina Kublickiene, Henrik Ullman, Heike Kielstein, Peter Damberg, Alessandra Bianco, Rainer Heuchel, Ying Zhao, Domenico Ribatti, Cristián Ibarra, Bertrand Joseph, Doris A. Taylor & Paolo Macchiarini

Experimental orthotopic transplantation of a tissue-engineered oesophagus in rats

After the CEPN decision, the publisher Springer Nature stopped short of retracting this paper and issued in October 2016 solely an expression of concern. Now, the retraction is unlikely to be avoided.

The full English-language decision letter by KI from today is available here (Swedish original here). Below a shortened version.

---

Suspicion of research misconduct

Decision

In a change to Karolinska Institutet’s decision  of 28 August 2015 in the matter with reference number  2-218412014,  Karolinska  Institutet finds that the research presented  in the article ‘Experimental orthotopic transplantation  of tissue-engineered oesophagus  in rats” published in the scientific  journal Nature Communications  (2014;5:3562)  is research  misconduct.

Paolo Macchiarini,  Philipp Jungebluth, Sebastian Sjöqvist and Mei Ling Lim are responsible  for research misconduct in this article.

Paolo Macchiarini  and Philipp Jungebluth  are today not employed  by KI. Actions on labour law grounds are therefore not applicable.

Sebastian Sjöqvist and Mei Ling Lim are both employed by Karolinska Institutet. They are junior researchers who were in a position of dependence  in relation to Paolo Macchiarini and other more senior researchers  in the research team. In addition  the process  has been very prolonged  in time. These are seen as mitigating circumstances, why they are only issued an admonition  (erinrarn). The responsible  department  head at Karolinska  Institutet shall in the next two years or, if their employment is shorter, during their time of employment  actively  follow up and support the research  they conduct at Karolinska  Institutet to ensure that the research is conducted in accordance  with good research  practice.

The journal Nature  Communications shall be notified with a request  that the article “Experimental  orthotopic  transplantation of tissue-engineered oesophagus  in rats” (2014;5 :3 562) be immediately withdrawn.

[…]

Investigation  in the now current matter ref. no. 2-72312016

Expert Group for research  misconduct  at the Central Ethical Review  Board

On 26 February 2016, Vice-Chancellor  Karin Dahlman-Wright  requested that the Expert Group for research misconduct at the Central Ethical Review Board (the Expert  Group) make a statement  on whether  what comes forth from the documents in the matter constitutes research misconduct.  The Expert  Group  should also particularly  shed light on whether the description  of the results in the article appears to be embellishment,  and the participating authors’  responsibilities for the content of the article both before and after publication.

On 18 March 2016,  the Expert Group commissioned  Eva Ekblad at the Department  of Experimental Medical Science  at Lund University  to review as an expert if research misconduct  occurred with regard to the scientific article now in question.

Eva Ekblad  worked based on the reported article in its entirety. In her report on 9 July 2016, she considers  that the article’s main author committed research misconduct  in several respects.  These are the refusal or inability to provide all raw data for the results presented  in the article, the refusal  or inability to provide animal  experiment journals  for the animals used in the article, a misleading  presentation,  interpretation  and description of results, a gross deviation from the animal ethics  permit  and misleading  the regional animal experiment  ethics review board.

Like Eva Ekblad,  the Expert Group had the whole article’s  contents as the starting  point for its assessment.  It also confirms that the deficient and incomplete documentation  are such a serious deviation from good research practice that it constitutes research misconduct. What is claimed in the article to be successful was not successful.  An example of this is that the rats that underwent surgery with the artificial oesophagus  and survived for 14 days during which they were examined,  in contrast to that stated in the article, had significant  weight loss which would not have been the case if the experiment had succeeded. This should have led to the discontinuation of the experiment.  Images and figures that should support the claimed research are in many cases inaccurate and misleading. The raw data that the Expert Group studied does not always agree with that in the article,  which leads to incorrect conclusions.  The methodology is summarily described, original data cannot be extracted from the material in the matter, there is a lack of systematic records and summaries of analyses and how data has been statistically processed and then analysed.  The conclusion is that the results in the article do not agree with the outcome of the research  conducted, which is research  misconduct.  In contrast to Eva Ekblad,  the Expert Group  considers that the shifting  information  on significant weight  losses among  the rats involved in the research  varies in such away that the deviation  from animal ethics permits in this case was not such that it in itself constitutes research  misconduct.

With regard  to the authors’  responsibility,  the Expert Group  believes  that a scientific article is a shared  product with a collective responsibility  and if the article is associated with such serious  shortcomings  that research misconduct  can be established,  the responsibility  falls on all authors. It is established  that the majority of the co-authors’ specific contributions  to the article are not presented,  and that it is word against word on crucial points.  The main author Paolo Macchiarini  bears the utmost responsibility and Sebastian Sjöqvist  and Philipp  Jungebluth  had central roles in the research process.  The Expert Group has some understanding  for the more junior co-authors’ position of dependence  on Paolo Macchiarini and the other leaders in the research  group, but believes  that it cannot  exempt them from responsibility  even  if it can be considered as a mitigating circumstance.

The authors’ opinions of the Expert  Group’s statement

The authors have been given an opportunity to express opinions on the Expert  Group’s statement. What the authors  who responded,  i.e. 17 of them, primarily stated is presented below.

Paolo Macchiarini, Philipp Jungebluth,  Johannes  Haag,  Silvia Baiguera, Constantino Del Gaudio  and Doris A. Taylor have submitted  a joint statement.  The circumstance  that some documentation  from the animal experiments  has not been able to be relocated  cannot  be blamed on the researchers, they write. Nobody from the research team was there when the unit was closed and moved. Nor can they be held responsible for the material that was there. It was KI’s responsibility that the documents were taken care of and archived in connection  with the move. In addition, they point out that it is not a matter of an artificial, but rather a biological oesophagus  and that nowhere in the article is it asserted  that this oesophagus  could be used as a model for a well-functioning transplant for an oesophagus  in a human. They also believe that it is difficult to respond to the opinions expressed  as they are vague and lack details, which means that the authors  question if the Expert Group understood the article. That it was unintentionally  rendered  incorrectly  in respect of the rats’ weight is admitted  as it concerns  weight loss and not weight gain, but otherwise,  the weigh losses are very well described in the article.  All weights were otherwise documented on paper,  in electronic  documents and in the analyses made. The fact that the joumal published the article contradicts the claim that the study’s results do not have support in what is presented  in the article. It is not stated in the article that the oesophagus was well-functioning and that the  transplants were successful.  Instead, they were very careful to be cautious in their description of the results through  word choices such as suggest, indicate and epithelium-like.

[…]

Bertrand Joseph [investigated by KI for irregularities in his own papers, -LS] describes his and Miguel Angel Burgillos’ contribution to the research and the article such that Miguel Angel Burgillos  analysed  the tissue  samples that Sebastian Sjöqvist  provided  him with and that they both contributed to the data analysis of the macrophages  presented  by Figures 5 d-g.

Doris A. Taylor has, in addition to that stated above, stated that she contributed suggestions on the experiments upon visiting Paolo Macchiarini’s laboratory.  Her involvement otherwise consists of her receiving a draft of the text in July 2013 and expressing quite a few editorial and critical  opinion  and proposing  more analyses.

Royal Swedish Academy of Sciences’ statement regarding co-author responsibility

On 3 October 2016, a work group within the Royal Swedish Academy of Sciences’ research policy committee submitted  its opinion in principle regarding  co-author responsibility  for misconduct  upon KI’s request.

Notification to the journal

On 18 July 2016,  KI notified the editorial  staff of the journal Nature  Communications  of the investigation  that is under way and asked them to consider an “Expression of concern”, which the journal did on 14 October 2016.

[…]

Main author’s  and co-authors’  responsibility

With regard  to the authors’  responsibilities, the Expert  Group  believes that a scientific article is a shared  product  with a collective responsibility and if the article is associated with such serious  shortcomings  that research misconduct  can be established, the responsibility  falls on all authors. It is established that the majority of the co-authors’ specific contributions to the article are not presented,  and that it is word against  word on crucial points.  The main author Paolo Macchiarini  bears  the utmost  responsibility and Sebastian  Sjöqvist  and Philipp  Jungebluth had central roles in the research process. The Expert Group has some understanding  for the more junior co-authors’ position of dependence  on Paolo Macchiarini  and the other leaders in the research  group, but believes that it cannot exempt them from responsibility  even  if it can be considered  a mitigating circumstance.

[…]

Based on all authors’ accounts of their input, KI divides them into three groups

1 .  Senior  authors  who have a chief responsibility  for the research.

Paolo Macchiarini and Philipp  Jungebluth  have both had a chief responsibility  for the design of the experiments, the execution  of the animal  experiments,  the interpretation  of the results and the authoring  of the article. They therefore have the chief responsibility for the misconduct  present  in the research  and the published  article.

2. Junior authors with a good insight into all or large parts of the research process.

Sebastian Sjoqvist was a doctoral student when the animal experiments were under  way. He is the first author and in “author contributions”  is stated  to be the one who, together with Philipp  Jungebluth, operated  on the animals and was responsible  for data collection. In addition, Sebastian  Sjöqvist  together with Paolo Macchiarini  wrote the text for the article. Mei Ling Lim was a postdoc in the research team and in “author contributions”  is said to be the one who carried out cell and molecular biology  analyses  and contributed,  together with Philipp  Jungebluth and Paolo Macchiarini,  to data analyses  and interpretations/explanations  of collected  data. According to her statement, she was only responsible for statistical analyses of her own analyses. However, “author contributions” and other documentation, including the appended  text with traceable changes, provides support for her having a more central role in the work. Sebastian  Sjöqvist and Mei Ling Lim are therefore  deemed to have had a good insight into all or large parts of the research  and therefore have a shared responsibility  for the misconduct.  They are both junior researchers  who were in a position  of dependence  on Paolo Macchiarini  and other more senior researchers  in the research  team, which may be seen as a mitigating  circumstance, and they therefore  cannot  be considered to have the same responsibility  for the misconduct  as Paolo Macchiarini  and Philipp Jungebluth.

3. Other authors

The other authors have  carried out and had responsibility only for a limited part of the research.  In their replies, they have presented what they did and none of the external reviewers  have criticised their contributions. According to the journal’s rules, as co-authors, they have been able to see and review the final version  of the text so all authors of course  have a certain responsibility  for the final text. Many of the co-authors also mention that they do not have the expertise to assess the whole of the text, but rather that they have expertise in their own limited area. In modem medical research,  cooperation between many researchers  with widely different  competencies is very common.  In light of this, KI assesses  that the other authors cannot be held responsible  for the misconduct in the research  in question.

Sanctions and other consequences

Paolo Macchiarini has earlier been dismissed by KI’s Personnel Disciplinary Board  in March 2016. Philipp Jungebluth is not employed at KI today. Actions against them on law grounds are therefore not applicable.

Sebastian Sjöqvist and Mei Ling Lim are both employed by Karolinska Institutet. They are both junior researchers who were in a position of dependence in relation to Paolo Macchiarini and other more senior  researchers in the research team. In addition the process has been very prolonged in time. These are seen as mitigating circumstances; why they only are issued an admonition (erinran) instead of being considered for disciplinary actions by KI’s Personnel Disciplinary Board. The responsible department head at Karolinska Institutet shall in the next two years or, if their employment is shorter, during  their time of employment actively follow up and support  their research to ensure that the research that they conduct at Karolinska Institutet  is done in accordance  with good research practice.

KI will immediately contact the joumal Nature  Communications and request  that the article  “Experimental  orthotopic transplantation  of tissue-engineered oesophagus  in rats” Nature Communications (20 I 4;5 :3 5 62) be immediately  withdrawn.

Decisions in this matter  have been made by acting Vice-Chancellor  Karin Dahlman-Wright  after presentation  by acting Pro-Vice-Chancellor  Henrik Grönberg  and General Counsel Helén Törnqvist.

---