About these ads

Retraction Watch

Tracking retractions as a window into the scientific process

Former Mount Sinai postdoc faked gene therapy data: ORI

with 27 comments

ori logoA former postdoc at the Mount Sinai School of Medicine in New York faked data in four published papers, one submitted manuscript, and four NIH grant applications, according to new findings by the Office of Research Integrity.

We reported on six retractions from Savio Woo’s Mount Sinai lab in 2010, from the Proceedings of the National Academy of Sciences (PNAS), the Journal of the National Cancer Institute, and two each from Human Gene Therapy and Molecular Therapy. The PNAS paper, as we noted then:

claimed to have discovered a possible cure for phenylketonuria, or PKU, in mice—a finding that was cited more than 30 times and trumpeted in the media.

At the time, Mount Sinai said that two of the lab’s postdocs had been dismissed for misconduct. Now, more than three and a half years later, the ORI reports that a former postdoc in that lab, Li Chen:

intentionally, knowingly, and recklessly fabricated and falsified data reported in four (4) publications, one (1) submitted manuscript, and four (4) grant applications…

The ORI report refers to the PNAS paper, one of the Human Gene Therapy papers, and the two Molecular Therapy papers, as well as an unpublished manuscript submitted to PNAS. (Chen is not an author of two of the papers — one from Human Gene Therapy, and the one from the JNCI.)

Mount Sinai had already conducted in investigation before the papers were retracted, and ORI conducted its own analysis after reviewing the report of that investigation. Here’s what the ORI found:

The Respondent fabricated figures reporting the chromosomal locations of integration sites, fabricated data reporting the use of polymerase chain reaction (PCR) to determine integration frequencies, falsified data representing the detection of chromosomal translocations in human cells, and fabricated figures by falsely reporting the results of High-Performance Liquid Chromatography (HPLC) assays. The Respondent also falsified experimental data for LacZ stained liver sections and for hematoxylin and eosin (H&E) stained liver sections.

The fakery involved 19 figures claiming that

phenylketonuria (PKU) gene therapy experiments were successfully completed, when the available evidence shows the experiments were not performed…

Chen — who “failed to take responsibility for the fabrication and falsification,” according to the ORI — will not be eligible for NIH grants for three years, and can’t serve on any NIH committees, including peer review committees, for the same amount of time.

About these ads

27 Responses

Subscribe to comments with RSS.

  1. A 3-year slap on the wrist? That’s a joke! Paid or sponsored to fabricate is how I would translate this story. The public should be enraged and demand that they pay back every penny received from tax-payers’ coffers. Raw justice must the way to go because petty penalties only encourage more dishonesty.

    JATdS

    April 25, 2014 at 8:19 am

  2. Please see Pubpeer comments:

    https://pubpeer.com/publications/18846115

    Full paper:

    http://www.nature.com/cgt/journal/v16/n3/pdf/cgt200874a.pdf

    Cancer Gene Ther. 2009 Mar;16(3):266-78. doi: 10.1038/cgt.2008.74. Epub 2008 Oct 10.
    Enhanced oncolytic potency of vesicular stomatitis virus through vector-mediated inhibition of NK and NKT cells.
    Altomonte J1, Wu L, Meseck M, Chen L, Ebert O, Garcia-Sastre A, Fallon J, Mandeli J, Woo SL.
    Author information

    1Department of Gene and Cell Medicine, Mount Sinai School of Medicine, New York, NY 10029-6574, USA.

    david hardman

    April 25, 2014 at 8:23 am

    • Could you please tell me how to publish on pubpeer? It appears to be heavily moderated. I commented something yesterday but they don’t publish it – this is not the first time. Has anyone experienced this?

      KK

      April 25, 2014 at 10:28 am

      • As long as your comments are factual you will not have any problem in posting. At least I did not. If you post after logging as a registered member the comments appear immediately but as unregistered it make take a couple of days.

        anoninform

        April 25, 2014 at 11:35 am

        • Why we need to give our identity for registration and we don’t even know who the moderators are – is it not a bit strange?

          KK

          April 25, 2014 at 12:15 pm

          • When posting new commet at Pubpeer, or posting a reply to a comment, you will get 2 options.
            Sign in, and Don’t sign in.

            Use the latter. After posting you will see “I understqnd, Post New Anonymous Comment”.

            Click on that blue area to submit comment.

            Success” appears after that.
            I usually click on that too.

            david hardman

            April 25, 2014 at 12:24 pm

          • Not strange at all. Any time you’re commenting online (including here at RW) your IP address is being logged. Unless you’re using TOR or another IP obscuring service, the folks who own RW know exactly where you are. It’s the same at PubPeer – the people who own the site know your IP. It’s up to you to decide how happy you are with that, and whether you’re willing in addition to give them your identity as a researcher. If you don’t want to do that – fine, but just accept that your comments will take longer to appear. That’s the price you pay.

            The longer term issue here is stability of PubPeer’s legal walls. If they were to be subpoenaed, then they could be forced to give up this sort of information. There’s no evidence that’s happened, and they seem to be on a pretty solid footing legally, but then again, until last summer nobody knew the NSA was reading email! If someone has powerful lawyers and enough money, they will be able to find the information they want.

            Regarding turnaround speeds, for brand-new unregistered comments on previously un-discussed papers, they tend to take a few days, but if your comments are being made on a paper that’s on the front page or already has a substantial base of comments, those tend to be moderated and approved faster, to keep the thread current.

            Paul Brookes

            April 25, 2014 at 3:16 pm

          • There is a good reason to “sign in”. You remain anon – so “Peer1, peer2, etc, but the thread is easy to follow. If you don’t ‘sign in”, then you appear as “Unreg”, but then so does everyone who fails to sign in. the problem with this is that when there are many Unreg, it is dificult to follow the thread; even worse when oneof the Unreg is the authors or related to them and involved in ad hominems and gish gallops. So I always “sign in”,

            ferniglab

            April 30, 2014 at 4:38 pm

  3. Given the ORI findings 3yrs seems very light – why not a lifetime ban. This behaviour will continue unless the cost of being caught is so high that it is not worth the risk. The PI’s should also suffer as it all happened under their watch. Also the Institution/Department should feel some pain. Both the PI’s and Institutions are happy to bask in the glory of the original papers. What about any patents that have been filed on false claims? Finally what about the patients who were given false hope?

    Patrick

    April 25, 2014 at 8:43 am

    • While I have not checked the CFR, I suspect the three year ban is the ORI’s enabling legislation, so it would require Congress to change the law in order to enact a harsher penalty.

      StrongDreams

      April 25, 2014 at 3:55 pm

      • That is not correct. Bans can any length of time and are often 3, 5, 7 years–or a lifetime.

        Theresa Defino

        April 30, 2014 at 4:39 pm

  4. This is simply beyond comprehension. A lifetime ban is what is needed for EGREGIOUS and MALICIOUS faking, and in addition, failure to acknowledge the fabrication and cheating.

    Statistical Observer

    April 25, 2014 at 9:23 am

  5. who is going to give him a job in science after this? A lifetime ban would be superfluous.

    Colin MacDiarmid

    April 25, 2014 at 9:33 am

    • You would be surprised how these individuals are still able to get a job after the most egregious actions. Thats why the Institution department needs to feel the financial pain. Payback all indirect and indirect costs back to the various funding bodies that were cited as supporting this fake work. Effectively put a funding lien on the Institution.

      Patrick

      April 25, 2014 at 9:51 am

    • Uh…he’ll get a job the same way Anil Potti got (and has kept) a job; the same way Evan Dreyer got (and has kept) a job; the same way Terry Reisine got (and has kept) a job; the same way Terry Elton has kept his job.

      D Cameron

      April 25, 2014 at 5:51 pm

      • Potti is a physician. No one ever said that his physician skills were problematic. He was a terrible administrator. I have never seen any evidence that Potti was DELIBERATELY screwing up. Rather, what seems to have happened was that he hired and employed incompetent data managers. No one has ever been identified, TTBOMK, as the data manager/data analyst who actually did the screw-ups.

        Statistical Observer

        April 25, 2014 at 7:20 pm

        • “Rather, what seems to have happened was that he hired and employed incompetent data managers”

          Do we know he employed him/her/them? Can we be sure if other papers from Duke may have been affected by same problem – ie were the data managers (possibly employed by Potti) shared between a number of groups?

          It was odd and unpleasant how this came personalized on Potti. The biggest problem is not that errors were made, but that the response was handled appallingly. And although Potti was the last/first author on most of the papers, he was not the most senior member of the institution in the authors list.

          littlegreyrabbit

          April 26, 2014 at 5:39 am

          • Right you are. Nevins, who was the senior author, has made amazing statements in public, absolutely washing his hands of the whole thing, even though he raised Potti up himself from the clinical services ranks. Duke is one of the most pre-eminent medical research places in the country, top 10 for sure. These papers had biostats folks involved – what were they doing? The other folks in the papers who actually did the data management and made the serious errors – which of those are still involved in medical research at Duke? WHO SPECIFICALLY did the one-off errors, the mislabelling, the duplication of samples in the calibration set?

            Potti has taken the entire heat on this, and it’s a real question as to who the actual perp on the mistakes was. Because I do not believe, for a single minute, that Anil Potti was involved in spreadsheet management in these experiments and trials.

            Statistical Observer

            April 26, 2014 at 11:08 am

  6. 3 years for a such blatant misconduct, expecially for someone who “failed to take responsibility for the fabrication and falsification”, clearly shows the inadequacy of NIH’s policies. I was wondering what level of indignation scientific community should reach to see them changing.

    Dario

    April 25, 2014 at 10:12 am

  7. “The ORI report refers to the PNAS paper, one of the Human Gene Therapy papers, and the two Molecular Therapy papers, as well as an unpublished manuscript submitted to PNAS. (Chen is not an author of two of the papers — one from Human Gene Therapy, and the one from the JNCI.)” Does this mean that there are other papers with data falsification but Chen is not an author on those? If so then does it mean that Chen is being made a fall guy or he did the work but was not credited with authorship?

    Another related question is when the post doc/ graduate student gets the blame for images/ data on the basis of which the PI gets the grant then why they never get any credit for the grants awarded? Is it sufficient that they generate data to ensure funding for the lab and their salary/ stipend? In my opinion it should be a part of their training where they should be more engaged in grant preparation and appropriately credited for grants awarded rather than using them as mere data generators which seems to be the case in majority of the places.

    WB

    April 25, 2014 at 11:32 am

    • I think you are confused by Ivan’s writing :-)

      Maybe this helps clarify: Woo’s lab has retracted 6 papers. On 4 of those Chen is the first author, but not on 2 others. Likely the culprit in those two papers is the second postdoc that was fired (Zhiyu Li). The ORI report does not mention those other two retracted papers. Maybe that will come later in a separate report.

      Marco

      April 25, 2014 at 1:57 pm

      • Thank you Marco for clarifying. I had not followed up on this but now that I looked at earlier posts I came across another comment pointing to a paper where either of these 2 fired post docs are not an author suggesting of some thing more going on in the lab https://pubpeer.com/publications/BC40D2F699B06F6934A46DBA075131
        It may be worthwhile to remember that Savio Woo was cleared by an internal investigation carried out by the university and I am not sure if ORI investigated beyond the recommendations of the university report. A question that has been asked numerous times – when these kind of issues arise why is the investigation not handed over to an outside party with no vested or conflicting interests?

        WB

        April 25, 2014 at 5:37 pm

  8. He should be banned from Science and arrested.

    DT

    April 25, 2014 at 2:30 pm

  9. FWIW, 45 CFR Part 76:

    § 76.320 Period of debarment.
    (a) Debarment shall be for a period commensurate with the seriousness of
    the cause(s). If a suspension precedes a debarment, the suspension period shall
    be considered in determining the debarment period.
    (1) Debarment for causes other than those related to a violation of the requirements
    of Subpart F of this part generally should not exceed three years. Where circumstances warrant, a
    longer period of debarment may be imposed.
    (2) In the case of a debarment for a violation of the requirements of Subpart
    F of this part (see § 76.305(c)(5)),the period of debarment shall not exceed five years.

    (Subpart F has to do with a drug free workplace)

    So likely for the federal bureaucracy, 3 years is the path of least resistance. They could probably get taken to federal court for imposing more than 3 years, since so many other offenders have only gotten 3 years. Those who feel strongly can always write to HHS and ask them to revise the regulations.

    StrongDreams

    April 25, 2014 at 4:21 pm

    • You are correct to some extent, Strong Dreams — the “standard” or typical debarment from federal funding that is imposed by ORI and other federal agencies for serious research misconduct is 3 years (85 by ORI since 1992).

      However, ORI has imposed or negotiated 1 year (13 by ORI) or 2 years (3 by ORI) in lesser cases. But ORI has also imposed in 40 other cases considerably more years (to lifetime in 3 “medical” cases):

      — 4 years (Washbaugh 1996, Hajra 1997, Tracy 2002, Xu 2004, Sperber 2008)
      — 5 years (Lee 1993, Rosner 1993, Tewari 1994, London 1997, Angelides 1999, Simmons 2000, Garey 2001, Arnold 2002, Handa 2002, Ruggerio 2002, Smith 2002, Yao 2002, Eagan 2003, Ganz 2003, Muenchen 2003, Hiserodt 2004, Jacoby 2005, Aronica 2006, Leadon 2006, Robinson 2007, Roovers 2007, Uzelmeier [Marcus] 2007, Van Parijs 2009, Sezen 2010, Zach 2012)
      — 7 years (Brodie 2010, Smart 2012, Thiruchelvam 2012)
      — 8 years (Poisson 1993)
      — 10 years (Dreyer 2001, Gelband 2004, Thomas 2009)
      — lifetime (Poehlman 2005, Kornak 2006 [and prison], Sudbo 2007).

      According to federal counsels, the federal administrative actions are not intended to be “punitive” but instead are imposed to protect the federal interest (i.e., taxpayer dollars) for a given period — expecting that those who are so publicly outed (at least by ORI) for research misconduct will hopefully not repeat their bad acts (a few have been recidivists in ORI cases, however, and they were re-debarred (e.g., Lin 2001 and 2004, 3 years each).

      [see case summaries in the ORI Annual Reports since 1993 at http://ori.hhs.gov/annual_reports ]

      Alan Price

      April 26, 2014 at 3:07 am

      • Alan

        Protecting the taxpayer dollar is fair enough – but what about protecting the public from the potential harm of using dubious compounds in phase I medicinal trials, clinical trial data being manipulated and being further used in phase II/III trials?

        We need an oversight foundation consisting of panels of independent experts to establish whether there is any science-fraud history of any molecule prior to its use in clinical trials. As we can see from the Potti case, such individuals/groups are indeed prepared to use compounds, only shown to be effective in less than honest data in preclinical work, on human beings.

        In any other system it would be called poisoning, which is illegal and a criminal offence.

        Stewart

        April 26, 2014 at 6:46 pm

  10. The ORI finding mentions “evidence and findings of an investigation report by Mount Sinai School of Medicine (MSSM) transmitted to [ORI] in April 2010″. So it took 4 years for this action to be taken, even with an institutional report in hand.

    Stefan

    April 27, 2014 at 10:04 am


We welcome comments. Please read our comments policy at http://retractionwatch.wordpress.com/the-retraction-watch-faq/ and leave your comment below.

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 35,693 other followers

%d bloggers like this: