Retraction Watch

Tracking retractions as a window into the scientific process

Malpractice case against Duke, Anil Potti settled

with 22 comments

pottiA lawsuit filed in October 2011 against Duke University and Anil Potti, who has retracted 11 papers and corrected a number of others amidst investigation into his work, has been settled, Retraction Watch has learned.

Potti resigned from Duke in 2010 following questions about his work, and revelations that he had lied on grant applications about being awarded a Rhodes Scholarship. He now works at a cancer center in North Dakota.

The lawsuit was filed by subjects in clinical trials based on Potti’s work. Plaintiff’s attorney Thomas W. Henson, who confirmed the settlement, tells Retraction Watch:

Our clients and Duke University Health System have resolved and settled all claims against the current defendants arising out of the genomics-based clinical trials for cancer treatment.  The specific terms of the settlement are confidential, by mutual agreement of the parties.

In January, the case was postponed because of a case of the flu among the plaintiff’s attorneys. Also that month, a judge ordered Duke to release more documents.

This isn’t the first time Potti has faced legal issues — he has settled at least 11 cases of malpractice for$75,000 or more, as we reported in 2011.

We’ve contacted Duke for comment, and will update if we hear back.

Like Retraction Watch? Consider supporting our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, and sign up on our homepage for an email every time there’s a new post.

Written by Ivan Oransky

May 1st, 2015 at 2:37 pm

Comments
  • Foobar May 1, 2015 at 3:22 pm

    Where is the settlement money coming from? If even partially from federal grants or tuition fees, then doesnt the public have a right to know?

    • Derec Avery May 1, 2015 at 6:40 pm

      I would doubt that federal grants or tuition fees would be involved in the payouts. I would think that Duke University specifically maintains liability insurance for just such malpractice lawsuits and the settlement(s) came from the insurance company and not the university. (Though I don’t no how high their premiums are going to rise as a result.)

  • Leonid Schneider May 1, 2015 at 4:54 pm

    I remember discussing with others on RW against Potti to be allowed to continue working in clinic. The argument was, if he doesn’t do research, he shouldn’t be banned from working clinically as doctor. Now I read here about malpractice issues. But even then, would you trust your doctor’s medical opinions and skills if you knew there were serious questions about his academic activities?

    • lhac May 1, 2015 at 7:45 pm

      Having trust in your doctors abilities is a good thing – it will help the healing process via the placebo effect.
      Your feeling of trust may not reflect his/her true abilities, though. Better let the hospital decide about those, they have more data to let them judge the quality of their personnel.

    • NVP May 2, 2015 at 6:08 am

      I would not trust him or any other ethically challenged people with my or anyone else’s health. He should move on to a field where that trust is not needed.

  • Scrutineer May 2, 2015 at 9:33 am

    “have resolved and settled all claims against the current defendants”

    Does that mean that there could be another batch of defendants in future? So still no closure for Duke, then?

    Since the good doctor is in the news again, figure sleuths might enjoy a hitherto unretracted paper dating from 2006 in the early-middle Duke phase, before the statisticians had begun sharpening their pencils.

    Chen G1, Sircar K, Aprikian A, Potti A, Goltzman D, Rabbani SA. Expression of RANKL/RANK/OPG in primary and metastatic human prostate cancer as markers of disease stage and functional regulation. Cancer. 2006 Jul 15;107(2):289-98.

    http://www.ncbi.nlm.nih.gov/pubmed/?term=16752412

    The paper is open access. Your attention is specifically drawn to the immunocytochemical analysis in Fig. 2.

    http://i.imgur.com/IguuIew.png

    There is more, much more.

    Note: In fairness to the Rhodes Scholar manqué, he is listed as the fourth MD out of six.

    • herr doktor bimler May 3, 2015 at 3:25 am

      http://i.imgur.com/IguuIew.png
      There is more, much more.

      The two lower right-hand panels in that image — ‘RANK’ and ‘OPG’ in the PC-3 column — look very similar, apart from a 180-deg. rotation.

    • herr doktor bimler May 3, 2015 at 8:13 am

      figure sleuths might enjoy a hitherto unretracted paper dating from 2006
      How’s this?
      http://i.imgur.com/HKkq0lQ.png

      • Scrutineer May 3, 2015 at 3:59 pm

        A decent start, oh Good Doktor Bimler!

        Though there is plenty more. For example all three of the fields in the left column have areas of cells that are related.

        The mystery though is why some sections in these fields are certainly not the same when most of the cells are identical in position and shape. Using red and blue false colours and transparency, see how the left pair that you suggested were related do indeed superpose after rotation, but with many red and blue cells being unmatched.

        http://i.imgur.com/2gGkged.png

        Since these cells have been stained with antibodies, they are fixed and quite dead, so cannot move around. Cells cannot arbitrarily come and go in these stainings. And as far as I know, cells are never sequentially reprobed with another antibody (unlike western blots, though cells can be simultaneously double-labelled in one experiment using labels of different colour).

        What do you think can be going on here?

        • herr doktor bimler May 3, 2015 at 4:46 pm

          Can you do the same thing with the panels in the right-hand column?

          • Scrutineer May 4, 2015 at 4:21 am

            I can 🙂 Here are the right two fields similarly superposed.

            http://i.imgur.com/KtRItif.png

            Observe that most of the centrally placed cells superpose very well. Most of the cells around the edges do not. Cells in the top right corner are clearly unrelated for example.

            Any theories as to how this could have happened?

          • herr doktor bimler May 4, 2015 at 7:10 am

            Peer 2 at PubPeer notices similar phenomena potentially linking other panels within the Figure.
            https://pubpeer.com/publications/A91F660B800C6CEF5529278BD95235#fb29306

            I am tempted to try the monochrome version of your color-coded super-imposition trick — that is, turning one image into its photographic negative, and superimposing on the second with partial transparency.

          • Scrutineer May 4, 2015 at 1:19 pm

            It will be interesting to see how you fare with the monochrome superposition. Do show us. There may well be better schemes than the quickly bodged false colour that I used.

            Meanwhile I made a small addition to the last figure. One of them right side fields may also have an in-image duplication.

            http://i.imgur.com/33xRYsW.png

          • herr doktor bimler May 4, 2015 at 7:14 pm

            An experiment in image cancellation.
            http://i.imgur.com/XMtGqKQ.png

          • Scrutineer May 5, 2015 at 2:29 am

            Doktor Bimler, your image cancellation seems to work very well! I think if you rotated DU145/RANKL by the usual 180 degrees and superposed it on DU145/RANK, cancellation would be almost complete.

            This leaves DU145/OPG as the single unique field of twelve, does it not?

          • herr doktor bimler May 5, 2015 at 7:31 pm

            I think if you rotated DU145/RANKL by the usual 180 degrees and superposed it on DU145/RANK, cancellation would be almost complete.

            OK, I added DU145/RANKL and RANK; also PrEC RANKL and RANK (unrotated). I have also included the Pc-3 / OPG self-comparison you mentioned.

            http://i.imgur.com/TbembYA.png

          • Scrutineer May 6, 2015 at 2:23 am
  • DrGeneNelson May 3, 2015 at 4:21 am

    See also: “Duke University settles suit with cancer patients over clinical trials” 02 May 2015 http://www.newsobserver.com/news/local/counties/durham-county/article20103387.html

    While reprimands against Anil Potti M.D. by the Medical Boards of North Carolina and Missouri exist, the Medical Board of North Dakota policy is to NOT display those adverse actions. See: https://www.ndbomex.org/public/find_verify/verify.asp Use last name of Potti or License Number 12312

  • fernando pessoa May 4, 2015 at 8:20 am

    Chen G1, Sircar K, Aprikian A, Potti A, Goltzman D, Rabbani SA. Expression of RANKL/RANK/OPG in primary and metastatic human prostate cancer as markers of disease stage and functional regulation. Cancer. 2006 Jul 15;107(2):289-98.

    https://pubpeer.com/publications/16752412

    Figure 2.

    http://imgur.com/PJRdghA

    http://i.imgur.com/wDF2JOm.jpg

    http://i.imgur.com/9KZnO5H.jpg

    • Scrutineer May 4, 2015 at 1:35 pm

      The PCRs in Fig. 1 show, at a quick glance, 16 bands. How many are there really?

  • Steven McKinney May 4, 2015 at 10:00 pm

    Scrutineer:

    Only a review of original source materials will reveal how many bands were imaged during the study. It’s not exactly clear from Fig. 1.

    http://i.imgur.com/r5fvTJq.png

    • Scrutineer May 5, 2015 at 2:15 am

      If the upper three gel slices are adjusted exposures of the same original, then the number of unique bands would be six, given the clear duplication of GAPDH controls. As you say, original source materials would indeed be required for unambiguous assignment.

  • Post a comment

    Threaded commenting powered by interconnect/it code.