The only chink of light - that there are still great journalists (like Shelley Wood) doing the job that the mainstream medical journals, regulators and leadership of medicine are manifestly failing to do. Well done Shelley.
The Heartwire article of yesterday is reproduced intact below. For my collated postings on this scandal click here
MIST correction published; no new information on disputed echo review for PFO/migraine trial
Sep 9, 2009
Dallas, TX - Authors of the controversial Migraine Intervention with STARflex Technology (MIST) trial paper  in Circulation have now published a lengthy correction  to the manuscript that originally appeared online in March 2008. As previously reported by heartwire, journal editors requested a correction from the authors after questions were raised—most notably by one of the original investigators for MIST—about details purportedly missing from the publication.
The correction by Dr Andrew Dowson (Kings College Hospital, London, UK) and colleagues appeared online Monday August 31, 2009, and the online version of the article has been updated to reflect the changes.
MIST was the first randomized, sham-controlled trial to investigate patent foramen ovale (PFO) closure as a treatment for migraine; the study failed to show a benefit of PFO closure. Garnering even more attention than the study results, however, have been questions from physicians both within and outside of the study about the design and conduct of the study and the airing of its results. The study sponsor, NMT Medical, is currently suing one of the original co-PIs for MIST, Dr Peter Wilmshurst (Royal Shrewsbury Hospital, UK), for concerns he raised to heartwire about how MIST data were being managed.
The disputed echo reviewsWilmshurst, along with his Shrewsbury Hospital colleague Dr Simon Nightingale, refused to sign off on the manuscript before it was submitted for publication and are not included as authors. Following the publication of MIST in Circulation, Wilmshurst contacted the journal about his concerns.
Conspicuously absent from the correction is information about an "off-protocol" review of the echocardiographic information that suggested a higher residual shunting rate following PFO closure than that reported by Dowson at the TCT 2007 meeting. Dowson reported a 94% closure rate—derived, as per trial protocol, from the follow-up conducted by the implanting cardiologist. Wilmshurst, however, says he conducted an after-the-fact review of echocardiograms, at NMT's behest, which suggested residual shunting in approximately one-third of patients. A residual shunt can point to either incomplete device closure of the PFO or to the presence of pulmonary shunts not properly identified at baseline.
Both Dowson and NMT originally denied to heartwire that this echo review had ever taken place (repeatedly insisting that it was not part of the trial protocol) and rejected Wilmshurst's suggestion that NMT had asked investigators to send their echos to him to review. Another trial investigator, however, Dr Michael Mullen (Royal Brompton Hospital, London, UK), subsequently confirmed to heartwire that a review had indeed been undertaken but insisted the information did not belong in the MIST paper.
Asked by heartwire why information about the disputed echo review is not mentioned in the correction, Circulation issued a statement saying, "The Circulation editors were unaware of this controversy during the review of the paper; however, they investigated the matter at the same time that they investigated Dr Wilmshurst's other claims, and they decided that this information did not belong in the paper."
Also contacted by heartwire, Wilmshurst responded with a written statement:
"The correction of the MIST paper states that I did not sign the Copyright Transfer Agreement because of an internal disagreement about the conduct of study. In an email to Dr Dowson in October 2007, I gave two reasons for refusing to sign the Copyright Transfer Agreement. First, that the paper contained errors, which are confirmed by this correction. Second, that I could not in conscience sign the declaration, [which reads] 'The authors had full access to the data and take responsibility for its integrity.' To sign a declaration knowing it to be false would make me guilty of a criminal offence in the United Kingdom. In November 2007, NMT's lawyer wrote to me: 'No investigator is given a full detailed set of all the data.' For 18 months I pressed Circulation for a correction. I sent the editor hundreds of pages of documents, including a copy of my email to Dr Dowson. All of the errors and omissions in the correction are ones that I reported to Circulation, but not all errors and omissions that I reported have been corrected. In addition, the contributions of some who did little in the study, including an 'author' who died in May 2004, before the trial started, are credited, but the contributions of Dr Nightingale and I, who helped design the trial, were members of the steering committee, wrote the early draft of the paper, and did over 600 of the 2000 patient visits in the trial, are not acknowledged."
NMT chief financial officer Richard Davis, also commenting on the correction for heartwire, emphasized that the "clarifications were made on request of the editors of the journal.
"The correction that appears in the journal Circulation includes a single error in the reference to a histogram contained in the original manuscript. No changes to the results, analysis, or conclusions of the study are made. The remainder of the published correction contains clarifications, many of which were not included in the original manuscript due to the word-count limitations.
"To comply with these word-count requirements, the manuscript focused specifically on the critical elements of the trial design and results. All serious adverse events were listed; other information was excluded that was deemed irrelevant to the findings regarding the effect of PFO closure on migraine. One procedural embolization in a cohort of 74 patients is not uncommon for these types of devices, and retrieval of a device due to poor positioning followed by replacement with a second device, as happened in two cases, is also not unusual. It is important to note that these procedural difficulties were not considered serious by the investigator or the safety committee of the study."
Clearing the MISTThe correction runs over 700 words long, with a four-page data supplement, and addresses what Dowson and colleagues term "a number of errors and omissions," including:
- An explanation of why Wilmshurst and Nightingale were not listed as authors.
- An expanded description of how intracardiac shunts were assessed.
- Clarification on specific adverse events related to failure to find or cross a PFO, two cases of device embolization, and one nondeployment.
- New information on patients who withdrew from the study and why.
- A correction to the units used for number of migraine days per month for Figure 3.
- Reiteration of the reasons for excluding two patients from an analysis looking at median total migraine days and the calculations used.
- Clarification that the authors' conflicts of interest information provided in the article was accurate.
Of note, at the time of Dowson's October 2007 TCT presentation, he disclosed that he had received grant/research support from NMT, and Davis estimated in an interview with heartwire shortly after Dowson's presentation that up to 10 study investigators and/or steering committee members had received consulting fees from the company. Wilmshurst once told heartwire that he and Nightingale had turned down offers for "large" consultancy payments from NMT.
- Dowson A, Mullen MJ, Peatfield R, et al. Migraine Intervention with STARFlex Technology (MIST) trial. A prospective, multicenter, double-blind, sham-controlled trial to evaluate the effectiveness of patent foramen ovale closure with STARFlex septal repair implant to resolve refractory migraine headache. Circulation 2008; 117:1397-1404.
- Correction for Dowson et al. Circulation 2009; DOI: DOI: 10.1161/CIRCULATIONAHA.109.192626. Available at: http://circ.ahajournals.org.