Monday, November 05, 2007

NMT Medical and a new UK research scandal

An important new scandal involving inappropriate research procedure, misconduct in procedure, and possible misconduct in data analysis has blown up in the UK this week.
  1. The saga involves a medical device, and the conduct and conclusions of the Migraine Intervention with STARflex Technology (MIST) I trial. A brief summary is below (a detailed press report is here)
  2. MIST I was the first large trial to test whether the use of a new device to close a patent foramen ovale would reduce the incidence of migraine headaches. As it turns out, the results of the trial may have implications for the company which extend far beyond the issue of headache.
  3. The study was sponsored by NMT Medical (of Boston, MA) and had two co-principal investigators who are as different as chalk and cheese a) Dr Peter Wilmshurst, a cardiologist and b) a so-called headache specialist Dr Andrew Dowson.
  4. Dowson: Although Dowson described himself as a neurologist, he is not. More importantly, Dowson has restrictions on his medical practice following previous research misconduct (interesting fiddling of research records here, GMC restrictions here). Dowson states that NMT was fully aware of the research misconduct "undertakings" when he was chosen to be co-principal investigator.
  5. Wilmshurst: By contrast, Wilmshurst is a highly respected cardiologist, and has worked in the field of PFO closure for 20 years. Wilmshurst was on the international planning committee that designed the MIST trial, and was originally selected to head up the trial before Dowson was put forward by the company as coinvestigator. NMT's website has multiple press releases naming Wilmshurst and Dowson as co-principal investigators. The protocol as submitted to the MREC (Multicentre IRB) in the UK states that "the study principal investigator will be shared by two physicians due to the nature of the trial".
  6. According to Wilmshurst, his troubles with the sponsor began after he and Dowson presented the MIST I trial results at the ACC 2006 meeting. To the disappointment of many, MIST I was negative: PFO closure was no better at curing migraine than a sham procedure among people with PFOs and migraine with aura. One hypothesis explaining failure was that there may have been a high rate of residual shunts among patients. "Although a higher rate of residual shunts would help explain the negative outcome in MIST, it has much larger implications for a device already marketed in Europe for the treatment of cryptogenic stroke".
  7. According to Wilmshurst (who has the emails to prove it) NMT officials instructed MIST investigators to send all patient echocardiograms to Wilmshurst for independent review immediately following the ACC. According to Wilmshurst, he ultimately reviewed the contrast echoes for all but one of the patients who underwent the scan at follow-up. At the TCT 2007 meeting last week, Dowson reported that the closure success rate was 94%, according to the implanting physician--in other words, a residual shunt rate of just 6% for the trial. Wilmshurst's review of the echocardiograms, however, suggested that the residual shunt rate was far higher--approximately one-third of the echocardiograms--a number he says was rejected by NMT. A second review of the echocardiograms (by Dr Luke Missault in Belgium) apparently yielded similar findings. NMT then ceased all communication with Wilmshurst and apparently refused further review of the echocardiograms stating that "to have echocardiograms reviewed at a single center would be a violation of study protocol, since it was not something that was specified in the patient consent process". !!!
  8. Wilmshurst states that NMT has repeatedly "blocked his attempts to view the complete MIST I data set, has lied about whether echocardiograms from the MIST patients have been independently reviewed, and has massaged the data to portray its PFO device in the best possible light". Dr Andrew Dowson presented the "findings" at the TCT 2007 meeting (TCT 2007: Transcatheter Cardiovascular Therapeutics October 20 - 25, 2007, Washington, DC). It is not clear whether Dowson has the Echo data either, or whether and has the expertise to interpret it. "In fact, the first Wilmshurst learned that these data were going to be presented at TCT was when he arrived in Washington for the meeting and saw the trial listed in the TCT program."
  9. Wilmshurst aired concerns about irregularities in the MIST trial and the activities of its sponsor, NMT Medical, during a presentation at this week's TCT 2007 meeting. NMT retaliated by claiming Wilshurst is not a principal investigator, cannot see the study data and has dropped Wilmshurst from the trial for so called "protocol violations" and so called "violations of the investigator's agreement". Asked for examples of these violations, the company declined to provide any "at this stage". NMT's lawyers also accused him of "breaching confidentiality agreements". Wilmshurst states that it is his duty to speak out "I cannot permit a confidentiality agreement to be used to allow NMT to make statements that are untrue in the belief that I or others who know that they are lying will remain silent". Wilmshurst also notes that he and other investigators at his center have "declined large consultancy payments from the sponsor".
  10. Several other trial investigators are asking questions. One stated "There should be complete transparency of the MIST echo data with an independent core lab review of the studies that have already been done". The company provided a different excuse to this investigator implying that his proposal of transparency is somehow inappropriate because he "receives consultancy fees from [another] company, AGA". (? logic)
  11. Wilmshurst and another colleagues (including Dr Simon Nightingale, who is also on the MIST steering committee) have repeatedly asked for the complete data collected in the MIST trial. According to Wilmshurst, the partial data they have seen, is "simply inconsistent". Wilmshurst also states that he and Nightingale have repeatedly asked for a full explanation as to why two patients, described as outliers, were excluded from secondary analyses of MIST I.
  12. "For the past year, Wilmshurst says the company has responded to his emails only by threatening him with legal action and telling him that he is forbidden to speak publicly about the trial. Attempts were made by NMT to drop his center as a trial site (despite the fact that it had been the second-largest enroller). This was vetoed by the ethics board. Later however the center withdraw from after learning that the CRO running the trial (Matrix) had been ordered by the company not to provide clinical record forms to the center.
  13. THE MREC (IRB) ordered NMT to meet with the steering committee and with the data safety and monitoring board. However, according to Wilmshurst, the company has more than once announced that it would hold this meeting, only to cancel at the last minute.
  14. Dowson, the other investigator with headache expertise has now submitted the study publication to Circulation who have apparently accepted it. It is apparently going to be published "in a matter of weeks".
  15. The company states that "outside independent auditors" have reviewed the data in the paper. "They seem to be comfortable at least with what's been presented to them . . . . Ethically it's important to get the data out there so we can do further studies."
  16. So it seems OK for unknown "outside independent auditors" to (supposedly) see some data ("at least with what's been presented to them"), but not the clinical investigators or study authors.
The procedure in this trial may constitute research misconduct irrespective of any manipulation of the data. In fact it seems hard to know whether this is science or black magic. The corrupted (and unreformed) UK regulator the MHRA is now apparently also on the case, so perhaps yet another amusing obfuscation is heading our way. As assessors of black magic for the homeopathy industry, the MHRA are probably well placed to comment on this stuff. After a few years of silence and poor leadership the bleeding starts.

See NMT medical's code of ethics. There isn't much there about patients or research integrity, or the hiring of shonky investigators to bypass those who care.

The integrity of UK science isn't looking too good either. It's deja vu again and again for this type of problem.

Earlier|Later|Main Page


Anonymous said...

GMC links have moved. Slippery toads who don't keep a consistent web interface

Anonymous said...

This site has a link to the GMC report:
which is here:

Anonymous said...

The paper on this trial has now ben published in Circulation 2008;117:1397-1404:

Note the statement: The authors had full access to and take full responsibility for the
integrity of the data. All authors have read and agree to the manuscript as written.

Anonymous said...

All authors signed that except the ones who were not prepared to see science bulldozed and colleagues bullied into submission through legal threats, and who refused to allow "outliers" to be excluded with resulting misleading conclusions. Several authors who signed were apparently untruthful about their stock-holdings - untruthful about one thing = untruthful about others.

Anonymous said...

So what has happened regarding this issue?

Anonymous said...

The NYT has an account of this case: Device Maker Sues a Doctor Who Called Its Product Flawed.