The trial that launched millions of mesh implant procedures: did money compromise the outcome?BMJ 2018; 363 doi: https://doi.org/10.1136/bmj.k4155 (Published 10 October 2018) Cite this as: BMJ 2018;363:k4155
- Jonathan Gornall, investigative journalist, Suffolk, UK
In March 1997, Swedish obstetrician and gynaecologist Ulf Ulmsten received an offer he couldn’t refuse. A year earlier Ulmsten, the head of obstetrics and gynaecology at Uppsala University Hospital, had published a paper reporting the results of a revolutionary surgical procedure to treat stress urinary incontinence in women.
The standard surgical treatment at the time was colposuspension, a procedure little changed since it was first developed in 1959, in which the neck of the bladder is lifted, compressing the urethra, and sutured to the pelvic bone. It required open abdominal surgery and involved several days in hospital and lengthy recovery. By contrast, Ulmsten’s mid-urethral sling procedure, in which a narrow length of plastic mesh tape is inserted through the vagina to act as a sling, or hammock, to raise and support the urethra, could be done under local anaesthetic as an outpatient procedure.
A study of 75 women treated in Ulmsten’s department at Uppsala University Hospital with what became known as the tension-free vaginal tape (TVT) procedure gave impressive results—84% (63) of the women with stress incontinence were completely cured throughout a two year follow-up period, and another 8% (5) were “significantly improved.”1 The results suggested TVT was at least as successful as colposuspension in treating stress urinary incontinence,2 with the added benefits that patients could get back to their lives more quickly and surgeons could perform more and easier procedures at less cost to their hospitals.
Ulmsten, aware that his results might be considered, in his words, “too positive” because all 75 operations had been carried out by experienced urogynaecologists in his department, organised a larger, multicentre study to find out how easy, effective, and safe the procedure could be in “ordinary” gynaecological units.
But what few if any others knew at the time, including Ulmsten’s research collaborators, was that even before this second study got under way Ulmsten had signed an agreement with Ethicon, a subsidiary of global medical giant Johnson and Johnson, that would make him a very rich man—provided the results of the second trial echoed those of the first.
Details of this shocking conflict of interest emerged in 2014, during a US product liability case brought against Ethicon and Johnson and Johnson by Linda Batiste, who claimed she had been left in severe pain by the insertion of mesh tape derived from the original TVT device, and that this transobturator tape (TOT) had been designed defectively. Batiste was awarded $1.2m in damages by a Texas jury. This was overturned on appeal in 2015 but the case led to an undisclosed settlement by Johnson and Johnson in 2016.3
During the trial in 2014 the jury was shown a licensing agreement signed in March 1997 between Johnson and Johnson and Medscand, the company Ulmsten had set up to exploit his invention. Ulmsten had filed a patent application in the US listing himself and colleague Jan Clarén as inventors on 25 February 1997 and assigned the patent to Medscand.4 The following month Johnson and Johnson agreed to pay Medscand a series of payments that amounted to $1m (£770 000; €870 000) provided that the proposed second trial upheld the findings of the first. It is not clear what if anything Jan Clarén knew of the deal. Attempts were made to contact him through Invent Medic, a Swedish company with which he is involved, but he did not respond to questions or requests for an interview.
Giving evidence during the Batiste trial, Michael Margolis, an assistant clinical professor in the department of obstetrics and gynaecology at University of California, Los Angeles, said that the message to Ulmsten implicit in the deal with Johnson and Johnson was “Prove … that this procedure works and it is safe, and we'll pay you money … If you don’t prove it, you don’t get paid.” What happened next, Margolis told the court, was “human nature. What do you think he is going to do?” This, he said, had been “wallet driven research.”
The paper reporting the results of the follow-up study, in which 131 women in six hospitals in Sweden and Finland were fitted with the device, appeared in the International Urogynecology Journal in 1998.5 The results were even better than before: 119 patients (91%) were declared cured of stress incontinence, which meant “they did not leak urine postoperatively, either objectively or subjectively,” and a further nine (7%) were “significantly improved.”
“Most encouraging,” reported Ulmsten and his coauthors, was “the low complication rate in ‘less experienced’ hands.” There was one bladder perforation, which was fixed during the operation, and a single case of a wound becoming infected, but both patients were among those cured of their incontinence. Ulmsten and his coauthors concluded that “TVT can be considered a safe and effective procedure.”5
A spokesperson for Johnson and Johnson confirmed to The BMJ that it had paid Ulmsten $1m. She did not confirm that the payment was agreed to be conditional on the follow-up study proving successful. She also said that in 1999 the company had “paid Medscand a total of $24 525 000 to purchase all assets associated with the TVT business.”
Medscand and its US arm, Medscand (USA), was sold in 2001 to another US company, Cooper Surgical, for $12m.6 Johnson and Johnson said it accepted that its lucrative financial offer to Ulmsten had been on the table before the second trial was carried out but rejected any suggestion that this had compromised the results of that trial.
“As part of our due diligence in licensing the TVT product from Medscand, Johnson & Johnson was interested in evaluating evidence that the TVT device would be safe and effective and that Dr Ulmsten’s results … could be replicated in the hands of other surgeons in other institutions,” its spokesperson said. She added that none of the trial centres had received any financial support from Ethicon for conducting this study, but did not comment on the claim that Ulmsten, the lead investigator, had been promised considerable sums of money if the second trial replicated the results of the first.
Johnson and Johnson said it had been satisfied that the second trial showed that “the TVT device and the procedure to implant it held immense value to the broader medical community separate and apart from the surgical skills of its inventor.” Both studies, it added, had been published in the International Urogynecology Journal, “one of the pre-eminent journals in this field.”
The journal included an editorial comment with Ulmsten’s second paper in 1998 noting that “many questions remain regarding the diagnostic criteria they used in selecting their patients, as well as the degree of testing performed postoperatively.” Furthermore, additional studies were needed to confirm the results.
But Ulmsten’s twin papers alone would serve as the launch pad for a procedure that, despite no knowledge of long term outcomes and the compromised nature of the evidence supporting its efficacy, was quickly nodded into play by regulatory bodies and rapidly adopted by surgeons around the world.
Twenty years on from Ulmsten’s controversial multicentre study, it is difficult to determine whether the outcome of the trial, which since 2000 has led to an estimated 200 000 procedures in England and Wales alone and three million around the world, was influenced by the vast amount of money that was apparently at stake for its inventor. Ulmsten, who died in 2004, can’t answer that question, though his obituary in the journal that had carried the crucial 1998 paper recalled his “great personal qualities of modesty and integrity.”7
Ulmsten’s financial relations with Johnson and Johnson were not acknowledged in the paper, though declarations of conflicts of interest were not common at that time. Johnson and Johnson insists subsequent research has confirmed the authenticity of Ulmsten’s trials. “In the 20 years that have passed since the study was published in 1998, hundreds of clinical studies with no connection to Dr Ulmsten or Ethicon, including over 100 randomised, controlled trials, have evaluated the clinical performance of TVT, further validating its safety and effectiveness,” it said. “Scientists from around the world who have conducted and reviewed independent research on pelvic mesh agree it is an important treatment option for women [and] several medical societies comprised of physicians practising in the field of female pelvic medicine have published position statements recognising mid-urethral slings (such as TVT) as the gold standard for the treatment of stress urinary incontinence.”910
It is not clear how many of Ulmsten’s coauthors on the 1998 paper knew of his deal with Johnson and Johnson. Attempts to contact all but one of them were unsuccessful. But Christian Falconer, one of the coauthors, says that he had known nothing about the transfer of TVT rights from Medscand to Johnson and Johnson—“this was the domain of Ulf Ulmsten”—nor about any payments to Medscand, in which he had not been a shareholder. “To the best of my knowledge,” he adds, “I have never been involved in any ‘wallet driven’ research.”
Whatever the effect of Ulmsten’s deal with Johnson and Johnson, it is now accepted that no matter how “hands off” industry backing seems to be, nor how independently investigators believe they are acting, it affects the outcome of research. A systematic review by Cochrane in 2017 analysed 75 papers and concluded that drug and device studies sponsored by industry were “more often favourable to the sponsor’s products than non-industry sponsored drug and device studies, due to biases that cannot be explained by standard ‘risk of bias’ assessment tools.”11
The cloud of suspicion hanging over the events of 1997 is not dissipated by the fact, revealed in evidence in various court hearings, that some of the documentation relating to the early days of Medscand and TVT, including what one lawyer described as “the core Ulmsten data,” was destroyed in a fire in an independent storage facility in Lausanne, Switzerland, in September 2009.1213 According to testimony in 2014, only a single binder of patient data from Ulmsten’s Scandinavian multicentre study survived.14
Competing interests: I have read and understood BMJ policy on declaration of interests and have no relevant interests to declare.
Provenance and peer review: Commissioned; not externally peer reviewed.