BMJ renews pressure on Roche to release all Tamiflu trial data

Nov 2, 2012 (CIDRAP News) – The British journal BMJ this week renewed its campaign to persuade Roche to release unpublished clinical trial data on the flu drug oseltamivir (Tamiflu)—and used the occasion to announce that in the future it will publish only those drug trial reports whose sponsors agree to release all trial data when requested.

The journal made its case in an article, an editorial, and a press release earlier this week. The publication released a letter that its editor-in-chief, Fiona Godlee, MBBChir, BSc, sent to a Roche board member about 2 weeks ago, urging him to press the company to release the data. In addition, the journal announced the launch of a Web site dedicated to the issue.

In a statement yesterday, Roche officials told CIDRAP News that almost 80% of oseltamivir clinical data has been published in the literature or made available to scientists online, and that most of the remaining data come from studies only recently completed.

The controversy dates to 2009, when BMJ (formerly the British Medical Journal) published a review by Cochrane Collaboration researchers saying they found inadequate evidence that oseltamivir reduces the risk of flu complications like pneumonia. The authors said that 8 of 10 trials examining the issue were never published and that Roche refused to supply the full data from the trials.

According to an Oct 29 BMJ article, Roche promised in December 2009 to supply full reports of all 10 trials, but it has not done so. Earlier this month, Godlee wrote to John Bell, PMedSci, a Roche board member and Oxford University professor, urging him to seek release of the data.

Bell told Godlee he referred the matter to Roche and was awaiting a response, according to an Oct 29 BMJ editorial by Godlee. The editorial argues that medical journals should publish clinical trial results only when the sponsor promises to release relevant, anonymous patient-level data from the trials in response to reasonable requests.

"The BMJ will require this commitment for all clinical trials of drugs and devices—whether industry-funded or not—from January 2013," the editorial states.

In a press release, BMJ said Godlee is one of 28 people who signed a letter published in The Times of London calling on drug companies to make clinical trial data for all currently used drugs available to healthcare professionals.

The 2009 Cochrane review published in BMJ was an update of a similar review published in 2006. The update analyzed 20 published trials of neuraminidase inhibitors used to treat seasonal flu, but eight trials that had been included in the earlier review were dropped because they were never published and the reviewers couldn't verify the results. According to a BMJ report published with the review, Roche refused to send the authors the raw data without a signed confidentiality agreement.

In its press release this week, BMJ said Roche has released some additional data to the Cochrane reviewers since 2009, "but the full data set has still not been provided."

The Cochrane authors have determined that there are at least 123 trials of oseltamivir and that 60% of patient data from completed phase 3 trials sponsored by Roche remain unpublished, according to the BMJ release.

Roche officials, in their statement yesterday, rejected the assertion that they have not fully shared data on oseltamivir. "Almost 80% of Roche's Tamiflu clinical data has been published as primary publications or made available to the scientific community on the web as complete Core Reports," the company said. "Of the remaining data, most come from studies that have recently completed. Roche is working to make them publicly available."

The company said it complies with all legal requirements regarding the publication of data but does not generally make patient-level data available because of legal and confidentiality constraints. "Roche provided the Cochrane group with access to 3,200 pages of very detailed information, enabling their questions to be answered," the statement said. "The Cochrane Group asked more questions and requested further data, but declined to sign a confidentiality agreement."

Roche officials also commented that, as has been reported previously, other researchers in 2011 reanalyzed and supported a 2003 oseltamivir meta-analysis that the Cochrane group had criticized.

In the 2003 analysis, published in Archives of Internal Medicine (see link below), a Swiss group (L. Kaiser and others) had analyzed data from 10 randomized controlled trials and concluded that the drug reduced the risk of lower respiratory tract complications in flu patients by 55%. But in their 2009 review, the Cochrane group omitted 8 of those 10 trials, saying they were never published and they couldn't obtain the full data from them.

In their statement yesterday, Roche officials noted that two Harvard researchers, Miguel Hernan, MD, PhD, and Marc Lipsitch, PhD, had included the 10 trials from the 2003 review in a new meta-analysis, which was published in Clinical Infectious Diseases in June 2011 (see link below). The results were "congruent with the Kaiser publication," the company said, adding that the Harvard group was given access to the full study reports after signing an agreement.

The report by Hernan and Lipsitch says their reanalysis of 11 randomized controlled trials showed that oseltamivir reduced the risk of lower respiratory tract complications by 37% in patients with confirmed flu.

In other comments, Roche officials said they have provided full clinical study data, including the studies from the Kaiser et al paper, to national health authorities according to their various requirements, so they can conduct their own analyses as part of the drug licensing process. The company also said it continuously monitors for side effects and notifies regulatory authorities.

Further, the statement said, "All completed Roche-sponsored clinical studies on the safety and efficacy of Tamiflu are available as peer-reviewed publications or in summary form on More detailed clinical trial reports, including all the studies from the Kaiser et al paper, are available for use by investigators on a password-protected site, enabling researchers to verify the findings of these studies."

In the BMJ editorial this week, Godlee praised GlaxoSmithKline (GSK), maker of the flu drug zanamivir (Relenza), for recently promising to provide anonymous patient-level data from its clinical trials to researchers on certain conditions. The conditions include a "reasonable scientific question," a research protocol, and a promise to publish the findings.

But Godlee added that the GSK action highlights the "rank absurdity" of the fact that, as things stand, all clinical trial data are not routinely released after regulatory authorities have made their decisions.

See also:

Oct 29 BMJ press release

Oct 29 BMJ editorial

Oct 29 BMJ story about efforts to obtain the Roche data and launch of the "open data campaign" Web site

BMJ dedicated site

Godlee letter to John Bell

Dec 9, 2009, CIDRAP News story on Cochrane review of Tamiflu

Related Jan 19 CIDRAP News story on oseltamivir review by the Cochrane group

Kaiser et al Arch Intern Med study

Hernan and Lipsitch Clin Infect Dis report

This week's top reads