Deborah Cohen on the attempts to track down unpublished oseltamivir trial data

Deborah Cohen“The same standard of openness should apply to all (drug) trial data, whether sponsored by industry, investigator-initiated, or sponsored by public grant-giving bodies.” That’s the view of representatives from the European Medicines Agency and the regulatory bodies from France, the UK, and the Netherlands writing in PLoS Medicine.

Their statement comes as accompaniment to an article detailing the Cochrane Collaboration’s attempts to track down unpublished oseltamivir (Tamiflu) trial data after Roche, the drug’s manufacturer, promised to provide the full clinical studies.

The PLoS Medicine paper is the latest twist in the longstanding saga around oseltamivir. A series of investigations by the BMJ and the Cochrane Collaboration over the past two years have found a litany of concerns over the integrity of the published literature about the effectiveness of Roche’s blockbuster drug.

The problems run the gamut of dubious publication practices—unpublished trials, ghostwritten articles, lead authors not having ready access to data, and missing adverse events from published literature.

The Cochrane group say that their lack of trust in the published literature has hampered their ability to fully assess the effectiveness of the widely stockpiled drug. They also say that Roche have not honoured their very public promise to make the full clinical study reports available.

The inability to access key trial data meant the Cochrane Collaboration was unable to verify claims made in a Roche supported meta-analysis by Laurent Kaiser and Fred Hayden that Tamiflu prevented serious complications of influenza in healthy people.

This review underpinned the guidance given by major health organisations to stockpile the drug in case of a pandemic influenza outbreak. (Further investigations also found that Roche funded some of the experts who advised the likes of the WHO and the European Medicines Agency on the benefits of oseltamivir in pandemic planning.) And governments—in the fear of being caught unprepared—duly followed the recommendations, and spent billions of dollars of public money on the drug.

Roche, meanwhile, is trying to claw back money owed to it by EU countries embroiled in the economic crisis engulfing Europe.

The PLoS Medicine paper comes as a sharp reminder of the perils of scientists and the media stoking fear about the potential of killer infections to wreak havoc globally—good science falls by the wayside. When newspapers create a health scare through the misreporting of side-effects of drugs or vaccines, they’re rightly condemned for irresponsible scaremongering.

However, every death from avian flu virus, H5N1 is announced on Twitter, while top ranking science journals publicly debate with the US security services about the dangers of publishing the details of the latest incarnation of the deadly virus.

The millions of people with very real infectious diseases, such as drug resistant tuberculosis, malaria and trypanosomiasis, must wonder what they have to do to figure so prominently in scientific consciousness.

While there is so much concern—which may well be justified—about the potential for an influenza pandemic, Roche can maintain its stance on not sharing the entirety of the data. They have provided a range of reasons for their refusal.

Reasons include: Cochrane has not agreed to enter into a confidentiality agreement with the company, a similar meta-analysis is commencing and there are concerns that the request may conflict with this, the initial request for data comes from the media, and some 3,200 pages of data have been made available already.

Authors of the PLoS paper, Tom Jefferson, Peter Doshi, and Chris Del Mar, unpick Roche’s excuses. Tamiflu is a global public health drug and the media has a legitimate reason for helping independent reviewers obtain data, they say.

They add that “3,200 pages of data is a fraction of the full study reports for the ten Kaiser trials Roche promised to make available” and state that “It is unclear why another group of independent researchers would prevent Roche from sharing the same data with our group.”

“They [Roche] are aided in this stance by public health agencies—WHO, the European Centre for Disease Prevention Control, and the US Centers for Disease Control and Prevention—who dismissed the results and implications of the latest Cochrane review,” Jefferson says.

This describes a possible alternative mode of action for the drug and, according to Jefferson, this is “our most important finding for a whole series of reasons.” They also say that oseltamivir has not been shown to reduce transmission of the influenza virus.

The PLoS Medicine paper states that regulators are not always in possession of full trial reports of all studies on a given intervention—as described in the BMJ earlier this year.

But the European drug regulators are starting to realise the current situation is no longer tenable. “We consider it neither desirable nor realistic to maintain the status quo of limited availability of regulatory trials data.”

They propose working toward revising the current system to allow greater public access to drug company data. Though they argue that, for it to work, it would have to be balanced with measures to protect the privacy of commercially sensitive data from drug firms.

The Australian regulators, the Therapeutic Goods Administration, have also announced that they intend to change how they work.

“The TGA has looked closely at the recent Cochrane review conclusions about its effect on complications or transmission. The TGA expects full study reports containing study protocol, reporting analysis plan, statistical analysis plan and individual patient data to clarify outstanding issues. These full clinical study reports are at present unavailable to us.”

Until they do this, US advocacy group, Alliance for Human Research Protection are clear about what should happen. “Nobody should participate in clinical trials unless the data generated by such research is made public,” they said in response to Jefferson and colleagues.

“Human beings who volunteer to serve as research subjects have been led to believe that their sacrifice–undertaking risk and discomfort—will serve a humanitarian purpose. If sponsors of research are not compelled by law to disclose the data from those trials—the integrity of the research is thrown into doubt.”

Deborah Cohen is investigations editor, BMJ.