BMJ has published the latest volley in a battle over one of the most controversial drugs of the 21st century: the anti-influenza compound oseltamivir, better known as Tamiflu. A working group of the Cochrane Collaboration, an international network of scientists that performs systematic reviews of the medical literature, has carried out the most exhaustive meta-analysis yet of the drug’s efficacy —and its conclusions are, once again, pretty damning.
Tamiflu can make flu symptoms disappear a little sooner than they would otherwise, the authors say, but there is no evidence that it can prevent serious complications from flu, or keep people out of the hospital. The group questions the wisdom of buying massive stockpiles of the drug to prepare for influenza pandemics, as many countries have done.
The review comes after a long, drawn-out fight to obtain all available data from Tamiflu trials from Roche, the company that produces the drug. The Cochrane group, with active support from BMJ, eventually won that tussle, and in doing so made Tamiflu the poster child for a successful broader campaign  to ensure access to clinical trial data. (The European Medicines Agency has already said that it will make the information it receives from drug companies publicly available , and several companies—including Roche and GlaxoSmithKline—have pledged to become far more transparent.)
There’s little disagreement that Tamiflu, which came on the market in 1999, is active against the flu virus—at least somewhat. Studies have pretty consistently shown that it can shorten the duration of symptoms, which normally last about a week, by up to 1 day. But the current debate is whether Tamiflu can also prevent serious complications and hospitalizations. (The data have never been strong enough to warrant conclusions about whether it prevents death.)
The Cochrane Acute Respiratory Infections Group, led by Tom Jefferson, an independent researcher in Rome, published its first review of Tamiflu’s efficacy in 2006 , based on published papers. In it, they confirmed the then-conventional wisdom that Tamiflu could reduce flu complications and hospitalizations risks in adults. But in 2009, the group received a letter from a Japanese pediatrician, Keiji Hayashi, who questioned the data.
To the group’s surprise, eight out of the 10 studies reviewed in a 2003 meta-analysis co-authored by Roche scientists —and favorable to Tamiflu—had never been published. Roche initially refused to hand over these data. “I don’t like being made a fool,” Jefferson says in a story in the current issue of BMJ by science journalist Julia Belluz . “I trusted literature. I trusted people who were doctors and researchers. I trusted the archive. I trusted Roche.”
In an updated review published in December 2009 , the group concluded that, lacking these privately held data, there was no evidence that Tamiflu prevented complications or hospitalizations. The findings changed public views about Tamiflu, and led to buyer’s remorse in some of the countries that stockpiled the drug.
Hayashi’s letter also marked the start of the Cochrane group’s quest to obtain the full data from Roche . The company promised to send the so-called clinical study reports—massive documents that contain the patient data underlying the conclusions in scientific papers—as early as 2009, but it wasn’t until last year that the full data sets actually arrived. The Cochrane group says that Roche was stonewalling them; the company says the delay was due to logistical issues, legal and privacy concerns, and miscommunication with the Cochrane scientists.
Now that the group has crunched the hard-fought numbers, the conclusions aren’t much different: Patients get better a little sooner with Tamiflu, but there is not enough evidence to say that they are at lower risk of complications or hospitalizations. What’s more, the authors write, side effects such as nausea and vomiting are more common than was known so far. “We believe these findings provide reason to question the stockpiling of oseltamivir, its inclusion on the [World Health Organization] list of essential drugs, and its use in clinical practice as an anti-influenza drug,” the group writes. (In a separate paper, the group finds similarly disappointing results for another flu drug, zanamivir, also known as Relenza , which sold far less than Tamiflu.)
Many flu scientists have disagreed with the Cochrane group’s previous analysis and dispute the new findings as well. Shortening the duration of symptoms, for one, isn’t useless, wrote Wendy Barclay, chair in influenza virology at Imperial College London, in a statement distributed by the Science Media Centre (SMC) in London. “Although one day does not sound like a lot, in a disease that lasts only 6 days, it is. In the community this gets people back to work and school,” she wrote.
But perhaps more important, some researchers object to the group’s conclusions about Tamiflu’s usefulness in an influenza pandemic. During the most recent pandemic, in 2009 and 2010, scientists didn’t carry out randomized clinical trials of Tamiflu, the gold standard in medicine, and the Cochrane group did not take into consideration so-called observational studies. But a review of such studies done during the pandemic, published last year  by Jonathan Nguyen-Van-Tam of City Hospital in Nottingham, U.K., showed that Tamiflu did reduce the risk of severe outcomes. “We now know that antivirals saved lives during the pandemic and we risk losing one of the few weapons we have because of overly negative publicity,” said Peter Openshaw, director of the Centre for Respiratory Infection at Imperial College London, in a statement to the SMC.
Dissatisfied with the work of the Cochrane Collaboration, another group of scientists has already said that it will repeat the work independently . The Multiparty Group for Advice on Science, which has an unrestricted grant from Roche, has said it will take the observational data into account. Virologist Albert Osterhaus of Erasmus MC in Rotterdam, the Netherlands, who launched the initiative last year, says its results will be announced at an influenza meeting in Riga  in September.
One way to settle the questions once and for all would be to set up new clinical trials with oseltamivir. The question is who would fund them. There’s little incentive for Roche to do so; the patents on the drug will expire in most countries in 2016, clearing the way for generic competitors to produce the compound, Financial Times journalist Andrew Jack writes in a story also published in BMJ this week .
That may mark the end for an amazingly successful run for Roche. Since the drug was launched in 1999, Tamiflu has generated more than $18 billion in sales, Jack writes. Roughly one-half of that was for pandemic stockpiles, the vast majority of which were never used and will soon expire; the U.S. government alone spent $1.3 billion on its Tamiflu cache.