Is Tamiflu a waste of money?
A new review questions whether Tamiflu and Relenza can halt the spread of the disease and says governments should reconsider stockpiling the drugs. But there are many doctors and scientists who disagree. Elizabeth Finkel reports.
On 19 March the prestigious journal Lancet Respiratory Medicine published a study showing that the influenza drugs Tamiflu and Relenza save lives. This week another prestigious publication, the British Medical Journal (BMJ), published a review saying they don’t.
So has there been a “Tamiflu turnaround” in the space of a few weeks? It’s certainly not an academic question. If a deadly new flu pandemic emerges, will it be of any benefit that governments have spent hundreds of millions of dollars to stockpile neuraminidase inhibitors, such as Tamiflu, as they have in the past?
The authors of the BMJ article, a network of independent clinicians and researchers who carry out extremely thorough analyses known as Cochrane Reviews, say they shouldn’t.
“We now have the most robust, comprehensive review on neuraminidase inhibitors that exists. Initially thought to reduce hospitalisations and serious complications from influenza, the review highlights that Tamiflu is not proven to do this,” said David Tovey, Cochrane editor-in-chief. It can also cause harmful side effects, he adds.
But some experts vehemently disagree. “This new report, taken alongside a lot of other data collected in different settings, does not convince me that the risks of taking Tamiflu or Relenza would outweigh the benefits,” said Wendy Barclay, an influenza virologist at Imperial College London. “We know that antivirals saved lives during the H1N1 pandemic,” adds Peter Openshaw, who heads respiratory infection research at Imperial.
Jodie McVernon, a public health expert at the University of Melbourne, also finds the Cochrane Review’s conclusion flawed. “We have synthesised many reviews and carried out modelling for the Australian government. In a pandemic, Tamiflu is useful in reducing the clinical burden and deaths.”
A comparison between patients who received the drug early, and those who received it late, suggested early treatment halved the death rate.
So how is it that eminent journals and scientists have reached diametrically opposite conclusions? McVernon points out that the Lancet and BMJ studies were reporting on two different things: one looked at extremely sick people hospitalised during the outbreak of a deadly pandemic virus. The other looked at generally well people infected with seasonal influenza. “The findings of each study are equally valid but pertain to different things,” she said.
The Lancet researchers, known as the PRIDE Consortium Investigators and headed by Jonathan Nguyen van-Tam, focused on the 2009 H1N1 global Swine flu epidemic that claimed, according to the US Centers for Disease Control, somewhere between 151,700 to 575,400 lives worldwide. They asked whether Tamiflu and other neuraminidase inhibitor drugs prevented deaths during the pandemic. Looking at data from over 29,000 patients in hospital, they concluded the drugs saved lives. A comparison between patients who received the drug early, and those who received it late, suggested early treatment halved the death rate.
By contrast the authors of the Cochrane Review looked at general populations recovering from mild to moderate seasonal influenza. In their study the drugs had only a minor effect, reducing the length of illness from seven to six and a half days. They say there was no reduction in hospital admissions but there was an increased risk of side effects. These included a 4% increase in the risk of vomiting and nausea and a 1% increase in the risk of “psychiatric events”.
'The key issue is not whether one study is more valid than the other, but about how governments should spend their money to prepare for a pandemic.'
Besides focusing on different things, the two reports also describe different kinds of studies.
The Cochrane Review limited itself to the data obtained from randomised control studies. Considered the gold standard of clinical studies, they carefully allocate comparable groups of people to receive either the real drug or a placebo. The PRIDE study, on the other hand, observed what took place in hospitals during the pandemic. Observational studies like this are generally more open to criticism, and indeed the PRIDE study has had its share. But as McVernon points out, randomised control studies can’t be done during an epidemic on hospitalised patients at risk of dying from influenza.
It’s not the first time a Cochrane Review has examined the effectiveness of Tamiflu; its first was back in 2006. What’s different about this one is that the authors, after spending five years pressuring the manufacturer Roche, finally had got access to previously undisclosed data on the results of Tamiflu trials. While McVernon applauds the success in gaining access to the raw clinical trials data, she points out that their analysis did not uncover anything new. “The milder effect of the drugs on the severity of seasonal influenza is well-known.”
According to David Spiegelhalter, Winton professor for the public understanding of risk at Cambridge University, the key issue is not whether one study is more valid than the other, but how governments should spend their money to prepare for a pandemic. “Treating people in hospital does not cost so much - it's the mass medication that costs millions of dollars. Even if PRIDE is correct, it does not tell governments that the drugs should be given to people outside hospital.”