Tamiflu shortens flu symptoms by a day, study finds

Oseltamivir (Roche’s Tamiflu) shortens the duration of flu symptoms by approximately a day, in infected patients, according to an analysis published in The Lancet^[1]
.

The analysis also claims that oseltamivir lowers the risk of lower-respiratory-tract infections that require antibiotics, such as pneumonia, and reduces the need for hospitalization. But the authors of the first complete systematic review^[2]  of oseltamivir for the Cochrane Collaboration, published in April 2014, dispute these benefits.

The Cochrane reviewers found instead that oseltamivir, an inhibitor of the neuraminidase enzyme, which has a role in the infectivity of influenza A and B viruses, reduces the duration of symptoms in adults by 16.8 hours. But they found there was no good evidence that it reduces serious complications of flu, including confirmed pneumonia, bronchitis or sinusitis, or the number of people admitted to hospital
^[2].

The Cochrane reviewers and The BMJ had fought for four-and-a-half years to gain access to all the data held by Roche on trials of oseltamivir, and the conclusions of the Cochrane review were based on documents relating to 20 of these trials.

In The Lancet study published on 29 January 2015, an independent group of investigators analysed “all published and unpublished Roche-sponsored randomised placebo-controlled, double-blind trials of 75 mg twice a day oseltamivir in adults”. Their conclusions were based on data from nine trials that took place between 1997 and 2001 involving 4,328 adults.

This analysis, the investigators say, is the first on Tamiflu to use individual patient data made available by Roche, rather than looking at aggregate study results. The investigators applied to Roche and were awarded a grant to fund the study.

The researchers found that symptoms were alleviated 21% faster in infected patients who were treated with oseltamivir rather than a placebo (time ratio 0·79, 95% CI 0·74–0·85; p<0·0001). Flu symptoms resolved after a median of 97·5 hours in patients given oseltamivir compared with 122·7 hours in patients given a placebo (difference –25·2 h, 95% CI –36·2 to –16·0). Participants were deemed to be infected if a culture from a nasal or throat swab was positive or they had a four-fold or greater rise in antibody production.

When patients with flu symptoms in whom an infection had not been confirmed were included in the analysis, oseltamivir still shortened the duration of symptoms but to a lesser extent: symptoms resolved 17·8 hours sooner than in patients given a placebo.

Fewer lower-respiratory-tract complications requiring antibiotics were reported in infected patients treated with oseltamivir: 4.9% of patients required antibiotics compared with 8·7% of those who took a placebo (risk ratio [RR] 0·56, 95% CI 0·42–0·75; p=0·0001). Oseltamivir also reduced the likelihood of hospital admission: 0·6% of oseltamivir-treated patients were admitted, compared with 1·7% of those in the placebo group (RR 0·37, 95% CI 0·17–0·81; p=0·013).

However, use of oseltamivir was found to increase the risk of nausea and vomiting. Nausea was reported in 9·9% of patients in the oseltamivir group and 6·2% of those in the placebo group (RR 1·60, 95% CI 1·29–1·99; p<0·0001), and vomiting was reported in 8·0% and 3·3%, respectively (RR 2·43, 95% CI 1·83–3·23; p<0·0001). No effects on neurological or psychological disorders were recorded.

The lead researchers were Arnold Monto, professor of epidemiology at the University of Michigan, and Stuart Pocock, professor of medical statistics at the London School of Hygiene and Tropical Medicine. “Our meta-analysis provides compelling evidence that oseltamivir therapy reduces by one day the typical length of illness in adults infected with influenza and also prevents complications and reduces the number of people needing hospital treatment,” said Monto. “Whether the magnitude of these benefits outweigh the harms of nausea and vomiting needs careful consideration.”

But one of the Cochrane reviewers, Peter Doshi, assistant professor of pharmaceutical health services at the University of Maryland in Baltimore, told The Pharmaceutical Journal: “This appears to be a Roche-sponsored reanalysis that confirms the benefits of Tamiflu.”

Monto’s team trusted that what was recorded by the original investigators as “pneumonia” really was pneumonia, which would have resulted in patients being given antibiotics. “I would challenge that assumption. Our interpretations are in line with the US Food and Drug Administration (FDA), which does not allow Roche to claim that Tamiflu reduces the risk of complications,” Doshi said. “Our interpretations were driven by our access to over 100,000 pages of Tamiflu clinical study reports, something these authors did not access, according to their methods section, but something that the FDA did.”

Doshi added that the authors of The Lancet study wrote that their results “for complications and admittance to hospital are broadly consistent with those of observational studies and some previous meta-analyses of randomised trials” — but this is “misleading”, he said. One of the meta-analyses they pointed to concluded that “it was not possible to assess the potential benefit for high-risk participants who are hospitalised, because the sample size of most studies was too small,” he added.

Discussing their results in The Lancet, the investigators said that the balance of benefits and harms becomes less favourable if more non-infected participants are treated with oseltamivir, so treatment strategies need to avoid this, for example through the use of rapid diagnostic testing.

In a linked comment piece in The Lancet, Heath Kelly from the Australian National University in Canberra and Benjamin Cowling from The University of Hong Kong point out^[3]
: Advice about testing will be different in a pandemic or a severe epidemic. In these situations it will not be practical to test everyone, and the population benefit will depend on preliminary assessments of the proportion of the population with an influenza-like illness that is attributable to influenza, the risk of admission to hospital because of influenza, and the potential costs of treating or not treating patients.”