Three randomised controlled trials (RCTs; n=873) were included in the review. Another double-blind placebo-controlled RCT (n=1,256) was found while the journal was in press, and this was summarised briefly as a footnote to the review.
The review's authors noted several methodological limitations of the included studies: a lack of consistency in the definition of influenza; a lack of detail in relation to the clinical and demographic characteristics of the trial participants; inappropriate outcome measures; and too much weight on retrospective subgroup analyses.
The intention to treat analysis of the North American and European trials indicated that the mean time required for clinical recovery was reduced by 0.7 days (5.3 days on average instead of 6 days) in the group receiving 10 mg of inhaled zanamivir plus placebo nasal spray (p=0.04). The oral inhalation plus nasal spray regimen was not significantly more effective than oral inhalation alone. Consumption of symptomatic treatments (especially acetaminophen) was not significantly different between the three groups.
The subgroup analysis of 262 patients with virologically-confirmed influenza in these two trials yielded similar results: there was a mean reduction of 0.9 days with inhaled zanamivir relative to placebo (p=0.05). Zanamivir was no more effective than placebo for patients in whom influenza was not confirmed virologically. The time required to resume normal activities, and the incidence of complications requiring antibacterial chemotherapy, were not significantly different between those taking 10 mg of inhaled zanamivir and those taking placebo.
In the third trial, the median time to clinical recovery was significantly shorter with zanamivir than with placebo, whether the analysis included all patients or only those with virologically-confirmed influenza. The median time for resumption of daily activities was 2 days shorter with zanamivir (7 versus 9 days, p=0.001). According to a retrospective subgroup analysis, zanamivir seemed similarly effective whether patients were infected with influenza virus type A or type B. Patients at risk of complications (mostly patients with moderate asthma in this trial) had a significantly lower risk of complications (usually pulmonary disorders) with zanamivir than with placebo.
The incidence of adverse events was not significantly different with inhaled zanamivir and with placebo in the three trials. However, there are few data on the safety profile of zanamivir in patients at risk. A press release from the US Food and Drug Administration was reported, which states that cases of bronchospasm following zanamivir administration have been reported in patients with asthma and chronic obstructive airways disease.
The fourth trial that was appended to the review reported a statistically-significant reduction in time to clinical recovery of one day, no prevention of complications in people at risk, and a statistically-significant reduction in the number of disturbed nights and global symptom score with zanamivir. The authors question whether the latter two results are clinically relevant.