ABSTRACT
During the first wave of the 2009 pandemic, caused by a H1N1 influenza virus (pH1N1) of swine origin, antivirals were the only form of therapeutic available to control the proliferation of disease until the conventional strain-matched vaccine was produced. Oseltamivir is an antiviral that inhibits the sialidase activity of the viral neuraminidase (NA) protein and was shown to be effective against pH1N1 viruses in ferrets. Furthermore, it was used in humans to treat infections during the pandemic and is still used for current infections without reported complication or exacerbation of illness. However, in an evaluation of the effectiveness of oseltamivir against pH1N1 infection, we unexpectedly observed an exacerbation of disease in virus-infected mice treated with oseltamivir, transforming an otherwise mild illness into one with high morbidity and mortality. In contrast, an identical treatment regime alleviated all signs of illness in mice infected with the pathogenic mouse-adapted virus A/WSN/33 (H1N1). The worsened clinical outcome with pH1N1 viruses occurred over a range of oseltamivir doses and treatment schedules and was directly linked to a reduction in NA enzymatic activity. Our results suggest that the suppression of NA activity with antivirals may exacerbate disease in a host-dependent manner by increasing replicative fitness in viruses that are not optimally adapted for replication in that host. IMPORTANCE: Here, we report that treatment of pH1N1-infected mice with oseltamivir enhanced disease progression, transforming a mild illness into a lethal infection. This raises a potential pitfall of using the mouse model for evaluation of the therapeutic efficacy of neuraminidase inhibitors. We show that antiviral efficacy determined in a single animal species may not represent treatment in humans and that caution should be used when interpreting the outcome. Furthermore, increased virulence due to oseltamivir treatment was the effect of a shift in the hemagglutinin (HA) and neuraminidase (NA) activity balance. This is the first study that has demonstrated that altering the HA/NA activity balance by reduction in NA activity can result in an increase in virulence in any animal model from nonpathogenic to lethal and the first to demonstrate a situation in which treatment with a NA activity inhibitor has an effect opposite to the intended therapeutic effect of ameliorating the infection.