Correspondence Seasonal Influenza

To the Editor—We were pleased to see the editorial by Glezen [1], which out-lined a hypothesis to explain the obser-vation made by Cowling and colleagues [2] in the same issue of the Journal that receipt of the seasonal trivalent influenza vaccine (TIV) appeared to increase the risk of infection with pandemic in-fluenza 2009 (pH1N1). This observation has been made previously, but the ef-fect size has varied. Four studies from Canada reported odds ratios (OR) for pH1N1 infection following receipt of seasonal TIV in the range 1.4–2.5 (con-fidence interval [CI] 5 1.03–2.74) [3], whereas the Hong Kong investigators reported a nonsignificant increased risk, with an OR 5 1.11 (.54–2.26) [2]. In-vestigators in Australia reported an OR 5.97 (.60–1.56) [4]. The timing of the pandemic compared to the previous in-fluenza season provides a possible ex-planation of these apparently discrepant findings. Although not yet a fully accepted phe-nomenon, there is a growing body of evi-dence that infection with one strain of influenza gives a ‘‘short-lived strain tran-scending immunity’ ’ [5]. This concept of heterosubtypic temporary immunity is a necessary inclusion for models to repro-duce the observed influenza phylogenetic structures [5, 6] and has recently been used by us to model and explain the apparent increased risk of infection with pH1N1 following receipt of seasonal TIV [7, 8]. There is evidence for temporary immunity from human studies during the 1957 H2N2 pandemic [9] and the reemergence of H1N1 in 1977–1978 [10]. Temporary heterosubtypic immunity is thought to last 3–6 months, but the evidence for its du-ration is derived from the modeling liter-ature [5] and needs additional investigation. During the first wave of the pandemic (April–August 2009, before pandemic vaccine was available), an individual with a recent seasonal influenza infection would have acquired some degree of temporary immunity to pH1N1 and would therefore have had a lower chance of pH1N1 in-fection. Individuals who had received the 2009 seasonal TIV had a lower probability of infection with the seasonal strain [3] and hence had an apparent higher risk of pH1N1 infection, having foregone the protection against pH1N1 infection re-sulting from seasonal influenza infection. This risk varied with seasonal vaccination coverage in the population