Canadian investigators published the results of 4 observational studies in April 2010 that purported to show an association between the 2008 TIV and nH1N1 illness. The 4 studies were a test‐negative case‐control study based on the Canadian sentinel vaccine effectiveness monitoring system in 4 provinces, a conventional case‐control design in Quebec that used population control subjects who were contacted by telephone about influenza vaccination, a test‐negative case‐control study in Ontario, and the Quebec household transmission study. The second (Quebec) and third (Ontario) studies included hospitalized as well as ambulatory patients. The sentinel vaccine effectiveness system estimated that the risk of nH1N1 infection was increased 1.4–2.5‐fold by vaccination with seasonal 2008 vaccine after adjustment for comorbidities, age, and geography. A perspective on these studies was provided by Viboud and Simonsen. They noted that TIV remained protective (53%) against seasonal viruses circulating in April and May 2009. The complementary studies in Ontario and Quebec confirmed the increased risk of nH1N1 infection attributed to TIV; however, this risk did not extend to the subsets of patients who were hospitalized. Hospitalized patients tended to have the same vaccination rates as did the control subjects. This inconsistency is difficult to explain if seasonal vaccine increased the risk of infection, because enhancement of disease might be expected. Furthermore, the household transmission study did not show increased risk of nH1N1 infection for 2008 vaccine recipients.
One variable that the previous observational studies had overlooked was the effect of natural infection by seasonal influenza virus. Cowling et al had the advantage of a study structure in place at the time of the emergence of nH1N1 that allowed them to determine infection with seasonal viruses as a factor to compare with seasonal TIV vaccination and nH1N1 infection. They found that a recent seasonal influenza A infection protected against nH1N1 infection and that TIV administration did not affect nH1N1 infection rates. Persons who received seasonal TIV were protected against the seasonal viruses and, therefore, were less likely to benefit from the cross protection afforded by natural infection. Without knowledge of recent natural infection, observers might conclude that seasonal vaccine is the risk factor. At this time, the best hypothesis for explaining the distribution of nH1N1 infections by vaccine status is that natural seasonal influenza A infection provides cross protection against nH1N1. Persons with 2008 TIV vaccination were less likely to benefit from the protection afforded by natural seasonal virus infection because of the protection afforded by the TIV.
The novel H1N1 virus is so classified because it is antigenically distinct from previously circulating viruses. Perhaps, therefore, the protection provided by natural infection is mediated by innate or nonspecific mechanisms rather than by specific immunity. Examination of Figure 2 in the article by Cowling et al shows that seasonal influenza viruses were co‐circulating with the nH1N1 virus during the period May–October 2009. Infections with seasonal viruses of any type would stimulate production of interferon and other cytokines that might transiently prevent infection with nH1N1. Evidence for interference among the major respiratory viruses has been suggested by observations from Houston, Texas, and Chapel Hill, North Carolina. LAIV infection mimics natural infection by stimulating immunity by multiple mechanisms. LAIV provides almost immediate protection against respiratory viruses, including influenza, that is evident for at least 14 days after administration, which allows time for specific immunity to develop. Experimental studies in animal models and involving human volunteers support this hypothesis. LAIV should be the vaccine of choice for healthy children and adults, especially when influenza is prevalent in the community.