Background: Antibody titers to influenza antigens decrease with time following vaccination raising concerns about the potential for waning influenza vaccine efficacy (VE). However, the relationship between time since vaccination and protection is unclear.
Methods: Healthy subjects 18-49 years of age were enrolled in a placebo controlled trial evaluating the efficacies of inactivated (IIV) and live-attenuated (LAIV) influenza vaccines prior to the 2007-08 influenza season. Subjects had blood collected for serologic studies immediately prior to vaccine receipt, 30 days after and at season's end, and were followed for identification of symptomatic acute respiratory illnesses. Throat swab specimens were collected from ill subjects and tested for influenza by RT-PCR or isolation in cell culture. Hemagglutination inhibition (HAI) and neuraminidase inhibition (NAI) antibody titers were measured at three points. The time varying coefficient for the effect of each vaccine (β(t)) was estimated by fitting a smooth Loess function to scaled Schoenfeld residuals summed with the coefficient estimated in Cox proportional hazards models; time varying VE (VE(t)) was calculated as 100*(1-eβ(t)). The test for proportional hazards was used to test for waning.
Results: Influenza, primarily type A (H3N2), circulated from early January through mid-April (Figure1). Overall VE was estimated to be 70% (95% Confidence Interval [CI]: 50% - 82%) for IIV and 38% (95% CI: 5% - 59%) for LAIV. Statistically significant waning was detected for IIV (P=0.03) but not LAIV (P=0.37); however, clear downward trajectories were observed for both vaccines in plots of VE(t) (Figure 2). Despite observed waning, IIV remained significantly efficacious for ~6 months until data became sparse in the tail end of the season. VE(t) results were consistent with serologic data; HAI and NAI antibody titers significantly decreased over the influenza season among IIV recipients but not placebo or LAIV recipients.
Conclusion: Although significant waning of IIV VE was detected, the vaccine maintained a high level of protection for the majority of the influenza season in subjects vaccinated approximately 3 months prior to peak influenza activity. These results do not suggest a benefit to delaying vaccination until just prior to the influenza season.
J. G. Petrie,
A. Monto, Sanofi Pasteur: Grant Investigator , Research grant