AS03-adjuvanted versus non-adjuvanted inactivated trivalent influenza vaccine against seasonal influenza in elderly people: a phase 3 randomised trial

Background We aimed to compare AS03-adjuvanted inactivated trivalent influenza vaccine (TIV) with non-adjuvanted TIV for seasonal influenza prevention in elderly people. Methods We did a randomised trial in 15 countries worldwide during the 2008-09 (year 1) and 2009-10 (year 2) influenza seasons. Eligible participants aged at least 65 years who were not in hospital or bedridden and were without acute illness were randomly assigned (1:1) to receive either AS03-adjuvanted TIV or non-adjuvanted TIV. Randomisation was done in an interne-based system, with a blocldng scheme and stratification by age (65-74 years and 75 years or older). Participants were scheduled to receive one vaccine in each year, and remained in the same group in years 1 and 2. Unmasked personnel prepared and gave the vaccines, but participants and individuals assessing any study endpoint were masked. The coprimary objectives were to assess the relative efficacy of the vaccines and lot-to-lot consistency of the AS03-adjuvanted TIV (to be reported elsewhere). For the first objective, the primary endpoint was relative efficacy of the vaccines for prevention of influenza A (excluding A H1N1 pdm09) or B, or both, that was confirmed by PCR analysis in year 1 (lower limit of two-sided 95% CI had to be greater than zero to establish superiority). From Nov 15, to April 30, in both years, participants were monitored by telephone or site contact and home visits every week or 2 weeks to identify cases of influenza-like illness. After onset of suspected cases, we obtained nasal and throat swabs to identify influenza RNA with real-time PCR. Efficacy analyses were done per protocol. This trial is registered with ClinicalTrials.gov, number NCT00753272. Findings We enrolled 43802 participants, of whom 21893 were assigned to and received the AS03-adjuvanted TIV and 21802 the non-adjuvanted TIV in year 1. In the year 1 efficacy cohort, fewer participants given AS03-adjuvanted than non-adjuvanted TIV were infected with influenza A or B, or both (274 [1.27%, 95% CI 1.12-1.43] of 21573 vs 310 [1.44%, 1.29-1.61] of 21482; relative efficacy 12-11%, 95% CI -3.40 to 25.29; superiority not established). Fewer participants in the year 1 efficacy cohort given AS03-adjuvanted TIV than non-adjuvanted TIV were infected with influenza A (224 [1.04%, 95% CI 0.91-1.18] vs 270 [1 26, 1.11-1.41]; relative efficacy 17.53%, 95% CI 1.55-30.92) and influenza A H3N2 (170 [0.79, 0.67-0.92] vs 205 [0 95, 0.83-1.09]; post-hoc analysis relative efficacy 22.0%, 95% CI 5.68-35.49). Interpretation AS03-adjuvanted TIV has a higher efficacy for prevention of some subtypes of influenza than does a non-adjuvanted TIV. Future influenza vaccine studies in elderly people should be based on subtype or lineage-specific endpoints.