Pathogenic avian influenza A virus H5N1 has caused outbreaks in poultry and migratory birds in Asia, Africa, and Europe, and caused disease and death in people. Although person-to-person spread of current H5N1 strains is unlikely, the virus is a potential source of a future influenza pandemic. Our aim was to assess the safety and immunogenicity of a vaccine against the H5N1 strain.

We did a randomised, open-label, non-controlled phase I trial in 300 volunteers aged 18–40 years and assigned one of six inactivated split influenza A/Vietnam/1194/2004 (H5N1) influenza vaccine formulations, comprising 7·5 μg (with adjuvant n=50, without adjuvant n=49), 15 μg (n=50, n=50), or 30 μg (n=51, n=50) of haemagglutinin with or without aluminium hydroxide adjuvant. Individuals received two vaccinations (on days 0 and 21) and provided blood samples (on days 0, 21, and 42) for analysis by haemagglutination inhibition and microneutralisation. We recorded all adverse events. Analyses were descriptive.

All formulations were well tolerated, with no serious adverse events, few severe reactions, and no oral temperatures of more than 38°C. All formulations induced an immune response, and responses were detectable in some individuals after only one dose. The adjuvanted 30 μg formulation induced the greatest response (67% haemagglutinin-inhibition seroconversion rate after two vaccinations). Adjuvant did not improve the response to the lower doses. Two vaccinations of non-adjuvanted 7·5 μg, adjuvanted 15 μg, or non-adjuvanted 15 μg seroconverted more than 40% of participants (haemagglutinin-inhibition test only). Haemagglutinin inhibition and neutralising results were comparable.

A two-dose regimen with an adjuvanted 30 μg inactivated H5N1 vaccine was safe and showed an immune response consistent with European regulatory requirements for licensure of seasonal influenza vaccine. We noted encouraging responses with lower doses of antigen that need further study to ascertain their relevance for the choice of the final pandemic vaccine.