Development and evaluation of influenza pandemic vaccines.

The spread of highly pathogenic avian influenza H5N1 among poultry and migratory birds, and its transmission to more than 140 human beings since 2004 has prompted concerns that the next influenza pandemic is imminent. Vaccination is the principal strategy to combat pandemic influenza; however, a pandemic will create huge surges in vaccine demand. Although annual influenza vaccine production is a well-planned process, rapid bulk manufacture of avian influenza vaccines of unknown growth and processing characteristics may be challenging. New approaches in immunisation strategies may be required to optimise protection of immunologically unprimed individuals.

On Nov 2–3, 2005, WHO held a meeting to review progress in development of vaccines for pandemic influenza. Clinicians, scientists, manufacturers, and regulators presented plans for production, evaluation of safety and immunogenicity, and licensing of vaccines against pandemic influenza. Here we present a summary of discussions at the meeting.

Highly pathogenic avian influenza viruses can be manipulated by reverse genetics to produce viruses that retain the original virus's novel surface antigens, but which are non-pathogenic and suitable for bulk production of vaccine. Reverse genetics systems are well established in a number of WHO-affiliated laboratories and are reliable for the manipulation of H5N1 viruses; however, experience with other influenza subtypes is limited and may be less successful. Before release from high biocontainment, reassortant viruses derived by reverse genetics from highly pathogenic avian influenza viruses must undergo stringent safety testing. Currently it can take 8–12 weeks following receipt of wild-type virus to the release of a safe reference vaccine virus to manufacturers. In some countries, guidelines governing “genetically modified organisms” impose additional regulation on work with viruses generated by reverse genetics. This time could be reduced substantially if regulatory authorities agree to allow release of these viruses earlier in the process. Because of intellectual property rights surrounding this technology, vaccine manufacturers may need to obtain appropriate licences in advance. To reduce delays in developing reference viruses, it is essential that avian influenza isolates from poultry and human beings are shared quickly with WHO laboratories for antigenic characterisation.

The European Medicines Agency and US Federal Drug Administration both recognise that approval for pandemic influenza vaccines may be granted during the early phase of an evolving pandemic. The European approach is the development of “core dossiers” that contain preclinical and safety data pertinent to existing influenza vaccines. Dossiers can be submitted in advance, and the strain or subtype variation added once the pandemic strain is known. Safety and immunogenicity assessments of adjuvanted or new formulations should be done during interpandemic periods. Since any fast-track approval must not compromise safety, robust post-marketing surveillance must be an integral part of licensing procedures.

Following the first H5N1 outbreak among human beings in Hong Kong in 1997, two doses of conventional surface-antigen H5N3 vaccine were found to be poorly immunogenic in unprimed human beings. The addition of the MF59 adjuvant substantially boosted immune responses and induced cross-reactive antibodies to antigenically unrelated H5 strains.6, 7 A range of clinical trials assessing 2004–05 H5N1 vaccine candidates is planned. Inactivated egg-grown subunit and whole-virion vaccine, cell-culture-derived vaccine, adjuvanted (aluminium hydroxide, aluminium phosphate, and MF59), and live attenuated virus formulations are planned for clinical evaluation in 2006. Additionally, assessment of subunit, adjuvanted, virosomal, and whole-virus H7 and/or H9 vaccines candidates will be completed in 2006.

A reverse genetics subunit H5N1 vaccine, based on A/Vietnam/1203/04 (H5N1), produced by Sanofi Pasteur using their regular production process, was assessed in a randomised trial among healthy adults in the USA. Additional studies are underway in children and elderly adults. Early analysis indicates that although the vaccine was well tolerated, immune responses were induced only at higher doses, with 11%, 4%, 48%, and 67% of 117 recipients seroconverting by neutralising antibody after two doses containing 7·5 μg, 15 μg, 45 μg, or 90 μg of antigen, respectively. It is clear from these data that dose-sparing formulations are urgently needed.

Hungarian investigators (Omnivest, Budapest) of an aluminium phosphate adjuvanted whole-virion H5N1 vaccine reported promising results. A single dose containing 30 μg H5 antigen induced seroconversions by haemagglutinin inhibition in 18 (90%) of 20 recipients assessed to date. Larger studies are underway. Alum as an adjuvant is widely available, and when combined with whole virus H2N2 and H9N2 vaccine, allowed an eight-fold reduction in antigen content while maintaining immunogenicity. Randomised trials of plain and alum-adjuvanted reverse-genetic H5N1 vaccine are underway in France and Australia and data analysis is expected by early 2006.

Live attenuated H5N1 and H9N2 influenza vaccines are protective in animal models and may have the advantage of inducing humoral and mucosal immunity. Early results from the clinical evaluation of a live attenuated H9N2 candidate suggest most recipients develop serum antibodies after two doses.

Finally, clinical evaluation of dose-sparing and adjuvanted vaccine formulations is urgently required to provide information to optimise antigen use. Future studies of adjuvanted vaccines should contain a non-adjuvanted comparator arm, and compare leading formulations in head-to-head studies. Wherever possible, serum samples from vaccine recipients should be stored so that cross-reactivity with emerging variant strains can be explored in future studies. The international cooperation required to successfully control the severe acute respiratory syndrome epidemic should be maintained when planning for an influenza pandemic. The clinical studies underway will provide important information to rationalise immunisation strategies for pandemic influenza, and WHO will review additional data when available in 2006.