Lessons learned from the H1N1 2009 pandemic

Abstract

On June 11, 2010 approximately 80 delegates drawn from government, industry and academia gathered in Singapore for the first World Influenza Congress Asia held in association with the fourth Annual World Vaccine Congress Asia 2010. A major focus of the meeting was sharing of experiences relating to the recent H1N1 2009 pandemic.

Influenza Communication Issues

There is a need for better influenza education of health professionals and the public to achieve the goal of the World Health Assembly resolution made in May 2003 which sought to attain a minimum of 75% vaccination coverage of the elderly by 2010 (Bram Palache, Solvay Pharmaceuticals, Netherlands). This target has been achieved by very few countries in Europe with Spain and the UK having rates approaching this target but other countries including Austria, Czech Republic and Poland achieving coverage in the over 65 year old population of less than half this target. The concern was expressed that despite many endorsed guidelines to improve influenza prevention, these recommendations are often not trusted or acted on by the public, representing a communication failure. Poor communication may also underlie poor influenza vaccine uptake by health care workers despite recommendations that they be vaccinated to protect the high-risk patients with whom they may have contact. To improve influenza prevention, communication and education from public bodies driving these recommendations needs to be improved, including the adoption of new communication modalities, using social media channels on the internet.

H1N1 2009 Epidemiology

The recent experience with the H1N1 2009 pandemic highlights future challenges for influenza. Hong Kong being an island with a strong medical epidemiology tradition presents an ideal environment in which to characterise the epidemiology of infectious diseases such as influenza. Seroprevalence data from Hong Kong as assessed by microneutralization assays after the peak of the H1N1 pandemic had passed, showed that the largest burden of pandemic infections occurred in young children with H1N1 2009 seroprevalence rates of ∼50% for 5–14 year olds, ∼15% for 15–19 year olds, 10% for 20–29 year olds and 5–8% for >30 year olds (J.S. Malik Peiris, University of Hong Kong). Hospital admissions represented approximately 1:100 of the case rate, and intensive care unit admission and fatality rates increased with increasing subject age. Mortality cases were 70 times higher in adults than children despite the far higher case rate in children. Overall, there was a comparable fatality rate in the elderly from H1N1 2009 to what is seen in Hong Kong with a typical influenza season. The H1N1 2009 viral load and symptom scores were comparable with seasonal influenza and H1N1 2009 tissue tropism in human cell lines was not different to seasonal strains. It was noted that H1N1 2009 has now re-entered the pig population and as it is genetically unstable in pigs is generating reassortments that may re-emerge into the human population in the future. The H1N1 2009 pandemic highlighted the need for better surveillance and epidemiological systems to generate more reliable early estimates of pandemic severity to assist planning including quantification of likely vaccine requirements. There is an ongoing problem of mismatch between pandemic vaccine demand and supply, with high early demand met with insufficient supply followed by excess supply with little demand after the pandemic peak had passed. The message of the need for more disease burden research and better systematic laboratory-based surveillance recurred throughout the meeting (Mark Simmerman, Sanofi Pasteur, Thailand).

The Southern Hemisphere was the first to experience a full H1N1 2009 season providing some useful epidemiological insights. The Australian experience was that the H1N1 virus effectively displaced all other circulating seasonal H1N1, H3N2 and B influenza strains, such that >90% of laboratory influenza isolates were H1N1 2009 (Nikolai Petrovsky, Flinders Medical Centre, Adelaide). The major impact on the health system had been on Intensive Care Units with extremely heavy utilization of ventilator beds and with the majority of seriously ill subjects being younger than those typically being seen during a seasonal influenza outbreak. The majority of seriously ill subjects had chronic medical co-morbidities, were pregnant or young children. In all, Australia had only recorded ∼180 deaths from influenza-related illness throughout the entire winter season, which is dramatically lower than the estimates of 3,000 deaths in a typical Australia influenza season. The explanation for this may lie in the fact that the elderly who are particularly vulnerable to seasonal influenza deaths, were spared by the H1N1 2009 virus to which they had underlying immunity and in addition were saved from exposure to the normally lethal seasonal strains thanks to the effect of the fast-spreading H1N1 2009 blocking the spread of these other influenza viruses in the community.

Inactivated Influenza Vaccines

The Asia-Pacific region remains poorly served with influenza vaccine manufacturing capacity relative to population, with current influenza vaccine production restricted to a small number of countries including Japan, China, South Korea and Australia. Nevertheless initiatives are underway to improve this situation with recent influenza vaccine projects commenced in India and Taiwan, amongst others. Green Cross in Korea presented their experience with introducing a H1N1 2009 vaccine in response to the pandemic. Production of the Korean vaccine commenced with the initial seed stock being received from NIBSC on June 8th 2009 (BG Reid, Green Cross, South Korea). Clinical studies commenced on September 15, 2009 and approval for marketing of the new H1N1 2009 vaccine was received on October 21, 2009. The vaccine itself is a traditional egg-grown inactivated virus vaccine that, similar to the data published on other inactivated H1N1 vaccines, induced seroprotection rates of 75–95% at doses from 15–30 ug HA. Children under the age of three had a much lower rate of seroconversion and required two vaccine doses to achieve a seroprotection rate of >70%. Green Cross also developed an MF59-adjuvanted H1N1 2009 vaccine called GreenFlu-S+ with a reduced 3.7 ug HA dose plus 0.125 ml of MF59 adjuvant. The MF59-adjuvanted vaccine achieved similar seroconversion and seroprotection to the unadjuvanted 15 ug HA vaccine.

Pre-pandemic Influenza Vaccines

Alternative vaccine approaches are clearly needed to avoid the current problem of pandemic vaccines being supplied too late to have any impact on a pandemic The H1N1 pandemic strain spread globally extremely quickly entering most countries in less than 1 to 2 months. Despite excess global seasonal vaccine production capacity, pandemic vaccine capacity provided less that 10% of world population coverage with most vaccine supply secured by a small number of developed world countries. The supply of H1N1 pandemic vaccine was less than previously foreseen by expert committees with only 534 million doses being supplied within the first six months of the pandemic against prior estimates that approximately 2.5 billion doses should be available by this time into a pandemic (Klaus Stohr, Novartis Vaccines and Diagnostics). The peak incidence of H1N1 cases in the United States had passed by November 2009 at a time when vaccine supply was still minimal. By the time that plentiful H1N1 vaccine was available, the new case rate had already dropped to low levels. Therefore over 40% of H1N1 cases in the USA occurred at a time when no meaningful vaccine supply was available. Ultimately, one solution advanced by Novartis and others to alleviate pandemic vaccine supply constraints is the use of pre-pandemic vaccines to pre-empt the pandemic. The difficulty in the prepandemic approach is the need to correctly predict the likely pandemic strains in advance. The fact that the influenza virus is inherently unpredictable and continues to surprise was shown by the fact that H1N1 was not even on the list of WHO pandemic influenza priorities which included H5, H7, H9, H6, H4 and H2 strains. This makes it extremely difficult for a pre-pandemic vaccine strategy to work although data from Novartis and others does suggest that priming with an unmatched pre-pandemic vaccine may improve later responsiveness to a pandemic vaccine dose, enabling only a single rather than a double dose of pandemic vaccine to achieve protection. Interestingly, in the case of H1N1 2009 only a single immunization was required to achieve robust seroprotection, consistent with H1N1 2009 behaving more like a typical moderately drifted seasonal influenza strain than a true pandemic strain caused by a completely new serotype to which the human population has no prior immunity.

Universal Influenza Vaccines

In addition to consideration of pre-pandemic immunization, measures to improve pandemic protection include technologies for faster vaccine manufacture, stimulating increased vaccine manufacturing capacity through improving seasonal vaccine uptake and research into new vaccine approaches. Given the challenge of predicting the next pandemic strain, a universal cross-protective influenza vaccine remains the Holy Grail. An Australian company, Gamma Vaccines, is developing a nasally administered flu vaccine under the name of Gammaflu™ based on gamma irradiated influenza virus (Tim Hirst, Gamma Vaccines). Mice were protected against PR8 virus challenge by administration of just one dose of gamma-irradiated virus whereas Fluvax, the inactivated CSL seasonal vaccine, provided no protection. Gammaflu™ induced cross-protection against H5N1, H3N2 and H2N2 despite an H1N1 strain being used for the vaccine antigen. Protection by intranasal Gammaflu was not transferrable to naive animals by transfer of serum antibody or B cells but only by transfer of T cells. Protection induced by the gamma irradiated virus was shown to rely on a CD8⁺, perforin-secreting, cytotoxic T-lymphocyte population. An argument in favour of gamma irradiation for viral inactivation is that this allows the virus to remain functional and still invade cells, as compared to the use of formalin or beta propriolactone which completely inactivate the virus by cross-linking the viral proteins and RNA.

Another more advanced universal influenza vaccine approach is the DepoVaccine FP-01 being developed in the UK (Carlton Brown, Immune Targeting Systems). This universal vaccine approach uses 6 different C8F17 fluorocarbon-conjugated long peptides (35 amino acids) which are highly conserved and known to contain high density clusters of CD4⁺ T-helper and CD8⁺ CTL epitopes from the major HLA sub-types to provide broad cross protection. A bioinformatic approach was taken to identify high affinity HLA-binding sequences from PA, PB1, PB2, M1 and nuclear proteins conserved across all H1, 2, 3, 5, 7 and 9 flu strains circulating pre- and post-2,000 in human, swine and avian species. Out of 20 initial antigens that met the required profile, 14 were eliminated due to manufacturing, stability or co-formulation related issues. The final vaccine is a freeze-dried formulation containing six different DepoVaccine antigens (>95% purity) known to be highly stable and readily characterized using industry standard methods. When injected into mice, this resulted in a 5–6 fold increase in γ interferon ELISPOT counts and was equivalent to the increase when the native peptides were injected in complete Freund's adjuvant. As human immune responses to peptides are highly HLA-restricted, it was reassuring that peripheral blood mononuclear cells from 97% of more than 200 human subjects recognized at least one of the antigens included in the vaccine and on average the response was 5 out of 6 antigens with multi-epitope responses seen per peptide per subject. The rationale for the fluorocarbon attachment in the vaccine is to create a soluble formulation which forms a short-term antigen depot upon injection, improves antigen half-life, enhances antigen endocytosis and antigen cross presentation to CD8 T cells. The company filed its Clinical Trial Application (UK) in June 2010 to commence human clinical trials of this vaccine.

Live Attenuated Influenza Vaccines

Live attenuated influenza vaccines (LAIV) are an alternative to the use of inactivated influenza vaccines, albeit only in individuals from 2–49 years, due to a high rate of respiratory side effects in children <2 years and poor efficacy in adults >50 years. In response to the H1N1 2009 pandemic, Medimmune produced a monovalent nasal LAIV vaccine for use in eligible individuals from 2–49 years old (Filip Dubovsky, MedImmune, USA). The seasonal LAIV vaccine has previously demonstrated efficacy in young children of ∼70% against culture-confirmed influenza-like illness caused by match strains but lower rates of efficacy have been seen in adult studies. The first round of vaccine reassortments for the H1N1 2009 LAIV vaccine showed suboptimal egg growth and the vaccine also stuck in the filters used in purification. In May–June 2009 MedImmune therefore generated a number of mutant viruses, introducing mutations into the wildtype HA sequence. The mutant California-09 V5119E/186D virus grew well in eggs, was filterable and reacted with wild type immune sera. V5119E/186D was therefore selected for the H1N1 2009 LAIV vaccine. Although the resulting LAIV vaccine induced HA inhibition titres in ferrets, subsequent placebo-control clinical trials performed in healthy adults and children showed a failure of the LAIV vaccine to induce seroprotection in the human subjects. After one LAIV dose, a maximum seroconversion rate of just 6.1% was seen in 18–49 year olds and 11.1% in 2–17 year olds. Seroconversion rates remained extremely low even after a second LAIV immunization. Thus, the Medimmune H1N1 2009 LAIV vaccine did not meet the current WHO or EMEA HAI-based criteria for influenza vaccine efficacy. MedImmune propose that seroconversion doesn't correlate with efficacy of the LAIV vaccine and suggest a possible T-cell mediated mechanism of LAIV protection. Seasonal Flumist has no efficacy in adults over 50 years and the H1N1 2009 LAIV vaccine was restricted to healthy 2–49 year olds. Ongoing non-randomized community studies may shed more light on the significance, in respect of efficacy against H1N1 2009, of the failure of the LAIV vaccine to induce H1N1 2009 seroconversion in most immunized human subjects.

Recombinant Influenza Vaccines

Given the problems with of slow and inadequate pandemic vaccine supply using traditional inactivated virus vaccines grown in egg or cell culture, and questions of efficacy and age range limitations on use of LAIV vaccine, alternative approaches are clearly needed. Medicago, a company based in Canada, has developed a HA-based vaccine genetically engineered in plants (Andy Sheldon, Medicago). The expression system uses agrobacteria for transfection of tobacco plants with the gene for HA. The plant cells then excrete the HA as a stable virus-like particle. This HA VLP vaccine when combined with aluminium hydroxide adjuvant induced cross protection against an H5N1 A/Vietnam strain when ferrets were immunized with the A/Indonesian strain. A Phase 1 human study of the same H5N1 vaccine, used two intramuscular doses of 5, 10 or 20 ug HA plus aluminium hydroxide adjuvant, and achieved a seroconversion rate of 58% and seroprotection rate of 50%, with the 20 ug dose. This was said to compare favourably with published seroconversion rates of Novavax's H5N1 VLP vaccine and Baxter's H5N1 vaccine.

In July 2009, Australia was host to the world's first clinical trials of two very different H1N1 2009 vaccines. Whilst one of these was a traditional inactivated egg-manufactured vaccine produced by CSL, the other vaccine used in a trial conducted by Vaxine Pty Ltd., was a recombinant HA antigen produced by Protein Sciences Corporation that was combined with a novel polysaccharide adjuvant (Nikolai Petrovsky, Vaxine). Whilst head-to-head data of recombinant HA versus inactivated influenza vaccines remains limited, both vaccine approaches were clearly successful against H1N1 2009, with subjects in both trials achieving seroconversion after one dose, in contrast to the poor seroconversion response with Medimmune's H1N1 2009 LAIV. The speed with which the recombinant HA pandemic vaccine was able to be produced in response to the declaration of a pandemic, its lack of dependency on egg-adapted recombinant virus stocks and its ready scalability, must make this a leading contender technology for production of future pandemic influenza vaccines, a fact recognized by recent US government funding initiatives to help implement this technology on a commercial scale.

Influenza Vaccine Adjuvants

A major barrier to development of influenza vaccines based on recombinant HA has been the perception that they are insufficiently immunogenic, given that they can require higher doses of HA to achieve the same antibody titres as inactivated virus vaccines. Aside from the immunogenicity issue, recombinant HA vaccines offer numerous advantages over inactivated vaccines for pandemic use, being faster to produce, more scalable and not vulnerable to disruptions in the egg supply. In a major positive finding, the immunogenicity disadvantage of recombinant HA can be overcome by the inclusion of Advax™, a highly effective polysaccharide adjuvant developed by Vaxine, that in animal and human trials markedly increased seroconversion and seroprotection rates to a H1N1 2009 vaccine based on recombinant HA (Nikolai Petrovsky, Vaxine, Australia). An advantage of the Advax-adjuvanted influenza vaccine approach that is not visible from seroprotection data, is its ability as shown in animal challenge studies to induce protective T-cell responses that enhance influenza protection over and above the level of neutralizing antibody alone. As shown in clinical studies, Advax was also able to enhance HAI titers to seasonal trivalent influenza vaccine.

Assessing Influenza Vaccine Efficacy

One of the biggest challenges for new influenza vaccines is to demonstrate efficacy. This is particularly true of new vaccines utilizing novel mechanisms of protection, e.g., those like Immune Targeting Systems's vaccine that induce T-cell immunity rather than neutralizing antibody. Retroscreen Virology has expertise in undertaking human influenza challenge studies to more accurately measure vaccine induced protection (Robert Lambkin-Williams, Retroscreen Virology, UK). As well as having tested the efficacy of traditional inactivated influenza vaccine products, Retroscreen tested the efficacy of a DNA-based influenza vaccine developed by PowderMed. This study confirmed the DNA vaccine had an efficacy of ∼50% against a homologous viral challenge. Retroscreen has also developed human challenge models for respiratory syncytial virus, rotavirus and norovirus.

Conclusions

This was a timely influenza meeting given the diminishing H1N1 2009 pandemic and the increasing availability of data providing retrospective insight into the nature of this atypical influenza pandemic. The overall conclusion was that there was still a lot of work remaining to be done to bolster defences against the next pandemic that may not be nearly as benign as H1N1 2009. Areas identified for attention included better influenza surveillance systems, tools to provide earlier assessment of disease severity, faster vaccine manufacturing technologies, improved vaccines based on recombinant antigens and new adjuvants and development of a universal influenza vaccine, if this can realistically be achieved. Whilst, the Asian region remains grossly undersupplied with respect to pandemic influenza vaccine manufacturing capacity, new regional vaccine initiatives show this does not need to remain the case providing sufficient government support is forthcoming, and regional seasonal influenza vaccine usage is increased from its current low base to a level that makes these new influenza vaccine manufacturing operations viable and sustainable.