No description available.

No description available.

Summary Introduction: This thesis reports the surveillance response of the EUROCAT network of population based registries for surveillance of congenital anomalies to the 2009 H1Nl influenza pandemic. Due to the controversy around vaccinating 1st trimester pregnant women, the relationship between pandemic influenza outbreaks and congenital anomalies (CA) deserves attention. The purpose of this thesis is to improve public health practice by investigating the effect of the 2009 HiNl influenza pandemic on CA prevalence in Europe and exploring the pandemic vaccination policy and policy making with respect to pregnancy. Methods: Systematic review of 44 studies and meta-analysis of 21 studies on influenza and CA was performed. Since seasonality can confound analyses of influenza and CA, pre-pandemic original analysis of seasonality of births and CA was performed according to estimated month of conception, enrolling 20 EUROCAT congenital anomaly registries from 14 countries covering 366,000 births annually over 2000-2009. The effect of 2009 H1N1 pandemic influenza and seasonal influenza exposure during critical periods of organogenesis on CA prevalence was analysed by an ecological time series study, enrolling 15 registries from 10 countries covering 364,000 births annually over 2007-2010 and using WHO influenza surveillance data. Case-malformed control analysis was performed to compare exposure to pandemic influenza and its co-exposures between CA subgroups, enrolling 11 registries from 9 countries covering 213,000 births annually over 2009-2010. European pandemic policies with respect to pregnancy were surveyed by sending questionnaires to EUROCAT registry leaders and European Departments of Health. The results of the policy survey were investigated by policy analysis using content analysis of the minutes and related documents describing British and Dutch decision making processes. Results: Meta-analysis of literature associated a wide range of CA with influenza including any CA (OR 2.25, 95% CI: 1.77-2.85), congenital heart disease (1.84, 1.50-2.26), neural tube defects (3.61, 2.23-5.82) and cleft lip (3.00, 2.12-4.25). Seasonality analysis detected seasonal peaks in the prevalences of all births (+4% compared to average in late September conceptions) and a number of CA including CA previously associated with influenza (+3%, June). Solely situs inversus prevalence was increased during influenza season (+35%, December). Ecological time series analysis did not detect an increased prevalence of CA 24 among pregnancies conceived during the 2009 influenza pandemic (ARR 1.03, 0.97-1.09). Situs inversus prevalence was increased during 2007-2010 influenza outbreaks (ARR 2.06, 1.02-4.15) and statistically non-significant increases were witnessed for all situs anomalies. Case-malformed control analysis detected an increase in upper respiratory tract exposure among pregnancies resulting in neural tube defects (AOR 2.00, 1.03-2.90), but this was based on data from a single registry and contradicted the findings of the ecological time series analysis (ARR 0.91, 0.70-1.19). Very few individual exposures had been recorded for pandemic influenza vaccine (n=5) or neuraminidase inhibitors (n=6) among CA registrations. Replies to the policy survey were received from 24 European countries of which 12 solely vaccinated 2nd and 3rd trimester pregnant women while 8 vaccinated pregnant women of any trimester. These differences were further investigated by policy analysis and limitations in evidence (in particular with respect to pandemic influenza vaccine and CA) were identified to be an important factor in causing these varying vaccination policies. Discussion and conclusion: The ecological time series analysis of the pandemic season did not find the increase in CA expected on the basis of the meta-analysis results, possibly due to biases leading to dilution and/or differences between populations, or lack of an underlying relationship with influenza. Ascertaining influenza exposure and unravelling influenza exposure from its co-exposures such as antivirals is a challenge for epidemiological studies. Based on the results of the seasonality study and ecological time series study, a hypothesis that situs anomalies can be related to either influenza or its co-exposures has been generated. The unknown effect of pandemic influenza vaccine on the fetus was a problem for European pandemic policy making and investigating the relationships between antivirals, influenza vaccine and CA using improved exposure data are urgent research priorities.

Emerging infectious diseases (EID) are increasing in frequency with zoonoses originating in wildlife posing the greatest threat to global health. Highly pathogenic avian influenza (HP AI) strain H5N1 is the most expensive and widespread zoonotic disease to emerge recently. First detected in China in 1996, the virus subsequently spread across Asia, Europe, Africa and the Middle East resulting in tens of millions of animal deaths, primarily poultry as well as 329 fatal human cases. This thesis utilises a range of techniques from multiple disciplines to address questions relating to EID epidemiology and control through to the impacts of HPAI H5N1 at the household level within Vietnam. The methodologies employed include adapting an analytical framework to address a public health problem, semi-structured interviews within central Hanoian and rural Vietnamese households, structured questioning, direct surveys of the live bird markets and key-informant interviews. This thesis has identified rapid growth in the trade and exploitation of birds for cultural and recreational human practices within Vietnam which involve several HP AI H5N1- susceptible species and promote ideal conditions for pathogen transmission. We estimate that three million birds annually are extracted from the wild to supply religious merit release practices in Vietnam alone. At the household level, poultry was found to be an important protein source for urban Vietnamese households and kept primarily for consumption by the majority of rural households. We found urban poultry consumers choose to take protective actions to limit direct exposure to HP AI H5N1 whilst rural households choose to persist with the keeping of household poultry flocks despite the potential risks to household health and livelihood stability. We also identify substantial under-reporting of HPAI H5N1 outbreaks to global surveillance databases and consider the implications of this for HPAI H5N1 surveillance programmes. The thesis concludes by bringing together the different aspects of HPAI H5N1's impacts within Vietnam and emphasises the value of multidisciplinary approaches to studying the impacts of EIDs.

Influenza virus causes a highly contagious respiratory disease whose morbidity and mortality is best prevented by vaccination and the need for a rapid and scalable vaccine response to emerging influenza virus strains is widely recognised. In this study, the haemagglutinins (HAs) of human HI and H3 influenza viruses and avian H5 influenza virus were produced as recombinant fusion proteins with the human immunoglobulin Fe domain. Recombinant HA-human immunoglobulin Fe domain (HA-HuFc) proteins were secreted from baculovirus-infected insect cells as glycosylated oligomeric HAs of the anticipated molecular mass, which agglutinated red blood cells and were purified by protein A chromatography. Purified protein was used to immunise mice in the absence of adjuvant. Immunogenicity was demonstrated for all subtypes, with the serum samples demonstrating subtype- specific haemagglutination inhibition. A degree of heterosubtypic cross reaction between H5 antibodies and HI HA was also shown in this study, plausibly due to structural relatedness of the HAI domains. A Th2 immune response was dominant with some ThI response, typical of the immune responses described for other influenza vaccines and for soluble oligomeric HA. Dose ranging immunisation studies demonstrated significant seroconversion down to 100ng per dose, the human equivalent dose of 24ugs as calculated by body surface area, which is lower than that observed for a recombinant HA in human studies. The polyvalent serum response was directed to the RBS of the HA as reported for experimental influenza infection and included neutralising antibodies. Mutations introduced into the RBS of the H5 HA altered the antigenicity of the virus protein and appeared to elicit antibodies more efficient for binding to human origin influenza subtypes. Epitope specificity studies identified the top of the HAl domain as the target of the induced antibodies, which broadly correlated with the epitope mapping for the H5 HA of several influenza strains. Thus, HuFc-tagged HAs expressed in insect cells are potential candidates for gene-to-vaccine approaches to influenza vaccination.

Highly pathogenic avian influenza (HP AI) H5Nl viruses cause devastating fatal disease in chickens, reaching nearly 100% mortality rate within 48 hours. In contrast, ducks show little or no clinical signs. This suggests that there are host species-specific differences in the innate immunity beMeen these two species. Large quantities of HP Al H5Nl viruses have been reported in skeletal muscle of infected chickens and ducks. An in vitro model comprising primary skeletal muscle cells from these two avian species was established to study host immune response 'to avian influenza. Infection with a low pathogenicity avian influenza (LP AI) virus in differentiated chicken muscle cells resulted in a vigorous induction of interferon-fJ and several interferon-inducible genes while two HP Al H5N 1 viruses effectively inhibited this antiviral response. Duck muscle cells did not appear to produce large amounts of type 1 interferons independent of the virus used.

Concomitant with the Spanish influenza and extending to the 1920s encephalitis lethargica (EL), also termed von Economo's disease or sleepy sickness, occurred in epidemic proportion. A previous epidemic of encephalitis lethargica also appeared at the time of 1889 influenza pandemic. Certain clinical and pathological features of encephalitis lethargica of 1916-1920 strongly suggested a viral aetiology and subsequently it was suggested that influenza virus was the responsible pathogen (Ravenholt & Foege, 1982). Sensitive molecular techniques are now available which can be applied both to fresh tissue and old formalin fixed and paraffin blocked sections using RT-PCR. Recent genetic analyses of haemagglutinin (HA) and neuraminidase (NA) genes of 1918 influenza virus have failed to identify virulence motifs (Taubenberger et al., 1997 & 1999; Reid et al., 1999,2000 & 2002). These methods were applied to eight brain samples from patients who died of encephalitis lethargica in 1916-1920 for the presence of influenza genes. The sections of brain from eight archival, unique cases of EL had been neuropathologically reviewed and the diagnosis confirmed prior to analysis using established reverse transcription-polymerase chain reaction (RT-PCR). Although our genetic study detected ß-actin genes in the archived brains, we found no evidence of influenza genes. An animal model of influenza neuropathogenesis was established for investigations of the localisation of influenza genes in brain tissue using in situ hybridisation (ISH). Influenza gene sequences were detected in ependymal cells, choroid plexus, hippocampal neurons, periventricular regions of the lateral ventricle and cerebral aqueduct, and cells of the raphe nucleus. A persistent influenza infection in the brains of infected mice is reported. Therefore short influenza sequences could, in theory and based on a simple longevity comparison of man and mouse, be detected in brain tissues of infected subjects for a period of 1.9 -6.7 years after its initial exposure to influenza infection.

Limited understanding of influenza transmission has been a frequent obstacle during the development of pandemic influenza infection prevention and mitigation strategies. The science is hotly debated, especially the relative importance of transmission via large droplets or aerosols. Clarification of the relative importance of different modes of transmission is critical for the refinement of evidence-based infection control advice and has been called for by the European Center for Disease Control (ECDC), the World Health Organization (WHO), and the US Institute of Medicine. The primary aims of this thesis were to investigate influenza transmission; i) by obtaining data concerning viral shedding and the presence of influenza virus in the near environment of infected individuals and ii) through the exploration of a human challenge model to study transmission. Two major clinical studies have been performed; • Shedding and environmental deposition of novel A (H1N1) pandemic influenza virus. The primary aims of the study were to correlate the amount of virus detected in a subject’s nose with that recovered from his/her immediate environment (on surfaces and in the air) and with symptom duration and severity. Adults and children, both in hospital and from the community, who had symptoms of influenza infection were enrolled. Information about symptoms was collected and samples were taken including nose swabs, swabs from surfaces and air samples. Forty two subjects infected with influenza A(H1N1)pdm09 were recruited and followed up. The mean duration of nasal viral shedding was 6.2 days (by PCR) and 4.6 days (by culture). Over 25% of cases remained potentially infectious for at least 5 days. Symptom scores and viral shedding were poorly correlated. From surface swabs collected in the vicinity of 40 subjects, 15 (38%) subject locations were contaminated with virus. Overall 36 of 662 (5.4%) surface swabs taken were positive for influenza, two (0.3%) yielded viable virus. Subjects yielding positive surface samples had significantly higher nasal viral loads on illness Day 3 and more prominent respiratory symptom scores. Room air was sampled in the vicinity of 12 subjects and PCR positive samples were obtained from five (42%). Particles small enough to reach the distal lung (≤4µm) were found to contain virus. • Use of a human influenza challenge model to assess person-to-person transmission: Proof-of-concept study. The primary aim of this study was to establish that an experimentally induced influenza infection is transmissible. Healthy subjects deemed sero-susceptible to influenza A/H3N2/Wisconsin/67/2005 were intranasally inoculated (Donors) and when symptoms began, further sero-susceptible subjects (Recipients) were exposed to Donors during an ‘Exposure Event’. Subjects were in close contact, e.g. playing games and eating meals together, for a total of 28 hours during a 2 day period. Samples were collected to confirm infection status. Among 24 healthy adult subjects, nine were randomised to the ‘Donor’ group and 15 to the ‘Recipient’ group. Following inoculation 5 out of 9 Donors (55%) developed illness and 7 out of 9 (78%) were proven to be infected. After exposure, 5 out of 15 Recipients developed symptoms and 3 out of 15 were proven to be infected. Three others were found to be non sero-susceptible prior to exposure. The overall attack rate in Recipients was 20% but was 25% after adjustment for pre-exposure immunity. The contact, droplet and aerosol routes of influenza transmission are all likely to have a role. This thesis shows that transmission of influenza via surfaces may be less important than current infection control policies and public guidance documents imply. Air sampling results add to the accumulating evidence that supports the potential for aerosol transmission of influenza. The human challenge model could be used to investigate routes of influenza transmission further and a study funded by the Centers for Disease Control (CDC) is planned.