| Summary: | Introduction: Repeat infections with influenza A occur because of the continual evolution of the virus. A specific humoral response occurs after each infection and this develops into an immunological profile for each individual based on their prior exposure history and the crossreaction between antigenically similar viruses. Understanding how this profile changes following acute infection is important for interpretation of seroepidemiological studies. Methods: This observational study was designed to investigate non-severe influenza and influenza like illness in a tropical, urban primary care setting. A prospective, observational study of patients with ILI in Ho Chi Minh City, Vietnam has been running since August 2013. Influenza A & B PCR and antibody testing to a panel of 11 human and 5 avian strains is performed using a novel protein microarray technique. A subset of subjects are followed up clinically and serologically for 7 months. This doctoral research is designed to address questions around the effect of cross-immunity from previous influenza infection on susceptibility to currently circulating strains, understand the short term antibody dynamics of both current and historical strains post infection and determine optimal serological determinants of recent infection. Results: 953 ILI patients were recruited between August 2013 & May 2015. 274 and 136 subjects had influenza A and B respectively. Three peaks of influenza activity were detected, H3N2 peak between April and June 2014, influenza B peak in July to December 2014 and mixed H3N2 and H1N1 peak March to May 2015. Lower baseline titre was associated with higher risk of influenza but response to H3N2 2005 and 2009 were most predictive of current susceptibility even when original antigenic sin was taken into account. Levels of protection offered by a fixed titre changed depending on force of infection. 186 ILI patients were recruited to the longitudinal study. The largest response was within subtype which peaked at around 30 days. A boosting of historic response was also seen. The response in historic strains waned quicker than response to recent strains. A smaller but significant between subtype increase in titre was also detected. Following an acute rise after infection within six months most individuals had returned to a standard rate of decline of 1 log2 titre unit each one to two years. The optimal serological marker of recent infection was investigated using 470 sample. For the most recent H3N2 2011 strain the optimal threshold was greater than log2 5.5 for all age groups at all time points. Sensitivity was greater than 90% for all thresholds but specificity was poor. Specificity was improved by using a multi-strain approach as measured by diversity index. The measured sensitivity and specificity will lead to a significant over estimation of influenza sero-prevalence unless test accuracy is adjusted for. Conclusion: Influenza in southern Vietnam has complex transmission dynamics including periods of intense influenza activity. Titre rise is seen within and across subtypes which would lead to repeat boosting of titre levels across many years. Further work should be performed to establish if this titre rise is accompanied by a change in neutralisation activity. Seroepidemiology of influenza is challenging because of cross-reaction. Methods exist to improve the estimation from these methods and should be employed for non-pandemic influenza serosurveillance.
|
|---|
