Influenza in Australian Children: burden of disease, advances in diagnosis, management and prevention through vaccination

Influenza is an almost universal respiratory viral infection which worldwide is reported to be responsible for 3 to 5 million cases of severe illness, and about 250,000 to 500,000 deaths per year. A systematic review I contributed to (Appendix #1) suggests an incidence of influenzaassociated influenza-like illness (ILI) of at least 1650 to 23400 (median 10180) per 100,000 globally per year. This indicates a substantial ongoing burden which is potentially amenable to vaccination strategies. In Australia, a targeted vaccination program exists aimed at individuals at higher risk of complications, including the elderly and individuals of all ages with co-morbid conditions. However internationally, increasing evidence from developed countries suggests a high burden in healthy young children, who contribute substantially to transmission in the community. This has led to a move towards universal vaccination strategies aimed at this group. Influenza, like other vaccine-preventable diseases, is a “moving” target affected by many factors including frequent mutation, changing epidemiology and control strategies such as vaccination; therefore, it requires constant re-evaluation to assess the success of interventions. The purpose of this thesis was to explore the current landscape of influenza in Australian children by establishing what is the extent of the problem and how can it be better managed. To this end, the questions being addressed were: 1. What is the current burden of influenza in Australian children? 2. In children who acquire infection, how can it be better diagnosed and how does earlier diagnosis help in managing children? 3. How well can influenza be prevented through vaccination in young children? 4. What are the safety issues surrounding influenza vaccination in children? 5. When children are vaccinated, how often does fever occur, and what is the pattern of fever and antipyretic use? How do these factors affect the vaccine’s immunogenicity in children? These issues are addressed through this thesis and the publications #1-7 contained within. Each publication is preceded by a brief overview of the issues and literature relevant to the publication and followed by a summary of the key findings and how it relates to the subsequent chapter of the thesis. Publications #1 and #2 examined the burden of influenza disease in children through analysis of administrative hospitalisation data (#1) and then through clinical review of influenza hospitalisations (#2). These found a substantial burden in young children, highest in infants, similar to other international developed countries. Publication #2 built upon findings in #1 and added clinical context showing the high proportion of hospitalised children with influenza who were previously healthy (~60%) and demonstrated that influenza was not a benign illness in hospitalised children (8.5% requiring intensive care unit admission). The same data helped show encephalitis to be an important neurological complication of influenza (Appendix #2). Vaccine uptake was too low, even in those at-risk children funded for vaccination (5%) and was much lower (0.7%) in unfunded healthy children. These findings support wider vaccination recommendations and funding to increase vaccine uptake. Appendix #3 and Publication #3 retrospectively investigated how children with influenza were investigated and managed through paediatric emergency departments. Appendix #3 showed that children with influenza had significantly higher rates of invasive investigations than children with other respiratory viruses, including lumbar punctures (LP, 18.3% vs 6.3% respectively) and blood cultures (71% vs 55.5%). Publication #3 then explored whether earlier diagnosis of influenza through rapid point-of-care testing (POCT) in this setting was beneficial. It found that early diagnosis was associated with reduced hospital admission and increased antiviral prescription. Rapid POCT, compared to standard influenza testing (multiplex respiratory virus panel), facilitated this earlier diagnosis (2.4 vs. 24.4 h, P<0.001), and had significant benefits in terms of reduced length of hospital inpatient stay (by 1 day, P=0.006) and increased antiviral prescription (odds ratio 4.54, P<0.001). These findings indicated POCT could have considerable utility in diagnosing influenza in a paediatric emergency department setting with impacts on hospitalisation rate and management approach. I then looked at vaccination to prevent infection. To advocate its use more widely in young children, vaccination must be shown to be efficacious in this population. Publication #4 confirmed the efficacy of influenza vaccination through a randomised controlled trial (RCT) in young children attending childcare. Vaccine efficacy of 100% (95%CI 16–100%) was demonstrated against laboratory-confirmed influenza in those aged 24-48 months. Vaccine safety concerns by parents limit vaccine uptake in children. It is a major ongoing concern especially in the context of a 2010 vaccine safety scare during which increased rates of febrile convulsions were detected after influenza vaccination in children aged <5 years. These occurred following vaccination with only one brand of vaccine, Fluvax® by bioCSL (now Seqirus). Publication #5 was a systematic review and meta-analysis of fever and febrile convulsions after inactivated influenza vaccine in children, conducted in the wake of these events. It found, reassuringly, that non-bioCSL brands of inactivated trivalent influenza vaccine (TIV) in RCTs had acceptable pooled fever estimates of 6.7% in young children (aged 6–35 months). Febrile convulsions were reported in 1.1 per 1000 vaccinated children in these RCTs. This contrasted with uncontrolled bioCSL studies, completed at that time, which reported fever rates between 22.5–37.1% for similarly aged children, though fortunately without increased febrile convulsions. Publication #6 reviewed the root cause analysis by bioCSL which identified the manufacturing process, virus-strain, and patient genetic factors as contributing to the increased risk of febrile convulsions seen with their vaccine. I critically analysed their conclusions, and discussed limitations in their analysis and why manufacturing issues, above others, should remain the key focus for increased reactogenicity with their TIV. Knowing determinants of vaccine immunogenicity in children could help improve vaccine efficacy/effectiveness. Publication #7 was the first study that used individual-level data from vaccine manufacturer clinical trials to explore fever, its patterns, frequency, and associated antipyretic use and how these affect immunogenicity in children. I documented evidence, from pooling three GlaxoSmithKline paediatric influenza trials (n=5902), of a 21-39% increase (p≤0.01) in adjusted post-vaccination geometric mean titre (GMT) when fever was experienced (versus no fever) after influenza vaccination (day 0-3) in children, and conversely a 13-20% decrease (p