Influenza Vaccines in the Elderly ?Why don't they work

Influenza viruses circulate world-wide and cause flu outbreaks each winter. These influenza strains continuously evolve to evade antibodies of our immune system. These antibodies are usually specific for only one strain, so people can get flu year after year. Disease ranges from nothing, through relatively mild respiratory symptoms leading to a few days in bed, to severe disease with hospital admission and serious complications such as pneumonia or even death. Elderly over 65 years are among those at highest risk of most severe disease and death.

Getting flu or indeed having a flu vaccine protects by stimulating antibodies to virus proteins, the most important of which is called haemagglutinin, but antibodies against other flu proteins may also contribute to protection. Annual influenza vaccination is the most effective way to protect people from flu and prevent outbreaks. The key to a good flu vaccine campaign each winter is therefore to match the vaccine to the flu strains that are circulating just prior to and therefore hopefully during each winter season. This ensures people have antibodies in their body ready to protect against circulating flu strains, but this involves a bit of guess work during annual vaccine production. Recent studies have demonstrated that for elderly people, the effectiveness of the annual flu vaccine is less than 20%. The reasons are unclear, but probably include guessing wrongly about which viruses will be dominant each winter, the type of vaccine used, the number of previous vaccinations or episodes of flu, and potentially a declining immune system with age. The consequences are that each winter emergency departments are very busy, wards are full with elderly patient with flu, operations can be cancelled and flu can spread in hospital. This is an enormous burden on the NHS.

The main focus of our research is to work out why vaccines aren't working in the elderly and to work out how we can improve the vaccine campaign to protect the individual patients and help the NHS during winter. We will study elderly patients (over 65) admitted to hospital with laboratory confirmed flu, over at least 2 winter seasons. We know from previous studies, that about three quarters of elderly patients who come into hospital with flu have been vaccinated and new flu vaccines are starting to be used in the UK.

We want to study the fine detail of the antibodies in flu patients and relate this information to the severity of illness, the type of vaccine they received and the genetic code of the virus that caused the flu. We will also compare their antibody responses with those from a broad range of people including elderly patients who did not get flu, and together this will help us design better vaccines each year. It will also help us forecast flu outbreaks and the likely burden to GPs and the NHS each year, so they can better prepare themselves each winter.

We also propose to use some new technologies to see if they can help manage flu cases in hospital or GP practices. One of these is a point of care test which gives a result in 15 minutes on a finger prick of blood. It helps determine whether the patient has a viral infection like flu or some other infection due to a bacteria, so the right treatment is given earlier. This might help reduce unnecessary antibiotic prescribing which is a major cause of increasing antibiotic resistance.

A second new technology we will study is a small rapid genetic sequencer (Nanopore) which can determine the fine genetic code of the flu strain in clinical samples. Our research will assess whether these devices can distinguish different strains and therefore identify whether and where the flu strains are spreading in either the hospital or the community, so infection control teams can target those areas and prevent spread.

Technical Summary

The hypotheses to be tested in a cohort of over 65 admitted to GSTT with laboratory confirmed influenza are:

1 The influenza antibodies (Ab) profile at admission predicts severity of outcome and reflects vaccine exposure 2 Nanopore sequencing can be used as a tool to confirm infecting influenza strain and to correlate sequence variation at key viral antigenic sites with serological profiling 3 Biomarkers (CRP and MxA) assessed at admission can provide prognostic information.

Study population and sampling: Serum and whole blood will be collected at admission/recruitment and at four weeks. Case note review and follow up will complete the assessment of clinical outcomes. Verification of vaccination history will be obtained from GP records. Severe influenza is defined as admission to ITU or death or pneumonia. Rapid testing for biomarker detection using a licensed PoC device (FebriDx RPS Diagnostics) will be recorded on admission samples.

Serology: Influenza serological work will be carried out at PHE and will focus on measurement of influenza strain specific anti HA Ab using Haemagglutination Inhibition, total neutralising Ab, Neuraminidase Inhibiting Ab (ELLA) and anti-HA stem Ab using serum from patients, and comparing the results between these samples and to the infecting strain.

Subtyping and sequencing of infecting strains: RNA will be extracted from clinical samples followed by RT-PCR with a single set of primers for the eight influenza gene segments. The PCR amplicons will be then purified and library preparation and sequencing by using the 1D2Sequencing Kit (Oxford Nanopore) for MinION or GridION platforms. Analysis will be done using by in-house bioinformatics pipelines developed by the CIDR bioinformatics team. Nanopore will be compared with influenza sequences derived by Illumina platforms at PHE.

Statistical methods: Pair wise comparisons will be calculated. Multivariable logistic regression model will be carried out by KCL statisticians