2 Feb 2015
David Rendle
Job Title
Influenza vaccination is so routine in equine practice there is perhaps a tendency to take the disease for granted. However, the Australian outbreak in 2007 highlighted the devastating effect the virus can have on equine welfare and industries – albeit in a naïve population.
The outbreak, which resulted from the ineffective quarantine of vaccinated horses that were shedding virus, was estimated to have cost Australia around AU$1b (£538m). However, vaccination was also an important part of the solution.
Given the frequency of horse movements into – and within – the UK equine population, in which it is estimated from sales data only one in three horses are vaccinated, the threat of a major equine influenza (EI) outbreak in the UK is ever-present and 2014 saw an increase in the number of (small and controlled) outbreaks compared to previous years. At practice level, vaccines need to be chosen carefully and protocols ought to be in place to ensure horses exhibiting clinical signs consistent with EI are isolated effectively and diagnosed promptly.
Since the initial recognition of H7N7 viruses (Prague 1956) and H3N8 viruses (Miami 1963), there have been considerable antigenic changes in EI viruses. H3N8 viruses diverged in the late 1980s into what became known as “American” and “Eurasian” strains. Later, further divergence of American strains occurred into three groups known as Argentina, Kentucky and Florida sublineages.
More recently, further divergence within the Florida sublineage has given rise to two clades; one and two. Since 2006, all EI strains isolated in the UK have been of the H3N8 Florida sublineage. In the UK, surveillance data indicates clade one viruses have assumed lesser importance and have not been identified as a cause of disease since 2010.
Genetic shifts in circulating influenza viruses are monitored by the expert surveillance panel of the World Organisation for Animal Health (OIE), which then makes recommendations on when vaccine strains need to be updated. The current recommendation is for vaccines to contain both a clade one and clade two virus strain of the Florida sublineage, with there being no need for a representative of the H7N7 viruses or Eurasian strains of the H3N8 viruses. Only one of the vaccines available currently is compliant with OIE recommendations (Table 1).
Recent marketing drives appear to have been aimed at persuading us vaccines not containing recommended strains can offer protection against those same strains selected by the OIE. In the author’s view, the less cynical might even be led to believe these vaccines were compliant with the recommendations of the OIE. While it is reassuring very old vaccine strains offer cross-protection against less old strains, the point of vaccination is to maximise protection against continually evolving strains. Challenge studies cited in marketing literature do demonstrate horses vaccinated with outdated strains have less severe clinical signs and shed less virus than naïve controls injected with saline. However, vaccinated horses still develop clinical signs and they still shed virus.
So, what are our goals? Are we looking for a vaccine that provides better protection than a saline injection at the peak of its efficacy, or a vaccine that comes as close as possible to providing complete protection and halting disease spread in field conditions and against a virus that is yet to evolve? Modelling data from experimental studies and field data from outbreaks in Newmarket in 1995 and 1998 all demonstrate the risk of an epidemic increases if the vaccine strain becomes more distant from the circulating strain1–4. If cross-protection were sufficient to ensure clinical protection and prevent disease outbreaks then why would the OIE repeatedly recommend strains are updated?
The 2007 outbreak of EI in Australia ultimately cost around AU$1b to resolve. The Callinan report (www.daff.gov.au/about/publications/eiinquiry) into the outbreak concluded it was caused by imported vaccinated animals and “if there are commercially available vaccines that contain representatives of currently circulating strains, the import conditions should specify that the horses be vaccinated using that vaccine or one of those vaccines”.
Furthermore, there have been recent cases of EI in the UK and elsewhere in northern Europe, in horses that have complete vaccination histories. Cases in horses reported to be vaccinated comprise a small minority – highlighting the benefits of vaccination – but also illustrate we cannot afford to be complacent when it comes to optimising the efficacy of vaccines.
The earliest vaccines simply stimulated a circulating immunoglobulin G (IgG) antibody response and much of the cited evidence for vaccine efficacy remains focused on antibody responses to vaccination. However, EI is not a viraemic condition and the commonly quoted IgG levels are likely to be less important to the immune response than immunoglobulin A responses on mucosal surfaces and cell-mediated immunity.
Intuitively, killed virus and subunit vaccines are less likely to mimic natural infection than vectored vaccines. However, recent evidence from Ireland indicated a primary course of a whole virus vaccine resulted in a significantly greater antibody response than its Irish competitors, which included immune-stimulating complexes (ISCOM) and vectored vaccines similar or the same as those used in the UK5.
Unfortunately, mucosal and cell-mediated immunity and protection against challenge with currently circulating virus strains were not assessed. Vaccines with ISCOM have been shown to stimulate a greater antibody response using an equivalent amount of antigen6,7 and also stimulate cell-mediated immunity8.
The major drive for influenza vaccination in the UK is compliance with the regulations laid down by the British Horse Racing Authority (BHA) and Fédération Equestre Internationale (FEI), with most major equestrian events falling under the vaccination recommendations of one or other of these associations. Recommended vaccination intervals were developed from work performed in the 1960s and 1970s when the first multivalent adjuvanted EI vaccines were developed in horses and recommendations were extrapolated from experimental studies investigating serological responses in naïve Welsh mountain ponies.
The equine population in the field is very different from the populations used in experimental studies. Horses may not be totally naïve, may have varying degrees of maternally derived immunity at the time of vaccination and may be vaccinated with different products. The vaccines themselves may have been exposed to suboptimal storage temperatures and may be administered at suboptimal intervals. These factors and differences between vaccine strains and field strains of EI all are all likely to contribute to the vaccine failures that do occur – particularly in young racehorses.
Modelling techniques have demonstrated a small proportion of non-responders can have a major impact on the spread of disease in a Thoroughbred training scenario. Concerns regarding vaccine failures have led to the modification of vaccination schedules by vets in practice and the administration of extra vaccines in an attempt to reduce the risk of disease. Such modifications have had demonstrable benefits in antibody responses9.
The risk of outbreaks between second and third vaccinations is particularly high (from 12 weeks after the second vaccination). In one investigation yearlings in Newmarket had antibody levels below protective levels within four months of booster vaccination, at the time when horses are mixing at the autumn sales10. Mathematical models have also been used to demonstrate the additional benefits of six-monthly vaccination in young horses11.
The FEI adopted a strategy of six-monthly vaccination in 2005. However, considering developments in vaccine technology and the experimental and epidemiological data that has become available since mandatory vaccination was introduced, it may be time for a wholesale rethink of vaccination schedules to bring them in line with current best evidence. The timing of the third vaccination in a primary course, the use of six-monthly vaccination and the need for restarting a primary course if a 12-month interval is exceeded all merit discussion.
The Hurlingham Polo Association’s (HPA) rules with respect to EI vaccination are less stringent than those of the FEI or BHA. The absence of a third vaccination at around six months as part of the primary course leaves a huge window of susceptibility between the second vaccination and the first annual vaccination and is likely to compromise the degree of immunity that is induced. The HPA also permits ponies to compete before they have received the second vaccination of the primary course without stipulation for any interval between vaccination and competition. Considering the extensive international movement of polo ponies, the author believes these omissions are cause for concern.
Another paradox of equine vaccination is the discrepancy between data sheet recommendations and the regulatory requirements of the equestrian governing bodies. Under the VMRs we are obliged to administer vaccinations in accordance with their marketing authorisations. Any divergence from the data sheets should be based on sound clinical reasoning and be accompanied by informed consent.
The regulatory authorities allow a far greater window for the second vaccinations of a primary course than the data sheets allow and experimental evidence would suggest is prudent. Practitioners administering vaccines within the regulatory authorities’ window, but outside the data sheet recommendation, have no clinical justification for doing so and are potentially committing an offence. In the author’s experience, at the third booster, regulatory requirements cannot be reconciled with some of the data sheet recommendations.
When multiple vaccine brands are used interchangeable we may also be in contradiction of the data sheets as some vaccine brands recommend they are not used interchangeably unless they contain the same vaccine strains. Little is known of the effects of mixing vaccine strains and technologies within a course.
In the UK it is estimated from sales data approximately 40 per cent of horses are vaccinated yet many of these horses have been injected with EI strains that are more than 10 years out of date. For there to be confidence that an EI epidemic will not occur, it is estimated 70 per cent of the equine population needs to be vaccinated (Animal Health Trust, personal communication).
Most experts are agreed it is not a case of if a major EI outbreak occurs, but when. Should this happen then, as things stand, in the author’s opinion the veterinary pharmaceutical industry and veterinary profession are both likely to be very exposed for seeming to worry more about profit margins than the health and welfare of the equines and the equine industries upon which they depend. Efforts need to be made to increase the uptake of vaccination with relevant EI strains and to ensure vaccines are administered according to their data sheets in order that they confer robust and sustained immunity.
Figure 1. Horses with equine influenza typically have a hacking cough and nasal discharge, are markedly pyrexic and become dehydrated as a result of reduced food and water intake.
Figure 2. Vaccinated horses may be infected with EI – especially if there is significant divergence between the vaccine and field strains. A soft cough and serous nasal discharge typify EI infection in vaccinated horses.
Table 1. Virus strains in current UK vaccines and strains the OIE would like them to contain. Recommended strains are shown in red and strains the OIE considers no longer necessary are shown in blue.
