The changing face of Hendra virus: what do we really know for certain?

Background

It has been 30 years since the discovery of a novel virus responsible for an outbreak of sudden death and severe respiratory disease in horses in the Brisbane suburb of Hendra.^1,2 This first outbreak resulted in the death of 20 horses, as well as infection of two people in close contact with the affected horses. Subsequently, one of these people died. This, and the retrospective diagnosis of a previous infection that caused the death of a veterinary assistant in Mackay on the Queensland central coast,³ Australian equine veterinarians and the horse owning public were faced with a serious viral zoonoses of equine origin for the first time.

Since 1994 there has been a considerable research effort to improve our understanding of Hendra virus (HeV), the pathogenesis of HeV infection in horses⁴ and in laboratory animal species,^5–10 and the ecology of HeV in its natural host species.^11–16 Although an uncommon infection in horses, serological studies suggest HeV is a common infection in the four Pteropid bat species of mainland Australia. However, the detection of a novel HeV variant in an acutely ill horse that was not detected by routine HeV diagnostic tests¹⁷ and the confirmation that this variant is present in flying fox populations^18,19 has led to the question: what do we really know about HeV?

Equine HeV infection and disease

The initial outbreak of HeV in 1994 was characterised by peracute, fulminant respiratory disease in affected horses, although neurological signs were also observed.¹ For many years after this, HeV was reported as occasional, sporadic cases involving individual horses, usually either moribund or found dead, with the diagnosis made post-mortem.^20,21 A small increase was reported in the frequency of HeV cases in horses between 2006 and 2010, but, in contrast, 18 incidents of equine HeV infection occurred within just 12 weeks in 2011.¹⁵

Equine HeV cases often present as an acute onset of severe respiratory or neurological signs, although early signs, such as fever, tachycardia and inappetence, were noted in experimentally infected horses.⁴ Necropsy reports of HeV cases describe pulmonary oedema, congested and consolidated lungs, with blood-tinged foam in the airways and dilated lymphatic vessels.^1,4,9,22 Histological examination of tissues from naturally and experimentally infected horses has revealed vasculitis of the small blood vessels in a wide range of tissues, including the lungs, central nervous system, and lymphoid and renal tissues.²³

Human cases have also been seen in the years following the original outbreak,²⁴ although these have been uncommon and the people involved survived. However, an outbreak of HeV in a veterinary clinic involving multiple equine and human cases in Brisbane in 2008,²⁵ and the death of another veterinary practitioner the following year, highlighted the serious zoonotic potential of HeV and also prompted questions to be asked about the state of our knowledge about the pathogenesis of disease caused by HeV infections. Although this virus had been recognised for over a decade, veterinary authorities were unable to provide practitioners with evidence-based advice on a minimum sample set to collect in suspected HeV cases in order to rule out HeV. Subsequently, experimental studies conducted at the Australian Animal Health Laboratory (AAHL) showed that viral RNA was detectable in nasal samples prior to the onset of clinical signs and that the amount of RNA increased during the incubation period to a maximum during the clinical phase of infection, consistent with replication of HeV in the upper respiratory tract.⁴ In live horses, nasal or nasopharyngeal samples and blood samples collected using ethylenediamine tetraacetic acid (EDTA) as an anticoagulant, as well as oral and rectal swabs, are recommended as suitable for polymerase chain reaction assay testing, where it is safe to collect these samples. Necropsy of acute HeV cases is hazardous, as high loads of infectious HeV were detected in many tissues. Necropsies should be reserved for experienced pathologists in suitable facilities where workplace health and safety can be ensured. Field necropsy samples can be limited to the same array of samples that would be collected from the live horse. These experimental challenge studies provided practitioners with a risk-based approach to sampling that aimed to safeguard the health of veterinarians and other people involved in further HeV cases.

A recombinant protein based sub-unit vaccine was developed and released for use in horses in an effort to protect horses from HeV infection, and in doing so protect humans that come into contact with infected horses.^26,27 This vaccine stimulates high levels of serum antibody against the HeV G protein, which acts as the adhesin of the virus.²⁸ Vaccination is the most effective way to protect horses against HeV, and its use has been quite widespread. However, there is some level of opposition to vaccine use among horse owners.²⁹ Notably, although HeV cases have been reported in horses since the release of the vaccine in 2012, none of these cases have been in a vaccinated horse.

Epidemiology

The surge in the number of HeV cases reported in the years immediately following the outbreak in Redlands Shire, Qld, in 2008 highlighted that there were still significant gaps in our knowledge of the ecology and epidemiology of HeV in the natural reservoir host, flying foxes, and in the mechanisms by which HeV was transmitted from bats to horses. Serological surveys of bats confirmed that HeV infection was a relatively common, and non-lethal, occurrence in flying foxes,¹¹ and further studies showed that high loads of virus were detectable in bat urine collected under bat roosts.¹⁶ To date, there is still no definitive data supporting the route of transmission from bats to horses, although exposure of horses to urine from infected bats is a compelling hypothesis. Interestingly, all human cases to date have been infected following contact with horses, rather than contact with bats.³⁰

A range of ecological studies have informed our understanding of the geographical distribution of HeV in the natural host species and enabled an evaluation of the risk of spillover to horses. An important study of nearly 15,000 pooled urine samples from bat roosts in Queensland and New South Wales collected over a 3-year period examined the seasonality of HeV excretion and other risk factors.³¹ Although HeV had been detected previously in all four Pteropid bats species in mainland Australia, high loads of HeV were most likely to be detected in urine samples from under bat roosts in southern Queensland and northern NSW, with moderate levels of HeV detected in samples collected in North Queensland. The higher viral loads were associated with the presence of black flying foxes in South East Queensland and spectacled flying foxes in North Queensland, suggesting that these species played a more important role in potential spillover events. These results, and the geographical restriction of equine cases of HeV to areas of Queensland and Northern NSW, suggested that black and spectacled flying foxes were more likely to be involved in spillover events where horses were infected, and thus that they posed a greater threat to human health through the risk of subsequent infection of humans from horses. Unfortunately, the identification of variant HeV strains that were not detectable because of mutations at the primer binding site used for the routinely employed polymerase chain reaction (PCR) assay ^17,19,32 has again raised more questions about our knowledge about HeV. Are the new variants more, or less, likely to be detected in different flying fox species? Can the results of previous ecological studies that could not detect the variants be relied upon when making risk-based equine health and public health decisions? The original detection of the variant HeV genotype 2 was from a grey-headed flying fox and this variant HeV strain has been detected in other grey-headed and black flying foxes,^18,19 as well as from the most southern reported equine HeV case.^17,33 Importantly, antibodies against HeV genotype 1 are cross reactive with glycoproteins of HeV genotype 2, suggesting that the existing recombinant HeV vaccine will protect horses against the variant strains.³⁴

Our knowledge of HeV infection in its natural reservoir hosts and in horses has come a long way since this virus was first discovered 30 years ago. Although there are many questions that still require answers, the undeniable truth is that we know a lot more than we did in 1994 and we have a vaccine that can protect horses, and therefore people, from a serious viral disease.