Standing at the crossroads: the emergence of Japanese encephalitis in Australia

Background

Arboviruses are maintained in nature in a cycle of transmission between invertebrate hosts, such as mosquitos and midges, and vertebrate hosts. For viruses such as dengue and yellow fever viruses the vertebrate reservoir host are humans, but more often the reservoir vertebrate host species are water birds or small mammals. These arboviruses are only of significance to humans or domestic animals when conditions are such that the invertebrate or reservoir host population expands and the virus spills over to infect incidental hosts, such as humans and domestic animals such as horses.

Japanese encephalitis virus (JEV) is a mosquito-borne flavivirus that, until 2022, was considered exotic to mainland Australia. Although JEV had been detected previously in the Torres Strait in the 1990s,¹ it had not established a sustainable transmission cycle, as further human cases were not reported. The virus implicated in the Torres Strait incidents was characterised as genotype 2 and it was hypothesised that infected windborne mosquitos from Papua New Guinea were the source of the virus.² Subsequently, no human cases of JEV were reported in Australia until 2021, when a human case of JEV was reported in a resident of the Tiwi Islands,³ although the virus implicated in the Tiwi Islands case was genotype 4, which was circulating in Indonesia previously.⁴

Between February and March 2022, JEV was identified as a cause of mortality and reproductive losses in piggeries in four mainland Australian states: Queensland, New South Wales, Victoria and South Australia. Within weeks of the first detection near Bundaberg in Queensland, there were cases reported in piggeries in central Victoria, throughout central western NSW and in South Australia. Importantly, pigs are not routinely transported between farming facilities, unlike other production animal species such as cattle and sheep. Thus, spread of this outbreak was not likely to be due to movement of infected pigs (Sus scrofa), but rather movement of other infected animals, most likely water birds, with subsequent infection of local populations of invertebrate hosts. These outbreaks of disease in pigs were followed by reports of human cases of Japanese encephalitis. In total, 25 confirmed cases were reported in these states, as well as an additional 12 probable cases, including 3 deaths (2 confirmed, 1 probable).⁵ The virus identified in samples from humans, pigs and mosquitos in south-east Australia was sequenced and was closely related (>99.7% nucleotide identity) to the virus from the Tiwi Islands case.³

Japanese encephalitis serogroup

JEV belongs in the Flavivirus genus of the family Flaviviridae, where it is grouped with West Nile virus (WNV) and Murray Valley encephalitis virus (MVEV) in the Japanese encephalitis serogroup, which reflects the close antigenic relatedness of these viruses. These viruses share a similar natural life cycle. Mosquitos are the competent invertebrate host species and are infected by biting and ingesting blood from an infected vertebrate reservoir host.⁶ The endemic transmission cycle of WNV and MVEV involves water birds acting as the major amplifying host, and Culex mosquito species. Birds are efficient amplifying hosts, as they develop a prolonged high titre viraemia but generally remain asymptomatic. Endemic JEV transmission also involves infection of water birds and mosquitos, but JEV can also establish a transmission cycle between mosquitos and pigs, as pigs have sufficient viraemic titres to reinfect mosquitos and thus are considered amplifying hosts.⁷ All three of these arboviruses can spill over into other animal populations, with disease manifestations especially obvious in horses and humans. Horses and humans are considered to be ‘dead end’ or incidental hosts, with infected individuals producing insufficient viraemic titres to reinfect subsequent biting mosquitos.

Japanese encephalitis

Serious clinical disease in domestic animals due to JEV infection is reported most commonly in horses and pigs, although other species can be infected. Widespread vaccination of horses is common in Japan, but, prior to this, outbreaks of Japanese encephalitis in horses occurred seasonally and was associated with human epidemics, with the prevalence ranging between 10 and 50 cases per 100,000 horses. The case fatality rate was estimated to be 50%.⁸ The majority of infected horses and humans remain asymptomatic after infection, with the ratio of symptomatic to asymptomatic infections in humans reportedly ranging from 1:25 to 1:1000, depending on host immunity, virus virulence, vector abundance and bite frequency.⁹ The clinical signs of JEV infection in horses are neurological signs consistent with encephalitis, progressing to blindness, coma and death. Clinical signs of Japanese encephalitis in horses can be vague and non-specific, and include transient fever, lethargy and ataxia, but a proportion of infected horses will have a high fever and show signs of CNS disease, such as aimless wandering, violent and demented behaviour, blindness, profuse sweating, muscle tremors, collapse and death. The main manifestations of the disease in pigs are reproductive, with expulsion of litters of stillborn or mummified foetuses, usually at term. Viable piglets may have tremors and convulsions before dying.¹⁰ Aspermia has also been reported in boars.¹⁰

In Australia, outbreaks of disease caused by flaviviruses in horses and humans are very sporadic in nature and tend to occur during La Niña-associated weather patterns, with the most recent cases occurring in 2011.^11–13 Warm, wet conditions with extensive flooding, widespread inundation and expanses of standing water favour expansion of the invertebrate host population, as well as the natural vertebrate reservoir host for WNV and MVEV. In 2022, the first cases of JEV infection were diagnosed in pigs¹⁴ and humans, with several unconfirmed cases in horses.

Diagnosis of JEV, WNV and MVEV infection in horses is complicated by the significant antigenic cross reactivity between these viruses, which makes serological diagnosis difficult with current diagnostic test modalities. Definitive diagnosis of any of these flavivirus diseases requires detection of virus in tissue, such as the brain or spinal cord.¹⁵ A rising antibody titre between paired serum samples is indicative of infection and may facilitate a presumptive diagnosis. Detection of viraemic horses is unusual, because of the low titres of virus in the blood of horses, although whole blood collected into ethylenediamine tetraacetic acid (EDTA) anticoagulant is a minimally invasive sample to collect from animals in the acute stage of disease and these samples should be collected ante mortem for virus detection. In Australia, an acutely ill horse with neurological signs could also be infected with Hendra virus, and thus could be a significant zoonotic risk. Veterinarians who suspect flavivirus encephalitides should take all precautions to minimise the risk of zoonotic infection with Hendra virus (HeV), including determining the HeV vaccination status of the affected horse(s). Exclusion testing to rule out HeV infection should be undertaken prior to any invasive or post-mortem sample collection.

Looking ahead

The emergence of JEV in Australia highlighted the generally good level of cooperation between different animal and public health agencies in Australia, but it also highlighted some of the deficiencies in our capacity to respond to complex infectious disease problems that transcend state boundaries and affect multiple animal species, including humans. Driving the response to an outbreak through the lens of human health was problematic when humans are incidental hosts and the vast majority of cases were in the animal industries. Establishment of disease surveillance programs in the north of Australia did not result in early detection of the JEV incursion, as the first report of this virus in Australia was from a commercial pig farm in Queensland. However, this is most likely a reflection of the focus of these programs on the range of significant transboundary diseases affecting countries to the north of Australia, such as foot and mouth disease, rabies, lumpy skin disease and African swine fever.

Disease preparedness in Australia is currently at a crossroads. Looking back, there are decades of under-investment in animal health laboratory capacity and a concentration of expertise in or around major capital cities. Looking forward, there is a high likelihood of further disease incursions, as changes in climate alter the geographical distribution of wild animal and insect species, and increased movement of people and animals further facilitates the movement of animal diseases. Several of these diseases, such as rabies, require a significant level of One Health collaboration to control, whereas others will have a devastating impact on the capacity of Australia to trade internationally in animals and animal products. The choice facing Australia is whether to continue the policies of economic rationalism of the past, or to choose the other road, of increased investment in animal health, environmental ‘health’ and a true One Health approach to the likelihood of future disease occurrences in Australia.