West Nile virus encephalitis

What it is

West Nile virus is a Flavivirus (family Flaviviridae) that circulates between wild birds and mosquitoes, mainly of the genus Culex. Horses and people are infected through mosquito bites and are dead-end hosts: viraemia is low and brief and not enough to infect new mosquitoes, so the horse does not transmit the virus to other horses or to people.

After the bite, the virus multiplies in tissues and lymph nodes and, in a small proportion of infected horses, reaches the central nervous system and causes polioencephalomyelitis: inflammation of the grey matter of the brain and spinal cord, with neuronal death.

In Europe lineages 1 and 2 circulate, with cases mainly in the Mediterranean basin and in central and eastern Europe; in the Americas it spread across the continent from 1999. It is a WOAH-listed disease and notifiable in many countries. The horse acts as a sentinel for virus circulation, which can also cause severe disease in people.

Who it affects

Unvaccinated horses in areas where the virus circulates, in late summer and autumn, when mosquitoes are abundant. Older horses are at greater risk of dying.

Anatomy involved

• Brain: lesions are concentrated in the grey matter of the thalamus, midbrain and hindbrain (pons and medulla oblongata), and in the ventral horns of the spinal cord, mainly in the thoracic and lumbar segments. Brainstem involvement explains cranial nerve deficits and behavioural changes, and involvement of spinal motor neurons explains weakness, muscle tremors and paralysis.

Clinical signs

• Fever, not always present, depression and inappetence.
• Ataxia and weakness, often asymmetric and more marked in the pelvic limbs, which can progress to recumbency.
• Muscle fasciculations and tremors, especially of the face, neck and trunk, which are very characteristic.
• Behavioural changes (drowsiness or hyperexcitability), hypersensitivity to touch and sound.
• Cranial nerve deficits: tongue weakness, facial paralysis, difficulty swallowing.
• Most survivors recover completely, although convalescence can last weeks.

Diagnosis

• IgM-capture ELISA on serum: indicates recent infection and is the most widely used test in the live horse.
• Virus neutralisation (plaque reduction test) for confirmation, interpreted with care in vaccinated horses.
• RT-PCR and immunohistochemistry on brain and spinal cord at necropsy.
• Other causes of acute neurological signs must be ruled out: rabies, equine herpesvirus 1 myeloencephalopathy, other viral encephalitides and equine protozoal myeloencephalitis. Until rabies has been ruled out, the horse should be handled with gloves and protection.

Treatment and prevention

• There is no antiviral treatment. Supportive care is given: non-steroidal anti-inflammatory drugs, fluid therapy, protection against falls, a support sling and nursing care for recumbent horses.
• Vaccination: inactivated and recombinant vector vaccines are available. They are given before the mosquito season, with boosters depending on the area and the vaccine.
• Mosquito control: remove standing water, use repellents and keep horses stabled at dawn and dusk.
• Drug use in horses is subject to each country's regulations and to withdrawal periods. In the EU the horse is a food-producing species, and its identification document (passport) states whether it is excluded from slaughter for human consumption, which determines which drugs may be given.

Key studies

• Salazar et al. (2004), cross-sectional study of 484 equids with confirmed infection in Nebraska and Colorado, with data obtained from owners by telephone. 28.4% died or were euthanised (137 of 482). Unvaccinated animals were twice as likely to die as those vaccinated at least once, those recumbent and unable to rise 78 times more likely, and females 2.9 times more likely. 79.9% of survivors (271 of 339) recovered fully, in a mean of 22.3 days. The data depend on owners' recall.
• Schuler et al. (2004), retrospective study of 569 affected horses in the 2002 North Dakota outbreak. 22% died. The odds of death in unvaccinated horses were 3 times those of horses given 1 or 2 doses and 16 times those of horses vaccinated according to the manufacturer's instructions. Recumbency, hind limb paresis and age over 5 years were associated with death. Vaccination status was unknown in 19% of cases.
• Epp et al. (2007), case-control study of 300 equids on 46 premises during the 2003 Saskatchewan outbreak. 64% had antibodies from natural infection. Unvaccinated animals were 23 times more likely to become ill (95% CI 3.0–168.5), and estimated field vaccine efficacy was 96% (95% CI 67–99%). The confidence interval is very wide.
• Bowen et al. (2014), randomised, blinded trial in 20 horses: inactivated lineage 1 vaccine or no vaccine, followed by intrathecal challenge with a lineage 2 strain from Greece. 60% of controls had to be euthanised and none of the vaccinated horses; 80% of controls became viraemic versus a single vaccinated horse, and 8 of 10 vaccinated horses had no lesions. It is an experimental infection that does not reproduce the natural route.
• Metz et al. (2021), systematic review and meta-analysis of 39 seroprevalence studies in 28,089 equids from 16 European countries (2001–2018). Pooled seroprevalence was 8% (95% CI 5–12%), slightly higher in Mediterranean countries. Heterogeneity between studies was very high (I² 99.3%), partly because many were carried out after outbreaks.