The equine encephalitides are clinically similar, usually cause diffuse encephalomyelitis (see Meningitis, Encephalitis, and Encephalomyelitis), and are characterized by signs of CNS dysfunction and moderate to high mortality. Arboviruses are the most common cause of equine encephalitis, but rabies virus, Sarcocystis neurona (see Equine Protozoal Myeloencephalitis), Neospora sp (see Neosporosis), and nematodes may also cause encephalitis. Arboviruses are transmitted by mosquitoes or other hematophagous insects, infect a variety of vertebrate hosts (including humans), and may cause serious disease. In the western hemisphere, most pathogenic arboviruses use a mosquito to bird or rodent cycle.
Etiology and Epidemiology
The most pathogenic viruses for horses are alphaviruses of the family Togaviridae. Endemic species in the USA include Eastern, Western, Highlands J, and Venezuelan equine encephalitis (Everglades) viruses. Other alphaviruses associated occasionally with equine encephalitis are Semliki Forest, Ross River, and Una viruses. These latter viruses are not found in the New World. Eastern equine encephalitis virus (EEEV) has 2 distinct antigenic variants that are separated longitudinally. The North American variant is the most pathogenic and is found in eastern Canada; all states within the USA east of the Mississippi River and in Arkansas, Minnesota, South Dakota, and Texas; and in the Caribbean Islands. The South American variant is less pathogenic and confined to central and South America.
Eastern and Western equine encephalitis (WEE) viruses are separated in North America primarily latitudinally; however, WEE virus is relatively heterogeneous, with several subtypes consisting of WEE, Sindbis, Aura, Ft. Morgan, and Y 62–33. WEE is found in western Canada, states in the USA west of the Mississippi, and in Mexico and South America. WEE previously isolated in the south and eastern USA has been shown to belong to the HJ virus serogroup. Venezuelan equine encephalomyelitis (VEE) has 6 antigenically related subtypes: subtype I (VEE), Everglades, Mucambo, Pixuna, Cabassou, and AG80–663. Subtype I serovars AB and C primarily cause epizootics; subtype I and serovar IE caused a large outbreak in Mexico in 1993. Epizootic strains are not generally found in the USA, although there was an epizootic of VEE in 1971. Sylvatic subtype II (Everglades) has been isolated from humans and mosquitoes in Florida; subtype III has been isolated in the Rocky Mountains and northern plains states.
The principal means of transmission and amplification of EEE is a mosquito-avian-mosquito cycle. EEE has been isolated from 27 different mosquito species in the USA. The primary mosquito vector for the enzootic or sylvatic cycle of EEE is Culiseta melanura, which inhabits swampland. Epizootics in equids, epornitics in pheasant and quail, and human cases are seen when virus infection rates are high in birds. Aedes vexans and A canadensis mosquitoes (which breed in containers) and Culex erraticus may be responsible for bridge transmission to mammals.
Seasonal changes in C melanura biology and their relationship to EEE virus transmission vary with the geographic location and its associated climate. In subtropical areas (eg, Florida), transmission occurs throughout the year with a peak in summer. In more temperate regions, there is a distinct summer transmission season. In South America, serologic studies suggest that forest-dwelling rodents and marsupials are the vertebrate hosts, and EEE is readily recovered from sentinel mice and hamsters.
WEE is transmitted primarily by Culex tarsalis, which is found just west of the Mississippi river and throughout the West. This mosquito breeds in sunlit marshes and in pools of irrigation water in pastures. WEE can also be transmitted by the tick Dermacentor andersoni. Epizootics of WEE are associated with increased rainfall in early spring followed by warmer than normal temperatures.
Sylvatic VEE viruses are found throughout North, Central, and South America in jungle or swampy environments with persistent fresh or brackish water. The mosquitoes that serve as the primary vectors for the bird- or rodent-mosquito life cycle are members of the subgenus Melanoconion (Culex cedecci). Epizootics are associated with a mutation to a subtype I (AB, C, and possibly E), a change in mammalian pathogenesis, and a change to several bridge vectors.
In general, viruses belonging to the Flaviviridae and Bunyaviridae families are less pathogenic than the Togaviridae; however, viral encephalomyelitis caused by any of these pathogens is a potentially catastrophic illness for any vertebrate host. St. Louis encephalitis virus is a flavivirus present in the USA prior to 1999 that has been associated with encephalitis in horses. It is primarily a human pathogen found from central Canada to Argentina and is transmitted among birds by Culex mosquitoes. Encephalitis can be produced experimentally in horses, but most naturally occurring infections in horses are asymptomatic. Japanese B virus is a flavivirus associated with clinical disease in Japan, India, Nepal, and Australia, although mortality is low.
West Nile virus (WNV) has the widest geographic distribution of all of the flaviviruses. Prior to 1999, WNV was recognized in Africa, the Middle East, Asia, and occasionally in European countries. In 1999, WNV infection was first recognized in North America. Since then, the virus has spread throughout the USA and parts of Canada and Mexico. WNV isolated from the outbreak in New York in 1999 appears to be closely related to an isolate recovered from geese in Israel in 1998.
WNV is maintained in an enzootic transmission cycle between wild birds and mosquitoes. It has been recovered from a wide range of North American mosquito species, with Culex spp thought to play the largest role in natural transmission. In the eastern and Midwestern regions of the USA, C pipiens is one of the major vectors, while in the panhandle region and western regions of the USA, C tarsalis is thought to be one of the major vectors.
Both wetland and terrestrial birds may be involved in the natural cycle of WNV, with migratory birds thought to introduce the virus into a geographic region. While a wide range of infected birds (~326) have high, sustained viremia and little or no clinical disease (passerines), fatal infections among corvids (eg, crows, blue jays, and magpies) has been the hallmark of WNV infection in the USA. Ticks have been demonstrated to be infected with WNV, but their role in the natural transmission is unknown. Experimentally, transmission has been documented between cohabitating birds and from oral exposure to WNV in drinking water in birds. Oral transmission has been demonstrated experimentally in several types of raptors.
In humans, other important routes of infection include blood transfusions, organ donation, breast milk, and across the placenta. Sporadic infections and illness have also occurred in several other mammalian species, including dogs, cats, camelids, sheep, and squirrels. Oral transmission has been demonstrated experimentally in cats. Farmed alligators have demonstrated disease and mortality due to WNV, and there has also been a report of WNV-induced disease in crocodiles. Alligators are susceptible to oral infection. In addition to birds, only alligators have consistently demonstrated high enough viremia (104–105 plaque-forming units) to amplify virus, serve as reservoir hosts, and transmit virus back to mosquitoes.
WNV is now endemic throughout the USA, Canada, and Central America and has been detected in several South American countries (Brazil, Peru, French Guiana, Trinidad, and Colombia). Over time, mutations of the virus may have resulted in changes in pathogenesis associated with geography. An increase in virulence has been predicted in isolates from the northeastern USA, while a possible decrease in pathogenesis appears to be occurring in Central and South American isolates. Longterm epizootics are likely to continue.
Cache Valley virus (transmitted by mosquitoes and Culicoides sp among rabbits), Main Drain virus (transmitted by Culicoides varipennis to hares and rodents in the western USA), and snowshoe hare virus (transmitted by Culiseta and Aedes mosquitoes among rabbits in southern Canada and northern USA) have all been identified, although infrequently, as the cause of encephalitis in horses.
The initial clinical signs are similar for the arboviruses; progression of clinical signs and severity of disease are the differentiating features. Initially, horses are quiet and depressed with clinical neurologic signs generally occurring 5 days after infection. Compared to WEE and West Nile encephalitis (WNE), clinical signs of EEE (and VEE) more frequently include altered mentation, impaired vision, aimless wandering, head pressing, circling, inability to swallow, irregular ataxic gait, paresis and paralysis, seizures, and death. Many horses progress to recumbence within 12–18 hr of onset of neurologic abnormalities. Most deaths occur within 2–3 days after the onset of signs. Mortality of horses showing clinical signs from EEE is 50–90%, from WEE 20–50%, and from VEE 50–75%.
The clinical signs and course of disease are highly variable in WNE. There are 2 lineage groups into which isolates of WNV can be placed, based on genetic sequencing. Horses are most susceptible to lineage type I and demonstrate little abnormalities to most African lineage type II. Presenting complaints most often include neurologic abnormalities; other common initial complaints include colic, lameness, anorexia, and fever. Initial systemic signs include a mild fever, feed refusal, and depression.
Neurologic signs are highly variable, but spinal cord disease and moderate mental aberrations are most consistent. Spinal cord disease manifests as asymmetric, multifocal or diffuse ataxia and paresis. Severe manifestations may occur independently in the front- or hindlimbs, unilaterally, or in a single limb. In all clinical studies published to date, >90% of affected horses developed some type of spinal cord signs, while 40–60% developed behavioral changes characterized by periods of hyperesthesia, ranging from mild apprehension to overt hyperexcitability with fractious reactions to aural, visual, and tactile stimuli. Fine and coarse tremors of the face and neck muscles are common and are described in 60–90% of horses. Some horses have periods of cataplexy or narcolepsy that may render them temporarily or permanently recumbent. Coma, blindness, head pressing, and other signs of forebrain disease are seen but are not as common as in alphavirus encephalitides.
Cranial nerve deficits can be seen in all arbovirus infections of horses; these include most cranial nerves with cell bodies located in the mid- and hindbrains. Weakness and/or paralysis of the face and tongue are most frequent. Horses with facial and tongue paresis can be dysphagic, and overt signs of quidding or even esophageal choke can develop. Many horses with severe mental depression and facial paresis will keep their heads low, resulting in severe facial edema. Occasionally, head tilt may be seen. Infrequently in WNE, urinary dysfunction ranging from mild straining to stranguria has been reported.
Horses infected with EEEV, WEEV, and WNV are considered dead-end hosts. Experimental infection of EEEV in horses suggests that viremia can be higher than that achieved from WEEV and WNV. Some horses in these experiments had up to 104 plaque-forming units, which may be close to the threshold for transmission. (The exact threshold for transmission to C melanura has not been determined.) The reservoir and transmission potential of horses infected with EEEV is not fully understood. Horses infected with the sylvatic subtypes of VEEV are also dead-end hosts; however, horses infected with epizootic strains of VEEV have a persistent and significant viremia that results in virus shedding in body fluids. Infection may pass from horse to horse via aerosolized respiratory secretions or direct contact.
During the neurologic phase, horses frequently thrash and injure themselves. Sepsis from trauma in recumbent horses also occurs. Prolonged recumbency leads to pulmonary infections, especially in foals, in which a long duration of slinging and treatment may be pursued more frequently than in large, recumbent animals. Dysphagia leads to decreased water and food intake with renal damage due to concurrent use of anti-inflammatory drugs. Skin and muscle necrosis are common in recumbent horses. Life-threatening trauma can also occur, including a ruptured diaphragm and fractures.
Gross lesions are rare and are limited to small multifocal areas of discoloration and hemorrhage throughout the brain and spinal cord. The brain should be examined microscopically for the presence of nonsuppurative meningoencephalitis. There may be congestion in the meninges of acutely affected animals. Microscopically, there is a non-necrotizing lymphohistiocytic poliomeningoencephalitis. Slight to severe inflammation, characterized by perivascular cuffing of lymphocytes and monocytes, is present. In the neuropil, dying neurons often are surrounded by microglial cells. In EEE, a severe gliosis with necrosis of the neuropil is observed in the cerebrum and extending through the corona radiata to the thalamus. There is severe gliosis and perivascular cuffing throughout all regions of the mid- and hindbrain and cervical spinal cord.
By comparison, there is usually relatively mild gliosis and perivascular cuffing in the cerebrum with WNE, with the most severe disease located in the thalamus and hindbrain. The most severe sites of the WNE lesions can be focally distributed, but usually there is generalized inflammation. In WNE cases with spinal cord disease, the lumbar spine can be severely affected.
No consistent changes in clinical pathology have been found in equine viral encephalitis. With EEEV and WNV in horses, peripheral lymphopenia is common. Hyponatremia is seen and may be due to inappropriate antidiuretic hormone secretion as occurs in encephalitic humans. Horses are frequently azotemic, likely from decreased food and water intake.
A presumptive diagnosis may be made on the basis of clinical signs, the location of the affected horse(s), and season of the year. In terms of antemortem diagnostic testing, analysis of CSF can be a valuable adjunct to presumptive clinical assessment. CSF analyzed from acutely EEEV-infected horses typically has a neutrophilic pleocytosis with markedly increased total solids. Partially immune horses may have predominately mononuclear cells, but nondegenerate neutrophils are still present. While WNV-infected horses can have normal CSF, if abnormal, there is a mononuclear pleocytosis with moderately to markedly increased total solids. In a few horses, virus may be isolated from the CSF of horses with acute infections. By the time neurologic signs are seen, viremia has ended and antigen detection from the plasma of clinically affected horses is of no value.
Serology is the key test for antemortem diagnosis of recent alphavirus and flavivirus infection in horses showing clinical signs. IgM rises sharply and is elevated in 85–90% of clinical arboviral encephalitis patients. Thus, the IgM capture ELISA is the test of choice to detect recent exposure to the virus. Neutralizing antibody titers (primarily IgG) develop slowly during this time and stay elevated for several months. Although neutralizing antibody tests will differentiate between subtypes of these viruses, and are thus considered the gold standard for confirmatory serology, paired serum samples are essential to differentiate field from vaccine exposure. Because virus neutralizing antibodies appear at the end of viremia and may precede the appearance of neurologic signs, paired samples may not show a 4-fold increase in horses with neurologic signs. Paired samples from febrile herdmates may be more diagnostic. Maternal antibodies may interfere with neutralizing responses in young foals. In general, however, they produce IgM similar to adult levels for diagnosis.
Hemagglutination inhibition and complement fixation are fairly unreliable and are only useful for large-scale screening. They must be followed up with more specific (and sensitive) tests. Most horses are sampled once antemortem, and in such cases the IgM capture format is the most reliable and confirmatory test for horses showing clinical signs for the purposes of diagnosis and public health monitoring. Caution should be used in diagnosing VEE in regions where the sylvatic subtypes of the virus are found, because subtypes cross-react in serologic tests.
Several postmortem diagnostic assays are available and, while specific, they vary in sensitivity, depending on the virus. The midbrain and brain stem have the highest concentrations of encephalitic viruses, including rabies virus. Antigen detection methods on fresh tissue include viral isolation and PCR. EEE has higher levels of virus than WNE. Several sections of brain stem and cervical, thoracic, and lumbar spinal cord should be tested for WNV. For formalin-fixed tissue, which is safer for most laboratories, immunohistochemistry and fixed tissue PCR can be used. Both are more sensitive for EEEV than WNV due to higher concentrations and more diffuse distribution of virus throughout the lesions in EEE.
Infectious and noninfectious causes of brain and spinal cord diseases should be considered as differential diagnoses. Infectious causes include alphaviruses, rabies, equine protozoal myeloencephalitis (see Equine Protozoal Myeloencephalitis), and equine herpesvirus 1; less likely causes are botulism and verminous meningoencephalomyelitis (eg, Halicephalobus gingivalis, Setaria spp, Strongylus vulgaris). Noninfectious causes include hypocalcemia, tremorigenic toxicities, hepatoencephalopathy, and leukoencephalomalacia.
Treatment of viral encephalitis is supportive, as there are no specific antiviral therapies. Management is focused on controlling pain and inflammation, preventing injuries associated with ataxia or recumbency, and providing supportive care. Intervention does not appear to significantly affect the outcome of most fulminate EEEV infections. For WNV, flunixin meglumine (1.1 mg/kg, IV, bid) early in the course of the disease decreases the severity of muscle tremors and fasciculations within a few hours of administration.
Recumbent horses that are mentally alert frequently thrash, causing self-inflicted wounds and posing a risk to personnel. Variable responses are observed to tranquilizers and anti-seizure medications, depending on the virus and severity of disease. A sling and hoist may be used to assist horses that are recumbent and have difficulty rising; however, recumbent horses with EEE generally are too comatose for slinging. Dysphagic horses require fluid and nutritional support.
Until equine protozoal myeloencephalitis is ruled out, prophylactic antiprotozoal medications may be instituted. Other supportive measures (eg, oral and parental fluids and nutrition for dehydrated and dysphagic horses) are also important. Broad-spectrum antibiotics should be given for treatment of wounds, cellulitis, and pneumonia. Horses with intermittent or focal neuropathies have a better prognosis than those with complete flaccid paralysis or that appear comatose. Efficacy of specific antiviral agents for the treatment of naturally occurring WNV or EEEV infection is unknown, even in humans. Recent work with passive immunotherapy indicates possible benefit after the onset of clinical signs in WNV models.
Horses with clinical neurologic signs from alphavirus infection that recover have a high incidence of residual neurologic deficits, whereas many horses that recover from WNE have been reported to have no residual neurologic deficits. In EEE, death is frequently spontaneous. With WNE, horses are euthanized due to humane reasons, but spontaneous death does occur. In EEE, most surviving horses exhibit longterm neurologic signs. In WNE, overt clinical signs in horses that recover can last from 1 day to several weeks; improvement usually occurs within 7 days of the onset of clinical signs. While 80–90% of owners report that the horse returns to normal function 1–6 mo after disease, at least 10% of owners report longterm deficits that limit athletic potential and resale value. Deficits include residual weakness or ataxia in one or more limbs, fatigue with exercise, focal or generalized muscle atrophy, and changes in personality and behavioral aberrations.
Formalin-inactivated whole viral vaccines for EEE, WEE, and VEE are commercially available in mono-, bi-, or trivalent form. Previously nonvaccinated adult horses require 2 injections. For adult horses in temperate climates, an annual vaccine within 4 wk of the start of the arbovirus season is recommended. However, for horses that travel between northern and southern areas affected by the virus, injections should be given 2 or even 3 times yearly in very active seasons. Mares should be vaccinated 3–4 wk before foaling to induce colostral antibody.
In foals that have received adequate colostrum from vaccinated dams, vaccination should begin at 5–6 mo of age; foals should receive 2 additional injections at 30 and 90 days after the first injection. It is unclear whether maternal antibody interferes with foal vaccine responses; however, epidemiologic evidence strongly indicates that horses between 4 mo and 4 yr of age are highly susceptible to EEE. If there is early spring (March) activity in the southeastern USA, the two additional injections may be administered 3–4 wk apart. In foals born to nonvaccinated or minimally vaccinated mares, maternal immunity may wane and vaccination should be performed at 3, 4, and 6 mo of age.
Regardless of age, vaccination for young animals should be performed twice in the spring, before the highest levels of EEEV activity (May to June). In Florida, where the highest numbers of EEE cases occur, horses of all ages should receive injections in January and again in April before the peak of the season. A third injection should be administered late in the summer if the season is particularly active.
Four vaccine formulations are available for prevention of WNV in equids. A killed, adjuvanted, whole WNV vaccine is approved for prevention of viremia from WNV. A series of 2 vaccinations given 3–6 wk apart, prior to the period when vectors are active, is recommended in adult horses. In foals, an initial series of 3 immunizations should follow a schedule of vaccination similar to that for EEE and WEE. Foals born to immunized or previously infected mares can start receiving vaccines at 4–6 mo of age.
WNV challenge studies indicate that ~10% of properly immunized horses do not produce neutralizing antibodies to WNV and that 2.3–3% of equine WNV cases seen in the field are in fully vaccinated horses. The duration of immunity from vaccination with the killed, adjuvanted WNV vaccine is unknown. It has been recommended that a booster be given every 4 mo in regions where the virus is exceptionally active throughout the year. The frequency of injection may be minimized to once or twice per year in climates that have short mosquito seasons and limited activity has historically been reported.
A recombinant canarypox vaccine carrying protective prM/E genes of WNV is available and labeled for protection against viremia. In horses previously vaccinated with another product, this vaccine has been shown to induce an antibody response, so vaccines can be interchanged without primary inoculation. Protection against viremia was demonstrated with one dose of the vaccine in naive horses. Ninety percent of vaccinated horses were protected against an experimental intrathecal challenge model with live virus administered within 6 wk of vaccination.
A modified live virus chimera vaccine that has the protective prM/E genes of WNV expressed in a flavivirus vector is also marketed. This vaccine does not contain any adjuvant and is marketed as a single injection product for induction of primary immunity. It is currently the only vaccine labeled for a 12-mo duration of protection against clinical disease and viremia. Induction of immunity is quite rapid, and within 10 days 83% of horses survived intrathecal challenge.
A DNA vaccine in which the prM/E proteins of WNV are expressed in a plasmid is also licensed and marketed.
Protection of horses from arboviruses must also include efforts to minimize exposure to infected mosquitoes. Mosquito mitigation includes applying an insect repellent that contains permethrin on the horse at least daily during vector season, especially at times of day when mosquitoes may be most active. Environmental management is also essential and includes keeping the barn area, paddocks, and pastures cleared of weeds and organic material, such as feces, that might harbor adult mosquitoes. Cleaning water tanks and buckets at least weekly will reduce mosquito breeding areas. Removal of other containers such as flower pots and used tires that may hold stagnant water is essential for reducing the number of mosquitoes in the area.
Options for control of arbovirus infection in other animals emphasize reduction of exposure. Keeping dogs and cats indoors or in a screened area, especially during the time when mosquitoes are most active, reduces exposure. Disposal of dead birds or other small prey that might be eaten may reduce oral exposure. Few arbovirus infections have been documented in dogs and cats; however, exposure to EEEV and WNV is detected in both species during arbovirus seasons. In active years, young dogs have been reported to be susceptible to EEEV infection. Dogs can respond to vaccination with neutralizing antibody.
Clinical cases of both of these diseases have also been recorded in other domestic and exotic animals during active seasons. Emus are exceptionally susceptible to EEEV. These animals are used for the commercial food industry and infection results in high viremia and high amounts of virus shedding rectally, orally, and in regurgitated material in affected animals. Vaccination will prevent viremia, shedding, and disease. Camelids are susceptible to WNV, and numerous reports of disease and pathology were recorded as the virus spread across North America. The killed adjuvanted vaccine marketed for use in horses has been used in camelids without any reports of major adverse effects, and animals have demonstrated production of neutralizing antibody.
People may be infected by all 4 of the arboviruses that commonly cause viral encephalitis in horses. Clinical signs in humans vary from mild flu-like symptoms to death. Children, the elderly, and immunosuppressed people are the most susceptible. People with neurologic disease due to arboviruses usually have permanent neurologic impairment on recovery. Human disease is reported infrequently and generally follows equine infections by ∼2 wk. Veterinarians should be aware of the possibility of human infection and use repellents and other procedures to protect themselves from hematophagous insects when working in sylvatic virus habitats or handling viremic horses.
Veterinarians should take biosecurity precautions when performing necropsies, especially on birds infected with WNV. Live birds may also pose a handling risk due to the very high viral load in cloacal fluid in some species. Appropriate barrier precautions are indicated.
Last full review/revision July 2011 by Maureen T. Long, DVM, PhD, DACVIM