Science Wiki

West Nile fever is an infectious disease caused by West Nile virus (WNV), a mosquito-borne orthoflavivirus of the family Flaviviridae. The virus is maintained in a transmission cycle between birds and Culex mosquitoes. Humans and horses are infected by mosquito bites but are dead-end hosts, because they usually do not develop enough viremia to infect feeding mosquitoes. About 80 percent of human infections are asymptomatic. About 20 percent cause a self-limited febrile illness, which is properly called West Nile fever, and fewer than 1 percent (about 1 in 150) cause West Nile neuroinvasive disease, which includes meningitis, encephalitis, and acute flaccid paralysis. The term "West Nile virus disease" covers both the febrile and the neuroinvasive forms.

WNV was isolated in 1937 in the West Nile District of Uganda. It was recognized for decades as a cause of sporadic and epidemic fever in Africa, the Middle East, South Asia, and Europe. Its introduction to New York City in 1999 was followed by spread across the Americas, and it is now the leading cause of mosquito-borne viral encephalitis in the continental United States. No human vaccine or specific antiviral drug is licensed, and management is supportive. Prevention relies on mosquito control, personal protection, and screening of blood and organ donations.

Classification and genome

WNV belongs to the genus Orthoflavivirus (formerly Flavivirus) and is a member of the Japanese encephalitis serocomplex, together with Japanese encephalitis virus, St. Louis encephalitis virus, Murray Valley encephalitis virus, and Usutu virus. Its genome is a single molecule of positive-sense, single-stranded RNA of about 11 kilobases, with a type 1 5′ cap, no 3′ poly(A) tail, and structured 5′ and 3′ untranslated regions. A single open reading frame is translated into a polyprotein of about 3,400 amino acids, which host and viral proteases cleave into three structural proteins and seven non-structural proteins:

• Structural proteins: capsid (C), precursor membrane (prM, cleaved to M), and envelope (E).
• Non-structural proteins: NS1, NS2A, NS2B, NS3, NS4A, NS4B, and NS5.

The virion is an enveloped particle about 50 nm in diameter. The E protein mediates receptor binding and pH-dependent membrane fusion and is the main target of neutralizing antibodies. Cleavage of prM by host furin in the trans-Golgi network produces the mature, infectious virion. NS3 is a serine protease (with NS2B as cofactor), an RNA helicase, and an RNA triphosphatase. NS5 is a methyltransferase and an RNA-dependent RNA polymerase. NS1 is found in replication complexes, on the cell surface, and as a secreted protein, and it contributes to immune evasion. A subgenomic RNA (sfRNA) derived from the 3′ untranslated region resists degradation by the host exonuclease XRN1, antagonizes antiviral responses, and contributes to replication in mosquitoes and birds.

Most orthoflaviviruses of the Japanese encephalitis serocomplex also have a conserved slippery heptanucleotide sequence and an RNA pseudoknot in the NS2A region. These cause a programmed −1 ribosomal frameshift in a fraction of translation events, which produces an extended form of NS1 called NS1′. NS1′ has been implicated in neuroinvasiveness and in efficient infection of mosquitoes.

Lineages

Phylogenetic analysis separates WNV into several lineages. Only lineages 1 and 2 are commonly associated with human disease.

WNV is a single serotype, and antibodies against one lineage cross-neutralize others, though cross-reactivity with other orthoflaviviruses complicates diagnosis. After its introduction to North America, the NY99 genotype was replaced by a genotype called WN02, which carries a V159A substitution in the envelope protein. In experiments WN02 has a shorter extrinsic incubation period in Culex pipiens and Culex tarsalis, especially at warm temperatures, and it infects the salivary glands earlier, which explains its rapid spread.

Replication cycle

The virus attaches to cells through factors that bind phosphatidylserine on the virion and through lectins such as DC-SIGN, then enters by receptor-mediated endocytosis. Acidification of the endosome triggers a conformational change in E that drives fusion of the viral and endosomal membranes and releases the genome into the cytoplasm. The RNA is translated at the endoplasmic reticulum (ER), and the non-structural proteins remodel ER membranes into invaginated vesicle packets in which replication is hidden from cytosolic sensors such as RIG-I and MDA5. Virions assemble on the ER, mature in the secretory pathway, and are released by exocytosis.

Discovery and early outbreaks

WNV was first isolated in 1937 from the blood of a woman with fever in the West Nile District of northern Uganda. In the 1950s, studies in Egypt showed that infection was common among children, that birds could carry the virus, and that Culex mosquitoes could transmit it. Outbreaks followed in Israel in the 1950s, in South Africa in 1974 (a large epidemic associated with heavy rain), and in France and other countries. Large epidemics with neurological disease began to appear in the 1990s, including outbreaks in Romania in 1996 (centered on Bucharest, with several hundred cases and a high fatality rate) and in Russia in 1999.

Arrival in the Americas

In late summer of 1999, an unusual die-off of crows and an outbreak of encephalitis in humans in Queens, New York, led to the identification of WNV in North America. The source of introduction is unknown. By 2004, the virus had spread west across the contiguous United States, and it reached Canada in 2001 to 2002, Mexico and the Caribbean in 2001 to 2003, and South America in the mid-2000s. In 2002, the United States had the largest recorded outbreak of WNV neuroinvasive disease until then, and the same year provided the first evidence of transmission by blood transfusion and organ transplantation, which led to screening of blood donations by nucleic acid testing beginning in 2003. The number of human cases peaked in 2003, and large seasons have recurred, including in 2012, when more than 5,000 cases were reported in the United States.

Europe and elsewhere

Southeastern Europe had a series of outbreaks beginning in the late 1990s. The introduction of lineage 2 to Hungary in 2004 and its spread through Austria, Greece, Italy, Serbia, Romania, and neighboring countries established that virus transmission had become endemic in much of southern and central Europe. The 2018 season, which was warm, was exceptional, with about 2,000 locally acquired human cases reported in Europe and neighboring countries. Transmission was first documented in Germany in birds and horses in 2018 and in humans in 2019, and later in the Netherlands. Large seasons have occurred in recent years, and the geographic range continues to change.

Enzootic cycle

WNV is maintained by transmission between Culex mosquitoes and birds. Mosquitoes become infected when they feed on a viremic bird. After the virus has infected the midgut, passed into the hemolymph, and reached the salivary glands, the mosquito can transmit it with each subsequent bite. This period is the extrinsic incubation period (EIP), and it shortens as temperature rises. The main vectors differ by region:

• North America: Culex pipiens (northeastern and north-central United States), Culex quinquefasciatus (southern United States), Culex tarsalis (western United States and the Great Plains), and Culex restuans and Culex nigripalpus (early-season and southeastern amplification).
• Europe, Africa, and the Middle East: Culex pipiens, Culex modestus, Culex perexiguus, and Culex univittatus.
• Australia: Culex annulirostris (Kunjin virus).

Birds of the order Passeriformes and the family Corvidae have the highest reservoir competence, with plasma viremia that can exceed 10⁸ plaque-forming units per mL for several days. Species such as American crows, blue jays, house finches, and house sparrows are highly competent, whereas pigeons, chickens, and most mammals are poor reservoirs. The virus has caused high mortality in some bird species, particularly corvids, and mass die-offs of crows are used as a sign of virus activity. Some bird populations in North America declined after 1999.

Mosquitoes of several Culex species feed mostly on birds in spring and early summer and take a larger share of blood meals from mammals, including humans, in late summer and autumn. Studies in North America suggest that the American robin accounts for a share of blood meals far larger than its share of the bird community. The seasonal shift toward mammals coincides with the period when infection prevalence in mosquitoes is highest, and this may contribute to the late-summer peak of human cases. The pattern varies by region and species, and changes in the availability of hosts may explain part of it.

In temperate regions, Culex pipiens overwinters as mated adult females in diapause, and virus has been detected in overwintering females. The extent to which vertical transmission and overwintering allow the virus to persist from one season to the next is debated, and repeated introduction by migratory birds may also contribute.

Non-vector transmission

• Blood transfusion and organ transplantation. Transmission is well documented. Screening of blood donations by nucleic acid testing has greatly lowered the risk but has not eliminated it. Transplant recipients, who are immunosuppressed, are at high risk of severe disease if the organ donor was infected.
• Mother to child. Intrauterine infection has been reported, and the risk of congenital abnormalities is thought to be low. Transmission through breast milk has been described, but is rare.
• Laboratory and other exposures. Needlestick and aerosol infections in laboratory workers and exposure of workers handling infected poultry have been reported. Direct person-to-person spread by casual contact does not occur.

Global distribution

WNV has the widest geographic distribution of any orthoflavivirus transmitted by arthropods. Transmission has been documented in Africa, Europe, the Middle East, South and Central Asia, Australia (as Kunjin virus), North America, Central America, the Caribbean, and South America. In the Americas, severe disease has been much more commonly reported in the United States and Canada than in Latin America, and the reason is not clear. It may involve under-ascertainment, cross-protection from other orthoflaviviruses, or differences in vectors and strains.

United States and Canada

WNV has been detected in all 48 contiguous states and in the southern Canadian provinces. Since 1999, more than 50,000 human cases have been reported in the United States, and about half of them were neuroinvasive. Because most infections are not apparent, modeling studies estimate that several million people have been infected. The annual number of reported cases varies widely, from fewer than 1,000 to several thousand, with large seasons in 2002 and 2003, 2012, 2018, 2021 (a large Arizona outbreak centered on Maricopa County), and 2023. Case numbers in a year depend on the weather, immunity in birds, and the number of infectious mosquitoes in populated areas, and they differ greatly by region, with the highest incidence in the Great Plains and mountain states.

Europe

Human cases in Europe have occurred each summer in several countries, mainly Italy, Greece, Romania, Hungary, Serbia, Spain, and France, with spread to Austria, Germany, and the Netherlands. The number of cases varies strongly from year to year, with peaks in 2010, 2018, and 2022 and 2024 in particular. The European Centre for Disease Prevention and Control publishes weekly reports during the transmission season.

Transmission dynamics

For a vector-borne pathogen transmitted in a single host species, the basic reproduction number R₀ in the classical Ross-Macdonald model is:

where m is the number of mosquitoes per host, a is the biting rate, b is the probability that an infectious bite infects a host, c is the probability that a bite on an infectious host infects a mosquito, μ is the daily mosquito mortality, n is the EIP in days, and r is the host recovery rate. The mosquito-specific part is the vectorial capacity:

For example, if μ = 0.1 per day and the EIP is 7 days, a mosquito survives the EIP with probability e^−0.7 ≈ 0.50. If cool temperatures lengthen the EIP to 14 days, this falls to e^−1.4 ≈ 0.25, which shows how strongly temperature affects transmission.

In the avian community, the multi-host form of the model sums over bird species k. Each species contributes in proportion to the fraction of blood meals φ_k that it supplies and to its reservoir competence:

where b_k is susceptibility, c_k is infectiousness to mosquitoes, and 1/_γ__k is the duration of infectious viremia. The combination of a few highly competent, heavily fed-upon species (such as robins, house finches, and corvids) can sustain amplification. Poorly competent species, including pigeons, chickens, and mammals, divert bites without amplifying the virus, a dilution effect that may reduce risk in diverse communities.

Human spillover risk at time t depends on the number of infectious mosquitoes, the biting rate, and the fraction of bites directed at humans, φ_human(t):

where I_v is the fraction of mosquitoes that are infectious. Risk is highest when early-season bird feeding builds I_v and a later increase in feeding on mammals raises φ_human.

Because most infections are not recognized, reported neuroinvasive cases can be used to estimate total infections. If about 1 in 150 infections causes neuroinvasive disease, then:

For 1,500 reported neuroinvasive cases in one season, this would indicate on the order of 200,000 infections. The ratio varies with age, and neuroinvasive disease is more completely reported than febrile disease.

Temperature, drought, and climate

Mosquito and viral traits respond nonlinearly to temperature, so transmission is a unimodal function of temperature. Trait-based models estimate lower thresholds of about 14 to 19 °C, optima of about 24 to 25 °C, and upper limits of about 32 to 35 °C. Estimated optima are about 24.5 °C for Culex pipiens, about 25.2 °C for Culex quinquefasciatus, and about 23.9 °C for Culex tarsalis. These are several degrees cooler than the optimum for Aedes aegypti transmitting dengue and Zika viruses, and they imply that further warming may lengthen the season and raise the peak risk in cool regions while reducing summer transmission in the hottest ones. Warm years, as in Europe in 2018, are associated with large outbreaks.

In several studies, large outbreaks in cities have been associated with warm weather and with drought or low rainfall in preceding months, even though many mosquitoes increase after heavy rain. Proposed mechanisms, which are hypotheses at varying levels of support, include the flushing of larvae from catch basins by heavy rain and the concentration of nutrients in stagnant water during drought, reduced predation on mosquito larvae when ponds shrink, the aggregation of birds and mosquitoes at scarce water, and stress-related changes in the immune response of birds.

Asymptomatic infection and West Nile fever

The incubation period is typically 2 to 6 days and ranges from 2 to 14 days, and it can be up to about 21 days in immunocompromised people. About 80 percent of infections are asymptomatic. In about 20 percent, a febrile illness develops, with fever, headache, myalgia, arthralgia, fatigue, nausea, and vomiting. A transient maculopapular rash occurs in a minority of patients, and lymphadenopathy may be seen. The illness usually lasts a few days to a week, although fatigue and weakness can last weeks or months.

Neuroinvasive disease

Fewer than 1 percent of infections (about 1 in 150) cause neuroinvasive disease. The risk rises steeply with age, with a risk of about 1 in 50 or higher in people older than 65 years, and it is also increased by immunosuppression, organ transplantation, hypertension, diabetes, and chronic kidney disease. Three main forms are recognized, and they can overlap:

The case fatality among patients with neuroinvasive disease is about 10 percent, and it is higher with encephalitis and in older adults. Death usually results from brain disease or respiratory failure. Among survivors, many have persistent fatigue, weakness, memory and concentration problems, and movement disorders for months or years, and some have permanent paralysis or cognitive deficits. Rare manifestations include Guillain-Barré syndrome, chorioretinitis and other ocular disease, myocarditis, hepatitis, pancreatitis, and rhabdomyolysis. Chronic kidney disease has been reported in some follow-up studies, and the persistence of virus in the kidney in some patients is a subject of research.

After a mosquito bite, the virus replicates in skin cells, including keratinocytes and dendritic cells such as Langerhans cells, then travels to regional lymph nodes and spreads through blood. Viremia is brief and low in titer in humans and usually cleared within about a week, so by the time symptoms of neuroinvasive disease appear, virus is rarely detectable in blood.

The virus enters the central nervous system by several proposed routes, including crossing of the blood-brain barrier after disruption by inflammatory cytokines such as TNF-α, infection of endothelial cells, transport inside infected leukocytes (the "Trojan horse" mechanism), and retrograde axonal transport from peripheral nerves. Neurons are the main target, especially in the brainstem, thalamus, basal ganglia, cerebellum, and the anterior horn of the spinal cord, and damage results from direct viral injury and from inflammation and immune-mediated injury. Both innate immunity (type I interferon, the RIG-I and MDA5 sensors, and Toll-like receptors 3 and 7) and adaptive immunity, in particular antibodies and CD8⁺ T cells, are required to control infection and to limit spread to the brain. IgM and neutralizing antibodies appear within a week, and they are important in clearance.

Host genetic factors influence severity. Homozygosity for the _CCR5_Δ32 deletion has been associated with an increased risk of symptomatic disease and neuroinvasion in several studies. Variants in the interferon-induced gene OAS1 and related genes have also been linked to susceptibility, and resistance to flaviviruses in some inbred mouse strains has been traced to the Oas1b gene. Immunosuppression, and the decline in immune function with age, increase the risk of severe disease.

Clinical diagnosis is difficult because the febrile illness is not specific. In neuroinvasive disease, the main differential diagnoses include herpes simplex virus encephalitis, enteroviral meningitis, St. Louis encephalitis, Japanese encephalitis, tick-borne encephalitis, Powassan virus infection, bacterial meningitis, and non-infectious causes. Laboratory diagnosis includes:

• Serology: Detection of IgM in serum or, preferably, in CSF by IgM antibody capture enzyme-linked immunosorbent assay (MAC-ELISA) is the standard method. IgM in CSF indicates infection of the central nervous system because IgM does not cross the blood-brain barrier. IgM can persist in serum for months to more than a year, so a positive result may reflect an earlier infection. Cross-reactivity with other orthoflaviviruses (such as St. Louis encephalitis virus, dengue virus, or vaccine-induced antibodies) is common, and a plaque reduction neutralization test (PRNT) is used to confirm the result.
• Nucleic acid amplification: RT-PCR of serum, plasma, or CSF has high specificity but low sensitivity in immunocompetent patients because viremia is brief and low-titer and is usually over by the time symptoms occur. It is more useful in immunocompromised patients, who can have prolonged viremia and delayed antibody responses, and in screening of blood donations.
• Other tests: CSF analysis shows a lymphocytic pleocytosis and raised protein. Magnetic resonance imaging may show abnormalities in the deep gray matter, brainstem, or spinal cord, but is normal in many patients. Electromyography can distinguish anterior horn cell disease from Guillain-Barré syndrome.

No specific antiviral treatment has been proven to be effective. Management of both the febrile illness and neuroinvasive disease is supportive. It includes fluids, analgesics, antipyretics, and, in severe cases, hospital care with respiratory support, treatment of raised intracranial pressure and seizures, and rehabilitation for weakness and paralysis. Several agents have been tried, including ribavirin, interferon, and intravenous immunoglobulin (IVIG), and some patients have received IVIG or plasma that contains high-titer anti-WNV antibodies. The evidence from randomized trials is limited, and no treatment is established. Candidate antivirals and neutralizing monoclonal antibodies have been effective in animal models.

Mosquito control and surveillance

Public health agencies monitor the activity of WNV with mosquito collections, dead-bird reports, sentinel animals, equine cases, and human cases. Mosquitoes are collected with CDC light traps (baited with carbon dioxide) and gravid traps (baited with fermented organic infusions) and tested in pools by RT-PCR. For M pools of equal size s of which Y are positive, the maximum likelihood estimate of infection prevalence is:

The vector index combines abundance and infection prevalence, summed over the competent species i:

where N_i is the mean number of females of species i collected per trap-night and P_i is the estimated infection prevalence. It estimates the number of infected mosquitoes collected per trap-night. Programs use local threshold values to trigger intensified control, and human cases commonly follow high values after a delay of a few weeks.

Control includes:

• Source reduction and larviciding: Management of stormwater catch basins, drainage, and water storage, with larvicides such as Bacillus thuringiensis israelensis (Bti), Lysinibacillus sphaericus, methoprene, and spinosad applied to breeding sites.
• Adulticiding: Ultra-low-volume spraying of pyrethroids or organophosphates from trucks or aircraft, usually at night when Culex are active. It is generally reserved for situations in which surveillance indicates elevated risk to humans. Evaluations of aerial spraying during outbreaks have reported lower incidence in treated areas, but these comparisons are observational. Resistance to insecticides is common in Culex populations.

Personal protection

Individual measures include repellents containing DEET, icaridin, IR3535, or oil of lemon eucalyptus, long clothing, particularly at dusk and dawn, when Culex are most active, window screens, and elimination of standing water around the home. Older adults and people with immunosuppression have the most to gain from these measures.

Blood and organ safety

Blood donations in the United States and in other countries are screened by nucleic acid testing, either individually or in minipools, in the transmission season. Donors and recipients of organs are screened in some settings, and testing of transplant recipients with unexplained neurological illness is recommended.

Vaccines

No vaccine is licensed for use in humans. Candidates include inactivated, live attenuated and chimeric, subunit, DNA, and mRNA vaccines, and some have been tested in phase 1 and 2 clinical trials. Large efficacy trials are difficult because of the unpredictable timing and location of outbreaks. Several vaccines are licensed for horses, including an inactivated whole-virus vaccine and a recombinant canarypox-vectored vaccine, and they greatly reduce the incidence of equine neuroinvasive disease. A vaccine for birds has been tested in endangered species, such as the California condor.

Horses are the domestic animals most commonly affected by clinical disease. Most equine infections are inapparent, but a minority develop encephalomyelitis with ataxia, weakness, and recumbency, and the case fatality in clinical cases is about one third. Large outbreaks in horses occurred in the United States in 2002 and 2003 and in Australia in 2011. Wild birds, especially corvids, raptors, and some owls, can develop fatal disease, and declines in some populations were reported after the introduction of the virus to North America. Other animals, including alligators, dogs, and some zoo animals, have been infected.

Active research addresses the long-term effects of neuroinvasive disease, including cognitive and renal outcomes, the mechanism of entry to the central nervous system, the host genetic factors that determine severity, the contribution of vertical transmission and overwintering to persistence, the roles of different bird species in amplification, and the influence of climate and land use on risk. Other areas include the development of a human vaccine and effective antivirals, improved surveillance based on mosquito and bird data, and methods for forecasting outbreaks from weather and environmental data.

Why are humans and horses considered dead-end hosts for West Nile virus?

In passerine reservoir birds (such as corvids, robins, and house finches), West Nile virus replicates to extreme plasma viremia levels exceeding 10⁸ plaque-forming units (PFU)/mL, which readily infects blood-feeding *Culex* mosquitoes. In humans and horses, peak viremia is brief and low-titer (typically <10³ PFU/mL), falling well below the oral infection threshold of the mosquito midgut.

Why does urban drought paradoxically increase West Nile fever outbreaks?

Unlike floodwater mosquitoes, urban *Culex pipiens* and *Culex quinquefasciatus* breed in organically rich subterranean stormwater catch basins. Heavy rainfall flushes larvae out of these drains, whereas drought concentrates decaying organic nutrients in stagnant basins, eliminates aquatic predators, and forces birds and mosquitoes to congregate around scarce residual water sources, accelerating enzootic amplification.

How does West Nile poliomyelitis differ from Guillain-Barré syndrome?

West Nile poliomyelitis is caused by direct cytolytic infection and destruction of anterior horn lower motor neurons in the spinal cord, producing sudden, asymmetric flaccid paralysis during acute fever with CSF pleocytosis and preserved sensory nerve conduction. Guillain-Barré syndrome is a post-infectious autoimmune demyelinating polyneuropathy that typically presents weeks later with symmetric ascending weakness and elevated CSF protein without white blood cell elevation.

• Zika fever
• Dengue fever
• Japanese encephalitis
• St. Louis encephalitis
• Arbovirus
• Culex