West Nile Virus in the United States (1999–2005)
The accidental introduction of West Nile Virus into New York City from the Old World in 1999 resulted in an epidemic in humans, horses, and birds that swept to the west coast in just 3 years. The virus is transmitted by infective mosquitoes among susceptible native birds, which serve as amplifying hosts. Clinical disease occurs in humans and horses, but not enough virus is produced in their blood to infect other mosquitoes; therefore, humans and horses are considered dead-end hosts. Humans can best protect themselves by remaining indoors during periods of high mosquito activity and/or by using recommended repellents. Effective vaccines are available for horses.
Introduction
The introduction of West Nile Virus (WNV) into New York City in 1999 was the start of the most significant invasion of the United States by an arbovirus (i.e., arthropod-borne virus) since this method of disease organism transmission was discovered at the turn of the 20th century. The resulting epidemic has far surpassed the extent and duration of the Venezuelan equine encephalitis and the indigenous St. Louis encephalitis outbreaks of the 1970s.1,2 West Nile Virus is a flavivirus related to the St. Louis and Japanese encephalitis viruses; it was first isolated from a human patient in the Ugandan province of West Nile in 1937.3 It has been responsible for a number of scattered outbreaks in humans and some animals, notably in Israel (1951 and 1957), France (1962 to 1965), South Africa (1974), India (1980 to 1981), and the Ukraine (1985); however, it was not considered a significant public health problem in its Old World range (i.e., Africa, Europe, and western Asia).4–10 At times, humans, birds, and mammals seropositive for WNV were reported in seroprevalence surveys, but such findings were not associated with human or animal disease.11
More significant outbreaks began to occur in the 1990s, especially in the Mediterranean basin and eastern Europe, that resulted in both human and equine mortality [Table 1].12 The human outbreak in Romania was the first in an urban setting, where risk factors included adult mosquitoes in the houses and mosquito larvae in the basements of affected patients.13,14 Significant mortality in birds was first documented in Israel (1997 to 1998), with deaths in storks, raptors, and domestic geese.15 The purpose of this paper is to provide a general overview of the epidemiology, introduction, and spread of WNV in the United States.
Introduction and Spread in the United States
The first human cases of West Nile encephalitis in the United States occurred in Queens, New York City in August 1999.16,17 The disease was first thought to be St. Louis encephalitis, which has occurred sporadically in many areas of the United States.18 Bird mortality, which is not a normal occurrence with St. Louis encephalitis, was noted among both native birds (crows) and some captive zoo birds.16,19 These findings led to further testing, which demonstrated that the disease agent was WNV, specifically an antigenic strain related to that isolated from the domestic goose outbreak in Israel (1997 to 1998).20–22 It is not known how the virus was introduced into New York, but it is theorized that infected adult mosquitoes were brought from the Old World in airplanes or shipping containers. The phenomenon of airport-associated malaria occurs when adult mosquitoes infected with human malaria parasites are transported by plane from malarious areas and infect people working or living at or near airports.23,24
The virus was also detected in crows and horses in Connecticut, New Jersey, and Maryland in 1999. All human cases (n=62) and fatalities (n=7) were confined to New York state.25,26 In 2000, the virus was detected in nonhuman hosts throughout upstate New York; all of Connecticut and Rhode Island; parts of Vermont, New Hampshire, New Jersey, and Pennsylvania; scattered counties in Maryland, Delaware, and Virginia; and one county in North Carolina. All 21 human cases and the two deaths in 2000 were confined to New York, New Jersey, and Connecticut.27 The year 2001 saw a significant expansion of the virus to the west and south. Nonhuman isolations occurred sporadically in the eastern third of the United States, with human cases (n=66; 29 deaths) in New York, Massachusetts, Connecticut, Pennsylvania, Maryland, Georgia, Florida, and Alabama.27 In 2002, the virus spread throughout the southeastern and midwestern United States and the plains states. Over 3000 human cases were reported, with 211 deaths.27 Most cases were reported in Illinois and Michigan, and significant outbreaks occurred in Louisiana and Mississippi.
To date, 2003 was the high watermark, with >8900 human cases and 232 deaths.28 Most human cases occurred in the northern plains and mountain states (49% in Colorado and Nebraska). Fewer cases and fatalities were reported for 2004 (cases=2470; deaths=88) and 2005 (cases=2148; deaths=59) to date. California led the nation in human cases in 2004 and 2005, with most cases shifting from southern California to northern parts of the state in 2005. South Dakota was second in the number of human cases in 2005 and is an exception to the general course of events, where large numbers of cases occur in a given region and then subside in subsequent years [Figures 1, 2].29,30
Epidemiology
West Nile Virus is primarily transmitted to humans, birds, and other vertebrates by the bites of competent, infective mosquitoes [Figure 3]. As many as 59 species of mosquitoes have been found with WNV in their bodies.31–33 Some of these isolations are probably the result of the female mosquito having blood in her digestive tract from a viremic host at the time of capture. Whether or not the mosquito becomes a competent vector (i.e., able to infect subsequent hosts) depends on the titer of virus in the original host and the ability of the virus to replicate in the mosquito and migrate to the salivary glands.34,35 Mosquito ecology, host preference, behavior, and abundance also play roles in how important a particular mosquito species is in maintaining and transmitting the virus.
Culex (Cu.) species are the primary vectors of WNV among birds and between birds, humans, and horses.36 Culex quinquefasciatus and the Cu. pipiens complex are considered the primary vectors in the eastern United States, while Cu. tarsalis and Cu. nigripalpus are prominent vectors in the west. The transmission cycle is primarily between susceptible birds and competent mosquito species. Humans and horses become infected when infective mosquitoes bite them; but they do not themselves contribute to the maintenance of the virus in nature, because they lack sufficient virus titer to infect subsequent mosquitoes that feed on them.36–38
Corvids (e.g., crows, magpies, jays) are particularly susceptible to WNV and have high viremia and subsequent mortality. In fact, crows and blue jays killed by the virus are used as an early indicator of WNV activity and generally precede human cases. Mortality in horses is not a reliable warning system for WNV compared to Venezuelan equine and St. Louis encephalitis, because horse morbidity often develops concurrently with human WNV disease (not before). In 72% of the counties reporting activity during the 2002 outbreak, WNV-infected birds were the first indicators of activity in the county, with a median lead time of 38.5 days (range 2 to 252 days) before the first human cases. The lead time exceeded the estimated maximum incubation time in humans (14 days) in 85% of those counties.39
The seasonal incidence of WNV in temperate areas is similar to that of other mosquito-transmitted disease agents, with peak transmission occurring from July through October.12,39 Environmental factors, particularly temperature and moisture, play important roles in the initiation and progress of mosquito-borne arbovirus outbreaks. Mosquito larval development, female blood feeding, and egg maturation all speed up as the temperature increases, resulting in higher overall mosquito populations and increased frequency of contact with hosts. The development of the virus in the mosquito is also temperature-dependent, with higher temperatures resulting in a shorter extrinsic incubation period (i.e., the time between ingestion of virus and virus being present in the salivary secretions to allow infection of new vertebrate hosts). In order for mosquitoes in a particular population to effectively transmit virus, a significant proportion of the females must survive longer than the extrinsic incubation period. Moisture in the form of rainfall contributes to the number, extent, and duration of larval habitats, and thus overall adult populations. High relative humidity contributes to increased adult survival.12,39–42
West Nile Virus in Humans
Table 2 presents an overview of the demographic and general clinical picture for the 2002 outbreak in humans in the United States.39 Neurological cases (e.g., meningitis, encephalitis, combined meningoencephalitis) accounted for 71% of the diagnosed cases, while 28% had only a fever. Among the neurological cases, 9% were fatal, with mortality increasing with age. Acute flaccid paralysis increased proportionally between 2002 and 2004. Granwehr et al. provided an excellent review of the varied clinical features and sequelae of WNV during this time period.31 Other patients who developed movement disorders during the course of WNV infection generally experienced recovery of some or all of their neuromotor function. Acute flaccid paralysis patients, however, have had a poor prognosis, with continuing social and financial burdens.
Exposure to infective mosquitoes is the primary risk factor for human WNV infection; however, several other means of transmission have been documented. Human blood intended for transfusion is now subject to molecularly based tests because of several incidences of WNV in recipients of WN-viremic blood.43,44 Intrauterine transmission was documented in one woman in 2002; however, >70 other women who contracted WNV during pregnancy delivered apparently WNV-free and healthy children.33,43 Transmission by organ transplantation, laboratory infection, and possible aerosol transmission to turkey handlers have also been documented.43–46
Bird Hosts
Avian hosts vary greatly in their response to WNV infection in terms of susceptibility, maintenance of virus titer, and their role in the maintenance and spread of the disease organism. Passerines (especially corvids) suffer the highest mortality, while no clinical signs were seen in selected members of the eight other avian orders, including Falconiformes, Galliformes, and Columbiformes. High viral titers (>105 plaque-forming units/mL) in common birds such as the crow, blue jay, house sparrow, and house finch in the geographical vicinity of human outbreaks, suggest that these species act as reservoirs or amplification hosts for vector mosquitoes.47 While WNV has been detected in dead birds of at least 138 species, low viremia in some species suggests that WNV may be only a contributory factor in many of the natural deaths observed in the New World.47,48 Recent Christmas bird counts (i.e., counts taken at a specified time around Christmas) in the northeastern United States indicate that WNV has not had a lasting influence on the winter populations of 10 selected bird species.49
Horses
Over 23,000 equine cases of WNV have been reported in the United States since its introduction in 1999 [Figure 4]. Most affected horses have fever and central nervous system involvement with ataxia, weakness of limbs, and difficulty rising. Fatality rates, including euthanized animals, have been between 38% and 57% in unvaccinated animals.50,51 The geographic spread of WNV in horses in the United States was similar to that seen for humans [Figures 1, 2], but morbidity and mortality in horses have declined since the introduction of a vaccine in 2002. Starting with the 2006 mosquito season, several vaccines with different modes of action will be available for horses. Vaccination of horses appears to be successful, as evidenced by the substantial drop in equine cases after 2002 compared to the continued rise of human cases through 2003.
Personal Protection
The only effective way for humans to protect themselves from infection by WNV is through personal protection. Most of the effective mosquito vectors are most active at dusk and dawn, with some activity through the night. Mosquitoes can be avoided by remaining in a well-screened residence during peak activity times or by using an effective insect repellent. According to the US Center for Disease Control, Division of Vector-borne Infectious Diseases, DEET (N, N-diethyl-m-toluamide) and picaridin (KBR 3023) provide the longest-lasting protection against blood-feeding arthropods. The plant-based repellent, oil of lemon eucalyptus (p-menthane 3,8-diol [PMD]), provides protection similar to that of low concentrations of DEET. Picaridin and DEET are safe when used as directed, but the protection provided varies with the concentration of active ingredient, body temperature, and perspiration. The label for oil of lemon eucalyptus states that it should not be used on children <3 years of age. The application and use of picaridin and DEET are not limited by age, pregnancy, or lactation.52–54
Because several common native mosquitoes are competent vectors and wild birds all over North America are susceptible, WNV has not been stopped or contained by public health officials. Most public health activities have been confined to surveillance, research, and public education. Communities with effective mosquito abatement programs have only been marginally better protected than other communities. The best advice given by public health officials to date has been to avoid outdoor activity when mosquitoes are active and to use repellents.
Conclusion
Increases in the speed and volume of trade and human travel have changed many aspects of people’s lives and professions. The introduction of a relatively obscure arbovirus from the Old World to New York City in 1999 resulted in an epidemic in humans, horses, and birds that swept throughout the United States, reaching the west coast in only 3 years. To date, there have been >18,000 human cases of WNV, with >600 fatalities. The WNV epidemic has resulted in increased awareness of arthropod-transmitted diseases by both the general public and public health agencies. This has, in turn, prompted increased surveillance, research, and education efforts that will be invaluable when the next exotic, insect-transmitted pathogen is introduced in the United States.
Acknowledgments
The author gratefully acknowledges the invaluable assistance provided by Sonya Dexter for manuscript preparation; David Paulsen for figure and table composition and preparation; and David Trently for providing sources concerning bird survival.
Citation: Journal of the American Animal Hospital Association 42, 3; 10.5326/0420170
Citation: Journal of the American Animal Hospital Association 42, 3; 10.5326/0420170
Citation: Journal of the American Animal Hospital Association 42, 3; 10.5326/0420170
Citation: Journal of the American Animal Hospital Association 42, 3; 10.5326/0420170
Distribution of human and nonhuman West Nile Virus reports in the United States, 1999–2002. (Courtesy of Centers for Disease Control and Prevention.)
Distribution of human West Nile Virus cases in the United States, 2004 and 2005 (through October 25, 2005). (Courtesy of United States Geological Survey.)
Generalized West Nile Virus transmission cycle. (Courtesy of Centers for Disease Control and Prevention.)
Reported cases of West Nile Virus in equines, 1999–2005. (Original figure by D. Paulsen, using data from the Centers for Disease Control and Prevention and the United States Geological Survey.)
