In the decade since the emergence of West Nile virus (WNV) in North America, our collective understanding of the clinical manifestations, prognostic factors and short- and long-term outcomes of illness has increased dramatically. Previous historic outbreaks of WNV in sub-Saharan Africa and the Middle East had typically been characterized as mild, self-limited febrile illness with few or no sequelae [1]. More recently, beginning in the 1990s, more cases of neurologic illness, particularly “neuroinvasive” disease, in which the virus directly infects the nervous system, began to be recognized [1,2]. However, the ongoing outbreak of WNV in North America has greatly expanded our understanding of the spectrum of illness associated with WNV infection in humans, and a number of previously under-recognized syndromes have been characterized [3,4,5,6].

It is generally estimated that about 80% of persons infected with WNV remain asymptomatic. Of those who develop symptoms, the vast majority develop an acute, systemic febrile illness (“West Nile fever”, WNF). Data suggest that less than 1% of infected persons develop neurologic illness, which is primarily attributed to neuroinvasive disease, in which the virus breaches the intrathecal space and produces infection of central nervous system (CNS) structures. Neuroinvasive disease includes aseptic meningitis (“West Nile meningitis”, WNM), encephalitis (“West Nile encephalitis”, WNE) or an acute poliomyelitis-like syndrome (“West Nile poliomyelitis”, WNP) [1,2]. WNM involves infection of the meninges (the outer covering of the brain and spinal cord) and makes up the largest percentage of neuroinvasive disease cases in younger age groups. WNE involves viral infection of the brain parenchyma itself and is more typically manifested in older persons or individuals with compromised immune systems. WNP results from viral infection of the anterior horn cells of the spinal cord, leading to acute flaccid limb weakness. When discussing the various manifestations of WNV infection, it is important to keep in mind that the clinical picture may not always be as clear-cut as delineated above, and sometimes, the clinical features of WNF, WNM and WNE may overlap. For example, patients may develop an altered mental status, due to severe systemic illness, without true histopathologic or radiologic evidence of cerebral inflammation or “encephalitis” in the pathophysiologic sense. Similarly, patients presenting with fever, headache and “neck stiffness” may not undergo lumbar puncture to demonstrate pleocytosis, and consequently, a diagnosis of WN “meningitis” may not be reported. Despite these limitations, the clinical syndrome in most persons with WNV illness can be diagnosed based upon clinical grounds.

As the clinical experience with WNV infection has extended over a period of years, our understanding of the short- and long-term outcomes of WNV infection has grown. However, detailed data regarding these long-term neurologic and functional outcomes of WNV is still relatively sparse. As with other viral encephalitides, initial severe neurologic illness does not necessarily correlate with eventual outcome, and some patients with initial severe encephalopathy with associated coma may experience good recovery and minimal sequelae [3]. However, others experience persistent neurologic dysfunction, including movement disorders, headaches, fatigue and cognitive complaints. Large hospital-based series suggest that patients with severe WNE frequently require assistance with daily activities following acute care discharge [4,5]. Patients frequently report functional and cognitive difficulties for over a year following acute infection, and only 37% of patients in the 1999 New York City outbreak achieved full recovery at one year [71]. Of 265 persons developing symptomatic WNV infection in Idaho between 2006 and 2008, 53% reported one or more persistent symptom six months or more following acute illness; the most frequent complaints were fatigue, muscle aches and difficulties with memory and concentration [72]. Cognitive complaints, including difficulties with attention and concentration, have been described among patients recovering from WNE and suggest a subcortical type of cognitive dysfunction based on prominent thalamic and basal ganglia involvement [3]. However, limited formal and objective neuropsychometric assessments have been performed. A few studies have shown that persons recovering from WNV illness demonstrate measurable neurocognitive deficits on standardized testing as long as one year after acute illness [73,74]. Other studies, however, have shown that persons recovering from WNV illness do not perform significantly differently on standardized neurocognitive assessments based upon the nature of clinical illness (e.g., WNF compared to WNM or WNE) or in comparison to unaffected persons [15]. However, self-reported fatigue, somatic and cognitive complaints are common among persons recovering from WNV illness, and subjective complaints and poorer performance on self-reported functionality indices have been seen in patients months or years following acute illness [71,75]. One study has suggested the normalization of self-reported symptoms within one year of acute illness [76]. As previously noted, neuropsychiatric symptoms, including depression and severe anxiety, have been reported by patients recovering from WNE [3,35,77].

Most children with symptomatic WNV infection present with WNF; of those who develop neuroinvasive disease, it most frequently manifests as meningitis [78,79,80]. However, severe and fatal encephalitis [81], poliomyelitis [82,83], rhombencephalitis [84] and hepatitis [85] have all been described in children with WNV infection. Similar to adults, immunocompromised children may be more susceptible to more severe illness [21].

Of all infected persons, less than 1% develop West Nile virus neuroinvasive disease (WNND). Although WNND has been reported among all ages, the proportion of persons who progress to WNND is greater among older compared to younger persons [8]. Serologic surveys in Romania and New York City indicate that WNV infection incidence is constant across all age groups during outbreaks [86,87], although serosurvey data are limited. Surveillance data from the United States indicate that age is the most important host risk factor for the development of WNND after infection. The incidence of neuroinvasive disease increases approximately 1.5-fold for each decade of life, resulting in a risk approximately 30 times greater for persons 80–90 years old compared to children younger than 10 years old [8]. During outbreaks, hospitalized persons older than 70 years of age had case fatality rates of 15% in Romania [86] and 29% in Israel [88]. Encephalitis with severe muscle weakness and change in the level of consciousness were also prominent clinical risk factors predicting fatal outcome [88,89].

Based upon a limited number of cases, patients who acquire WNV infection from infected donor organs are likely at a higher risk for severe neurologic disease and death compared with patients infected through the natural route of mosquito bite inoculation [90,91]. The risk of severe neurologic disease among other organ transplant recipients is not well-defined and may be related to the interval between infection and transplantation or the type of post-transplant immunosuppressive therapy. A seroprevalence study carried out in a Canadian outpatient transplant clinic following a WNV epidemic in 2002 indicated that the risk of neuroinvasive disease following infection was 40% (95% confidence interval 16%–80%) [92]. During that epidemic, transplant patients were approximately 40 times more likely than the general population to develop WNND [93]. However, a subsequent study assessed seropositivity and incidence of WNND among 194 solid organ transplant recipients and 195 controls and found no significant difference in seropositivity for WNV IgG between the groups and determined that the incidence of WNND among the transplant recipients was low among the seropositive transplant patients [94].

In addition to increased age and organ transplantation, hypertension, cerebrovascular disease, renal disease and diabetes have also been identified as possible risk factors for WNND, and prior immunosuppression has been associated with a fatal outcome [88,89,95,96,97,98,99]. The incidence of neuroinvasive disease and the probability of death after acquiring neuroinvasive disease are slightly higher in men than in women [8].

There is currently no definitive treatment for WNV infection. Prevention of infection through protection from mosquito bites is therefore critical and the single most important public health measure. In the absence of definitive antiviral treatment, management of illness due to WNV infection remains supportive. Patients with otherwise uncomplicated WNF generally do not require specific intervention, though control of headache and rehydration may sometimes be needed. However, persons with documented WN viremia and patients with WNF in which other risk factors, including older age and underlying immunosuppression, are present should be observed for progression to more severe neuroinvasive disease. Patients with severe WNM may also require pain control for severe headache, and dehydration due to associated nausea and vomiting may require hospitalization for rehydration. In patients with WNE, attention to the level of alertness and airway protection is important. While seizures and increased intracranial pressure have been infrequently reported with WNE, if present, they should be managed appropriately.

Patients with WNP may not have concurrent meningitis or encephalitis, and thus, WNV infection may not initially be suspected. This may result in the implementation of inappropriate diagnostic procedures or treatment modalities, including anticoagulation for suspected acute stroke or muscle biopsy for suspected myopathy. WNV infection should be suspected in persons developing acute asymmetric paralysis, particularly if accompanied by other signs of infection. Patients developing early dysarthria and dysphagia are at a higher risk for subsequent acute respiratory failure [18]; for this reason, hospitalization and observation of patients with poliomyelitis is prudent, and the development of dysarthria and dysphagia should be viewed with concern. Management of poliomyelitis due to poliovirus suggests that the initiation of aggressive physical activity during the acute febrile period of illness is associated with more profound and persistent weakness [100]; in the absence of additional data, the avoidance of aggressive physical activity during the acute febrile illness or during the initial 48–72 h of weakness in WNP would be a reasonable approach, with subsequent application of physical and occupational therapy.

Viremia in humans is short-lived, and WNV is usually cleared from the system by the time of clinical presentation; this fact presents a substantial theoretical obstacle for the targeting and design of specific anti-viral therapies. The recent use of several therapeutic modalities, including antiviral agents, nucleic acid analogues, missense sequences, immunomodulating agents and angiotensin-receptor blockers, has been outside the setting of carefully controlled, randomized, blinded, placebo-controlled trials; thus, anecdotal reports of the effectiveness of these agents are unsubstantiated. However, such anecdotal reports of effectiveness and the clinical desire to provide an intervention have led to empiric use.

The immunomodulating agent, interferon-α, while again showing in vitro inhibition of cytotoxicity due to WNV [101], has not been fully evaluated in animal models, and data from an open-label, non-blinded trial in the United States have not suggested a clear benefit. The use of interferon-α in the treatment of the closely related Japanese encephalitis virus suggested no benefit [102].

Animal models and anecdotal reports have suggested the efficacy of high-titer WNV-specific intravenous immune globulin (IVIG) from pooled donors (Omr-IgG-am®) [103,104,105], and humanized monoclonal WNV antibodies targeting the envelope protein of the virus (MGAWN1) [106,107,108,109]. However, animal models suggest that efficacy is greatest if these therapeutics are given prior to or very shortly after the onset of clinical illness, and attempts at human randomized clinical trials to assess the efficacy of these therapeutic agents have been unsuccessful, largely due to the challenge of enrolling a sufficient number of subjects within a likely therapeutic window. Neither of these products is licensed or available for use in the United States. As a result, the prevention of initial infection is the most critical component of any approach for the management of WNV illness.

Over the past decade, the understanding of the clinical spectrum of illness, as well as the immediate and longer-term outcomes associated with human WNV infection has increased substantially. However, there are remaining clinical questions that require further elucidation. Data on the long-term neurocognitive impact on patients recovering from WNE are scant, and further information is needed to ascertain long-lasting cognitive impairment following encephalitis from WNV. The parkinsonian features associated with acute WNV illness appear in most cases to be transient and resolve over time; however, recurrent- or early-onset parkinsonism in such patients due to the senescence of dopaminergic neurons remains a hypothetical possibility. Similarly, whether patients recovering from WNP will develop recurrent limb weakness in previously affected limbs years after their acute illness, akin to ‘post-polio syndrome’ seen with poliovirus, is unknown at this point, but needs assessment. In the future, additional assessment of these and other clinical manifestations of WNV infection will be critical in aiding our understanding of the pathogenesis of WNV disease and hopefully will guide management and treatment options.