Historical Data on Outcomes of Wnv Infection

Reports of human WNV infection from areas of endemicity in Africa and the Middle East suggest that WNV infection in these areas has historically resulted in mild, uncomplicated disease in the small proportion of individuals (primarily children) developing symptomatic illness [2]. Long-term outcomes in these areas have not been detailed but could be presumed to be favorable and unremarkable. More recent epidemics of WNV infection in Israel and Eastern Europe have been more consistent with the North American experience, with more frequent development of neurologic illness, particularly in older persons or in those who are immunocompromised [3, 4]. However, data on long-term outcomes and sequelae following acute illness during these outbreaks have not been well-described.

Because of the sheer number of cases and the ongoing nature of the epidemic, the North American experience has resulted in a tremendous increase in our understanding of short-term clinical features of WNV-associated syndromes. Most human infections are subclinical, and most persons developing illness develop febrile illness (e.g., West Nile fever [WNF]), accompanied by other systemic symptoms, including myalgias, malaise, headache, and gastrointestinal distress. The minority of patients develop neuroinvasive disease, including West Nile meningitis (WNM), West Nile encephalitis (WNE), and West Nile poliomyelitis (WNP) [2, 5]. The acute features of each of these syndromes are shown in table 1.

Short-Term and Long-Term Outcomes of Wnv Infection

The distinction between a “short-term” and a “long-term” outcome of illness is somewhat arbitrary. For the purposes of this discussion, a “short-term” outcome is defined as including the acute illness and the immediate posthospitalization and convalescent period; a “long-term” outcome is defined as including events that occur for months to years following acute illness. Although much has been described about the acute and short-term features of human WNV infection, far less is known about possible long-term or chronic sequelae.

WNF. During the early years of the outbreak of WNV infection in North America, WNF was frequently described as a mild, benign, self-limited illness. Indeed, this manifestation tends to occur in younger individuals and is generally associated with few physical or neurological sequelae. However, recent data suggest that WNF may be associated with more-significant ongoing problems than were initially appreciated. WNF in older adults may be associated with significant mortality [6]; although the ultimate cause of death in fatal cases of WNF is not entirely clear, the cause may be related to exacerbation of underlying comorbid conditions [6,7–8]. Older age is an independent risk factor for poorer outcome of WNF [6]. Even among young, previously healthy individuals, WNF may be associated with considerable short-term morbidity, with the occurrence of persistent systemic symptoms and absenteeism [9]. A particularly notable persistent symptom following WNV infection, both WNF and neuroinvasive disease, is extreme fatigue. Although fatigue may occur after resolution of other viral illnesses, the frequency and severity with which WNV-associated fatigue is reported is impressive. Nearly 96% of patients with WNF in 1 series described postillness fatigue that lasted a median of 36 days [9]. Self-assessed somatic complaints, such as fatigue, weakness, and concentration problems, may be described as frequently by persons recovering from WNF as by those recovering from neuroinvasive disease [10, 11]. Although the underlying mechanisms resulting in persistent morbidity following WNF remain unclear, mounting data suggest that “mild” WNV infection may be associated with significant short-term and long-term morbidity.

WNV-associated neuroinvasive disease. Although few analyses have assessed the outcomes of patients with WNM specifically, available data suggest that the short-term and long-term outcomes of this manifestation are similar to those of WNF; indeed, there is significant clinical overlap between these syndromes. WNM probably accounts for ∼40% of cases of West Nile neuroinvasive disease (WNND); the median age of patients with WNM is similar to that of patients with WNF overall, although among hospitalized patients, the age of persons with WNM may be considerably less than that of those with WNF or WNE [7]. Patients with WNM frequently require hospitalization for pain control for acute severe headache or rehydration because of prolonged nausea, vomiting, and poor oral intake [5]. Outcomes and disposition at hospital discharge among patients with WNM are generally favorable, and most patients return home to independent living after relatively short hospital stays. However, similar to patients with WNF, persons recovering from WNM frequently have persistent subjective complaints of fatigue, weakness, and memory and concentration problems [12].

WNE is associated with considerable short-term and long-term morbidity and mortality. WNE occurs more frequently in older persons (aged >55 years) and in those with underlying immunosuppression [2]. Patients with WNE are hospitalized longer [13] and experience medical complications (e.g., bronchopneumonia and cardiac arrhythithmiae) at greater rates than do patients with other forms of WNV illness [7, 13]. Acute mortality associated with WNE has been surprisingly consistent among studies, with most in-hospital case-fatality rates estimated to be ∼20% [7, 14, 15]. Acute cause of death in patients with WNE has not been adequately delineated to date, but in many cases, death appears to be a result of acute respiratory failure or sudden cardiac complications [7, 14, 16]. Even after hospitalization, patients with WNE continue to experience long-term mortality at higher rates than the general population, with 1-year postinfection mortality rates comparable to those of patients with severe, chronic noninfectious diseases [17].

Persistent neurologic sequelae can occur in patients with WNE who survive acute illness. In many cases, the movement disorders and extrapyramidal involvement associated with acute WNE are transient, lasting days or weeks [5, 18, 19]. However, these movement disorders may persist for months or years. Six (38%) of 16 subjects observed after the 2002 Louisiana epidemic of WNV infection continued to display tremors and parkinsonism at 8 months after resolution of acute illness [5]. In a study in North Dakota [10], new or persistent tremor was observed or reported in 20% of patients 1 year after resolution of acute illness, and 18% of patients with WNND from a cohort in northern Colorado continued to have persistent tremors, parkinsonism, and ataxia >1.5 years after resolution of acute illness. In many instances, the tremors and parkinsonism were mild and not functionally disabling; however, in some cases, tremors and bradykinesia impacted daily activities, such as eating and dressing.

Patients with WNE frequently require placement in assisted living situations after hospitalization for acute illness. Among patients hospitalized with WNE in Colorado, 75% required some amount of assisted care after hospitalization for acute illness [7]; a similar percentage was reported in a study of patients in Ontario [14]. Physical, occupational, or speech therapy is frequently required following acute WNE, with estimates of the percentage of patients requiring ⩾1 form of rehabilitation ranging from 47% to 65% [12, 13]. However, initial severity of neurological status in patients with WNE is not necessarily indicative of ultimate functional outcome, with several studies suggesting that even patients with severe encephalopathy and poor acute neurologic status may fully recover; such individuals tend to have better pre-illness health status [5]. Although no formal assessments have been performed on the economic impact and long-term costs associated with prolonged institutionalization and rehabilitation following WNND, some patients require months of intensive care and ventilatory support [20], and the economic impact from those with the most-severe illness could be considerable.

WNP. There is increasing data on the long-term outcomes of West Nile flaccid paralysis, or poliomyelitis, suggesting that it is associated with tremendous morbidity and mortality. WNP occurs as a result of viral involvement of the lower motor neurons or anterior horn cells of the spinal cord, resulting in an acute flaccid paralysis more typically associated with poliovirus infection. Although reliable estimates have been difficult to obtain, at least 1 study suggests that WNP accounts for ∼10% of persons hospitalized with WNND [7]; however, this is possibly an underestimate, because WNP is difficult to assess in acutely moribund or comatose patients. Short-term morbidity associated with WNP can be substantial. A wide range of presentations and degrees of limb weakness may occur [16, 20,21–22], ranging from mild monoplegia to flaccid quadriparesis and respiratory failure. Cranial nerve involvement, particularly facial nerve palsy, has also been observed, and although the exact frequency with which this condition occurs in cases of WNP is difficult to ascertain from published case series, it may be as high as 70% of cases [16]. The most severe manifestation of WNP is acute neuromuscular respiratory failure resulting from involvement of diaphragmatic and intercostal respiratory muscle innervation. Although successful extubation and recovery can occur, respiratory involvement is associated with high mortality, with fatality rates of >50%. Patients who are successfully extubated may experience dyspnea and require supplemental oxygen for years following acute illness [16, 20]. Because WNP can occur in the context of any of the other conditions associated with WNV infection, it is likely that some patients with WNE and altered mental status who require emergent intubation have, in fact, developed acute respiratory failure attributable to anterior horn cell involvement. Early dysarthria and dysphagia appear to be predictive of subsequent respiratory failure, and all patients with WNV infection should be closely monitored for these conditions [16].

Most strength recovery by patients with WNP appears to occur during the first 6–8 months following weakness onset, with a subsequent plateauing, during which additional strength recovery may occur (but to a less notable degree) (figure 1) [20, 23]. This pattern is consistent with recovery from poliovirus poliomyelitis [24]. Milder weakness (<2 unit decrements by manual muscle testing) and fewer involved limbs tend to be associated with better recovery. Importantly, however, initial severity of paralysis does not necessarily predict eventual strength outcome, and some patients with profound quadriparesis can experience complete strength recovery [16, 23]. The pathophysiology underlying this phenomenon may be substantial yet reversible damage to some motor neurons that can eventually recover, and this recovery may result in strength improvement. Determination of the number of surviving spinal motor units using electrophysiologic techniques shows promise to be a way of potentially predicting eventual strength recovery [23]. The role of physical and occupational therapies in eventual WNP outcomes remains largely anecdotal. Various degrees of improvement following physical and occupational therapy regimens have been documented [25, 26], but limited numbers of patients and absence of data on control subjects preclude any meaningful comments on the possible role of physical and/or occupational therapy for strength improvement. Clearly, physical and/or occupational therapy is important in patients with persistent profound weakness, to prevent severe contractures, atrophy, and joint problems. It is likely that experience gained from management of poliovirus poliomyelitis will continue to guide physical and occupational therapy regimens for rehabilitation from WNP.

Cognitive and functional outcomes following WNV infection. There is a limited but growing body of literature on the long-term cognitive and functional outcomes following WNV infection (table 2). All studies appear to substantiate a significant amount of self-reported disability following both WNF and WNND. Persistent symptoms, including fatigue, muscle pain, muscle weakness, and headache, were self-reported by >20% of patients with WNF and by nearly 50% of patients with WNE at >90 days postillness among 656 surveyed patients in northern Colorado [13]. The median interval to full recovery following otherwise uncomplicated WNF among 91 patients in Chicago, Illinois, was >60 days, and persistent fatigue, muscle weakness, and headaches were frequently described [9]. Among 35 patients with acute WNV infection during the 1999 outbreak in New York City, only 37% had achieved self-reported recovery by 1 year; persistent physical, cognitive, and functional difficulties were reported by over one-half of patients up to 18 months following acute illness [12]. Self-reported somatic complaints, including fatigue, memory problems, and excessive sleepiness, were frequently described among 49 patients recovering from WNV infection in North Dakota >1 year after acute illness [10]. Surprisingly, most of these studies have found that patients with milder acute illness appear to be just as likely to describe persistent problems, such as fatigue, memory difficulties, and muscle pain, as are patients with WNND, and nonhospitalized patients are just as likely to describe ongoing functional difficulties as are hospitalized patients [10, 12]. One study has suggested that patients recovering from WNND self-describe cognitive complaints at a slightly greater frequency than do patients recovering from WNF [11]. Overall, however, persistent subjective, functional impairment appears to be frequently reported among all WNV syndromes; younger age is more predictive of eventual functional recovery than is initial illness severity [12].

Although self-reported subjective cognitive and functional impairment following WNV infection occurs frequently, to date, measurements of neurocognitive performance using standardized, objective neuropsychological tests have been limited. Because of the prominent involvement of deep cortical brain structures in WNE, it might be expected that subcortical cognitive impairment, involving difficulties in registering new information, in immediate recall, and in attending to tasks, might result. One assessment of neurocognitive performance suggested that patients with WNND performed somewhat more poorly on the Telephone Interview for Cognitive Status—an easily administered, validated objective assessment for cognitive performance—compared with patients with WNF; however, the differences noted were subtle [11]. A study assessing neurocognitive impairment among 49 patients recovering from WNF and WNND in North Dakota used a battery of standardized tests of attention, executive function, learning and memory, and other cognitive domains [10]. The only clearly significant abnormalities found in the results of neurocogitive testing of patients with both WNND and mild disease were in the categories of manual dexterity and motor speed; results of tests of cognitive ability were generally within normal limits and were comparable between hospitalized and nonhospitalized patients. We observed similar findings among a cohort of 54 patients recovering from WNF, WNM, and WNE in Colorado, using computer-based neurocognitive tests. The impairment of motor speed and dexterity was consistent with the observed striatal-thalamic involvement in WNND; the finding of these abnormalities in patients without overt encephalitis raises the question of whether pathologic involvement of these structures may occur in less-severe clinical illness [10].

Collectively, these studies suggest that persistent subjective difficulties with memory, concentration, and cognition occur frequently following WNV infection, but these subjective symptoms do not necessarily correspond with poor performance on objective neurocognitive testing. Results of these studies should be interpreted with caution; however, the neurocognitive scores reflected in these studies were all obtained from patients recovering well enough to return home to independent living. Patients requiring continued institutionalization or intense assistance with daily activities have not been assessed and may have profound cognitive deficits. Additionally, the findings of difficulties with dexterity and subtle cognitive deficits in patients with WNND in some studies suggest that, in some cases, subtle abnormalities may result in significant functional impairment in some patients. Additional studies of the overall neurocognitive impact of WNV infection are needed.

Wnv Infection: Future Perspectives

Over time, as ongoing studies gather additional data and greater numbers of patients are systematically assessed, long-term outcomes of WNV infection will become more apparent. There are several important clinical issues that are particularly relevant to long-term outcomes of WNV infection that should be evaluated over time.

WNV and parkinsonism. Researchers have commented on the frequent involvement of WNV infection in the basal ganglia with resultant parkinsonian features. Although in most patients, parkinsonism appears to be transient and self-limited, persistent parkinsonism has been observed in some patients >1.5 years following acute WNV infection [10]. Although not well-characterized, experience with other closely related flaviviruses has suggested that features of parkinsonism can recur up to 15 years after acute infection [28], and postencephalitic parkinsonism presumed to be attributable to other viral infections may occur years after acute encephalitis [29, 30]. Neuroimaging and histopathologic assessments have demonstrated that WNV and other flaviviruses frequently display distinct neurotropism for the basal ganglia, including the substantia nigra, resulting in neuronophagia and cell death [5, 31, 32]. In addition, animal models have demonstrated a decrease in dopaminergic cells, with resultant parkinsonism, following infection with Japanese encephalitis virus [33]. It is possible that substantia nigra involvement in WNND may predispose to the development of parkinsonism in the context of natural senescent loss of dopaminergic neurons due to ageing [34]; continued studies on the potential role of WNV in parkinsonian syndromes, which include histopathologic studies to assess for the typical neuropathology of postencephalitic parkinsonism, will be important future endeavors.

WNV and postpolio syndrome. Because WNP is clinically and pathophysiologically identical to poliovirus poliomyelitis, the possibility of development of a delayed “postpolio syndrome” following partial or complete recovery from WNP will require future assessment. Postpolio syndrome is characterized by the recurrence—years later—of paresis or paralysis in limbs previously affected by poliovirus infection. The presumed mechanism is thought to be progressive denervation of muscle fibers because of motor unit degeneration among motor neurons previously affected by poliomyelitis. Some estimates have suggested that postpolio syndrome may occur in up to 78% of survivors of poliovirus poliomyelitis [35]. The development of a similar delayed onset or recurrence of limb weakness among survivors of WNP is possible and even likely. Continued long-term assessment of persons recovering from limb weakness due to WNP will be critical in assessing the epidemiology and frequency of this potential syndrome.

Summary

As the future pattern of the epidemic of WNV infection declares itself, it is likely that WNV infection will continue to be a source of considerable morbidity and mortality. As greater numbers of affected individuals survive WNV infection, the possible long-term sequelae will likely become an increasingly important public health issue. The nature of these sequelae will be an important focus of subsequent investigations of the effects of WNV infection in humans.