Severe West Nile Virus Disease in Healthy Adults

In 2001, Weiss et al. [1] summarized the clinical findings for 19 cases of West Nile virus (WNV) infection in New York and New Jersey that occurred during the summer of 2000. They reported that patients >50 years of age were more likely to present with meningoencephalitis and had an increased mortality rate compared with patients <50 years of age. Clinical WNV infection usually manifests as West Nile fever, aseptic meningitis, or encephalitis. More unusual manifestations include poliomyelitis-like syndrome [2, 3, 4] optic neuritis [5], and extrapyramidal movement disorders [6]. Although some reports included cases of WNV infection with clinical Guillain-Barré syndrome [7], it now appears that flaccid paralysis in WNV infection is caused predominately by involvement of the anterior horn cells [8].

In 2002, the United States experienced a dramatic spread of the WNV. A total of 4156 patients were documented to be infected with the virus, and 284 patients died [9]. Michigan documented the second highest number of cases in the United States, with 614 confirmed cases of WNV infection and 51 deaths. Kent County, Michigan, experienced its first human cases of WNV infection in 2002; 45 patients were hospitalized with WNV infection. We initially observed severe manifestations in adults <50 years of age, a group that previously had been thought to be at risk only for uncomplicated WNV meningitis [10]. Focal neurologic disease appeared to be more prevalent than in previous studies [11], with cases of monoparesis, paraparesis, and quadriparesis.

We reviewed the medical records of patients hospitalized in Kent County with WNV infection during 2002 to better characterize the epidemiology, clinical characteristics, and short-term outcomes for WNV infection. Because persons >50 years old had been shown previously to be at increased risk of complicated WNV infection [12], we divided our case patients into 3 age groups: ⩽50 years, 51–64 years, and ⩾65 years of age.

Methods. This study was conducted in Kent County, Michigan, which has a population of 575,334 and is located in the western portion of Lower Michigan. Kent County includes Grand Rapids, a city with a population of 197,800 [13]. There are 4 acute-care hospitals serving Kent County, all located in Grand Rapids.

Laboratories in the 4 acute-care hospitals submitted all clinical specimens (serum and CSF) to a single laboratory, the Michigan Department of Community Health (MDCH) Laboratory, for detection of WNV. Specimens were tested for WNV IgM antibody using an antibody capture ELISA (MAC-ELISA). Positive results were confirmed in duplicate. Positive test results were defined as any of the following: (1) detection of CSF WNV IgM antibody; (2) a 4-fold increase in WNV IgM titer from acute to convalescent serum; or (3) a single serum sample positive for WNV IgM with confirmatory plaque reduction neutralization test result [14].

Case patients were identified on the basis of positive WNV test results from the MDCH laboratory. Additional patients were identified by review of medical records from the sole Infectious Diseases practice in Kent County and by active reporting to the Kent County Health Department. Patients were included if they were admitted to the hospital in 2002 for 24 h of observation or with inpatient status and had a test result positive for WNV. Patients who were tested in the outpatient setting or who were tested in the emergency department and never admitted were excluded from our analysis.

The Institutional Review Board at each participating institution approved the study protocol. The protocol adhered to Health Insurance Portability and Accountability Act guidelines for protection of patient privacy [15].

Case patients were classified, in accordance with CDC criteria [16], as having aseptic meningitis or encephalitis. The categories of aseptic meningitis and encephalitis were mutually exclusive, so that no patient was counted twice. Aseptic meningitis was defined as fever, headache, stiff neck, and CSF pleocytosis without mental status change. Encephalitis was defined as altered mental status (ranging from confusion to coma with or without additional signs of brain dysfunction) with fever and headache. Case patients with meningoencephalitis were classified in the encephalitis group. The strength of patients with either meningitis or encephalitis was further classified, on the basis of documented examination findings, as follows: normal, weakness of 1 limb (monoparesis), or weakness of >1 limb (paraparesis or quadriparesis).

The inpatient medical records of hospitalized WNV case patients were reviewed. Information was collected regarding demographic characteristics, clinical signs and symptoms at presentation, date of onset of symptoms, laboratory and radiologic study results, neurodiagnostic findings, and clinical status at discharge from the hospital.

Student's t test was used to compare laboratory parameters for selected clinical groups. A P value of <.05 was considered to be statistically significant.

Results. During 2002, there were 45 patients hospitalized in Kent County with WNV infection. Forty-four charts were reviewed; 1 chart was not available for review. Thirty patients (68%) were female; 14 (32%) were male. Seventeen (39%) patients were aged 0–50 years; 6 patients (14%) were aged 51–64 years, and 21 (47%) were aged ⩾65 years.

The reported duration of symptoms prior to admission was 1–21 days (median, 4 days). The peak incidence of onset of symptoms was between 16 August and 15 September 2002 (70% of cases). The earliest date of symptom onset was 1 August, and the latest date was 15 October (figure 1).

Symptoms at presentation are listed in table 1. The most common physical findings at admission were temperature of >38.0 °C (70% of patients), altered mental status (36%), motor weakness (23%), abnormal reflexes (18%), cerebellar abnormality (9%), and cranial neuropathy (2%).

CSF samples were obtained from 42 (95%) of the 44 patients with WNV infection. Results of CSF analysis are documented in table 2. It is of interest that 34% of the CSF samples revealed >49% neutrophils on a differential count. The mean percentage of neutrophils in CSF was 23.4%. There was no correlation between the duration of symptoms prior to lumbar puncture and the neutrophils percentage revealed by the CSF analysis (correlation coefficient, 0.09).

Case patients with aseptic meningitis had a higher level of CSF pleocytosis than did case patients with encephalitis (WBC count, 289 vs. 182 cells/mm³), but this difference did not reach statistical significance. There was no significant difference in the ratio of CSF protein level to WBC count for patients with meningitis versus patients with encephalitis, for patient with focal weakness versus patients with encephalitis without weakness, or for patients with fatal cases of infection versus patients with nonfatal cases.

Thirteen patients (30%) received a clinical diagnosis of aseptic meningitis; 29 patients (66%) received a diagnosis of encephalitis (table 3). Two patients could not be classified into either category. Although the encephalitis group was older than the aseptic meningitis group (mean age, 55 vs. 47 years), this difference was not statistically significant (P > .2).

Ten patients (23%) had focal weakness. All of those with neuromuscular weakness were in the encephalitis group. Thus, 34% of the patients with encephalitis had focal weakness. Six of these 10 patients had paraparesis or quadriparesis. Five of the 6 patients (83%) with paraparesis or quadriparesis were aged ⩾65 years. Four of the 10 had monoparesis (table 3), and all 4 patients with monoparesis were in the 2 younger age groups (i.e., they were 0–64 years of age). Subsequent electrodiagnostic testing, which was performed after hospital discharge for survivors with focal weakness, demonstrated involvement of the anterior horn cells consistent with a poliomyelitis-like syndrome (Christian VandenBerg, personal communication).

The average length of the acute-care hospital stay was 9.8 days, with a median of 6 days (range, 2–46 days). All patients were accounted for at the time of discharge. Twenty-eight patients (68%) were able to return home at discharge. Ninety percent of patients aged 0–50 years were discharged to home, compared with 35% of those aged ⩾65 years. Five patients were discharged to acute rehabilitation facilities. Five patients were transferred to extended-care facilities. One patient left the hospital against medical advice and did not require readmission for complications of the WNV infection.

The other 6 patients (14% of the total) died of WNV infection. Three case patients with paraparesis or quadriparesis died of respiratory failure, suggesting involvement of the respiratory muscles, similar to the respiratory involvement associated with poliomyelitis [17]. All fatalities occurred in the group of patients aged ⩾65 years. Thus, the mortality rate in the group of patients aged ⩾65 years was 35%. The mortality rate among patients with encephalitis aged ⩾65 years was 46%.

Two of the 6 patients that died of WNV infection were known to be immunocompromised. Two additional immunocompromised patients survived their infection with WNV.

Autopsies were performed on 2 of the 6 patients with fatal cases. Histopathologic examination of brain specimens from both patients [18] demonstrated microglial nodules in the deep gray-matter structures, the brain stem, and the cerebellum. For both of these patients, histopathologic examination of the spinal cord showed bilateral myelitis confined to the anterior horns.

Discussion. A greater proportion of WNV encephalitis cases occurred in adults aged ⩽50 years than has been described previously [19]. Sixteen of 29 case patients (55%) with encephalitis were aged <65 years, and 12 of these (41% of encephalitis case patients) were aged ⩽50 years.

Adults aged ⩽65 years with focal weakness tended to have monoparesis, suggesting limited spread of viral infection within the cord. Adults aged ⩾65 years were more likely to develop paraparesis or quadriparesis, presumably because of more diffuse involvement of the anterior horn cells. Sejvar et al. [20] have reported that there is little to no recovery of strength in those with acute flaccid paralysis 8 months after initial infection. Thus, although 7 of the 10 patients with focal weakness survived, the chance for recovery of strength amongst these survivors is low.

Paraparesis and quadriparesis were associated with an increased mortality rate among the group of patients with encephalitis. However, once corrected for age of ⩾65 years, there was no difference in mortality rates for patients who had encephalitis with and without focal weakness

The mortality rate in our case series (14%) is higher than the national average of 6.7% for WNV in 2002 [9]. Our study reviewed only patients with WNV infection who required hospitalization. The mortality rate among our patients was similar to that reported previously for hospitalized patients with WNV infection [21]. Twenty-nine of our patients (66%) had encephalitis, which has been associated with a higher mortality rate than has WNV meningitis [22].

Adults aged ⩽50 years represented 43% of our hospitalized case patients. Adults aged <65 years were only slightly less likely to manifest WNV encephalitis than were those aged ⩾65 years (64% and 76%, respectively, of each age group had encephalitis). Of patients with focal neuromuscular weakness, adults aged <65 years tended to have monoparesis, while those aged ⩾65 years tended to have paraparesis or quadriparesis. Controlled trials of treatment for WNV infection are greatly needed, both to decrease the mortality rate for untreated WNV infection and to decrease the chance of long-term neuromuscular weakness in survivors.

• Received June 10, 2003.
• Accepted August 25, 2003.

• © 2004 by the Infectious Diseases Society of America