CSF findings in 250 patients with serologically confirmed West Nile virus meningitis and encephalitis

Objective: To provide a large, comprehensive evaluation of the CSF findings in patients with serologically confirmed West Nile virus (WNV), CNS disease, and their correlation with outcome. Methods: CSF samples from 334 WNV-infected hospitalized patients were analyzed. Information was available and extracted from the medical records of 250 of these patients, and CSF parameters correlated with clinical and epidemiologic features of disease ( e. g., patient age, sex, outcome). Results: Patients with meningitis had a mean of 226 cells/mm(3), and those with encephalitis had a mean of 227 cells/mm(3). Three percent of meningitis patients and 5% of encephalitis patients had fewer than 5 cells/mm(3), and approximately 8% of both groups had more than 500 cells/mm(3). Patients with meningitis had a mean of 41% neutrophils, and those with encephalitis had 45%. Forty-five percent of meningitis patients and 37% of encephalitis patients had at least 50% neutrophils in their initial CSF specimen. Neither the mean percent neutrophils nor their distribution differed significantly between groups. Forty-seven percent of encephalitis patients and 16% of meningitis patients had CSF protein of 100 mg/dL or greater (p < 0.01). Although specific CSF parameters, including nucleated cell count and protein concentration, correlated significantly with outcome, multivariate analysis suggested that their total predictive value was modest. Age was an additional predictor of outcome independent of CSF variables in all patients. Conclusions: Serologically confirmed West Nile virus meningitis and encephalitis produce similar degrees of CSF pleocytosis and are frequently associated with substantial CSF neutrophilia. Patients with encephalitis have higher CSF protein concentrations and are more likely to have adverse outcomes, including admission to long-term care facilities or even death after their acute illness. CSF findings were only a modest predictor of disease outcome, with patient age adding important independent prognostic information.