PubMed1991
Nonherpetic encephalitis outside the newborn period is usually a self-limited disease. The majority of patients will recover without significant sequelae, and require only supportive therapy during the acute illness. Though the underlying viral etiology frequently will escape detection, identification of the infecting agent has considerable prognostic value which can complement clinical measures of severity of disease. The most important initial task of the clinician faced with a case of presumptive viral encephalitis is to eliminate the possibility of a treatable illness. Once this has been done, the diagnosis of viral encephalitis can be supported by documenting the characteristic slow-wave background activity on EEG, and a mild lymphocellular pleocytosis in the CSF. Because viral encephalitis can be caused by such a large number of organisms, the search for an etiology can be daunting. Realizing that all the agents described above can, at times, cause encephalitis without any clue to their identity, one nevertheless may use several pieces of historical information to narrow the possibilities. Travel history, animal exposures, immunization history, and seasonality all may help to steer the search in a particular direction and, indeed, may point to a nonvirologic cause as well. In addition, detection of extraneurological signs and symptoms may strongly indicate a specific virologic diagnosis. Finally, knowledge of concurrent community epidemic patterns, and of surveillance data routinely collected by local and state health departments, can help to increase or decrease the likelihood of a given pathogen. The causative viral agent usually can be identified by serological testing and viral culture. Occasionally, single serological determinations are diagnostic: in rabies (when the patient has not received immune prophylaxis), eastern equine encephalitis, and HIV, since seropositivity is strongly associated with symptomatic illness; and in Epstein-Barr virus, if a panel of antibody determinations which can time the infection is available. In addition, high CSF: serum titers for antibody against any neurotropic agent is usually diagnostic, though the absence of a high central nervous system antibody titer does not eliminate any potential viral pathogen. With these few exceptions, a single serological determination for a given pathogen is almost always impossible to interpret; paired sera (one obtained upon diagnosis, and one obtained 10 to 14 days later, either just prior to hospital discharge or at a follow-up visit) are far more helpful. Many viruses that directly infect the central nervous system are difficult to recover from the CSF; therefore, viral isolation from the nasopharynx and stool also should be sought.(ABSTRACT TRUNCATED AT 400 WORDS)