[Retrospective study of viral causes of central nervous system infections in Tunisia (2003-2009)]. / Étude rétrospective des étiologies virales des infections neuroméningées en Tunisie (2003-2009)

OBJECTIVES: To determine the role of enteroviruses, Herpesviridae, West Nile virus and Sandfly Toscana virus in central nervous system (CNS) infections in Tunisia. METHODOLOGY: 847 cerebrospinal fluid (CSF) samples, 427 serum samples and 23 stool samples were collected from 1071 patients hospitalized for CNS viral infections from January 2003 through December 2009. All CSF samples were first tested by PCR to detect enteroviruses and Herpesviridae. In specific epidemic contexts in patients negative for these viruses, arbovirus infection was tested by ELISA. RESULTS: Virological testing was positive in 17.5% of cases. West Nile virus and enteroviruses accounted for 58% of them, enteroviruses 23.5%, Herpesviridae 8.5%, and Toscana virus 10%. West Nile virus infection was observed only in 2003, during an outbreak in coastal regions. Toscana virus circulated regularly throughout the study period. Enteroviruses were responsible for grouped cases of aseptic meningitis in both 2003 and 2005. Arboviruses and enteroviruses were detected mainly in summer and autumn. Herpesviridae were associated with sporadic cases of meningoencephalitis. CONCLUSION: This report on viral causes of CNS infections in Tunisia shows that West Nile virus and enteroviruses appear to circulate mainly during epidemics, while the circulation of Toscana virus seems continuous. Negative virus findings may be due, at least in part, to late sampling, inappropriate sample collection and transportation to the virology lab, or failure to test for the right virus. It is essential to promote collaboration between clinicians and biologists to maximize the likelihood of diagnosis.

Enterovirus detection in stool specimen: relevance for poliovirus and enterovirus surveillance.

Detection of enterovirus genome by PCR in clinical samples is now extensively used for the diagnostic of enterovirus infections given its rapidity and high sensitivity. In contrast, its use in surveillance programs targeting specific enterovirus serotypes remains less frequent. The most sensitive protocols are those amplifying in the 5'untranslated region (5'UTR). However the possibility to use sequence analysis of the 5'UTR amplicons for serotype identification is not yet well established. In this report, stool samples from polio suspected cases and their healthy contacts were tested. The results of direct detection of enterovirus genome by PCR and serotype identification based on sequence analysis of the PCR products in the 5'UTR were compared to those of standard cell-culture-based protocols. Standard protocols detected enterovirus isolates in 7.4% of cases while 9.8% of samples were positive by PCR. Serotype identification based on sequence analysis of amplicons showed concordant results with serotypes determined on virus isolates by seroneutralisation or sequencing in the VP1 gene in 39% of cases only. These results confirm that the use of PCR amplification from stool samples improves the sensitivity of enterovirus detection but do not recommend the use of sequence analysis of the 5'UTR PCR product to determine enterovirus serotype.