Evaluation of treponemal serum tests performed on cerebrospinal fluid for diagnosis of neurosyphilis.

Abstract

We evaluated the use of treponemal serum tests in cerebrospinal fluid (CSF) to diagnose neurosyphilis since CSF-Venereal Disease Research Laboratory (VDRL) is specific but lacks sensitivity. We tested CSF specimens using the following treponemal serum tests: INNO-LIA, Treponema pallidum particle agglutination (TP-PA), Trep-Sure, and Maxi-Syph. The reference standard to calculate sensitivity and specificity was having two or more reactive/positive tests on CSF. The reference standard group included 11 cases that fulfilled the definition of neurosyphilis (reactive CSF-VDRL plus symptoms) and three cases that did not fulfill the definition: two cases had neurologic symptoms but a nonreactive CSF-VDRL, and one had several positive CSF syphilis tests (reactive VDRL and positive treponemal and syphilis polymerase chain reaction) but no history (referred sample). Controls included 18 patients in whom a CSF-VDRL was performed the same week as patients in the reference group. The sensitivity was 85.7% (12/14) for CSF-VDRL, 92.9% (13/14) for Trep-Sure, 100% (10/10) for Maxi-Syph, 92.3% (12/13) for INNO-LIA, and 83.3% (10/12) for TP-PA. Specificity was 100% for all tests. Treponemal serum tests performed on CSF were useful in identifying two patients with nonreactive CSF-VDRL.