A-266 Evaluation of an Antibody Index Assay for Assessment of Intrathecal Anti-Treponema pallidum IgG Synthesis for the Diagnosis of Neurosyphilis

Abstract

Abstract Background Diagnosis of neurosyphilis (NS) is challenging due to nonspecific clinical symptoms and the lack of a sensitive and specific laboratory test. The non-treponemal Venereal Disease Research Laboratory (VDRL-CSF) assay is a commonly used test in diagnosing NS. An important limitation of the VDRL-CSF is limited specificity, in part due to the inability to differentiate between intrathecally produced antibodies vs passive antibody diffusion into the CSF due to blood-brain barrier (BBB) disruption. Here, we compared the intrathecal synthesis antibody index (AI) of anti-Treponema pallidum IgG antibodies to results from the VDRL-CSF assay in routinely submitted clinical samples for evaluation of suspected NS. Methods Paired serum and CSF residual clinical samples from 98 patients with previous VDRL-CSF results were collected. For each pair, the serum and CSF were drawn within 24 h. No additional clinical information and limited supporting clinical test results were available. Thirty-one positive and 67 negative paired samples were tested by an anti-T. pallidum IgG ELISA (Euroimmun, Lübeck, Germany). The AI was calculated using Reiber’s method, with intrathecal total IgG synthesis corrected using the Reiber recommended albumin quotient calculation. Results were compared for positive, negative and overall agreement between methods. Results The anti-T. pallidum AI showed a positive, negative and overall agreement of 71.0% (22/31), 100% (67/67) and 90.8% (89/98), respectively, compared to VDRL-CSF. Of note, all nine discordant samples showed abnormally elevated albumin levels in CSF (range: 27.4 to 215 mg/dl; median = 41.7 mg/dl; normal value =<27 mg/dl), indicating a compromised BBB which likely resulted in false-positive VDRL-CSF results. The range and median value for CSF albumin in the 22 concordant samples was 12.4 to 168 mg/dl and 37.2 mg/dl, respectively. In total, 83.9% (26/31) of VDRL-CSF positive samples showed elevated CSF albumin levels which suggests a high percentage of BBB impairment in this patient population. This demonstrates the advantage in specificity that is offered by AI testing compared to use of the VDRL-CSF assay, given the ability of AI to differentiate intrathecally produced vs passively diffused syphilis-specific antibodies. No potential increase in sensitivity was observed in AI testing as 100% negative agreement was obtained. Conclusion Our data suggests that establishment of an AI using a treponemal assay to detect intrathecal antibody synthesis provides improved specificity compared to the current commonly performed VDRL-CSF antibody test as an aid in the diagnosis of NS. Further evaluation of sensitivity is warranted using samples from patients with clinically diagnosed NS.