Clinical prediction and diagnosis of neurosyphilis in HIV-negative patients: a case-control study

Background Early diagnosis and treatment of neurosyphilis is of great significance for regression. There is no gold standard for the diagnosis of neurosyphilis. We did this study to explore the factors associated with the clinical diagnosis of neurosyphilis and assess their accuracy for the diagnosis of neurosyphilis. Methods We retrospectively reviewed 100 cases of syphilis patients who underwent lumbar puncture at a major dermatology hospital in Guangzhou, China between April 2013 and November 2016. Fifty patients who were clinically diagnosed with neurosyphilis were selected as case group. Control group consisted of 50 general syphilis patients who were matched with age and gender. The records of patients were reviewed to collect data of socio-demographic information, clinical symptom, and laboratory indicators. Multivariable logistic regression was used to explore diagnostic indictors, and ROC analysis was used to assess diagnostic accuracy. Results Neurological symptoms (odds ratio (OR) = 59.281, 95% CI:5.215-662.910, P = 0.001), cerebrospinal fluid (CSF) Treponema pallidum particle agglutination (TPPA) titer (OR = 1.004, 95% CI:1.002-1.006, P < 0.001), CSF protein (OR = 1.005, 95% CI:1.000-1.009, P = 0.041), and CSF white blood cell (WBC) (OR = 1.120, 95% CI:1.017-1.233, P = 0.021) were found to be statistically associated with neurosyphilis. In ROC analysis, CSF TPPA titer had a sensitivity of 90%, a specificity of 84%, and an area under curve (AUC) of 0.941. Conclusion CSF TPPA can potentially be considered as an alternative test for diagnosis of neurosyphilis. Combining with neurological symptoms, CSF protein, CSF WBC, the diagnosis would have a higher sensitivity.