Syphilis and neurosyphilis: HIV-coinfection and value of diagnostic parameters in cerebrospinal fluid

Background: Neurosyphilis might be difficult to diagnose particularly in asymptomatic patients and patients with HIV-coinfection. The objective of this study was to evaluate current diagnostic standards for neurosyphilis in HIV-positive and -negative patients. Methods: We studied retrospectively patients with an active syphilis infection who had additionally undergone lumbar puncture. Patients where the criteria for the diagnosis of a definite or probable neurosyphilis were applicable were further analyzed for clinical symptoms, CSF, HIV-status as well as Treponema pallidum testing in serum and CSF. Correlation analysis of categorical variables was done by using the Chi-square test or in cases of small sample sizes the exact test of Fisher. p values <= 0.05 were considered significant. Results: Eighty-nine patients were diagnosed with syphilis. All necessary criteria for the diagnosis of a neurosyphilis were available in 67 of them including 35 HIV-positive and 32 HIV-negative patients. A definite neurosyphilis could be retrospectively diagnosed in 13 and a probable in another 25 cases. Normal CSF results were more likely in HIV-negatives (p = 0.016). A neurosyphilis was correlated to a CSF pleocytosis > 5 cells/mu l and to an albumin quotient > 7.8 mg/dl regardless of a parallel HIV infection. HIV-positives had more frequently a CSF-RPR titre > 1: 4 than HIV-negatives (p = 0.031). However, the RPR test in CSF in definite or probable neurosyphilis had a sensitivity of only 21 %. Discussion: Our data show that a pleocytosis and an elevated albumin quotient correlate with neurosyphilis. However, the CSF-RPR test as gold standard in neurosyphilis diagnostics has a very low sensitivity.