Objectives: The objectives of this study were to determine predictive factors for neurosyphilis in HIV-infected patients with syphilis and optimize the use of lumbar puncture.
Study design: The authors reviewed 112 cases of HIV-infected patients with syphilis who underwent a lumbar puncture. Diagnosis of neurosyphilis was based on a cerebrospinal fluid white blood cells count > or =20/microL, and/or a reactive cerebrospinal fluid-Venereal Disease Research Laboratory, and/or a positive intrathecal T. pallidum antibody (ITPA) index.
Results: Twenty-six of 112 had neurosyphilis. Neurologic manifestations and serum rapid plasma reagin (RPR) were associated with neurosyphilis (P = 0.036, P = 0.018, respectively). In multivariate analysis, log(2)RPR was still associated with neurosyphilis (P = 0.005). In patients without neurologic manifestations, the risk of neurosyphilis increases gradually with log(2)RPR. A serum RPR of 1/32 seems to be the best cutoff point to decide the performance or not of a lumbar puncture (sensitivity 100%, specificity 40%).
Conclusion: In HIV-infected patients with syphilis, lumbar puncture could be restricted to those with neurologic manifestations or a serum RPR > or =1/32.