Neurosyphilis: diagnosis and response to treatment.

Received 11 June 2008; accepted 19 June 2008; electronically published 20 August 2008. Reprints or correspondence: Dr. Daniel M. Musher, Infectious Disease Section, Rm. 4B-370, Michael E. DeBakey Veterans Affairs Medical Center, 2002 Holcombe Blvd., Houston, TX 77030 ([email protected]). Clinical Infectious Diseases 2008; 47:900–2 This article is in the public domain, and no copyright is claimed. 1058-4838/2008/4707-0006 DOI: 10.1086/591535 In this issue of Clinical Infectious Diseases, an important article examines the serological response to treatment of neurosyphilis. Marra et al. [1] show that normalization of the results of the serum antibody test for cardiolipin (rapid plasma reagin [RPR]) is a strong indicator of success after treatment of neurosyphilis. Most of their patients had HIV infection, but their findings are likely to apply to HIVuninfected persons, as well. This article [1] provides great practical help to physicians who treat patients for sexually transmitted diseases. Performing a lumbar puncture in a clinical setting is logistically difficult, and furthermore, patients often refuse it. Thus, it is reassuring to learn that normalization of the serum RPR titer is highly predictive of a good response to therapy, even though this finding is somewhat less likely to apply to untreated patients with AIDS. A full understanding of the article by Marra et al. [1] requires further discussion of 2 important issues—one relating to the diagnosis of neurosyphilis and the other to treatment. In the prepenicillin era [2, 3], neurosyphilis was diagnosed clinically; the diagnosis was supported by serologic test results positive for syphilis and detection in CSF of antibody to cardiolipin (initially by the Wassermann reaction, then by the Hahn and other more sensitive modifications, and ultimately by the venereal disease research laboratory [VDRL] test). If the CSF VDRL test result was negative, increased WBC count or protein concentration in CSF provided laboratory support; except in forms of neurosyphilis that are now rare, this was decidedly uncommon [2]. Asymptomatic neurosyphilis was diagnosed on the basis of CSF VDRL test results, although in an occasional case, other CSF abnormalities, in addition to a high serum RPR titer, might have been regarded as diagnostic. Simpy stated, a diagnosis of neurosyphilis or the exclusion of this diagnosis depended largely on the CSF VDRL test result. In 1972, Hooshmand et al. [4] reported a series of cases in which they diagnosed neurosyphilis on the basis of (1) suggestive neurologic findings, in addition to a positive serum fluorescent treponemal antibody-absorption test result (this highly sensitive test, now replaced by the equivalently sensitive microhemagglutination Treponema pallidum [MHA-TP] test, detects antibody to outer cell wall proteins of T. pallidum, and once the result is positive, it remains so for life) or (2) a positive CSF fluorescent treponemal antibody-absorption test result in addition to other CSF abnormalities or neurologic abnormalities for which other causes had been excluded. Only 57% of the patients in the study by Hooshmand et al. [4] had a positive CSF VDRL test result. This article [4] is often cited to support the notion that a reactive CSF VDRL test is not a regular feature of neurosyphilis. However, MHA-TP testing of CSF samples is not accepted as a diagnostic tool, because it is overly sensitive; passive diffusion of plasma proteins with positive serum MHA-TP yields a positive result even when neurosyphilis is not present [5]. Some European authorities [6] use the CSF MHA-TP assay, but they report the result after calculating the ratios of CSF to serum protein concentration and CSF to serum MHA-TP titer to determine whether its detection reflects passive diffusion from plasma or local synthesis of antibody in the CNS. Hooshmand et al. [4] stated that 100% of their patients had positive CSF fluorescent treponemal antibody-absorption test results, as if to assure the reader of the correctness of their diagnoses. In fact, this has never been a valid basis for diagnosis of neurosyphilis, and it continues to astonish me that this article [4] was ever published in that form. If the authors overdiagnosed neurosyphilis, which I believe they most certainly did, the true percentage of patients with negative CSF VDRL results should be much lower. There are other reasons to be suspicious