Syphilis is discussed in Chapter 15. CNS syphilis can be diagnosed clinically but is much more accurately diagnosed through tests on CSF. In two studies, cell counts were normal (<5 cells) in 19%-62%, between 5 and 10 cells in 24%-69%, and more than 10 cells in 12%-14%. The majority of the cells were mononuclear in 60%-80% of the cases and polymorphonuclear in 20%-40% of the cases. CSF protein was normal in 14%-61% of patients, between 45 and 100 mg/dl in 34%-61%, and more than 100 mg/dl in 5%-25%. The Veneral Disease Research Laboratory (VDRL) test on CSF specimens was reactive in about 55% of the patients (literature range, 10%-70%). A serum VDRL test was reactive in 49%-86% of the patients. It was noted that before penicillin was discovered, CSF studies recorded much higher incidence and degree of abnormalities.

CNS syphilis usually requires serologic tests for diagnosis. The standard serologic tests for syphilis such as the VDRL usually, but not always, give positive results on peripheral blood specimens when they are positive on CSF. A lack of relationship is most often found in the tertiary stage, when the peripheral blood VDRL test result may revert to normal. Conversely, the CSF response is very often negative when the peripheral blood VDRL result is positive. CNS syphilis usually is a tertiary form with symptoms appearing only after years of infection, and in many patients with syphilis the CNS is not clinically involved at all. Despite lack of clinical CNS symptoms, actual CNS involvement apparently is fairly common (in at least 30%-40% of syphilis cases), beginning in the primary and secondary stage; although specific syphilitic syndromes, if they develop, are not seen until the tertiary stage years later. There is some evidence that concurrent infection by the HIV-1 virus may increase the risk of active CNS syphilis. The best criteria of CNS disease activity are elevated CSF cell count and protein levels (however, CSF WBCs are reported to be elevated in 38%-81% of cases and CSF protein elevated in 39%-86%). A reactive CSF VDRL result indicates disease that has been present for a certain length of time, without necessarily being currently active. The CSF VDRL usually is normal in patients with biologic false positive (BFP) serologic test for syphilis (STS) reactions. After adequate treatment CSF pleocytosis usually disappears within 3 months (range, 1.5-6 months). CSF protein elevation, however, may persist as long as several years.

The three most important forms of CNS clues are general paresis, tabes dorsalis, and vascular neurosyphilis. In general paresis, the CSF serology almost always is normal in untreated patients. In tabes dorsalis, the CSF serology is said to be abnormal in most early untreated patients, but may be normal in up to 50% of late or “burnt-out” cases. In vascular neurosyphilis, approximately 50% of results are abnormal.

The FTA-ABS is nearly always reactive in peripheral blood when CSF shows some laboratory abnormality suggestive of neurosyphilis. The FTA-ABS has been shown to be more frequently reactive than the VDRL when testing the CSF of patients with syphilis. Several studies reported 100% of patients with the diagnosis of CNS syphilis had a reactive CSF FTA-ABS response. However, a substantial number of patients who were asymptomatic and had normal CSF cell counts and protein also had a reactive CSF FTA-ABS response. Therefore, the U.S. Centers for Disease Control (CDC) and an important segment of other investigators currently believe that a reactive CSF FTA-ABS response does not necessarily represent active CNS syphilis and that the clinical importance of a reactive FTA-ABS test on spinal fluid is uncertain when the cell count, protein, and spinal fluid VDRL results are normal. There is also the problem of spinal fluid contamination by blood, which could produce false positive results if the contaminating blood contained serum antibodies.

Based on current CDC recommendations, the FTA-ABS test is usually not done on CSF, since there is no problem of BFP reactions in the spinal fluid VDRL (unless the CSF is contaminated with blood), and the peripheral blood FTA-ABS test is reactive in almost all cases of CNS syphilis. Unfortunately, there is currently no gold standard to determine which patients actually have CNS syphilis and therefore how accurate the various CNS laboratory tests really are.

Up to 75% of patients with active CNS syphilis may demonstrate oligoclonal bands on CSF electrophoresis (similar to those seen in multiple sclerosis).