Background:  Central nervous system infection is a well-known complication of HIV disease and occurs early in the course of systemic infection. However, the critical phase of central nervous system infection for an individual patient is unknown. Thus, to further clarify the value of cerebrospinal fluid (CSF) analysis in HIV/AIDS patients, 91 HIV-1⁺ individuals in different disease stages without overt neurological deficits have been recruited consecutively.

Methods:  Of 91 early-stage patients, 15 were without treatment and 21 were on HAART; of 91 late-stage individuals, 8 were without treatment and 40 were on HAART; of 91 intravenous drug users, 3 were without treatment and 4 were on HAART. All patients underwent venous and lumbar puncture as well as neuropsychological testing (AIDS dementia scale, grooved peg-board test, and a motor test battery). CSF was analyzed for routine parameters, viral load, and immunological values (MCP-1 and GAL3).

Results:  In early-stage patients without HAART, CSF viral load correlated with duration of HIV⁺, an elevated CSF lymphocytic cell count, immunoglobulin G (IgG) index, as well as with motor and neuropsychological abnormalities, whereas in treated early-stage patients, there was only a correlation with motor abnormalities. In late-stage patients without HAART, CSF viral load correlated with the same parameters as in untreated early-stage patients, but additionally with CSF protein content, CSF lactate, plasma viral load, MCP-1, GAL3, and negatively with CD4⁺ cell count, whereas in late-stage patients on HAART, there was only a weak correlation of CSF viral load with CSF cell count, MCP-1, GAL3, and plasma viral load. Of 91 individuals, 20 had a higher CSF than plasma viral load; CSF viral load correlated strongly with all routine parameters and plasma viral load, and weakly with motor and neuropsychological deficits and negatively with duration of HIV-1-positivity and the CD4⁺ cell count.

Conclusions:  The data provide 2 main messages. First, HIV-1/AIDS patients seem to have an individual “set-point” for CSF disease, which provokes an inflammatory process only incompletely influenced by ART; in treated late-stage patients there is no correlation of central nervous system deficits with CSF viral load. Second, a subgroup of patients exhibits a “primary” central nervous system disease manifestation. Factors contributing to the higher CSF than plasma viral load in some individuals must be clarified in further studies.