Prolactin levels in the cerebrospinal fluid of patients with HIV infection and AIDS.

The objective of this study was to assess the levels of prolactin (PRL) in cerebrospinal fluid (CSF) of HIV-infected patients with regard to nonHIV-infected patients, and to assess the levels of prolactin in the CSF of HIV-infected patients with and without neurological HIV-involvement. Seventeen HIV-infected patients with different degrees of immunological and neurological involvement were studied. A second group of six HIV-seronegative patients with varying clinical conditions requiring lumbar punctures were included as controls. CSF was collected from patients and controls. Patients were studied neurologically and neuropsychologically, and computed tomography of the brain were performed. They were staged according to CDC clinical classification for HIV infection, and on the basis of tomographic findings into one of five stages. An additional classification for neurological involvement in AIDS was used. CD4+ cell counts, CSF studies, serum-prolactin levels and CSF-prolactin levels were performed as principal laboratory tests. CSF PRL concentrations were significantly higher in the HIV-infected group (n = 17) than the nonHIV infected control group (n = 6) (mean +/- s.d.; 5.77 +/- 2.22 vs. 3.53 +/- 0.69 x 10(-6) g l-1, respectively; p = 0.009, Mann-Whitney U-test). Moreover, even CSF-PRL concentration was higher in HIV-infected patients without cognitive impairment (stage 0 of the clinical classification), (n = 12) in comparison with nonHIV infected controls (n = 6) (mean +/- s.d.; 5.51 +/- 2.31 vs. 3.53 +/- 0.69 x 10(-1) g l-1, respectively; p = 0.028, Mann-Whitney U-test). There was a good correlation between serum and CSF-PRL levels in HIV-infected patients when measured by the Spearman Rank Test (rs = 0.773; p = 0.005). PRL raised serum levels were found in 4 out of 13 patients (30.73%). We conclude that higher levels of CSF-PRL are more frequently found in HIV-infected patients in comparison to uninfected controls. High levels of circulating PRL were also found in HIV-infected patients corroborating results from other work. A good correlation coefficient was found between circulating and CSF-PRL levels in HIV-infected patients, suggesting that disruption of the blood-brain barrier might account for a possible pathogenic mechanism.