A prospective radiologic and neurologic follow-up study of 61 HIV-1 -infected subjects: early beginning and slow progression of brain atrophy

Elevated levels of soluble Fas and Fas ligand in cerebrospinal fluid of patients with AIDS dementia complex

We measured the levels of sFas and sFasL in CSF and serum of HIV-1 infected patients and related them to AIDS dementia complex (ADC). Specimens were obtained from 51 HIV-1 infected individuals (29 with ADC) and 39 HIV negative individuals. The sFas was detectable in all sera and 98% of CSF specimens. Measurable levels of sFasL were found in 79% of the CSF and 98% of sera samples. According to the presence or absence of ADC, we observed significant differences in CSF sFas (median and IQR 116, 132 vs. 30, 23 pg/ml, P<0.001) and sFasL (median and IQR 127, 290 vs. 15, 73 pg/ml, P<0.001) levels. The sFas in serum differed significantly between HIV-1 infected subjects and non-infected controls (P<0.001), with no correlation to ADC. On the contrary, sFasL in serum differed among HIV-1 infected subjects according to clinical signs of ADC. In the cross-sectional study, the number of cells present in CSF and CD4+ T cell counts in blood did not correlate to the levels of CSF sFas and sFasL. Interestingly, the number of HIV RNA copies in CSF correlated significantly to the levels of CSF sFasL (P=0.001) but not to sFas in the same compartment. Antiretroviral therapy reduced viral load and sFas levels in CSF in the majority of patients. sFas is a useful marker for ADC diagnosis and follow-up during antiviral treatment.

Preferential loss of large neocortical neurons during HIV infection: a study of the size distribution of neocortical neurons in the human brain

The infection with human immunodeficiency virus (HIV) is associated with a global and severe loss of neocortical neurons. However, there is limited knowledge concerning whether all neurons are equally susceptible to damage during HIV infection. Other studies have reported low vulnerability of small interneurons and high vulnerability of large motor neurons. Thus, it is natural to suggest that HIV infection, which causes damage to neurons in several ways, may predominantly affect large neurons in the neocortex. In this study we have used three unbiased stereological probes: Cavalieri's principle, the optical dissector and the rotator method, to obtain both total neocortical neuron number and their size distribution in formalin-fixed brains from six male acquired immunodeficiency syndrome (AIDS) patients and six male controls. The material is a selection of a large material choosing the youngest. The number of neurons in neocortex was reduced by 25% from 24.4 x 10(9) in controls to 18.3 x 10(9) in the AIDS patients; the reduction is similar to that of 27% found in the large material. In the normal size distribution of the neocortical neurons most neurons were smaller than 5000 micron3 and no sampled neurons were larger than 28,000 micron3. In addition, the absolute size distribution of neocortical neurons showed a significant decrease of the largest group of neurons by 50% (2p = 0.01) in the AIDS group, while there was no significant difference between controls and AIDS patients in the number of small neurons. The pattern of reduction in the number of large neocortical neurons was found in frontal, temporal, parietal as well as in occipital regions. This suggests that large neurons are more sensitive than small neurons to the destruction caused by the HIV infection.