Sv. Tan et al., AIDS-ASSOCIATED VACUOLAR MYELOPATHY AND TUMOR-NECROSIS-FACTOR-ALPHA (TNF-ALPHA), Journal of the neurological sciences, 138(1-2), 1996, pp. 134-144
The spinal cords from 15 patients with AIDS-associated vacuolar myelop
athy (VM). 4 AIDS patients without VM, and 5 HIV-seronegative controls
, were studied with immunocytochemistry for TNF alpha. CSF and blood f
rom HIV-seropositive patients with VM (n = 16), non-vacuolar myelopath
ies (n = 8). CNS infection but no clinical myelopathy (n = 31), no cli
nical or radiological evidence of CNS disease (n = 9), and from 7 HIV-
seronegative controls with motor neurone disease were assayed for TNF
alpha using an ELISA technique. TNF alpha was present on immunostainin
g in all the 15 cords with VM studied. The stained cells were macropha
ges, microglia and endothelial cells. The amount of immunostaining was
higher in cords with VM compared with cords from HIV-seropositive pat
ients without VM (p = 0.001). the distribution of staining corresponde
d to the areas of pathology but did not correlate with the severity of
the VM. Immunostaining was also higher in the HIV-seropositive group
compared to the HIV-seronegative controls (p = 0.001). There was no si
gnificant difference in the levels of TNF alpha in the CSF of patients
with VM compared to any of the other groups studied. Blood levels of
TNF alpha were lower in the HIV-seropositive controls without CNS dise
ase and in the HIV-seronegative MND controls, than in patients with VM
, non-vacuolar myelopathies and CNS disease. CSF TNF alpha levels did
not appear to a reliable indicator of intramedullary levels. The findi
ngs support the hypothesis that TNF alpha may be relevant in the patho
genesis of vacuolar change in VM.