The neuropathology of adult HIV infection.

Since the onset of the acquired immune deficiency syndrome (AIDS) epidemic fifteen years ago, much has been learned about the effects of the human immunodeficiency virus (HIV) in the nervous system. This review summarizes the pathology findings in the central nervous system (CNS). There is now abundant evidence that HIV can infect the CNS directly, leading to a characteristic HIV encephalitis (HIVE) which occurs in 10-50 p. 100 of AIDS autopsy series. Multinucleated giant cells are the pathognomonic feature of HIVE and are found predominantly in the central white matter and deep grey matter. Evidence of productive HIV infection in the CNS is confined to cells of the microglial/macrophage lineage, from which the giant cells are almost certainly derived. These cells are known to express both CD4 and beta-chemokine receptors, which act in conjunction to permit HIV entry. Restricted infection of astrocytes has also been identified by a variety of methods. HIVE is frequently associated with white matter damage ranging from inflammatory (microglia, macrophages and sparse lymphocytes) to degenerative (myelin loss and axonal damage) pathology. Although giant cells are seen less frequently in neocortical grey matter, significant neuronal loss has been established in a number of studies. Recent investigations using markers of apoptosis, (including TUNEL, Bcl-2 and BAX), have established the presence of DNA damage in some neurons and in other cell types. Axonal damage has also been confirmed by evidence of amyloid precursor protein expression. The CNS is also vulnerable to opportunistic infections and high grade B-cell lymphomas as a result of the immune suppression of advanced HIV infection. Cytomegalovirus (CMV) infection is reported in 10-30 p. 100 of AIDS cases at autopsy, toxoplasma in 10-25 p. 100, progressive multifocal leucoencephalopathy in about 5 p. 100 and lymphomas, usually primary, in up to 10 p. 100. A wide variety of other infections has also been reported. These may coexist with HIVE and may be difficult to diagnose in life. CMV gives rise to microglial nodular encephalitis, ventriculitis, necrotising encephalitis and myelo-radiculitis. Presymptomatic HIV positive patients do not show HIVE or opportunistic infections or lymphomas in the CNS. They frequently display a low-grade T-cell infiltrate in the leptomeninges and parenchyma, particularly around vessels. This lymphocytic infiltrate has been attributed to presumed early invasion of the CNS by HIV although the exact timing of entry is uncertain. It is possible that reported abnormalities in presymptomatic cases such as gliosis, microglial activation and rising proviral load may anticipate the onset of HIVE but most studies show that significant CNS damage and HIV-related pathology is confined to patients with AIDS. HIV-related pathology in the spinal cord includes not only HIV myelitis, opportunistic infections and lymphomas, but also vacuolar myelopathy (VM) which affects predominantly the dorsolateral white matter tracts. The cause of VM is not understood and has not been unequivocally linked with HIV infection. It is noted that none of these neuropathological features (including HIVE) correlates exactly with the clinical expression of AIDS-related dementia (ARD). The exact contribution of macrophage activation and cytokine release, astrocytic infection, neuronal loss and axonal damage to the neuropsychiatric syndromes of advanced HIV infection remain to be determined. While the current understanding of the pathogenesis of HIVE and ARD is beyond the scope of this review it is axiomatic that accurate documentation of neuropathology findings will help to resolve the outstanding dilemmas relating to HIV infection of the CNS. There is considerable optimism that progress in therapeutic regimes for HIV-infected patients will succeed in eliminating the virus from the blood and from lymphoid tissue. (ABSTRACT TRUNCATED)