Variable progression of HIV-associated dementia.

A consecutive series of 71 patients diagnosed with HIV-associated dementia (HAD) (1984-1994) were studied to characterize the clinical course of HAD, and to identify predictive markers of rapid neurologic progression. Neurologic progression rate was determined from the change in the Memorial Sloan Kettering (MSK) dementia severity score from diagnosis to death. Those with the most rapid progression in neurologic disability were compared with those with slow or no progression. Autopsy material was immunostained for macrophage activation markers and gp41 in 30 individuals. Median survival was 3.3 months and 6.1 months for rapid-progression and no-progression patients, respectively. Rapid progression was associated with injection drug use but not with race, gender, or age. CD4+ cell counts were lower at diagnosis among rapid-progression than no-progression patients but no differences in AIDS-defining illnesses or patterns of antiretroviral therapy were found. At presentation, rapid-progression patients had more prominent symptoms of mental slowing than those with no progression; however, no other clinical features, CSF, or imaging features distinguished the groups. Less abundant macrophage activation in both basal ganglia and midfrontal gyrus regions, as judged by HAM56 immunostaining, was noted in 9 no-progression patients, compared with 12 rapid-progression patients. Neurologic progression and survival with HAD is highly variable. A significant proportion of individuals with dementia have prolonged survival of more than 12 months and remain cognitively stable. A history of injection drug use and presentation with prominent psychomotor slowing is associated with more rapid neurologic progression, and these patients tend to show more abundant macrophage activation within the CNS.

Sustained cognitive decline in HIV infection: relationship to CD4+ cell count, plasma viremia and p24 antigenemia.

To determine the clinical and virological correlates of neuropsychological test performance decline in HIV infection, we measured viral burden in blood in 272 HIV-seropositive men without dementia in the Baltimore arm of the Multicenter AIDS Cohort Study (MACS). These measures were then related to neuropsychological (NP) decline, defined as a decline relative to prior best performance of 2.0 standard deviations or more on one or more neuropsychological tests. A short battery of NP tests (Mini-Screen Battery) was administered to all 272 men. NP test performance decline was identified in 53/272 (19.5%) of participants on the Mini-Screen Battery. Follow-up NP data were available for 204 participants who had undergone the Mini-Screen. The frequency of sustained NP test performance decline was 7.8% for the Mini-Screen Battery. A lower CD4+ cell count was weakly associated with sustained NP test performance decline. After adjustment for CD4+ cell count, hemoglobin, body mass index, and presence of AIDS, none of the viral burden measures (p24 antigenemia, plasma viremia, quantitative culture) correlated with sustained NP test performance decline. We conclude that these measures of blood HIV viral burden are not markers for NP decline, but that a lower CD4+ cell count is.

Relationship between human immunodeficiency virus-associated dementia and viral load in cerebrospinal fluid and brain.

Cerebrospinal fluid (CSF) human immunodeficiency virus (HIV) RNA levels were measured with the Nucleic Acid Sequence-Based Amplification (NASBA) assay to determine the relationship with neurological status; 37 subjects with HIV dementia (HIV-D) were compared with 77 with HIV with minor neurological signs (HIV-MCMD) and 93 neurologically normal HIV-seropositive individuals (HIV-NML). The NASBA assay had a lower limit of detection of 100 copies per milliliter. Mean CSF log HIV RNA levels were significantly higher in those with dementia after adjusting for CD4 count and were correlated with dementia severity. Plasma levels did not distinguish comparably immunosuppressed subjects with or without dementia. CSF and plasma RNA levels were significantly intercorrelated for subjects with CD4 counts <200/mm3 and also correlated inversely with CSF beta2-microglobulin. CSF RNA levels were independent of CSF pleocytosis or antiretroviral exposure. Brain RNA levels were consistently higher than CSF but correlated with CSF values for dementia subjects. The NASBA assay can be used reliably to determine HIV RNA levels in CSF, brain, and plasma samples. CSF HIV RNA may be a surrogate marker for brain infection, based on the observed correlation with brain levels. The association between plasma HIV RNA and CSF levels of HIV and beta2-microglobulin suggests that both viral load and CNS immune activation are important determinants of neurological disease.

Psychomotor slowing in HIV infection: a predictor of dementia, AIDS and death.

The objective of this study was to determine if sustained decline in psychomotor speed tests is associated with an increased risk of progression to dementia, acquired immunodeficiency syndrome (AIDS), or mortality in human immunodeficiency virus (HIV)-1-infected homosexual men in the Baltimore site of the Multicenter AIDS Cohort-Study (MACS). Clinical and neuropsychological data were obtained on 291 HIV+ homosexual men seen semi-annually over a nine year period (1986-1994). A proportional hazards model was used to assess the predictive value of sustained psychomotor slowing (defined as a 2.0 standard deviation (s.d.) decline in performance on either the Symbol Digit Modalities test or Trailmaking test at two consecutive evaluations). Time-dependent co-variates included in the model were sustained psychomotor slowing, number of attended visits, CD4+ lymphocyte count, hemoglobin and antiretroviral medication use. HIV+ participants with and without sustained psychomotor slowing were compared. Outcome variables were the development of dementia, AIDS and death. HIV+ subjects with sustained psychomotor slowing had an increased hazard of dementia (Risk ratio (RR) = 5.0, P = 0.008), AIDS (RR = 2.4, P = 0.02), and death (RR = 2.0, P = 0.04). A similar analysis using sustained cognitive decline in one domain from a more extensive neuropsychological test battery failed to show any predictive value. Sustained decline in psychomotor performance in HIV infection was predictive of dementia, AIDS and death. This brief neuropsychological test battery may be useful for early detection of HIV+ individuals with a poorer prognosis who may benefit from more aggressive treatment to prevent HIV dementia.

Cerebrospinal fluid human immunodeficiency virus type 1 (HIV-1) p24 antigen levels in HIV-1-related dementia.

Human immunodeficiency virus type 1 (HIV-1) p24 antigen, a putative marker of virus load, was assayed in 79 blood and 83 cerebrospinal fluid (CSF) samples from 90 HIV-1-seropositive individuals with or without dementia. Twenty-eight subjects had no evidence of neuropsychological impairment, 17 had mild impairment without objective evidence of dementia, and 45 were demented. HIV-1 p24 antigen was detected more frequently in CSF samples from demented (19/40) than normal (1/26) or mildly impaired (1/17) subjects and in 67% of individuals with significant dementia (MSK stages 2-4). p24 Antigen was detected less frequently in CSF from demented subjects on antiretroviral drugs than untreated demented individuals. Overall, the sensitivity of the antigen capture assay in CSF among demented individuals was 47.5%; the specificity, 95.0%; positive predictive value, 90.4%; negative predictive value, 66.1%; and the efficiency, 72.2%. A direct relationship was also noted between the degree of cognitive impairment and blood p24 antigen detection frequency and antigen concentration. CD4+ blood lymphocyte counts were lower for demented individuals, and HIV-1 p24 antigen was detected more frequently and p24 antigen concentration was higher in blood and CSF from individuals with low CD4+ blood lymphocyte counts. beta 2-Microglobulin levels were higher in CSF from demented subjects and correlated directly with CSF p24 antigen concentration. However, in contrast to CD4+ blood lymphocyte counts and beta 2-microglobulin levels, only p24 antigen concentration correlated with dementia severity. Therefore, p24 antigen can be a useful marker for dementia related to HIV-1 infection.

Regional brain atrophy in HIV-1 infection: association with specific neuropsychological test performance.

Quantified magnetic resonance imaging (MRI) was related to neuropsychological (NP) test scores in an asymptomatic HIV-1 seropositive group, a non-demented AIDS/ARC group, a group of subjects with HIV-1 dementia, and a seronegative control group. The MRIs were quantified using three planimetric measures of brain structure: the bicaudate ratio (a measure of caudate region atrophy), the bifrontal ratio (a measure of frontal region atrophy), and the ventricle to brain ratio (a measure of overall cerebral atrophy). Cognitive performance was assessed with standard NP tests. Significant correlations between the MRI ratios and many of the NP tests were observed. Of the tests grooved pegboard, part B of the trail making test, the verbal fluency test, and the digit span forward were associated with MRI abnormalities. The bicaudate ratio was most closely associated with the NP tests. These findings indicate that ventricular enlargement, especially in the region of the caudate, is closely related to poor NP test performance in HIV-1 infection.

Patterns of cerebral atrophy in HIV-1-infected individuals: results of a quantitative MRI analysis.

Cerebral atrophy is a common radiologic manifestation of HIV dementia. To evaluate the relationship between cognitive impairment and cerebral atrophy, adjusting for age and immune status, we used standardized planimetry to measure the ventricle-brain ratio (VBR) and the bifrontal (BFR) and bicaudate (BCR) ratios, three measures of cerebral atrophy. We analyzed cranial MRIs of 23 HIV-1-seronegative controls (SN) and 116 HIV-1-infected individuals. Of the HIV-1-seropositive individuals, 37 had HIV dementia (DM group), 40 had neurologic or neuropsychological abnormalities insufficient for HIV dementia (NP+ group), and 39 were neurologically normal (NML group). We performed comparisons using analysis of covariance with correction for multiple comparisons. Both the VBR, a general measure of overall cerebral atrophy, and the BCR, a measure of atrophy in the region of the caudate nucleus, are significantly associated with dementia. The association is stronger for BCR enlargement than for VBR enlargement, suggesting that selective caudate region atrophy is associated with HIV dementia. These results indicate that overall cerebral atrophy and prominent caudate region atrophy are important radiographic features of HIV dementia.

The diagnostic utility of elevation in cerebrospinal fluid beta 2-microglobulin in HIV-1 dementia. Multicenter AIDS Cohort Study.

We measured serum and CSF beta 2-microglobulin (beta 2M) levels in HIV-1 seropositive individuals with and without dementia to determine the frequency and diagnostic utility of elevation of CSF beta 2M. We compared 34 samples from 27 patients with HIV-1 dementia with 110 samples from 54 HIV-1 seropositive participants in the Multicenter AIDS Cohort Study, none of whom had progressive dementia. Neurosyphilis and CNS opportunistic processes were excluded in all subjects. We stratified the nondemented subjects by duration of HIV seropositivity and peripheral blood CD4 count. Compared with the nondemented group, demented subjects had significantly higher CSF total protein, IgG%, and CSF albumin/serum albumin ratios. A highly significant association was found between elevated CSF beta 2M and reduced CD4 count (p less than 0.0001). No significant differences were noted between the demented and nondemented groups in CSF WBC count or in the frequency of CSF HIV-1 isolation. The mean CSF beta 2M was 1.9 mg/l in the nondemented subjects compared with 4.2 mg/l in those with dementia (p less than 0.0001). We derived a cutoff of 3.8 mg/l from the distribution of CSF beta 2M in the nondemented group. The determination of CSF beta 2M had a sensitivity of 44%, specificity of 90%, and a positive predictive value of 88% for diagnosis of HIV dementia when compared with nondemented subjects with CD4 counts less than 200. In those without dementia, there was a strong correlation between serum and CSF beta 2M (r = 0.50, p less than 0.0001), but in demented subjects CSF beta 2M was elevated independently of serum levels, suggesting that CSF beta 2M is produced within the brain in HIV dementia. In the absence of CNS opportunistic processes, elevated CSF beta 2M greater than 3.8 mg/l is a clinically useful marker for HIV dementia.