Cognitive impairment still common but ART reduces risk

Cognitive impairment remains common amongst people with HIV and is linked to more severe immune deficiency and absence of treatment, researchers reported at the International AIDS Society Conference (IAS 2011) this week in Rome. However, antiretroviral drugs that penetrate the central nervous system do not appear to improve overall outcomes.

Prevalence and risk factors

Valerio Tozzi with the Italian National Institute of Infectious Disease and colleagues looked at prevalence of and risk factors for HIV-associated neurocognitive disorders, or HAND, from the advent of combination antiretroviral therapy (ART) in 1996 through 2010. They also assessed severity and qualitative changes in cognitive function.

This observational study included 1375 HIV-positive patients on antiretroviral therapy at a single site in Italy. About 75% were men, the median age was 42 years, and they had a median 13 years of education. People with psychiatric or neurological conditions and active drug users were excluded.

Participants had been infected with HIV for a median of six years and about one-third had an AIDS diagnosis. The current median CD4 cell count was near 400 cells/mm³, but the nadir or lowest-ever count was 165 cells/mm³.

Participants underwent a series of about a dozen neuropsychological tests to measured cognitive function in five domains. These tests measured abilities such as verbal learning and ability to draw a complex figure from memory. Performance was compared to normal values for the HIV negative general population matched for sex and age.

The researchers used statistical methods to determine factors associated with higher or lower risk of HAND, looking at three clinical categories of cognitive impairment: HIV-associated dementia, mild neurocognitive disorder, and asymptomatic neurocognitive impairment.

Overall, HAND prevalence decreased only slightly over the course of the study, falling from 46% during 1996-1998 to 38% during 2008-2010. Severity also shifted over time. During 1996-1998, 30% of participants had dementia or mild neurocognitive impairment. Over time this fell steadily, to 18% by 2008-2010, but the proportion of people with asymptomatic impairment rose.

In an unadjusted analysis, people with HAND were significantly older on average, older when they had their first HIV test, and had been infected for a longer duration than those without; they also had about two years less education. . People with HAND were much more likely to have received an AIDS diagnosis than those without (49% vs 24%). They had lower current (349 vs 478 cells/mm³) and nadir (158 vs 231 cells/mm³) CD4 counts and were more likely to have hepatitis C co-infection (51% vs 31%).

In a multivariate analysis that controlled for others factors, five factors remained "strongly and significantly" associated with HAND: older age, higher education level, AIDS diagnosis, and current CD4 cell count.

Nadir CD4 count and hepatitis C status were no longer significant predictors of HAND, conflicting with some other studies. Tozzi explained that the strong association with AIDS diagnosis indicated that history of immune deficiency had an influence, but suggested that persistent rather that past immunodeficiency is most relevant.

Looking only at 569 people with dementia or mild symptomatic impairment, age, education, AIDS diagnosis, and current CD4 count were again significant predictors in a multivariate analysis. But here, having cardiovascular risk factors was also associated with a higher likelihood of HAND. Tozzi suggested this link might be related to chronic inflammation.

Impairment in specific domains was variable. The researchers noted some evidence of reduced impairment in measures of visual memory and spatial perception, but most areas¾including attention and memory, motor skills, and processing speed¾remained about the same, indicating "very limited evidence" for a change in HAND neurocognitive profile.

CNS penetration

If HIV infection is associated with neurocognitive impairment, it makes sense to ask whether antiretroviral drugs that enter the central nervous system (CNS) and inhibit viral replication in the brain might lead to improvement.

Sean Rourke's team from Canada hypothesized that people using antiretroviral regimens with a higher CNS penetration effectiveness (CPE) score would do better in overall neurocognitive function and specific domains.

CPE is a measure developed by Scott Letendre and colleagues to rank how well drugs enter the brain, based on chemical properties, levels in cerebrospinal fluid (CSF) and other factors. The index was first published in 2006 and revised in 2010 to include newer antiretrovirals.

This analysis included 529 participants at two hospital clinics in Ontario. Again, most were men, two-thirds were white, the mean age was about 48 years, and the average education level was about 14 years. About half had a current CD4 count below 500 cells/mm³, but two-thirds had a CD4 nadir below 200 cells/mm³.

Using 2006 CPE scores, just under half of participants were taking regimens with high CNS penetration. Using 2010 scores, the proportion rose to 60%.

Looking at global neuropsychological impairment and a number of domain-specific cognitive measures, the researchers found that more than 50% of participants showed some degree of impairment.

Overall, there was no significant relationship between CPE score and neurocognitive outcomes using either the 2006 or 2010 criteria. This remained the case after adjusting for demographics and disease-related factors.

Although most individual tests and domains showed no difference, one test (spatial span) showed significant improvement with higher CPE scores, and one (digit-symbol test) showed significant worsening when using 2006 scores. Rourke suggested that it is important to look at specific tests and domains as well as global cognitive measures.

Taken together, these studies indicate that cognitive impairment is a common problem among people with HIV. While antiretroviral treatment is associated with improved functioning, specific drugs that enter the brain do not appear to make much difference. This finding adds support for the concept that the detrimental effects of HIV in the brain are related to inflammatory processes that can be triggered by even a small amount of virus.

Produced in collaboration with HIVandHepatitis.com