Objective: To evaluate whether baseline levels of plasma and CSF HIV RNA, tumor necrosis factor alpha (TNFalpha), monocyte chemoattractant protein-1 (MCP-1), matrix metalloproteinase-2 (MMP-2), or macrophage colony stimulating factor (M-CSF) are predictors of incident HIV-associated dementia (HIVD) in a cohort with advanced HIV infection.
Methods: A total of 203 nondemented subjects with CD4 lymphocyte counts less than 200/muL, or <300/microL but with cognitive impairment, underwent semiannual neurologic, cognitive, functional, and laboratory assessments. HIVD and minor cognitive motor disorder (MCMD) were defined using American Academy of Neurology criteria. The cumulative incidence of HIVD was estimated using Kaplan-Meier curves. Cox proportional hazards regression models were used to examine the associations between biologic variables and time to HIVD, adjusting for age, sex, years of education, duration of HIV infection, type of antiretroviral use, premorbid IQ score, and presence of MCMD.
Results: After a median follow-up time of 20.7 months, 74 (36%) subjects reached the HIVD endpoint. The dementia was mild in 70% of cases. The cumulative incidence of HIVD was 20% at 1 year and 33% at 2 years. Highly active antiretroviral therapy (HAART) was used by 73% of subjects at baseline. A plasma HIV RNA level was undetectable in 23% of subjects and a CSF HIV RNA level was undetectable in 48% of subjects. In adjusted analyses, neither plasma nor CSF HIV RNA levels (log10) were associated with time to HIVD; log10 levels of plasma TNFalpha (HR 3.07, p = 0.03) and CSF MCP-1 (HR = 3.36, p = 0.06) tended to be associated with time to HIVD.
Conclusion: The lack of association between baseline plasma and CSF HIV RNA levels and incident dementia suggests highly active antiretroviral therapy may be affecting CNS viral dynamics, leading to lower HIV RNA levels, and therefore weakening the utility of baseline HIV RNA levels as predictors of HIV-associated dementia.