RESUMEN
OBJECTIVES: HIV-associated dementia (HAD) is the most severe clinical expression of HIV-mediated neuropathology, and the processes underlying its development remain poorly understood. We aimed to exploit high-dimensional metabolic profiling to gain insights into the pathological mechanisms associated to HAD. DESIGN: In this cross-sectional study, we utilized metabolomics to profile matched cerebrospinal fluid (CSF) and plasma samples of HAD individuals ( n â=â20) compared with neurologically asymptomatic people with HIV (ASYM, n â=â20) and healthy controls (NEG, n â=â20). METHODS: Identification of plasma and CSF metabolites was performed by liquid-chromatography or gas-chromatography following a validated experimental pipeline. The resulting metabolic profiles were analyzed by machine-learning algorithms, and altered pathways were identified by comparison with KEGG pathway database. RESULTS: In CSF, HAD patients displayed an imbalance in glutamine/glutamate ratio, decreased levels of isocitrate and arginine, and increased oxidative stress when compared with ASYM or NEG. These changes were confirmed in matched plasma samples, which in addition revealed an accumulation of eicosanoids and unsaturated fatty acids in HAD individuals. Pathway analysis in both biological fluids suggested that alterations in several metabolic processes, including protein biosynthesis, glutamate and arginine metabolism, and energy metabolism, in association to a perturbed eicosanoid metabolism in plasma, may represent the metabolic signature associated to HAD. CONCLUSION: These findings show that HAD may be associated with metabolic modifications in CSF and plasma. These preliminary data may be useful to identify novel metabolic biomarkers and therapeutic targets in HIV-associated neurological impairment.