Objective: We investigated whether HIV brain latency was associated with brain injury in virally suppressed HIV infection.
Design: Observational cross-sectional and longitudinal study.
Methods: The study included 26 virally suppressed HIV-infected men (61.5% with HIV-associated neurocognitive disorder) who undertook cerebrospinal fluid (CSF) analyses at baseline. They also completed a proton magnetic resonance spectroscopy (1H MRS) and neuropsychological assessments at baseline and 18 months. To quantify whether there was residual brain HIV transcription, we measured CSF HIV-tat. As an HIV brain latency biomarker, we used concentrations of CSF BcL11b - a microglia transcription factor that inhibits HIV transcription. Concurrently, we assessed neuroinflammation with CSF neopterin, neuronal injury with CSF neurofilament light-chain (NFL), and in-vivo neurochemistry with 1H MRS of N-acetyl aspartate (NAA), choline (Cho), creatine, myo-inositol (MI), glutamine/glutamate (Glx) in the frontal white matter (FWM), posterior cingulate cortex (PCC), and caudate nucleus area.
Results: Baseline adjusted regression models for neopterin, NFL, and tat showed that a higher CSF BcL11b was consistently associated with lower FWM creatine (when adjusted for neopterin: β = -0.30, P = 0.15; when adjusted for NFL: β = -0.47, P = 0.04; and when adjusted for tat: β = -0.47, P = 0.02). In longitudinal analyses, we found no time effect, but a consistent BcL11b altering effect on FWM creatine. The effect reached a significant moderate effect size range when corrected for CSF NFL (β = -0.36, P = 0.02) and CSF tat (β = -0.34, P = 0.02).
Conclusions: Reduced frontal white matter total creatine may indicate subclinical HIV brain latency-related injury. H MRS may offer a noninvasive option to measure HIV brain latency.