There is controversy as to whether Asymptomatic Neurocognitive Impairment (ANI) in HIV-Associated Neurocognitive Disorders (HAND) solely represents a statistical neuropsychological entity with no neurobiological underpinning and no HIV causation. We hypothesized that in a sample of non-confounded virally suppressed HIV+ persons versus demographically comparable controls, HIV-related grey and white matter (WM) atrophy could be observed in ANI.
85 HIV+ (plasma & CSF HIV RNA <50cp/mL, median nadir CD4=180, CD4=528) and 44 demographically comparable HIV- men (mean age=55; mean education=14.5 years; 90% men who have sex with men, 95% White Australian) underwent anatomical MRI, neuropsychological evaluation, and HIV laboratory tests. Volumes of interest (VOI) from MR images were extracted using Freesurfer to yield grey and WM in regions linked to HIV-related brain injury (total cortical volume, basal ganglia, lateral ventricles, fronto-striatal and fronto-parietal WM, all relative to Total Intracranial Volume, TIV). HAND status was ANI=38%, MND=13%, HAD= 3% based on the Global Deficit Score (GDS≥0.5) and functional decline; others were neuropsychologically (NP)-normal. A history of HAND occurred in 17.6%. We used multivariate analyses controlling for family-wise error rate to assess the effects of HIV status group on VOI. Next, ANI (N=32), MND (N=10) vs. NP-normal (N=40) and HIV biomarker (HIV duration, nadir CD4, current CD4 & CD8) effects on VOI were assessed with multiple regression controlling for a history of HAND.
Relative to the HIV- group, the HIV+ group demonstrated subcortical grey (d=0.50-0.60) and WM (d=0.43-0.69) atrophy, with relative cortical sparing (d=0.23). ANI showed reduced medial-orbitofrontal WM compared to NP-normal cases (p=.04, β=-.31). MND showed enlarged lateral ventricles (p=.02, β=.34), reduced caudal-middle-frontal WM (p=.04, β=-.32), reduced caudal-anterior-cingulate WM (p=.006, β=-.42), and reduced inferior-parietal WM (p=.04, β=-.33) compared to NP-normal cases. HIV disease duration predicted greater medial-orbitofrontal WM atrophy only in ANI (p=.002, β=-.51; Fig.1). Higher CD8 were independently associated with atrophy in the inferior-parietal WM (p=.003, β=-.41).
ANI is associated with specific frontal WM atrophy. HIV disease duration is a unique contributor to ANI related brain atrophy. These findings give neurobiological validity to ANI and may serve as an ANI biomarker.