Correlating brain volume and callosal thickness with clinical and laboratory indicators of disease severity in children with HIV-related brain disease.

Abstract

Objective MRI markers of central nervous system disease severity may precede subjective features of HIV encephalopathy in children. Previous work in HIV-infected adults shows that brain atrophy was associated with low CD4 and with neuropsychological impairment. Significant thinning of the corpus callosum (CC), predominantly anteriorly, was also found in HIV-infected adults and correlated with CD4 levels. These findings have not been tested in children. The aim of this study was to determine if brain volume and midsagittal CC linear measurements (thickness and length) on MRI in children with HIV-related brain disease correlate with clinical and laboratory parameters of disease severity. Retrospective MRI analysis in children with HIV-related brain disease used a volumetric analysis software and a semi-automated tool to measure brain volume and callosal thickness/length, respectively. Each measure was correlated with clinical parameters of disease severity including Griffiths Mental Development scores (GMDS), absolute CD4 counts (cells/mm(3)), nadir CD4 (the lowest CD4 recorded, excluding baseline), duration of HAART, and decreased brain growth. Thirty-three children with HIV-related brain disease were included. Premotor segment of the CC mean thickness correlated with age (p = 0.394). Motor CC maximum thickness correlated significantly with general developmental quotient (p = 0.0277); CC length correlated with a diagnosis of acquired microcephaly (p = 0.0071) and to CD4 level closest to date of the MRI scan (p = 0.04). Length of the CC and the "motor CC segment" may represent surrogate clinical biomarkers of central nervous system disease severity and with decreased level of immunity in HIV-infected patients that precede established HIV encephalopathy.