Volumetric brain analysis in neurosurgery: Part 2. Brain and CSF volumes discriminate neurocognitive outcomes in hydrocephalus

OBJECT The evaluation of hydrocephalus remains focused on ventricular size, yet the goal of treatment is to allow for healthy brain development. It is likely that brain volume is more related to cognitive development than is fluid volume in children with hydrocephalus. This study tests this hypothesis by comparing brain and fluid volumes with. neurocognitive outcome in pediatric patients with hydrocephalus. METHODS Warf and colleagues previously acquired CT scans for pediatric patients in Uganda with myelomeningocele, measured frontal-occipital horn ratio (FOHR), and administered the modified Bayley Scales of Infant Development, third edition (BSID-III) to measure neurocognitive outcome that did not correlate with FOHR. In this present study, brain and fluid volumes were measured in 33 of these patients, 26 of whom required surgical treatment for hydrocephalus. Linear discrimination analysis (LDA) was used to test whether age-normalized brain and fluid volumes can discriminate neurocognitive outcome. RESULTS Hydrocephalic patients show normal to small brain volumes and substantially larger fluid volumes compared with normal values. FOHR correlates highly with fluid volume (r = 0.84, p < 0.001) and substantially less with brain volume (r = -0.37, p = 0.03), while brain and fluid volumes do not correlate with each other (p = 0.99). Brain and CSF volumes correlated best with fine motor (p = 0.03, p = 0.01), cobnitive (p = 0.05, p = 0.09), and expressive communication (p = 0.08, p = 0.08) scores. A combination of these 3 scores was used as a multivariate measure of neurocognitive outcome. Brain volume alone, unlike fluid volume, could discriminate high from low cognitive outcome (by t-test and ANOVA). It was shown that a combination of age-normalized brain and fluid volumes can discriminate neurocognitive outcome by 2-way LDA (p < 0.01) and 3-way LDA (p < 0.01). The multivariate LDA demonstrated the contribution of large fluid volume to a decrement in cognition. CONCLUSIONS Hydrocephalus is treated by normalizing CSF, but normal brain development depends on brain growth. A combination of brain and CSF volumes appears to be significantly more powerful at predicting good versus poor neurocognitive outcomes in patients with hydrocephalus than either volume alone.