Independent contributions of cortical gray matter atrophy and ventricle enlargement for predicting neuropsychological impairment in multiple sclerosis
Introduction
Recent studies showing that brain atrophy is a better predictor of neuropsychological deterioration in multiple sclerosis (MS) than lesion burden (Benedict et al., 2004c, Bermel and Bakshi, 2006, Christodoulou et al., 2003, Zivadinov et al., 2001) have raised questions about the traditional view of MS as a subcortical or white-matter (WM) dementia (Filley et al., 1989, Swirsky-Sacchetti et al., 1992). Slowed processing speed is a hallmark of MS-associated cognitive dysfunction and the disruption of white matter tracts interconnecting association cortices has long been considered a viable explanation of such slowing. However, in MS, gray matter (GM) volume appears to have unique (Sanfilipo et al., 2006), if not greater (Bakshi et al., 2005, Dalton et al., 2004) clinical relevance than WM lesions or volume. GM atrophy is more strongly associated with cognitive decline, especially when the cerebral cortex is assessed (Amato et al., 2004, Benedict et al., 2006a). Thus, it would appear that cognitive impairment in MS can be attributed to cortical as well as subcortical cerebral pathology.
It is also well established that structural imaging measures such as whole brain volume, tissue compartment (e.g. GM) volumes, ventricle size and lesion volume, are strongly inter-correlated. Thus, when stepwise regression models are employed to predict neuropsychological outcomes on the basis of MRI, usually only one variable remains in the model (Benedict et al., 2006a, Benedict et al., 2004c, Benedict et al., 2004a) and measures of third or lateral ventricle volume (i.e. central atrophy) yield the strongest effects (Benedict et al., 2004a, Christodoulou et al., 2003). In other words, when investigators attempt to learn which structural imaging measure is most closely aligned with MS-associated cognitive impairment, models accounting for a general measure such as ventricle enlargement leave little residual variance for understanding more specific brain–behavior relationships. Previously, we measured regional lobar atrophy (regional brain parenchymal fraction of frontal, temporal, parietal and occipital lobes in each hemisphere) to predict cognitive performance in MS patients (Benedict et al., 2005, Locatelli et al., 2004). The results revealed that temporal atrophy was a significant predictor of both auditory/verbal and visual/spatial memory whereas learning consistency was predicted by frontal atrophy. This study did not control for the effects of ventricle enlargement, a powerful predictor of NP impairment in MS (Benedict et al., 2006a, Benedict et al., 2004c, Benedict et al., 2004a).
In the present study, we investigated the null hypothesis that regional cortical atrophy does not contribute to the prediction of neuropsychological impairment after accounting for third ventricle width. We used semi-automated brain region extraction (SABRE) to parcellate the brain into 26 regions in accordance with previous research (Carone et al., 2006, Dade et al., 2004, Feinstein et al., 2004). We measured GM atrophy in those regions of the brain where GM corresponds to the cerebral cortex rather than the deep gray matter. Regression models predicting cognitive outcomes included these GM volumes after controlling for the influence of third ventricle width.
Section snippets
Participants
We retrospectively studied 59 patients with clinically definite MS (McDonald et al., 2001) using brain MRI and neuropsychological testing assessments within a 180-day window (mean difference 67.46 ± 50.93, range 0–175 days). Patients with preexisting medical conditions associated with NP compromise were excluded, as were patients receiving steroid treatment or who had relapsed in the 3 months prior to participation. Demographic data (mean ± SD) were as follows: age 44.7 ± 8.6 years, education 14.2 ±
Results
Partial correlation coefficients controlling for age between third ventricle width and neuropsychological tests were significant for CVLT2-TL (r = − 0.41, p < 0.001), CVLT2-DR (r = − 0.44, p < 0.001), BVMTR-TL (r = − 0.29, p < 0.5), PASAT (r = − 0.29, p < 0.05) and SDMT (r = − 0.59, p < 0.001). Although BVMTR-DR was negatively correlated with third ventricle width, this correlation did not reach the level of statistical significance. There were no significant correlations between BDIFS and either rGMFs or NP scores.
Discussion
To the best of our knowledge, this is the first study showing that both central and cortical atrophy contribute independently to predicting the presence of slowed processing speed and memory disorder in MS. Various structural MRI measures of brain volume and pathology are highly intercorrelated. It is now well documented that cortical volume is correlated with cognitive impairment in MS (Amato et al., 2004, Benedict et al., 2006a), and in one study, there was a regionally specific relationship
References (53)
- et al.
Voxel-based morphometry—The methods
NeuroImage
(2000) - et al.
Predicting neuropsychological abnormalities in multiple sclerosis
J. Neurol. Sci.
(2006) - et al.
The measurement and clinical relevance of brain atrophy in multiple sclerosis
Lancet Neurol.
(2006) - et al.
Semiautomatic brain region extraction (SABRE) reveals superior cortical and deep gray matter atrophy in MS
NeuroImage
(2006) - et al.
Semiautomatic brain region extraction: a method of parcelling brain regions from structural magnetic resonance images
NeuroImage
(2004) - et al.
Evidence for a direct association between cortical atrophy and cognitive impairment in relapsing-remitting MS
NeuroImage
(2006) - et al.
Correction for intracranial volume in analysis of whole brain atrophy in multiple sclerosis: the proportion vs. residual method
NeuroImage
(2004) - et al.
Short-term brain atrophy changes in relapsing-remitting multiple sclerosis
J. Neurol. Sci.
(2004) - et al.
California verbal learning test: performance by patients with focal frontal and nonfrontal lesions
Brain
(2003) - et al.
Neocortical volume decrease in relapsing-remitting MS patients with mild cognitive impairment
Neurology
(2004)