Mapping limbic network organization in temporal lobe epilepsy using morphometric correlations: Insights on the relation between mesiotemporal connectivity and cortical atrophy
Introduction
Temporal lobe epilepsy (TLE) is the most common medically intractable partial epilepsy in adults. Although hippocampal atrophy is a hallmark of the disorder, magnetic resonance image (MRI) volumetry and voxel-based morphometry have shown extensive temporal (Bernasconi et al., 1999, Bernasconi et al., 2003, Bernasconi et al., 2001, Bernasconi et al., 2004, Jutila et al., 2001) and extra-temporal pathology (Bernasconi et al., 2004, Coste et al., 2002, Jutila et al., 2001, McMillan et al., 2004, Moran et al., 2001, Natsume et al., 2003a, Seidenberg et al., 2005). These findings suggest that the evaluation of the hippocampus in isolation is insufficient to understand the pathophysiology of this condition.
Entorhinal cortex atrophy is found in a large spectrum of patients, including those with hippocampal atrophy (Bernasconi et al., 1999, Jutila et al., 2001) and normal volumes (Bernasconi et al., 1999, Bernasconi et al., 2001), supporting the view that this structure plays a pivotal role in the epileptogenic network of TLE (Bartolomei et al., 2005, Roch et al., 2002). From a functional point of view, the entorhinal cortex is the gatekeeper for the reciprocal flow of information between the hippocampus and the neocortex (Insausti et al., 1987b, Van Hoesen and Pandya, 1975, Van Hoesen et al., 1975, Witter, 1993). Assessing the relationship between the entorhinal periallocortex and the neocortex promises to provide new insights into the pathophysiology of TLE.
Quantitative MR image analysis techniques provide metrics for in vivo assessment of structural integrity of the human brain. Voxel-based morphometry enables an automated whole-brain analysis of estimated gray matter (GM) concentration surrounding a given voxel (Ashburner and Friston, 2000, Ashburner and Friston, 2001). Although frequently utilized to characterize structural abnormalities in TLE (Bernasconi et al., 2004, Bonilha et al., 2004, Keller et al., 2001), this technique may not be the most appropriate to assess cortical pathology, since its smoothing step neglects anatomical relationships across a folded surface (Singh et al., 2006). Furthermore, anatomical variability in gyrification and sulcation may reduce the sensitivity to detect significant effects, even after non-linear warping and smoothing (Ashburner and Friston, 2001, Bookstein, 2001, Tisserand et al., 2002). Measuring cortical thickness is a more direct and biologically meaningful technique to quantify atrophy as it allows studying continuous changes with respect to the anatomy of the folded cortical surface. Advanced image processing methods enable automatic and reliable measurement of cortical thickness by calculating the distance between corresponding points on both the GM and white matter (WM) surfaces across the entire cortical mantle (Dale et al., 1999, Kim et al., 2005, MacDonald et al., 2000, Thompson et al., 2004). Analysis of cortical thickness in TLE has been so far limited to two studies based on small sample of patients (Lin et al., 2007, McDonald et al., 2008). Although TLE is considered primarily a limbic disorder, the study of the limbic network has been done only by a few diffusion tractography studies (Concha et al., 2005). Diffusion-based tractography techniques (Mori et al., 1999) would allow a precise mapping of connections and their alterations in TLE. However, tractography is problematic in the vicinity of mesiotemporal and neocortical regions, where fibers are of small diameter, branch and cross. Despite promising advances to overcome these drawbacks (Behrens et al., 2003), tractography alone may not capture the entire extent of pathological interactions, which may involve intermediate relay structures.
Morphometric correlational analysis has been proposed as an alternative mean of studying connectivity patterns between various cerebral regions (Rykhlevskaia et al., 2008). This approach depends upon the assumption that positive correlations indicate connectivity. Indeed, our previous cortical thickness correlational analysis of the language network showed a pattern remarkably similar to diffusion tensor imaging data (Lerch et al., 2006). In a diseased population, such correlation analyses have the potential to map networks that undergo common pathological processes.
Our purpose was to map limbic network organization in TLE and to statistically relate it to neocortical atrophy. Such systematic analysis has not been carried out in TLE and would allow assessing the relationship between mesiotemporal connectivity and overall cortical damage. We chose the entorhinal cortex as the seed region since this structure is considered the main link between the cerebral cortex and the hippocampal formation. Limbic networks were statistically inferred by correlating the mean thickness of the entorhinal cortex with thickness measurements across the entire cortical surface in healthy controls and patients with pharmacologically intractable TLE.
Section snippets
Subjects
We randomly selected from our database 110 patients referred to our hospital for the investigation of medically intractable TLE. Demographic and clinical data were obtained through interviews with the patients and their relatives. TLE diagnosis and lateralization of the seizure focus were determined by a comprehensive evaluation including detailed history, neurological examination, review of medical and EEG records, and neuropsychological assessment in all patients. Based on these criteria, the
Group analysis
Fig. 1a shows areas of significant cortical thinning in TLE. In LTLE, we observed thinning in bilateral fronto-central (i.e., in the superior, middle, and medial frontal gyrus, the precentral gyrus and paracentral lobule), cingulate and contralateral medial occipito-temporal regions. The most severe frontal thinning occurred in ipsilateral precentral regions with an absolute decrease of more than 0.3 mm (corresponding to > 10% decrease in cortical thickness). Ipsilateral to the seizure focus,
Discussion
We analysed cortical thickness in a large group of patients with pharmacologically intractable TLE using a validated surface-based method that has been extensively applied to the healthy and diseased brain (Charil et al., 2007, Lerch and Evans, 2005, Shaw et al., 2006, Shaw et al., 2007, Singh et al., 2006). Compared to other surface extraction softwares, the CLASP algorithm has shown a good reproducibility and accuracy. Importantly, by avoiding surface self-intersection, CLASP provides the
Acknowledgments
This research was funded by a grant from the Canadian Institutes of Health research. B.C.B. is funded by the German National Merit Foundation and by the German Academic Exchange Service. The authors would like to thank Andrea Bernasconi for his insightful comments, and U. Yoon, O. Lyttelton, and C. Lepage for their help in the image processing.
References (123)
- et al.
Evidence for a direct projection from the superior temporal gyrus to the entorhinal cortex in the monkey
Brain Res.
(1983) - et al.
Voxel-based morphometry—the methods
NeuroImage
(2000) - et al.
Why voxel-based morphometry should be used
NeuroImage
(2001) - et al.
Whole-brain voxel-based statistical analysis of gray matter and white matter in temporal lobe epilepsy
NeuroImage
(2004) Propagation of human complex-partial seizures: a correlation analysis
Electroencephalogr. Clin. Neurophysiol.
(1991)- et al.
Functional anatomy of limbic epilepsy: a proposal for central synchronization of a diffusely hyperexcitable network
Epilepsy Res.
(1998) - et al.
Selective frontal, parietal, and temporal networks in generalized seizures
NeuroImage
(2003) - et al.
Voxel-based morphometry of the thalamus in patients with refractory medial temporal lobe epilepsy
NeuroImage
(2005) “Voxel-based morphometry” should not be used with imperfectly registered images
NeuroImage
(2001)- et al.
Does dysplasia cause anatomical dysconnectivity in schizophrenia?
Schizophr. Res.
(1998)