Hippocampal Atrophy Networks Mirror Morphological and Cognitive Compromise in Temporal Lobe Epilepsy
Abstract number :
2.197
Submission category :
5. Neuro Imaging / 5A. Structural Imaging
Year :
2023
Submission ID :
411
Source :
www.aesnet.org
Presentation date :
12/3/2023 12:00:00 AM
Published date :
Authors :
Presenting Author: Sara Larivière, PhD – Brigham and Women's Hospital / Harvard Medical School
Frederic Schaper, MD, PhD – Brigham and Women's Hospital / Harvard Medical School; Jessica Royer, PsyD – McGill University; Raul Rodriguez-Cruces, MD, PhD – McGill University; Ke Xie, MSc – McGill University; Jordan DeKraker, PhD – McGill University; Shahin Tavakol, MSc – McGill University; William Drew, BSc – Brigham and Women's Hospital / Harvard Medical School; Yifei Weng, MD, PhD – Nanjing University; Andrea Bernasconi, MD, PhD – McGill University; Neda Bernasconi, MD, PhD – McGill University; Luis Concha, MD, PhD – Universidad Nacional Autónoma de México; Zhiqiang Zhang, MD, PhD – Nanjing University; Birgit Frauscher, MD, PhD – McGill University; Boris Bernhardt, PhD – McGill University; Michael Fox, MD, PhD – Brigham and Women's Hospital / Harvard Medical School
Rationale:
Temporal lobe epilepsy (TLE) has long been associated with hippocampal atrophy. Some patients also show cortical atrophy, but it remains unclear how this cortical atrophy is related to hippocampal atrophy or clinical symptoms. We tested whether functional connectivity with locations of peak hippocampal atrophy in patients with TLE would identify cortical regions that are also atrophied or associated with clinical symptoms.
Methods:
We studied 94 individuals with drug-resistant unilateral TLE and 120 age- and sex-matched healthy controls from three independent sites. From the T1-weighted MRIs, we segmented the hippocampus, measured hippocampal thickness, and identified locations of peak hippocampal atrophy in each patient and control. We computed the brain network functionally connected to each participant’s hippocampal atrophy maps using human brain connectome data (n=1000; Fig. 1A), a technique known as “atrophy network mapping.” We then identified differences in atrophy network mapping between TLE and controls. Finally, atrophy network maps were related to cortical thickness measures, memory scores, and duration of epilepsy. Findings were corrected for multiple comparison using random field theory (pRFT).
Results:
Atrophied hippocampal vertices were heterogenous at the single-subject level, as no single vertex was atrophied in more than 5% of controls and 40% of patients (Fig. 1B). Whole-brain connectivity with the location of hippocampal atrophy differed between TLE and controls, with increased anticorrelation to the precentral gyrus and supplementary motor area, and increased positive connectivity to the superior parietal lobule, intraparietal sulcus, visual motion area, and fusiform gyrus in TLE (all pRFT < 0.001; Fig. 1C). Testing for morphological associations, areas of connectivity changes - predominantly areas of increased anticorrelation - were related to cortical atrophy in TLE (Fig. 1D). On the other hand, higher connectivity strength in areas of positive correlations were related to worse memory scores (Fig. 2A-C). Repeating our connectivity analysis in patients with prolonged epilepsy duration (more than 20 years) versus those with shorter duration (less than six years) further revealed trend-level effects in similar areas, suggesting an influence of hippocampal atrophy networks on morphological and cognitive compromise as the disease unfolds (Fig. 2D). Our findings were robust across a series of analytical methods and control analyses.
Conclusions:
Connectivity with sites of peak hippocampal atrophy identifies a cortical network that is atrophied and associated with clinical symptoms in patients with TLE.
Funding: CIHR
Neuro Imaging