Authors :
Presenting Author: Ankeeta A., PhD – Thomas Jefferson University
Qirui Zhang, MD – Post-Doc Fellow, Department of Neurology, Thomas Jefferson University; Sam Javidi, PhD – Post-Doc Fellow, Department of Neurology, Thomas Jefferson University; Kapil Chaudhary, PhD – Post-Doc Fellow, Department of Neurology, Thomas Jefferson University; Salvatore Citro, MD – Department of Neurology – Thomas Jefferson University; Yolanda Kry, BA – Department of Neurology – Thomas Jefferson University; Michael Sperling, MD – Professor, Department of Neurology, Thomas Jefferson University; Joseph Tracy, PhD – Professor, Department of Neurology, Thomas Jefferson University
Rationale:
The mechanisms by which language laterality (LL) measures establish functional asymmetries in the brain are not understood. The following remain unclear: (1) are changes in the hemispheric asymmetry of language associated with changes in the asymmetry of gray matter (GM) structure, (2) do changes in LL affect the hemispheric representation of functions such as verbal memory, and (3) are LL indices customized to an individuals’ activation more sensitive to laterality changes than standard LL indices reliant are broad areas of cortex not active during language. This project addressed three key questions in the setting of temporal lobe epilepsy (TLE): (1) Are fMRI measures of LL associated with linked structural measures of GM asymmetry? (2) Are fMRI LL measures that capture activation unique to the individual better than standard LL measures at discriminating TLE from healthy controls (HCs), (3) Do asymmetries in LL predict asymmetries in the brain organization of other cognitive functions, such as memory?
Methods:
A total of 189 unilateral TLE patients (left=115; right=74) and matched HCs (115) underwent fMRI verb generation (VG) assessment for LL. Laterality indices (LIs: frontal, temporal, parietal) were calculated using the SPM Laterality Toolbox. Individualized LIs (ILIs) were created based on frontal, temporal, and parietal VG clusters. Structural laterality indices (SLIs) were derived from GM volume beneath the activation. Comparable fMRI verbal memory ILIs indices were calculated. Functional (ILI, SPM-LIs) and structural (SLIs) metrics were correlated. Discriminant function analysis determined the best TLE/HC discriminators; partial least squares (PLS) analyzed ILIs' association with fMRI memory task ILIs. This study was approved by the Thomas Jefferson University Human Subjects Research Committee.
Results:
Functional/structure associations were stronger for the ILIs/SLIs compared to the SPM LIs/SLIs, with these associations strongest in LTLE. The best discriminators of LTLE/HC were the ILI and SLI measures in the frontal and temporal lobes, with five of seven involving SLIs. There were few discriminators of RTLE/HC status, with two of three involving the SMA ILI and SLI measures. PLS models showed the SPM LIs did not reliably predict verbal memory laterality measures. In contrast, the ILIs and SLIs reliably predicted the verbal memory ILIs, explaining 47% of their variance.Conclusions:
Findings showed that individualized functional and linked-structural measures of LL differentiated TLE from HCs better than standard SPM measures. Individualized SLIs were the strongest discriminators. Correlations between ILIs and linked SLIs differed between our groups, indicating that coupled functional/structural measures that are individualized are sensitive to TLE pathology. ILI patterns predicted the reorganization of other cognitive functions such as verbal memory. These findings provide insights into the structural and functional mechanisms behind language laterality and cognitive reorganization in TLE.
Funding:
Joseph Tracy (PI) JIT (PI), NIH/NINDS, R01 NS112816-0.