Authors :
Presenting Author: Salvatore Citro, MD – Farber Institute for Neuroscience, Department of Neurology, Thomas Jefferson University, Philadelphia, PA
Presenting Author: Joseph Tracy, PhD – Thomas Jefferson University
Sam Javidi, PhD – Farber Institute for Neuroscience, Department of Neurology, Thomas Jefferson University, Philadelphia, PA; Ankeeta A, PhD – Farber Institute for Neuroscience, Department of Neurology, Thomas Jefferson University, Philadelphia, PA; Xiaosong He, PhD – University of Science and Technology of China, Department of Psychology, Hefei, Anhui, P.R. China; Qirui Zhang, MD – Farber Institute for Neuroscience, Department of Neurology, Thomas Jefferson University, Philadelphia, PA; Yolanda Kry, B.S. – Farber Institute for Neuroscience, Department of Neurology, Thomas Jefferson University, Philadelphia, PA; Michael Sperling, MD – Farber Institute for Neuroscience, Department of Neurology, Thomas Jefferson University, Philadelphia, PA; Joseph Tracy, PhD – Farber Institute for Neuroscience, Department of Neurology, Thomas Jefferson University, Philadelphia, PA
Rationale: Reorganization of language networks in epilepsy is well established, but the degree to which it is dependent upon corresponding structural pathways is not well known. In this project, we test whether alterations in the density of white matter streamlines are associated with task-fMRI patterns of functional language asymmetry, focusing on whether abnormal shifts in language are predicted by the underlying structural connectivity in right and left temporal lobe epilepsy patients (RTLE, LTLE), as well as matched healthy controls (HCs).
Methods: Forty-one LTLE, twenty-four RTLE, and fifty-one HCs underwent structural (high resolution Tl; diffusion tensor imaging) and functional acquisitions based on a verb generation fMRI task. Hemispheric and lobe-based frontal, temporal and parietal language indices (LI) were computed through the SPM toolbox. Structural measures involved extraction of lobe and hemisphere-based streamline density (LSD and HSD, respectively) measures involving 19 AAL regions for each hemisphere (Figure 1 A-D). Partial least squares (PLS) regression was performed to specify the streamline features most predictive of language laterality (latent factor variable scores >1) within each group.
Results: The three groups did not differ by age, handedness, gender, nor SPM LIs, with left hemisphere dominance prevalent in the sample. The PLS revealed two to three latent factors derived from SD reliably predicted language laterality within each group for both the LSD and HSD models (Figure 2A,2B). HCs showed PLS SD predictors of LIs with predominantly inverse relationships between LI and SD in the left hemisphere, and concordant LI/SD relationships in the right hemisphere. This indicated normative left hemisphere dominance was closely associated with increased SD amongst the left-sided language hubs. In LTLE, both LSD and HSD models revealed abnormal right frontal and left parietal LI/SD associations that suggested language reorganization (rightward shift) was accompanied by SD changes in those areas. Key language areas (e.g., Broca’s, Heschl’s) demonstrated abnormal rightward biases in LI that were associated with heightened SD. RTLE showed abnormal LI/SD associations in the left parietal, right frontal, and right temporopolar regions (Figure 2A, 2B). RTLE also lacked several LI/SD associations in the left temporal lobe present in HCs. These data suggested that even in the setting of right temporal disease and largely left hemisphere language dominance, LI/SD relationships are perturbed.
Conclusions: Functional language laterality was related to a selective set of structural connectivities in both TLE and HCs. Distinct language laterality/white matter structural connectivity associations were evident in all the three groups, with the associations generally weaker in the lobe housing epileptic pathology and stronger in the unaffected hemisphere. This suggested that epileptic pathology has structural impact outside the epileptogenic temporal lobe, contributing to the basis of reorganized language responses and dominance in the brain.
Funding: Joseph I Tracy (PI), NIH/NINDS, R01 NS112816-01