Structural Effects of Right-hemispheric Language-reorganization after Perinatally Acquired Left-hemispheric Stroke

Background: Language is one of the most lateralized brain functions, and cortical language areas are similarly asymmetrical. In a previous study, we presented evidence that perinatal left-hemispheric stroke is usually associated with functionally successful right-hemispheric language reorganization. It remains unclear, however, whether language-reorganization is associated with anatomical differences in the reorganized hemisphere.

Hypothesis: Exploratory voxel-wise gray matter comparison of patients with right-hemispheric language-reorganization after perinatally acquired left hemispheric stroke to normal controls with left-hemispheric language-organization limited to the right hemisphere.

Methods: Prospective data collection of 8 patients (age 9 to 26, 3 female) with perinatally acquired left-hemispheric brain lesions as well as 29 normally developing controls with left-hemispheric language-organization (age 8 to 29, 14 female). We used a Siemens 1.5 Avanto scanner to acquire whole brain functional echoplanar images while patients and controls performed the vowel identification task, reflecting language production. Anatomical images were acquired in the form of a T1-weighted three-dimensional dataset (1x1x1 mm³). Functional and anatomical data was preprocessed and analyzed with SPM12, CAT12 and custom scripts. We determined the frontal lateralization indices and performed a voxel-wise comparison of the right hemisphere for gray matter group differences including age and sex as covariates. For the resulting significance clusters we additionally calculated the mean hemispheric gray matter volume in homotopic regions of the left hemisphere.

Results: Six of eight patients showed right-hemispheric language reorganization (LI -0.66; [-0.89 to -0.34]) whereas the 29 controls had an evident left-dominance (LI 0.85; [0.62 to 0.96]). There was one controls>patients cluster in right-hemispheric gray matter indicating a significant group difference (cluster forming threshold p = 0.001, extent threshold pFEW = 0.05). More specifically, the difference cluster was situated in the medial temporal gyrus with a cluster-size of 795 voxels and mean cluster-specific gray matter volumes of 1.81 [1.45–2.15] vs. 1.37 cm³ [1.18–1.57] (RH) and 1.34 [1.02–1.67] vs. 0.89 [0–1.29] cm³ (LH). Hence, gray matter volume of this right-hemispheric region in reorganized patients (1.37 cm³) corresponds to the homotopic left-hemispheric region in left-dominant controls (1.34 cm³) and is significantly smaller (p < 0.001, Mann-Whitney) compared to the right hemisphere in controls (1.81 cm³).

Discussion: Right-hemispheric language-reorganization seems to lead to a similar cortical volume in a posterior-temporal area as in homotopic regions of the left hemisphere in left-dominant controls. The functional relevance of these findings will be investigated further.

No conflict of interest has been declared by the author(s).