Rearrangement of the posterior language area with lesion progression

28 

Abstract Submission 

Riho Nakajima¹, Takahiro Osada², Masashi Kinoshita³, Akitoshi Ogawa², Hirokazu Okita⁴, Seiki Konishi², Mitsutoshi Nakada⁵

¹Department of Occupational therapy, Kanazawa University, Kanazawa, Japan, ²Department of Neurophysiology, Juntendo University School of Medicine, Tokyo, Japan, ³Department of Neurosurgery,Kanazawa University, Kanazawa, Japan, ⁴Department of Physical Medicine and Rehabilitation, Kanazawa University Hospital, Kanazawa, Japan, ⁵Department of Neurosurgery, Kanazawa University, Kanazawa, Japan

Riho Nakajima  
Department of Occupational therapy, Kanazawa University
Kanazawa, Japan

Takahiro Osada  
Department of Neurophysiology, Juntendo University School of Medicine
Tokyo, Japan

Masashi Kinoshita  
Department of Neurosurgery,Kanazawa University
Kanazawa, Japan

Akitoshi Ogawa  
Department of Neurophysiology, Juntendo University School of Medicine
Tokyo, Japan

Hirokazu Okita  
Department of Physical Medicine and Rehabilitation, Kanazawa University Hospital
Kanazawa, Japan

Seiki Konishi  
Department of Neurophysiology, Juntendo University School of Medicine
Tokyo, Japan

Mitsutoshi Nakada  
Department of Neurosurgery, Kanazawa University
Kanazawa, Japan

Damaged functional areas are compensated by other cortical regions to maintain brain function. Although cortical rearrangement can occur in either hemisphere when the tumor grows to the posterior language area (PLA), the rearrangement of Wernicke's area has rarely been reported.1, 2) This study investigated the characteristics of cortical rearrangement of the PLA within the ipsilateral hemisphere using direct electrical stimulation (DES) during awake brain surgery and resting-state functional connectivity.

Sixty-two patients with left hemispheric glioma who underwent awake brain surgery were included in the study. All the patients underwent structural magnetic resonance imaging (MRI) and resting state functional MRI (rsfMRI) preoperatively. Each structural MRI was spatially normalized to the Montreal Neurological Institute (MNI) template via the segmentation routine using SPM12. Patients were divided into PLA and non-PLA groups based on lesion extension to the PLA. Cortical brain mapping was performed using DES. During intraoperative language assessment, patients were asked to name the pictures presented on the screen. When incorrect (impaired) responses were elicited in the naming task during DES, the point was considered an area with a language function (positive points). The region was considered an area with a non-language function (normal points) if the response was correct (unimpaired). We defined incorrect responses as "positive responses" and correct responses as "normal responses." Spatial locations of the positive and normal responses were retrospectively plotted on the corresponding original 3D T1 images for each patient using operative reports and intraoperative video records with the iPlan software. Each positive point on the original T1 images was transferred to the corresponding point on the normalized T1 images. The positive points were then mapped onto the corresponding lattice cells on the lateral view of the brain. The frequency of the positive responses was calculated as the ratio of the number of patients with positive responses to the number of stimulated patients. RsfMRI analysis was conducted using SPM12, FSL, and the Human Connectome Project (HCP) pipelines. We calculated the betweenness centrality (BC) among the 360 cerebrocortical parcels,3) which were averaged across the vertices in each parcel from HCP.4) The frequencies of positive responses were compared between the groups using the chi-square test. One- and two-way analysis of variance (ANOVA) was performed to examine the BC characteristics. This study was approved by the Medical Ethics Committee of Kanazawa University (approval numbers: 1797 and 3322).

In awake language mapping, 48 positive and 488 normal points were identified. In the non-PLA group, areas with a high proportion of positive responses were detected in the posterior part of the superior temporal gyrus (pSTG) and middle temporal gyrus, while no positive responses were detected in the supramarginal gyrus (SMG). In the PLA group, a high proportion of positive responses were distributed widely in the cortical area, including the pSTG and temporoparietal junction. The proportion of positive responses in the posterior SMG (pSMG) was significantly higher in the PLA group (47%) than in the non-PLA group (0%) (P=0.0091). RsfMRI revealed that BC in the pSMG was significantly higher than that in the surrounding areas in both the non-PLA and PLA groups (t(41)=4.95, P=8.73 × 10-6; t(19)=3.09, P=0.0047, respectively). Finally, we performed a mixed ANOVA with the areas (pSTG/pSMG) and patient groups (non-PLA/PLA groups) as the main effects. A significant interaction was observed between the areas and the patient groups (F(1,38)=6.50, P=0.013).

This study suggests that the posterior language area can be rearranged from pSTG to pSMG, which is a hub region in the brain, with lesion progression to the posterior superior temporal gyrus.

Direct Electrical/Optogenetic Stimulation ¹

Language Other ²

Cortex

Language

Plasticity

Other - glioma, awake surgery, posterior language area, resting state functional MRI