Exploring the Relationship between Brain Tumors and Accelerated Aging by Disconnection Maps

2300 

Abstract Submission 

Ying-Hong Yao¹, Shin Tai Chong¹, Chih-Chin Heather Hsu¹, Chen-Yuan Kuo¹, Xinrui Liu², Ching-Po Lin^1,3

¹Institute of Neuroscience, National Yang Ming Chiao Tung University, Taipei, Taiwan, ²Department of Neurosurgery, First Hospital of Jilin University, Changchun, Jilin, ³Department of Education and Research, Taipei City Hospital, Taipei, Taiwan

Ying-Hong Yao  
Institute of Neuroscience, National Yang Ming Chiao Tung University
Taipei, Taiwan

Shin Tai Chong  
Institute of Neuroscience, National Yang Ming Chiao Tung University
Taipei, Taiwan

Chih-Chin Heather Hsu  
Institute of Neuroscience, National Yang Ming Chiao Tung University
Taipei, Taiwan

Chen-Yuan Kuo  
Institute of Neuroscience, National Yang Ming Chiao Tung University
Taipei, Taiwan

Xinrui Liu  
Department of Neurosurgery, First Hospital of Jilin University
Changchun, Jilin

Ching-Po Lin  
Institute of Neuroscience, National Yang Ming Chiao Tung University|Department of Education and Research, Taipei City Hospital
Taipei, Taiwan|Taipei, Taiwan

Brain tumors significantly disrupt neural network integrity, leading to both local and distal functional impairments due to compromised cerebral communication pathways. Utilizing diffusion Magnetic Resonance Imaging tractography, researchers can investigate the impact of brain tumors on white matter (WM) tracts and assess structural 'disconnectome' [1]. However, these approaches are hindered by prolonged scanning durations and intricate post-processing requirements, necessitating substantial technical expertise and time investment. To overcome these limitations, the incorporation of lesion mapping methodologies, in tandem with predefined WM and gray matter atlases, has emerged as a streamlined alternative, enhancing the efficiency and effectiveness of population-level investigations. Prior studies employing lesion mapping have focused on the structural disconnection-cognitive impairment nexus in cases of brain tumors, stroke, and epilepsy [2]. Building upon this foundation, the current study integrates WM disconnection maps with a brain age prediction model to delve into the relationship between tumor-induced localized damage and distant network disconnections. The primary aim is to elucidate the effects of localized tumor lesions on accelerated aging across cerebral hemispheres, thereby contributing to a more comprehensive understanding of the global ramifications of neural network disruptions caused by brain tumors.

In this study, we analyzed structural imaging data from 6 brain tumor patients, aged 27 to 68 at Jilin University Hospital. Each patient had undergone preoperative 3D T1-weighted (T1w) imaging with the following parameter: TR/TE/T1 3500/2.3/1100 ms, voxel size 1×1×1 mm³, and FOV 256×256x176 mm². First, we performed manual segmentation with ITK-SNAP, followed by validation from a 20-years experienced neurosurgeon. Lesion segmentation maps were registered to the MNI152 space using the ANTs. Next, we applied the Lesion Quantification Toolkit in conjunction with the HCP-842 population-averaged streamline tractography atlas and AAL2 brain parcellation template to quantify the WM disruption [3-5]. The disconnections were measured by assessing parcel-wise disconnection severity matrices, encapsulating the pattern of WM disconnections between pairs of grey matter parcels (Fig. 1A) [6]. For brain age prediction, we utilized a support vector regression model with a radial basis function kernel implemented through the Scikit-learn library (Fig. 1B) [7]. We strategically excluded brain regions directly impacted by tumors, and then compared the predicted age discrepancies between the disrupted areas and their corresponding healthy contralateral counterparts, as well as with bilaterally healthy regions (Fig. 1C). The Wilcoxon-Mann-Whitney U-test was used for the statistical comparison, considering p-values less than 0.05 as indicative of statistical significance.

We analyzed 6 patients with brain tumor, the detail of clinical information and the mean values of brain age differences of each patient were summarized in Fig. 2A. Fig 2B represents the case of patient P01, displaying the lesion mask overlaid on T1w in the MNI152 space, along with computed voxel-wise and parcel wise-disconnection map. We found a significant increase (p<0.03) in the brain age differences within disrupted brain regions, averaging 17.76±5.21 years, compared to the healthy brain regions, where the difference was 10.64±2.95 years (Fig. 3C).

Our study revealed that brain lesions not only affect local areas but also cause WM disconnections, impacting distant regions within and across brain hemispheres. By integrating WM disconnection maps with a brain age prediction model and successfully highlighted significant age disparities between the hemispheres. The efficient use of lesion mapping in our study provides valuable insights into the impact of brain tumors, potentially enhancing surgical planning and treatment strategies alongside cognitive assessments.

Image Registration and Computational Anatomy

White Matter Anatomy, Fiber Pathways and Connectivity ²

Anatomical MRI ¹

MRI

Other - Brain tumor, disconnectome, lesion mapping, brain age prediction