ESTRO 2024 - Abstract Book

S940

Clinical - CNS

ESTRO 2024

2. Whybra, P. et al. Assessing radiomic feature robustness to interpolation in 18 F-FDG PET imaging. Scientific reports 2019. 9(1):9649. doi: 10.1038/s41598-019-46030-0.

3. www.spaarc-radiomics.io

4. Kolozsi, E. et al. Multiparametric MRI radiomics model to predict overall survival in Glioblastoma Multiforme, Radiother. Oncol., 2020 152, Supp 1: S847-8.

5. Zwanenburg, A., et al. "The Image Biomarker Standardization Initiative: Standardized Quantitative Radiomics for High-Throughput Image-based Phenotyping." Radiology 2020 May;295(2):328-338. doi: 10.1148/radiol.2020191145.

2688

Digital Poster

Molecular Signatures and Immune Cell Interactions in Primary and Recurrent Glioblastoma

Hyejo Ryu 1 , Bum-Sup Jang 1 , Joo Ho Lee 1 , Xue Li 1 , Chul-Kee Park 2 , Seok-Gu Kang 3

1 Seoul National University College of Medicine, Radiation Oncology, Seoul, Korea, Republic of. 2 Seoul National University College of Medicine, Neurosurgery, Seoul, Korea, Republic of. 3 Yonsei University College of Medicine, Neurosurgery, Seoul, Korea, Republic of

Purpose/Objective:

We aimed to elucidate the molecular signatures and immune cell interactions associated with primary and recurrent glioblastoma (GBM) tumors, with a focus on understanding GBM recurrence and its impact on prognosis.

Material/Methods:

We analyzed 52 matched pairs of primary and recurrent IDH wild-type glioblastoma bulk RNA sequencing data from two institutions, collected between 2008 and 2021. We revealed 687 differently expressed genes (DEGs) based on primary vs recurrent status, recurrent timing, and recurrent pattern. Integrating clinical factors including age, sex, surgery, chemotherapy, and radiotherapy into DEGs, we established random forest model and variable importance analysis to identify significant genes and pathways associated with the clinical outcome. Additionally, we identified differential interactions among detected immune cells by using the ‘CODEFACS and ‘LIRICS’ analysis pipeline.

Results:

Recurrent tumors exhibited enrichment of genes involved in the calcium signaling pathway and glial cell development compared to primary tumors. Notably, tumor-associated macrophage type 2 (TAM 2) showed a significant increase in recurrent tumors. Immune cell interactions involving TAM 1 were more prevalent in recurrent tumors. Stratifying the tumors based on relapse interval, we observed enrichment of the metabolic pathway in both