Radiogenomic approach to glial tumors imaging under conditions of initial diagnostic measures: adaptation principles development

Introduction. Radiomics is a rapidly developing field in oncology visualization aimed at searching for prognostically effective imaging features associated with specific genetic events that determine various characteristics of the disease course. According to numerous studies, the presence of IDH mutations in glial tumors determines a longer overall survival. Despite the fact that biopsy is considered to be the «gold standard» for brain tumors differential diagnosis, it is though quite difficult to perform due to the complexity of surgical access, common cases of the repeat procedure impossibility, serious complications and mortality.

Aim: a search for imaging features providing prognostic data on the presence of certain mutations and gene expression in gliomas, obtained using traditional pulse sequences and characterized by the absence of restrictions on applicability depending on the tumors visible morphological features.

Material and methods: retrospective analysis of 49 eligible patients' primary brain MRI data between 2021 and 2023 from Almazov National Medical Research Centre (n = 31) and Napalkov Oncological Centre (n = 18) with glial tumors and subsequently identified status of the target variable; preprocessing of MR images using the histogram matching; regions of interest determination and semi-automated slice-by-slice segmentation with subsequent extraction of radiomics features; search for predictive radiomics features regarding the status of target variable using statistical analysis tools.

Results. Dependence Entropy was found to be highly effective as a predictor of IDH mutations (area under the ROC-curve – 0.766 [0.627–0.880]).

Conclusions. We determined a target variable for the development of a predictive model (IDH status), a pulse sequence (T2-Tirm), a tool for initial imaging data preprocessing (histogram matching), regions of interest (tumor-associated T2-Tirm-hyperintensity including cystic and/or necrotic lesions). As a result, a statistically significant relationship between the Dependence Entropy feature and IDH status of glial tumors was found. In the course of further work it is planned to increase the size of a database, improve the accuracy of the existing statistical model, search for relevant radiomic features extracted using other traditional pulse sequences, create a comprehensive predictive radiogenomics model and develop a software.

1. 1 Di Bonaventura R, Montano N, Giordano M, et al. Reassessing the Role of Brain Tumor Biopsy in the Era of Advanced Surgical, Molecular, and Imaging Techniques-A Single-Center Experience with Long-Term Follow-Up. J Pers Med. 2021;11(9):909. Published 2021 Sep 12. https://doi.org/10.3390/jpm11090909

2. 2 Kelly PJ, Daumas-Duport C, Kispert DB, Kall BA, Scheithauer BW, Illig JJ. Imaging-based stereotaxic serial biopsies in untreated intracranial glial neoplasms. J Neurosurg. 1987;66(6):865-874. https://doi.org/10.3171/jns.1987.66.6.0865

3. 3 Lin K, Cidan W, Qi Y, Wang X. Glioma grading prediction using multiparametric magnetic resonance imaging-based radiomics combined with proton magnetic resonance spectroscopy and diffusion tensor imaging. Med Phys. 2022;49(7):4419-4429. https://doi.org/10.1002/mp.15648

4. 4 Han S, Liu Y, Cai SJ, et al. IDH mutation in glioma: molecular mechanisms and potential therapeutic targets. Br J Cancer. 2020;122(11):1580- 1589. https://doi.org/10.1038/s41416-020-0814-x

5. 5 Foltyn M, Nieto Taborda KN, Neuberger U, et al. T2/FLAIR-mismatch sign for noninvasive detection of IDH-mutant 1p/19q non-codeleted gliomas: validity and pathophysiology. Neurooncol Adv. 2020;2(1):vdaa004. Published 2020 Jan 10. https://doi.org/10.1093/noajnl/vdaa004

6. 6 Patel SH, Batchala PP, Muttikkal TJE, et al. Fluid attenuation in non-contrast-enhancing tumor (nCET): an MRI Marker for Isocitrate Dehydrogenase (IDH) mutation in Glioblastoma. J Neurooncol. 2021;152(3):523-531. https://doi.org/10.1007/s11060-021-03720-y

7. 7 Zinn PO, Mahajan B, Sathyan P, et al. Radiogenomic mapping of edema/cellular invasion MRI-phenotypes in glioblastoma multiforme [published correction appears in PLoS One. 2012;7(2). https://doi.org/10.1371/annotation/b5267cb3-6aa7-47fc-a648-47f30a7cff3e.

8. 8 Majadan, Bhanu [corrected to Mahajan, Bhanu]]. PLoS One. 2011;6(10):e25451. https://doi.org/10.1371/journal.pone.0025451

9. 9 Aghi M, Gaviani P, Henson JW, Batchelor TT, Louis DN, Barker FG 2nd. Magnetic resonance imaging characteristics predict epidermal growth factor receptor amplification status in glioblastoma. Clin Cancer Res. 2005;11(24 Pt 1):8600-8605. https://doi.org/10.1158/1078-0432.CCR-05-0713

10. 10 Diehn M, Nardini C, Wang DS, et al. Identification of noninvasive imaging surrogates for brain tumor gene-expression modules. Proc Natl Acad Sci U S A. 2008;105(13):5213-5218. https://doi.org/10.1073/pnas.0801279105

11. 11 Tymchuk AI. Textural signs in the problem of segmentation of aerial photographs based on luminance dependence matrices. Cybernetics and programming. 2018;6:31-39. (In Russ). https://doi.org/10.25136/2644-5522.2018.6.28395