IDH Genotyping and Glioma Prognosis Research Based on an Interpretable Transformer Learning Framework

Abstract

Accurate genotyping and prognosis of glioma patients present significant clinical challenges, often dependent on subjective judgement and insufficient scientific evidence. This study aims to develop a robust, noninvasive preoperative multi-modal MRI-based transformer learning model to predict IDH genotyping and glioma prognosis. This multi-centre study included 563 glioma patients to develop an interpretable classification model utilising various preoperative imaging sequences, including T1-weighted, T2-weighted, fiuid-attenuated inversion recovery, contrast-enhanced T1-weighted, and diffusion-weighted imaging. The model employs a multi-task learning framework to extract and fuse radiomic, deep learning, and clinical features for IDH genotyping and glioma prognosis. Additionally, a multi-modal transformer strategy is integrated to analyse structural and functional MRI, thereby enhancing model performance. Experimental results indicate that the model demonstrates superior performance, surpassing previous research and other state-of-the-art methods. The model achieves an AUC of 91.40% in the IDH genotyping task and 93.37% in the glioma prognosis task. Group analysis reveals that the model exhibits higher sensitivity to IDH-mutant cases and more accurately identifies low-risk groups compared to medium-or high-risk groups. This study aims to achieve accurate IDH genotyping and glioma prognosis through effective classification method, offering valuable diagnostic insights for clinical practice and expediting treatment decisions.

Keywords

glioma prognosis / IDH genotyping / image analysis / image classification / multi-modal MRI / multi-task transformer learning

References

[1] M. Weller, P. Y. Wen, S. M. Chang, et al. “Glioma ” , Nature Reviews Disease Primers 10, no. 1 (2024): 33, https://doi.org/10.1038/s41572-024-00516-y.

[2] J. Bai, J. Varghese, and R. Jain, “Adult Glioma WHO Classification Update, Genomics, and Imaging: What the Radiologists Need to Know,” Topics in Magnetic Resonance Imaging 29, no. 2 (2020): 71-82, https://doi.org/10.1097/rmr.0000000000000234.

[3] M. J. Mair, M. Geurts, M. J. van den Bent, and A. S. Berghoff, “A Basic Review on Systemic Treatment Options in WHO Grade II-III Gliomas,” Cancer Treatment Reviews 92 (2021): 102124, https://doi.org/10.1016/j.ctrv.2020.102124.

[4] A. Poff, A. P. Koutnik, K. M. Egan, S. Sahebjam, D. D'Agostino, and N. B. Kumar, “Targeting the Warburg Effect for Cancer Treatment: Ketogenic Diets for Management of Glioma,” Seminars in Cancer Biology 56 (2019): 135-148, https://doi.org/10.1016/j.semcancer.2017.12.011.

[5] C. J. Pirozzi and H. Yan, “The Implications of IDH Mutations for Cancer Development and Therapy,” Nature Reviews Clinical Oncology 18, no. 10 (2021): 645-661, https://doi.org/10.1038/s41571-021-00521-0.

[6] K. Yang, Z. Wu, H. Zhang, et al., “Glioma Targeted Therapy: Insight Into Future of Molecular Approaches,” Molecular Cancer 21, no. 1 (2022): 39, https://doi.org/10.1186/s12943-022-01513-z.

[7] F. S. Varn, K. C. Johnson, J. Martinek, et al., “Glioma Progression Is Shaped by Genetic Evolution and Microenvironment Interactions,” Cell 185, no. 12 (2022): 2184-2199.e16, https://doi.org/10.1016/j.cell.2022.04.038.

[8] M. Smits, “MRI Biomarkers in Neuro-Oncology ” Nature Reviews Neurology 17, no. 8 (2021): 486-500, https://doi.org/10.1038/s41582-021-00510-y.

[9] A. Roux, P. Roca, M. Edjlali, et al., “MRI Atlas of IDH Wild-Type Supratentorial Glioblastoma: Probabilistic Maps of Phenotype, Man-agement, and Outcomes,” Radiology 293, no. 3 (2019): 633-643, https://doi.org/10.1148/radiol.2019190491.

[10] W. L. Bi, A. Hosny, M. B. Schabath, et al., “Artificial Intelligence in Cancer Imaging: Clinical Challenges and Applications,” CA: A Cancer Journal for Clinicians 69, no. 2 (2019): 127-157, https://doi.org/10.3322/caac.21552.

[11] G. Li, L. Li, Y. Li, et al., “An MRI Radiomics Approach to Predict Survival and Tumour-Infiltrating Macrophages in Gliomas,” Brain 145, no. 3 (2022): 1151-1161, https://doi.org/10.1093/brain/awab340.

[12] J. Cheng, J. Liu, H. Kuang, and J. Wang, “A Fully Automated Multimodal MRI-Based Multi-Task Learning for Glioma Segmentation and IDH Genotyping,” IEEE Transactions on Medical Imaging 41, no. 6 (2022): 1520-1532, https://doi.org/10.1109/tmi.2022.3142321.

[13] J. Guiot, A. Vaidyanathan, L. Deprez, et al., “A Review in Radio-mics: Making Personalized Medicine a Reality via Routine Imaging,” Medicinal Research Reviews 42, no. 1 (2022): 426-440, https://doi.org/10.1002/med.21846.

[14] T. Lin, Y. Wang, X. Liu and X. Qiu, “A Survey of Transformers,” AI Open 3 (2022): 111-132, https://doi.org/10.1016/j.aiopen.2022.10.001.

[15] J. Maynard, S. Okuchi, S. Wastling, et al., “World Health Organi-zation Grade II/III Glioma Molecular Status: Prediction by MRI Morphologic Features and Apparent Diffusion Coefficient,” Radiology 296, no. 1 (2020): 111-121, https://doi.org/10.1148/radiol.2020191832.

[16] K. Clark, B. Vendt, K. Smith, et al., “The Cancer Imaging Archive (TCIA): Maintaining and Operating a Public Information Repository,” Journal of Digital Imaging 26, no. 6 (2013): 1045-1057, https://doi.org/10.1007/s10278-013-9622-7.

[17] S. Bakas, H. Akbari, A. Sotiras, et al., “Advancing the Cancer Genome Atlas Glioma MRI Collections With Expert Segmentation La-bels and Radiomic Features,” Scientific Data 4, no. 1 (2017): 170117, https://doi.org/10.1038/sdata.2017.117.

[18] M. Ceccarelli, F. P. Barthel, T. M. Malta, et al., “Molecular Profiling Reveals Biologically Discrete Subsets and Pathways of Progression in Diffuse Glioma,” Cell 164, no. 3 (2016): 550-563, https://doi.org/10.1016/j.cell.2015.12.028.

[19] T. W. Macfarland and J. M. Yates, Mann-Whitney U Test (Springer International Publishing, 2016).

[20] R. Tibshirani, “Regression Shrinkage and Selection Via the Lasso,” Journal of the Royal Statistical Society: Series B 58, no. 1 (1996): 267-288, https://doi.org/10.1111/j.2517-6161.1996.tb02080.x.

[21] K. Simonyan and A. Zisserman, “Very Deep Convolutional Net-works for Large-Scale Image Recognition,” Computer Science (2014), https://doi.org/10.48550/arXiv.1409.1556.

[22] K. He, X. Zhang, S. Ren, and J. Sun, “Deep Residual Learning for Image Recognition,” in 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2016), 770-778, https://doi.org/10.1109/CVPR.2016.90.

[23] G. Huang, Z. Liu, L. Van Der Maaten, and K. Q. Weinberger, “Densely Connected Convolutional Networks,” in 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2017), 2261-2269, https://doi.org/10.1109/CVPR.2017.243.

[24] Sun Y., Dong L., Huang S., “Retentive Network: A Successor to Transformer for Large Language Models,” (2023), https://doi.org/10.48550/arXiv.2307.08621.

[25] B. Taha, T. Li, D. Boley, C. C. Chen, and J. Sun, “Detection of Isocitrate Dehydrogenase Mutated Glioblastomas Through Anomaly Detection Analytics,” Neurosurgery 89, no. 2 (2021): 323-328, https://doi.org/10.1093/neuros/nyab130.

[26] Y. S. Choi, S. Bae, J. H. Chang, et al., “Fully Automated Hybrid Approach to Predict the IDH Mutation Status of Gliomas Via Deep Learning and Radiomics,” Neuro-Oncology 23, no. 2 (2021): 304-313, https://doi.org/10.1093/neuonc/noaa177.

[27] A. Gao, H. Zhang, X. Yan, et al., “Whole-Tumor Histogram Analysis of Multiple Diffusion Metrics for Glioma Genotyping,” Radiology 302, no. 3 (2022): 652-661, https://doi.org/10.1148/radiol.210820.

[28] C. De Looze, A. Beausang, J. Cryan, et al., “Machine Learning: A Useful Radiological Adjunct in Determination of a Newly Diagnosed Glioma's Grade and IDH Status,” Journal of Neuro-Oncology 139, no. 2 (2018): 491-499, https://doi.org/10.1007/s11060-018-2895-4.

[29] Y. S. Choi, S. S. Ahn, J. H. Chang, et al., “Machine Learning and Radiomic Phenotyping of Lower Grade Gliomas: Improving Survival Prediction,” European Radiology 30, no. 7 (2020): 3834-3842, https://doi.org/10.1007/s00330-020-06737-5.

[30] H. G. Yoon, W. Cheon, S. W. Jeong, et al., “Multi-Parametric Deep Learning Model for Prediction of Overall Survival After Postoperative Concurrent Chemoradiotherapy in Glioblastoma Patients,” Cancers 12, no. 8 (2020): 2284, https://doi.org/10.3390/cancers12082284.

[31] L. Pasquini, A. Napolitano, M. Lucignani, et al., “AI and High-Grade Glioma for Diagnosis and Outcome Prediction: Do All Machine Learning Models Perform Equally Well?,” Frontiers in Oncology 11 (2021): 601425, https://doi.org/10.3389/fonc.2021.601425.

[32] J. Yan, B. Zhang, S. Zhang, et al., “Quantitative MRI-Based Radio-mics for Noninvasively Predicting Molecular Subtypes and Survival in Glioma Patients,” NPJ Precision Oncology 5, no. 1 (2021): 72, https://doi.org/10.1038/s41698-021-00205-z.

[33] L. Macyszyn, H. Akbari, J. M. Pisapia, et al., “Imaging Patterns Predict Patient Survival and Molecular Subtype in Glioblastoma Via Machine Learning Techniques,” Neuro-Oncology 18, no. 3 (2016): 417-425, https://doi.org/10.1093/neuonc/nov127.

[34] J. N. Mandrekar, “Receiver Operating Characteristic Curve in Diagnostic Test Assessment,” Journal of Thoracic Oncology 5, no. 9 (2010): 1315-1316, https://doi.org/10.1097/jto.0b013e3181ec173d.

[35] A. Lai, S. Kharbanda, W. B. Pope, et al., “Evidence for Sequenced Molecular Evolution of IDH1 Mutant Glioblastoma From a Distinct Cell of Origin,” Journal of Clinical Oncology 29, no. 34 (2011): 4482-4490, https://doi.org/10.1200/jco.2010.33.8715.

[36] Cancer Genome Atlas Research Network, D. J. Brat R. G. Verhaak K. D. Aldape, et al., “Comprehensive, Integrative Genomic Analysis of Diffuse Lower-Grade Gliomas,” New England Journal of Medicine 372, no. 26 (2015): 2481-2498, https://doi.org/10.1056/nejmoa1402121.

[37] G. Reyes-Botero, C. Dehais, A. Idbaih, et al., “Contrast Enhancement in 1p/19q-Codeleted Anaplastic Oligodendrogliomas Is Associated With 9p Loss, Genomic Instability, and Angiogenic Gene Expression,” Neuro-Oncology 16, no. 5 (2014): 662-670, https://doi.org/10.1093/neuonc/not235.

[38] M. Smits and M. J. van den Bent, “Imaging Correlates of Adult Glioma Genotypes,” Radiology 284, no. 2 (2017): 316-331, https://doi.org/10.1148/radiol.2017151930.

[39] N. Upadhyay and A. D. Waldman, “Conventional MRI Evaluation of Gliomas,” British Journal of Radiology 2, no. 2 (2011): S107-S111, https://doi.org/10.1259/bjr/65711810.

[40] M. P. G. Broen, M. Smits, M. M. J. Wijnenga, et al., “The T2-FLAIR Mismatch Sign as an Imaging Marker for Non-Enhancing IDH-Mutant, 1p/19q-Intact Lower-Grade Glioma: A Validation Study,” Neuro-Oncology 20, no. 10 (2018): 1393-1399, https://doi.org/10.1093/neuonc/noy048.

[41] S. Wu, X. Zhang, W. Rui, et al., “A Nomogram Strategy for Identi-fying the Subclassification of IDH Mutation and ATRX Expression Loss in Lower-Grade Gliomas,” European Radiology 32, no. 5 (2022): 3187-3198, https://doi.org/10.1007/s00330-021-08444-1.

[42] J. Yu, Z. Shi, Y. Lian, et al., “Noninvasive IDH1 Mutation Estima-tion Based on a Quantitative Radiomics Approach for Grade II Glioma,” European Radiology 27, no. 8 (2017): 3509-3522, https://doi.org/10.1007/s00330-016-4653-3.

[43] Y. Bai, Y. Lin, J. Tian, et al., “Grading of Gliomas by Using Monoexponential, Biexponential, and Stretched Exponential Diffusion-Weighted MR Imaging and Diffusion Kurtosis MR Imaging,” Radiology 278, no. 2 (2016): 496-504, https://doi.org/10.1148/radiol.2015142173.

[44] M. I. de la Fuente, R. J. Young, J. Rubel, et al., “Integration of 2-hydroxyglutarate-Proton Magnetic Resonance Spectroscopy Into Clin-ical Practice for Disease Monitoring in Isocitrate Dehydrogenase-Mutant Glioma,” Neuro-Oncology 18, no. 2 (2016): 283-290, https://doi.org/10.1093/neuonc/nov307.

[45] B. Suchorska, A. Giese, A. Biczok, et al., “Identification of Time-to-Peak on Dynamic 18F-FET-PET as a Prognostic Marker Specifically in IDH1/2 Mutant Diffuse Astrocytoma,” Neuro-Oncology 20, no. 2 (2018):279-288, https://doi.org/10.1093/neuonc/nox153.

[46] J. Zhao, Y. Huang, Y. Song, et al., “Diagnostic Accuracy and Po-tential Covariates for Machine Learning to Identify IDH Mutations in Glioma Patients: Evidence From a Meta-Analysis,” European Radiology 30, no. 8 (2020): 4664-4674, https://doi.org/10.1007/s00330-020-06717-9.

[47] M. Li, X. Ren, X. Chen, et al., “Combining Hyperintense FLAIR Rim and Radiological Features in Identifying IDH Mutant 1p/19q Non-Codeleted Lower-Grade Glioma,” European Radiology 32, no. 6 (2022): 3869-3879, https://doi.org/10.1007/s00330-021-08500-w.

[48] M. Mladenovsk, I. Valkov, M. Ovcharov, N. Vasilev, and I. Duh-lenski, “High Grade Glioma Surgery—Clinical Aspects and Prognosis,” Folia Medica 63, no. 1 (2021): 35-41, https://doi.org/10.3897/folmed.63.e55255.

[49] A. Chattopadhay, A. Sarkar, P. Howlader, and V. N. Balasu-bramanian, “Grad-Cam++: Generalized Gradient-Based Visual Expla-nations for Deep Convolutional Networks,” in Proceedings of the 2018 IEEE Winter Conference on Applications of Computer Vision (WACV) (2018), 839-847.

Funding

National Natural Science Foundation of China(Grants 82441022)

National Natural Science Foundation of China(82371934)

Medical Science and Technology Research Project of Henan Province(SBGJ202101002)

Joint Fund of Henan Province Science and Technology R&D Programme(225200810062)

Henan Provincial Medical Science and Technology Research Joint Construction Project(LHGJ20240053)

Henan Provincial Medical Science and Technology Research Joint Construction Project(LHGJ20240036)