Statistical Comparisons of Individual Glioblastoma Samples Can Reveal Molecular Subtypes with Clinical Phenotypes
Background: Glioblastoma (GBM) is a malignant tumor of astrocytes that occurs in approximately 3.22 per 100,000 people1. Unfortunately, the underlying pathophysiology is poorly understood and physicians lack targeted treatments to offer patients1. In recent years, there has been a push to further classify GBM into discrete subtypes that harbor similar mutations or respond similarly to treatments1. Currently, there are only two subtypes of GBM that are widely accepted to have clinical significance3,6. These are MGMT methylation status as well as IDH1 methylation status3,6. To further elucidate clinically distinct subtypes, researchers have begun to turn to statistical analyses to harness the power of big datasets2-8. These analyses use a variety of statistical tests to help identify relationships between patients, mutations, and clinical outcomes2-8. This literature review seeks to explore the specific statistical tests used, as well as clinical phenotypes that correspond to the identified subtype2-8.
Objectives: To explore the statistical tests being utilized to further subclassify GBM. The identified subtypes will also be examined with an emphasis on clinical relevance.
Search Methods: Pubmed was used to query publications ranging from 2017 – 2023. Search terms included “Glioblastoma”, “Statistical”, “Pan-Cancer”, “Computational”, “Biomarkers”, “Classification”, and “Bioinformatics”.
Results: Both pan-cancer and cancer specific analyses can be used to identify important subtypes in GBM2-8. The most common statistical tests used were T-tests, Mann-Whitney, Cox’s Regression, and Kaplan-Meir2-8. These statistical tests were able to identify important subtypes in GBM. The first subtype is PARP17. GBM tumors that include this mutation were found to be correlated with increased microsatellite instability7. This is clinically relevant because microsatellite has been associated with increased response to immunotherapies2. The second subtype is PDC1 mutation status8. Contrary to other cancer types, GBM had a worse prognosis with a positive mutation status8. This result warrants further research, but could possibly be due to the status of the brain as an immune privileged site8. The next subtype is a six gene expression signature that is able to accurately classify GBM patients as high risk or low risk5. The underlying pathology of this signature is undergoing investigation, but many of the genes have been shown to have roles in DNA repair5. The last subtype is a novel MGMT fusion3,6. The role of this fusion’s impact on prognosis is its ability to repair DNA damage3,6. This means that any damage inflicted by chemotherapy would be rapidly repaired by tumors harboring this fusion3,6.
Conclusions: Computational and statistical techniques are invaluable tools in discovering populations of GBM patients that may respond well to specific therapies1-8. While many of these tumor populations require further research into their underlying pathology, the clinical relevance lies in a physician’s ability to know which patients require a more aggressive therapy1-8. Future research lies in larger datasets as well as clinical trials proving increased survival when stratifying patients by these subtypes.
- Wen PY, Weller M, Lee EQ, et al. Glioblastoma in adults: a Society for Neuro-Oncology (SNO) and European Society of Neuro-Oncology (EANO) consensus review on current management and future directions. Neuro Oncol. 2020;22(8):1073-1113. doi:10.1093/neuonc/noaa106
- Sareen H, Ma Y, Becker TM, Roberts TL, de Souza P, Powter B. Molecular Biomarkers in Glioblastoma: A Systematic Review and Meta-Analysis. Int J Mol Sci. 2022;23(16):8835. doi:10.3390/ijms23168835
- Butler M, Pongor L, Su YT, et al. MGMT Status as a Clinical Biomarker in Glioblastoma. Trends Cancer. 2020;6(5):380-391. doi:10.1016/j.trecan.2020.02.010
- Zhou L, Tang H, Wang F, et al. Bioinformatics analyses of significant genes, related pathways and candidate prognostic biomarkers in glioblastoma. Mol Med Rep. 2018;18(5):4185-4196. doi:10.3892/mmr.2018.9411
- Zuo S, Zhang X, Wang L. A RNA sequencing-based six-gene signature for survival prediction in patients with glioblastoma. Sci Rep. 2019;9(1):2615. doi:10.1038/s41598-019-39273-4
- Oldrini B, Vaquero-Siguero N, Mu Q, et al. MGMT genomic rearrangements contribute to chemotherapy resistance in gliomas. Nat Commun. 2020;11(1):3883. doi:10.1038/s41467-020-17717-0
- Zhang X, Wang Y, A G, Qu C, Chen J. Pan-Cancer Analysis of PARP1 Alterations as Biomarkers in the Prediction of Immunotherapeutic Effects and the Association of Its Expression Levels and Immunotherapy Signatures. Front Immunol. 2021;12:721030. doi:10.3389/fimmu.2021.721030
- Miao Y, Wang J, Li Q, et al. Prognostic value and immunological role of PDCD1 gene in pan-cancer. Int Immunopharmacol. 2020;89(Pt B):107080. doi:10.1016/j.intimp.2020.107080