Toward Precision Oncology in Glioblastoma with a Personalized Cancer Genome Reporting Tool and Genetic Changes Identified by Whole Exome Sequencing

Abstract

Precision/personalized medicine in oncology has two key pillars: molecular profiling of the tumors and personalized reporting of the results in ways that are clinically contextualized and triangulated. Moreover, neurosurgery as a field stands to benefit from precision/personalized medicine and new tools for reporting of the molecular findings. In this context, glioblastoma (GBM) is a highly aggressive brain tumor with limited treatment options and poor prognosis. Precision/personalized medicine has emerged as a promising approach for personalized therapy in GBM. In this study, we performed whole exome sequencing of tumor tissue samples from six newly diagnosed GBM patients and matched nontumor control samples. We report here the genetic alterations identified in the tumors, including single nucleotide variations, insertions or deletions (indels), and copy number variations, and attendant mutational signatures. Additionally, using a personalized cancer genome-reporting tool, we linked genomic information to potential therapeutic targets and treatment options for each patient. Our findings revealed heterogeneity in genetic alterations and identified targetable pathways, such as the PI3K/AKT/mTOR pathway. This study demonstrates the prospects of precision/personalized medicine in GBM specifically, and neurosurgical oncology more generally, including the potential for genomic profiling coupled with personalized cancer genome reporting. Further research and larger studies are warranted to validate these findings and advance the treatment options and outcomes for patients with GBM.

Introduction

Glioblastoma (GBM) remains a highly aggressive adult brain tumor, characterized by extensive inter/intratumor heterogeneity (DeCordova et al., 2020; Perrin et al., 2019; Soeda et al., 2015; Tirosh and Suvà, 2020). Despite an increasing understanding of the underlying pathophysiology and the use of combined chemotherapy and radiotherapy after maximal safe resection as standard therapy, GBM remains an incurable disease with high mortality. In patients treated with standard therapy, the median survival is only 15 months; less than 25% of patients survive up to 2 years, and the 5-year survival rate after initial diagnosis is less than 10% (Ostrom et al., 2017). Whether the prognosis of patients with recurrent GBM is better or similar to that of patients with primary GBM remains controversial.

New therapeutic candidates have shown promise in preclinical and early phase studies and have allowed several clinical trials to be conducted; however, there have been no significant improvements in therapy to date. Moreover, less than 10% of GBM patients participate in prospective clinical trials (Yabroff et al., 2012). Considering the numerous biological differences within individual tumor cells, the one-size-fits-all treatment strategy is not successful in tumors with large inter/intratumor heterogeneity, such as GBM (Prados et al., 2015).

In recent years, the elucidation of the genomic profiles of cancers has enabled the development of biomarker-driven therapies in cancer treatment and has greatly improved tumor prognosis in many different cancers. However, this is not yet the case for GBM.

Genome sequencing of GBM tumors has identified several genetic alterations, some of which have been shown to be clinically relevant and to drive GBM pathology and recurrence. These include mutations in the IDH1, IDH2, TP53, NF1, ERBB2, PTN, PIK3R1, CDK4, TERT, and MDM2 genes, loss of chromosome arm 10q, amplifications of the growth factor receptors EGFR and PDGFR, and aberrations in the RTK/Ras/PI3K signaling pathways (Turkalp et al., 2014; Verhaak et al., 2010). Based on mutations observed in the isocitrate dehydrogenase genes (IDH1 and IDH2), GBM tumors have been classified into two subtypes that are histologically indistinguishable but differ in their clinical features (Ohgaki and Kleihues, 2007). In primary GBM cases, 90% are of the IDH-wild-type, whereas secondary GBM have mutations in IDH. In addition, extensive studies of genomic, epigenetic, and transcriptomic profiles provided by The Cancer Genome Atlas (TCGA) have identified several molecular subtypes for GBMs with distinct prognosis and different response to therapies (Brennan et al., 2013; McLendon et al., 2008).

The perspective of precision medicine in GBM has become an important necessity due to limited treatment options and clinical outcomes. Therefore, more comprehensive molecular profiling of tumor tissue is required for each GBM patient. It is now standard practice in many leading academic hospitals to perform in-depth molecular profiling of the tumor and potentially administer therapeutic drugs for these specific genetic alterations (Mahlokozera et al., 2018).

In this study, we performed whole exome sequencing (WES) of tumor tissue samples from the temporal, frontal, or occipital lobes and matched nontumor samples from six newly diagnosed GBM patients. We identified single nucleotide variations (SNVs), insertions or deletions (indels), copy number variations (CNVs), microsatellite instability (MSI) prediction, and mutational signatures in the tumor and evaluated the results from a precision oncology perspective by generating personalized reports that linked genomic information to potential therapeutic options. This article presents the promising applications of precision/personalized medicine for patients with GBM.

Materials and Methods

Tumor samples

The procedures and protocols of the study were approved by and in accordance with the rules of the study site at Marmara University School of Medicine Ethics Committee (Approval number: 2019-891). Six adults with newly diagnosed, nontherapy-naive GBM patients were enrolled in the study. Written informed consent was obtained from all participants before the study. Only patients with noneloquent tumor location were selected for the study. For GBM that did not present on the surface of the brain, pia-arachnoid was coagulated and incised, and a shortest transcortical route was used to reach tumor mass. Tumor tissue was harvested during routine surgery from the cell-rich areas of the temporal, frontal, or occipital lobes. Normal (nontumorous) control brain tissue from the surgical corridor was also collected as comparative material for each patient. Tissues were stored in liquid nitrogen at −196°C. All tumor samples were examined by a neuropathologist, and all tumor samples were diagnosed as IDH wild-type GBM.

DNA extraction and WES

Genomic DNA samples from the tumor and matching nontumorous samples of each patient were obtained using QIAmp^® genomic DNA extraction kit (Qiagen, Hilden, Germany) with optimized protocol. The quality of genomic DNA was determined by estimating A260/A280 and A260/A230 ratios using Nanodrop (Thermo Fisher Scientific, Waltham, MA, USA) both >1.8, and by 1% agarose gel electrophoresis for each sample. All samples that met the thresholds for library preparation were sequenced using the NovaSeq6000 platform (Illumina, Inc.) and the IDT xGEN Hyb panel (Integrated DNA Technologies, Inc.). In addition to the exome panel, the GOAL probes were added during enrichment hybridization (Kuo et al., 2020). This method aims to generate a regular exome with tumor depth (∼150 × ) along with amplified coverage (∼500 × ) for 527 cancer-related genes, GOAL consortium fusion regions, relevant viral sequences, MSI probes, and inherited cancer genes. The protocol yielded single-end reads with a read length of 150 bp for each of the matched normal and tumor in the cohort.

Somatic variant discovery

FASTQC (v.0.11.9) (Andrews, 2010) was used for quality control of the raw Fastq data. Reads were aligned to the reference genome GRCh38.p13 using the default mode of Burrows–Wheeler Aligner (BWA) (Li, 2013) and Picard (version 2.26.8) (Broad Institute). Before variant detection, the data were preprocessed using GATK preprocessing tools (version 4.2.6.1), for instance, by marking duplicates, sorting the SAM file by coordinates, and filtering by coverage (> 30) to correct technical biases and make the data suitable for further analysis with Picard (version 2.26.8) and GATK (version 4.2.6.1) (McKenna et al., 2010).

Somatic SNVs and indels were called by using Mutect2 (included in GATK) for each sample using tumor-normal mode where a tumor sample is matched with a normal sample (Cibulskis et al., 2013). As recommended by GATK, FilterMutectCalls was used to further filter the raw output of Mutect2 for poor mapping quality, local haplotypes with too many variants, and sites in a “panel of normal” blacklist (Benjamin et al., 2019). Somatic mutations were annotated using Variant Effector Predictor (release 106) and GATK-Funcotator (McLaren et al., 2016; Van der Auwera et al., 2020). For somatic CNVs, the GATK CNV-calling and ModelSegments workflow was used. In the first step, alignment data were denoised against a panel of normal to detect copy ratios. In the second step, ModelSegments was used to perform segmentation for both copy ratios and for allelic fractions jointly when both data types are present together.

Mutational signatures

Mutational signatures of samples were analyzed using SigProfiler to decipher signatures based on the SigProfiler Bioinformatic Tools (version 3.2) in COSMIC (release v93 9). These tools include matrix generation, extraction, and plotting. Contribution of each signature for each sample was statistically calculated.

Personal cancer genome reporting

Personal Cancer Genome Reporter (PCGR) software (Nakken et al., 2018) was used for functional cancer-related annotation and interpretation of cancer genomes for precision medicine. PCGR is designed with a strong clinical focus, aiming to provide actionable insights that can guide treatment decisions. It prioritizes variants that are known to be clinically relevant and have potential therapeutic implications. Unlike some existing tools that may primarily focus on variant calling and annotation, PCGR places a strong emphasis on clinical interpretation. It incorporates knowledge from curated databases, clinical trials, and published literature to provide personalized treatment recommendations based on the genomic profile of the patient's tumor. This ensures that the reported findings are clinically relevant and actionable. A clinically interpretable report was generated for each patient, contributing to clinical translation and precision oncology.

Results

Clinical features of the patient cohort

In this study, we enrolled six adults (4 males, 2 females) with newly diagnosed GBM (Table 1). Mean and median ages of patients were 56.1 and 56.5 years old, respectively. Since only newly diagnosed patients were included in study, their Karnofsky performance scores were above 80 which shows that they are able to carry on normal activity and work without special care. All samples were IDH-wild-type. Half of the tumor samples were taken from the temporal lobe, two of them were from the frontal lobe, and one tumor sample was from the occipital lobe. Mean tumor volume was 28.3 cm³. Clinical characteristics are shown in Table 1.

Table 1.

Clinical Features of the Patient Cohort and Characteristics of the Tumor Samples

Patient ID	Age	Sex	Tumor location	IDH status	Max diameter (cm)	Tumor volume (cm³)	KPS
P2	67	M	Temporal	WT	4.7	28,2	90
P3	42	M	Frontal	WT	4.7	30,1	100
P4	62	F	Occipital	WT	3.3	18,9	100
P5	56	M	Temporal	WT	4.9	32,1	90
P6	57	F	Temporal	WT	4.7	26,3	100
P7	53	M	Frontal	WT	5.2	34,7	80

Sequencing data

WES with boosted coverage for tumor-related loci was performed for tumor and matched nontumorous samples and resulted in single-end reads (an average read length of 100 bp) with 125 million read pairs for tumor samples and 62.5 million read pairs for normal samples. The mean coverages of tumor samples for cancer-related exomes and regular exomes were 500 × and 150 × , respectively. The mean coverage for matched nontumorous samples was 150 × .

SNVs and indels

To call short somatic variants (SNVs and Indels), MuTect2 was preferred considering its high accuracy in detecting short somatic variants and controlling the false positives in comparison to other somatic variant callers (Cai et al., 2016). As a result, we identified a hypermutated tumor sample (P5) with 1634 SNVs and 67 indels. Excluding this hypermutated tumor, 26 to 81 (median: 48.2) SNVs and 2 to 5 indels (median: 3.6) were detected per tumor sample. Of the genes including these SNVs, in average, 39.2 genes per tumor were protein-encoding. The most common genetic alterations were PTEN missense mutation (n = 3), EGFR amplification (n = 3), and ARHGEF10 and CNOT3 deletion (n = 3). The mutational profile of cohort including 17 most commonly mutated genes is shown as a oncoprint in Figure 1.

FIG. 1.

Oncoprint of cohort.

Copy number variations

We used GATK-CNVcalling and ModelSegments workflow to CNVs. The CNV analysis yields a mean of 174 (range = 80–299) segments alteration per tumor sample. We used a threshold of log(2) ratio ≥0.8 for filtering CNVs (mean = 32, range = 6–87) and identified proto-oncogenes and tumor suppressor genes subject to CNVs (Table 2).

Table 2.

Proto-Oncogenes and Tumor Suppressor Genes Subject to Copy Number Alterations

Patient ID	Proto-oncogene	Tumor suppressor
P2	—	RAD17, DAXX, ARHGEF10, HNF1B, PRPF31, CNOT3
P3	—	KANK1, PTPRD, CDKN2A
P4	—	—
P5	EGFR, KIT, KDR, MDM4	SRD5A3
P6	EGFR, KIT, KDR, MDM4	SRD5A3
P7	—	BRCA2, STARD13, FOXO1, ZC3H13, ALDH1A2, PRPF31, CNOT3, RAD17, ARHGEF10, DAB2IP, AHNAK, MEN1, BATF2, LATS2, PDX1, CDX2, HMGB1

Tumor mutational burden

Tumor mutation burden (TMB), which is an indicator of response to immunotherapy, was expressed as mutations per megabase (mutation/mb). In our small cohort, the median TMB was 1.45 mutations/Mb for five tumor samples, excluding the hypermutated sample (P5) with 44.53 mutations/Mb (Fig. 2A, B). Figure 2B indicates how the tumor mutational burden estimated for P5 sample (blue dotted line) compares with the distributions observed for tumor samples in TCGA.

FIG. 2.

(A). Mutation spectrum of the patient cohort. (B). The plot indicates how the mutational burden estimated for the hypermutated P5 sample (blue dotted line) compares with the distributions observed for tumor samples in TCGA. (C). First node of diagram depicts Tier 3 genes (unknown clinical significance) of each patient, third node shows Tier 1-2 gene (strong/potential clinical significance) in cohort in terms of Personal Cancer Genome Report Tiered Scheme (D). Sankey diagram to represent individualized, targeted therapy options for each patient. First node shows patient IDS which connected to the mutated genes in the second node. Third node shows possible targeted therapy in clinical trials for these gene alterations. TCGA, The Cancer Genome Atlas.

PCGR for precision medicine

For functional cancer-related annotation and interpretation of cancer genomes for precision medicine, we use the PCGR tool. Figure 2C shows a diagram depicting a tiered scheme for the PCGR. The analysis of the cohort using the PCGR tool revealed the presence of tier 1 mutations in the following genes: DNMT3A, EGFR, and BRAF. (Fig. 2C) These mutations are considered to have strong clinical significance in the context of cancer risk and tumorigenesis and possible therapeutic options. We interpreted both somatic SNVs/indels and CNVs and integrated them with comprehensive cancer-related literature to identify potential therapeutic targets (Table 3) and candidate treatment options (Fig. 2D) for each individual.

Table 3.

Personal Cancer Genome Report Tiered Scheme

	P2	P3	P4	P5	P6	P7
Tier 1 strong clinical significance		BRAF
Tier 2potential clinical significance	DNMT3A			EGFR DNMT3A	EGFR	EGFR

Discussion

Glioblastoma (GBM) is the most common adult malignancy encountered in the daily practice of neurosurgical clinics. Understanding the molecular mechanisms underlying GBM is crucial for diagnosis and treatment. In the current contemporary moment, molecular pathology informed by systems science and multiomics technology platforms is replacing classical pathology, and central nervous system tumors are the area where this change is occurring most rapidly. Therefore, advanced molecular diagnosis is becoming the standard instead of classical histopathological diagnosis (Louis et al., 2021).

The current standard treatment for GBM is maximal safe resection followed by radiotherapy in combination with concomitant temozolomide chemotherapy (Stupp et al., 2005). Despite standard treatment, GBM eventually progresses. In case of recurrence, another alkylating agent such as lomustine, a nitrosourea compound, may be used in addition to standard treatment, except for a small group of patients with the option of a second surgical resection and reirradiation (Taal et al., 2014). However, these treatment modalities have been shown to fail in recurrent disease.

Given the limited success of standard GBM therapy, personalized molecular therapy has recently become increasingly important. The most commonly targeted genetic alterations in GBM are the PIK3/AKT/mTOR (phosphoinositide 3-kinase/protein kinase B/mammalian target of rapamycin) pathway, mutation, or amplification of the epidermal growth factor receptor (EGFR) gene and other growth factor receptors (KIT, PDGFRA, FGFR genes, MET), BRAF mutation, fibroblast growth factor receptor (FGFR) gene fusions, cell cycle pathways (TP53, MDM2, CDK4/6, RB1), and higher TMB (Le Rhun et al., 2019; Touat et al., 2017).

The PI3K/AKT/mTOR pathway is an intracellular signaling pathway involved in a wide range of cellular mechanisms, including survival, proliferation, regulation of a wide range of cellular mechanisms, including cell metabolism, growth, survival, angiogenesis, and motility (Courtney et al., 2010; Fruman and Rommel, 2014; Manning and Cantley, 2007). The PTEN gene, one of the most frequently altered tumor suppressor genes in GBM, is an important negative regulator of the PI3K/AKT/mTOR pathway. Preclinical studies of targeted therapies against this pathway have shown promising results, but clinical trials have been insufficient to inhibit the target at doses tolerated by patients (Chang et al., 2005; Wen et al., 2019). In our cohort, five of six patients have mutations in this potentially targetable pathway-PTEN (n = 4/6, %66), PIK3CB (n = 1/6, %16), PIK3R1 (n = 1/6, %16).

Receptor tyrosine kinases are a subclass of tyrosine kinases that possess high-affinity cell surface receptors for many polypeptide growth factors, cytokines, and hormones (Blume-Jensen and Hunter, 2001). EGFR, VEGFR, PDGFR, FGFR, and MET, which are members of the RTK family, are frequently altered in GBM. Thus, RTK amplifications and/or mutations occur in 66% of primary GBMs in the TCGA cohort (Brennan et al., 2013; Colardo et al., 2021). Several RTK inhibitors and antibodies have been used as targeted therapy for GBM (Joshi et al., 2012; Martinho et al., 2013; Qin et al., 2021). Compensatory upregulation of RTKs after inhibition of a specific RTK may lead to tumor resistance and consequently limit clinical efficacy (Qin et al., 2021). Two samples (P5 and P6) in the cohort have potentially actionable EGFR amplification. EGFR, although frequently altered in GBM, is difficult to target effectively due to low activity against EGFRvIII and low CNS activity (Taylor et al., 2012). BRAF V600E mutations are rarely detected in GBM. vemurafaenib and dabrafenib, small molecule BRAF inhibitors, have been used in clinical trials for BRAF V600E mutated GBM and showed promising results (Burger et al., 2017; Johanns et al., 2018; Schreck et al., 2018). P3, which has a BRAF V600E mutation, is a potential candidate for BRAF inhibitor treatment in the cohort.

In this study, we applied whole-exome sequencing to a small cohort of six GBM patients to clarify molecular diagnosis, which is the first step in personalized genomics applications. We have introduced the first personalized medicine at our institution using advanced molecular genetics and bioinformatics approaches on a shoestring budget. As tumor sequencing costs decrease, we aim to apply precision medicine to every patient and achieve better outcomes in a disease with poor prognosis such as GBM. To leverage genomic alteration for clinical translation, a multidisciplinary molecular tumor board should be established. Multidisciplinary tumor boards typically consist of medical oncologists, radiation oncologists, surgical oncologists, pathologists, genetic counselors, cancer biologists, and bioinformaticians. Clinicians on these panels need to be more familiar with genomic data and techniques to translate them into clinical care (Bryce et al., 2017; Harada et al., 2017). This study provided us with this opportunity for the first time at our hospital and helped us to broaden our experience. We used whole-exome sequencing with increased coverage for cancer-related genes to generate a personalized cancer clinical genomic report. The patient-specific somatic mutations and copy number changes obtained as a result of our study are compatible with those obtained from large cohort studies (Brennan et al., 2013).

Our study has several limitations. The main limitation of our study was the small sample size. This study was designed as a hypothesis generation and feasibility study to introduce next-generation sequencing in GBM and to interpret the results in a clinically relevant way. The second limitation was the lack of transcriptomic (RNA-seq) data. Recent research has shown that tumor sequencing on two platforms—whole genome sequencing, WES, and transcriptome sequencing (RNA-seq)—results in higher accuracy and enables comprehensive reporting in a clinically relevant time frame (Chang et al., 2016; Rusch et al., 2018).

Conclusions

This study demonstrates the prospects of precision/personalized medicine in GBM specifically, and neurosurgical oncology more generally, including the potential for genomic profiling coupled with personalized cancer genome reporting. Further research and larger studies are warranted to validate these findings and advance the treatment options and outcomes for patients with GBM.

Footnotes

Acknowledgments

WES experiments were performed at Yale Center for Genome Analysis. The numerical calculations reported in this article were partially performed at TUBITAK ULAKBIM, High Performance and Grid Computing Center (TRUBA resources).

Authors' Contribution

Conception and design: O.E., F.B. Acquisition of data: O.E., F.B., K.B. Analysis and interpretation of data: O.E., F.B., K.Y.A., K.B. Drafting the article: O.E., F.B., K.Y.A. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Statistical and bioinformatics analyses: O.E., K.Y.A. Administrative/technical/material support: O.E., F.B., S.C.O., C.E., K.B. Study supervision: F.B. All authors have read and approved the final manuscript.

Author Disclosure Statement

The authors declare they have no conflicting financial interests.

Funding Information

No funding was received for this article.

Abbreviations Used

References

Andrews Simon. Babraham Bioinformatics - FastQC A Quality Control Tool for High Throughput Sequence Data. n.d. https://www.bioinformatics.babraham.ac.uk/projects/fastqc/ [Last accessed: July 29, 2023].

Benjamin

, Sato

, Cibulskis

, et al. Calling Somatic SNVs and Indels with Mutect2. bioRxiv, 2019; 861054; doi: 10.1101/861054.

Blume-Jensen

, Hunter

. Oncogenic kinase signalling. Nature, 2001; 411(6835):355–365; doi: 10.1038/35077225.

Brennan

, Verhaak

RGW

, McKenna

, et al. The somatic genomic landscape of glioblastoma. Cell, 2013; 155(2):462; doi: 10.1016/J.CELL.2013.09.034

Broad Institute. Picard Tools - By Broad Institute. n.d. Available from: https://broadinstitute.github.io/picard/ [Last accessed: July 29, 2023].

Bryce

, Egan

, Borad

, et al. Experience with precision genomics and tumor board, indicates frequent target identification, but barriers to delivery. Oncotarget, 2017; 8(16):27145–27154; doi: 10.18632/ONCOTARGET.16057

Burger

, Ronellenfitsch

, Lorenz

, et al. Dabrafenib in patients with recurrent, BRAF V600E mutated malignant glioma and leptomeningeal disease. Oncol Rep, 2017; 38(6):3291–3296; doi: 10.3892/OR.2017.6013

Cai

, Yuan

, Zhang

, et al. In-depth comparison of somatic point mutation callers based on different tumor next-generation sequencing depth data. Sci Reports, 2016; 6(1):1–9; doi: 10.1038/srep36540

Chang

, Wen

, Cloughesy

, et al. Phase II study of CCI-779 in patients with recurrent glioblastoma multiforme. Invest New Drugs, 2005; 23(4):357–361; doi: 10.1007/S10637-005-1444-0

10.

Chang

, Brohl

, Patidar

, et al. Multidimensional clinomics for precision therapy of children and adolescent young adults with relapsed and refractory cancer: A report from the center for cancer research. Clin Cancer Res, 2016; 22(15):3810–3820; doi: 10.1158/1078-0432.CCR-15-2717

11.

Cibulskis

, Lawrence

, Carter

, et al. Sensitive detection of somatic point mutations in impure and heterogeneous cancer samples. Nat Biotechnol, 2013; 31(3):213–219; doi: 10.1038/nbt.2514

12.

Colardo

, Segatto

, Di Bartolomeo

. Targeting RTK-PI3K-MTOR Axis in Gliomas: An Update. Int J Mol Sci, 2021; 22(9):22; doi: 10.3390/IJMS22094899

13.

Courtney

, Corcoran

, Engelman

. The PI3K pathway as drug target in human cancer. J Clin Oncol, 2010; 28(6):1075–1083; doi: 10.1200/JCO.2009.25.3641

14.

DeCordova

, Shastri

, Tsolaki

, et al. Molecular heterogeneity and immunosuppressive microenvironment in glioblastoma. Front Immunol, 2020; 11:535080; doi: 10.3389/FIMMU.2020.01402/BIBTEX

15.

Fruman

, Rommel

. PI3K and cancer: Lessons, challenges and opportunities. Nat Rev Drug Discov, 2014; 13(2):140–156; doi: 10.1038/NRD4204

16.

Harada

, Arend

, Dai

, et al. Implementation and utilization of the molecular tumor board to guide precision medicine. Oncotarget, 2017; 8(34):57845–57854; doi: 10.18632/ONCOTARGET.18471

17.

Johanns

, Ferguson

, Grierson

, et al. Rapid clinical and radiographic response with combined dabrafenib and trametinib in adults with BRAF-mutated high-grade glioma. J Natl Compr Canc Netw, 2018; 16(1):4–10; doi: 10.6004/JNCCN.2017.7032

18.

Joshi

, Loilome

, Siu

, et al. Evaluation of tyrosine kinase inhibitor combinations for glioblastoma therapy. PLoS One, 2012; 7(10):e44372; doi: 10.1371/JOURNAL.PONE.0044372

19.

Kuo

, Paulson

, Hempelmann

, et al. Validation and implementation of a modular targeted capture assay for the detection of clinically significant molecular oncology alterations. Pract Lab Med, 2020; 19:e00153; doi: 10.1016/J.PLABM.2020.E00153

20.

Le Rhun

, Preusser

, Roth

, et al. Molecular targeted therapy of glioblastoma. Cancer Treat Rev, 2019; 80:101896; doi: 10.1016/J.CTRV.2019.101896

21.

. Aligning Sequence Reads, Clone Sequences and Assembly Contigs with BWA-MEM. 2013. https://doi.org/10.48550/arXiv.1303.3997

22.

Louis

, Perry

, Wesseling

, et al. The 2021 WHO classification of tumors of the central nervous system: A summary. Neuro Oncol, 2021; 23(8):1231–1251; doi: 10.1093/NEUONC/NOAB106

23.

Mahlokozera

, Vellimana

, Li

, et al. Biological and therapeutic implications of multisector sequencing in newly diagnosed glioblastoma. Neuro Oncol, 2018; 20(4):472–483; doi: 10.1093/NEUONC/NOX232

24.

Manning

and Cantley

. AKT/PKB Signaling: Navigating downstream. Cell, 2007; 129(7):1261–1274; doi: 10.1016/J.CELL.2007.06.009

25.

Martinho

, Silva-Oliveira

, Miranda-Gonçalves

, et al. In vitro and in vivo analysis of RTK inhibitor efficacy and identification of its novel targets in glioblastomas. Transl Oncol, 2013; 6(2):187-IN20; doi: 10.1593/TLO.12400

26.

McKenna

, Hanna

, Banks

, et al. The genome analysis toolkit: A mapreduce framework for analyzing next-generation DNA sequencing data. Genome Res, 2010; 20(9):1297–1303; doi: 10.1101/GR.107524.110

27.

McLaren

, Gil

, Hunt

, et al. The ensembl variant effect predictor. Genome Biol, 2016; 17(1):1–14; doi: 10.1186/S13059-016-0974-4/TABLES/8

28.

McLendon

, Friedman

, Bigner

, et al. Comprehensive genomic characterization defines human glioblastoma genes and core pathways. Nature, 2008; 455(7216):1061–1068; doi: 10.1038/nature07385

29.

Nakken

, Fournous

, Vodák

, et al. Personal cancer genome reporter: Variant interpretation report for precision oncology. Bioinformatics, 2018; 34(10):1778–1780; doi: 10.1093/BIOINFORMATICS/BTX817

30.

Ohgaki

and Kleihues

. Genetic pathways to primary and secondary glioblastoma. Am J Pathol, 2007; 170(5):1445–1453; doi: 10.2353/AJPATH.2007.070011.

31.

Ostrom

, Gittleman

, Liao

, et al. CBTRUS Statistical Report: Primary brain and other central nervous system tumors diagnosed in the United States in 2010–2014. Neuro Oncol, 2017; 19(Suppl_5):v1–v88; doi: 10.1093/NEUONC/NOX158

32.

Perrin

, Samuel

, Koszyca

, et al. Glioblastoma heterogeneity and the tumour microenvironment: Implications for preclinical research and development of new treatments. Biochem Soc Trans, 2019; 47(2):625–638; doi: 10.1042/BST20180444

33.

Prados

, Byron

, Tran

, et al. Toward precision medicine in glioblastoma: The promise and the challenges. Neuro Oncol, 2015; 17(8):1051–1063; doi: 10.1093/NEUONC/NOV031

34.

Qin

, Musket

, Musich

, et al. Receptor tyrosine kinases as druggable targets in glioblastoma: Do signaling pathways matter?. Neurooncol Adv, 2021; 3(1):1–12; doi: 10.1093/NOAJNL/VDAB133

35.

Rusch

, Nakitandwe

, Shurtleff

, et al. Clinical cancer genomic profiling by three-platform sequencing of whole genome, whole exome and transcriptome. Nat Commun, 2018; 9(1):1–13; doi: 10.1038/s41467-018-06485-7.

36.

Schreck

, Guajardo

, Lin

DDM

, et al. Concurrent BRAF/MEK inhibitors in BRAF V600-mutant high-grade primary brain tumors. J Natl Compr Canc Netw, 2018; 16(4):343–347; doi: 10.6004/JNCCN.2017.7052

37.

Soeda

, Hara

, Kunisada

, et al. The evidence of glioblastoma heterogeneity. Sci Reports, 2015; 5(1):1–7; doi: 10.1038/srep07979

38.

Stupp

, Mason

, van den Bent

, et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med, 2005; 352(10):987–996; doi: 10.1056/NEJMOA043330/SUPPL_FILE/987SA1.PDF

39.

Taal

, Oosterkamp

, Walenkamp

AME

, et al. Single-agent bevacizumab or lomustine versus a combination of bevacizumab plus lomustine in patients with recurrent glioblastoma (BELOB Trial): A randomised controlled phase 2 trial. Lancet Oncol, 2014; 15(9):943–953; doi: 10.1016/S1470-2045(14)70314-6

40.

Taylor

, Furnari

, Cavenee

. Targeting EGFR for treatment of glioblastoma: Molecular basis to overcome resistance. Curr Cancer Drug Targets, 2012; 12(3):197; doi: 10.2174/156800912799277557

41.

Tirosh

, Suvà

. Tackling the many facets of glioblastoma heterogeneity. Cell Stem Cell, 2020; 26(3):303–304; doi: 10.1016/J.STEM.2020.02.005

42.

Touat

, Idbaih

, Sanson

, et al. Glioblastoma targeted therapy: Updated approaches from recent biological insights. Ann Oncol, 2017; 28(7):1457–1472; doi: 10.1093/ANNONC/MDX106

43.

Turkalp

, Karamchandani

, Das

. IDH Mutation in Glioma: New insights and promises for the future. JAMA Neurol, 2014; 71(10):1319–1325; doi: 10.1001/JAMANEUROL.2014.1205

44.

Van der Auwera

, O'Connor

, Safari an

OMCompany

. Genomics in the Cloud: Using Docker, GATK, and WDL in Terra. Genomics in the Cloud, 2020; 300. ISBN: 9781491975190

45.

Verhaak

RGW

, Hoadley

, Purdom

, et al. Integrated genomic analysis identifies clinically relevant subtypes of glioblastoma characterized by abnormalities in PDGFRA, IDH1, EGFR, and NF1. Cancer Cell, 2010; 17(1):98–110; doi: 10.1016/J.CCR.2009.12.020

46.

Wen

, Touat

, Alexander

, et al. Buparlisib in patients with recurrent glioblastoma harboring phosphatidylinositol 3-kinase pathway activation: An open-label, multicenter, multi-arm, Phase II Trial. J Clin Oncol, 2019; 37(9):741–750; doi: 10.1200/JCO.18.01207

47.

Yabroff

, Harlan

, Zeruto

, et al. Patterns of care and survival for patients with glioblastoma multiforme diagnosed during 2006. Neuro Oncol, 2012; 14(3):351; doi: 10.1093/NEUONC/NOR218