Dataset Information

Radiogenomic modeling predicts survival-associated prognostic groups in glioblastoma.

ABSTRACT:

Background

Combined whole-exome sequencing (WES) and somatic copy number alteration (SCNA) information can separate isocitrate dehydrogenase (IDH)1/2-wildtype glioblastoma into two prognostic molecular subtypes, which cannot be distinguished by epigenetic or clinical features. The potential for radiographic features to discriminate between these molecular subtypes has yet to be established.

Methods

Radiologic features (n = 35 340) were extracted from 46 multisequence, pre-operative magnetic resonance imaging (MRI) scans of IDH1/2-wildtype glioblastoma patients from The Cancer Imaging Archive (TCIA), all of whom have corresponding WES/SCNA data. We developed a novel feature selection method that leverages the structure of extracted MRI features to mitigate the dimensionality challenge posed by the disparity between a large number of features and the limited patients in our cohort. Six traditional machine learning classifiers were trained to distinguish molecular subtypes using our feature selection method, which was compared to least absolute shrinkage and selection operator (LASSO) feature selection, recursive feature elimination, and variance thresholding.

Results

We were able to classify glioblastomas into two prognostic subgroups with a cross-validated area under the curve score of 0.80 (±0.03) using ridge logistic regression on the 15-dimensional principle component analysis (PCA) embedding of the features selected by our novel feature selection method. An interrogation of the selected features suggested that features describing contours in the T2 signal abnormality region on the T2-weighted fluid-attenuated inversion recovery (FLAIR) MRI sequence may best distinguish these two groups from one another.

Conclusions

We successfully trained a machine learning model that allows for relevant targeted feature extraction from standard MRI to accurately predict molecularly-defined risk-stratifying IDH1/2-wildtype glioblastoma patient groups.

SUBMITTER: Nuechterlein N

PROVIDER: S-EPMC7883769 | biostudies-literature | 2021 Jan-Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Radiogenomic modeling predicts survival-associated prognostic groups in glioblastoma.

Nuechterlein Nicholas N Li Beibin B Feroze Abdullah A Holland Eric C EC Shapiro Linda L Haynor David D Fink James J Cimino Patrick J PJ

Neuro-oncology advances 20210101 1

<h4>Background</h4>Combined whole-exome sequencing (WES) and somatic copy number alteration (SCNA) information can separate <i>isocitrate dehydrogenase (IDH)1/2</i>-wildtype glioblastoma into two prognostic molecular subtypes, which cannot be distinguished by epigenetic or clinical features. The potential for radiographic features to discriminate between these molecular subtypes has yet to be established.<h4>Methods</h4>Radiologic features (<i>n</i> = 35 340) were extracted from 46 multisequence ...[more]

PMID: 33615222

Similar Datasets

Project description:Simple Summary Glioblastoma is the most common tumour that originates in the brain in adults. Most of the published data on glioblastoma are from patients in clinical trials who tend to be younger and fitter than the average patient. We therefore looked at patient demographic, tumour characteristics, and treatments received in a group of 490 real-world patients with glioblastoma to evaluate their survival, and to investigate whether we could find any factors that were associated with longer survival. Overall, the average survival of patients was 9 months. Patients tended to live longer if they were younger, had surgery, if they had further treatment after surgery (chemo- or radio-therapy), or if they had a tumour marker called MGMT promotor methylation. Abstract Background: IDH-wildtype glioblastoma is the most common malignant primary brain tumour in adults. As there is limited information on prognostic factors outside of clinical trials; thus, we conducted a retrospective study to characterise the glioblastoma population at our centre. Methods: Demographic, tumour molecular profiles, treatment, and survival data were collated for patients diagnosed with glioblastoma at our centre between July 2011 and December 2015. We used multivariate proportional hazard model associations with survival. Results: 490 patients were included; 60% had debulking surgery and 40% biopsy only. Subsequently, 56% had standard chemoradiotherapy, 25% had non-standard chemo/radio-therapy, and 19% had no further treatment. Overall survival was 9.2 months. In the multivariate analysis, longer survival was associated with debulking surgery vs. biopsy alone (14.9 vs. 8 months) (HR 0.54 [95% CI 0.41–0.70]), subsequent treatment after diagnosis (HR 0.12 [0.08–0.16]) (standard chemoradiotherapy [16.9 months] vs. non-standard regimens [9.2 months] vs. none [2.0 months]), tumour MGMT promotor methylation (HR 0.71 [0.58–0.87]), and younger age (hazard ratio vs. age < 50: 1.70 [1.26–2.30] for ages 50–59; 3.53 [2.65–4.70] for ages 60–69; 4.82 [3.54–6.56] for ages 70+). Conclusions: The median survival for patients with glioblastoma is less than a year. Younger age, debulking surgery, treatment with chemoradiotherapy, and MGMT promotor methylation are independently associated with longer survival.

Project description:PurposeTo derive quantitative image features from magnetic resonance (MR) images that characterize the radiographic phenotype of glioblastoma multiforme (GBM) lesions and to create radiogenomic maps associating these features with various molecular data.Materials and methodsClinical, molecular, and MR imaging data for GBMs in 55 patients were obtained from the Cancer Genome Atlas and the Cancer Imaging Archive after local ethics committee and institutional review board approval. Regions of interest (ROIs) corresponding to enhancing necrotic portions of tumor and peritumoral edema were drawn, and quantitative image features were derived from these ROIs. Robust quantitative image features were defined on the basis of an intraclass correlation coefficient of 0.6 for a digital algorithmic modification and a test-retest analysis. The robust features were visualized by using hierarchic clustering and were correlated with survival by using Cox proportional hazards modeling. Next, these robust image features were correlated with manual radiologist annotations from the Visually Accessible Rembrandt Images (VASARI) feature set and GBM molecular subgroups by using nonparametric statistical tests. A bioinformatic algorithm was used to create gene expression modules, defined as a set of coexpressed genes together with a multivariate model of cancer driver genes predictive of the module's expression pattern. Modules were correlated with robust image features by using the Spearman correlation test to create radiogenomic maps and to link robust image features with molecular pathways.ResultsEighteen image features passed the robustness analysis and were further analyzed for the three types of ROIs, for a total of 54 image features. Three enhancement features were significantly correlated with survival, 77 significant correlations were found between robust quantitative features and the VASARI feature set, and seven image features were correlated with molecular subgroups (P < .05 for all). A radiogenomics map was created to link image features with gene expression modules and allowed linkage of 56% (30 of 54) of the image features with biologic processes.ConclusionRadiogenomic approaches in GBM have the potential to predict clinical and molecular characteristics of tumors noninvasively. Online supplemental material is available for this article.

Dataset Information

Radiogenomic modeling predicts survival-associated prognostic groups in glioblastoma.

Background

Methods

Results

Conclusions

Publications

Radiogenomic modeling predicts survival-associated prognostic groups in glioblastoma.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets