Dataset Information

A Preliminary Study on Machine Learning-Based Evaluation of Static and Dynamic FET-PET for the Detection of Pseudoprogression in Patients with IDH-Wildtype Glioblastoma.

ABSTRACT: Pseudoprogression (PSP) detection in glioblastoma remains challenging and has important clinical implications. We investigated the potential of machine learning (ML) in improving the performance of PET using O-(2-[18F]-fluoroethyl)-L-tyrosine (FET) for differentiation of tumor progression from PSP in IDH-wildtype glioblastoma. We retrospectively evaluated the PET data of patients with newly diagnosed IDH-wildtype glioblastoma following chemoradiation. Contrast-enhanced MRI suspected PSP/TP and all patients underwent subsequently an additional dynamic FET-PET scan. The modified Response Assessment in Neuro-Oncology (RANO) criteria served to diagnose PSP. We trained a Linear Discriminant Analysis (LDA)-based classifier using FET-PET derived features on a hold-out validation set. The results of the ML model were compared with a conventional FET-PET analysis using the receiver-operating-characteristic (ROC) curve. Of the 44 patients included in this preliminary study, 14 patients were diagnosed with PSP. The mean (TBRmean) and maximum tumor-to-brain ratios (TBRmax) were significantly higher in the TP group as compared to the PSP group (p = 0.014 and p = 0.033, respectively). The area under the ROC curve (AUC) for TBRmax and TBRmean was 0.68 and 0.74, respectively. Using the LDA-based algorithm, the AUC (0.93) was significantly higher than the AUC for TBRmax. This preliminary study shows that in IDH-wildtype glioblastoma, ML-based PSP detection leads to better diagnostic performance.

SUBMITTER: Kebir S

PROVIDER: S-EPMC7690380 | biostudies-literature | 2020 Oct

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

A Preliminary Study on Machine Learning-Based Evaluation of Static and Dynamic FET-PET for the Detection of Pseudoprogression in Patients with IDH-Wildtype Glioblastoma.

Kebir Sied S Schmidt Teresa T Weber Matthias M Lazaridis Lazaros L Galldiks Norbert N Langen Karl-Josef KJ Kleinschnitz Christoph C Hattingen Elke E Herrlinger Ulrich U Lohmann Philipp P Glas Martin M

Cancers 20201022 11

Pseudoprogression (PSP) detection in glioblastoma remains challenging and has important clinical implications. We investigated the potential of machine learning (ML) in improving the performance of PET using O-(2-[<sup>18</sup>F]-fluoroethyl)-L-tyrosine (FET) for differentiation of tumor progression from PSP in IDH-wildtype glioblastoma. We retrospectively evaluated the PET data of patients with newly diagnosed IDH-wildtype glioblastoma following chemoradiation. Contrast-enhanced MRI suspected P ...[more]

PMID: 33105661

Similar Datasets

Project description:Objectives: In the present study, we aimed to determine the candidate genes that may function as biomarkers to further distinguish patients with isocitrate dehydrogenase (IDH)-wildtype glioblastoma (GBM), which are heterogeneous with respect to clinical outcomes. Materials and Methods: We selected 41 candidate genes associated with overall survival (OS) using univariate Cox regression from IDH-wildtype GBM patients based on RNA sequencing (RNAseq) expression data from the Chinese Glioma Genome Atlas (CGGA, n = 105) and The Cancer Genome Atlas (TCGA, n = 139) cohorts. Next, a seven-gene-based risk signature was formulated according to Least Absolute Shrinkage and Selection Operator (LASSO) regression algorithm in the CGGA RNAseq database as a training set, while another 525 IDH-wildtype GBM patient TCGA datasets, consisting of RNA sequencing and microarray data, were used for validation. Patient survival in the low- and high-risk groups was calculated using Kaplan-Meier survival curve analysis and the log-rank test. Uni-and multivariate Cox regression analysis was used to assess the prognosis value. Gene oncology (GO) and gene set enrichment analysis (GSEA) were performed for the functional analysis of the seven-gene-based risk signature. Results: We developed a seven-gene-based signature, which allocated each patient to a risk group (low or high). Patients in the high-risk group had dramatically shorter overall survival than their low-risk counterparts in three independent cohorts. Univariate and multivariate analysis showed that the seven-gene signature remained an independent prognostic factor. Moreover, the seven-gene risk signature exhibited a striking prognostic validity, with AUC of 78.4 and 73.9%, which was higher than for traditional "age" (53.7%, 62.4%) and "GBM sub-type" (57.7%, 52.9%) in the CGGA- and TCGA-RNAseq databases, respectively. Subsequent bioinformatics analysis predicted that the seven-gene signature was involved in the inflammatory response, immune response, cell adhesion, and apoptotic process. Conclusions: Our findings indicate that the seven-gene signature could be a potential prognostic biomarker. This study refined the current classification system of IDH-wildtype GBM and may provide a novel perspective for the research and individual therapy of IDH-wildtype GBM.

Dataset Information

A Preliminary Study on Machine Learning-Based Evaluation of Static and Dynamic FET-PET for the Detection of Pseudoprogression in Patients with IDH-Wildtype Glioblastoma.

Publications

A Preliminary Study on Machine Learning-Based Evaluation of Static and Dynamic FET-PET for the Detection of Pseudoprogression in Patients with IDH-Wildtype Glioblastoma.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets