IDH Mutation Status Prediction Research Articles

Two-Stage Training Framework Using Multicontrast MRI Radiomics for IDH Mutation Status Prediction in Glioma.

Purpose To develop a radiomics framework for preoperative MRI-based prediction of isocitrate dehydrogenase (IDH) mutation status, a crucial glioma prognostic indicator. Materials and Methods Radiomics features (shape, first-order statistics, and texture) were extracted from the whole tumor or the combination of nonenhancing, necrosis, and edema regions. Segmentation masks were obtained via the federated tumor segmentation tool or the original data source. Boruta, a wrapper-based feature selection algorithm, identified relevant features. Addressing the imbalance between mutated and wild-type cases, multiple prediction models were trained on balanced data subsets using random forest or XGBoost and assembled to build the final classifier. The framework was evaluated using retrospective MRI scans from three public datasets (The Cancer Imaging Archive [TCIA, 227 patients], the University of California San Francisco Preoperative Diffuse Glioma MRI dataset [UCSF, 495 patients], and the Erasmus Glioma Database [EGD, 456 patients]) and internal datasets collected from the University of Texas Southwestern Medical Center (UTSW, 356 patients), New York University (NYU, 136 patients), and University of Wisconsin-Madison (UWM, 174 patients). TCIA and UTSW served as separate training sets, while the remaining data constituted the test set (1617 or 1488 testing cases, respectively). Results The best performing models trained on the TCIA dataset achieved area under the receiver operating characteristic curve (AUC) values of 0.89 for UTSW, 0.86 for NYU, 0.93 for UWM, 0.94 for UCSF, and 0.88 for EGD test sets. The best performing models trained on the UTSW dataset achieved slightly higher AUCs: 0.92 for TCIA, 0.88 for NYU, 0.96 for UWM, 0.93 for UCSF, and 0.90 for EGD. Conclusion This MRI radiomics-based framework shows promise for accurate preoperative prediction of IDH mutation status in patients with glioma. Keywords: Glioma, Isocitrate Dehydrogenase Mutation, IDH Mutation, Radiomics, MRI Supplemental material is available for this article. Published under a CC BY 4.0 license. See also commentary by Moassefi and Erickson in this issue.

Development of End-to-End AI-Based MRI Image Analysis System for Predicting IDH Mutation Status of Patients with Gliomas: Multicentric Validation.

Radiogenomics has shown potential to predict genomic phenotypes from medical images. The development of models using standard-of-care pre-operative MRI images, as opposed to advanced MRI images, enables a broader reach of such models. In this work, a radiogenomics model for IDH mutation status prediction from standard-of-care MRIs in patients with glioma was developed and validated using multicentric data. A cohort of 142 (wild-type: 32.4%) patients with glioma retrieved from the TCIA/TCGA was used to train a logistic regression model to predict the IDH mutation status. The model was evaluated using retrospective data collected in two distinct hospitals, comprising 36 (wild-type: 63.9%) and 53 (wild-type: 75.5%) patients. Model development utilized ROC analysis. Model discrimination and calibration were used for validation. The model yielded an AUC of 0.741 vs. 0.716 vs. 0.938, a sensitivity of 0.784 vs. 0.739 vs. 0.875, and a specificity of 0.657 vs. 0.692 vs. 1.000 on the training, test cohort 1, and test cohort 2, respectively. The assessment of model fairness suggested an unbiased model for age and sex, and calibration tests showed a p < 0.05. These results indicate that the developed model allows the prediction of the IDH mutation status in gliomas using standard-of-care MRI images and does not appear to hold sex and age biases.

IDH mutation status prediction by a radiomics associated modality attention network

IDH mutation status prediction by a radiomics associated modality attention network

NI-12 The ratio of T1-Weighted to T2-Weighted Signal Intensity and IDH mutation in glioma

Introduction: Prediction of IDH mutation status for Lower-grade glioma (LrGG) is clinically significant. The purpose of this study is to test the hypothesis that the T1-weighted image/T2-weighted image ratio (rT1/T2), an imaging surrogate developed for myelin integrity, is a useful MRI biomarker for predicting the IDH mutation status of LrGG. Methods: Twenty-five LrGG patients (IDHwt: 8, IDHmt: 17) at Asahikawa Medical University Hospital (AMUH) were used as an exploratory cohort. Twenty-nine LrGG patients (IDHwt: 13, IDHmt: 16) from Osaka International Cancer Institute (OICI) and 103 patients from the Cancer Imaging Archive (TCIA) / Cancer Genome Atlas (TCGA) dataset (IDHwt: 19, IDHmt: 84) were used as validation cohorts. rT1/T2 images were calculated from T1- and T2-weighted images using a recommended signal correction. The region-of-interest was defined on T2-weighted images, and the relationship between the mean rT1/T2 (mrT1/T2) and the IDH mutation status was investigated. Results: The mrT1/T2 was able to significantly predict the IDH mutation status for the AMUH exploratory cohort (AUC = 0.75, p = 0.048). The ideal cut-off for detecting mutant IDH was mrT1/T2 < 0.666 ~ 0.677, with a sensitivity of 58.8% and a specificity of 87.5%. This result was further validated by the OICI validation cohort (AUC = 0.75, p = 0.023) with a sensitivity of 56.3% and a specificity of 69.2%. On the other hand, the sensitivity was 42.9% and the specificity was 68.4 % for the TCIA validation cohort (AUC = 0.63, p = 0.068). Conclusion: Our results supported the hypothesis that mrT1/T2 could be a useful image surrogate to predict the IDH mutation status of LrGG using two domestic cohorts. The decline of the accuracy for the TCIA cohort should be further investigated.

Perfusion parameters derived from MRI for preoperative prediction of IDH mutation and MGMT promoter methylation status in glioblastomas

PurposeTo investigate the feasibility for preoperative prediction of IDH mutation and MGMT promoter methylation status in glioblastomas(GBMs) by intravoxel incoherent motion(IVIM) and dynamic susceptibility contrast(DSC). MethodsPreoperative IVIM and DSC images of 71 patients(IDH mutation:45, IDH wildtype: 26; MGMT methylation: 31, MGMT unmethylation:40) with glioblastomas were analyzed retrospectively. Perfusion parameters including microcirculation perfusion coefficient(D*), perfusion fraction(f), cerebral blood volume(CBV) and cerebral blood flow(CBF) were measured. Corrected perfusion parameters containing corrected perfusion coefficient(ADCperf) and simplified perfusion fraction(SPF) were from the simplified IVIM with 3 b values. Correlations among parameters were analyzed by Spearman correlation. All parameters were compared with Mann-Whitney U test. Univariate and multivariate logistic regression models were constructed. The receiver operating characteristic(ROC) curve was analyzed. ResultsThe IVIM parameters showed merely moderate correlations with CBV and showed no correlation with CBF. IDH mutation GBMs showed lower D*, ADCperf, SPF, CBV and higher f than IDH wildtype GBMs(all p < 0.05). D* was the independent predictor for IDH mutation with the highest AUC of 0.912(95%CI: 0.821–0.966). The D*, ADCperf, SPF and CBV of MGMT promoter methylation GBMs were lower than unmethylation GBMs while f was higher(all p < 0.05). Multivariate model showed the highest prediction efficacy for MGMT promoter methylation with an AUC of 0.915(95%CI: 0.824–0.968). The CBF was not useful in distinguishing IDH mutation and MGMT promoter methylation status(p = 0.055, 0.215). ConclusionIDH mutation and MGMT promoter methylation status in GBMs can be assessed effectively by IVIM and DSC. Besides, D* was the independent predictor of IDH mutation status.

Diagnostic performance between MR amide proton transfer (APT) and diffusion kurtosis imaging (DKI) in glioma grading and IDH mutation status prediction at 3 T

PurposeAccurate glioma grading and IDH mutation status prediction are critically essential for individualized preoperative treatment decisions. This study aims to compare the diagnostic performance of magnetic resonance (MR) amide proton transfer (APT) and diffusion kurtosis imaging (DKI) in glioma grading and IDH mutation status prediction. MethodFifty-one glioma patients without treatment were retrospectively included. APT-weighted (APTw) effect and DKI indices, including mean diffusivity (MD), fractional anisotropy (FA), mean kurtosis (MK), and kurtosis FA (KFA) were obtained from APT and diffusion-weighted images, respectively. DKI indices in tumors were normalized to that in contralateral normal appearing white matter (CNAWM) and the APTw difference (ΔAPTw) between the two regions was calculated. Student’s t-test, one-way ANOVA and ROC analyses were conducted. ResultsAmong the enrolled 51 patients, 13 had glioma-II, 17 had glioma-III and 21 had glioma-IV. 25 patients were diagnosed as IDH-mutant, and 26 as IDH-wild type. MD and MK differed significantly between glioma-IV and glioma II/III (P < 0.05), but not between glioma-II and glioma-III. FA and KFA showed no significant difference among the three groups (P > 0.05). IDH-mutant group exhibited significantly higher MD and lower FA, MK and ΔAPTw than IDH-wild type (P < 0.05), whereas the two groups showed comparable KFA values. In contrast, ΔAPTw differed significantly across tumor grades and IDH mutation status (P < 0.05), with consistently better discriminatory performance than DKI indices in glioma grading and IDH mutation status prediction. ConclusionsAPT imaging was superior to DKI in glioma grading and IDH mutation status prediction, benefiting accurate diagnoses and treatment decisions.

Vascular habitat analysis based on dynamic susceptibility contrast perfusion MRI predicts IDH mutation status and prognosis in high-grade gliomas.

The current study aimed to evaluate the clinical practice for hemodynamic tissue signature (HTS) method in IDH genotype prediction in three groups derived from high-grade gliomas. Preoperative MRI examinations of 44 patients with known grade and IDH genotype were assigned into three study groups: glioblastoma multiforme, grade III, and high-grade gliomas. Perfusion parameters were analyzed and were used to automatically draw the four reproducible habitats (high-angiogenic enhancing tumor habitats, low-angiogenic enhancing tumor habitats, infiltrated peripheral edema habitats, vasogenic peripheral edema habitats) related to vascular heterogeneity. These four habitats were then compared between inter-patient with IDH mutation and their wild-type counterparts at these three groups, respectively. The discriminating potential for HTS in assessing IDH mutation status prediction was assessed by ROC curves. Compared with IDH wild type, IDH mutation had significantly decreased relative cerebral blood volume (rCBV) at the high-angiogenic enhancing tumor habitats and low-angiogenic enhancing tumor habitats. ROC analysis revealed that the rCBVs in habitats had great ability to discriminate IDH mutation from their wild type in all groups. In addition, the Kaplan-Meier survival analysis yielded significant differences for the survival times observed from the populations dichotomized by low (< 4.31) and high (> 4.31) rCBV in the low-angiogenic enhancing tumor habitat. The HTS method has been proven to have high prediction capabilities for IDH mutation status in high-grade glioma patients, providing a set of quantifiable habitats associated with tumor vascular heterogeneity. • The HTS method has a high accuracy for molecular stratification prediction for all subsets of HGG. • The HTS method can give IDH mutation-related hemodynamic information of tumor-infiltrated and vasogenic edema. • IDH-relevant rCBV difference in habitats will be a great prognosis factor in HGG.

COMP-03. QUANTITATIVE IMAGE FEATURE ANALYSIS IN DIFFUSE GLIOMA – A VALUABLE MR IMAGING BIOMARKER FOR PREOPERATIVE IDH MUTATION CLASSIFICATION

Abstract INTRODUCTION Pre-operative differentiation of IDH mutant gliomas from similar appearing pathologies on imaging prior to definitive surgical diagnosis may aid treatment navigation, maximize the surgical approach, and provide diagnostic support for inoperable tumors. Quantitative image feature analysis offers a potential non-invasive method to identify diagnostic, prognostic, and predictive imaging biomarkers. We investigated the use of radiomic MR imaging features to classify tumors based on IDH mutation status. METHOD Pre-operative T1-weighted (T1W), T2-weighted (T2W), T1-contrast enhanced (T1CE), and fluid attenuated inversion recovery (FLAIR) MR brain images, along with patient IDH mutation status (mutant/wildtype) were obtained for 128 glioma patients from The Cancer Genome Atlas (TCGA). Enhancing tumor was delineated by GLISTRboost. GlistrBoost is a hybrid-discriminative model that segments tumors based on an expectation-maximization framework with a classification scheme and uses a probabilistic Bayesian strategy for segmentation refinement. MR studies for 78 glioma patients from six institutions were used for training and 50 glioma patients from a different institution were used for validation. Pre-processing included registration, resampling, and normalization. Cancer Imaging Phenomics Toolkit (CaPTK) extracted 938 radiomic image features per sequence for the enhancing tumor contour. Relevance of each individual feature was determined by the least absolute shrinkage and selection operator (LASSO). The ability of relevant radiomic image features to identify mutation status of IDH was assessed by logistic regression. RESULTS LASSO identified one highly informative radiomic imaging feature, the minimum of the mean absolute histogram deviation on T1 MR images, which was able to predict IDH mutation status with an accuracy of 0.74, precision of 1.0, and recall of 0.32. CONCLUSION Non-invasive prediction of IDH mutation status from pre-surgical MR images offers potential diagnostic, therapeutic, and prognostic benefits for glioma patients. Quantitative image feature analysis is a feasible method for identifying potential radiomic imaging features.

NIMG-67. DEVELOPMENT OF VERSATILE MACHINE-LEARNING APPROACHES FOR RADIOGENOMICS OF GLIOMA IN DIFFERENT COHORTS

Abstract BACKGROUND Radiogenomics aims to analyze clinical images and information, and to predict key molecular profiles of tumors. However, imaging protocol is usually different in facilities, and it has been rarely examined whether the performance of developed methods in a dataset is robustly sustained even in other independent datasets. We explored machine learning and matrix decomposition methods using preoperative magnetic resonance images (MRIs) of glioma patients to establish versatile platform regardless of the heterogeneity of the datasets. METHODS Preoperative glioma MRIs and clinical information were obtained from public dataset of The Cancer Imaging Archive (TCIA, N=159) and National Cancer Center Hospital (NCC, N=166). More than 16,000 radiomic features were applied for the prediction of tumor grading and IDH mutation status. Accuracy of prediction was evaluated by AUROC (area under the receiver operating characteristic curves). RESULTS The performances were comparable between the image features regardless of dimension reduction methods (the best accuracy for tumor grading and IDH status prediction was 0.91 and 0.88, respectively), but they were drastically decreased in the transfer learning (0.70 and 0.69). On the other hand, they were successfully improved by applying matrix decomposition and brain embedding (0.86 and 0.79). CONCLUSION Our result and pipeline can be a global benchmark for future studies in heterogeneous datasets. Further evaluation in larger cohorts are planned.

Texture analysis- and support vector machine-assisted diffusional kurtosis imaging may allow in vivo gliomas grading and IDH-mutation status prediction: a preliminary study

We sought to investigate, whether texture analysis of diffusional kurtosis imaging (DKI) enhanced by support vector machine (SVM) analysis may provide biomarkers for gliomas staging and detection of the IDH mutation. First-order statistics and texture feature extraction were performed in 37 patients on both conventional (FLAIR) and mean diffusional kurtosis (MDK) images and recursive feature elimination (RFE) methodology based on SVM was employed to select the most discriminative diagnostic biomarkers. The first-order statistics demonstrated significantly lower MDK values in the IDH-mutant tumors. This resulted in 81.1% accuracy (sensitivity = 0.96, specificity = 0.45, AUC 0.59) for IDH mutation diagnosis. There were non-significant differences in average MDK and skewness among the different tumour grades. When texture analysis and SVM were utilized, the grading accuracy achieved by DKI biomarkers was 78.1% (sensitivity 0.77, specificity 0.79, AUC 0.79); the prediction accuracy for IDH mutation reached 83.8% (sensitivity 0.96, specificity 0.55, AUC 0.87). For the IDH mutation task, DKI outperformed significantly the FLAIR imaging. When using selected biomarkers after RFE, the prediction accuracy achieved 83.8% (sensitivity 0.92, specificity 0.64, AUC 0.88). These findings demonstrate the superiority of DKI enhanced by texture analysis and SVM, compared to conventional imaging, for gliomas staging and prediction of IDH mutational status.