Reappraising the 2016 WHO classification for diffuse glioma

Deep learning identified glioblastoma subtypes based on internal genomic expression ranks

Background

Glioblastoma (GBM) can be divided into subtypes according to their genomic features, including Proneural (PN), Neural (NE), Classical (CL) and Mesenchymal (ME). However, it is a difficult task to unify various genomic expression profiles which were standardized with various procedures from different studies and to manually classify a given GBM sample into a subtype.

Methods

An algorithm was developed to unify the genomic profiles of GBM samples into a standardized normal distribution (SND), based on their internal expression ranks. Deep neural networks (DNN) and convolutional DNN (CDNN) models were trained on original and SND data. In addition, expanded SND data by combining various The Cancer Genome Atlas (TCGA) datasets were used to improve the robustness and generalization capacity of the CDNN models.

Results

The SND data kept unimodal distribution similar to their original data, and also kept the internal expression ranks of all genes for each sample. CDNN models trained on the SND data showed significantly higher accuracy compared to DNN and CDNN models trained on primary expression data. Interestingly, the CDNN models classified the NE subtype with the lowest accuracy in the GBM datasets, expanded datasets and in IDH wide type GBMs, consistent with the recent studies that NE subtype should be excluded. Furthermore, the CDNN models also recognized independent GBM datasets, even with small set of genomic expressions.

Conclusions

The GBM expression profiles can be transformed into unified SND data, which can be used to train CDNN models with high accuracy and generalization capacity. These models suggested NE subtype may be not compatible with the 4 subtypes classification system.

Molecular subtyping of diffuse gliomas using magnetic resonance imaging: comparison and correlation between radiomics and deep learning

OBJECTIVES

The molecular subtyping of diffuse gliomas is important. The aim of this study was to establish predictive models based on preoperative multiparametric MRI.

METHODS

A total of 1016 diffuse glioma patients were retrospectively collected from Beijing Tiantan Hospital. Patients were randomly divided into the training (n = 780) and validation (n = 236) sets. According to the 2016 WHO classification, diffuse gliomas can be classified into four binary classification tasks (tasks I-IV). Predictive models based on radiomics and deep convolutional neural network (DCNN) were developed respectively, and their performances were compared with receiver operating characteristic (ROC) curves. Additionally, the radiomics and DCNN features were visualized and compared with the t-distributed stochastic neighbor embedding technique and Spearman's correlation test.

RESULTS

In the training set, areas under the curves (AUCs) of the DCNN models (ranging from 0.99 to 1.00) outperformed the radiomics models in all tasks, and the accuracies of the DCNN models (ranging from 0.90 to 0.94) outperformed the radiomics models in tasks I, II, and III. In the independent validation set, the accuracies of the DCNN models outperformed the radiomics models in all tasks (0.74-0.83), and the AUCs of the DCNN models (0.85-0.89) outperformed the radiomics models in tasks I, II, and III. DCNN features demonstrated more superior discriminative capability than the radiomics features in feature visualization analysis, and their general correlations were weak.

CONCLUSIONS

Both the radiomics and DCNN models could preoperatively predict the molecular subtypes of diffuse gliomas, and the latter performed better in most circumstances.

KEY POINTS

• The molecular subtypes of diffuse gliomas could be predicted with MRI. • Deep learning features tend to outperform radiomics features in large cohorts. • The correlation between the radiomics features and DCNN features was low.

Multi-parametric arterial spin labelling and diffusion-weighted magnetic resonance imaging in differentiation of grade II and grade III gliomas

Purpose

To assess arterial spin labelling (ASL) perfusion and diffusion MR imaging (DWI) in the differentiation of grade II from grade III gliomas.

Material and methods

A prospective cohort study was done on 36 patients (20 male and 16 female) with diffuse gliomas, who underwent ASL and DWI. Diffuse gliomas were classified into grade II and grade III. Calculation of tumoural blood flow (TBF) and apparent diffusion coefficient (ADC) of the tumoral and peritumoural regions was made. The ROC curve was drawn to differentiate grade II from grade III gliomas.

Results

There was a significant difference in TBF of tumoural and peritumoural regions of grade II and III gliomas (p = 0.02 and p =0.001, respectively). Selection of 26.1 and 14.8 ml/100 g/min as the cut-off for TBF of tumoural and peritumoural regions differentiated between both groups with area under curve (AUC) of 0.69 and 0.957, and accuracy of 77.8% and 88.9%, respectively. There was small but significant difference in the ADC of tumoural and peritumoural regions between grade II and III gliomas (p = 0.02 for both). The selection of 1.06 and 1.36 × 10^-3 mm²/s as the cut-off of ADC of tumoural and peritumoural regions was made, to differentiate grade II from III with AUC of 0.701 and 0.748, and accuracy of 80.6% and 80.6%, respectively. Combined TBF and ADC of tumoural regions revealed an AUC of 0.808 and accuracy of 72.7%. Combined TBF and ADC for peritumoural regions revealed an AUC of 0.96 and accuracy of 94.4%.

Conclusion

TBF and ADC of tumoural and peritumoural regions are accurate non-invasive methods of differentiation of grade II from grade III gliomas.

CHI3L1, NTRK2, 1p/19q and IDH Status Predicts Prognosis in Glioma

The aim of this study was to identify relevant biomarkers for the prognosis of glioma considering current molecular changes such as IDH mutation and 1p19q deletion. Gene expression profiling was performed using the TaqMan Low Density Array and hierarchical clustering using 96 selected genes in 64 patients with newly diagnosed glioma. The expression dataset was validated on a large independent cohort from The Cancer Genome Atlas (TCGA) database. A differential expression panel of 26 genes discriminated two prognostic groups regardless of grade and molecular groups of tumors: Patients having a poor prognosis with a median overall survival (OS) of 23.0 ± 9.6 months (group A) and patients having a good prognosis with a median OS of 115.0 ± 6.6 months (group B) (p = 0.007). Hierarchical clustering of the glioma TCGA cohort supported the prognostic value of these 26 genes (p < 0.0001). Among these genes, CHI3L1 and NTRK2 were identified as factors that can be associated with IDH status and 1p/19q co-deletion to distinguish between prognostic groups of glioma from the TCGA cohort. Therefore, CHI3L1 associated with NTRK2 seemed to be able to provide new information on glioma prognosis.