Visualization of tumor heterogeneity and prediction of isocitrate dehydrogenase mutation status for human gliomas using multiparametric physiologic and metabolic MRI

Abstract

This study aimed to differentiate isocitrate dehydrogenase (IDH) mutation status with the voxel-wise clustering method of multiparametric magnetic resonance imaging (MRI) and to discover biological underpinnings of the clusters. A total of 69 patients with treatment-naïve diffuse glioma were scanned with pH-sensitive amine chemical exchange saturation transfer MRI, diffusion-weighted imaging, fluid-attenuated inversion recovery, and contrast-enhanced T1-weighted imaging at 3 T. An unsupervised two-level clustering approach was used for feature extraction from acquired images. The logarithmic ratio of the labels in each class within tumor regions was applied to a support vector machine to differentiate IDH status. The highest performance to predict IDH mutation status was found for 10-class clustering, with a mean area under the curve, accuracy, sensitivity, and specificity of 0.94, 0.91, 0.90, and 0.91, respectively. Targeted biopsies revealed that the tissues with labels 7-10 showed high expression levels of hypoxia-inducible factor 1-alpha, glucose transporter 3, and hexokinase 2, which are typical of IDH wild-type glioma, whereas those with labels 1 showed low expression of these proteins. In conclusion, A machine learning model successfully predicted the IDH mutation status of gliomas, and the resulting clusters properly reflected the metabolic status of the tumors.

Figure 1

Component planes with a SOM for CE-T1WI, FLAIR, MTR_asym at 3.0 ppm, and ADC colorized from blue to red according to each value, with red indicating a higher weight. The inter-class borderlines obtained by K-means clustering with K = 10 are shown on the SOM component planes as black lines between the nodes. Detailed profiles can be seen on the K-means clustering map from labels 1 to 10 shown at the far right.

Figure 2

(a) Box-whisker plots and (b) radar charts of labels by 10-class clustering. (a) The box shows the interquartile range between the 25th and 75th percentiles for log-ratio values; the lines within the boxes represent medians, and the whiskers represent measurements 1.5 times the interquartile range. The circles represent outliers beyond 1.5 times the interquartile range. * P < 0.05, ** P < 0.01, *** P < 0.001. (b) Radar charts of four variables (CE-T1WI, FLAIR, MTR_asym at 3.0 ppm, and ADC) in each label categorized into three groups.

Figure 3

Representative cases of IDH mutant and wild-type gliomas with 10-class clustering. The CE-T1WI and FLAIR images, MTR_asym at 3.0 ppm, and ADC maps are shown for each patient. Each color within the tumor ROIs corresponds to each label in the 10-color bar and each category in the 3-color bar. The ratios of each label and category are shown in pie charts. In these examples, labels in category M (label 1) occupied about half of the tumor ROIs in IDH mutant gliomas, while labels in category W (label 7–10) occupied less than a quarter of the tumor ROIs. In IDH wild-type gliomas, labels in category W occupied the majority of the tumor ROIs.

Figure 4

Box-whisker plots of histological measurements, namely, HIF1a-, GLUT3-, HK2-, MCT1-, LDHA-, and Ki67-positive cell percentages. The box shows the interquartile range between the 25th and 75th percentiles for positive cell percentage; the lines within boxes represent medians, and the whiskers represent measurements 1.5 times the interquartile range. The circles represent outliers beyond 1.5 times the interquartile range.

Figure 5

MR images and corresponding hematoxylin and eosin (H&E) and immunohistochemistry staining for MRI-guided biopsy targets (circles). (a) IDH mutant glioma for which an area with labels categorized as M, indicating the IDH mutant feature, was biopsied. Expressions of HIF1a, GLUT3, and HK2 are low in the slides from a 5-mm radius sample taken from the MRI-guided biopsy target. (b) IDH wild-type glioma for which an area with labels categorized as W, indicating IDH wild-type feature, was biopsied. Expressions of HIF1a, GLUT3, and HK2 are high.

Figure 6

Graphical overview of the processing pipeline.