Imaging-genomics reveals driving pathways of MRI derived volumetric tumor phenotype features in Glioblastoma

Abstract

Background: Glioblastoma (GBM) tumors exhibit strong phenotypic differences that can be quantified using magnetic resonance imaging (MRI), but the underlying biological drivers of these imaging phenotypes remain largely unknown. An Imaging-Genomics analysis was performed to reveal the mechanistic associations between MRI derived quantitative volumetric tumor phenotype features and molecular pathways.

Methods: One hundred fourty one patients with presurgery MRI and survival data were included in our analysis. Volumetric features were defined, including the necrotic core (NE), contrast-enhancement (CE), abnormal tumor volume assessed by post-contrast T1w (tumor bulk or TB), tumor-associated edema based on T2-FLAIR (ED), and total tumor volume (TV), as well as ratios of these tumor components. Based on gene expression where available (n = 91), pathway associations were assessed using a preranked gene set enrichment analysis. These results were put into context of molecular subtypes in GBM and prognostication.

Results: Volumetric features were significantly associated with diverse sets of biological processes (FDR < 0.05). While NE and TB were enriched for immune response pathways and apoptosis, CE was associated with signal transduction and protein folding processes. ED was mainly enriched for homeostasis and cell cycling pathways. ED was also the strongest predictor of molecular GBM subtypes (AUC = 0.61). CE was the strongest predictor of overall survival (C-index = 0.6; Noether test, p = 4x10(-4)).

Conclusion: GBM volumetric features extracted from MRI are significantly enriched for information about the biological state of a tumor that impacts patient outcomes. Clinical decision-support systems could exploit this information to develop personalized treatment strategies on the basis of noninvasive imaging.

Keywords: Glioblastoma; Imaging-genomics; Neuro-imaging; Noninvasive; Pathways; Prediction; Radiation Oncology; Radiomics; Volumetric.

Fig. 1

Examples of volumetric tumor phenotype features. Glioblastoma (GBM) tumors show strong phenotypic differences, which can be objectively quantified with volumetrics. This figure shows examples of GBM tumors exhibiting high (top) and low (bottom) volumetric feature values for Necrosis, Contrast Enhancement, Edema, and Tumor Bulk (columns) as they appear on T1 weighted (columns 1,2, and 4) or T2-FLAIR (column 3) magnetic resonance images for different patients

Fig. 2

Volumetric phenotype features within the same tumor. Detailed example of a glioblastoma tumor in a patient. (a,b) On T1-weighted post-Gadolinium contrast (T1C) images, a central area of Necrosis is typically surrounded by a Contrast Enhancing ring, both of which can be derived from dark and light regions, respectively. Tumor Bulk represents the addition of these tumor features. (c) The Total Tumor Volume is represented by hyperintensity extracted from T2-FLAIR images. Edema is the difference of Tumor Bulk from Total Tumor Volume

Fig. 3

Correlation map. Pairwise Pearson correlation coefficients of volumetric features. Only few volumes were highly correlated (blue) or highly anti-correlated (anti-correlated)

Fig. 4

Pathway enrichment analysis. In total, 64 biological processes (rows) were significantly (FDR < 0.05) enriched for at least one volumetric feature (columns) as indicated by an asterisk. Heatmap shows normalized enrichment scores (NES) calculated with Gene Set Enrichment Analysis. Positive NES (blue) correspond to correlated pathways and negative NES (yellow) correspond to anti-correlated pathways

Fig. 5

Size distribution of volumetric tumor features across molecular subtypes of GBM. (a) Compared to the Total Volume, Edema had the largest median size across all molecular GBM subtypes. (b) Classical and neural tumors showed larger Edema areas than mesenchymal and proneural tumors. Size variation of volumetric feature areas other than Edema was generally low across subtypes

Fig. 6

Prognostic value of volumetric tumor features. Necrosis, Contrast Enhancement, Tumor Bulk, and Total Tumor Volume were significantly (asterisk) prognostic (p < 0.05). The Contrast Enhancement feature showed the highest prognostic performance as measured by the C-index