MR Imaging Predictors of Molecular Profile and Survival: Multi-institutional Study of the TCGA Glioblastoma Data Set

Abstract

Purpose:To conduct a comprehensive analysis of radiologist-made assessments of glioblastoma (GBM) tumor size and composition by using a community-developed controlled terminology of magnetic resonance (MR) imaging visual features as they relate to genetic alterations, gene expression class, and patient survival. Materials and Methods:Because all study patients had been previously deidentified by the Cancer Genome Atlas (TCGA), a publicly available data set that contains no linkage to patient identifiers and that is HIPAA compliant, no institutional review board approval was required. Presurgical MR images of 75 patients with GBM with genetic data in the TCGA portal were rated by three neuroradiologists for size, location, and tumor morphology by using a standardized feature set. Interrater agreements were analyzed by using the Krippendorff a statistic and intraclass correlation coefficient. Associations between survival, tumor size, and morphology were determined by using multivariate Cox regression models; associations between imaging features and genomics were studied by using the Fisher exact test. Results:Interrater analysis showed significant agreement in terms of contrast material enhancement, nonenhancement, necrosis, edema, and size variables. Contrast-enhanced tumor volume and longest axis length of tumor were strongly associated with poor survival (respectively, hazard ratio: 8.84, P = .0253, and hazard ratio: 1.02, P = .00973), even after adjusting for Karnofsky performance score (P = .0208). Proneural class GBM had significantly lower levels of contrast enhancement (P = .02) than other subtypes, while mesenchymal GBM showed lower levels of nonenhanced tumor (P , .01). Conclusion:This analysis demonstrates a method for consistent image feature annotation capable of reproducibly characterizing brain tumors; this study shows that radiologists' estimations of macroscopic imaging features can be combined with genetic alterations and gene expression subtypes to provide deeper insight to the underlying biologic properties of GBM subsets.q RSNA, 2013

MR Imaging Characteristics Associate with Tumor-Associated Macrophages in Glioblastoma and Provide an Improved Signature for Survival Prognostication

“…29,37,38 Sharper margins and less contrast enhancement by MR imaging have been associated with features of ALTϩ tumors, including mutant IDH1, mutant tumor protein p53 (TP53), or proneural tumors. 13,16,18,20,39,40 Considering that most ALTϩtumors in this study had mutant IDH1 or IDH2, MR imaging features of ALT may be due to mutant IDH. The expansive T1/FLAIR VASARI feature had greater significance when tumors were compared based on IDH mutant status rather than telomere maintenance subtype.…”

“…13,[15][16][17][18][19] That tumor biology influences imaging features is further evident in studies that investigate subgroups of glioblastomas based on multiple gene expression differences. 20 This study aimed to determine whether MR imaging features were associated with the telomere maintenance mechanism and tumor-associated macrophage content-based subtypes. The inclusion of MR imaging parameters with the molecular subtyping and MGMT promoter methylation status 21 was also investigated to determine whether it could better predict outcome for patients with glioblastoma.…”

Whole Slide Imaging for Analytical Anatomic Pathology and Telepathology: Practical Applications Today, Promises, and Perils

“…68,69 Recently, using multimodal, multiscale approaches and machine-based classification, researchers have devised ways to mine scanned histologic data on glioblastoma in The Cancer Genome Atlas Project and other sources; the resulting quantitative morphometric analysis findings have been further integrated with molecular data to provide in silico cancer research [70][71][72][73][74][75][76][77][78][79][80] and with radiologic data to provide clinicopathoradiologic correlation. 81 Insights from this work also include findings on the importance of tumor-infiltrating lymphocytes in glioblastoma, 72 and in silico approaches from these studies have uncovered novel findings regarding the regulation of asymmetric cell division in glioblastoma by such mediators as the human Brat ortholog TRIM3. 80 …”

Multi-modal Glioblastoma Segmentation: Man versus Machine

“…Necrosis was defined as a region within the tumor that did not enhance and had a hyper-intense signal on T2w and FLAIR. Edema was classified on FLAIR, T1w, T2w and contrast-enhanced T1w [27]. The average time for manually segmenting all subcomponents of one study patient was approximately one hour.…”