Presence of Fragmented Intratumoral Thrombosed Microvasculature in the Necrotic and Peri-Necrotic Regions on SWI Differentiates IDH Wild-Type Glioblastoma From IDH Mutant Grade 4 Astrocytoma.

Abstract

Background: Isocitrate dehydrogenase (IDH) wild-type (IDH_wt) glioblastomas (GB) are more aggressive and have a poorer prognosis than IDH mutant (IDH_mt) tumors, emphasizing the need for accurate preoperative differentiation. However, a distinct imaging biomarker for differentiation mostly lacking. Intratumoral thrombosis has been reported as a histopathological biomarker for GB.

Purpose: To evaluate the fragmented intratumoral thrombosed microvasculature (FTV) signs on susceptibility-weighted imaging (SWI) for distinguishing IDH_wt and IDH_mt tumors.

Study type: Retrospective.

Subjects: Ninety-seven treatment-naïve patients with histopathologically confirmed IDH_wt GB (54 males, 26 females) and IDH_mt grade 4 astrocytoma (13 males, 4 females).

Field strength/sequence: 3-T, SWI, fluid-attenuated-inversion-recovery (FLAIR), T₁-weighted, T₂-weighted, PD-weighted, post-contrast T₁-weighted and dynamic-contrast-enhanced (DCE)-MRI.

Assessment: SWI data were evaluated by three experienced neuroradiologists (S.S., 11 years; J.S., 15 years; R.K.G., 40 years of experience), who assessed FTV presence in necrotic and peri-necrotic regions. FTV was identified as intratumoral susceptibility signal having minimal or no interslice connections. Quantitative DCE-MRI parameters were derived using first-pass-analysis and extended Tofts model. FLAIR abnormal, contrast-enhancing, and necrotic regions were segmented using in-house developed U-Net architecture.

Statistical tests: Fleiss' Kappa, Cohen's Kappa, Shapiro-Wilk test, t tests or Mann-Whitney U test, receiver-operating characteristic (ROC) analysis, confusion matrix. A P-value <0.05 was considered statistically significant.

Results: Fleiss' kappa test provided 91% inter-rater agreement, and Cohen's kappa provided intrarater agreement ranged from 81% to 97%. The raters' accuracy in distinguishing IDH_wt from IDH_mt ranged from 92% to 94%. Some of the quantitative DCE-MRI parameters (CBV, Ve, and K^trans) provided statistically significant differences in differentiating IDH_wt and IDH_mt. K^trans demonstrated 80.3% sensitivity and 81.2% specificity, with ROC analysis showing an AUC of 0.77.

Data conclusion: FTV signs in necrotic and peri-necrotic regions on SWI demonstrated a high accuracy in distinguishing IDH_wt from IDH_mt. Qualitative assessment of FTV signs showed almost perfect inter-rater and intrarater agreement. Quantitative DCE-MRI metrics also showed statistically significant differentiation of IDH_wt and IDH_mt.

Plain language summary: This study demonstrates that preoperative imaging, particularly the visualization of the fragmented thrombosed vasculature (FTV) sign on susceptibility-weighted imaging (SWI), effectively differentiates isocitrate dehydrogenase (IDH) wild-type (IDH_wt) glioblastoma (GB) from IDH mutant (IDH_mt) grade 4 astrocytomas. Over 90% of IDH_wt GB patients displayed the FTV sign, a specific imaging biomarker absent in IDH_mt cases. Perfusion parameters such as cerebral blood volume, Ve, and K^trans were elevated in IDH_wt gliomas, reflecting distinct vascular profiles. SWI offers a noninvasive and accurate diagnostic method, overcoming limitations of histopathology. Despite limitations like unequal sample sizes and retrospective analysis, this study underscores the clinical potential of SWI in improving glioma characterization and aiding treatment planning.

Level of evidence: 4 TECHNICAL EFFICACY: Stage 2.