ASL-based Exchange Time Mapping for Differentiation of High-Grade Glioma and Brain Metastases
Abstract Background To evaluate whether blood-brain barrier (BBB) mapping via arterial spin-labeling (ASL)-derived exchange time (Tex) can differentiate high-grade gliomas from brain metastases and to compare regional BBB alterations with dynamic susceptibility contrast (DSC)-derived leakage parameters (K2). Methods 18 patients with therapy-naive cerebral masses (11 gliomas, 7 metastases) underwent multi-echo-ASL and DSC perfusion MRI. Tex maps were obtained using a validated extended two-compartment model. K2 maps were derived from single-echo DSC. Tex and absolute K2 (|K2|) were quantified in contrast-enhancing tumor (CET), peritumoral T2/FLAIR hyperintense region (PTR), gray matter (GM), and white matter (WM). Regional differences were assessed using Wilcoxon signed-rank and Mann–Whitney U tests. Receiver operating characteristic (ROC) analyses evaluated discrimination between gliomas and metastases. Results Tex was significantly lower in PTR of gliomas versus metastases (p = 0.046). In ROC analysis, Tex slightly outperformed |K2| alone (AUC=0.792, 95% CI [0.535, 0.986] vs. AUC=0.760, 95% CI [0.508, 0.959]), with further improvement upon combination (AUC=0.833, 95% CI [0.577, 1.000]). In gliomas, Tex was significantly lower in CET than in GM and WM (both p < 0.01) and PTR showed significantly lower Tex than WM (p < 0.01). |K2| was significantly higher in CET than all other regions (all p < 0.01) but did not differ between PTR and GM/WM. Conclusion In this exploratory study, ASL-derived Tex suggests stronger BBB alterations in peritumoral tissue of high-grade gliomas compared with brain metastases. Adding |K2| further improved discrimination. This supports ASL-based BBB mapping as a complement to DSC leakage imaging. Validation in larger multicenter cohorts is warranted.
Authors
- Benedikt Wiestler (ORCID: https://orcid.org/0000-0002-2963-7772)
- Gabriel Hoffmann (ORCID: https://orcid.org/0000-0002-5233-7532)
- Amnah Mahroo (ORCID: https://orcid.org/0009-0005-8480-0965)
- Philipp Raffler (ORCID: https://orcid.org/0009-0006-7269-4998)
- Stephan Kaczmarz (ORCID: https://orcid.org/0000-0001-7694-7012)
- Matthias J P van Osch
- Lena Václavů
- Matthias Günther
Institutions
- Leiden University (NL)
- University Health Network (CA)
- University of Bremen (DE)
- Leiden University Medical Center (NL)
- TUM Klinikum (DE)
- Mediri (Germany) (DE)
- Fraunhofer Institute for Digital Medicine (DE)
- Ontario Brain Institute (CA)
- Technical University of Munich (DE)
Publication Details
- Journal
- Neuro-Oncology Advances
- Published
- 2026-07-23
- DOI
- https://doi.org/10.1093/noajnl/vdag195
- Primary Topic
- Advanced MRI Techniques and Applications
- Type
- article
- Field-Weighted Citation Impact
- 0.00