Parametric Response Mapping of FLAIR MRI Provides an Early Indication of Progression Risk in Glioblastoma

Benjamin A Hoff; Benjamin Lemasson; Thomas L Chenevert; Gary D Luker; Christina I Tsien; Ghoncheh Amouzandeh; Timothy D Johnson; Brian D Ross

doi:10.1016/j.acra.2020.08.015

Parametric Response Mapping of FLAIR MRI Provides an Early Indication of Progression Risk in Glioblastoma

Acad Radiol. 2021 Dec;28(12):1711-1720. doi: 10.1016/j.acra.2020.08.015. Epub 2020 Sep 11.

Authors

Benjamin A Hoff¹, Benjamin Lemasson², Thomas L Chenevert¹, Gary D Luker¹, Christina I Tsien³, Ghoncheh Amouzandeh¹, Timothy D Johnson⁴, Brian D Ross⁵

Affiliations

¹ Department of Radiology, the Center for Molecular Imaging, University of Michigan, Ann Arbor, Michigan.
² Université Grenoble Alpes, Grenoble Institut Neuroscience, Inserm, U1216, 38000 Grenoble, France.
³ Department of Radiation Oncology and Molecular Radiation Sciences, Johns Hopkins University, Baltimore, Maryland.
⁴ Department of Biostatistics, University of Michigan, Ann Arbor, Michigan.
⁵ Department of Radiology, the Center for Molecular Imaging, University of Michigan, Ann Arbor, Michigan. Electronic address: bdross@med.umich.edu.

Abstract

Rationale and objectives: Glioblastoma image evaluation utilizes Magnetic Resonance Imaging contrast-enhanced, T1-weighted, and noncontrast T2-weighted fluid-attenuated inversion recovery (FLAIR) acquisitions. Disease progression assessment relies on changes in tumor diameter, which correlate poorly with survival. To improve treatment monitoring in glioblastoma, we investigated serial voxel-wise comparison of anatomically-aligned FLAIR signal as an early predictor of GBM progression.

Materials and methods: We analyzed longitudinal normalized FLAIR images (rFLAIR) from 52 subjects using voxel-wise Parametric Response Mapping (PRM) to monitor volume fractions of increased (PRM_rFLAIR+), decreased (PRM_rFLAIR-), or unchanged (PRM_rFLAIR0) rFLAIR intensity. We determined response by rFLAIR between pretreatment and 10 weeks posttreatment. Risk of disease progression in a subset of subjects (N = 26) with stable disease or partial response as defined by Response Assessment in Neuro-Oncology (RANO) criteria was assessed by PRM_rFLAIR between weeks 10 and 20 and continuously until the PRM_rFLAIR+ exceeded a defined threshold. RANO defined criteria were compared with PRM-derived outcomes for tumor progression detection.

Results: Patient stratification for progression-free survival (PFS) and overall survival (OS) was achieved at week 10 using RANO criteria (PFS: p <0.0001; OS: p <0.0001), relative change in FLAIR-hyperintense volume (PFS: p = 0.0011; OS: p <0.0001), and PRM_rFLAIR+ (PFS: p <0.01; OS: p <0.001). PRM_rFLAIR+ also stratified responding patients' progression between weeks 10 and 20 (PFS: p <0.05; OS: p = 0.01) while changes in FLAIR-volume measurements were not predictive. As a continuous evaluation, PRM_rFLAIR+ exceeding 10% stratified patients for PFA after 5.6 months (p<0.0001), while RANO criteria did not stratify patients until 15.4 months (p <0.0001).

Conclusion: PRM_rFLAIR may provide an early biomarker of disease progression in glioblastoma.

Keywords: FLAIR-MRI; Glioma; PRM; Progression; Response.

Publication types

Research Support, N.I.H., Extramural

MeSH terms

Brain Neoplasms* / diagnostic imaging
Contrast Media
Disease Progression
Glioblastoma* / diagnostic imaging
Humans
Magnetic Resonance Imaging
Neoplasm Recurrence, Local
Retrospective Studies

Substances

Contrast Media

Abstract

Publication types

MeSH terms

Substances

Grants and funding