Early imaging marker of progressing glioblastoma: a window of opportunity

Abstract

Purpose: Therapeutic intervention at glioblastoma (GBM) progression, as defined by current assessment criteria, is arguably too late as second-line therapies fail to extend survival. Still, most GBM trials target recurrent disease. We propose integration of a novel imaging biomarker to more confidently and promptly define progression and propose a critical timepoint for earlier intervention to extend therapeutic exposure.

Methods: A retrospective review of 609 GBM patients between 2006 and 2019 yielded 135 meeting resection, clinical, and imaging inclusion criteria. We qualitatively and quantitatively analyzed 2000+ sequential brain MRIs (initial diagnosis to first progression) for development of T2 FLAIR signal intensity (SI) within the resection cavity (RC) compared to the ventricles (V) for quantitative inter-image normalization. PFS and OS were evaluated using Kaplan-Meier curves stratified by SI. Specificity and sensitivity were determined using a 2 × 2 table and pathology confirmation at progression. Multivariate analysis evaluated SI effect on the hazard rate for death after adjusting for established prognostic covariates. Recursive partitioning determined successive quantifiers and cutoffs associated with outcomes. Neurological deficits correlated with SI.

Results: Seventy-five percent of patients developed SI on average 3.4 months before RANO-assessed progression with 84% sensitivity. SI-positivity portended neurological decline and significantly poorer outcomes for PFS (median, 10 vs. 15 months) and OS (median, 20 vs. 29 months) compared to SI-negative. RC/V ratio ≥ 4 was the most significant prognostic indicator of death.

Conclusion: Implications of these data are far-reaching, potentially shifting paradigms for glioma treatment response assessment, altering timepoints for salvage therapeutic intervention, and reshaping glioma clinical trial design.

Keywords: FLAIR signal intensity (SI); Imaging biomarker; Neurologic Assessment in Neuro-Oncology (NANO); Progressed glioblastoma; Response Assessment in neuro-Oncology (RANO); Signal Assessment in Neuro-Oncology (SANO).