Session Item

Glioblastoma (GBM) is a highly aggressive primary brain tumor that portends poor prognosis even with multimodality treatment. Maximally safe surgical resection is often aided by magnetic resonance-diffusion tensor imaging (MR-DTI) to navigate tumor-associated white matter (TA-WM) tracts and spare eloquent areas. We hypothesized that (1) TA-WM tracts harbor microscopic disease not targeted through surgery or radiotherapy (RT), and (2) the greater the extent of TA-WM involvement, the worse the survival outcomes. We accordingly investigated the prognostic significance of TA-WM tracts in GBM utilizing MR-DTI. 

We studied a retrospective cohort of 76 GBM patients, all of whom underwent pre-operative MR-DTI followed by maximally safe resection and standard-of-care adjuvant therapy. TA-WM tracts were identified by MR-DTI fractional anisotropy (FA) maps. For each patient, a tumor-associated white matter score (TA-WMS) was computed based on number of involved tracts and corresponding FA grade of involvement (0 = no involvement, 1 = displacement, 2 = partial disruption, 3 = complete disruption). Kaplan-Meier statistics were utilized to determine survival outcomes, and stratify results by molecular markers and TA-WMS. Log-rank test was used to compare survival between groups. Cox proportional hazards regression was utilized to determine prognostic variables. 

The median age was 61 years (range 28 – 77 years). 50 patients were male and 26 were female. All tumors were IDH-wildtype WHO Grade IV GBM; 51 tumors were MGMT- while 25 were MGMT+. All patients received adjuvant RT (median dose 60Gy) along with concurrent and adjuvant temozolomide (or other systemic therapy on clinical trial). The median number of TA-WM tracts was 3 (MGMT- range 0-8; MGMT+ range 0-10), and median TA-WMS was 6 (MGMT- range 0-14; MGMT+ range 0-20). Median overall survival (OS) was 14.1 months for the MGMT- cohort and 36.1 months for the MGMT+ cohort (p=0.0001). For the MGMT- cohort, there was a statistically significant decrease in OS for increasing TA-WMS (median OS 16.5 months for TA-WMS 0-4; 13.6 months for TA-WMS 5-8; 7.3 months for TA-WMS >9; p=0.0002, Fig. 1). This trend was not observed in the MGMT+ cohort (median OS 37.0 months for TA-WMS 0-4; 31.6 months for TA-WMS 5-8; 41.0 months for TA-WMS >9; p=0.6297). For MGMT- patients with TA-WMS >6 and involvement of corticospinal/corticobulbar tracts, corpus callosum, anterior commissure and/or bilateral involvement of any tracts, median OS was 8.3 months versus median OS 14.1 months with TA-WMS >6 but not involving aforementioned critical tracts (p=0.003, Fig. 2). For MGMT- patients, TA-WMS was predictive of overall survival in multivariate analysis (HR = 1.14, 95% CI 1.03 – 1.27, p=0.012) while age, gender, and largest tumor dimension were non-significant. 

Increased TA-WMS results in decreased overall survival in MGMT- but not MGMT+ glioblastoma. This bears significance toward RT field design in MGMT-unmethylated GBM patients.