Association of Frailty and Survival After Treatment in Older Patients with Glioblastoma
Sara Hardy1, Myla Strawderman1, Lauryn Hemminger1, Michelle Janelsins1, Sarah Cawley1, Jennifer Serventi2, Jacqueline Behr1, Michael Milano1, Allison Magnuson1, Nimish Mohile1
1University of Rochester, 2University Of Rochester
Objective:

We aimed to characterize frailty in patients ≥65 years old with glioblastoma using a deficit accumulation frailty index (DAFI) score and examine the association of frailty and treatment with overall survival (OS).

Background:

Older adults are a heterogeneous population. Frailty, the accumulation of deficits that increase vulnerability to stressors, may underly some of this heterogeneity. However, there are scant data on frailty in patients with glioblastoma.

Design/Methods:

We prospectively enrolled 30 patients with glioblastoma and calculated a DAFI score based on a 45-item scale. Using standard cutoffs, patients were categorized as non-frail, pre-frail, or frail. Patients were treated per multidisciplinary consensus and followed every 2 months. OS (months from diagnosis) was compared between groups using the log-rank test. Cox regressions assessed multiple predictors for survival. Model fit was described by AIC (smaller is better).

Results:

For all (median age 73.9, median follow up 11.1 months), median OS with conventional chemoradiation was 17.5 months, hypofractionated chemoradiation was 14.0 months, and radiation alone was 7.0 months (p=0.0002). Median OS was 17.5 months for non-frail patients (n=15), 11.2 months for pre-frail patients (n=11), and 7.9 months for frail patients (n=4, p=0.008). In non-frail patients, median OS for conventional chemoradiation was not reached, hypofractionated chemoradiation was 15.0 months, and radiation alone was 9.7 months (p=0.04). Non-frail patients (40% MGMT methylated) treated with chemotherapy (temozolomide) had a median OS of 25 vs 9.7 months without chemotherapy (p=0.02). Models using frailty alone predicted survival (AIC= 99) better than a combined model using KPS, MGMT methylation, and age (AIC= 112).

Conclusions:

Non-frailty was associated with longer survival; in older patients receiving conventional chemoradiation, survival approached that of younger patients. Frailty also predicted survival better than a combined model using KPS, MGMT methylation, and age. Frailty may be a superior to age and KPS in guiding decisions regarding glioblastoma treatment.

10.1212/WNL.0000000000205451