HOUT-21. REAL-WORD EVALUATION OF THE IMPACT OF RADIOTHERAPY AND CHEMOTHERAPY IN ELDERLY PATIENTS WITH GLIOBLASTOMA BASED ON AGE AND PERFORMANCE STATUS: A NATIONAL CANCER DATABASE ANALYSIS

Abstract

OBJECTIVES: This retrospective study evaluates the impact of Karnofsky performance status (KPS) in addition to age on the management and outcomes of elderly patients with glioblastoma (GBM). METHODS: The National Cancer Database (NCDB) was queried between 2004-2015 for GBM patients with age 60 and reported KPS. Four groups were created: Age 60/ KPS<70 (Gp1), Age 60-69/KPS 70 (Gp2), age 70-79/KPS 70 (Gp3), age 80/KPS 70 (Gp4). Univariate (UVA) and multivariable (MVA) modeling with Cox regression determined predictors of survival (OS), and propensity score-matched analysis was performed. RESULTS: A total of 30,530 patients were included. Median age at diagnosis was 69, and median followup was 7.3 months. Median survivals were 5.2, 11.3, 6.3 and 3.7 months in groups 1, 2, 3 and 4, respectively (p <0.001). Combined chemotherapy (CT) and radiotherapy (RT) was received in 65.5% of Gp2, 53.7% of Gp3, 48.2% of Gp1 and 33.1% of Gp4 (p<0.001). On both UVA and matched OS analysis, all patients benefited from combined CT and RT, except those in Gp1, who did better with CT alone (p<0.01). RT alone was associated with the worst OS in all groups (p<0.01). Predictors of worse OS on MVA were greater age, lower KPS, white ethnicity, higher comorbidity score, worse socioeconomic status, treatment in a community center, tumor multifocality, subtotal resection, non-IMRT use, and no adjuvant treatment (all p<0.01). CONCLUSIONS: Based on this NCDB review, patients with GBM aged 60 with poor KPS have better survival following CT alone, whereas elderly, even those 80 years old, with good KPS fared best with combined CT and RT. RT alone resulted in worse OS compared to CT or combined CT and RT. Functional status is an independent, key prognostic factor that should be incorporated with age for management decisions in the elderly GBM population.