Object This study was undertaken to evaluate the association between age at diagnosis, patterns of care, and outcome among elderly individuals with anaplastic astrocytoma (AA) and glioblastoma multiforme (GBM). Methods Using the Surveillance, Epidemiology and End Results database, the authors identified 1753 individuals with primary GBM and 205 individuals with primary AA (diagnosed between June 1991 and December 1999) who were 66 years and older and whose records were linked to Medicare information. To facilitate gathering of prediagnosis comorbidity and postdiagnosis treatment information, only those individuals were included who had the same Medicare coverage for 6 months before and 12 months after diagnosis. The odds of undergoing various combinations of treatments and the associations with outcome were calculated by tumor type and age and adjusted by various predictors. Results Age was not associated with treatment differences in individuals with AA. Very elderly individuals (≥ 75 years old) with GBM were more likely to have biopsy only (odds ratio [OR] 2.53, 95% confidence interval [CI] 1.78–3.59), surgery only (OR 1.47, 95% CI 1.15–1.87), or biopsy and radiation (OR 1.39, 95% CI 1.07–1.82) and were less likely to receive multimodal therapy. Regardless of patient age or lesion histological characteristics, survival was decreased in patients treated with biopsy only. Individuals with GBM who had surgery only or biopsy and radiation had worse outcomes than individuals treated with surgery and radiation. There were no differences in survival by lesion histological characteristics. Very elderly individuals with malignant astrocytomas were more likely to receive limited treatment (most pronounced in individuals with GBM). Survival variation correlated with treatment combinations. Conclusions These findings suggest that in clinical neurooncology patient age is associated with not receiving effective therapies and hence worse prognosis.

No abstract

Racial/ethnic differences in survival after primary glioblastoma diagnosis exist and may be partially explained by racial/ethnic differences in treatment and income.