Survival analysis for adult glioma in England and Wales.

Abstract

To investigate factors influencing survival for adult glioma at the national population level, data from the Cancer Registry in England and Wales were analyzed. Glioma was diagnosed in 32,267 patients during 1971-1990 and follow-up data were available for these patients until 1995. Median survival and crude survival rates (CSR) for 8 variables (age, gender, International Classification of Diseases for Oncology [ICD-O] morphology, World Health Organization [WHO] grade, tumor site, deprivation, geographical region, and period of diagnosis) were calculated and tested using the Kaplan-Meier method and the log-rank test. Relative survival rates (RSR) were calculated using the life table of 1981. Cox multivariate regression was performed for estimating hazards ratios (HR) and tested using the log likelihood ratio test. The median survival and the 1-, 5-, and 10-year CSR for this population were 0.42 years, 29.1%, 12.0%, and 7.7%, respectively. The 1- and 5-year RSR were 29.6% and 12.3%, respectively. Survival was influenced by all 8 variables tested (p </= 0.025). Age at diagnosis was the most influential factor (HR, 1.22/5 years). Females had better survival than males (HR, 0.94). Patients with unspecified gliomas had the worst survival, while those with ependymoma had the best survival. Tumors in the posterior fossa were associated with better survival than those in the cerebral hemisphere. Survival was associated with the WHO grade and the level of deprivation, and had regional differences. Survival improved by 16% between 1971-1975 and 1986-1990. Survival in adult glioma is influenced by multiple factors, including socioeconomic background. However, age at diagnosis and tumor characteristics are the most influential factors.