Survival analysis of 81 children with primary spinal gliomas: a population-based study

Primary spinal gliomas are rare. Most clinical studies are based on single centers with small numbers of patients and limited length of follow-up. Because data from the Cancer Registry cover larger numbers of patients and longer durations of follow-up, our objective was to define prognostic factors that might predict the survival at a national population level. From 1971 to 1995, data of 81 children (age < 15 years) with primary spinal gliomas from the Cancer Registry of England and Wales were analyzed. Median survival and crude survival rates in respect of 7 variables (age, sex, morphology, WHO grade, socioeconomic status, geographical region, and period of diagnosis) were calculated using the Kaplan-Meier method. The Cox regression was performed for estimating hazard ratios (HR) for death. Results showed that the 1-, 5-, and 10-year crude survival rates for this population were 72.84, 60.49, and 58.0%, respectively. Both univariate and multivariate analyses revealed that only morphology (HR 2.79 for nonependymoma, p = 0.05) and WHO grade (HR 6.74 for high grade, p = 0.01) were significant prognostic factors. Results from this population-based study are very helpful for comparison with other population-based studies and for public health purposes.