We used high-quality data from one population-based cancer registry in Yorkshire, UK to describe primary central nervous system tumours in teenagers and young adults (16-24 years) diagnosed between 1990 and 2009. The Birch classification scheme was used to identify differences by tumour subgroup. Incidence, patterns of care and survival trends were described using Poisson and Cox regression.
There were 163 cases comprising 98 astrocytomas, 17 ¡®other gliomas¡¯, 14 ependymomas, 11 medulloblastomas and 23 ¡®other intracranial and intraspinal neoplasms¡¯ yielding an overall incidence of 18.1 million person-years. Care varied significantly over time and by principal treatment centre (Leeds 77 % , Hull 23 % ), co-ordinating specialty (neurosurgery 53 % , clinical oncology 22 % , paediatrics 17 % , other adult services 8 % ) and treatment received. Cox regression showed no significant difference in survival by age, gender, treatment centre, level of deprivation, year of diagnosis or co-ordinating specialty, but a significant difference by tumour grade and diagnostic group. Survival improved for all diagnostic groups except astrocytoma, although only the medulloblastoma group showed a significant change over time.
The lack of any significant improvement in survival over time in most diagnostic groups warrants further investigation and provides justification for a more collaborative regional approach to the care of central nervous system tumours, perhaps through the development of regional guidelines for this unique population. More detailed analysis of relapse patterns and prediagnostic symptoms would also be informative for this cohort.