Introduction
Access to clinical trials is a common request for patients with brain tumours. However, opening clinical trials requires additional work per centre opened. We have previously shown that surgical and oncology workload varies between centres, and fluctuates over time. There is a trade-off between offering access to clinical trials and increasing costs associated with opening trials in centres that treat few patients.

Methods
We used two separate datasets from England covering 3 years – one for neurosurgical workload and one for radiotherapy. We only included adult patients and calculated cumulative proportions of the malignant primary brain tumour population (C71) by number of centres. We investigated stability by checking how many patients would have to be added/ removed from a centre to change their rank.

Results
There were 7061 surgical and 5060 radiotherapy patients. To capture 25% of patients, we would need to open trials in 4 surgical/5 radiotherapy centres; for 50%, 9 surgical/ 13 radiotherapy centres; for 75%, 16 surgical/ 24 radiotherapy centres. Centre rank was fluid: adding 16 surgical/9 radiotherapy patients would change the rank of a centre.

Discussion
These are the first data to allow for rational planning of trials in brain tumour patients. We have shown that we can reach 75% of the brain tumour population by opening trials in ~50% of surgical and radiotherapy centres. Centre rank alters over year, so we should be cautious about being too prescriptive. Nonetheless, these data should allow some rational planning of trial centre inclusion.