Aims To describe the presentation of epileptic seizures, the tumour-related risk factors, seizure treatments used and the timing of withdrawal of anti-epileptic drugs (AEDs).

Methods Retrospective case note review of 120 newly diagnosed brain tumour patients referred between 01/2010 and 12/2014 to the regional paediatric cancer service was carried out to determine patient, tumour, seizure characteristics and their treatment and outcomes. Statistical analyses using Fischer Test, Mann-Whitney U test and logistic regression test of association between variables were performed.

Results Two patients with tuberous sclerosis and associated seizures and 1 patient lost to follow up, were excluded. 67/117 (57.3%) were male, median age at tumour presentation was 8.1 years (IQR ^25°–75°: 3.6–12.7). Anatomical distribution was posterior fossa 36%, cerebral hemisphere 28%, supratentorial midline 24%, spinal cord 5%, metastatic 7%. Most common histological types include pilocytic astrocytoma 15%, visual pathway glioma 8%, other low grade astrocytoma 11%; high grade glioma 10%, medulloblastoma 11%.

With a median follow up of 33 months (IQR ^25°–75°: 24–56), there were 35 patients (29%) with seizures. Logistic regression analysis identified cerebral hemisphere localization as risk factor for seizures (OR: 7.07 IC 95% 2.89–17.3). In particular, neither age at diagnosis nor tumour type were associated with seizure risk, although 6/8 patients with glioneuronal tumours developed seizures.

With a median follow up of 24 months (IQR^25°-75°: 15–48) 37% were seizure free (SF) off AEDs, 43% were SF on AEDs and 20% experienced continuing seizures. One patient was untreated, seizures resolved.

12 patients withdrew AEDs, median duration of AED therapy before withdrawal was 11 months (IQR^25°–75° 5–14 months), and the median follow up after withdrawal was 15 months (IQR^25°–75° 5–34 months). 4 children had seizure relapse after further acute events (2 had acute hydrocephalus, 1 had sepsis, 1 had ventriculitis after shunt revision).

Conclusions Epileptic seizures affect about 1/3^rd of children presenting with brain tumours, most commonly in cortically-located tumours. The low risk of recurrent seizures after treatment justifies early withdrawal of AED after stabilisation of seizures. A 3 month seizure free interval prior to AED withdrawal is proposed as a service evaluation target for future studies.

Acknowledgement: on behalf of the East Midlands Children and Young People’s Integrated Cancer Service (EMCYPICS) and University of Nottingham Children’s Brain Tumour Research Centre (www.cbtrc.org)