Following the internal pilot, the study did not meet prespecified stop/go criteria for continuation.

Rolandic epilepsy (RE) is the most common type of epilepsy. Children with RE have seizures and can often find that their learning, sleep, behaviour, self-esteem and mood are affected.

As part of standard NHS care, children diagnosed with RE may be treated with standard anti-epileptic medicines, like carbamazepine, or no medicine at all. The medicines used to treat epilepsy often slow down a child's thinking and learning. In the past, doctors believed this was an acceptable price to pay to reduce seizures. However, with RE, where the seizures usually stop in teenage years, investigators do not know if it is better to treat these children with medicines or not, especially if the medicines might have a negative effect on their learning.

A newer medicine called levetiracetam has also been found to work in children with RE and has shown less problems with thinking and learning in adults. However, it is still no known if this is also the case for children and it has not been proven which of the three options (carbamazepine, levetiracetam or no treatment) would be best for RE patients. The CASTLE study aims to find this out.

In addition, it has been found that seizures often happen when a child has had poor sleep and they often come at night or early in the morning. It has been shown that sleep can be improved through practice without the need of medicines. There are established guidelines to help toddlers go to sleep, but nothing available that helps young people with epilepsy and their parents improve their sleep quality. In the CASTLE study, a sleep training plan has been developed for children with epilepsy and the trial aims to find out whether following this sleep training plan results in less seizures than using no sleep training at all.

The trial is a phase IV randomised factorial design controlled trial comparing carbamazepine, levetiracetam or active monitoring combined with or without sleep behaviour intervention. A factorial trial design has been used as this approach enables the efficient simultaneous investigation of AED (carbamazepine; levetiracetam; no AED) and sleep behaviour intervention (vs standard care) by including all participants in both analyses. In a factorial trial it is also possible to consider both the separate effects of each intervention and the benefits of receiving both interventions together (for example levetiracetam and sleep intervention).

The CASTLE trial will take place in NHS out-patient paediatric epilepsy and general paediatric clinics in the United Kingdom (UK).

Once consent has been obtained from the appropriate adult, and assent from the child if appropriate, by the delegated member of the research team the eligibility assessments will be completed, full eligibility confirmed (confirmation must be by a medically qualified doctor) and baseline data will be collected prior to randomisation.

Randomisation will be performed via a web based tool accessed by research team at site. This system is generated centrally by the Clinical Trial Research Centre (CTRC) using a computer algorithm concealed from the investigators and research teams/trial management group. In order to balance the groups, minimisation for variables believed to influence disease outcome and end points will be built into the randomisation algorithm.

Participants will be randomised to treatment with carbamazepine, levetiracetam or active monitoring. Where randomised to drug treatment, the randomised treatment should ideally begin on the day of randomisation or within 14 days of randomisation at the latest. Randomised treatment will continue for a minimum of 12 months and a maximum of 48 months. All treatments will be procured, prescribed and issued as per routine NHS practice.

Clinical data capture will be in the form of paper copies of Case Report Forms (CRFs) that will be returned as an on-going process from each centre to the CTRC. Patient/parent reported data will be collected directly on paper at each outpatient visit with the exception of CANTAB, which will be collected on iPads at the centre.

All trial documents (except raw Hospital Episode Statistics (HES) from NHS digital that will only be retained for 1 year) will be retained for 25 years from the End of Trial. The PI at each investigational centre must make arrangements to store the essential trial documents, (as defined in Essential Documents for the Conduct of a Clinical Trial (ICH E6, Guideline for Good Clinical Practice)) including the ISF, until the CTRC informs the investigator that the documents are no longer to be retained

Inclusion Criteria:

1. Children diagnosed with RE (see International League Against Epilepsy Diagnostic Manual at https://www.epilepsydiagnosis.org/syndrome/ects-overview.html)
2. EEG showing focal sharp waves with normal background (see International League Against Epilepsy Diagnostic Manual at https://www.epilepsydiagnosis.org/syndrome/ects-eeg.html)
3. Aged ≥5 years and <13 years at the time of randomisation
4. Currently untreated with antiepileptic drugs
5. Written informed consent received from person with parental responsibility/legal representative.
6. Family have an email address and regular internet access (for online sleep intervention)
7. Parent and child are to have a good understanding of the English language

Exclusion Criteria:

1. Known contraindication to any of the trial drugs
2. Previously treated for epilepsy with antiepileptic drugs

The current data sharing plans for this study are unknown and will be available at a later date