Short Versus Long-term Levetiracetam in Brain Tumors: A Phase 3 Randomized Controlled Trial (LIBRA)

Patients with prior history of seizure from primary brain tumor in the supratentorial location with controlled seizure on levetiracetam monotherapy for at least six months will be considered for the study. Patients more than 18 years of age with KPS ≥ 50 will be eligible. Patients will be randomized in one of the two arms (standard arm or experimental arm) in a 1:1 ratio and stratified based on seizure type, location, histology, tumor grade, and adjuvant therapy. Randomization will be done by the statistician via computerized software using a permuted block design. In the standard arm, patients will continue on the same dose and schedule of levetiracetam (typically prescribed in the range of 1000-3000 mg/ day in 2-3 divided doses) for a duration of 2 years. In the experimental arm, levetiracetam will be tapered by 250- 500 mg every week and stopped. Follow-ups will be done every 3-6 months as per standard practice for the given tumor histology. Neuroimaging will be done 6-12 monthly as per routine clinical practice. The quality-of-life assessment will be done every six months. The primary endpoint is 2-year seizure free survival calculated from the time of randomization.

Patients will continue to receive standard treatment, including adjuvant therapy as standard practice. In case in either arm, the patient develops a seizure episode after stopping levetiracetam will be restarted on levetiracetam monotherapy. If a patient develops a seizure episode while on levetiracetam monotherapy, further add-on antiepileptics will be considered as per standard practice by the responsible physician. Any complications arising from previous treatments (e.g., radio necrosis) or recurrent disease during the study period will be managed according to standard institutional practice without any influence of the study.

An interim analysis is planned for harm or futility to stop the trial when quarter of the expected events have occurred (25% information time), if the observed HR is equal to or greater than the non-inferiority margin (1.47).

At the interim analysis, the one-sided P-value is calculated for testing the hypothesis HR = 1 versus the alternative HR > 1.73 (the experimental treatment is doing worse than the standard treatment). If the P-value is <0.0110 at the interim analysis, the trial would stop with the conclusion that non-inferiority cannot be claimed.

The study will be conducted at Tata Memorial Centre and Manipal Academy of Higher Education with a total sample size of 604 patients for a duration of seven years.