Abstracts

There are an increasing number of antiepileptic drugs (AEDS) to choose from for initial monotherapy. Reliable evidence that informs a choice among AEDS comes from randomized controlled trials (RCTS) in which AEDs are compared head to head. Trials which compare the same AEDs (i.e. make a direct comparison) can be summarised in a meta-analysis. Indirect comparisons can also be made, for example, data from trials comparing drugs A and B, and trials comparing drugs A and C can be used to estimate how drugs A and C compare. Indirect comparisons can be valuable where direct comparisons either do not exist, comprise a limited amount of data of low quality. The aim of this project was to make the best use of existing data regarding comparative effects of AEDs in order to better inform clinical practice. We undertook a meta-analysis using individual patient data from 18 AED monotherapy RCTs representing 4496 patients. These trials made comparisons among carbamazepine, lamotrigine, oxcarbazepine, phenytoin, phenobarbital and valproate. Direct and idirect comparisons were summarised to provide a total summary of available evidence. Outcomes were time to treatment failure, time to 12 month remission from seizures and time to a first seizure. Selected results are expressed below as Hazard ratios (HR) with 95% confidence limits (95% CI). Results from direct and indirect comparisons are generally consistent. For time to first seizure, results consistently suggest valproate superior to other AEDs for generalized onset seizures (VPS- CBZ 1.21 (0.97-1.50), VPS-PB 1.29 (0.92-1.80), VPS-PHT 0.97 (0.76-1.24), VPS-LTG 1.17 (0.81-1.68), VPS-OXC 1.22 (0.79-1.98) and inferior for focal onset seizures (VPS-CBZ 0.81 (0.70-0.94), VPS-PB 0.61 (0.47-0.79), VPS-PHT 0.82 (0.67-1.00), VPS-LTG 1.11 (0.83-1.49), VPS-OXC 0.72 (0.51-1.02). For lamotrigine there is no clear trend for generealised onset seizures but it is inferior for partial onset seizures. Phenobarbitone is consistently superior to other AEDs for partial onset seizures and inferior for generalized onset seizures. For phenytoin there is no clear trend for partial onset seizures but it is superior to other AEDs for generalized onset seizures. Carbamazepine was not clearly superior or inferior to other drugs for parital or generalized onset seizures. This data work summarises the totality of evidence from RCTs regarding the comparative effects of AEDs when used as monotherapy. Results will inform clinical decision making as well as future RCTs. (Supported by NHS R[amp]D UK)