Emory University, Atlanta, Georgia, USA
A golden age of antiepileptic drug development has yielded over a dozen useful new compounds, but the nature of clinical trials has made translation to practical use in the clinic difficult. Most clinical trials are designed for regulatory purposes and fail to answer critical clinical questions. These questions include: which drug is best as initial therapy, which drugs work as monotherapy, what are good drug combinations, what is the best starting dose and titration schedule, what is a reasonable target dose, what is the shape of the dose-response curve and does it vary significantly between patients, what is the true incidence of side effects, and what is the long-term efficacy of the drug? Most of these questions could be answered by changing trial designs, but many changes would entail additional time and money. There are encouraging signs that trials with procedures more directly applicable to the clinic are becoming common. These include direct comparative trials, longer trials with emphasis on seizure freedom, and trials with more flexible dosing schedules. In the past, funding of longer and more naturalistic trials has fallen to government agencies, but commercial funding has been obtained for several recent studies. Better quality control, innovative endpoints, structured searching for side effects, and standardisation of data collection are also promising topics for development.