EVALUATION OF ANTIEPILEPTIC DRUG EFFICACY - A REVIEW OF CLINICAL-TRIAL DESIGN

Up to 30% of patients with epilepsy are not adequately treated with cu rrently available antiepileptic drugs. Despite a need for new agents, few were developed after valproic acid (sodium valproate), which appea red in the 1970s. This picture has changed recently. A number of new a ntiepileptic drugs have been approved for marketing, and additional ap provals are expected soon. The evidence in the form of adequate and we ll controlled studies supporting the efficacy of each of these new dru gs has included (and in some cases been wholly composed of) placebo-co ntrolled add-on trials, showing the viability of that design. Add-on t rials, which compare a new drug with placebo in the presence of a stab le regimen of antiepileptic drug therapy, can be conducted as parallel or crossover designs. These designs have been considered insensitive because they are conducted in patients refractory to available antiepi leptic drugs, but add-on trials have proved able to identify effective new drugs. They also permit long term evaluation and provide informat ion, including drug interaction data, in one of the major clinical con texts where a new antiepileptic drug may be used. New designs now allo w the evaluation of antiepileptic effectiveness in other clinical cont exts, including monotherapy, and provide alternatives when drug intera ctions obscure add-on trial interpretation. The key feature of this cl ass of monotherapy designs is basing patients' trial duration on their seizure activity rather than on a fixed time period. This is accompli shed by defining 'therapeutic failure' criteria which serve to assess efficacy of the test agent while protecting patient safety. It is hope d that investigators will continue to seek innovative de signs to yiel d information to guide the clinical use of these new antiepileptic dru gs.