Drug discovery and preclinical testing of drug candidates for developmental and epileptic encephalopathies.

Abstract

Drug development for developmental and epileptic encephalopathies (DEEs) follows different strategies on one hand including disease-targeting precision medicine approaches considering the genetic variants and pathomechanisms in DEEs and on the other hand including therapeutic approaches with novel targets or second-generation drug candidates that may be of interest beyond selected DEEs. Although the first group of approaches can only be tested in dedicated DEE models, assessment in induced non-specific seizure and epilepsy models may provide valuable information if the mechanism of action implies a broader spectrum of efficacy. Data from such models can inform about general anti-seizure efficacy, efficacy against different seizure types including a possible broad-spectrum potential, dose range, and "therapeutic" plasma/brain concentrations. However, only dedicated DEE models will guide selection of the best candidates with a favorable efficacy and tolerability spectrum for specific DEEs. Several DEE models have already been used for preclinical testing of therapeutic approaches. Testing in these specific models can provide information about the effects on seizure generation; spread of seizure activity; epilepsy development; survival; behavioral, cognitive, and motor function; and about tolerability. On the other hand, we still face several limitations and challenges including lack of models for many DEEs, incomplete penetrance of the phenotype, high mortality, low throughput, limited knowledge concerning pharmacology and predictive validity, and species differences in development and disease course. In this review, we provide an overview of the preclinical efficacy data of approved orphan drugs in both model types and discuss the current state-of-knowledge concerning predictive validity. In conclusion, testing strategies need to be carefully tailored to the candidate drug or therapeutic approach. In this context, there is an urgent need for development of further specific DEE models and for a comprehensive characterization of the face and predictive validity of existing and future DEE models.