THE MANAGEMENT OF EPILEPSY IN THE 1990S - ACQUISITIONS, UNCERTAINTIESAND PRIORITIES FOR FUTURE-RESEARCH

The pharmacological treatment of epilepsy has made considerable progre ss during the last decade, due to improved knowledge of the clinical p harmacology of individual drugs, acquisition of new information on the factors affecting response and need for drug treatment, and developme nt of promising new agents. Once a clinical diagnosis of epilepsy has been made (which generally requires the occurrence of more than one se izure), treatment should be started with a single drug selected on the basis of seizure type and tolerability profile. Although there are im portant regional differences in prescribing patterns and individual ci rcumstances may dictate alternative choices, carbamazepine is generall y regarded as the preferred treatment for partial seizures (with or wi thout secondary generalisation) while valproic acid (sodium valproate) is usually the first choice in most forms of generalised epilepsies. To achieve therapeutic success, the daily dosage must be tailored to m eet individual needs, and there is suggestive evidence that in some pa tients the dosage prescribed initially may be unnecessarily large. Pla sma antiepileptic drug concentrations may aid in the individualisation of dosage, but should not be regarded as a substitute for careful mon itoring of clinical response. Although overall about 70% of patients c an be completely controlled, response rate is influenced by a number o f factors, the most important of which are seizure type and syndromic form. The importance of a correct syndromic classification for rationa l drug selection has been poorly assessed and represents a major area for future research. Patients who do not respond to the highest tolera ted dose of the initially prescribed drug maybe switched to monotherap y with an alternative agent or may be given add-on treatment with a se cond drug. Appropriate prospective trials are required to assess the m erits of either strategy. If add-on therapy is selected and the patien t becomes seizure free, it may be possible to discontinue the drug pre scribed initially and reinstitute monotherapy. Only a minority of pati ents are likely to require multiple drug therapy, and it remains to be established whether specific drug combinations are more effective tha n others. Until further information becomes available, the new agents should be reserved for patients failing to respond to the conventional treatments of first choice. Patients whose seizures cannot be control led by available drugs should be reassessed, and polytherapy should be maintained only when there is clear evidence that benefits outweigh p ossible adverse effects. In many patients who have been seizure free f or at least 2 years it may be possible to gradually discontinue all me dications. The decision to withdraw treatment is determined largely by the risk of seizure relapse which, in turn, is primarily dependent on the syndromic form.