Proof of principle trials: EEG surrogate endpoints

Abstract

There is no doubt that the pathophysiological processes underlying the generation of epileptic spikes in the EEG are the same as those involved in clinical seizures, and therefore a drug which suppresses seizures is likely to do the same to EEG spikes. Indeed, phenytoin, the benzodiazepines, valproate and lamotrigine have been shown to do so when tested acutely in single doses in patients with frequent interictal spikes (Jawad et al., 1986) and it is reasonable to regard such studies as a useful method for screening new drugs for antiseizure activity. Also, the coincidence between spike-and-wave activity and clinical absences has made EEG monitoring an essential part of the short and long term evaluation of anti-absence drugs. In long term trials of drugs in partial and secondary generalised tonic clonic seizures, however, there is sometimes a discordance between interictal spike frequency and clinical seizures, and monitoring the EEG in such studies has proved to be unrewarding. EEG responses to photic stimuli have also been shown to be inhibited by many antiepileptic drugs, even those that do not have a role in patients who have photosensitive epilepsy (Binnie et al., 1986a). We are concerned here particularly with acute studies of interictal spikes and photic responses. Despite the evidence supporting their value, these surrogate measures of seizure activity have not been widely used in the early evaluation of new antiepileptic drugs, one exception being lamotrigine, which underwent assessment in both interictal spikes (Jawad et al., 1986) and photosensitivity (Binnie et al., 1986b). This is presumably because most advisers to the pharmaceutical industry consider that the acid test of a new drug is whether or not it suppresses seizures, and the earlier this is demonstrated, the better. However, definitive proof-of-efficacy trials in patients with seizures are time-consuming and expensive, and it can be argued that early evidence of efficacy from short studies of a small size before embarking on a major trials programme makes sense, particularly if a suitable window occurs in the development schedule for these studies to be done. If they yield positive results, encouragement will be given to the further development of the drug. If they are negative, however, a major problem will arise: should the drug be abandoned as being unlikely to be effective in suppressing seizures because it has failed in these surrogate measures? If the evidence concerning the power of these tests to identify antiseizure activity was robust, the answer to this question would be ‘yes’, but unfortunately this is not the case and few investigators would advise that a development programme was terminated because * Tel.: +44-1594-562220; fax: +44-1594-562220. E-mail address: alan.richens@btinternet.com.uk (A. Richens).