Predicting drug-resistant patients who respond to add-on therapy with levetiracetam

Levetiracetam (LEV) is approved for use as add-on therapy in adult patients with partial epilepsy. It is apparent from clinical trials that up to 8% of previously drug-resistant patients may be rendered seizure-free by adding-on levetiracetam. As yet there is no way of predicting these unexpectedly responsive patients. We set out to identify our previously refractory patients who had demonstrated unexpected responsiveness to add-on therapy with levetiracetam, and compared these to patients who had not responded to the drug. We then attempted to characterise any clinical features that differentiated these groups of patients.

We included all patients with a history of present or previous exposure to levetiracetam who had been unresponsive to at least two other prior anti-epileptic drugs (AEDs) and recorded their demographic and clinical data. We divided response into (a) 'seizure-free' (seizure-free for a minimum of 6 months after commencing LEV); (b) 'partial > 50%' (greater than 50% reduction in seizures for a minimum of 6 months after commencing LEV); (c) 'honeymoon' (seizure-free for less than 6 months after commencing LEV and then returned towards baseline frequency); and (d) 'no-response'. For the purpose of analysis we considered the 'seizure-free' and 'partial > 50%' groups as 'responders', and the 'no response' group as 'non responders'.

344 patients were included in the analysis. Fifty-six patients (16.3%) were rendered seizure-free on levetiracetam. Idiopathic generalised epilepsy and post-traumatic partial epilepsy were more common in the responder than the non-responder group (p = 0.005 and 0.05 respectively). Lamotrigine was used significantly more often in combination with levetiracetam in responders than non-responders (p = 0.003). The mean daily dose of levetiracetam was lower in responders than non-responders.

A higher than expected number of previously drug resistant patients was rendered seizure-free by add-on therapy with levetiracetam. Those who respond best appear to do so at relatively low doses and our data suggest the possibility of a beneficial pharmacodynamic interaction between levetiracetam and lamotrigine. We were unable to identify any clinical factors that clearly predicted which patients would become seizure-free and we hypothesise that response may be determined by genetic or molecular factors. All drug-resistant patients, including those being assessed for surgery, should be considered for a trial of levetiracetam, regardless of their epilepsy classification.