Anticonvulsant therapy in the elderly--a need for placebo controlled trials

Epileptic seizures are common in the elderly, yet data concerning the long-term clinical course and apparent impact of anticonvulsant therapy are scant. We studied 73 consecutive elderly patients with a diagnosis of seizures [remote symptomatic (52%), acute symptomatic (23%), progressive symptomatic (10%), cryptogenic (15%)] during a median period of clinical review of 33 (range 3-72) months. Sixty-seven patients received anticonvulsant drugs, 38 phenytoin (PHT), 21 carbamazepine (CBZ), 6 sodium valproate (VPA) and 2 phenobarbitone. Six patients were untreated with drugs and three of these had no further seizures over a median review period of 26 months. Forty-one (61%) treated patients remained seizure free and a further nine patients suffered less than three fits per year. Seventeen patients had poorer control (three to five seizures per year in six patients and more than five seizures per year in eleven patients). Mean daily dosage of anticonvulsants (PHT 248 mg, CBZ 320 mg, VPA 571 mg) and serum concentrations were modest. Anticonvulsant side effects were reported by 27% of all treated patients (22% of those who were seizure free). Both adverse effects and satisfactory seizure control were associated in the majority of patients with serum anticonvulsant concentrations at the lower limit or below recommended therapeutic ranges utilised in the young. This study suggests that placebo controlled studies are warranted to appraise the extent to which anticonvulsant drugs modify a generally favourable prognosis for seizure disorders in the elderly and to adequately define the benefit-risk ratio of such drugs.