Discussion
This study shows that replacing one anticonvulsant drug with another is beneficial for nearly 1 in 3 patients (18 of 59; 31%) with generalized tonic-clonic or partial seizures in whom a previous single-drug therapy up to individual tolerance has failed. In addition, side effects disappeared in 44% (8 of 18) of the patients with improved seizure control. Transfer to carbamazepine, phenytoin, or phenobarbital or primidone seemed to be equally successful, in agreement with reports of the similar efficacy of these major anticonvulsant drugs for partial epilepsy 14). Patients with a seizure reduction of 75% or more often had epilepsy of recent onset. Thus, when the previous drug has failed, immediate transfer to another single drug seems to be most promising.
Side effects disappeared in 27% of the patients. The transfer to carbamazepine seemed to be slightly, but not significantly, more effective in reducing side effects compared with phenytoin, phenobarbital, or primidone; this finding requires further exploration. Even though substitution is clearly very common in the management of epilepsy, few studies have been reported of failures of maximally tolerated daily dosage in single-drug therapy.
In controlled comparative trials no significant difference was found between the single-drug treatment results with carbamazepine and phenytoin 11). In individual cases, however, carbamazepine worked considerably better than phenytoin, and vice versa @}. In a single-blind trial comparing primidone and carbamazepine in 4 5 patients, 5 patients were entirely free of seizures while on primidone, but had attacks on carbamazepine; the opposite was true in 3 patients E57.
Apparently no systematic studies have addressed the effectiveness of alternative single-drug therapy in the failure of previous single-drug therapy. Our data suggest that alternative single-drug therapy is valuable and should be considered before the permanent addition of a second drug. Further prospective studies are needed to outline clinical features that will help to predict the effect of alternative single-drug therapy with an individual anticonvulsant drug.
For financial support we thank the Deutsche Forschungsgemeinschaft (Schm. 448-6-1).