[Treatment of hemifacial spasm with type A botulinum toxin (AGN 191622): a dose finding study and the evaluation of clinical effect with electromyography].

Forty-one patients with hemifacial spasm had an injection of type A botulinum toxin (AGN 191622; Allergan Co. Ltd., Irvine, CA). Patients were randomly divided into 3 groups by the injection dose: group L (1 unit; 14 patients), group M (5 units; 14 patients), and group H (10 units; 13 patients). Half of the dose was injected into the orbicularis oculi and the rest into the zygomaticus major muscles on the affected side. The clinical effect and electromyogram were evaluated at 2 weeks after the injection. The clinical benefit was dependent on the injection dose, and group H showed the highest rate of improvement (84.6%). No adverse effect related to the toxin was demonstrated except one patient in group H who showed mild and transient lagophthalmos. For 81.8% of group H patients, the final judgement was "useful" or "very useful", which was 9.1% for group L and 50.0% for group M. On the other hand, electromyography disclosed no consistent dose-finding relationship. We conclude that at least 10 (preferably more) units of botulinum toxin are necessary for effectively treating hemifacial spasm. Electromyography has only limited value for the evaluation of clinical effect.