Rasmussen's encephalitis in surgery for epilepsy.

Nineteen patients with Rasmussen's encephalitis were studied, using light-microscopy, immunohistochemistry and in some cases electron-microscopy. Although there was inflammation, no causative agent was identified. 10 patients underwent hemispherectomy and with one exception did well, while local resection for seven produced little or no improvement. For two patients biopsy with subpial transection was beneficial, although a third had a poor response, necessitating subsequent hemispherectomy. Surgical treatment of these patients is justified by a beneficial outcome. Neuropathological study of the resected material is essential for diagnosis and management of this condition.