One of the most common inflammatory lesions encountered in corticectomies performed for intractable epilepsy is the chronic inflammatory reaction (often with a giant cell or granulomatous component) that is left by depth electrodes implanted for preoperative monitoring purposes within brain parenchyma. Leaving aside the obvious fact that any inflammatory disorder of the brain (especially viral encephalitides, e.g., caused by Herpes simplex) may be accompanied by seizure activity, evidence is emerging of an increasingly important role for brain inflammation in epilepsy. Steroids and adrenocor-ticotropic hormone (ACTH) have powerful anticonvul-sant effects, especially in children with infantile spasms or West syndrome. Seizure activity is regularly associated with a cerebrospinal fluid (CSF) pleocytosis and elevated CSF proinflammatory cytokines. Emerging evidence suggests that patients who develop temporal lobe epilepsy with hippocampal sclerosis (TLE-HS) following febrile convulsions may have been at risk for developing TLE-HS because of polymorphisms in the interleukin (IL)-1p-511T allele (234). Temporal lobes resected from patients with TLE have demonstrated overexpression of NFkappaB, a transcription factor involved in acute inflammation (235). There is indirect evidence that new-onset refractory status epilepticus (NORSE) may have an inflammatory basis in that many patients have an antecedent inflammatory or infectious illness and CSF usually shows a pleocytosis (236).
Rasmussen Syndrome (RS)
Rasmussen syndrome (or Rasmussen encephalitis, RE) is characterized by intractable focal seizures (usually epilepsia partialis continua) with progressive hemiparesis, symptoms attributed to chronic (pathogen-free) inflammation of gray and white matter, with progressive unihemi-spheric atrophy (237-239). A recent European consensus meeting on RS has formulated diagnostic criteria, which incorporate clinical, EEG, MR, and pathologic findings (240). Although the pathologic appearances of brain tissue affected by RE suggest a chronic viral infection, no virus has ever been consistently isolated from RS brain tissue or discovered within it using modern molecular or microbial culture techniques (241, 242). Circulating anti-glu-R3 antibodies were reported to be of etiologic importance in some patients with RS (243, 244), but subsequently these antibodies were found to occur in other seizure disorders (245, 246). Very recently, autoantibod-ies against the N-methyl-D-aspartate glutamate receptor (NMDA-type GluR) epsilon2 subunit and its epitopes were reported in RS patients (247). However, although the initial report is highly promising, the diagnostic specificity of GluRepsilon2 for RS remains to be confirmed. Any explanation for RS will ultimately have to account for its unihemispheric topography. Rarely have pathologic studies demonstrated bihemispheric involvement (248).
The neuropathologic findings of RS/RE (Figure 5-10) are said to comprise four merging stages (240, 249, 250), the earliest of which is characterized by inflammation, especially perivascular lymphocytes, and microglial nodule formation within brain parenchyma, but little morphologic evidence of neuronal injury. In stage 2, lymphocytic infiltration increases in density and both astrocytes and microglial cells become more widespread, tending to involve all cortical layers—a "panlaminar" pattern of cortical inflammation and gliosis. Patchy neuronal loss may be present. In stage 3 the neuronal population is depleted in either a patchy or panlaminar pattern, with
Was this article helpful?