Abstracts

Rationale: Gamma Knife radiosurgery (RS) may be an alternative to open surgery for mesial temporal lobe epilepsy (MTLE), but the anticonvulsant mechanisms of RS are unclear, with some data suggesting a necrotic lesion at the target and some suggesting neuromodulatory effects. Examination of the visual field defects (VFD) after RS may provide evidence of the extent of a postoperative fixed lesion. VFD occur in ~50% of patients undergoing open surgery for MTLE. Methods: This multicenter prospective trial of RS enrolled patients with unilateral hippocampal sclerosis and concordant video-EEG findings. Main results of this US Pilot Study have been published before. An exclusion criterion was presence of pre-operative VFD. Patients were randomized to low (20Gy) or high (24Gy) volumes delivered to the amygdala, hippocampal head, and parahippocampal gyrus. Exposure to optic nerve was limited to <8Gy. Postoperative perimetry (Goldman or Humphries) were obtained at 24m after RS. Perimetry VFD were classified by two independent reviewers blinded to other study data. Results were contrasted with RS dose (20Gy/24Gy), 50% isodose volume, seizure remission (1 year seizure free between postoperative 24m-36m), volume of peak effect (at 12 m) radiation-induced change at the surgical target, and side of RS (language-dominant/nondominant). Results: Postoperative VF results were available for 26 of the 30 patient sample. No patients complained of VFD or poor acuity, and confrontation examinations were normal. None had scotoma or arcuate defects. 17 (65%) had postoperative VFD. VFD were homonymous superior quadrantanopias reaching to the horizontal midline [7/26 (27%)] or sparing the horizontal midline [n=10/26 (38%)]. No VFD reached below the horizontal midline. Although the proportion of those with VFD was higher in the high dose [10/12 (83%] than the low dose arm [7/14 (50%)], the difference, probably because of the sample size, was not statistically significant (Fisher's exact test p=0.08). Volumes of T2-weighted radiation-induced change at 12m were higher in those with VFD (403±350ml vs 182±250, p=0.11 t-test). The side of RS, isodose volumes, and seizure remission were not associated with presence of or severity of VFD. Conclusions: VFD appeared after RS in proportions similar to those published for open surgery for MTLE. The nature of VFD was consistent with lesions of the optic radiations, sparing optic nerve. These findings, along with previously published findings on spectroscopic changes following RS, support the hypothesis that the anti-epileptic effects of RS are destructive rather than modulatory.