Abstracts

Rationale: Radiosurgery (RS) is a novel and highly promising treatment for intractable mesial temporal lobe epilepsy (MTLE). In this study, we characterize the time-course and dose-response of radiographic alterations after RS and relate them to clinical response. Methods: Thirty patients were enrolled in a prospective, multi-center trial designed to investigate the safety and efficacy of Gamma Knife® radiosurgery for MTLE. Routine magnetic resonance imaging (MRI) was obtained at 12- and 24-months following treatment, and at additional time-points per physician discretion. MRIs were evaluated by blinded central reviewers for qualitative and quantitative changes, specifically for contrast-enhancement, T2 abnormality, mass effect, spectroscopy, and diffusion changes. MR features were analyzed for potential association to radiation dose (20 or 24 Gy) and seizure remission (seizure-free at 24-36 months). Results: RS produced dramatic alterations of MRI features in all patients. At 0-8 months, only subtle MRI changes were evident. At about 9 months, acute changes appeared with “ring” contrast-enhancement surrounded by high T2 signal of the white matter. While enhancement was localized to the medial temporal lobe, T2 signal appeared diffusely and extended as far posterior as the parietal and occipital lobes in some cases. These changes occurred at a similar time to the onset of seizure reduction. Apparent diffusion coefficients (ADC) were increased to approximately twice that measured from the contralateral temporal lobe. There was a reduction in the NAA-to-creatine ratio and a later increase in the choline-to-creatine ratio that appeared to be coincident with the maximal T2 and ADC changes. T2 hyperintensity and contrast-enhancement appeared to peak between 10-15 months, and usually fully resolved by 30 months with mesial temporal atrophy. Despite individual variability in timing and magnitude, MRI changes measured at 12-months were quantitatively more extensive in patients that received the higher radiation dose (Volume of T2 hyperintensity: For 20GY (mean+/-SD)= 148cc+/-226 vs. 24 GY= 416+/-348cc, p=0.03/ Volume of contrast-enhancement: For 20GY=14.5+/-17cc vs. 24 GY= 50.6+/- 26cc, P=0.0003, unpaired t-test). Furthermore, eventual seizure remission also appeared to be statistically associated with greater radiographic response at 12-months. (Volume of T2 hyperintensity= REMISSION: 370+/-334cc vs NO-REMISSION 92+/-170cc, P=0.02/ Volume of contrast-enhancement= REMISSION: 39.2+/26cc vs. NO-REMISSION 16+/-25cc, P=0.04). At 24-months, while MR alterations were still present, they did not correlate with dose or seizure remission. Conclusions: Marked alterations in MRI occur after RS for MTLE. Most changes are dose-dependent and occur in a delayed fashion, paralleling the onset of seizure reduction. The volume of vasogenic edema and and enhancement at 12 months is predictive of seizure remission. These results have implications for understanding the specific structural and biochemical alterations after RS, and prognosticating outcome based upon individual susceptibility to radiation.