Abstracts

Rationale: Up to 70% of patients with intractable temporal lobe epilepsy(TLE) and hippocampal sclerosis(HS) continue to experience complex partial seizures(CPS) after temporal lobe surgery when followed up two years post-surgery.¹This figure falls to 48% ten years post-surgery.²Why some patients continue to experience postsurgical seizures despite being excellent surgical candidates is unknown. In this study, we sought to investigate whether regional atrophy of the hippocampus on presurgical MRI is associated with postsurgical CPS. We hypothesised that posterior ipsilateral and contralateral hippocampal atrophy would be associated with postsurgical CPS based on previous work.³Methods: We analysed T1-weighted MRI data(Magnetom Trio, Siemens) of 86 patients with TLE(mean age 42 yrs, SD 14; 57 left, 29 right) with neuroradiological evidence of ipsilateral HS. All patients had comprehensive routine presurgical investigations⁴ and underwent standardised amygdalohippocampectomy. HS was confirmed in all with histopathological assessment of resected specimens. Engel s outcome scores were obtained at a mean of 20.8 months(12.2 SD) post-surgery. Quantitative presurgical MRI data was compared between patients who became free from CPS(Engel s Class Ia-d; EI) and those who were not completely free from CPS(Class II-IV; EII-IV). We estimated the volume of the whole hippocampus using methods adapted from our previous work(Fig 1),⁵ and divided the hippocampus into sections approximating the head (first 35%), body (next 45%) and tail (final 20%) based on a previously described protocol.⁶ Intracranial volume(ICV) was automatically determined using Freesurfer software. Volumes were also analysed with respect to presurgical clinical variables.Results: Fifty-four(63%) patients achieved EI outcome. There was modest reduction in ipsilateral and contralateral hippocampal head, body and tail volume(both raw and normalised to ICV) in the EII-IV group relative to E1, but these differences were not significant(ANOVA, p>0.05). There was no significant relation between raw and normalised volumes, age of onset of TLE, presurgical CPS frequency or a history of childhood febrile seizures. However, duration of epilepsy negatively correlated with volume of the ipsilateral hippocampal head(raw p=0.05, normalised p=0.03) and contralateral hippocampal tail(raw p=0.01, normalised p=0.005).Conclusions: All patients studied here were excellent surgical candidates and had visually indistinguishable differences on clinical MRI during presurgical evaluation. We report that outcome of temporal lobe surgery is not quantitatively related to the gross volume of the ipsilateral or contralateral hippocampal head, body or tail when the hippocampus is measured using robust manual methods. Acknowledgements This research was supported by a Medical Research Council award to SSK. References 1.Engel J, Jr.In: Engel J, Jr., ed. New York: Reven Press,1993:609. 2.de Tisi J. Lancet.2011;378:1388. 3.Keller SS. Epilepsy Res.2007;74:131. 4.Kral T. Zentralblatt fur Neurochirurgie.2002;63:106. 5.Keller SS. Neuroimage.2002;16:23. 6.Hackert VH. Neuroimage.2002;17:1365.