Abstracts

MR spectroscopic imaging (MRSI) has previously been shown to be sensitive in lateralisation of medial temporal lobe epilepsy (MTLE). However, its predictive value for surgical outcome remains unclear. This paper focuses on how measurements of hippocampal dysfunction may predict surgical outcome in a cohort of patients with MTLE., N=21 consecutive MTLE patients (mean age 39+/-12 yrs) who underwent anteromedial temporal lobe resection were studied; minimum follow-up 14mos. MRSI data were acquired at 4T using a voxel-shifting technique to select the hippocampal spectra. For the hippocampal loci (Fig 1), the p-value for significant abnormality of the measured NAA/Cr was determined using a linear regression and a threshold p-value of 0.025. Grouping the data into anterior (loci 1,2) and posterior (loci 3,4), it was classified as abnormal if either of the included loci were abnormal. Comparing the MRSI data with outcomes using ILAE class (I-V), a contingency table (Table 1) was constructed and analyzed using a [chi][sup2] analysis., At a median follow-up period of 26.5 months, 14 (67%) patients were completely seizure free (class I). Three patients (14%) continued to experience auras without clinical seizures (class II) and 4 continued to experience seizures (classes III-V, 19%) (Table 1). Notably, 3 patients with normal bilateral anterior hippocampal values all had poor outcomes (n=2 class III, n=1 class V). All 9 patients who demonstrated unequivocal unilateral anterior hippocampal abnormality had class I outcome. The complex group was those who demonstrated bilateral anterior hippocampal abnormalities. [chi][sup2] testing demonstrated these categories to be significantly different (p[lt]0.001). Notably, the equivalent contingency table derived from posterior hippocampal data was not significant (p=0.12)., Anterior hippocampal MRSI findings can be predictive of post-surgical outcomes. Patients with bilaterally normal anterior hippocampal NAA/Cr are associated with surgical failure; those with unilaterally abnormal hippocampal values are seizure free. The complex group with bilaterally abnormal NAA/Cr (43% of our group) includes seizure free and non-seizure free outcomes. Based on the finding by Spencer, that recurrence occurs in 25% of the patients after 2 years, follow up in this group may shed additional understanding of the significance of presurgical bilateral anterior defects.[figure1], (Supported by NIH R01 EB-000473.)