Abstracts

This study was performed to evaluate the effect of seizures on the bilateral hippocampus in medial temporal lobe epilepsy and neocortical epilepsy by single voxel proton magnetic spectroscopy.
Forty-one patients with intractable medial temporal lobe epilepsy having unilateral hippocampal sclerosis and 43 patients with a neocortical epileptic focus who underwent subsequent epilepsy surgery were recruited. Twenty healthy volunteers formed the normal control group. Ninety-five percent confidence intervals of N-acetyl aspartate/choline (NAA/Cho) and NAA/creatine (NAA/Cr) ratios in control subjects were used as threshold values to determine abnormal NAA/Cho and NAA/Cr ratios in the ipsilateral and contralateral hippocampus of patients. The asymmetry indices (AIs) for the evaluation of the lateralizing ability of MRS for each pair of voxels were calculated. The relationships between the results of MRS and the duration of epilepsy, frequency of seizure, effect of secondary generalized tonic clonic seizures (2GTCS) were examined.
NAA/Cho and NAA/Cr were significantly lower in the ipsilateral hippocampus of medial TLE (paired t-test, p[lt]0.0001 and p=0.01 respectively). NAA/Cho of the hippocampus in neocortical epilepsy was also significantly lower than in the side ipsilateral to the epileptogenic hemisphere (paired t-test, p[lt]0.05). The NAA/Cho ratio was abnormally low in the ipsilateral hippocampus in 12 of 43 neocortical epilepsy patients. A bilateral hippocampal abnormally low NAA/Cho ratio was also found in 18 of 43 patients. Ipsilateral or bilateral abnormally low NAA/Cr ratio was detected in 14 patients. Of AIs using NAA/Cho and NAA/Cr ratios, at least one AI correctly lateralized seizure focus in 27 of 41 patients with mTLE (65.9%). This analysis also correctly lateralized epileptogenic hemisphere in 21 of 43 neocortical epilepsy patients (48.8%). Both of the two AIs falsely lateralized the epileptogenic hemisphere in only two patients with neocortical epilepsy. Bilateral NAA/Cho ratio abnormality was significantly related to poor surgical outcome in medial TLE (Fisher[rsquo]s exact test, p=0.039, one-sided). The mean NAA/Cr of the contralateral hippocampus in mTLE was significantly lower in patients with a history of secondary generalized tonic clonic seizure (2GTCS) than in patients without 2GTCS (t-test, p=0.002).
Our results demonstrate functional abnormality of the hippocampus in neocortical epilepsy and the relation between 2 GTCS and NAA/Cr of the contralateral hippocampus in mTLE. Abnormal NA/Cho or NAA/Cr in neocortical epilepsy were not found to be related to surgical outcome, suggesting that the abnormal ratios do not mean the presence of epileptogenic tissue. This is in line with the presence of seizure effect on hippocampus in neocortical epilepsy as well as in mTLE.