Abstracts

RATIONALE: Less than 50% of patients with temporal lobe epilepsy (TLE) and a normal brain MRI (NMRI)have been reported to become seizure-free following standard antero-temporal lobectomies. In addition, resection of functional hippocampus can result in post-surgical memory deficits. Tailoring the extent of the resection to the epileptogenic zone may increase the proportion of seizure-free patients, while minimizing the risk of memory deficits. The necessary extent of resection of mesial temporal structures in patients with NMRI and TLE is yet to be established. This study was undertaken to answer this question.
METHODS: We compared the extent of resection of mesial-temporal structures between 12 patients (7 men and 5 women) with TLE and NMRI and 16 patients (7 men and 9 women) with unilateral hippocampal atrophy. Presurgical MRI studies included volumetric measurements of amygdala and hippocampal formation. To be included in the study, all patients had to undergo a tailored antero-temporal resection and all had to be seizure-free (Engel Class I) or have at the most 1 seizure/year (Engel Class II) during the post-surgical follow-up period of at least 2 years duration. Three investigators evaluated the extent of resection of mesial structures in a brain MRI done 6 months after surgery. Analysis of each MRI was carried out by each investigator, independently and blinded to patient identity and pre-surgical data. We used a semi-quantitative and a descriptive method to assess the extent of resection of mesial structures. The semi-quantitative method, developed by Awad et al (Awad et al., Epilepsia, 1989), is based on a 20 compartment model of the temporal lobe (5 planes, each divided into 4 quadrants, mesial, basal, infero-lateral and supero-lateral). The amount of resection of each quadrant resulted in 5 scores, one for each quadrant and a total score. In the descriptive method, a rating of (i) intact, (ii) partially resected or (iii) totally resected was generated for amygdala, head, body and tail of hippocampus by each investigator. Comparison of both groups were carried out with ANOVA and Chi-square statistics
RESULTS: Semi-quantitative method: Patients with NMRI had significantly lower scores for total resection (F=14.3[1,26], p=0.001) and for resection of mesial (F=10.1[1,26], p=0.004), basal (F=6.9[1,26], p=0.014), infero-lateral (F=3.9[1,26], p = 0.05) and supero-lateral quadrants (F=10.9[1,26], p=0.003). The descriptive method revealed that patients with NMRI were significantly less likely to have a resection of body (X2=15.2, p=0.001) and tail of hippocampus (X2=12.4, p=0.002).
CONCLUSIONS: In patients with NMRI and TLE, resection of body and tail of hippocampus is not always necessary to result in a good postsurgical outcome.