Abstracts

Rationale: Quantification of hippocampal T2 relaxation time can identify the side of hippocampal sclerosis (HS) in patients without hippocampal atrophy detectable by volumetric MR imaging. However, the majority of patients with mesial temporal lobe epilepsy (TLE) and normal MR imaging are not necessarily associated with HS. In this study, the use of T2 relaxometry was investigated to detect non-sclerotic epileptogenic abnormality of the hippocampus in presurgical evaluation of TLE. Methods: This prospective study included 30 patients who underwent hippocampectomy as part of surgical treatment of refractory TLE. Ten patients had structural epileptogenic lesions in the extra-hippocampal temporal lobe. Twelve patients underwent intracranial electroencephalography (iEEG) study before surgery. Visual assessment of atrophy and increased T2 signal intensity, volumetry, and T2 relaxometry of hippocampus were performed pre-operatively using 3 T magnetic resonance imaging, and compared with the neuropathological findings and iEEG findings. Magnetic resonance imaging of 30 age- and sex-matched healthy controls was used to establish normal values, which were defined as z score within 2. Results: Visual assessment, volumetry, and T2 relaxometry detected hippocampal abnormalities on the surgical side in 16 (53%), 16 (53%), and 26 (87%) patients, respectively. Hippocampal volume loss was always associated with prolonged T2 relaxation time, and supported by histopathological diagnosis of HS in all cases except one. Hippocampal abnormality was detected only by T2 relaxometry in nine patients (30%). Pathological diagnosis of these cases included mild HS in one, microdysgenesis in one, and granule cell pathology in three. Four patients with normal hippocampal volume and T2 relaxation time had no HS or granule cell pathology. Prolonged T2 relaxation time was associated with medial temporal seizure onset in iEEG (p < 0.05). Conclusions: T2 relaxometry can detect hippocampal abnormality in epilepsy patients without hippocampal atrophy. Moreover, normal T2 relaxation time may provide a negative marker of epileptogenicity. The acquisition time of T2 relaxometry is shorter and more acceptable in routine radiologic practice using 3 T rather than 1.5 T MR scanners because of the improved pulse sequences and signal to noise ratio afforded by the increased field strength. T2 relaxometry may lead to better selection of surgical candidates and improved resection strategies. Funding: This research was partially supported by Grant-in-Aid for Scientific Research No. 23591747 from the Japan Society for the Promotion of Science. Author S.S. was supported by Grant-in-Aid for Scientific Research from the Miyagi Medical Women's Association, Sendai, Japan.