Abstracts

Rationale: Reduced signal on [11C]FMZ PET has repeatedly been reported in the medial temporal lobe of patients with temporal lobe epilepsy (TLE) and hippocampal sclerosis (HS) or normal MRI, and is associated with epileptogenic foci. FMZ PET continues to be in widespread use in the presurgical evaluation. Within the same individual, usually healthy controls, linear correlations between parametric and non-parametric representations of FMZ binding have been shown. Methods of varying complexity have been used clinically, but the question of how these impact on diagnostic yield in individual patients has not been investigated. Patients with HS can serve as test cases for investigating the yield of an image analysis method.Methods: We formally compared the diagnostic yield of parametric and semiquantitative images, all used previously in the literature. Parametric images were obtained either with a parent tracer arterial plasma input function and spectral analysis (yielding volume-of-distribution (VD) images), or with an image-based input function and the simplified reference tissue model (binding potential images, BP-SRTM). Semiquantitative images were obtained by integrating images from 10-20 or 20-40 mins (ADD1020 and ADD2040). The dynamic 90 min [11C]FMZ PET datasets following injection of ~370 MBq of [11C]FMZ and arterial plasma input functions necessary to construct all four types of images were available for 15 patients with medically refractory TLE and histologically verified unilateral HS and for 13 healthy controls. Identical analyses were performed using Statistical Parametric Mapping (SPM2) for all images.Results: The coefficients of variation (CV; SD/mean) of global average image values ranged from 10-15% for controls across all groups. Patients had a similar CV (14%) for the VD group, but markedly higher CVs in the BP-SRTM (28%) and ADD1020 and ADD2040 groups (34%). While the means of images of delivery/blood flow (RI, derived from the SRTM) were identical, their CV in patients was nearly four times higher than in controls (15% vs 4%). ADD1020 and ADD2040 images detected decreased FMZ uptake ipsilateral to the epileptogenic hippocampus in 13/15 individual patients, with bilateral decreases in 6/13 (ADD1020) and 5/13 (ADD2040). BP-SRTM images detected ipsilateral decreases in 12/15 cases, with bilateral decreases in three. In contrast, VD images showed ipsilateral hippocampal decreases in all 15 patients, with bilateral decreases in three. Bilateral decreases in the ADD images tended to be more symmetrical; in one case they were more marked contralaterally.Conclusions: Semiquantitative analyses or those with an internal reference show high but imperfect sensitivity to detect medial temporal changes in TLE. Full quantification with an image-independent input offers a further advantage, with 100% sensitivity. Much greater heterogeneity of blood flow and/or tracer delivery between patients than between controls is the likely physiological basis of differences between analysis strategies. At least in TLE; intrasubject correlations do not predict equivalent clinical usefulness.