RT Journal Article
SR Electronic
T1 Correlation of metabolic and volumetric metrics of the structures of the temporal lobe on MRI and FDG-PET in patients with mesial temporal sclerosis
JF Journal of Nuclear Medicine
JO J Nucl Med
FD Society of Nuclear Medicine
SP 1866b
OP 1866b
VO 57
IS supplement 2
A1 Carlos Leiva Salinas
A1 Nathan Fountain
A1 Mark Quigg
A1 W Jeffrey Elias
A1 James Patrie
A1 Patrice Rehm
YR 2016
UL http://jnm.snmjournals.org/content/57/supplement_2/1866b.abstract
AB 1866 (b)Objectives To correlate metrics of metabolic activity on FDG PET and volume on MRI of different temporal lobe structures in patients with mesial temporal sclerosis (MTS) who were seizure-free for at least 2 years after surgery. To correlate the degree of temporal hypometabolism on FDG PET and volume on MRI with total epilepsy duration.Methods We retrospectively reviewed the pre-surgical brain FDG PET-CT and MRI of 18 patients with intractable temporal lobe epilepsy and MRI findings of MTS, who were seizure-free for at least 2 years after temporal lobectomy. On MRI we automatically segmented the bilateral hippocampi and amygdalae and calculated their volume. On PET we automatically segmented the bilateral hippocampi, amygdalae, and whole temporal lobes. We calculated the mean SUV for each of those brain structures and compared their metabolic activity to a population of normal patients in terms of Z-scores, and calculated the volume of the tissue that was hypometabolic as compared to the normalized dataset, using 2 SD as threshold. We compared the SUV, Z-scores, and volume values in the bilateral structures using the paired t test. We correlated the volume and SUV and Z-scores metrics with the total epilepsy duration (years from onset) by means of the Spearman Rank correlation coefficient.Results 14 patients were female (77.8%). 12 subjects had left MTS (66.7%). Mean age was 37.2±12.9 years. Mean total epilepsy duration was 22.9±12.9 years. Mean total volume of hypometabolic temporal tissue was 34.3±29.8 ml. Mean SUV in the diseased side was significantly lower than the contralateral side for hippocampus (4.13±1.76 vs 4.93±1.85), amygdala (3.9±1.85vs 4.45±1.76), and whole temporal lobe (4.91±2.16 vs 5.66±1.88) (p<0.001 for all). Mean volume in the diseased side was significantly lower than the contralateral side for hippocampus (2.84±0.49 vs 3.52±0.4) and amygdala (1.49± 0.24 vs 1.72± 0.3) (p<0.001 and p=0.02, respectively). No significant correlation was identified between SUV and volume for the diseased hippocampus (rs=-0.36, p=0.15) or amygdala (rs=-0.1, p=0.69). There was a significant positive moderate correlation between total epilepsy duration and the degree of the diseased hippocampal (rs=0.49, p=0.037) and amygdalar (rs=0.47, p=0.047) hypometabolism as compared to normal subjects; but none was found for the whole temporal lobe (rs=0.06, p=0.83). There was also a significant positive moderate correlation between total epilepsy duration and contralateral hippocampal (rs=0.59, p=0.01) hypometabolism as compared to normal subjects. No significant relationship between total epilepsy duration and SUV was identified for any of the temporal lobe structures (rs=0.35, rs=0.29 and rs=-0.1, and p=0.25, p=0.21 and p=0.74 for the hippocampus, amygdala and global temporal lobe respectively). There was no significant correlation between the total volume of hypometabolic temporal tissue and total epilepsy duration (rs=-0.23, p=0.36). No significant relationship between epilepsy duration and volume was identified for the hippocampus or amygdala (rs=0.31 and rs=-0.1, and p=0.21 and p=0.7, respectively).Conclusions There was no interdependence between metabolic activity and volume of the medial temporal lobe structures, which suggests that FDG PET and MRI add unique and separate information. Total epilepsy duration correlated with the degree of hippocampal and amygdalar hypometabolism relative to normals, but not absolute SUV values; although this could be due to small sample, it highlights the importance of comparing PET data of epileptic patients to normalized datasets and suggests that cortical hypometabolism may be associated with the duration of epilepsy. There was also a significant relationship between epilepsy duration and the degree of hypometabolism of the contralateral hippocampus, which may indicate remote neuronal damage in patients with longstanding seizures.