| ObjectivesIt keeps unknown for a long time about the role of the glucose metabolic change intemporal lobe in temporal lobe epilepsy (TLE). The relation of interictal hypometabolicregions in temporal lobe with the epileptogenic focus is not known either. Thus, the purposeof this research is to explore the energy metabolic features of temporal lobe in TLE andanalyze their relation with the clinical presentations, features of electroencephalography,status of hippocampus sclerosis and analysis result of Magnetic Resonance Spectroscopy(MRS), in order to provide the theorical supports for the preoperative seizure onset zonelocation and postoperative evaluation using positron emission tomography/Computedtomography (PET/CT).Methods18F-fluorodeoxyglucose positron emission tomography (18F-FDG-PET) was performed in82 consecutive patients suffering TLE without intracranial mass, all of who subsequentlyunderwent surgical resection. The clinical data, the EEG findings and the postoperativepathologic analysis confirmed the diagnosis of TLE. The equipment applied was DiscoveryLS PET/CT (GE, CA, USA). The parameters were set as: 150mA, 120 Kv, 4.25mm layerthickness. After 45～60 min rest following intravenous injection of contrast medium 18F-FDGat 0.15～0.18mCi/kg, the PET/CT scanning was performed. Image-collection and imagingreconstruction patterns were three-dimension and OS-EM respectively. After computerprocess, the images of coronal, transversal and sagittal sections were all obtained. Theappearance of temporal lobe's foci in PET/CT was described by visual method and thestandard uptaking values (SUV) were calculated and analyzed semi-quantitatively. Definitivehigh or low metabolic zone in two consecutive sections was regarded abnormal. The area with 15% decreasing or increasing radiation activity in comparison with corresponding areas wasregarded as epilepsy foci. Based on the different distributional features of hypometabolic zonedisplayed in PET/CT, a classification of abnormal metabolic zone in TLE was established asfollows:ⅰunilateral interior temporal hypometabolism, whose hypometabolic zones werelocated in medial part of unilateral temporal lobe, with definite boundary to surroundingnormal metabolic zones;ⅱunilateral lateral temporal hypometabolism, whosehypometabolic zones were located in unilateral temporal cortex, with definite boundary tosurrounding normal metabolic zones of unilateral temporal lobe;ⅲunilateral median andlateral temporal hypometabolism, whose hypometabolic zones were located in median aslateral temporal lobe;ⅳbilateral temporal hypometabolism, whose features were thesummary of all of the typeⅰ,ⅱandⅲ;ⅴextensive hypometabolism, in addition totemporal hypometabolism, there were other low metabolic zones in the lobes of the other side,but un-adjacent to the foci of temporal lobe. Statistical discrepancies between these fivegroups and their relations with the patient age, age at epilepsy onset, epilepsy duration, thekinds of seizure, seizure duration and seizure frequency were all assessed using the SAS8.1.MRI(GE 2.0T Prestige super-conductive MR scanning system) was performed in 48 of theseTLE cases. The standard head coil was served as emission and reception coil, oblique coronalsection scanning vertical to HC longitudinal axis was performed. Scanning parameters wereset as follows: spin-echo(SE) sequence, T1W1 (TR=500ms, TE=12ms), enhanced 2 times,4ms layer thickness, 0mm layer inter-space. 1H- Magnetic resonance spectroscopy (1H-MRS)was performed in 34 TLE patients adopting point resolving spectroscopy sequence (PRESS)to localize the interesting origin of bilateral hippocampus with volume of 2×2×2cm3. Afterlocal homogenization and dehydrodization, spectrum was collected by PRESS, withTR1500ms and TE36ms, reinforcing in 200 times. Computers performed all proceduresautomatically. Late-stage processing of spectrum included: phase correction, base-linecorrection, measurement of all metabolic product concentration based on spectrum coveringareas. After base-line correction of spectrum, areas below the peak of acetyl aspartic acid(NAA), Creatine (Cr), and Choline (Cho) waves were calculated by the integrity method,corresponding to the concentrations of them respectively. 95% reference value ((?)±1.96s) of NAA/Cho+Cr in normal group was defined as standard value. MRI and MRS scanreviewed the different metabolic subtypes of TLE. The Hippocampus sclerosis (HS) and thevalue of NAA/(Cho+Cr) were evaluated. 11 TLE cases underwent VideoElectroencephalography to evaluate intictal regional delta slowing. Drowsy, asleep, andpostictal EEG recordings (12 h after seizures) were excluded from the evaluation. Temporalregional interictal delta activity was then compared with the regional hypometabolism in thesame area. During the operation, we observe the slow waves in Electrocorticogram (EcoG)and depth electroencephalogram (DEEG).ResultsInterictal 18F-FDG-PET hypometabolism was detected in all 82 patients, the sensibility is100%. Among these hypometabolism, typeⅰ13cases(16%), typeⅱ4 cases (5%), typeⅲ10cases (12%), typeiv28cases (34%), typeⅴ27 cases (33%). There was no significantdifference in patients' ages, epilepsy onset ages, patterns of seizure, seizure duration, andseizure frequency among these 5 subtypes of hypometabolism. However, the duration ofepilepsy appeared longer in typeⅰ. MRI was performed in 48 patients to detect 9 unilateralHS (18.75%), 13 bilateral HS (27.08%), 26 cases without HS (54.17%). 34 TLE underwentMRS analysis to detect 3 cases (8.82%) with normal value of NAA/(Cho+Cr), 8 cases(23.53%) with unilateral decreased value of NAA/(Cho+Cr), 23 cases (67.65%) with bilateraldecreased value of NAA/(Cho+Cr). Incidence of HS in different subtypes of abnormalmetabolic zones in TLE was uneven: no HS was observed in typeⅱand the unilateral HS andbilateral HS were seen in typeⅳ. There were no significant correlations between HS andtemporal hypometabolism. We found intermittent rhythmic delta activity (IRDA) in 8 out of11 observation cases and among them 3 showed slow waves in EcoG scanning in temporalarea, where PET detected hypometabolism correlated with the regional interictal delta activityin EEG in the same side.ConclusionsThe 18F-FDG-PET/CT reveals the interictal focal glucose metabolic changes in TLE,that is hypometabolism. Based on the patterns of hypometabolism, TLE can be classified intofive subtypes: unilateral interior temporal hypometabolism, unilateral lateral temporal hypometabolism, unilateral median and lateral temporal hypometabolism, bilateral temporalhypometabolism, and extensive hypometabolism. Although the sensibility of 18F-FDG-PET/CT for the TLE hypometabolism is high, the explanation of the image must bethoughtful. The location of epileptogenic focus should be made based on the pattem ofhypometabolism, EEG and other examinational data. The pathologic mechanisms for TLEhypometabolism cannot only be explained with HS and loss of neurons, while electricaldysfunction might play an important role in the procedure. |
PDF Full Text Request