Free Papers: Neuroimaging


*Taketoshi Maehara, *Tadashi Nariai, *Youji Tanaka, †Yuji Murata, ‡Kenji Ishii, ‡Kiichi Ishiwata, and *Kikuo Ohno Departments of *Neurosurgery and †Radiology, Tokyo Medical and Dental University ; and ‡Positron Medical Center, Tokyo Metropolitan Institute of Gerontology, Tokyo, Japan .

Purpose: We investigated the clinical usefulness of the subtraction of interictal from ictal single-photon emission computed tomography (SPECT) coregistered to magnetic resonance imaging (MRI) (SISCOM), in detection of a seizure focus in patients with intractable frontal lobe complex partial seizures (CPSs). As a radiopharmaceutical, we used technetium-99m-ethyl cysteinate diethylester (99mTc-ECD), because it is more stable than technetium-99m-hexamethylpropylene amine oxime (99mTc-HMPAO) and, therefore, more convenient for observing ictal or periictal events.

Subjects and Methods: Three patients who had typical frontal CPSs and achieved significant improvement after focus resection were subjects of this study. After presurgical examinations, including repeated EEGs, video-EEG monitoring, MRI, neuropsychological testing, [11C]flumazenil (FMZ) positron emission tomography (PET), and [18F]fluorodeoxyglucose (FDG) PET, two patients underwent invasive monitoring for focus detection. The remaining patient underwent focus resection without invasive monitoring, because MRI revealed cortical dysplasia in the right orbitofrontal (OF) gyrus, and EEG demonstrated an ictal onset in the region of the corresponding lesion. Ictal SPECT images were obtained by recognizing ictal onset with video monitoring. The time lag between seizure onset and injection of the radiopharmaceutical was recorded. The SISCOM images were reconstructed by using the image-analysis software Dr. View (Asahi Kasei Joho, Tokyo, Japan) running on a personal computer.

Results: The SISCOM images demonstrated increased ECD uptake in the left anterior cingulate and OF gyrus in patient 1, in the right anterior cingulate gyrus in patient 2, and in the right OF gyrus in patient 3. The time lag between seizure onset and the isotope injection was 59 s in patient 1, 37 s in patient 2, and 53 s in patient 3. In each patient, laterality of the seizure onset in the frontal limbic systems showed good agreement with the area showing increased ECD uptake. We then analyzed the diagnostic value for focus detection of interictal FMZ-PET and FDG-PET in the same patients. In the focus area determined by electrocorticography or EEG, FMZ-PET detected decreased binding only in patient 3, but not in patients 1 and 2, whereas FDG-PET detected hypometabolism in patients 1 and 3, but not in patient 2. As for the focus detection of patients with frontal CPS, SISCOM seemed to have a superior diagnostic value over interictal FMZ-PET and FDG-PET. We also found that SISCOM might be a suitable indicator to determine the position of intracranial electrodes placement.

Conclusions: We demonstrated that SISCOM is an important presurgical examination for focus detection in patients with intractable frontal lobe CPSs.


*Keisuke Toda, *Hiroshi Baba, *Tomonori Ono, and *Masahiro Yonekura *Department of Neurosurgery, National Nagasaki Medical Center, Nagasaki, Japan .

Purpose: Single-photon emission computed tomography (SPECT) is a noninvasive functional neuroimaging study used as part of presurgical evaluation when attempting to localize an epileptogenic region. It is well recognized that the interictal SPECT in patients with mesial temporal lobe epilepsy demonstrates hypoperfusion areas in the temporal lobe ipsilateral to the epileptic focus. We evaluated the pattern of regional cerebral blood flow (rCBF) by visual inspection of SPECT images in epilepsy patients, but it is difficult to evaluate rCBF quantitatively in the current clinical SPECT systems.

Subjects and Methods: In this study, we analyzed changes of the rCBF in temporal lobe epilepsy by using a fully automated rCBF quantification program; the three-dimensional stereotaxic region of interest (ROI) template (3D-SRT). 3D-SRT has been reported to measure automatically the area-weighted mean value of the rCBF for each of 24 segments arbitrarily defined as the anatomic area supplied from major arteries from values of 536 ROIs of the whole brain. The 12 paired segments are as follows: anterior, precentral, central, parietal, angular, temporal, occipital, pericallosal, lenticular nucleus, thalamus, hippocampus, and cerebellum segments. Twenty-one patients with unilateral mesial temporal lobe epilepsies who received anterior temporal lobectomy with amygdalohippocampectomy were included in this study. The ages of patients ranged from 14 to 49 years (mean, 27 years), and the duration of seizures, from 1 to 31 years (mean, 13.4 years). Histopathologic diagnoses were as follows: hippocampal sclerosis (n = 15), cortical dysplasia (n = 3), dysembryoplastic neuroepithelial tumor (DNT) (n = 1), ganglioglioma (n = 1), and arteriovenous malformation (AVM) (n = 1). 99mTc-ECD SPECT was obtained during an interictal period in all patients, and rCBF was calculated by using 3D-SRT. To compare the values of rCBF between 21 patients, we determined the laterality indices (LIs) in all 12 segments of each patient (LI = |(left – right)|/|(left + right)| × 100). The mean LIs were as follows: 1.17 in anterior, 2.91 in precentral, 2.64 in central, 3.19 in parietal, 3.47 in angular, 4.92 in temporal, 2.19 in occipital, 1.57 in pericallosal, 2.23 in lenticular nucleus, 4.99 in thalamus, 4.82 in hippocampus, and 1.74 in cerebellum. The LIs of thalamus, temporal, hippocampus, and angular were significantly higher (p < 0.01) compared with that of the anterior. These results demonstrated that the laterality of rCBF is obvious in the temporal lobe and hippocampus as well as the thalamus in patients with mesial temporal epilepsy. The outcome after surgery at the time of follow-up according to Engel's classification was class I and II in 18 patients (85.7%) and class III in three (14.3%). No correlation was found between the mean LIs of class I/II and those of class III in all segments except thalamus. In the thalamus, LI is significantly higher (p < 0.01) in class I/II (5.58 ± 1.12) than in class III (1.45 ± 0. 17), suggesting that patients showing laterality of rCBF in thalamus are expected to achieve good surgical outcome from temporal lobectomy. The 3D-SRT is a useful modality to quantify the rCBF in the evaluation of an epileptogenic area.


*Tasuku Miyajima, *Yoshiaki Watanabe, *Mitsuhiro Nakajima, *Shingo Oana, *Gaku Yamanaka, *Naoto Yamada, *Shiori Nemoto, *Masaaki Ogihara, *Tetsuhiko Matsuno, and *Akinori Hoshika *Department of Pediatrics, Tokyo Medical University, Tokyo, Japan .

Background: Early delineation of the pathologic changes in the acute stage of convulsive status in infants is important to predict prognosis. We report five cases during the acute stage after convulsive status to demonstrate that important findings readily are detected with T2-weighted imaging, by using a fluid-attenuated inversion recovery (FLAIR) pulse sequence.

Methods: Five infants [one with severe myoclonic epilepsy in infancy (SMEI), one with complex partial seizure (CPS), and three with influenza encephalopathy (IE)] who manifested Todd's palsy after unilateral convulsive status were studied. Informed consent to releases the research results was obtained from all parents.

Results: The SMEI case (female) was admitted because of convulsions at age 6 months. The first brain MRI performed 2 weeks afterward showed no abnormality on standard T1-weighted images (T1WI), T2-weighted images (T2WI) and T2-weighted imaging using FLAIR sequence [repetition time (TR) 7,280 ms, inversion time (TI) 1,700 ms, echo time (TE) 110 ms]. Subsequently, she frequently manifested convulsive status with right or left seizures triggered by high fever or photic and pattern stimulation. At age 23 months, after an episode of convulsive status with high fever, she had a right body unilateral seizure. T2-weighted images using FLAIR pulse sequence (TR/TI/TE 7280/1700/132) clearly depicted diffuse and multiple high-intensity areas in the left hemispheric white matter and right hippocampus. However, these finding were not detected by standard T1WI and T2WI. The CPS case was a 4-year-old female infant with developmental delay who had CPSs and left Todd's palsy that continued for ∼2 h. After 1 week, brain MRI was performed. Lesions in the white matter of the right frontal lobe were more clearly detected on high-contrast images provided by T2-weighted images, using a FLAIR pulse sequence (TR/TI/TE 9000/2200/138) than by T2WI and T1WI. However, these findings disappeared after 3 months. Among three cases of unilateral convulsion and Todd's palsy caused by influenza encephalopathy, only one case showed hyperintense areas of frontotemporal cortical lesions on T2-weighted images with FLAIR pulse sequence (TR/TI/TE 9000/2300/137) and diffusion-weighted imaging (TR/TE 270/137). This case had no after effects and the findings on the diffusion-weighted imaging disappeared at follow-up MRI. However, the hyperintense lesions in the left hippocampus and white matter of the left frontal cortex were still depicted by T2-weighted images, with a FLAIR pulse sequence (TR/TI/TE 9000/2300/137).

Conclusions: In cases of convulsive status epilepticus during infancy, subsequent development of brain lesions, including hippocampal sclerosis, is a potential concern. Heavily T2-weighted images, using a FLAIR pulse sequ strong MR contrast of pathologic tissues. The results that we obtained suggest the usefulness of the MRI-FLAIR for estimating prognosis of patients with convulsive status.


*Kazumi Matsuda, *Tadahiro Mihara, *Takayasu Tottori, *Toshiaki Ohtsubo, *Koichi Baba, *Fumihiro Nakamura, *Yushi Inoue, and *Tateki Fujiwara *National Epilepsy Center, Shizuoka Medical Institute of Neurological Disorders, Shizuoka, Japan .

Purpose: Changes in the central-type benzodiazepine receptors (BZRs) can be detected by positron emission tomography (PET) with flumazenil, an antagonist of BZR, or by single-photon emission computed tomography (SPECT) with iomazenil (IMZ), a partial inverse agonist of BZR. A decrease of the BZR density in an epileptic focus has been demonstrated, and BZR imaging is expected to depict the focus as a decreased uptake area. This study investigated the usefulness of IMZ-SPECT for the detection of epileptic foci in patients with refractory epilepsy showing no obvious MRI abnormality.

Subjects and Methods: Twenty-three patients who underwent surgery for medically intractable partial seizures without definite abnormality on MRI were studied. The mean age at seizure onset was 12.7 (4–27) years, and the mean age at operation was 28 (20–42) years. All patients underwent interictal IMZ-SPECT and [123I]iodeamphetamine (IMP) SPECT for cerebral blood flow before operation. IMZ-SPECT was performed 3 h after intravenous injection of IMZ (167 MBq), and IMP-SPECT was performed immediately after intravenous injection of IMP (111–222 MBq). Nineteen patients underwent prolonged intracranial EEG recording before surgery. Frontal resection was performed in 12 patients, anterior temporal lobectomy in eight, amygdalohippocampectomy in two, and frontoparietal resection in one. The pathological diagnosis was focal cortical dysplasia (FCD) in eight patients, mesial temporal sclerosis (MTS) in nine, others in two, and no abnormality in four. The abilities of IMZ- and IMP-SPECT to detect abnormalities and epileptic foci were compared. Sensitivity was assessed in all patients, and specificity was assessed in 14 patients who were seizure free for >1 year after operation. Moreover, to investigate the difference in detectability according to pathological background, the detection rates of MTS and FCD by both methods were compared.

Results: In all patients, the sensitivity was 61% (14 patients) for IMZ-SPECT and 28% (eight patients) for IMP-SPECT. Neither IMZ- nor IMP-SPECT could detect epileptic foci in seven (30%) patients. In fourteen patients who were seizure free for >1 year after operation, the specificity was 86% (12 patients) for IMZ-SPECT and 36% (five patients) for Iological findings, the detection rates of FCD and MTS were, respectively, 75% (six patients) and 56% (five patients) for IMZ-SPECT, and 28% (one patient) and 56% (five patients) for IMP-SPECT. Despite the absence of definite abnormal pathology in four patients, IMZ- and IMP-SPECT detected an epileptic focus in one patient each.

Conclusions: These results show that, compared with IMP-SPECT, IMZ-SPECT is more useful for the detection of epileptic foci in refractory partial epilepsy showing no obvious MRI abnormality. IMZ-SPECT especially is effective to detect FCD. These findings indicate that IMZ-SPECT can play an important role in presurgical evaluation as a noninvasive interictal imaging technique.


*Jun Natsume, †Neda Bernasconi, †Masami Nishikawa, †Marco Fedi, †Francois Dubeau, †Frederick Andermann, †Andrea Bernasconi, and †Mirko Diksic *Department of Pediatrics, Japanese Red Cross Nagoya First Hospital, Nagoya, Aichi, Japan ; and †Department of Neurology and Neurosurgery, Montreal Neurological Institute and Hospital, McGill University, Montreal, Quebec, Canada .

Background: Periventricular nodular heterotopia (PVNH) is a commonly identified malformation of cortical development. Although the association between PVNH and epilepsy is well known, controversy exists concerning the location of the epileptogenic area. Some evidence suggests an involvement of the neocortex. We recently demonstrated that(α-[11C]methyl-l-tryptophan (α-MTrp) positron emission tomography (PET) is a useful tool in the localization of the epileptogenic area in partial epilepsy, especially in patients with tuberous sclerosis or cortical dysplasia.

Purpose: To determine whether α-MTrp PET can detect neocortical changes in PVNH secondary to abnormal serotoninergic or kynurenine pathway metabolism.

Methods: We studied four patients (mean age, 28 years; range, 23–36 years) with PVNH and intractable seizures by using α-MTrp PET. Three had bilateral PVNH, and one had unilateral PVNH. The net α-MTrp uptake constant K* was calculated from the time–activity curve, and the input function estimated from the venous sinus radioactivity and blood samples. We compared the K* image of each patient with those of 21 normal control subjects (mean age, 35 ± 14 years) by using the statistical parametric mapping (SPM) program. We compared the location of increased uptake of α-MTrp with the location of epileptic EEG abnormalities. We defined a significant cluster as a cluster with a height of p < 0.005 and an extent threshold of 100. When no clusters met this definition, we examined again with a height of p < 0.05.

Results: In patient 1, EEG localized seizure onset in the right posterior temporal neocortex overlying the PVNH. PET revealed increased α-MTrp uptake in the same area, as well as in the right frontotemporal and the left frontopolar areas. In patients 2 and 3, α-MTrp PET showed increased uptake in the areas of the additional seizure foci, but not the areas of seizure generation depicted on EEG. In patient 4 with unilateral PVNH, multifocal increases of α-MTrp uptake were observed. α-MTrp PET did not show increased uptake in the PVNH itself in all the patients studied.

Conclusions: In PVNH, α-MTrp PET detected abnormal metabolism of tryptophan in the neocortex. In some patients, the increase was seen in several neocortical areas but never included the heterotopic nodules. Based on our previous observations, this finding suggests epileptogenicity in the neocortex overlying PVNH. The pathologic mechanism of the abnormality underlying the widespread alteration of cortical tryptophan metabolism extending beyond the MRI-visible lesion remains unclear. Neocortical areas overlying PVNH display abnormal tryptophan metabolism. This abnormality is suggestive of a widespread epileptogenic lesion extending beyond the MRI-visible lesion.