By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Due to essential maintenance the subscribe/renew pages will be unavailable on Wednesday 26 October between 02:00- 08:00 BST/ 09:00 – 15:00 SGT/ 21:00- 03:00 EDT. Apologies for the inconvenience.
Address correspondence and reprint requests to Dr. H. Shiraishi at Nuclear Magnetic Resonance Center, Massachusetts General Hospital, Room 2301, Building 149, 13th Street, Charlestown, MA 02129-2060, U.S.A. E-mail: firstname.lastname@example.org
Summary: Purpose: To determine whether magnetoencephalography (MEG) has any clinical value for the analysis of seizure discharges in patients with medial frontal lobe epilepsy (FLE).
Methods: Four patients were studied with 74-channel MEG. Interictal and ictal electroencephalographic (EEG) and MEG recordings were obtained. The equivalent current dipoles (ECDs) of the MEG spikes were calculated.
Results: In two patients with postural seizures, interictal EEG spikes occurred at Cz or Fz. The ECDs of interictal MEG spikes were localized around the supplementary motor area. In the other two patients with focal motor or oculomotor seizures, interictal EEG spikes occurred at Fz or Cz. The ECDs of interictal MEG spikes were localized at the top of the medial frontal region. The ECDs detected at MEG ictal onset were also localized in the same area as those of the interictal discharges.
Conclusions: In medial FLE patients, interictal and ictal MEG indicated consistent ECD localization that corresponded to the semiology of clinical seizures. Our findings demonstrate that MEG is a useful tool for detecting epileptogenic focus.
In addition to polymorphic seizure manifestations, the variability of electroencephalographic (EEG) findings makes the diagnosis of frontal lobe epilepsy (FLE) often difficult. The true epileptogenic focus is especially difficult to determine in patients who have epileptogenic focus in the medial cerebral cortex and show EEG spikes in the midline region.
Magnetoencephalography (MEG) has several advantages over EEG in determining epileptogenic foci. Compared with an electrical field, a magnetic field used in MEG is less influenced by the differences in conductivities of various types of brain tissues. Moreover, MEG is capable of demonstrating epileptogenic foci located in deEPIy fissured tissues.
Several studies have analyzed the MEG localization of epileptogenic foci in the lateral cerebral cortex (1–7), and ictal MEG findings have been reported to express the true epileptogenic foci in patients with focal motor seizures (8,9). Nakasato et al. (1) examined the epileptogenic focus by MEG, electrocorticography (EcoG), and scalp EEG, and suggested that MEG is a powerful tool for the definition of epileptogenic focus of the brain. Furthermore Stefan et al. (10) concluded that the MEG recording is indispensable for the evaluation of candidates with extratemporal foci for epileptic surgery.
The purpose of the present study was to determine whether MEG has any clinical value for the analysis of epileptogenic foci in cases showing EEG spikes on the midline as the main interictal discharges.
PATIENTS AND METHODS
We studied four patients (three girls and one boy, aged 3–16 years) over a period of 16 to 31 months. In addition to routine diagnostic examinations [computed tomography (CT), magnetic resonance imaging (MRI), single-photon emission CT (SPECT), EEG, and neuropsychological testing], MEG studies were conducted. The subjects' parents or guardians gave their written informed consent for the MEG studies. The MEG was recorded with a 74-channel magnetometer (Magnes II; Biomagnetic Technologies Inc., San Diego, CA, U.S.A.) containing two 37-channel superconducting quantum intererference devices (SQUIDs). To ensure whole-brain mapping, the sensor positions were changed several times. EEG was recorded simultaneously for visual screening by using 20 scalp electrodes according to international 10-20 system, with two additional electrodes for electrocardiogram (ECG) monitoring. We performed the EEG recording at a time constant of 0.3 and no high-cut filtering. We usually recorded for a duration of >2 h for each patient. In case 4, we recorded for 5 h toward sleeping time to detect ictal discharge. The MEG and EEG were analyzed visually to capture the epileptiform discharges.
Brief MEG recordings were selected visually from each patient, and these recordings were sampled in 6-s blocks. The MEG data were digitally filtered with a bandpass of 4 to 100 Hz for offline analysis. We manually selected the segments containing abnormal paroxysms and calculated the equivalent current dipoles (ECDs) by the single-dipole model with a dipole-fit software (Biomagnetic Technologies). For a given MEG spike, we analyzed time slices for ∼100 ms at the vicinity of the MEG spike, and calculated the ECD continuously. We then selected the ECD that had the best correlation as the representative ECD of the particular MEG spike. Although we studied MEG spikes with dual SQUID sensor, we performed the dipole fit with 37-channel MEG recordings. A dipole fit was acceptable when there was a correlation of 0.98 in the comparison between an actual measurement and the theoretic value. The calculated ECDs were superimposed on each patient's MRI to map the location. Onset of seizure discharge on the MEG (MEG ictal onset) was defined as the beginning of low-amplitude spike rhythm. Because MEG ictal onset occurred several seconds before clinical seizure onset, the ictal MEG recording was not affected by seizure-induced body movements, and we were able to analyze the ictal MEG discharge.
SPECT was performed in all patients during the interictal and ictal periods, using a ring-type SPECT scanner (Headtome-SET070; Shimadzu Corp., Kyoto, Japan). We successfully injected the SPECT ligand at the ictal onset period, because we were at the side of the patient when we performed the ictal SPECT. 99mTc-ethyl-cysteinate dimer was injected intravenously at a dose of 370 MBq in patient 1 and 740 MBq in patients 2 to 4.
A 3-year-old boy was seen at age 6 months with daily seizures. At age 3 months, he began to have seizures with open eyes, deviation of the head and eyes to the left, and tonic extension of the left arm and leg, which was sometimes followed by clonic movement of both extremities. The duration of seizure was ∼15 s at the time of this study. In his seizure history, there were prolonged seizures for 30 s when his seizures were frequent.
He showed normal growth, with a developmental quotient of 84 (Tanaka–Binet test). No interictal neurologic deficits were observed, except temporary paralysis of the left leg when the seizures clustered.
Interictal EEG spikes occurred mainly at Cz (Fig. 1). In one ictal EEG recording, bilateral diffuse spikes appeared 10 s before the seizure. This rhythm was followed by Cz-dominant or bilateral anterior-dominant spike rhythm. The beginning of ictal discharge was not clearly localized on EEG.
On MEG recordings, ECDs derived from interictal MEG spikes were localized on the right frontomedial cortex (Fig. 1). In ictal MEG, the sensor located on the right central region detected spike rhythm before clinical seizure, which was suppressed at seizure onset, and was followed by recruiting spike rhythm. The ECDs analyzed from the spikes at MEG ictal onset were also localized around the right supplementary motor area (SMA) (Fig. 1). Because the patient's head was displaced from the MEG sensor 3 s from the seizure onset, discharges during the seizure could not be analyzed.
No abnormalities were found in MRI, and in the interictal and ictal SPECT.
A 16-year-old girl had daily seizures. At age 1 year, she began to have seizures with tonic extension and rotation of the extremities after bilateral perioral contraction. Seizures appeared almost symmetrical, and sometimes evolved to clonic seizure of both extremities. The duration of seizure was ∼20 s at the time of this study. Now the seizure frequency has decreased, but the seizure duration has not changed. There are seizures with only perioral contraction without impairment of consciousness.
Although she attended normal classes until age 15 years, she had to change to a special school for the handicapped after age 15 because of low intellectual performance. Her full-scale intelligence quotient (IQ) assessed by Wechler Adult Intelligence Scale–Revised (WAIS–R) was 45. She had no neurologic deficits.
Interictal EEG spikes occurred mainly at Fz (Fig. 2). On ictal EEG, spikes increased in frequency 8 or 9 s before a clinical seizure; they were attenuated at seizure onset, and were followed by bilateral diffuse spike rhythm (Fig. 2).
Interictal MEG spikes were clearly detected by the sensor located on the frontocentral region (Fig. 2). In the ictal MEG recording, spike rhythm was detected just before clinical seizure by the sensor located on the same region, but was strongly suppressed at seizure onset. The ECDs of the interictal MEG spikes clustered in the left frontal medial lobe, possibly near the SMA (Fig. 3). No significant ictal MEG spikes were available for ECD analysis.
A MRI fluid-attenuated inversion recovery (FLAIR) image detected a circumscribed lesion in the left medial cortex (Fig. 3). Although interictal SPECT showed no abnormal perfusion, ictal SPECT demonstrated greater increase in perfusion in the medial cortex of the left frontal lobe (Fig. 3). This finding supported the existence of an epileptogenic focus in the medial cortex.
A 15-year-old girl had daily seizures. At age 9 years, she began to have focal clonic seizures in the left leg while remaining conscious. The seizure sometimes evolved to clonic convulsion in both limbs with upward eye deviation, sometimes accompanied by falling. The duration of seizure was ∼20 s at the time we performed the study.
At age 11 years, her left leg became paralyzed, and she could not move her left foot to the flexor position. The paralysis was followed by claudication. Her early development was normal, but she had difficulty studying in elementary school and attended a special school for the handicapped after age 13 years. Her full-scale IQ assessed by Wechler Intelligence Scale for Children–III was 66.
Interictal EEG spikes were seen on Fz (Fig. 4). A recruiting spike rhythm during the ictal period also began on Fz (Fig. 4). This discharge propagated into a diffuse spike rhythm.
Interictal and ictal MEG spikes were clearly detected by the sensor located on the vertex. ECDs derived from interictal MEG spikes were localized in a tight cluster at the top of the right medial frontal area (Fig. 5), which is considered to control left foot movement. Sequential analysis of three 600-ms epochs of the discharges observed at MEG ictal onset showed that the ECDs were initially localized in the same area as the interictal discharges, and then moved to the left frontal medial cortex.
No abnormalities were observed in MRI, and in the interictal and ictal SPECT.
A 15-year-old girl had seizures. At age 12 years, she began to have oculomotor seizures with eye deviating to the left without saccadic movement, and dystonic posturing of the left hand, followed by upward eye deviation and blinking of the left eyelid. She remained conscious during the seizures. Some secondarily generalized tonic–clonic seizures were observed. The duration of the seizure was ∼15 s at the time of the present study.
Her development was normal, and her full-scale IQ assessed by WISC–R was 80. She had no neurologic deficits.
Interictal EEG spikes were seen on Cz and recruiting spike rhythm at ictus also appeared on Cz (Fig. 6). Ictal discharges were followed by bilateral diffuse spike rhythm.
Interictal and ictal MEG spikes were clearly detected by the sensor located on the frontocentral region. The ECDs derived from interictal MEG spikes were tightly localized at the right frontal medial cortex and neighboring dorsolateral cortex (Fig. 7). The ictal MEG discharges at three periods of the seizure were analyzed (Figs. 6 and 7). The ECDs derived from the spikes at MEG ictal onset were localized in the same region as the interictal discharges, and the ECDs for the ictal spikes 5 and 10 s after ictal onset showed a gradual shift to the left frontal medial cortex.
The MRI findings were unremarkable, and the interictal and ictal SPECT showed no abnormalities.
Although epileptogenic foci in the medial cortex of the brain can be localized to some extent by scalp-recorded interictal and ictal EEG, some studies suggested that the epileptogenic areas can be defined only by intracranial EEG recording (1). Because early ictal discharges originating from the SMA or the cingulate gyrus may be attenuated at all scalp electrodes, it can be difficult to identify the location of EEG onset of a seizure (11–13). The interictal and ictal MEG findings from individual patients can be summarized as follows. In patient 1, ECDs derived from interictal MEG discharges were located around the SMA or motor cortex in the medial frontal lobe. ECDs at MEG ictal onset also were located in this area. Clinical seizure manifestation in this patient usually began as left limb–dominant tonic posturing that sometimes evolved to left limb–dominant clonic seizure. These seizure symptoms correlated well with the seizure type usually generated from an epileptogenic focus in the cortex at the vicinity of SMA (11), as indicated by the ECD localization. In patient 2, the ECDs derived from interictal MEG discharges were located near the SMA, but ictal ECDs could not be analyzed because ictal MEG spike discharge was strongly suppressed. In this patient, MRI revealed a circumscribed lesion in the cortex, and ictal SPECT showed hyperperfusion in this area. In patient 3, both interictal and ictal ECDs were detected near the SMA or motor cortex of the medial portion in the right hemisphere. Clinically the patient had focal motor seizure of the left foot, which corresponded well to a focus in this region. However, we cannot explain the etiology of her paralysis. Left ankle paralysis appeared at age 11 years when seizures started to increase, and has become established to the extent that she has to use a device against spontaneous flexion. We speculate the etiology to be either focal cortical dysplasia together with impaired cortical function, or repetitive postictal paralysis that appears to be continuous because of frequent focal motor seizures. However, the etiology cannot be decided without pathological findings of the brain. In patient 4, the ECDs derived from MEG discharges during interictal period and at MEG ictal onset were both located near the SMA or premotor area. According to cortical stimulation studies using SPECT (14–-17), the majority of eye movements are elicited from the medial portion of the superior frontal gyrus and a few from the upper half of the anterior cingulate gyrus (15,17), although similar responses are also elicited by cortical stimulation of the lateral frontal eye field located in the dorsolateral cortex (11,16). The monoictal seizure symptoms of patient 4 might well originate from this area where clustering of the ECDs was observed.
In all cases, we performed simultaneous recording of both hemispheres with MEG sensors placed symmetrically. According to the results of this examination, we detected magnetic changes from the hemisphere in which the ECD localized during the interictal period for each case. However, we could not define any significant ECD because the correlation of dipole fits was very low. We therefore concluded that the source of epileptic discharge might not exist on the lateral convexity of the cerebral cortex.
In patients 3 and 4, ECDs determined from the interictal MEG discharges were located at the edge of the medial and lateral cortex of the frontal region, and ictal EEG showed initial recruiting spike rhythm at Fz and Cz followed by bilateral diffuse spike rhythm and diffuse slow waves. These findings suggest that, when patients have EEG patterns of interictal midline spikes and recruiting midline spike rhythm at ictal onset, the epileptogenic focus is located at the interhemispheric surface near the lateral cortex.
In MEG studies, analysis of ictal discharges has been complicated by body movements during the seizure. However, we detected ictal activity on the MEG in all our patients. The ECDs derived from the spikes at MEG ictal onset, which often preceded clinical seizure onset, and closely corresponded to the tightly clustered ECDs derived from the interictal discharges. Thus the ictal MEG supports findings obtained from the interictal MEG. Furthermore, we were able to show progressive shift of the ictal excitation by demonstrating progressive changes in ECD localization in a sequential analysis of the discharges at MEG ictal onset, and also in a periodic analysis of the discharges during seizure.
All patients in this study had EEG spikes on the midline as the major interictal discharge. Several researchers have argued over the etiology and clinical value of EEG spikes on the midline or vertex (11,12,18). Tükel and Jasper (18) used intracranial EEG monitoring to show that midline spikes were generated from an epileptogenic focus in the medial cortex of the brain. McLachlan et al. (19) found similarities between the scalp-recorded epileptiform discharges of a cat with an artificial epileptogenic focus in the interhemispheric region and the epileptiform discharges of a patient who had an epileptogenic lesion in the medial cortex of the brain. They concluded that midline spikes represent an abnormal discharge in the medial cortex. Our study confirmed these findings, because clusters of ECDs derived from these discharges were clearly located in the medial cortex of the frontal lobe in our patients, although midline spikes also occur in children with generalized seizures (20–22).