- Email: [email protected]
Epilepsy surgery, resection volume and MSI localization in lesional frontal lobe epilepsy
Brief report
www.elsevier.com/locate/ynimg NeuroImage 21 (2004) 444 – 449
Epilepsy surgery, resection volume and MSI localization in lesional frontal lobe epilepsy A. Genow, a,* C. Hummel, a G. Scheler, a R. Hopfenga¨rtner, a M. Kaltenha¨user, a M. Buchfelder, b J. Romsto¨ck, b and H. Stefan a a b
Neurological Clinic, Department of Neurology, Epilepsy Center, University Erlangen-Nu¨rnberg, D-91054 Erlangen, Germany Department of Neurosurgery, University of Erlangen-Nuremberg, Erlangen, Germany
Received 22 May 2003; revised 20 August 2003; accepted 22 August 2003
To verify whether interictal noninvasive information detected by magnetoencephalography (MEG) recordings can contribute to localize focal epileptic activity relevant for seizure generation in lesional frontal lobe epilepsy, magnetic source imaging (MSI) localizations of epileptic discharges were compared to the extent of neurosurgical resection and postoperative outcome. Preoperative MEG spike localizations were displayed in postoperative magnetic resonance imaging (MRI) scans to check whether dipole sites were located within the resection cavity. Moreover, MEG localizations were compared with results of prolonged video-EEG monitoring and, in three cases, with invasive EEG recordings. Our results in five cases with lesional frontal lobe epilepsy showed that good surgical outcome could be achieved in those patients where the majority of MEG spike localizations were located within the resected brain volume. D 2003 Elsevier Inc. All rights reserved. Keywords: Frontal lobe epilepsy; Magnetic source imaging; Brain resection volume
Introduction Epilepsy surgery in principle aims at the control of seizures in patients with pharmacoresistant epilepsies while avoiding neurological or neuropsychological deficits. Identification and delineation of the epileptogenic zone, defined as the brain region that generates a patient’s habitual focal seizures and has to be removed to abolish seizures completely (Lu¨ders, 1991), represents the main task in presurgical epilepsy evaluation. For this purpose, a variety of noninvasive and invasive diagnostic tools are applied. Noninvasive investigations include intensive video-EEG monitoring, high-resolution magnetic resonance imaging (MRI), single photon emission computerized tomography (SPECT), positron emission
* Corresponding author. Neurological Clinic, Department of Neurology, Epilepsy Center, University of Erlangen-Nu¨rnberg, Schwabachanlage 6, D-91054 Erlangen, Germany. Fax: +49-9131-8536469. E-mail address: [email protected] (A. Genow). Available online on ScienceDirect (www.sciencedirect.com.) 1053-8119/$ - see front matter D 2003 Elsevier Inc. All rights reserved. doi:10.1016/j.neuroimage.2003.08.029
tomography (PET) and neuropsychological testing. In many cases, these techniques provide ample evidence of brain regions involved in seizure generation. Nevertheless, in a significant number of patients, the epileptogenic zone cannot be identified conclusively by these methods, and invasive EEG recordings are required to detect focal onset of epileptic activity (Engel and Ojemann, 1993). Invasive procedures, however, not only bear considerable risks of intracranial infection or bleeding and often result in prolonged hospitalization and postponement of surgery (Engel and Ojemann, 1993) but the limited ‘‘tunnel view’’ of invasive recordings also requires detailed planning of electrode placement. Therefore, delineation of the epileptogenic zone with noninvasive techniques should be optimized. Magnetoencephalography (MEG), a neurophysiological method with high temporal and spatial resolution, is increasingly applied to improve noninvasive presurgical focus localization and delineation of functionally important areas (Baumgartner et al., 2000; Ebersole et al., 1995; Iwasaki et al., 2002; Pataraia et al., 2002). Magnetic source imaging (MSI), the combination of MEG source localization with co-registered MRI data, does not only permit localizations at the surface of the brain, as obtained by electrocorticography, but provides a three-dimensional view of source localizations, even in deeper brain structures or in association with structural lesions. Delineation of the epileptogenic zone in extratemporal epilepsies (ETE), especially frontal lobe epilepsies, is considered to be much more demanding than in temporal lobe epilepsies (TLE), as the frontal cortex covers 40% of the brain, and propagation of epileptic activity can be fast (Kotagal and Arunkumar, 1998). Despite these difficulties, MEG has been proven to be particularly useful in localization of the epileptogenic zone in ETE (Nakasato et al., 1994; Smith et al., 1995; Stefan et al., 2000). The aim of this study was to validate MEG localizations in frontal lobe epilepsy by analyzing congruency among presurgical focus localization by MEG, resection volume and postoperative outcome. Moreover, MSI results were compared to results of prolonged video-EEG monitoring in all patients, to recordings of invasive strip and depth electrodes (case 4) and intraoperative electrocorticogram (cases 2 and 3).
A. Genow et al. / NeuroImage 21 (2004) 444–449
Patients and methods So far, data of five patients suffering from frontal lobe epilepsy was analyzed. All patients underwent presurgical epilepsy evaluation in the Erlangen Epilepsy Center (ZEE). Case 1: Four weeks before admission to the hospital for epilepsy assessment, the 20-year-old patient developed simple partial seizures in the right arm, initiating with loss of control followed by clonic jerks. Once, speech arrest and a tonic clonic seizure occurred. MRI scans showed a calcified mass in the left frontal lobe. Case 2: The patient was suffering from symptomatic epilepsy since the age of 12 because of an arachnoidal cyst in the left frontal lobe. Resection in the same year resulted in temporary abolishment of seizures. From the age of 15, frequent seizures reoccurred, presenting with staring and fumbling, only rarely generalizing into tonic clonic seizures. A second course of epilepsy evaluation was performed at the age of 19. Case 3: After epilepsy onset at the age of 2, the patient was suffering from seizures initiating with a dubious feeling in the head, followed by tonic cramping of the facial muscles and elevation of both arms. Occasionally, the patient fell over but did not lose consciousness. Epilepsy surgery in the left supplementary motor area was performed at the age of 24, but after a seizure-free period of 6 months, seizures reoccurred nearly daily. Seizure semiology was resembling that before surgery. The patient was admitted for a second course of presurgical epilepsy evaluation at the age of 29. Case 4: From the age of 6, the patient was suffering from symptomatic epilepsy because of a cystic lesion in the left frontal lobe. Seizures started with a feeling of anxiety or brief twitching, followed by staring, fumbling and blurring of consciousness. Once a month the patient had seizures with tonic head movement to the right side, generalizing into tonic clonic seizures. At the age of 32, presurgical epilepsy evaluation was performed. Case 5: By the age of 17, the patient suffered from bleeding of an angioma in the left frontal lobe, requiring surgical treatment. Four years later, she started having complex partial seizures with tingling in the head, staring and oral automatism, occasionally generalizing into tonic clonic seizures. By the age of 24, the patient was admitted for presurgical epilepsy evaluation. Recording procedure All patients had MEG investigations during the routinely performed presurgical epilepsy evaluation. In all patients, spontaneous magnetic brain activity was continuously recorded by a 74channel dual unit MEG system (Magnes II, 4-D Neuroimaging, San Diego, CA, USA; bandpass 1 – 100 Hz, sampling rate 512.8). Data analysis Interictal MEG data was bandpass-filtered offline (3 – 70 Hz) and visually inspected for specific epileptic discharges. As a standard procedure, epileptic discharges detected in MEG data were localized using the model of an equivalent current dipole in a homogeneously conducting sphere (4-D Neuroimaging software, version Magnes 1.2.5). To obtain the earliest possible localizations, rising slopes and peaks of spikes and sharp waves were selected for localization procedures. Among a number of interictal spikes, only those dipole localizations with good fitting criteria (confidence volume < 3 mm3, correlation between the empirical and the
445
configuration calculated according to the dipole model >98%) were included in further analysis. Localized dipoles were displayed using the patients’ individual preoperative MRI scans. To merge MEG localizations with MRI scans, electrophysiological and image data were co-registered, using predefined anatomical landmarks (nasion, left and right preauricular points), serving to establish a common spatial coordinate system. Before each MEG investigation, anatomical landmarks were digitized, using a three-dimensional digitizing device (Polhemus, Colchester, VT, USA). These landmarks were traced in the MRI scans by adhesive contrasting markers. To exactly retrieve anatomical landmarks during electrophysiological and pre- and postoperative MRI investigations, digital photographs of the tagged landmarks were taken before the first investigation. Preoperative dipole localizations were displayed in corresponding postoperative MRI images, taken at least 5 months after neurosurgery, by superimposing pre- and postoperative anatomical landmarks. Dipole localizations were compared with the site and extent of surgery. Outcome was assessed according to Engel’s classification during routinely performed postoperative follow-up examinations by a physician who was blinded with respect to MEG localization results. The first postoperative checkups were performed approximately 6 months after surgery, the latest ones, at the time of data analysis, between 6 and 60 months after surgery.
Results Case 1: MSI yielded dipole sources at the border of the lesion in the left frontal lobe. During prolonged video-EEG monitoring, ictal recordings showed high-frequency seizure pattern in electrodes F3 and Fz. Lesionectomy revealed a calcified cavernoma. Two years after surgery, the patient was still completely seizure-free. Superimposed preoperative MSI results and postoperative MRI scans showed that preoperative dipole localizations clustered well within the resection cavity (Fig. 1). Case 2: MSI results showed a dipole cluster localizing epileptic discharges at the upper border of the resection cavity of the first epilepsy surgery. During prolonged video-EEG monitoring, seizure onset was recorded in left fronto – centro – parietal electrodes. Intraoperative electrocorticogram revealed major spiking activity in the region of the MSI localization. Tailored resection of tissue at the upper border of the lesion revealed a dysembrioplastic neuroepithelial tumor. The patient significantly benefited from his second surgery. Seizure frequency was markedly reduced from about 100 to 4 complex partial seizures per year after surgery. Superimposed preoperative MEG localizations and postoperative MRI scans showed that all dipoles were located within the resection cavity (Fig. 2). Case 3: Dipole localizations of MEG spikes revealed a center of epileptic activity at the anterior margin of the resection cavity of the first epilepsy surgery. Because of muscle artifacts, ictal EEG recordings did not yield worthwhile information about seizure onset. No epileptic discharges could be recorded during interictal EEG recordings. During intraoperative electrocorticography, major spiking could be recorded at the site of the MSI localization. In a second surgical approach, the previous resection was extended, the left frontal pole was resected, followed by multiple subpial transsections. The patient became completely seizure-free. The last follow-up investigation was performed 5 years after surgery.
446
A. Genow et al. / NeuroImage 21 (2004) 444–449
Fig. 1. (a) Preoperative MSI localizations; (b) resection cavity; (c) preoperative dipole localizations projected onto postoperative MRI scan.
Projection of preoperative MEG spike localizations on postoperative MRI scans showed that all dipoles were located inside the extended resection cavity (Fig. 3). Case 4: MSI findings showed sources of epileptic activity involving both frontal lobes. Spikes were not located at the very border of the lesion, but rather widely distributed towards superior – posterior regions. Ictal EEG recordings showed bifrontal rhythmic activity; during one seizure, early seizure pattern was recorded in left frontocentral electrodes. During invasive recordings, including frontolateral and frontomesial strip and intracerebral depth electrodes, a widely distributed seizure pattern was recorded. Extended
lesionectomy was performed. Histology revealed an ependymoma WHO grade 2. 16 months after surgery, the patient reported no significant reduction of seizure frequency. However, seizure intensity was reduced, having only complex partial seizures, just once generalizing into a tonic clonic seizure. Projection of MSI results onto postoperative MRI scans showed that all dipole localizations were located outside of the resection cavity (Fig. 4). Case 5: MEG localizations of early spike components of complex poly spikes where localized at the upper, mesial border of the lesion. Prolonged video-EEG monitoring showed seizure pattern pointing to deep frontal structures close to midline without clear
Fig. 2. (a) Preoperative MSI localizations at the upper border of resection cavity of the first surgery; MRI scan and three-dimensional rendering of cortex surface; (b) schematic drawing of electrode placement during intraoperative electrocorticography, hatched area indicating the lesion; red: major spike activity; yellow: little spike activity; (c) preoperative dipole localizations projected onto postoperative MRI scan.
A. Genow et al. / NeuroImage 21 (2004) 444–449
447
Fig. 3. (a) MSI results showing spike localizations at the anterior border of the first resection cavity; (b) schematic drawing of electrode placement during intraoperative electrocorticography, hatched area indicating the lesion; red: major spike activity; yellow: little spike activity; (c) MEG localizations projected into corresponding postoperative MRI scan showing the expanded resection cavity.
Fig. 4. (a) Schematic drawing and X-ray of invasive strip and depth electrodes. 1 – 6: frontolateral strip; 7 – 10: frontomesial strip; red: electrodes showing seizure pattern during seizure onset; (b) preoperative MRI scan showing ependymoma in the left frontal lobe; (c) preoperative MSI results: bilateral dipole clusters; (d) preoperative dilope localizations projected onto postoperative MRI scan, arrows indicating resection cavity.
448
A. Genow et al. / NeuroImage 21 (2004) 444–449
Fig. 5. (a) Preoperative MEG localization results at the upper-mesial border of the lesion; (b) preoperative MSI results projected onto postoperative MRI scans.
lateralization. Extended lesionectomy was perfomed. The patient, preoperatively having three to four seizures per month, is postoperatively seizure-free for 6 months now. Projection of preoperative spike localizations onto postoperative MRI scans showed that spike localizations were located within the resection cavity (Fig. 5). Merging preoperative dipole localizations with postoperative MRI scans showed that in those patients with good seizure outcome (Engel 1 or 2), the majority of dipole localizations were located Table 1 Number of spike localizations within/outside resection cavity and postoperative outcome (Engel’s classification) Patient
Spike localizations within resection cavity
Spike localizations outside resection cavity
Latest follow up (months after surgery)
Outcome
1 2 3 4 5
13 21 2 0 2
2 0 0 10 0
23 18 60 16 6
1A 2B 1A 4A 1A
within the resection cavity. In one patient with minor seizure outcome (Engel 3), dipole localizations were located outside of the resected volume (Table 1).
Discussion Definition of the exact resection volume is a major difficulty in postoperative evaluation of epilepsy surgery. As the extent of resection may be overestimated by the operating surgeon (Schwartz et al., 2002), evaluation of postoperative MRI scans indicating the correct resection volume is crucial. In this study, good correlation of MEG localizations with pre- and postoperative MRI was enabled by using a common spatial coordinate system for electrophysiological and imaging data. By matching MSI localizations with pre- and postoperative MRI data, dipole sites could be precisely checked to be located within or outside the resected brain volume. In four lesional frontal lobe epilepsy cases who had good surgical outcome, the exact correlation between source localization of epileptic activity by MEG and the resected brain volume could be demonstrated for the first time. This explains good seizure control after
A. Genow et al. / NeuroImage 21 (2004) 444–449
epilepsy surgery and is in agreement with findings of King et al. (2000) who were able to show that resection of MEG spike foci correlated strongly with good outcome. Despite pessimistic reports on reoperation after unsuccessful epilepsy surgery (Schwartz and Spencer, 2001), it could be demonstrated that the extent of the second resection can be guided by MSI, with good postoperative outcome. In the two reoperated cases, localization of focal epileptic activity was not found in the whole circumference of the resection cavity of the first surgery, but rather in a distinct predominant area at the border of the lesion, showing that MEG localizations contributed substantially to the second course presurgical evaluation which was followed by final successful resective surgery. In two patients (cases 3 and 5), MEG yielded crucial information about spatial relationship of epileptic activity to the epileptogenic lesion, whereas during prolonged video-EEG monitoring, localization of ictal activity was poor. The results presented here in five cases with lesional frontal lobe epilepsy are encouraging, as clinically relevant localizations could be established using MSI of interictal epileptic activity. Further investigations will be needed to confirm our results in a larger number of patients. Future research should concern the following topics: 1. In which cases of frontal lobe epilepsies is interictal MEG analysis sufficient to localize the relevant seizure focus? 2. Is it possible to show congruency between MSI localization, resection volume and postoperative outcome in other extratemporal epilepsies and in temporal lobe epilepsy? 3. Which minimum resection volume is required to control seizures and to avoid unwanted neurological or cognitive effects? 4. Can the integration of source localizations by MSI in intraoperative neuronavigation be used to improve determination of resection volumes in epilesy surgery, especially if intraoperative MRI (Ganslandt et al., 1999; Nimsky et al., 2001) is applied?
Acknowledgments Part of this research was supported by Deutsche Forschungsgemeinschaft (DFG) Grant STE 380/9.
449
References Baumgartner, C., Pataraia, E., Lindinger, G., Deecke, L., 2000. Magnetoencephalography in focal epilepsy. Epilepsia 41 (Suppl. 3), 39 – 47. Ebersole, J.S., Squires, K.C., Eliashiv, S.D., Smith, J.R., 1995. Applications of magnetic source imaging in evaluation of candidates for epilepsy surgery. Neuroimaging Clin. N. Am. 5, 267 – 288. Engel, J.J., Ojemann, G.A., 1993. The next step. In: Engel, J. (Ed.), Surgical Treatment of the Epilepsies. Raven Press, New York, pp. 319 – 329. Ganslandt, O., Fahlbusch, R., Nimsky, C., Kober, H., Moller, M., Steinmeier, R., Romsto¨ck, J., Vieth, J., 1999. Functional neuronavigation with magnetoencephalography: outcome in 50 patients with lesions around the motor cortex. J. Neurosurg. 91, 73 – 79. Iwasaki, M., Nakasato, N., Shamoto, H., Nagamatsu, K., Kanno, A., Hatanaka, K., Yoshimoto, T., 2002. Surgical implications of neuromagnetic spike localization in temporal lobe epilepsy. Epilepsia 43, 415 – 424. King, D.W., Park, Y.D., Smith, J.R., Wheless, J.W., 2000. Magnetoencephalography in neocortical epilepsy. Adv. Neurol. 84, 415 – 423. Kotagal, P., Arunkumar, G.S., 1998. Lateral frontal lobe seizures. Epilepsia 39 (Suppl. 4), 62 – 68. Lu¨ders, H.O. (Ed.), 1991. Epilepsy Surgery. Raven Press, New York, pp. 337 – 347. Nakasato, N., Levesque, M.F., Barth, D.S., Baumgartner, C., Rogers, R.L., Sutherling, W.W., 1994. Comparisons of MEG, EEG, and ECoG source localization in neocortical partial epilepsy in humans. Electroencephalogr. Clin. Neurophysiol. 91, 171 – 178. Nimsky, C., Ganslandt, O., Kober, H., Buchfelder, M., Fahlbusch, R., 2001. Intraoperative magnetic resonance imaging combined with neuronavigation: a new concept. Neurosurgery 48, 1082 – 1089 (discussion 1089-91). Pataraia, E., Baumgartner, C., Lindinger, G., Deecke, L., 2002. Magnetoencephalography in presurgical epilepsy evaluation. Neurosurg. Rev. 25, 141 – 159 (discussion 160-1). Schwartz, T.H., Spencer, D.D., 2001. Strategies for reoperation after comprehensive epilepsy surgery. J. Neurosurg. 95, 615 – 623. Schwartz, T.H., Marks, D., Pak, J., Hill, J., Mandelbaum, D.E., Holodny, A.I., Schulder, M., 2002. Standardization of amygdalohippocampectomy with intraoperative magnetic resonance imaging: preliminary experience. Epilepsia 43, 430 – 436. Smith, J.R., Schwartz, B.J., Gallen, C., Orrison, W., Lewine, J., Murro, A.M., King, D.W., Park, Y.D., 1995. Multichannel magnetoencephalography in ablative seizure surgery outside the anteromesial temporal lobe. Stereotact. Funct. Neurosurg. 65, 81 – 85. Stefan, H., Hummel, C., Hopfenga¨rtner, R., Pauli, E., Tilz, C., Ganslandt, O., Kober, H., Mo¨ller, M., Buchfelder, M., 2000. Magnetoencephalography in extratemporal epilepsy. J. Clin. Neurophysiol. 17, 190 – 200.
www.elsevier.com/locate/ynimg NeuroImage 21 (2004) 444 – 449
Epilepsy surgery, resection volume and MSI localization in lesional frontal lobe epilepsy A. Genow, a,* C. Hummel, a G. Scheler, a R. Hopfenga¨rtner, a M. Kaltenha¨user, a M. Buchfelder, b J. Romsto¨ck, b and H. Stefan a a b
Neurological Clinic, Department of Neurology, Epilepsy Center, University Erlangen-Nu¨rnberg, D-91054 Erlangen, Germany Department of Neurosurgery, University of Erlangen-Nuremberg, Erlangen, Germany
Received 22 May 2003; revised 20 August 2003; accepted 22 August 2003
To verify whether interictal noninvasive information detected by magnetoencephalography (MEG) recordings can contribute to localize focal epileptic activity relevant for seizure generation in lesional frontal lobe epilepsy, magnetic source imaging (MSI) localizations of epileptic discharges were compared to the extent of neurosurgical resection and postoperative outcome. Preoperative MEG spike localizations were displayed in postoperative magnetic resonance imaging (MRI) scans to check whether dipole sites were located within the resection cavity. Moreover, MEG localizations were compared with results of prolonged video-EEG monitoring and, in three cases, with invasive EEG recordings. Our results in five cases with lesional frontal lobe epilepsy showed that good surgical outcome could be achieved in those patients where the majority of MEG spike localizations were located within the resected brain volume. D 2003 Elsevier Inc. All rights reserved. Keywords: Frontal lobe epilepsy; Magnetic source imaging; Brain resection volume
Introduction Epilepsy surgery in principle aims at the control of seizures in patients with pharmacoresistant epilepsies while avoiding neurological or neuropsychological deficits. Identification and delineation of the epileptogenic zone, defined as the brain region that generates a patient’s habitual focal seizures and has to be removed to abolish seizures completely (Lu¨ders, 1991), represents the main task in presurgical epilepsy evaluation. For this purpose, a variety of noninvasive and invasive diagnostic tools are applied. Noninvasive investigations include intensive video-EEG monitoring, high-resolution magnetic resonance imaging (MRI), single photon emission computerized tomography (SPECT), positron emission
* Corresponding author. Neurological Clinic, Department of Neurology, Epilepsy Center, University of Erlangen-Nu¨rnberg, Schwabachanlage 6, D-91054 Erlangen, Germany. Fax: +49-9131-8536469. E-mail address: [email protected] (A. Genow). Available online on ScienceDirect (www.sciencedirect.com.) 1053-8119/$ - see front matter D 2003 Elsevier Inc. All rights reserved. doi:10.1016/j.neuroimage.2003.08.029
tomography (PET) and neuropsychological testing. In many cases, these techniques provide ample evidence of brain regions involved in seizure generation. Nevertheless, in a significant number of patients, the epileptogenic zone cannot be identified conclusively by these methods, and invasive EEG recordings are required to detect focal onset of epileptic activity (Engel and Ojemann, 1993). Invasive procedures, however, not only bear considerable risks of intracranial infection or bleeding and often result in prolonged hospitalization and postponement of surgery (Engel and Ojemann, 1993) but the limited ‘‘tunnel view’’ of invasive recordings also requires detailed planning of electrode placement. Therefore, delineation of the epileptogenic zone with noninvasive techniques should be optimized. Magnetoencephalography (MEG), a neurophysiological method with high temporal and spatial resolution, is increasingly applied to improve noninvasive presurgical focus localization and delineation of functionally important areas (Baumgartner et al., 2000; Ebersole et al., 1995; Iwasaki et al., 2002; Pataraia et al., 2002). Magnetic source imaging (MSI), the combination of MEG source localization with co-registered MRI data, does not only permit localizations at the surface of the brain, as obtained by electrocorticography, but provides a three-dimensional view of source localizations, even in deeper brain structures or in association with structural lesions. Delineation of the epileptogenic zone in extratemporal epilepsies (ETE), especially frontal lobe epilepsies, is considered to be much more demanding than in temporal lobe epilepsies (TLE), as the frontal cortex covers 40% of the brain, and propagation of epileptic activity can be fast (Kotagal and Arunkumar, 1998). Despite these difficulties, MEG has been proven to be particularly useful in localization of the epileptogenic zone in ETE (Nakasato et al., 1994; Smith et al., 1995; Stefan et al., 2000). The aim of this study was to validate MEG localizations in frontal lobe epilepsy by analyzing congruency among presurgical focus localization by MEG, resection volume and postoperative outcome. Moreover, MSI results were compared to results of prolonged video-EEG monitoring in all patients, to recordings of invasive strip and depth electrodes (case 4) and intraoperative electrocorticogram (cases 2 and 3).
A. Genow et al. / NeuroImage 21 (2004) 444–449
Patients and methods So far, data of five patients suffering from frontal lobe epilepsy was analyzed. All patients underwent presurgical epilepsy evaluation in the Erlangen Epilepsy Center (ZEE). Case 1: Four weeks before admission to the hospital for epilepsy assessment, the 20-year-old patient developed simple partial seizures in the right arm, initiating with loss of control followed by clonic jerks. Once, speech arrest and a tonic clonic seizure occurred. MRI scans showed a calcified mass in the left frontal lobe. Case 2: The patient was suffering from symptomatic epilepsy since the age of 12 because of an arachnoidal cyst in the left frontal lobe. Resection in the same year resulted in temporary abolishment of seizures. From the age of 15, frequent seizures reoccurred, presenting with staring and fumbling, only rarely generalizing into tonic clonic seizures. A second course of epilepsy evaluation was performed at the age of 19. Case 3: After epilepsy onset at the age of 2, the patient was suffering from seizures initiating with a dubious feeling in the head, followed by tonic cramping of the facial muscles and elevation of both arms. Occasionally, the patient fell over but did not lose consciousness. Epilepsy surgery in the left supplementary motor area was performed at the age of 24, but after a seizure-free period of 6 months, seizures reoccurred nearly daily. Seizure semiology was resembling that before surgery. The patient was admitted for a second course of presurgical epilepsy evaluation at the age of 29. Case 4: From the age of 6, the patient was suffering from symptomatic epilepsy because of a cystic lesion in the left frontal lobe. Seizures started with a feeling of anxiety or brief twitching, followed by staring, fumbling and blurring of consciousness. Once a month the patient had seizures with tonic head movement to the right side, generalizing into tonic clonic seizures. At the age of 32, presurgical epilepsy evaluation was performed. Case 5: By the age of 17, the patient suffered from bleeding of an angioma in the left frontal lobe, requiring surgical treatment. Four years later, she started having complex partial seizures with tingling in the head, staring and oral automatism, occasionally generalizing into tonic clonic seizures. By the age of 24, the patient was admitted for presurgical epilepsy evaluation. Recording procedure All patients had MEG investigations during the routinely performed presurgical epilepsy evaluation. In all patients, spontaneous magnetic brain activity was continuously recorded by a 74channel dual unit MEG system (Magnes II, 4-D Neuroimaging, San Diego, CA, USA; bandpass 1 – 100 Hz, sampling rate 512.8). Data analysis Interictal MEG data was bandpass-filtered offline (3 – 70 Hz) and visually inspected for specific epileptic discharges. As a standard procedure, epileptic discharges detected in MEG data were localized using the model of an equivalent current dipole in a homogeneously conducting sphere (4-D Neuroimaging software, version Magnes 1.2.5). To obtain the earliest possible localizations, rising slopes and peaks of spikes and sharp waves were selected for localization procedures. Among a number of interictal spikes, only those dipole localizations with good fitting criteria (confidence volume < 3 mm3, correlation between the empirical and the
445
configuration calculated according to the dipole model >98%) were included in further analysis. Localized dipoles were displayed using the patients’ individual preoperative MRI scans. To merge MEG localizations with MRI scans, electrophysiological and image data were co-registered, using predefined anatomical landmarks (nasion, left and right preauricular points), serving to establish a common spatial coordinate system. Before each MEG investigation, anatomical landmarks were digitized, using a three-dimensional digitizing device (Polhemus, Colchester, VT, USA). These landmarks were traced in the MRI scans by adhesive contrasting markers. To exactly retrieve anatomical landmarks during electrophysiological and pre- and postoperative MRI investigations, digital photographs of the tagged landmarks were taken before the first investigation. Preoperative dipole localizations were displayed in corresponding postoperative MRI images, taken at least 5 months after neurosurgery, by superimposing pre- and postoperative anatomical landmarks. Dipole localizations were compared with the site and extent of surgery. Outcome was assessed according to Engel’s classification during routinely performed postoperative follow-up examinations by a physician who was blinded with respect to MEG localization results. The first postoperative checkups were performed approximately 6 months after surgery, the latest ones, at the time of data analysis, between 6 and 60 months after surgery.
Results Case 1: MSI yielded dipole sources at the border of the lesion in the left frontal lobe. During prolonged video-EEG monitoring, ictal recordings showed high-frequency seizure pattern in electrodes F3 and Fz. Lesionectomy revealed a calcified cavernoma. Two years after surgery, the patient was still completely seizure-free. Superimposed preoperative MSI results and postoperative MRI scans showed that preoperative dipole localizations clustered well within the resection cavity (Fig. 1). Case 2: MSI results showed a dipole cluster localizing epileptic discharges at the upper border of the resection cavity of the first epilepsy surgery. During prolonged video-EEG monitoring, seizure onset was recorded in left fronto – centro – parietal electrodes. Intraoperative electrocorticogram revealed major spiking activity in the region of the MSI localization. Tailored resection of tissue at the upper border of the lesion revealed a dysembrioplastic neuroepithelial tumor. The patient significantly benefited from his second surgery. Seizure frequency was markedly reduced from about 100 to 4 complex partial seizures per year after surgery. Superimposed preoperative MEG localizations and postoperative MRI scans showed that all dipoles were located within the resection cavity (Fig. 2). Case 3: Dipole localizations of MEG spikes revealed a center of epileptic activity at the anterior margin of the resection cavity of the first epilepsy surgery. Because of muscle artifacts, ictal EEG recordings did not yield worthwhile information about seizure onset. No epileptic discharges could be recorded during interictal EEG recordings. During intraoperative electrocorticography, major spiking could be recorded at the site of the MSI localization. In a second surgical approach, the previous resection was extended, the left frontal pole was resected, followed by multiple subpial transsections. The patient became completely seizure-free. The last follow-up investigation was performed 5 years after surgery.
446
A. Genow et al. / NeuroImage 21 (2004) 444–449
Fig. 1. (a) Preoperative MSI localizations; (b) resection cavity; (c) preoperative dipole localizations projected onto postoperative MRI scan.
Projection of preoperative MEG spike localizations on postoperative MRI scans showed that all dipoles were located inside the extended resection cavity (Fig. 3). Case 4: MSI findings showed sources of epileptic activity involving both frontal lobes. Spikes were not located at the very border of the lesion, but rather widely distributed towards superior – posterior regions. Ictal EEG recordings showed bifrontal rhythmic activity; during one seizure, early seizure pattern was recorded in left frontocentral electrodes. During invasive recordings, including frontolateral and frontomesial strip and intracerebral depth electrodes, a widely distributed seizure pattern was recorded. Extended
lesionectomy was performed. Histology revealed an ependymoma WHO grade 2. 16 months after surgery, the patient reported no significant reduction of seizure frequency. However, seizure intensity was reduced, having only complex partial seizures, just once generalizing into a tonic clonic seizure. Projection of MSI results onto postoperative MRI scans showed that all dipole localizations were located outside of the resection cavity (Fig. 4). Case 5: MEG localizations of early spike components of complex poly spikes where localized at the upper, mesial border of the lesion. Prolonged video-EEG monitoring showed seizure pattern pointing to deep frontal structures close to midline without clear
Fig. 2. (a) Preoperative MSI localizations at the upper border of resection cavity of the first surgery; MRI scan and three-dimensional rendering of cortex surface; (b) schematic drawing of electrode placement during intraoperative electrocorticography, hatched area indicating the lesion; red: major spike activity; yellow: little spike activity; (c) preoperative dipole localizations projected onto postoperative MRI scan.
A. Genow et al. / NeuroImage 21 (2004) 444–449
447
Fig. 3. (a) MSI results showing spike localizations at the anterior border of the first resection cavity; (b) schematic drawing of electrode placement during intraoperative electrocorticography, hatched area indicating the lesion; red: major spike activity; yellow: little spike activity; (c) MEG localizations projected into corresponding postoperative MRI scan showing the expanded resection cavity.
Fig. 4. (a) Schematic drawing and X-ray of invasive strip and depth electrodes. 1 – 6: frontolateral strip; 7 – 10: frontomesial strip; red: electrodes showing seizure pattern during seizure onset; (b) preoperative MRI scan showing ependymoma in the left frontal lobe; (c) preoperative MSI results: bilateral dipole clusters; (d) preoperative dilope localizations projected onto postoperative MRI scan, arrows indicating resection cavity.
448
A. Genow et al. / NeuroImage 21 (2004) 444–449
Fig. 5. (a) Preoperative MEG localization results at the upper-mesial border of the lesion; (b) preoperative MSI results projected onto postoperative MRI scans.
lateralization. Extended lesionectomy was perfomed. The patient, preoperatively having three to four seizures per month, is postoperatively seizure-free for 6 months now. Projection of preoperative spike localizations onto postoperative MRI scans showed that spike localizations were located within the resection cavity (Fig. 5). Merging preoperative dipole localizations with postoperative MRI scans showed that in those patients with good seizure outcome (Engel 1 or 2), the majority of dipole localizations were located Table 1 Number of spike localizations within/outside resection cavity and postoperative outcome (Engel’s classification) Patient
Spike localizations within resection cavity
Spike localizations outside resection cavity
Latest follow up (months after surgery)
Outcome
1 2 3 4 5
13 21 2 0 2
2 0 0 10 0
23 18 60 16 6
1A 2B 1A 4A 1A
within the resection cavity. In one patient with minor seizure outcome (Engel 3), dipole localizations were located outside of the resected volume (Table 1).
Discussion Definition of the exact resection volume is a major difficulty in postoperative evaluation of epilepsy surgery. As the extent of resection may be overestimated by the operating surgeon (Schwartz et al., 2002), evaluation of postoperative MRI scans indicating the correct resection volume is crucial. In this study, good correlation of MEG localizations with pre- and postoperative MRI was enabled by using a common spatial coordinate system for electrophysiological and imaging data. By matching MSI localizations with pre- and postoperative MRI data, dipole sites could be precisely checked to be located within or outside the resected brain volume. In four lesional frontal lobe epilepsy cases who had good surgical outcome, the exact correlation between source localization of epileptic activity by MEG and the resected brain volume could be demonstrated for the first time. This explains good seizure control after
A. Genow et al. / NeuroImage 21 (2004) 444–449
epilepsy surgery and is in agreement with findings of King et al. (2000) who were able to show that resection of MEG spike foci correlated strongly with good outcome. Despite pessimistic reports on reoperation after unsuccessful epilepsy surgery (Schwartz and Spencer, 2001), it could be demonstrated that the extent of the second resection can be guided by MSI, with good postoperative outcome. In the two reoperated cases, localization of focal epileptic activity was not found in the whole circumference of the resection cavity of the first surgery, but rather in a distinct predominant area at the border of the lesion, showing that MEG localizations contributed substantially to the second course presurgical evaluation which was followed by final successful resective surgery. In two patients (cases 3 and 5), MEG yielded crucial information about spatial relationship of epileptic activity to the epileptogenic lesion, whereas during prolonged video-EEG monitoring, localization of ictal activity was poor. The results presented here in five cases with lesional frontal lobe epilepsy are encouraging, as clinically relevant localizations could be established using MSI of interictal epileptic activity. Further investigations will be needed to confirm our results in a larger number of patients. Future research should concern the following topics: 1. In which cases of frontal lobe epilepsies is interictal MEG analysis sufficient to localize the relevant seizure focus? 2. Is it possible to show congruency between MSI localization, resection volume and postoperative outcome in other extratemporal epilepsies and in temporal lobe epilepsy? 3. Which minimum resection volume is required to control seizures and to avoid unwanted neurological or cognitive effects? 4. Can the integration of source localizations by MSI in intraoperative neuronavigation be used to improve determination of resection volumes in epilesy surgery, especially if intraoperative MRI (Ganslandt et al., 1999; Nimsky et al., 2001) is applied?
Acknowledgments Part of this research was supported by Deutsche Forschungsgemeinschaft (DFG) Grant STE 380/9.
449
References Baumgartner, C., Pataraia, E., Lindinger, G., Deecke, L., 2000. Magnetoencephalography in focal epilepsy. Epilepsia 41 (Suppl. 3), 39 – 47. Ebersole, J.S., Squires, K.C., Eliashiv, S.D., Smith, J.R., 1995. Applications of magnetic source imaging in evaluation of candidates for epilepsy surgery. Neuroimaging Clin. N. Am. 5, 267 – 288. Engel, J.J., Ojemann, G.A., 1993. The next step. In: Engel, J. (Ed.), Surgical Treatment of the Epilepsies. Raven Press, New York, pp. 319 – 329. Ganslandt, O., Fahlbusch, R., Nimsky, C., Kober, H., Moller, M., Steinmeier, R., Romsto¨ck, J., Vieth, J., 1999. Functional neuronavigation with magnetoencephalography: outcome in 50 patients with lesions around the motor cortex. J. Neurosurg. 91, 73 – 79. Iwasaki, M., Nakasato, N., Shamoto, H., Nagamatsu, K., Kanno, A., Hatanaka, K., Yoshimoto, T., 2002. Surgical implications of neuromagnetic spike localization in temporal lobe epilepsy. Epilepsia 43, 415 – 424. King, D.W., Park, Y.D., Smith, J.R., Wheless, J.W., 2000. Magnetoencephalography in neocortical epilepsy. Adv. Neurol. 84, 415 – 423. Kotagal, P., Arunkumar, G.S., 1998. Lateral frontal lobe seizures. Epilepsia 39 (Suppl. 4), 62 – 68. Lu¨ders, H.O. (Ed.), 1991. Epilepsy Surgery. Raven Press, New York, pp. 337 – 347. Nakasato, N., Levesque, M.F., Barth, D.S., Baumgartner, C., Rogers, R.L., Sutherling, W.W., 1994. Comparisons of MEG, EEG, and ECoG source localization in neocortical partial epilepsy in humans. Electroencephalogr. Clin. Neurophysiol. 91, 171 – 178. Nimsky, C., Ganslandt, O., Kober, H., Buchfelder, M., Fahlbusch, R., 2001. Intraoperative magnetic resonance imaging combined with neuronavigation: a new concept. Neurosurgery 48, 1082 – 1089 (discussion 1089-91). Pataraia, E., Baumgartner, C., Lindinger, G., Deecke, L., 2002. Magnetoencephalography in presurgical epilepsy evaluation. Neurosurg. Rev. 25, 141 – 159 (discussion 160-1). Schwartz, T.H., Spencer, D.D., 2001. Strategies for reoperation after comprehensive epilepsy surgery. J. Neurosurg. 95, 615 – 623. Schwartz, T.H., Marks, D., Pak, J., Hill, J., Mandelbaum, D.E., Holodny, A.I., Schulder, M., 2002. Standardization of amygdalohippocampectomy with intraoperative magnetic resonance imaging: preliminary experience. Epilepsia 43, 430 – 436. Smith, J.R., Schwartz, B.J., Gallen, C., Orrison, W., Lewine, J., Murro, A.M., King, D.W., Park, Y.D., 1995. Multichannel magnetoencephalography in ablative seizure surgery outside the anteromesial temporal lobe. Stereotact. Funct. Neurosurg. 65, 81 – 85. Stefan, H., Hummel, C., Hopfenga¨rtner, R., Pauli, E., Tilz, C., Ganslandt, O., Kober, H., Mo¨ller, M., Buchfelder, M., 2000. Magnetoencephalography in extratemporal epilepsy. J. Clin. Neurophysiol. 17, 190 – 200.