Epilepsy Research (2012) 99, 147—155
journal homepage: www.elsevier.com/locate/epilepsyres
Paradoxical lateralization of non-invasive electroencephalographic ictal patterns in extra-temporal epilepsies Claudia B. Catarino a,b, Christian Vollmar a, Soheyl Noachtar a,∗ a b
Epilepsy Center, Department of Neurology, University of Munich, 81377 Munich, Germany Department of Clinical and Experimental Epilepsy, UCL Institute of Neurology, Queen Square, London WC1N 3BG, UK
Received 22 September 2011; received in revised form 31 October 2011; accepted 6 November 2011 Available online 30 November 2011
KEYWORDS Electroencephalography; Partial epilepsy; Epilepsy surgery; Dipole analysis; Parasagittal
Summary Video-electroencephalographic (EEG) ictal recordings play an important role in the pre-surgical evaluation of patients with medically refractory focal epilepsy. Paradoxical lateralization of the scalp EEG ictal onset patterns, consistently contralateral to the side of the proven epileptogenic lesion is rare but important to recognize, with possible implications on patient management. We searched the database of the University of Munich Epilepsy Monitoring Unit for patients with extratemporal epilepsies, with scalp EEG ictal patterns consistently contralateral to the proven epileptogenic zone. All available clinical, EEG and imaging data were reviewed. Dipole source analysis of EEG seizure onset was performed where possible. Four patients were identified, who had proven paradoxical lateralization of scalp EEG ictal patterns, demonstrated by seizure freedom after epilepsy surgery, data from invasive electroencephalography, or imaging and seizure semiology. Parasagittal lesions on MRI brain scan were found in three cases. Invasive recordings with subdural electrodes were performed in one patient. Dipole source analysis of EEG seizure onset was possible in two patients, helping to correctly lateralize the ictal EEG pattern in one patient. Patients with midline or near midline neocortical seizure foci may show paradoxical lateralization of the ictal EEG, likely due to the spatial orientation of the cortical generators in the medial regions of the cerebral hemispheres. These patients may have excellent surgical outcome despite the apparently discordant EEG findings, making this an important phenomenon to be recognized in clinical practice. © 2011 Elsevier B.V. All rights reserved.
Abbreviations: DNET, dysembryoplastic neuroepithelial tumor; EEG, electroencephalography; EP, evoked potentials; FCD, focal cortical dysplasia; FLAIR, fluid-attenuated inversion recovery; MPRAGE, magnetization prepared rapid acquisition gradient echo. ∗ Corresponding author at: Epilepsy Center, Department of Neurology, Klinikum Grosshadern, University of Munich, Marchioninistr. 15, 81377 Munich, Germany. Tel.: +49 89 7095 3691; fax: +49 89 7095 6691.E-mail address:
[email protected] (S. Noachtar). 0920-1211/$ — see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.eplepsyres.2011.11.002
148
Introduction Paradoxical lateralization of ictal onset patterns on scalp EEG, contralateral to the epileptogenic zone, has been rarely reported, and may be under-recognized in clinical practice. Ictal recordings in parasagittal foci were recognized as being of difficult localization on scalp EEG several decades ago. One patient with a precentral glioma, in a large series of parasagittal lesions, had ictal EEG pattern of spikeslow wave of higher amplitude on the side contralateral to the lesion (Tukel and Jasper, 1952). Similar findings were also described in patients with ‘‘destructive parasagittal lesions’’, and were attributed to the effects of tissue loss (Penfield and Jasper, 1954). Two patients with epilepsia partialis continua of one lower extremity were shown to have epileptiform discharges paradoxically lateralized to the side of the clinical activity (Adelman et al., 1982; Oishi et al., 2002). More recently, six children with unilateral hemispheric encephaloclastic lesions were reported with paradoxical lateralization of ictal scalp EEG (Garzon et al., 2009). A similar phenomenon has been described in the setting of visual evoked potentials (EP) after half-field pattern visual stimulation (Barett et al., 1976), somatosensory EP after stimulation of the posterior tibial nerve (Cruse et al., 1982), and event-related potentials during imagined foot movements (Osman et al., 2005), where paradoxical lateralization occurs due to the spatial orientation of cortical generators on the mesial surface of the cortex, perpendicular to the interhemispheric fissure; these generators project parallel to the scalp surface, so recorded potentials show lateralization ipsilaterally to the stimulated side (Barett et al., 1976; Cruse et al., 1982). This has been proven for somatosensory cortical evoked responses by simultaneous recording of scalp and invasive electrodes placed into the interhemispheric fissure (Lesser et al., 1987). A fundamentally different, unrelated, phenomenon is the ‘‘false EEG lateralization’’ reported in mesial temporal lobe epilepsy (Castro et al., 2008; Engel and Crandall, 1983; Mintzer et al., 2004; Sammaritano et al., 1987), where the mechanism is thought to be rapid ictal activity spread to the contralateral side. We report here proven paradoxical lateralization of ictal EEG patterns in four patients with extra-temporal epilepsies, and three with mesial, parasagittal, epileptic foci.
C.B. Catarino et al. with mainly ipsilateral and only occasional contralateral ictal EEG were not included, as this could most likely reflect different pathways of ictal propagation, and patients with temporal lobe epilepsy were excluded.
EEG-video monitoring Non-invasive EEG-video recording with scalp electrodes was performed in all patients. The EEG was recorded with 32, 40 or 64-channel digital EEG machines (Vangard® , Cleveland, OH/USA and XLTEK® , London, Ontario, Canada), amplified, digitized at a sampling rate of 200 Hz (12 bit) and stored electronically for off-line analysis. Patients were monitored for four to nine days; and one to sixteen habitual seizures with EEG seizure pattern were recorded. In all patients, a protocol for antiepileptic drug reduction was applied during monitoring. Patient 1 underwent additional invasive EEGvideo recording with subdural electrodes covering the left central and interhemispheric regions. Electrodes were visualized by image co-registration of CT and MRI brain scans and subsequent volume rendering, allowing anatomic localization of the irritative zone and ictal onset zone (Vollmar et al., 2008).
Dipole source analysis For patients 1 and 2, dipole source analysis of the ictal seizure onset was performed, using BESA® 5.1 (Brain Electric Source Analysis) software (Megis, Munich, Germany). The seizure pattern at EEG onset consisted of rhythmic delta activity in the right frontocentral region, for patient 1; and consisted of stereotyped repetitive left occipital spikes, for patient 2. We used a realistic head model based on an averaged brain and the finite element method (Scherg and Ebersole, 1994). The dipole source localization was then superimposed on an averaged brain MRI in standard space.
MRI brain scan
Ethics
All patients had a brain MRI, 1.5-Tesla (Magnetom Vision, Siemens, Erlangen, Germany) or 1.0-Tesla (Impact, Siemens), including T1, T2 and proton-density weighted techniques, in axial, coronal and sagittal planes. For patients 1 and 2, this also included fluid-attenuated inversion recovery (FLAIR) sequences, and a volumetric T1 acquisition (magnetization prepared rapid acquisition gradient echo, MPRAGE).
This study has been approved by the local Research Ethics committee.
Results
Study design
Demographic and clinical data
The patients were evaluated at the Epilepsy Monitoring Unit of the University of Munich. Approximately 950 consecutive patients were screened for ictal EEG recordings contralateral to the known or hypothesized ictal onset zone. Patients
Four patients were included in the study, for whom it was possible to prove paradoxical lateralization of the noninvasive ictal EEG onset seizure patterns. Demographic and clinical details are summarised in Table 1.
Methods
Clinical, imaging and electroencephalographic data of the cases included in the study.
Case ID
Gender/age at study (yr)
1
F/11
2
Age at sz onset (yr)
Epilepsy syndrome
Etiology
Seizure semiology
MRI brain scan findings
EEG, ictal onset
EEG, interictal
Epilepsy surgery/surgical outcome
7
Focal, L paracentral
DNET
L parasagittal precentral gyrus DNET
R central, and R frontal
Mid to R central IED
Yes, lesionectomy/sz free
M/38
19
Focal, R hemispheric
Post-traumatic
Aura, somatosensory R leg → Tonic, R foot, or clonic, R arm → GTC Aura, unspecific → Tonic, L → Clonic, L
L occipital
L occipital IED
No
3
M/12
10
Focal, L parietal
FCD
Aura, unspecific → Automotor
R temporoparietal and frontal encephalomalacia; R frontal, parietooccipital gliosis L paracentral lobule FCD
R mid-parietal
L temporal IED
Yes, lesionectomy/sz free
4
M/69
69
Focal, R hemispheric
Unknown
Clonic status, L leg
Non-lesional
L mid-central
R mid-central IED
No
Lateralization of ictal patterns in extra-temporal epilepsies
Table 1
Abbreviations: DNET, dysembryoplastic neuroepithelial tumor; EEG, electroencephalography; FCD, focal cortical dysplasia; GTC, generalized tonic-clonic; IED, interictal epileptiform discharges; L, left; R, right; sz, seizure; yr, years.
149
150
C.B. Catarino et al.
Figure 1 Case 1. MRI brain scan: (A) sagittal T1 and axial T2, show a parasagittal, left frontal, dysembryoplastic neuroepithelial tumor. (B) Three-dimensional reconstructed MRI images show the location of the subdural electrodes (white circles) in relation to the cortex relief. Seizures start in the mesial frontal region (black circle). The area of maximum interictal spiking (dotted black line) during the invasive EEG-video monitoring is also shown on the projection of the subdural electrodes. (C) Non-invasive ictal EEG (Cz referential montage) shows a right centro-parietal seizure pattern at seizure onset (open arrow), which 12 s later evolves into a left frontocentral seizure pattern (solid arrow).
Lateralization of ictal patterns in extra-temporal epilepsies
Patient 1 This 11-year-old right-handed girl had left focal epilepsy since the age of 7, due to a dysembryoplastic neuroepithelial tumor (DNET) in the left parasagittal region, also involving parts of the left precentral and superior frontal gyri (Fig. 1A). She had somatosensory auras in her right leg, on average five per week, which could evolve into tonic seizures of the right foot, or clonic seizures of the right arm, and occasional secondary generalization. Invasive EEG-video monitoring with subdural electrodes covering the left lateral and mesial central cortex revealed that the ictal onset patterns and most of the interictal spikes were localized on the left hemisphere, inferior and mesial to the lesion (Fig. 1B), while the non-invasive EEG-video monitoring had shown midline-to-right central interictal epileptiform discharges, and ictal recordings had right centro-parietal seizure pattern at onset (Fig. 1C). Guided by intraoperative electrical stimulation of the cortex for localization of the primary motor leg area, the resection included the lesion and the adjacent area of seizure onset and maximum interictal spiking. Postoperatively the patient has remained seizure free, with a follow-up of over 5 years.
Patient 2 This 38-year-old right-handed man had post-traumatic focal epilepsy since the age of 19, following a road traffic accident with severe head trauma, which also resulted in mild left spastic hemiparesis, left hemihypoesthesia and left homonymous hemianopia. He had auras of an unspecific sensation, tonic seizures of the left body, and sometimes clonic seizures of the left body, occurring on average once every 3 months. His brain MRI revealed post-traumatic right frontotemporo-parietal encephalomalacia and right frontal and parieto-occipital gliosis (Fig. 2A). Dipole source analysis of the seizure EEG onset allowed correct localization of the generator source to the right, ipsilateral to the lesion seen on brain MRI (Fig. 2B and C), even though the visual analysis of the seizure EEG onset showed rhythmical sharp waves and delta waves with left occipital maximum. Non-invasive EEG-video monitoring showed left-sided parieto-occipital spikes; one habitual left tonic seizure evolving into asymmetrical bilateral tonic and then left clonic seizure was recorded, with preserved consciousness during the seizure. Postictally, left Todd’s paralysis was documented. Ictal EEG showed right occipital repetitive spikes, which preceded clinical onset by 180 s; five seconds before clinical onset, there was a left occipital seizure pattern (Fig. 2D and E).
Patient 3 This 12-year-old right-handed boy had refractory focal epilepsy since the age of ten, due to left mesioparietal focal cortical dysplasia (FCD) (Fig. 3A). He had daily unspecific auras, and automotor seizures. Non-invasive EEG-video monitoring recorded sixteen habitual seizures, with no recognizable lateralizing semiological signs. The ictal onset EEG pattern was localized to the vertex and right parietal region, with rhythmical sharp waves at 1 Hz on the vertex shifting
151 to the right (Fig. 3D). A lesionectomy of the left FCD was performed, and the patient has remained seizure free since.
Patient 4 A 69-year-old man presented with epilepsia partialis continua involving the left leg for six weeks. Scalp EEG showed polyspikes paradoxically distributed over the left parasagittal region, consistently associated with clonic jerks of the left calf; he also had right parasagittal spikes, which were clinically silent (Fig. 4). His brain MRI showed diffuse cerebral atrophy and no evidence of a focal lesion. The etiology of the status remained unknown.
Discussion The evaluation of patients considered for resective epilepsy surgery relies on several tests to define the epileptogenic zone, which is the area of the cerebral cortex that is indispensable for generating the epileptic seizures (Lüders and Awad, 1992), and whose resection leads to a seizure-free outcome. Electroencephalographic studies, especially ictal EEG recordings, are frequently essential in correctly determining the location of the epileptogenic zone, but always need to be interpreted in the context of the other presurgical investigations (Miller and Cole, 2011). We report here proven paradoxical lateralization of ictal EEG patterns in four patients with extra-temporal epilepsies, and three with mesial parasagittal epileptic foci. The underlying mechanism explaining the paradoxical lateralization of the scalp ictal EEG patterns is postulated to be similar to the one described for EP: it can be explained by the spatial orientation of an EEG dipole close to the midline, with a cortical outside negative source situated on the mesial surface of the cerebral hemisphere, with an orientation such that the negative electrical activity is projected obliquely and the highest negative amplitude is recorded on the contralateral side (Gloor, 1985) (Fig. 5A), whereas usually the greatest amplitude is recorded on the electrodes ipsilateral to the lesion, when the generators are perpendicular to the scalp (Fig. 5B). The phenomenon of paradoxical lateralization of EEG seizure pattern, when the ictal EEG pattern on the scalp appears contralateral to the actual seizure onset zone, has been seldom reported in the literature, mostly in patients with parasagittal cortical tumors (Tukel and Jasper, 1952; Penfield and Jasper, 1954), and epilepsia partialis continua of one lower extremity (Adelman et al., 1982; Oishi et al., 2002). Paradoxical lateralization of scalp EEG findings should be considered when EEG lateralization is not concordant with other clinical and imaging data, particularly in neocortical epilepsy with parasagittal epileptogenic foci. It should be distinguished from the ‘‘false EEG lateralization’’ reported in mesial temporal lobe epilepsy, where the mechanism is thought to be rapid ictal activity spread to the contralateral side (Castro et al., 2008; Engel and Crandall, 1983; Mintzer et al., 2004; Sammaritano et al., 1987). Rapid spread to the contralateral hemisphere could theoretically be another explanation for paradoxical lateralization of ictal EEG. However, this would be expected to be accompanied by some change in seizure semiology, with
152
C.B. Catarino et al.
Figure 2 Case 2. (A) MRI brain scan findings. Axial FLAIR, post-traumatic right fronto-temporo-parietal encephalomalacia and right fronto-parieto-occipital gliosis. Source dipole analysis of the EEG seizure pattern at onset superimposed on the brain MRI of the patient (B), and superimposed on a standard brain MRI (C), showing location of the source dipole on the right mesial posterior border of the lesion. (D) Ictal EEG recording, longitudinal bipolar montage, with left parieto-occipital seizure pattern preceding the clinical onset by 180 s, contralateral to the right hemispheric lesion, and therefore paradoxically lateralized (open arrow). Five seconds before clinical onset, there was a right occipital seizure pattern (solid arrow). Dipole source analysis at EEG onset correctly lateralized the seizure pattern to the right (B and C).
lateralizing signs supporting secondary seizure propagation, which was not the case in our patients. We have presented evidence proving paradoxical lateralization of the EEG seizure pattern in four patients with focal neocortical epilepsies and mesial parasagittal foci, one non-lesional. The epileptogenic zone of these patients was shown to be located in or close to the mesial surface of
the cerebral hemispheres. This was confirmed by seizure freedom after epilepsy surgery in two patients, with postoperative follow-up of more than five years. The epileptogenic zone was determined in two patients by the combined findings of seizure semiology, structural lesion on brain MRI and interictal EEG findings. For patient 2, EEG-video monitoring documented his habitual seizures, with left clonic activity,
Lateralization of ictal patterns in extra-temporal epilepsies
153
Figure 3 Case 3. MRI brain scan: (A) sagittal T1 and axial T1, show the left mesial central cortical malformation. Ictal EEG shows a right centro-parietal seizure pattern, contralateral to the left central lesion at clinical seizure onset (B), characterized by bilateral tonic posturing (open arrow). Twenty-two seconds later, the seizure pattern was located in the left parietal region (P3, Cz) (C, solid arrow); at that time, the seizure clinically evolved into a right clonic seizure.
and left-sided Todd’s paralysis. These ictal signs are of high lateralizing value, as semiological studies showed high sensitivity of clonic activity and Todd’s paralysis in correctly lateralizing the epileptogenic zone, 92% for the former (Janszky et al., 2001), and 93% for the latter (Kellinghaus and Kotagal, 2004). This patient’s post-traumatic and postsurgical right skull defect was expected to cause increased amplitude of EEG activity recorded over the right hemisphere, the so-called breach effect. This would be expected
to enhance right lateralization of the electrical brain activity, and so cannot explain the paradoxical lateralization to the left. Patient 4 has two types of epileptiform discharges. The occurrence of clinically silent spikes over the right hemisphere indicates a right hemispheric generator (Fig. 4). However, all polyspikes associated with left leg jerking were paradoxically lateralized to the left because the main generator of these was located in the mesial aspect of the right
154
C.B. Catarino et al.
Figure 4 Case 4. (A) Ictal EEG, longitudinal bipolar montage, shows left parasagittal polyspikes (solid arrow, and vertical line), synchronous with the EMG artifact, i.e. associated with the left calf jerking. In contrast, isolated spikes (open arrows), with a right central maximum (C4) were not associated with the jerks. EMG L. M. Tib. Ant., EMG left anterior tibialis muscle. (B) Isopotential field map, with electrodes CZ and C4 showing 100% of the amplitude of the potential and the subsequent rings, potential field drops of 10%, demonstrates the left parasagittal polyspikes, and also the isolated spikes with a right central maximum (C4), which had no clinical correlate. Source: (A) Reprinted with permission, from Lüders HO and Noachtar S (1995) Atlas und Klassifikation der Elektroenzephalographie, Ciba Geigy Verlag, Wehr. © Reprinted by permission of Novartis Pharma Verlag, Nürnberg.
Figure 5 Schematic illustration of the proposed mechanism for paradoxical lateralization. (A) A cortical source situated on the mesial surface of a cerebral hemisphere (outside negative) is oriented with the negative electrical activity projecting obliquely; the highest negative amplitude is recorded on the contralateral side. (B) A source located on the cerebral convexity will be recorded ipsilaterally, given the orientation of the electrical activity vector.
Lateralization of ictal patterns in extra-temporal epilepsies
155
central cortex. This is the only location explaining both the left leg semiology and the paradoxically left EEG localization. Dipole source analysis was found to be useful for correctly localizing the seizure onset zone, and can be used as a complementary non-invasive method, helping to overcome some of the limitations of visual EEG analysis. Identifying and understanding possible paradoxical lateralization of EEG is important and becomes particularly relevant when evaluating patients with medically refractory epilepsy for epilepsy surgery, not to reach erroneous conclusions about the odds for seizure freedom after surgery. Two of our four patients have been rendered seizure free after epilepsy surgery, illustrating that patients with paradoxical lateralization of the ictal EEG may have excellent surgical outcome.
Janszky, J., Fogarasi, A., Jokeit, H., Ebner, A., 2001. Lateralizing value of unilateral motor and somatosensory manifestations in frontal lobe seizures. Epilepsy Res. 43, 125—133. Kellinghaus, C., Kotagal, P., 2004. Lateralizing value of Todd’s palsy in patients with epilepsy. Neurology 62, 289—291. Lesser, R.P., Lueders, H., Dinner, D.S., Hahn, J., Morris, H., Wyllie, E., Resor, S., 1987. The source of ‘paradoxical lateralization’ of cortical evoked potentials to posterior tibial nerve stimulation. Neurology 37, 82—88. Lüders, H.O., Awad, I.A., 1992. Conceptual considerations. In: Lüders, H.O. (Ed.), Epilepsy Surgery. Raven Press, New York, pp. 51—62. Miller, J.W., Cole, A.J., 2011. Is it necessary to define the ictal onset zone with EEG prior to performing epilepsy surgery? Epilepsy Behav. 20, 178—181. Mintzer, S., Cendes, F., Soss, J., Andermann, F., Engel Jr., J., Dubeau, F., Olivier, A., Fried, I., 2004. Unilateral hippocampal sclerosis with contralateral temporal scalp ictal onset. Epilepsia 45, 792—802. Oishi, A., Tobimatsu, S., Ochi, H., Ohyagi, Y., Kubota, T., Taniwaki, T., Yamamoto, T., Furuya, H., Kira, J., 2002. Paradoxical lateralization of parasagittal spikes revealed by back averaging of EEG and MEG in a case with epilepsia partialis continua. J. Neurol. Sci. 193, 151—155. Osman, A., Muller, K.M., Syre, P., Russ, B., 2005. Paradoxical lateralization of brain potentials during imagined foot movements. Brain Res. Cogn. Brain Res. 24, 727—731. Penfield, W., Jasper, H., 1954. Epilepsy and Functional Anatomy of the Human Brain. Little, Brown, Boston. Sammaritano, M., de Lotbinière, A., Andermann, F., Olivier, A., Gloor, P., Quesney, L.F., 1987. False lateralization by surface EEG of seizure onset in patients with temporal lobe epilepsy and gross focal cerebral lesions. Ann. Neurol. 21, 361—369. Scherg, M., Ebersole, J.S., 1994. Brain source imaging of focal and multifocal epileptiform EEG activity. Neurophysiol. Clin. 24, 51—60. Tukel, K., Jasper, H., 1952. The electroencephalogram in parasagittal lesions. Electroencephalogr. Clin. Neurophysiol. 4, 481—494. Vollmar, C., Noachtar, S., Winkler, P.A., 2008. Multimodal image processing in pre-surgical planning. In: Lueders, H.O. (Ed.), Textbook of Epilepsy Surgery. Informa Health Care, London, pp. 771—778.
References Adelman, S., Lueders, H., Dinner, D.S., Lesser, R.P., 1982. Paradoxical lateralization of parasagittal sharp waves in a patient with epilepsia partialis continua. Epilepsia 23, 291—295. Barett, G., Blumhardt, L., Halliday, A.M., Halliday, E., Kriss, A., 1976. A paradox in the lateralisation of the visual evoked response. Nature 261, 253—254. Castro, L.H., Serpa, M.H., Valerio, R.M., Jorge, C.L., Ono, C.R., Arantes, P.R., Rosemberg, S., Wen, H.T., 2008. Good surgical outcome in discordant ictal EEG-MRI unilateral mesial temporal sclerosis patients. Epilepsia 49, 1324—1332. Cruse, R., Klem, G., Lesser, R.P., Lueders, H., 1982. Paradoxical lateralization of cortical potentials evoked by stimulation of posterior tibial nerve. Arch. Neurol. 39, 222—225. Engel Jr., J., Crandall, P.H., 1983. Falsely localizing ictal onsets with depth EEG telemetry during anticonvulsant withdrawal. Epilepsia 24, 344—355. Garzon, E., Gupta, A., Sakamoto, A.C., Lüders, H., 2009. Paradoxical ictal EEG lateralization in children with unilateral encephaloclastic lesions. Epileptic. Disord. 11, 215—221. Gloor, P., 1985. Neural generators and the problem of localization in electroencephalography: application of the volume conduction theory to electroencephalography. J. Clin. Neurophysiol. 2, 327—354.
