Current address of Dr. Marusic: Department of Neurology, University Hospital Motol, 2nd Medical School, 150 06 V Uvalu 84, Prague, Czech Republic.
Focal Cortical Dysplasias in Eloquent Cortex: Functional Characteristics and Correlation with MRI and Histopathologic Changes
Article first published online: 2 APR 2002
Volume 43, Issue 1, pages 27–32, January 2002
How to Cite
Marusic, P., Najm, Imad M., Ying, Z., Prayson, R., Rona, S., Nair, D., Hadar, E., Kotagal, P., Bej, Mark D., Wyllie, E., Bingaman, W. and Lüders, H. (2002), Focal Cortical Dysplasias in Eloquent Cortex: Functional Characteristics and Correlation with MRI and Histopathologic Changes. Epilepsia, 43: 27–32. doi: 10.1046/j.1528-1157.2002.00801.x
- Issue published online: 2 APR 2002
- Article first published online: 2 APR 2002
- Revision accepted November 28, 2001.
- Cortical dysplasia;
- Balloon cells
Summary: Purpose: Focal cortical dysplasia (CD) is increasingly recognized as a common pathologic substrate of medically intractable epilepsy. As these lesions are often localized in the frontal lobe (therefore in potentially eloquent cortex), an understanding of the functional status of the involved region(s) and of its anatomic and pathologic correlates is of prime importance. The purpose of this study is to assess the function of focal CD in relation to magnetic resonance imaging (MRI) and histopathologic features.
Methods: Eight patients operated on for medically intractable epilepsy with histologically proven focal CD involving putative eloquent cortex in the frontal lobe (perirolandic and Broca's areas) were included in the study. Functional regions (motor and language) and epileptogenic areas were assessed by extraoperative electrocorticographic recording and electrical cortical mapping. Cortical functions were correlated with the extent of epileptogenicity on electrocorticographic recordings, MRI features, and histologic characteristics.
Results: Language or motor areas were colocalized with epileptogenic regions (n = 6 of 8, 75%), but were not mapped in regions of increased signal on fluid-attenuated inversion recovery (FLAIR) MRI (when they were identified) on preoperative MRI (n = 5 of 5, 100%). Histologically, balloon cells were almost exclusively found in nonfunctional regions with FLAIR MRI abnormalities. When resected, regions of motor cortex were characterized by cortical dyslamination, columnar disorganization, and dysmorphic neurons, but were devoid of balloon cells.
Conclusions: We found an absence of language or motor functions in perirolandic and Broca's areas that showed decreased epileptogenicity, histopathological evidence of CD with balloon cells and FLAIR MRI signal increase. Language and motor functions were present in epileptogenic and dysplastic areas with no balloon cells and no FLAIR signal abnormalities. These findings have implications on options for epilepsy surgery in patients with CD.
Focal cortical dysplasias (CDs) are increasingly recognized as pathologic substrates in patients with medically intractable epilepsy who undergo surgical resection. (1–3). These patients tend to have worse postoperative seizure outcome as compared with those with other types of focal lesions (4,5). Intrinsic epileptogenicity and preservation of eloquent functions were previously described in pathologically proven CD (6–9). The advent of magnetic resonance imaging (MRI) techniques enabled the recognition of small focal CD. Several reports described MRI characteristics of CD that may include increased signal on T2-weighted and fluid-attenuated inversion recovery (FLAIR) sequences, cortical thickening, abnormal gyral and sulcal anatomy, and/or indistinct gray–white matter junction (10–12). Despite the advances in MRI techniques, some patients with pathologically proven CDs have normal imaging studies. In other patients, MRI fails to delineate the full extent of the dysplastic lesion (s) (13). FLAIR sequence has been reported to be a sensitive MRI technique for the identification and delineation of abnormal cortical areas (14–16). Recent studies showed a correlation between increased signal intensity (ISI) on FLAIR sequences and the presence of balloon cells (BCs) associated with severe cortical disorganization in patients who underwent surgical cortical resection for the treatment of intractable focal epilepsy (12,17–19).
The purpose of this study is to assess the functional status of focal CD (as identified by direct cortical electrical stimulation) and its relation to imaging, in situ electrocorticographic (ECoG), and histopathologic characteristics in patients who underwent focal neocortical resection for the treatment of medically intractable epilepsy.
MATERIALS AND METHODS
Eight patients who underwent cortical resection between August 1997 and February 2000 were included in the study. All patients with medically intractable focal epilepsy arising from the frontal lobe (including the central area that was defined as containing both pre- and postcentral gyri) with the lesion located adjacent or inside the anatomic location of presumed eloquent cortex. Pathologic examination by a neuropathologist (R.P.) confirmed the presence of CD in all cases. Patients with other pathologic diagnoses were excluded from the study.
All patients underwent prolonged noninvasive scalp and invasive subdural cortical electrode video-EEG monitoring and MRI studies (including FLAIR sequence). Subdural grid electrodes covered the region of suspected seizure onset, anatomically defined eloquent regions, and areas of MRI identified abnormality. All subdural electrode placements and surgical resections were performed by the same neurosurgeon (W.B.).
Functional and epileptogenic areas (defined by ictal-onset zone) were assessed in all patients by extraoperative electrical cortical mapping and ECoG recording, respectively. The demographic, clinical, ECoG, functional, and imaging characteristics of patients included in the study are shown in Table 1.
|Age/ gender||Onset (yr)||Handed- ness||Speech||MRI abnormality||FLAIR signal||Epileptogenic functional area colocalization||Spatial correlation||Resection area(s)|
|Epileptogenic /increased signal intensity||Functional /increased signal intensity||Increased signal intensity||Epileptogenic area||Functional area|
|1||28/F||12||Right||Left||Left fronto centrotemporal||−||+||NA||NA||NA||±||+|
|2||14/M||2||Left||Bilateral||Left fronto central||−||+||NA||NA||NA||+||+|
|4||14/M||1||Right||Not determined||Right fronto central||+||+||Adjacent||Adjacent||+||±||+|
|6||11/M||2||Right||Left||Left fronto central||+||+||Adjacent||Adjacent||+||+||+|
|7||38/M||5||Left||Left||Left fronto central||+||+||Adjacent||Adjacent||+||+||+|
|8||25/F||10||Right||Left||Right fronto- centrotemporal||−||+||NA||NA||NA||±||+|
The location and extent of the ictal onset zone were assessed during prolonged extraoperative recording. Seizures were recorded in all eight patients. For the purpose of this study, the epileptogenic area was defined as only including the ictal-onset zone. This zone was defined as electrodes that exhibited a maximal amplitude sustained rhythmic (on referential recordings) or paroxysmal fast (beta and faster) activities for a duration of 10 s, with or without associated behavioral seizure manifestations.
All patients had MRI studies preoperatively, 24–48 h after implantation of subdural electrodes and after cortical resection. Preoperative MRI studies consisted of sagittal and coronal volume acquisition T1-weighted images, axial or coronal T2-weighted images, and FLAIR images. Cortical structures and sulcal–gyral organization were studied, and the presence of asymmetric focal FLAIR increased signal intensity was noted. As described later, the location and extent of the FLAIR signal increase area was assessed. Postimplantation MRI study with subdural electrodes served for the anatomic correlation of electrodes localization in relation to underlying cerebral cortex and to the area of FLAIR signal increase.
Positive and negative motor phenomena as well as speech arrest and dermatomal sensory dysfunction were assessed during direct stimulation of each subdural electrode. In addition, evoked potential studies were done when needed to delineate the spatial relation of subdural grids in relation to the central sulcus. The methodology of these procedures has been described elsewhere (20,21). A stimulation map showing the relations between cortical function and anatomic structures covered by subdural electrodes was generated for each patient.
MRI scans were performed on a 1.5-T whole-body MR imager (Magnetom VISION, Siemens, Erlangen, Germany). T1-weighted images were acquired by using a magnetization-prepared rapid acquisition gradient echo (MP RAGE) sequence with coronal 2-mm slices without interslice gap, repetition time (TR) 11.4, and echo time (TE) 4.4 ms; flip angle, 10 degrees; field of view, 75 mm2; matrix size, 256 × 256; resulting in a voxel size of 1 × 1 × 2 mm. FLAIR images were acquired by using a fast-FLAIR (RARE) sequence with coronal (temporal lobe epilepsy) or axial (extratemporal lobe epilepsy) cuts; slice thickness, 6 mm with 2.4 mm interslice gap; TR, 8,000, and TE, 105.0 ms; flip angle, 180 degrees; field of view, 75 mm2; matrix size, 256 × 256.
Correlation of spatial relation of subdural electrodes to MRI
As shown in Fig. 1, a three-dimensional surface reconstruction of each brain was created to provide visual correlation between subdural electrode position, FLAIR signal, and cortical surface topology. T1-weighted MRI images were acquired and combined to form a single three-dimensional volume with voxel dimensions of 0.9 × 0.9 × 2.0 mm. Volumes were processed to remove background elements (22) and to segment the brain (23) automatically. Electrodes on the 2D MRI images also were semiautomatically segmented and saved with the segmented brain volume. The surface of the brain was reconstructed by using an in-house computer program that interactively renders the volume and surface locations of the brain and electrodes, respectively. To achieve real-time response, a hardware-accelerated, back-to-front volume-rendering algorithm on a Silicon Graphics computer (Mountain View, CA, U.S.A.) was adopted (24). To display the electrodes, a 2-mm-diameter sphere (identical in size to the recording electrode diameter) was surface-rendered at the centroid of each segmented electrode and overlaid on the volume-rendered image. This hybrid display of brain and electrodes allows the accurate identification of electrode position for surgical resection of epileptogenic tissue.
Multimodal image analyses were performed individually for each patient on a Unix workstation (Onyx 2, Mountain View, CA, U.S.A.) with the aid of an in-house developed (EL) interactive post-processing software based on the AIR program. Analyses included the following steps: (a) manual segmentation and back-to-front hardware-accelerated volume-rendered 3D-reconstruction of preoperative T1-MP RAGE, and FLAIR images; (b) fusion of the two volumes, using a best-fit algorithm; (c) visual analysis of 2D images to detect signal abnormalities, using the “native” images as well as a threshold-based color-coded display method for FLAIR images; the low threshold used for display purposes was set at 1.5 standard deviations above the mean signal intensity count for the total volume; and (d) identification of subdural electrode positions and projection on 3D volume to assess the relations between the anatomic/structural abnormalities and electrocorticographic data.
The position of subdural electrodes in relation to the central sulcus was confirmed by the combination of electrical cortical mapping and evoked potential studies [Trigeminal (lip) and median nerve somatosensory evoked potentials (SSEPs)].
Based on prolonged extraoperative monitoring data analyses, maps of ECoG activities and their relations with both electrode positions and FLAIR signal abnormalities were generated. Before surgical resection, areas underlying electrodes of nonepileptogenic and epileptogenic areas were identified and labeled by using brilliant green solution. The resected tissues were divided into
Group I, nonepileptogenic cortex, which included areas of no ictal onset. This group was subdivided into two subgroups according to the MRI FLAIR data; and
Group II, Epileptogenic cortex, which included area(s) of ictal onset.
As the goal of the resection was to achieve optimal seizure control with minimal functional deficit, extraoperatively (and at times intraoperatively) identified functional areas were spared unless there was a significant overlap with epileptogenic areas. Functional cortex resection (if performed) was restricted to the face area in three patients and extended to the precentral region representing the hand or shoulder motor areas in another three patients (2, 4, and 7).
As previously reported (25), the resected blocs varied in size (depending on the number of involved subdural electrodes). The resected tissue was taken from a midline point between the two electrodes recording distinct electrical activities (∼0.5 cm from the center of the adjacent electrode). Representative parts of the specimens were submitted to the Department of Pathology of the Cleveland Clinic Foundation for independent pathologic interpretation, and other adjacent cubes were processed for further histologic characterizations.
Cubes of cortical specimens were labeled and immersion-fixed for 36–48 h in 4% paraformaldehyde at 4°C, and then cryoprotected with 20% buffered sucrose for processing as described later. As previously described, at least three sections (30 μm) at alternating 1-mm intervals from each cube were collected for cresyl violet (CV) Nissl staining (25,26). Various neocortical regions were analyzed at ×5 and ×40. The presence of focal CD was confirmed on CV staining in at least one of the resected samples from each patient included in the study. The following histopathologic features were characterized: (a) cortical architecture: laminar organization, columnar arrangement, the persistence of neurons in the molecular layer (layer I), and the presence of neurons in the subcortical white matter; and (b) cellular morphology: neuronal orientation, CV staining intensity, and for the presence of large cells with central nuclei and well-defined cytoplasmic membranes (giant neurons or meganeurons) and “balloon” cells (strikingly large opalescent cytoplasm with eccentric nuclei).
The diagnosis of focal CD was made on identification of disorganization in the cortical architecture (dyslamination and columnar disorganization) and the presence of dysmorphic neurons, and giant neurons with central nuclei with or without the identification of BCs.
Topographic and pathologic correlations
As shown in Figs. 2 and 3, topographic maps of the electrical stimulation as correlated with epileptogenic regions and areas of signal increase on FLAIR (when present) were generated for each patient. Pathologic correlations with the topographic maps were generated based on the following criteria: (a) normal cortex, (b) dysplastic cortex without the identification of BCs (CD+/BC–), and (c) dysplastic cortex with BCs (CD+/BC+).
Functional and electroencephalographic cortical mapping
Eloquent cortex (as defined earlier) was identified in all patients included in the study. The anterior speech area (Broca's) was identified in three (37%) of eight patients by using extraoperative electrical cortical stimulation. Hand, face, or tongue motor areas were identified in seven (88%) of eight patients. Somatosensory responses were elicited in all patients. ECoG seizures were recorded, and areas of ictal onset were identified in all patients studied. Epileptogenic areas were restricted to one region (adjacent electrodes) in four (50%) of eight patients and were multifocal (two or more areas separated by more than one electrode) in four (50%) of eight patients.
MRI signal characteristics
As shown in Table 1, focal signal increase on FLAIR sequences in the mapped cortical areas was identified in five (63%) of eight and was absent in three (37%) of eight patients.
All patients had evidence of focal CD (as defined earlier) in at least one resected cortical bloc. BCs were identified in five of eight of the patients studied. BCs were present mainly in the deep layers of the cortex and in the subcortical white matter. The presence of BCs was always associated with coexistent histopathologic evidence of severe dyslamination and neuronal dysmorphism.
Correlation between functional cortex and MRI, EcoG, and histopathologic characteristics
As shown in Tables 1 and 2, cortical areas that showed evidence of functional response on direct electrical stimulation were the sites of ictal onset in six (75%) of eight patients. In three of those six patients, the areas of colocalization (overlap) between functional responses and ictal-onset zones extended to more than two electrodes. Moreover, resected functional areas did not show evidence of any obvious focal ISI on FLAIR images and were pathologically characterized by the presence of CD without BCs. BC-containing cortical areas showed no evidence of ictal activity. Moreover, electrical stimulation of CD+/BC+ areas did not result in any clinical response. Conversely, all the BC-containing areas showed evidence of focal FLAIR signal increase. The ictal-onset zones were localized to the electrodes immediately adjacent to the ISI regions in all five patients with focal FLAIR signal abnormalities. Two patients had additional ictal-onset foci distant to the areas of ISI.
As shown in Table 1, the focal resections included eloquent cortex in six of eight patients. In five of those patients, the resection was limited to the motor face cortex. These patients showed mild facial droop postoperatively that recovered almost completely 3 to 6 months later. In patient 7, the resection extended to the hand and trunk primary motor areas with the subsequent development of right hemiparesis (Arm > leg) that slightly improved after extensive inpatient and outpatient physical therapy. The deficits occurred only after resections of the areas that showed functional activity preoperatively. There were no unexpected deficits.
This study directly assessed the function of pathologically confirmed CD. Our results suggest that the functional status of CD is directly correlated to their histologic characteristics. Dysplastic lesions that are histologically characterized by the presence of BCs (Taylor-type cells) did not show in situ evidence of electrically inducible function despite their location in otherwise anatomically functional areas such as the precentral motor cortex.
Our results are in agreement with previous reports that showed the persistence of eloquent function in CD that are devoid of balloon cells (8,9). In addition, our current data suggest that regions with CD are nonfunctional if they contain BCs.
Duchowny et al. (27) reported that language cortex overlapped or bordered the epileptogenic region in 12 of their patients with developmental tumors or pathologically confirmed CD. Although CD overlapped with epileptogenicity and function in some of their patients, the function and epileptogenicity were adjacent to the CD in others. Moreover, the cellular characteristics (i.e., the presence or absence of BCs) were not specified or correlated with the location of epileptogenicity and function.
BC-negative CDs showed evidence of epileptogenicity (ictal onset), as evidenced by the recording of ictal patterns from the same areas. These findings are in agreement with previous studies that reported on the intrinsic epileptogenicity of CD (6,7). Ictal-onset zones were not mapped to the BC-positive CD areas.
Interestingly, the epileptogenicity was mainly recorded from the immediately adjacent cortical areas in all five cases in which BCs were found. Similar ECoG patterns were reported in patients with low-grade glial tumors [e.g., dysembryoplastic neuroepithelial tumor (DNET) and ganglioglioma], whereas dysplastic and epileptic cortical areas were found in the immediate surroundings of these lesions (28–31). As a full resection of the epileptogenic areas was not possible in all cases studied, it is still possible that epileptogenic areas that may contain BCs were not resected.
The absence of function (and epileptogenicity) in dysplastic cortex containing BCs may be due to severe disruption of the neuronal circuits in these areas, therefore resulting in the lack of neuronal–neuronal communication in the same areas. The functional status and the role that BCs may play in the lack of expression of function and epileptogenicity remain unknown.
In summary, our results suggest distinct functional, ECoG, and imaging properties of FCDs that are correlated mainly with the presence or absence of BCs. Moreover, these results show that functional cortex may be displaced within the same hemisphere and therefore may have direct implications on the options for epilepsy surgery.
Acknowledgment: This study was supported by 1K08 NS02046 grant to IN from the National Institutes of Health.
- 7Dubeau F, Palmini A, Fish D, et al. The significance of electrocorticographic findings in focal cortical dysplasia: a review of their clinical, electrophysiological and neurochemical characteristics. Electrocorticography: current trends and future perspectives: EEG. 1998;48:77–96.
- 11Abnormalities of gyration, heterotopias, tuberous sclerosis, focal cortical dysplasia, microdysgenesis, dysembryoplastic neuroepithelial tumour and dysgenesis of the archicortex in epilepsy: clinical, EEG and neuroimaging features in 100 adult patients. Brain 1995;118:629–60., , , et al.
- 24Cabral B, Cam N, Foran J. Accelerated volume rendering and tomographic reconstruction using texture mapping hardware. In: Kaufman A, Krueger W, eds. Proceedings of the 1994 Symposium on Volume Visualization 1994;91–98.