The middle cingulate gyrus extends slightly caudally posterior to the vertical commissure posterior (VCP) line.10 The right T1-weighted magnetic resonance image illustrates the same landmarks on magnetic resonance imaging; the dark shade highlights a lesion in the posterior cingulate. AC indicates anterior commissure; aMCC, anterior middle cingulate cortex; pACC, posterior anterior cingulate cortex; PC, posterior commissure; PCC, posterior cingulate cortex; pMCC, posterior middle cingulate cortex; VCA, vertical commissure anterior.11
Representation of the magnetic resonance image lesions and corresponding distribution of interictal and ictal electroencephalographic findings on scalp. The interictal maps show electrodes with involvement greater than 50% of the maximum amplitude of the epileptiform discharges. The darker color, when present, marks the subset of electrodes that consistently exhibit maximal amplitude of the interictal activity. The highlighted ictal electrodes represent the ones with greater than 50% of maximum amplitude. The darker red color marks regions/electrodes with relatively localized and faster ictal rhythms (the most prominent ictal rhythm is in the left upper corner of the ictal map). Interictal maps have been left blank in patients who had no identifiable interictal discharges on scalp. Ictal maps have been left blank in patients whose scalp ictal rhythms were obscured. PF indicates paroxysmal fast activity.
Summary of clinical data, reconstruction of magnetic resonance image lesions, surgical resection, seizure outcomes, duration of follow-up, and pathology data. EEG indicates electroencephalography; F, female; L, left; M, male; R, right. > indicates followed by. Dialeptic = complex partial seizures without automatisms.
Alkawadri R, So NK, Van Ness PC, Alexopoulos AV. Cingulate EpilepsyReport of 3 Electroclinical Subtypes With Surgical Outcomes. JAMA Neurol. 2013;70(8):995-1002. doi:10.1001/jamaneurol.2013.2940
The literature on cingulate gyrus epilepsy in the magnetic resonance imaging era is limited to
case reports and small case series. To our knowledge, this is the largest study of surgically
confirmed epilepsy arising from the anterior or posterior cingulate region.
To characterize the clinical and electrophysiological findings of epilepsies arising from the
anterior and posterior cingulate gyrus.
Design, Setting, and Participants
We studied consecutive cingulate gyrus epilepsy cases identified retrospectively from the
Cleveland Clinic and University of Texas Southwestern Medical Center epilepsy databases from 1992 to
2009. Participants included 14 consecutive cases of cingulate gyrus epilepsies confirmed by
restricted magnetic resonance image lesions and seizure freedom or marked improvement following
Main Outcomes and Measures
The main outcome measure was improvement in seizure frequency following surgery. The clinical,
video electroencephalography, neuroimaging, pathology, and surgical outcome data were reviewed.
All 14 patients had cingulate epilepsy confirmed by restricted magnetic resonance image lesions
and seizure freedom or marked improvement following lesionectomy. They were divided into 3 groups
based on anatomical location of the lesion and corresponding seizure semiology. In the posterior
cingulate group, all 4 patients had electroclinical findings suggestive of temporal origin of the
epilepsy. The anterior cingulate cases were divided into a typical (Bancaud) group (6 cases with
hypermotor seizures and infrequent generalization with the presence of fear, laughter, or severe
interictal personality changes) and an atypical group (4 cases presenting with simple motor seizures
and a tendency for more frequent generalization and less-favorable long-term surgical outcome). All
atypical cases were associated with an underlying infiltrative astrocytoma.
Conclusions and Relevance
Posterior cingulate gyrus epilepsy may present with electroclinical findings that are suggestive
of temporal lobe epilepsy and can be considered as another example of pseudotemporal epilepsies. The
electroclinical presentation and surgical outcome of lesional anterior cingulate epilepsy is
possibly influenced by the underlying pathology. This study highlights the difficulty in localizing
seizures arising from the cingulate gyrus in the absence of a magnetic resonance image lesion.
The term anterior cingulate epilepsy was first used by Andy and Chinn1 in 1957 to describe an experimental epilepsy model in cats. In 1970, Mazars2 published an article titled “Criteria for identifying cingulate epilepsies,” which predated magnetic resonance imaging (MRI) and at a time in which localization by intracerebral depth electrodes was problematic by today’s standards. In 1992, Bancaud and Talairach3 described a typical syndrome characterized by complex motor manifestations, intense fright, and incomplete loss of consciousness. In the MRI era, anterior cingulate epilepsy cases have been described in the form of case reports or short case series.4- 7 Little is known about posterior cingulate epilepsy. In 2008, Koubeissi et al8 analyzed the propagation of parietal cingulate seizures with secondary mesial temporal involvement. Other reports of posterior cingulate epilepsy did not validate the localization with favorable surgical outcomes.9
Identification of patients with cingulate epilepsies based on electroclinical features alone presents a major challenge in clinical epileptology. To this end, we report our experience in a series of patients with pharmacoresistant epilepsy who underwent targeted resections involving the cingulate gyrus between 1992 and 2009. The time span chosen corresponds to the interval during which all patients underwent MRI routinely.
Patients were identified from the Cleveland Clinic and University of Texas Southwestern Medical Center epilepsy patient databases spanning from 1992 to 2009. We included patients who underwent resections of MRI-defined lesions confined to the cingulate gyrus.
We included cases with good surgical outcomes (Engel class I) and cases with greater than 95% reduction in seizure frequency only if the patients experienced seizure freedom for at least 6 consecutive months following surgery. We excluded patients who were younger than 2 years of age at the time of video electroencephalography (EEG) evaluation and patients with less than 1 year of follow-up. After careful screening, 14 consecutive patients were included, with a follow-up period ranging from 1 to 11 years after the resection.
Visual analysis of MRIs was used to determine the location of the lesion within the cingulate gyrus. For the purposes of this study, the cingulate gyrus was divided into 2 portions: anterior and posterior using Talairach and Tournoux coordinates.10 The border between the anterior cingulate gyrus (area 24) and posterior cingulate gyrus (area 23) is now believed to be at the vertical commissure posterior line, as supported by recent histological studies (Figure 1).11 The lesions were identified on MRI planes and were projected to a schematic brain map on the sagittal plane, as illustrated in Figure 2. If the lesions were not clearly visualized in the sagittal planes, then MRIcro Analyze viewer12 was used to reconstruct sagittal fluid-attenuated inversion recovery/T2 images from coronal or axial ones. In case of overlap, invasive data were used to confirm the location, and, if not performed, the bulk of the lesion was used to determine the relationship to the vertical commissure posterior line.
Demographics data, video EEG recordings, radiological findings, pathology data, and surgical outcomes were reviewed. Seizures recorded in invasive evaluations were also reviewed. The clinical seizures were classified according to a semilogical seizure classification.13
The median age at presentation was 24.5 years (range, 4.0-51.0 years). Eight of 14 patients (57%) were male, and 12 (86%) had left cingulate lesions. The average duration of epilepsy before surgery was 4.2 years (range, 0.5-13.0 years). Patients were divided into 2 groups based on the anatomic location of the lesion: anterior (n = 10) and posterior (n = 4) cingulate groups. We observed 2 clusters of electroclinical features in the anterior cingulate cases; therefore, this group was further divided into a typical anterior cingulate epilepsy group (n = 6) with complex motor seizures similar to the description by Bancaud and Talairach3 and an atypical anterior cingulate group (n = 4) with more frequent simple motor seizures and complete loss of consciousness. The findings of each subgroup will be presented and discussed separately. Figure 3 provides a summary of clinical findings, localization of the MRI lesions, surgical procedures, and long-term outcomes of the 14 cases.
Six patients were identified, of whom 5 were male. The median age at time of epilepsy onset was 27.0 years (range, 2.5-51.0 years).
Auras were reported in 3 patients (50%). The most frequently reported aura present in all 3 was fear. One patient had 2 other concurrent auras: contralateral indistinct somatosensory aura and a subjective sensation of freezing with or without lightheadedness. Clinical manifestations were marked predominantly by complex motor signs. All patients had hypermotor seizures (eg, twisting around, banging head, bicycling, flailing, violent sweeping movements, and running); one patient had equally frequent contralateral simple motor seizures along with the hypermotor activity (patient 2).
Early loud vocalization was seen in 4 patients (67%) and nonmirthful laughter was seen in 2 patients (33%). Interestingly, 1 patient had frequent ictal urination without generalization. None of these 6 patients reported complete loss of consciousness during seizures, and a quick recovery was seen in all. Four patients (67%) had never experienced secondarily generalized seizures, 1 had his only secondarily generalized seizure during video EEG evaluation while not taking antiseizure medications, and 1 patient had a total of 5 secondarily generalized seizures within a 3-year period, whereas she was having frequent habitual partial seizures daily (7/d).
Interictally, definite epileptiform abnormalities were seen only in 2 patients (33%), implicating the frontocentral regions in both. These discharges were frequent in 1 patient (patient 2; many/min), whereas they were rare in the other patient (patient 5; 2-4/d).
Ictal patterns were mostly obscured by muscle artifact and, when discernible, pointed to the ipsilateral frontocentral region (Figure 2).
We observed that 4 of the patients in the anterior cingulate group exhibited clinical features that differed from those seen in the typical group, as previously described, such as frequent simple motor seizures along with frequent complete loss of consciousness and secondary generalization. Hypermotor seizures were infrequent or absent. The median age at time of epilepsy onset was 23.5 years (range, 1.0-45.0 years). Three patients were female.
Aura of subjective sensation of freezing was reported in 1 patient (25%). Unlike the typical group, recorded seizures in all 4 patients were classified as focal simple motor seizures. Automatisms were observed in 1 patient without the violent component seen in the typical group. Early loud vocalization was seen in 2 patients (50%). One patient had ictal urination in partial seizures without generalization. All patients reported complete loss of consciousness in some or all of their seizures. Unlike the typical group, all patients had experienced secondarily generalized seizures that were frequent in 3.
Definite interictal abnormalities were seen in 1 patient (patient 10), implicating the temporal regions bilaterally. Ictal patterns pointed to the ipsilateral frontocentral region in 2 patients (50%) and involved both sides in 1 patient. Ictal patterns were obscured in another patient (Figure 2).
Four patients were identified, of whom 2 were male. The median age at time of epilepsy onset was 14 years (range, 9-19 years).
Auras were reported in all 4 patients (100%). The most frequently reported aura present in 3 was a dyscognitive aura (déjà vu, jamais vu, or depersonalization); an abdominal aura was seen in 2, gustatory in 1, and subjective feeling of falling in 1. Two patients had multiple auras.
Motor phenomena were observed in 3 patients (75%), who presented with different seizure types including simple motor and automotor seizures. Dialeptic seizures (complex partial sezures without automatisms) were seen in 2 patients. Secondary generalization was seen in all patients and was frequent in 3.
Three patients had unequivocal focal slowing in the ipsilateral temporal region. Of those, 1 also had focal slowing involving the contralateral temporal chain. Bifrontal slowing was seen intermittently in patient 14. Interictal epileptiform discharges were seen in this group at high frequency and implicated the ipsilateral temporal (n = 2 maximum in the anterior temporal electrodes and 1 in the posterior temporal electrodes) and frontal lobe (n = 1). Ictal patterns were widespread (lateralized to the ipsilateral hemisphere in 2 or nonlocalizable in 1 patient) with faster rhythms seen in the ipsilateral temporal region in 2 patients and bifrontally in 1. One patient had ictal patterns that were confined to the ipsilateral temporal region (patient 11) (Figure 2).
Two patients with anterior cingulate (patients 1 and 8) and 2 with posterior cingulate lesions (patients 12 and 14) underwent invasive evaluations with subdural electrodes to confirm the epileptogenicity of the lesion. As expected, cingulate coverage with subdural electrodes was challenging and incomplete in all patients owing to inherent anatomical and vascular considerations. Owing to the limited coverage of the lesion, the area of earliest EEG change involved several electrodes in all patients—likely reflecting a spread exit pattern rather than the actual seizure-onset zone.
The duration of 4 randomly selected seizures from each patient was significantly shorter in the typical anterior cingulate cases (medians of 30.0, 150.0, and 96.8 seconds for typical anterior, atypical anterior, and posterior cingulate cases, respectively; Kruskal-Wallis P < .001; Wilcoxon-paired tests; P < .001). A trend for higher frequency of seizures was apparent among the typical anterior cingulate cases. Figure 3 provides a summary of the pathology and surgical outcomes. It should be noted that any resection outside the cingulate gyrus in nontumoral cases was merely technical, and it was not intended to cure the underlying condition; on the other hand, resection in cases of astrocytoma was often extended outside the cingulate owing to the nature of the pathology. Semiology and interictal 18fluorodeoxyglucose–positron emission tomography scan studies (when performed) were more successful in lateralizing the epileptic focus when compared with interictal and ictal EEG (the focus lateralized correctly by semiology in 8 of 14 patients and by positron emission tomography in 8 of 9 patients vs 3 of 14 and 6 of 13 by interictal and ictal EEG, respectively). Only clinical signs with lateralizing value greater than 80% were used in the analysis.14 Interestingly, mild ipsilateral temporal hypometabolism by positron emission tomography with 18fluorodeoxyglucose was seen in 3 of the posterior cingulate cases.
Here we presented our experience with a group of 14 patients with lesional cingulate epilepsy who were treated surgically over an 18-year period. Cingulate gyrus epilepsy is a rare disorder. The purpose of this study was to better characterize epilepsies arising from the cingulate gyrus. Therefore, we only included patients with lesions confined to the cingulate gyrus and favorable seizure outcome.
Importantly, all 4 of these posterior cingulate cases masqueraded as temporal lobe epilepsies clinically and/or electrographically. Among the 10 patients in the anterior cingulate group, most presented with frequent, mostly nocturnal and predominantly hypermotor/hyperkinetic, seizures with rare generalization with or without an aura of fear, loud vocalization, or nonmirthful laughter. These findings are consistent with Bancaud and Talairach’s series.3 We also identified a smaller group of anterior cingulate cases, which we labeled for the purposes of this report as atypical given the propensity for frequent generalization, prominence of simple motor rather than hypermotor manifestations, and less favorable surgical outcome. Notably, the pathology in all of the atypical cases was astrocytoma.
A total of 4 patients in our series underwent invasive evaluations with subdural electrodes. The decision to proceed with an invasive evaluation in the anterior cases was influenced by lack of localizing ictal rhythms and paucity of interictal discharges on scalp recordings. In the posterior cases, implantation was driven by the apparent discrepancy between the location of the lesion and the electroclinical presentation; the latter was suggestive of a temporal focus. As expected, cingulate coverage with subdural electrodes was incomplete in all 4 patients. The deeper parts of the interhemispheric cortex are not readily accessible, and any surgical approach is challenging given the abundance of large caliber vessels.
Electroencephalography findings were not helpful in terms of localization in most of the cases, underlining the challenges of identification of these patients based on scalp EEG. In general, the ability of surface EEG to detect interictal activity depends on the extent of the irritative zone, the proximity of the generator to the scalp, and the orientation of the dipole.15 Furthermore, EEG activities originating from a deep-seated focus may present with a misleadingly widespread appearance (patients 10, 11, 12, and 14) because of the distance that separates the epileptogenic zone from the scalp (Figure 2).16 Mislocalized interictal and ictal EEG findings likely reflect propagated activity resulting from the underlying structural connectivity of the cingulate gyrus. Lastly, patients 10 and 12 brought to mind the concept of secondary bilateral synchrony as a result of a unilateral parasagittal epileptic focus (Figure 2).17,18
The dichotomy of the electroclinical behavior between the anterior and posterior cingulate cases is not unexpected, accounting for the differences in cytoarchitecture and connectivity of the anterior vs the posterior cingulate. The anatomofunctional heterogeneity of the cingulate has been known for a long time and is well reflected in Brodmann’s cytoarchitectural map.19 These differences were further delineated by the work of Vogt et al.11,20 The heterogeneity of the cingulate cannot be overlooked. However, from a practical standpoint, it is reasonable to divide the cingulate into anterior and posterior portions, as illustrated by our study.
There is significant overlap between the typical anterior cingulate cases in our series with the syndrome that Bancaud and Talairach3 characterized in their work. This syndrome consists of brief and frequent, usually daily, seizures, with partial loss of awareness. Seizures in this series manifested by hypermotor activity with or without aura of fear, mirthless laughter, prominent vocalization, and rare secondary generalization. Less frequently encountered were ictal urination and aura of subjective sensation of freezing.
Our current knowledge of the function of the anterior cingulate supports the central role of this structure in generating the previously mentioned symptoms. Talairach and colleagues21 reported that stimulation of the anterior cingulate leads to motor responses that are “different from those elicited by stimulation of the motor area”—these observations are consistent with the findings by Bancaud et al22 of behavioral changes in relation to anterior cingulate stimulation. The symptomatogenic zone for nonmirthful laughter is not entirely defined. Laughter was induced by electrocortical stimulation of subdural contacts covering the left mesial superior frontal gyrus,23 as well as depth electrode contacts within the right anterior cingulate gyrus.24 It is noteworthy to mention that reports of gelastic seizures originating from the frontal lobe have been linked to a cingulate focus in most published cases.25 In addition, gelastic seizures have been reported in 6 of 16 cases of anterior cingulate epilepsy based on our review of the literature.26,27 Early loud vocalization, independent of seizure generalization, was seen frequently in 6 of the 10 anterior cases in our study. Notably, the anterior cingulate plays a role in vocalization associated with expressing internal states.26 Mutism can result from large and/or bilateral lesions of the anterior cingulate cortex in humans.26,28 Meyer and colleagues29 observed alterations in speech with cingulate stimulation. Fear was seen as an aura in 3 cases of the anterior cingulate group in our series. The anterior cingulate modulates emotional responses including fear.3,29 Interestingly, involuntary ictal urination during seizures without secondary generalization was observed in 2 patients in the anterior group. The aura of urge to urinate has been reported in relationship to the anterior cingulate,26 but it has also been linked to the nondominant temporal lobe.30 Notably, a positron emission tomography study in healthy volunteers demonstrated hypometabolism involving the anterior cingulate gyrus during voluntary retention.31
Bancaud and others have reported a myriad of autonomic manifestations in relation to anterior cingulate seizures.3,7,32 Surprisingly, in our series, the only autonomic manifestation reported was tachycardia, which was difficult to attribute to primary autonomic activation in the presence of concomitant violent hypermotor seizures. It is possible that subtle autonomic features were overlooked during postictal examination, perhaps superseded by the striking clinical manifestations.
Imaging studies consistently report acute nociceptive responses in the cingulate.33,34 None of the patients in our series reported painful auras despite preservation of awareness. The absence of pain-related symptoms might be attributed to the role of the cingulate in mediating the affective component of pain rather than pain perception per se.
The most commonly reported persistent behavioral changes with anterior cingulate lesions include reduced aggression, diminished fear, decreased motivation, and inappropriate behavior.26 Autistic, reclusive, sociopathic, self-mutilating, and obsessive-compulsive personality changes have also been described.26,35 Three patients with anterior cingulate lesions in our study had striking interictal personality changes manifested as aggression and paranoid delusions that led to imprisonment for 1 patient. Nearly complete resolution of these symptoms was noted after resection of the epileptogenic lesion.
We hypothesize that the observed differences in the atypical group of anterior cingulate cases reflect early activation of extracingulate networks as a result of chronic changes in the cortex adjacent to the tumor,36 direct invasion by the tumor per se, or surrounding areas of cortical dysplasia with intrinsic epileptogenicity associated with the underlying tumors.37
Little is known about the electroclinical behavior of posterior cingulate epilepsy. Koubeissi et al8 reported a case of lesional posterior cingulate epilepsy with seizures characterized by staring and automatisms, with bilateral interictal mesial temporal discharges and a rapid propagation to the ipsilateral mesial structures on invasive studies. The posterior cingulate cortex is known to project to the hippocampus via the entorhinal and parahippocampal cortex and to receive direct and indirect connections from the hippocampus. Additionally, ictal propagation to the cingulate gyrus is frequently observed among patients with temporal lobe epilepsy.38 Therefore, it is no surprise that all of the posterior cases had features suggestive of temporal lobe origin. In our series, ictal patterns involved the ipsilateral temporal lobe; bilateral temporal ictal patterns have also been described.8 None of the patients had the typical organized theta rhythms associated with mesial temporal lobe epilepsy.39 This report highlights the importance of considering cingulate epilepsies, the posterior cingulate in particular, among the group of pseudotemporal epilepsies. In our lesional posterior cingulate patients, we could not identify definite EEG findings to suggest extratemporal localization with the exception perhaps of the paucity of sustained organized ictal theta rhythms and the nonlocalizable rhythms in 1 patient (patient 12). These findings have modulated our practice over the years. Specifically, we now resort to an extensive intracranial (stereo-EEG) exploration of the limbic and paralimbic network in patients who present with features of temporal lobe epilepsy but have no clear structural/functional abnormalities within the temporal lobe. Given their cytoarchitectural commonalities and rich connectivity to the orbitofrontal cortex, the insula and cingulate need to be explored in addition to the mesial temporal structures, temporopolar cortex, and temporal neocortex in such patients.40
In some patients, the resection was extended outside the cingulate; even in the absence of apparent lesion on neuroimaging (primarily in suspected tumoral cases), it is possible that some semiological features and favorable outcome were owing to activation and resection of the extracingulate structures, respectively. Currently, to our knowledge, there is no available biomarker of epileptogenicity to reliably identify relevant epileptogenic tissue. In this article, we used good surgical outcomes following resection as a surrogate gold standard of the epileptogenic zone. Although this approach is commonly used in surgical series, we acknowledge that the concept of viewing a complex epileptogenic process as a solitary focus that can be successfully removed is rather simplistic. This holds true especially for nonlesional or subtly lesional cases. Our understanding of epileptogenicity in the face of existing dysfunctional networks has been evolving in the past few years in parallel with our ability to study large neural networks and their interactions.
In summary, cingulate epilepsy is a rare disorder. Anterior cingulate epilepsy has spectacular motor manifestations frequently with fear, laughter, or personality changes. The electroclinical presentation and surgical outcome of lesional anterior cingulate epilepsy is possibly influenced by the underlying pathology. Posterior cingulate epilepsy presents with electroclinical findings that are highly suggestive of temporal lobe origin and can be considered an example of pseudotemporal epilepsies. This study highlights the difficulty in localizing seizures arising from the cingulate gyrus in the absence of an MRI-identifiable lesion.
Accepted for Publication: March 9, 2013.
Corresponding Author: Rafeed Alkawadri, MD, Department of
Neurology, Yale University School of Medicine, 15 York St, LCI 714B, New Haven, CT 06510 (email@example.com)
Published Online: June 10, 2013. doi:10.1001/jamaneurol.2013.2940.
Author Contributions: Dr Alkawadri had full
access to all the data in the study and takes responsibility for the integrity of the data and the
accuracy of the data analysis.
Study concept and design: Alkawadri, Van Ness, Alexopoulos.
Analysis and interpretation of data: Alkawadri, So, Alexopoulos.
Drafting of the manuscript: Alkawadri, Alexopoulos.
Critical revision of the manuscript for important intellectual content: So, Van
Study supervision: Van Ness.
Conflict of Interest Disclosures: None
Previous Presentations: This work was presented at the Annual Meeting of the
American Academy of Neurology; April 13, 2011; Honolulu, Hawaii.