Restricted Frontomesial Epileptogenic Focus Generating Dyskinetic Behavior and Laughter


Corresponding author: Dr P. Kahane, laboratoire de neurophysiopathologie de l'épilepsie, pavillon de Neurologie, Hôpital Nord, 38043 Grenoble Cedex, France. E-mail:


Summary: Purpose: Substantial data are missing about the anatomic location of frontal regions supporting gelastic seizures.

Methods: We report the results of stereo-electro-encep halographic recordings performed over several distinct functional premotor and executive fields in a patient whose seizures were characterized by dyskinetic behavior and ictal laughter, in the absence of cerebral MRI abnormalities.

Results: The epileptogenic zone was circumscribed in the anterior and ventral part of the supplementary motor area and the underlying dorsal cingulate cortex. There were no or little spreading to cortical neighboring areas. The patient is seizure-free (follow-up of 27 months) after a stereotactic electric radiofrequency lesion of the epileptogenic focus.

Conclusion: The present data suggest that pericingulate premotor areas are involved in the triggering of the motor component of laughter. In this case, the coexistence of paroxysmal dyskinesias during laughter might reflect the involvement of specific compartment(s) of the basal ganglia.


In contrast to a previous hypothesis suggesting that the supplementary motor area (SMA) constitutes the main premotor area of the medial wall, studies in monkeys have subsequently demonstrated morphological and functional properties distinguishing a rostral subdivision of the SMA, namely pre-SMA (1,2) and several cingulate motor areas (CMA) located on the banks and in the depth of the cingulate sulcus (1,3,4). In humans, the specific contribution of each of these fields to behavior is currently studied, but it remains a gap between the advance in neurophysiological knowledge and their application in frontal lobe epilepsy. Classically, SMA seizures are characterized by somato-sensitive feelings, asymmetric limb tonic posturing, adversive oculo-cephalic movements, speech arrest or pallilalia, with or without preservation of consciousness (5,6). Although there are few well documented cases of epileptogenic focus restricted to the SMA, the respective role of the SMA and of neighboring cortical areas for the occurrence of the ictal symptomatology remains a debatable issue (7–10).

Here, we report the case of a 24 years-old patient whose seizures, characterized by hypermotor and laughting behavior, were proven to arise from the anterior part of the SMA proper (SMAp) and the underlying dorsal cingulate gyrus. The patient was successfully treated by a focal electrocoagulation of these regions, thus confirming the localization of the epileptogenic focus.


A 24-year-old right-handed man, with no history of other neurological disorder, was referred to our institution for refractory partial epilepsy. The first unprovoked epileptic seizure occurred at age 4 years. He was treated by valproate and remained seizure-free until the age of 10, when seizures reappeared 4–5 times per day. Seizures were then controlled by carbamazepine until the age of 18. The relapsing seizures became intractable to antiepileptic drug treatment with a mean frequency of 30 seizures per day. The epileptic attacks remained highly stereotyped. They were brief (10 to 30 seconds) and occurred primarily during wakefulness. At onset, the patient described a slightly painful sensation of tightness in the right shoulder and in the region of the right sternocleidomastoid muscle. Then, involuntary raising of the right arm, bilateral tonic posturing of the arms, swinging movements of the trunk and laughter occurred. The patient remained conscious and afterwards did not describe any subjective feeling of cheeriness or cheerfulness. Interictal neurological examination did not reveal any abnormalities and repeated high-resolution MRIs of the brain were normal.

Many seizures were recorded during the video-EEG monitoring. Interictal scalp EEG showed paroxysmal slow waves, sharp waves, and spike-wave complexes over the vertex. Ictally, a low voltage fast activity followed by a spikes discharge could be seen bilaterally over the precentral regions and the vertex, with a slight predominance on the left hemisphere. Analysis of the ictal semiology confirmed the occurrence of rapid-onset bilateral tonic posturing of the upper limbs and swinging-twisting movements of the trunk, preceded by brief choreoathetosic movements of the right hand. Consciousness was never impaired and the patient was most often able to point out the onset of a seizure. The laughter was not present at the onset of the seizures but occurred simultaneously with the swinging of the trunk and the dyskinetic posturing of the arms. It was spasmodic and accompanied by a forced smile. The postictal examination failed to identify any emotional component associated to the laughter.

The WADA test confirmed the left hemispheric specialization for the language.

The patient underwent stereo-electro-encephalographic (SEEG) recordings for accurate delimitation of the epileptic focus and in order to perform a functional mapping of cortical areas prior to surgery. Eleven electrodes explored the left premotor regions (Fig. 1), whereas two additional right electrodes explored the SMA and the anterior cingulate gyrus (ACG).

Figure 1.

Left: Schematic drawing of a parasagittal view of the left hemisphere, showing the position of the most medial contact of each electrode in relation to the sulcal anatomy. The location of the electrode contacts were plotted on preoperative frontal and parasagittal slices MRI by superimposition of the postoperative frontal and lateral skull radiography in relation to the anterior commissure – posterior commissure (AC-PC) plane. Electrodes positions into premotor areas: A: SMAp; B and C: pre-SMA; D: cCMA. Anatomic data, low intensity 1 Hz and 50 Hz electrical stimulations (ES) of electrodes A, B, C, D and E and movement related potential (Fig. 2), confirmed the location of these electrodes in motor fields. Right: theorical boundaries of the medial premotor areas (11,14). Abbreviations: AC Anterior commissure, PC Posterior commissure, VCA and VCP: vertical line passing through the AC and PC, 1: sulcus rostralis inferior, 2: sulcus rostralis superior; 3: cingulate or callosomarginalis sulcus; 4: intracingulate sulcus; 5: ascending branch of the cingulate sulcus; 6: sulcus paracentralis; 7: sulcus precentralis; 8: sulcus centralis; 9: sulcus subparietalis. (Ref: Chassagnon et al. Restricted frontomesial epileptogenic focus generating dyskinetic behavior and laughter.)

Interictal epileptiform discharges were seen on the left hemisphere on contacts located in the SMAp, pre-SMA and cCMA but subclinical rhythmic spikes and sharp-waves were largely predominant in the SMAp (electrode A) and the cCMA (electrode D).

The occurrence of seizures was dramatically reduced after electrode implantation, so that only one spontaneous subjective seizure was recorded during the SEEG monitoring. During this seizure, the patient complained of a slight sensation of tightness in the right shoulder, while the ictal SEEG showed a discharge of spikes and polyspikes at 5–6 Hz restricted to the SMAp and lasting 55 sec (see Fig. 2).

Figure 2.

Spontaneous partial seizure characterized by a slight sensation of tightness in the right shoulder. The 5–7 Hz rythmic discharges of sharp waves and spikes were recorded in the SMAp, and at a lesser degree in the pre-SMA. (Ref: Chassagnon et al. Restricted frontomesial epileptogenic focus generating dyskinetic behavior and laughter.)

Furthermore, seizures clinically similar to those recor ded during the scalp-video-EEG monitoring and iden tified by the patient as his usual seizures were induced by low intensity stimulation (0,4 mA, frequency 50 Hz, pulse width 1 ms) of the SMAp. These electrically-induced ictal 25–30 Hz discharges of spikes involved mainly the cCMA and, at a lower degree, the SMAp and the primary motor cortex. No discharges were recorded in the pre-SMA, premotor lateral cortex and cingulate gyrus. The signal symptom (i.e., sensation of tightness in the right shoulder) was reproduced by electrical stimulations (3 mA) of the SMAp, CMAc and cingulate gyrus (electrode E), in the absence of postdischarge.

Based on the interictal and ictal SEEG recordings and effects of ES, stereotactic monopolar electric radiofrequency lesions were performed over four and two medial contacts of the SMAp and CMAc electrodes respectively, at a voltage of 30 to 40 V, intensity 10–20 mA. The postoperative neurological status was normal. Before the removal of the depth electrodes 24 hours later, the interictal SEEG epileptiform discharges previously recorded over the SMAp had disappeared and electrical stimulations of the SMAp and CMAc did not induce any discharges despite an intensity 8 times higher (3 mA) than the one responsible for seizures before surgery.

The patient is currently seizure free with a follow-up period of 27 months, and antiepileptic drugs have been reduced toward a monotherapy by topiramate.

The postoperative MRI (Fig. 3) performed 6 months after surgery showed circumscribed round signal changes in the left SMAp and underlying dorsal part of the cingulate gyrus (overlapping the CMAc).

Figure 3.

Post-operative MRI showing signal changes at the location of the stereotactic radiofrequency lesions performed 6 months before. Left: para-saggital view of the left hemisphere, T1 signal. Right: Horizontal view, T2 signal. (Ref: Chassagnon et al. Restricted frontomesial epileptogenic focus generating dyskinetic behaviour and laughter.)


Correlations between anatomical, SEEG and clinical data are rare in focal mesial premotor epilepsies because documented cases are most often lesional and/or because the epileptogenic area encompasses neighboring areas in addition to the SMA and/or the CMA (8,9). The present case deserves reporting because of the restricted size of the epileptogenic focus limited to the SMAp and the underlying dorsal aspect of the cingulate gyrus, as proven by the seizure free status after focal radiofrequency lesions.

The patient had only one spontaneous seizure during the SEEG monitoring, that could be related to a phenomenon previously reported as transient cessation of seizures after implantation of depth electrodes by Katariwala and colleagues (15). Nevertheless, the epileptogenic zone was restricted to the anterior part of the SMAp and the cCMA, for the following reasons: (i) the strict location of interictal and subclinical epileptiform discharges in the SMAp; (ii) the occurrence of electrically induced seizures at low-intensity stimulations of the SMAp; (iii) the disappearance of interictal epileptiform discharges, spontaneous and electrically-induced discharges after radiofrequency stereotactic lesions over the SMAp and CMAc electrodes.

Our data suggest that the epileptogenic zone lay ahead of the paracentral sulcus, just behind the VAC line and on both sides of the cingulate sulcus. The upper focus appears located in the anterior and ventral part of the SMAp, while the lower focus seems to be located in the caudal CMA, within or at the vicinity of the posterior part of the area 24c' or gigantopyramidal area of Braak (14). This latter dorsal part of the cingulate gyrus is functionally different from anterior affective, visceromotor, nociceptive, attention-to-action and cognitive cingulate fields previously described (16–18).

The ictal clinical semiology observed in the present patient did not have the typical semiology of the so-called SMA seizures (19–20), nor the instinctivo-motor or affective behavior reported in cases of stimulations and epilepsies of the ACG (17,21,22). By contrast, we observed a combination of bilateral dyskinetic movements characterized by a rapid succession of dystonic postures, usually reported in movement disorders involving the basal ganglia (23). Such ictal motor behavior was also observed in a case of cryptogenic epilepsy in which the seizures propagated from the SMA toward the caudate nucleus (24). This is in agreement with the existence of direct projections of the SMA not only toward the primary motor cortex, but also toward the striatum (bilateral projections), the red nucleus, the subthalamic nucleus, the claustrum, and the amygdala, as described in the monkey (25), suggesting that the basal ganglia could support the clinical dystonico-posturing expression of the present patient's seizures. It is noteworthy that in a recent publication, subthalamic nucleus stimulation inducing dyskinesia in a patient with Parkinson's disease, could also give rise to laughter (26). This supports the possible involvement of common neural networks during dyskinesias and laughter.

The present report also provides interesting informations about the possible role of the ACG in gelastic seizures. Pathological laughter has been reported in cases of lesions disrupting cerebro-ponto-cerebellar pathways at different levels and of diverse origins (see for review 27). Electrical stimulation and seizures arising from hypothalamic hamartoma, temporal and frontal regions were responsible for documented cases of gelastic seizures, with or without concomitant subjective feelings of mirth (28). By contrast to the case of electrically induced laughter reported by Fried et al. (29), the present patient did not have the appropriate affective tone. This absence of concomitant feeling of mirth nor sadness in case of frontomesial epilepsy have already been reported by Arroyo et al. in a case of a lesional epilepsy involving the ACG and the SMA (30) and is also in accordance with the ictal neutral affect of a patient with gelastic seizures in whom the dipole sources were located in the ACG (31). Our report provides further insights about the anatomical location of the frontal medial region involved in laughter's expression. The data suggest that the occurrence of laughter in the present patient required the involvement of the supra and infra cingulate-sulcus fields affected by ictal discharges and subsequently destroyed by radiofrequency, since the stimulation of each site separately was not able to elicit laughter. This area seems to be located in a pericingulate field neighboring the caudal CMA. We suggest that this field is critical for the expression of laughter, in relation to effector structures of diencephalic and brain stem localization. The spasmodic shape of the laughter and the absence of simultaneous appropriate affective tone suggest that it was a motor active phenomenon rather than the expression of an underlying emotion. Taking into account previous related papers (24,26,29,30), the present anatomo-electrographic-clinical data suggest that a restricted pericingulate field overlapping the ventral SMAp and the CMAc is located at the crossroads of executive pathways liable to support laughter and dyskinetic movements.


Dyskinetic behavior and the motor component of laughter can be triggered by transient functional disruption of restricted mesial frontal areas located in the anterior and caudal part of the SMA proper and the underlying dorsal zone of the cingulate area 24 c'. The functional specialization of this area with respect to the “intention to action” role of the anterior cingulate cortex remains to be studied (32).


Acknowledgments:  The authors wish to thanks Dr A. Depaulis and Pr N. Moshe for reviewing the English manuscrit.


ACG, anterior cingulate gyrus; CMA, cingulate motor area; SMA, supplementary motor area; SEEG, stereo-electro-encephalography.