Stereo-EEG exploration in a case of eating epilepsy with cutlery-induced seizures

A rare type of reflex seizures/epilepsy is eating seizures/epilepsy (ES/EE) with estimated prevalence of 0.1-0.05 % amongst epilepsies []. Eating seizures may be triggered by mastication, thought/sight/smell of food, presence of food in the mouth without chewing, swallowing, certain food types, and by texture, temperature, or food-related sounds. We present a case investigated with stereo-EEG (SEEG) where eating with cutlery was the seizure trigger.

An 18 year-old right-handed man with drug-resistant epilepsy was admitted for SEEG. Medical and family history was unremarkable. At primary school reading and spelling were poor.

He had daily and occasional nocturnal seizures, without warning, consisting of unresponsiveness, right deviation of eyes and head, right arm jerking, followed by post-ictal dysphasia with numbness and weakness of the right limbs. Occasionally they evolved to bilateral tonic-clonic seizures. Although the great majority of seizures would happen close to meals, no clear link between eating and seizures was established during the course of his epilepsy.

Epilepsy protocol MRI was normal. An FDG-PET showed left temporal hypometabolism, extending to occipital and parietal lobes (Fig. 1). Interictal scalp EEG showed frequent slow waves and high voltage spikes over the left mid-posterior temporal region. Seizure onset was seen at the same region.

Implantation of eleven 10-contact Dixi depth electrodes (DIXI Medical, Besancon, France) (A–K) in the left hemisphere was guided by PET and scalp EEG findings (Fig. 1). During 10 days of recording with drug reduction, five typical clinical seizures occurred, four related to eating and one during sleep. Frequent subclinical seizures consisting of bursts of fast activity at about 40 Hz lasting between 2–10 seconds occurred every time he was eating with cutlery. The initial ictal SEEG changes were similar in clinical and subclinical seizures (Fig. 2). The act of picking up cutlery and start eating produced ictal activity. There were no clinical or subclinical seizures when the patient had finger-food, drank beverages or when he used cutlery to prepare his food. The relationship to the emotional context of eating was tested by a detailed 15-minute discussion concerning his favourite food, and a series of photos of delicious-looking meals presented over 30 min. Food was delivered to him but before he started eating, the meal was taken away. None of these provoked ictal activity. Furthermore, we offered a meal consisting of food that the patient did not like and although there was no positive food anticipation, eating with cutlery again produced several episodes of ictal activity.

Receptive language disturbance was noted when stimulating the temporo-occipital junction (G), the deep angular gyrus (H), and the deepest contacts at the temporo-parietal junction (I). Habitual seizure onset was provoked by stimulation (50 Hz, 5 mA) at the deep contacts (2–3) of the inferior temporo-occipital area electrode (F). The results suggested regional seizure onset including the receptive language area and resection was not offered.

Published cases/series of eating seizures (ES) are heterogeneous. The aetiology in our case is unclear, but the combination of developmental language disorder, FDG-PET abnormality, normal MRI, frequent interictal epileptiform discharges, and electrical subclinical seizures suggest an underlying architectural cortical dysplasia.

Two categories of eating seizures (ES) have been proposed; ES with temporo-limbic and ES with extratemporal, mainly suprasylvian, onset []. The two reported cases of eating seizures assessed by SEEG found a seizure onset in the anterior insula [] and in the mesial temporal/inferior insular region [] and the patients were seizure free after resection. In our case Scalp-EEG and PET data pointed to a posterior perisylvian focus. As the link between eating and his seizures was unknown prior the implantation, the insular cortex was not sampled. SEEG showed ictal activity starting immediately after, or rarely exactly before the first mouthful, when cutlery was in patient’s hands and was moving to place food in the mouth (Fig. 2). The build-up of subclinical seizures could lead to a clinical seizure. By use of SEEG we were able to show in a case with focal epilepsy, a phenomenon similar to the build-up of myoclonic jerks preceding cognition-induced seizures.

Proprioceptive and somatosensory afferents from the oropharyngyal area are thought to trigger ES of extratemporal origin []. However, our patient would have a seizure only when using cutlery, suggesting that the combination of afferents from the hands and the oropharyngyal region, or the complex behaviour of eating with cutlery involving gustatory, somatosensory and visual-praxis afferents, was required to provoke a seizure. Our testing did not include the use of cutlery without eating but no subclinical seizures were noted on SEEG when the patient used cutlery to prepare the food (i.e. slicing) and they would appear only when he used the cutlery to place food to the mouth. No seizures were reported when he was writing or using other tools. A later behavioural modification by not using cutlery markedly reduced his seizure frequency. A similar case has been reported in a nine year old boy with focal seizures only when he ate using knife and fork. The accidental amputation of three fingers from his right hand rendered him seizure free, suggesting that the combination of sensory input from the right hand and oropharyngyal region was required to surpass the epileptogenic threshold []. Despite the absence of insula sampling and the lack of confirmation of the epileptogenic zone, our case is the first to illustrate the relation between trigger and ictal SEEG activity, suggesting that the proprioceptive and somatosensory activation includes extra-oral inputs arising from the body parts involved in the complex function of eating. This consideration is shared in a report of ES in a case where, paradoxically, chewing stopped the interictal epileptiform discharges and another case where ES did firstly appear when the patient started eating by herself []. Triggers in reflex epilepsies may involve combination of afferents causing the link to be missed from clinical history. Establishing such a link is important for implantation strategies using the knowledge from previous reports. Data show that sampling insula and posterior perisylvian areas should be considered when assessing cases with ES but more reports of cases with this condition undergoing SEEG will be helpful.

The authors received no financial support for the research, authorship, and/or publication of this article.