Les visages et leurs émotions

In this second part, we address particularly the question of the neural mechanisms and structures involved in the recognition of facial emotional expressions that are crucial in social cognition. Emotion recognition in others can be critically impaired in some neurodegenerative and neurovascular diseases. That dysfunction sometimes correlated to disabling behavioural disorders and interpersonal communication impairment must be further understood. The results of a series of scalp and intracranial event related potential recordings, as well as recent advances in the literature, are reported. ERPs to facial emotional expressions were thus recorded in multiple subcortical and cortical areas in drug refractory epileptical patients implanted with depth electrodes. The roles of amygdala, insula and prefrontal cortex located at crossroads between perceptive analysis and emotional conceptual knowledge are particularly underlined. Altogether, these studies demonstrate that facial expressions are widely processed in space and time, some structures reacting very early and automatically, others providing a sustained reaction depending on the attention.

Les visages et leurs émotions

Faces represent a crucial vector of interhuman communication. The message transmitted by the face has multiple features. Recognition of each feature can be impaired independently or in combination with others. In order to understand the behavioral consequences of such impairments, which can be a major social handicap, we first must specify the neural networks involved in face recognition. We propose in this first part to present the systems involved in face recognition, in particular the question of identity and prosopagnosia. Different neural networks are indeed implicated in the recognition of invariant facial features such as identity, gender, ethnicity, and recognition of variant features like facial expression and eye gaze. This paper is illustrated by some of our scalp and intracranial electrophysiological studies performed in humans allowing us to describe some aspects of face recognition dynamics combining an excellent spatial and temporal resolution. Intracranial recordings were performed in drug refractory epileptical patients implanted with depth electrodes. These studies demonstrate that numerous deep brain and cortical structures participate early and sometimes in a sustained manner in face recognition.

Temporo-mesial epilepsy surgery: outcome and complications in 100 consecutive adult patients

BACKGROUND

We studied the surgical outcome, and the complications in a group of 100 consecutive adult patients with medically refractory epilepsy arising from the temporo-mesial structures.

METHODS

Hundred patients were treated surgically between 1994 and 2003 for drug-resistant epilepsy involving the temporo-mesial structures. All of them underwent a comprehensive noninvasive presurgical evaluation. Fourty-eight of them underwent depth electrodes recordings (according to the Talairach's StereoElectroEncephaloGraphic (SEEG) methodology) because the noninvasive investigations were not congruent enough to identify the epileptic zone. The patients presenting with any space-occupying lesion, or with a cavernoma, or with a strictly lateral neocortical epileptic focus, were excluded. The MRI-examination was abnormal in 87 cases, displaying a hippocampal atrophy in 69 cases. The extent of temporal resection was planned according to the results of the presurgical investigation in each particular patient. Consequently, this "tailored" resection varied from selective amygdalo-hippocampectomy (6 cases), to anterior temporal lobectomy (76 cases), or to total temporal lobectomy (18 cases).

FINDINGS

The mean post-operative follow-up period was 53 months. 85 patients were found to be in Engel's class I post-operatively (free of disabling seizures), among them 74 were in class Ia (totally seizure free). Nine patients were in Engel's class II and six were in Engel's class III or IV (failures). There was no surgical mortality. Three patients had a postoperative hematoma; two patients required a shunt insertion; in three patients meningitis occurred; and two patients had postoperative ischaemia of the anterior choroidal artery territory, which resulted in a mild permanent hemiparesis. Neuropsychological complications are not addressed in detail in this article.

CONCLUSIONS

These data indicate that "tailored" resective surgery for temporo-mesial epilepsy can be performed with a low rate of morbidity, and is highly efficacious. The use of invasive presurgical investigation (SEEG) may explain this high rate of success.

[11C]-Methionine PET: dysembryoplastic neuroepithelial tumours compared with other epileptogenic brain neoplasms

BACKGROUND AND OBJECTIVES

Brain tumours responsible for longstanding partial epilepsy are characterised by a high prevalence of dysembryoplastic neuroepithelial tumour (DNT), whose natural evolution is much more benign than that of gliomas. The preoperative diagnosis of DNT, which is not yet feasible on the basis of available clinical and imaging data, would help optimise the therapeutic strategy for this type of tumour. This study tested whether [(11)C]-methionine positron emission tomography (MET-PET) could help to distinguish DNTs from other epileptogenic brain tumours.

METHODS

Prospective study of 27 patients with partial epilepsy of at least six months duration related to a non-rapidly progressing brain tumour on magnetic resonance imaging (MRI). A structured visual analysis, which distinguished between normal, moderately abnormal, or markedly abnormal tumour methionine uptake, as well as various regions of interest and semiquantitative measurements were conducted.

RESULTS

Pathological results showed 11 DNTs (41%), 5 gangliogliomas (18%), and 11 gliomas (41%). MET-PET visual findings significantly differed between the various tumour types (p<0.0002), regardless of gadolinium enhancement on MRI, and were confirmed by semiquantitative analysis (p<0.001 for all calculated ratios). All gliomas and gangliogliomas were associated with moderately or markedly increased tumour methionine uptake, whereas 7/11 DNTs had a normal methionine uptake, including all six located in the mesiotemporal structures. No DNT presented with a marked MET-PET abnormality.

CONCLUSION

Normal MET-PET findings in patient with an epileptogenic and non-rapidly progressing brain tumour are suggestive of DNT, whereas a markedly increased tumour methionine uptake makes this diagnosis unlikely.

Short and middle-latency Median Nerve (MN) SEPs recorded by depth electrodes in human pre-SMA and SMA-proper

OBJECTIVE

To analyse waveforms, latencies and amplitudes of Median Nerve (MN) SEPs recorded by stereotactically electrodes implanted in the SMA of 14 epileptic patients (9 in pre-SMA,3 in SMA-proper, 2 in both) in order to evaluate which short and middle-latency SEPs are generated in this area and which could be the physiological relevance of these responses.

METHODS

Short and middle-latency MN SEPs were recorded by chronically implanted electrodes in the fronto-temporal cortex and in particular in the mesial frontal region of 14 drug-resistant epileptic patients. MN stimulations of 100 micros were delivered by skin electrodes at the wrist; stimulus intensity was adjusted slightly above the motor threshold.

RESULTS

The main result of this study is that middle-latency SEPs were originated in pre-SMA but not in SMA-proper as demonstrated by both referential and bipolar recordings. In particular off-line computed bipolar traces between neighbouring contacts implanted in the pre-SMA and in the frontal external regions showed a phase reversal at the deepest contacts located in pre-SMA. Conversely, bipolar recordings between neighbouring contacts implanted in the SMA-proper and in the frontal external regions showed inversion recovery at more superficial contacts, implanted in area 6. Finally, we confirmed that no short-latency MN SEP (and in particular the N30) is originated in the whole SMA.

CONCLUSIONS

Among premotor areas, somatosensory inputs seem to reach pre-SMA and area 6 but not SMA-proper.

SIGNIFICANCE

This study assessed that no scalp SEP in the first 100 ms after MN stimulus could be generated in SMA-proper.

Somatosensory and pain responses to stimulation of the second somatosensory area (SII) in humans. A comparison with SI and insular responses

Somatosensory and pain responses to direct intracerebral stimulations of the SII area were obtained in 14 patients referred for epilepsy surgery. Stimulations were delivered using transopercular electrodes exploring the parietal opercular cortex (SII area), the suprasylvian parietal cortex (SI area) and the insular cortex. SII responses were compared to those from adjacent SI and insular cortex. In the three areas we elicited mostly somatosensory responses, including paresthesiae, temperature and pain sensations. The rate of painful sensations (10%) was similar in SII and in the insula, while no painful sensation was evoked in SI. A few non-somatosensory responses were evoked by SII stimulation. Conversely various types of non-somatosensory responses (auditory, vegetative, vestibular, olfacto-gustatory, etc.) were evoked only by insular stimulation, confirming that SII, like SI, are mostly devoted to the processing of somatosensory inputs whereas the insular cortex is a polymodal area. We also found differences in size and lateralization of skin projection fields of evoked sensations between the three studied areas, showing a spatial resolution of the somatotopic map in SII intermediate between those found in SI and insula. This study shows the existence of three distinct somatosensory maps in the suprasylvian, opercular and insular regions, and separate pain representations in SII and insular cortex.

Hippocampal-orbitofrontal connectivity in human: an electrical stimulation study

OBJECTIVE

The identification of the pathways involved in seizure propagation remains poorly understood in humans. For instance, the respective role of the orbitofrontal cortex (OFC) and of the commissural pathways in the interhemispheric propagation of mesial temporal lobe seizures (mTLS) is a matter of debate. In order to address this issue, we have directly tested the functional connectivity between the hippocampus and the OFC in 3 epileptic patients undergoing an intra-cranial stereotactic EEG investigation.

METHODS

Bipolar electrical stimulations, consisting of two series of 25 pulses of 1 ms duration, 0.2 Hz frequency, and 3 mA intensity, were delivered in the hippocampus. Evoked potentials (EPs) were analysed for each series, separately. Grand average of reproducible EPs was then used to calculate latency of the first peak of each individual potential.

RESULTS

Hippocampal stimulations evoked reproducible responses in the OFC in all 3 patients, with a mean latency of the first peak of 222 ms (range: 185-258 ms).

CONCLUSIONS

Our data confirm a functional connectivity between the hippocampus and the OFC in human.

SIGNIFICANCE

This connectivity supports the potential role of the OFC in the propagation of mTLS.

Intracranial EEG study of seizure-associated nose wiping

The authors studied the relation between seizure-associated nose wiping (NW) and intracerebral EEG data in 32 patients. NW was more frequent in mesial temporal lobe seizures (TLSs; 65%) than in other TLSs (36%; p < 0.05) and in frontal lobe seizures (3%; p < 0.0001). It was associated with the presence of an amygdala discharge at seizure onset (p < 0.05) and with the recording of an ictal low-voltage fast activity within that structure (p < 0.05), supporting the role of an amygdala dysfunction in the pathophysiology of NW.

[Polymorphism of epilepsy associated with the A3243G mutation of mitochondrial DNA (MELAS): reasons for delayed diagnosis]

INTRODUCTION

Mitochondrial disease is a potential diagnosis in patients with epilepsy beginning in childhood or adolescence with a typical polymorphic presentation and preponderant occipital lobe seizures. Diagnosis may however be delayed in some patients with long-standing disease, particularly when cardinal mitochondrial symptoms are missing; clinical manifestations may be dissociated over time leading to genetic diagnostic tests being prescribed long after disease onset.

OBSERVATION

We report the case of a 17 year old woman in whom the diagnosis of lipothymic episodes, migraine, idiopathic photo-sensitive generalized epilepsy, and partial occipital epilepsy complicated by occipital epileptic status were successively proposed because of the initial clinical presentation and the slow disease course. Eleven years after disease onset the diagnosis of progressive myoclonic epilepsy was made due to the occurrence of myoclonic jerks with giant SEPs associated with a cerebellar syndrome, deterioration of psychomotor performances and diffuse slowing of EEG activity with pseudo-periodic bursts of delta waves. Genetic analysis showed an A3243G mutation of mitochondrial DNA, usually correlated with the MELAS phenotype, while the clinical presentation of progressive myoclonic epilepsy was more suggestive of MERRF.

CONCLUSION

Although each of the symptoms successively observed in this patient has been reported in MELAS, the slow course of the disease, which is unusual in this mutation, the absence of stroke-like episodes, and the polymorphism of the epilepsy all contributed to delayed final diagnosis.

Le lobe de l’insula et les épilepsies partielles

The role of the insular lobe in temporal lobe epilepsy (TLE) has often been suggested but never directly demonstrated. In this article, we review data from recent literature and from our stereo-electroencephalographic (SEEG) recordings in patients referred for temporal lobe epilepsy surgery (TLE). Our description of the clinical features of insular lobe seizures is based on data from video and SEEG ictal recordings and direct electric cortical stimulation in a population of 50 consecutive patients whose seizures, on the basis of scalp video EEG recordings, were suspected to originate from, or to rapidly propagate to, the peri-sylvian cortex. A total of 144 intra-insular electrodes have been implanted in this series of patients. In six patients a stereotyped sequence of ictal symptoms could be identified on the basis of electro-clinical correlations. The clinical presentation of insular lobe seizures was that of simple partial seizures occurring in full consciousness, beginning with a sensation of laryngeal constriction followed by paresthesiae that were often unpleasant affecting large cutaneous territories. These initial symptoms were eventually followed by dysarthric speech and/or elementary auditory hallucinations, and seizures often ended with focal dystonic postures. The insular origin of these symptoms was supported by the data from functional cortical mapping of the insula using direct cortical stimulations. We were able to reproduce several of the spontaneous ictal symptoms in the six patients with insular seizures. Moreover, from the whole set of insular stimulations that we performed it could be concluded that the insular cortex is involved in somatic, vegetative and visceral functions to which spontaneous ictal insular symptoms are related. The observation of the insular symptoms sequence at the onset of seizures in patients who are candidates for TLE surgery strongly suggests that the epileptic focus is located in the insular lobe. It entails the risk of unsuccessful temporal lobectomy and should lead: i) to reconsider the indication of temporal lobectomy and; ii) to explore directly the ictal activity of both mesio-temporal and insular cortices before making any decision regarding epilepsy surgery.

[Drug-resistant partial epilepsies: introductory remarks on their classification, pathophysiology and surgical treatment]

There is no validated classification of drug-resistant partial epilepsies in adults that have proved helpful to individualize homogeneous groups of patients, whose clinical course would be predictable at the early stage of the disease and whose treatment, eventually surgical, could be standardized. The only entity fitting with this definition is the syndrome of mesio-temporal lobe epilepsy, which accounts for a high percentage of surgical treatment indications. However the objective to perform surgical treatment in all of these patients is far from being reached in France, as in many other developed countries. Apart from this syndrome, therapeutic decisions are mostly based on case by case clinical and electrophysiological analysis of seizures, in relation with the localization and anatomy of the causal lesion, when the latter can be identified using brain MRI. The pragmatic approach consists in aiming at a consensus regarding the nature and the reliability of presurgical strategies, including invasive electrophysiological investigations, rather than in elaborating a syndromic classification, the pertinence of which cannot be validated. Some concepts and data that might be helpful to reach a consensus are discussed in this brief review article.

[Functional neuroimaging in adults]

A consensus has not been reached on the proper role of the different functional neuroimaging techniques available for the exploration of drug-resistant partial epilepsy, particularly in preparation for surgery. Several studies have suggested the potential contribution of positron emission tomography (PET) with [18F]-fluoro-deoxyglucose (FDG), [11C]-flumazenil (FMZ), or [11C]-alpha-methyl-tryptophane (AMT), single-photon emission computed tomography (SPECT) to measure ictal cerebral blood flow, magnetic resonance spectroscopy (MRS), and functional magnetic resonance imaging (fMRI) of language and memory functions, but to date there has been no impact study validating the clinical contribution of these different exploration techniques. The most robust data in the literature indicate that [18F]-FDG PET can help to predict surgical failures in patients with temporal lobe epilepsy. Studies with lesser power have suggested that SPECT measurement of ictal hyperperfusion blood flow, and [11C]-FMZ and [11C]-AMT PET can be useful in symptomatic neocortical epilepsy, particularly in patients with tuberous sclerosis explored with [11C]-AMT PET. Use of these different exploration techniques in epileptic patients with a normal MRI is warranted when complementary information could help in deciding on when and how to perform an invasive EEG, but there are significant risks of erroneous identification of the epileptogenic zone. Functional MRI assessment of language regions can be proposed as an alternative to the Wada test to determine hemispheric dominance in patients with temporal lobe epilepsy. Multicentric impact studies will be needed before evidence-based guidelines can be developed for the use of functional neuroimaging techniques in drug-resistant partial epilepsy.

The prognostic value of evoked responses from primary somatosensory and auditory cortex in comatose patients

OBJECTIVE

To evaluate somatosensory and auditory primary cortices using somatosensory evoked potentials (SEPs) and middle latency auditory evoked potentials (MLAEPs) in the prognosis of return to consciousness in comatose patients.

METHODS

SEPs and MLAEPs were recorded in 131 severe comatose patients. Latencies and amplitudes were measured. Coma had been caused by transient cardiac arrest (n=49), traumatic brain injury (n=22), stroke (n=45), complications of neurosurgery (n=12) and encephalitis (n=3). One month after the onset of coma patients were classified as awake, still comatose or dead. Three months after (M3), they were classified into one of the 5 categories of the Glasgow outcome scale (GOS).

RESULTS

At M3, 41.2% were dead, 47.3% were conscious (GOS 3-5) and 11.5% had not recovered consciousness. None of the patients in whom somatosensory N20 and auditory Pa were absent did return to consciousness and in the post-anoxic group, reduced cortical amplitude too was always associated with bad outcome. Conversely, N20 and Pa were present, respectively, in 33/69 and 34/69 patients who did not recover.

CONCLUSIONS

The prognostic value of SEPs and MLAEPs in comatose patients depends on the cause of coma. Measurement of response amplitudes is informative. Abolition of cortical SEPs and/or cortical MLAEPs precludes post-anoxic comatose patients from returning to consciousness (100% specificity). In any case, the presence of short latency cortical somatosensory or auditory components is not a guarantee for return to consciousness. Late components should then be recorded.

The human supplementary motor area-proper does not receive direct somatosensory inputs from the periphery: data from stereotactic depth somatosensory evoked potential recordings

Somatosensory evoked potentials to median nerve (MN) stimulation were recorded by stereotactically implanted electrodes in the frontal lobe of two epileptic patients in order to evaluate whether short-latency cortical responses could be generated in the supplementary motor area (SMA)-proper. In both patients two potentials, called P20 and N30 responses, showed a decreasing amplitude from the most superficial to the deepest contacts and were recorded in the white as well in the grey matter of the frontal lobe. Furthermore, no phase reversal of both P20 and N30 potentials was identifiable along electrode trajectories. Our results suggest that short-latency somatosensory evoked responses recorded in the SMA-proper contralateral to MN stimulation are volume conducted from remote cortical generators.

[Cervical spondylotic myelopathy: motor and somatosensory evoked potentials, clinical and radiological correlation]

Somatosensory and motor evoked potentials (EPs) have been recorded in 38 patients with cervical spondylosis, documented by MRI. All were symptomatic, 23 presented with myelopathy. Somatosensory evoked potentials were abnormal in 66p.cent as well as motor evoked potentials. The N13 potential, generated by the posterior dorsal horn of the cervical spinal cord, was abnormal in 63p.cent of patients and was the only disorder detected in 12p.cent. Together, somatosensory and motor evoked potentials were abnormal in 82p.cent of patients. There was no correlation between EPs findings and radiological data. Similary, MRI and clinical data were agreeing in only 50p.cent of patients. When a spondylotic myelopathy is suspected, SEPs proved more sensitive to detect somatosensory dysfunctionning than clinical testing and radiological data were discordant with clinical status in 50p.cent of cases. In order to obtain a high sensitivity, both somatosensory and motor evoked potentials should be recorded on all limbs with a special attention to segmental cervical and cervico-medullary responses. EPs data help to identify patients with cervical cord dysfunction and thus contribute to the therapeutic decision for surgery.

Early secondary somatosensory area (SII) SEPs. Data from intracerebral recordings in humans

OBJECTIVE

To record somatosensory evoked potentials (SEPs) to median nerve stimulation by chronically implanted electrodes in the parieto-rolandic opercular area of 9 epileptic patients, in order to evaluate whether somatosensory evoked responses could be generated in the second somatosensory area (SII) earlier than 40 ms after stimulus.

METHODS

Nine patients (4 males, 5 females) with drug-resistant partial epileptic seizures were investigated using stereotactically implanted electrodes in the parietal cortex, posterior to vertical anterior commissure plane and in the frontal opercular region rostral to vertical anterior commissure (VAC).

RESULTS

The main finding of this study is the recording of an early somatosensory evoked potential, (N30op), by chronically implanted electrodes in the SII area of 8 epileptic patients. In 3 patients where SEPs were performed after ipsilateral median nerve (MN) stimulation, a N30op was recorded 5.8+/-2 ms later than contralateral one.

CONCLUSIONS

This is the first report of early SEPs recorded by electrodes implanted in SII area. The N30op potential, even if less consistent than later potentials, confirmed the important role of the SII area in the early processing of somatosensory inputs.

Role of operculoinsular cortices in human pain processing: converging evidence from PET, fMRI, dipole modeling, and intracerebral recordings of evoked potentials

Insular and SII cortices have been consistently shown by PET, fMRI, EPs, and MEG techniques to be activated bilaterally by a nociceptive stimulation. The aim of the present study was to refer to, and to compare within a common stereotactic space, the nociceptive responses obtained in humans by (i) PET, (ii) fMRI, (iii) dipole modeling of scalp LEPs, and (iv) intracerebral recordings of LEPs. PET, fMRI, and scalp LEPs were obtained from normal subjects during thermal pain. Operculoinsular LEPs were obtained from 13 patients using deep brain electrodes implanted for presurgical evaluation of drug-resistant epilepsy. Whatever the technique, we obtained responses which were located bilaterally in the insular and SII cortices. In electrophysiological responses (LEPs) the SII insular contribution peaked between 150 and 250 ms poststimulus and corresponded to the earliest portions of the whole cerebral response. Group analysis of PET and fMRI data showed highly consistent responses contralateral to stimulation. On single-subject analysis, LEPs and fMRI activations were concentrated in relatively restricted volumes even though spatial sampling was quite different for both techniques. Despite our multimodal approach, however, it was not possible to separate insular from SII activities. Individual variations in the anatomy and function of SII and insular cortices may explain this limitation. This multimodal study provides, however, cross-validated spatial and temporal information on the pain-related processes occurring in the operculoinsular region, which thus appears as a major site for the early cortical pain encoding in the human brain.

Cognitive effects of precentral cortical stimulation for pain control: an ERP study

Electrical stimulation of the motor cortex (MCS) is a promising and increasingly used neurosurgical technique for the control of refractory neuropathic pain. Although its mechanisms of action remain unknown, recent functional imaging data suggest involvement of the thalamus, brainstem and anterior cingulate/orbitofrontal cortex. Since some of these areas are also implicated in higher cognitive functions, notably attentional processes, we analysed cognitive ERPs and behavioural performance during an "oddball" auditory detection task in patients submitted to this procedure. Eleven consecutive patients undergoing MCS because of neuropathic refractory pain, ranging in age from 25 to 71 years, were included in the study. ERPs were obtained in all cases both during the application ("MCS-on") and within the 10 min that followed discontinuation of the procedure ("MCS-off"). In five patients, ERPs could also be obtained just before the start of MCS. When the patients' sample was taken as a whole, there were no consistent effects of MCS on the ERPs. There was, however, a significant interaction of MCS action with the patients' age, reflecting a significant delay during MCS of the cognitive responses N2 and P3 (N200 and P300) in the group of patients older than 50 years exclusively. This effect was rapidly reversible after MCS discontinuation. No MCS-related changes were observed in the N1 component. At the individual level, the effect of MCS on the endogenous ERPs was highly variable, ranging from a total stability of ERPs (mostly in younger subjects) to latency differences of tens of milliseconds in the older group. These results, together with recent experiments showing P300 alteration during repetitive transcranial stimulation, suggest that motor cortex stimulation may interfere with relatively simple cognitive processes such as those underlying target detection, and that the risk of abnormal cognitive effects related to cortical stimulation may increase with age. Although the procedure appears on the whole remarkably safe, complementary neuropsychological studies in this category of patients are advised, as well as caution to possible adverse cognitive effects when using MCS in the elderly, notably in the presence of pre-existent cerebral lesions.