Abstract

We studied painful and non-painful somaesthetic sensations elicited by direct electrical stimulations of the insular cortex performed in 43 patients with drug refractory temporal lobe epilepsy, using stereotactically implanted depth electrodes. Painful sensations were evoked in the upper posterior part of the insular cortex in 14 patients, mostly in the right hemisphere. Non-painful sensations were elicited in the posterior part of the insular cortex in 16 patients, in both hemispheres. Thus, painful and non-painful somaesthetic representa-tions in the human insula overlap. Both types of responses showed a trend toward a somatotopic organization. These results agree with previous anatomical and unit recording studies in monkeys indicating a participation of the posterior part of the insular cortex in processing both noxious and innocuous somaesthetic stimuli. In humans, both a posterior and an anterior pain-related cortical area have been described within the insular cortex using functional imaging. Our results help to define the respective functional roles of these two insular areas. Finally, lateralization in the right hemisphere of sites where painful sensations were evoked is coherent with the hypothesis of a preponderant role of this hemisphere in species survival.

Introduction

Based on numerous anatomical and microelectrode studies in monkeys, the involvement of the insular cortex in processing painful as well as non-painful somatosensory inputs has been increasingly contemplated (Burton and Jones, 1976; Robinson and Burton, 1980; Mufson and Mesulam, 1984; Friedman and Murray, 1986; Apkarian and Hodge, 1989; Hodge and Apkarian, 1990; Schneider et al., 1993; Apkarian and Shi, 1994; Craig et al., 1994; Dostrovsky and Craig, 1996; Blomqvist et al., 2000). Laser-evoked potentials in humans have shown bilateral dipolar sources in the second somatosensory area (SII) or the insular cortex (Tarkka and Treede, 1993; Bromm and Chen, 1995; Kakigi et al., 1995; Valeriani et al., 2000; Opsommer et al., 2001). Furthermore, functional imaging in humans have con-firmed these data by showing an activation of the insular cortex in response to painful and non-painful somaesthetic stimuli (Burton et al., 1993; Casey et al., 1994, 1996, 2001 Coghill et al., 1994; Hsieh et al., 1995b; Craig et al., 1996, 2000; Vogtet al., 1996 ; Andersson et al., 1997; Antognini et al., 1997; Derbyshire et al., 1997; Rainville et al., 1997; Svensson et al., 1997, 1998; Xu et al., 1997; Davis et al., 1998a,b; Derbyshire and Jones, 1998; Disbrow et al., 1998; Iadarola et al., 1998; May et al., 1998; Oshiro et al., 1998; Paulson et al., 1998; Bushnell et al., 1999; Gelnar et al., 1999; Peyron et al., 1999; Ploghaus et al., 1999; Tölle et al., 1999; Kwan et al., 2000) [for a review see Peyron et al. (Peyron et al., 2000)]. Studies on direct electrical stimulation of the insular cortex remain scarce. Respiratory, somato-motor and circulatory effects have been reported in primates and carnivores during insular cortex stimulation (Kaada et al., 1949; Babkin and Van Buren, 1951; Hoffman and Rasmussen, 1953). In humans, it has long been a challenge to stimulate the insular cortex during the presurgical assessment of epilepsy, due to its anatomic location, buried under the frontal, temporal and parietal opercular cortices and covered by a dense wall of vessels. Thus, only a few studies have reported non-nociceptive somaesthetic symptoms, cardiovascular effects as well as viscero-motor and viscero-sensitive sensations con-secutive to direct electrical stimulation of the insular cortex (Penfield and Faulk, 1955; Wieser, 1983; Oppenheimer et al., 1992; Ostrowsky et al., 2000). No painful sensation was so far described except in two patients we previously mentioned (Ostrowsky et al., 2000).

In this study, we report both painful and non-painful somaes-thetic sensations elicited by cortical stimulation of the insular cortex in patients undergoing a depth stereotactic recording (stereo-electroencephalography, SEEG) in the course of the pre-surgical evaluation of their temporal lobe epilepsies. Our data show that the posterior part of the insular cortex is involved in the processing of both painful and innocuous somaesthetic inputs. Furthermore, their respective topographical distribu-tions overlap within this cortical area.

Materials and Methods

Patients

The 43 patients included in this study (29 women and 14 men) were investigated at the Neurological Hospital in Lyon between January 1996 and November 2000. Language lateralization was assessed by intracarotid amobarbital test in all patients. Among the 36 right-handed patients, 35 had a left language representation and one a bilateral language representation. Among the seven left-handed patients, four had a left language representation, two had a bilateral language representation and one had a right language representation. All patients suffered from a drug-resistant temporal lobe epilepsy (TLE) and underwent a SEEG in the course of their presurgical investigation. Cortical structures to be explored were chosen according to video-EEG recordings of seizures, interictal [18F]fluoro-deoxyglucose positron emission tomography (18-FDG PET), interictal and ictal single photon emission tomography (SPECT) and cerebral magnetic resonance imaging (MRI) data. The decision of insular implantation in the patients of this study was based on the existence of ictal symptoms (such as lip and face paraesthesiae, auditory or gustatory hallucinations, laryngeal contraction, hypersaliva-tion, simple motor signs and post-ictal paresis) suggesting an early spread of the discharge to perisylvian structures (Isnard et al., 2000). Twenty-four patients presented a right TLE and 19 a left TLE.

Stereotactic Implantation and Insular Site Location

The stereotactic implantation technique used is that described by Talairach and Bancaud (Talairach and Bancaud, 1973). A cerebral angiography is first performed in stereotactic conditions using an X-ray source 4.85 m away from the patient's head, eliminating the linear enlargement due to X-ray divergence. In order to reach the pertinent cortical target, the stereotactic coordinates (Talairach and Tournoux, 1988) of each electrode [x for the medio-lateral axis, with x = 0 being the coordinate of the sagittal interhemispheric plane; y for the rostro-caudal axis, y = 0 being the coordinate of the frontal plane passing through the vertical anterior commissure (VAC); z for the vertical axis, z = 0 being the coordinate of the horizontal plane passing through the anterior and posterior commissures (AC–PC plane)] are calculated preoperatively on the individual cerebral MRI previously enlarged at scale 1. Cerebral MRI and angiographic images are easily superimposed to avoid any risk of vascular injuries during implantation.

At the end of the surgery, a post-implantation frontal X-ray was taken and superimposed on MRI images to verify the final position of each implanted electrode with respect to targeted anatomical structures.

Electrodes are implanted perpendicular to the mid-sagittal plane and can be left in place chronically up to 15 days. The electrodes have a diameter of 0.8 mm and 5–15 recording contacts, depending on their length. Contacts are 2 mm long and are separated by 1.5 mm from one another. The insular cortex was investigated using trans-opercular electrodes. In all cases, only the first and second deepest contacts of the trans-opercular electrode were found to be strictly located in the thickness of the insular cortex as checked on superimposed MRI and X-ray images (Fig. 1A). A total of 93 insular sites were explored, 49 in the right and 44 in the left insular cortex. Each of these insular sites was localized using the stereotactic coordinates of the deepest contact of the corresponding trans-opercular electrodes. The x coordinate (depth of implantation) was verified on the post-implantation frontal X-ray whereas the strict perpendicular trajectory of the electrode, verified on both the frontal and sagittal post-implantation X-rays, ensured that the preoperatively calculated y and z coordinates had been preserved. The estimated spatial error of this insular site location technique is ±2 mm in the three dimensions due to several factors: MRI slice thickness of 1 mm, MRI definition (pixel size) of 0.9 mm; pooling of sites of all patients on a single MRI saggital slice of the insular cortex, whether this slice was extracted from a standardized atlas (Montreal National Institute standard brain) or from the individual MRI of one of our patients (see below).

Stimulation Paradigm

A current regulated neurostimulator designed for a safe diagnostic stimulation of the human brain was used (Babb et al., 1980). Square pulses of constant polarity were applied between the two deepest contacts of the trans-opercular electrodes located in the insular cortex (see Fig. 1A). Two modes of stimulation were used; low frequency stimulation (LFS): 1 Hz, 1–3 ms, 1–5 mA applied during 5–10 s and high frequency stimulation (HFS): 50 Hz, 0.1–0.5 ms, 0.8–6 mA applied during 5 s. These parameters were chosen to avoid any tissue injury [charge density per square pulse <55 μC/cm2; see Gordon et al. (Gordon et al., 1990)]. These parameters, along with the bipolar mode of stimulation using adjacent contacts, ensured a good spatial specificity with respect to the desired structure to be stimulated (Nathan et al., 1993). HFS was used in all patients and at least once on each insular site. LFS was only performed in the first 20 patients and then abandoned due to its poor efficiency in terms of evoked clinical responses. Stimulation of cortical areas is a routinely applied procedure used in our unit during SEEG recordings to map functionally eloquent cortex and epileptogenic areas. Patients were fully informed of the aim of this investigation and gave their consent before entering this phase of the presurgical evaluation of their epilepsy. During stimulation, patients were sitting in bed and were asked to relax.

Results

A total of 188 stimulations (61 LFS and 127 HFS) were performed in the 93 insular sites in the 43 patients. Clinical responses (of different modalities) were evoked in only 35 of the patients (35/43; 81%), in 67 of the sites (67/93; 72%), and during 111 of these stimulations (59% of the stimulations, 14 LFS and 97 HFS). These responses were reported during the stimulation period and disappeared either immediately or very shortly (a few seconds) then after. The electrical stimulations could be followed by an after-discharge (AD) which either remained localized to the stimulation site (localized AD) or spread to other insular or extra-insular sites that were not directly stimulated (diffusing AD). In 52 sites, the clinical responses were evoked in the absence of AD. In 12 sites, the elicited responses were followed by a localized AD. The three sites where responses were followed by a diffusing AD were excluded from the topographical and somatotopic analysis, and therefore not represented on the figures.

Because only one to three insular sites were stimulated per patient, individual data remained poorly informative. Thus, the topographic and somatotopic analysis of our responses were performed by pooling sites where responses were obtained on a single sagittal insular image. In Figure 1 all stimulated sites, whether right or left, are plotted on a sagittal section of the insular cortex of the MNI (Montreal National Institute, Canada) standard brain (Fig. 1B) and on the corresponding slice of one of our patients (Fig. 1C). Due to inter-individual variations, the borders and anatomical details of the MNI brain slice are blurred as compared with those of the individual MRI slice. However, as can be seen in Figure 1, the sampling of the insular cortex was very similar using either of these two types of brain section. This plotting shows that most of the insular surface has been explored except the most anterior part of the insular cortex (y > to 15 mm). For illustration's sake, we plotted sites of stimulation on the individual sagittal MRI slice in other figures.

Painful Sensations Evoked by Insular Cortex Stimulation

Painful sensations were elicited in 17 insular sites (17/93; 18.2%) in 14 patients (14/43; 32.5%). Mean intensity of stimulation was 2.2 mA ± 0.9. Twelve sites were located in the right and five in the left insular cortex. Pain was elicited only during HFS, without AD on 13 sites (11 patients) and with a localized AD in two sites (in two patients). The two sites, both located in the left insular cortex of one patient where the painful sensations were followed by a diffusing AD, are excluded from the follow-ing discussion.

The 15 sites where pain was elicited in the absence of AD or with a localized AD were located in the upper posterior part of the insular cortex (Figs 2A and 4A). Pain intensity could vary from mild to intolerable but was not related to the presence or absence of a localized AD, nor to stimulation intensity. The occurrence of an intolerable painful sensation led us to im-mediately stop the stimulation, which then only lasted for about 1 s instead of 5 s. Pain disappeared as soon as the stimulation was interrupted except in a few cases where the intense pain was followed by a sore feeling that could last up to 1 min after the end of the stimulation. Painful sensations were located contra-lateral to the stimulation site or bilaterally when midline parts of the body were involved. Whereas limb sensations were elicited across the posterior part of the insular cortex, face, ear and nose painful sensations (the cephalic representation) were mostly elicited in a limited area in the upper anterior part of the posterior section of the insular cortex (Fig. 2B). Qualities of the evoked pain were described as a burning, a stinging or a disabling sensations. Electrical shocks or discharges were also reported (Fig. 2C). Since all elicited sensations comprised a part of the contralateral body and since the anatomical location of the right and left reactive sites within the insular cortex did not differ significantly (Mann–Whitney test; two-tailed P > 0.66), coordinates of insular sites of both hemispheres were pooled to calculate mean coordinates of the insular pain representation. These mean coordinates were x = 31.2 ± 4.9 mm, y = –16.4 ± 6.3 mm, z = 13.2 ± 7.6 mm (Table 1).

Most sites where painful sensations occurred were located in the right hemisphere (Fig. 2A). Although based on a limited number of sites (15), this higher proportion of sites inducing pain is statistically significant (Fisher's exact test; two-tailed P = 0.02). Moreover, the three sites located in the left insular cortex correspond to stimulation sites in three different patients: the first patient was left-handed and had a bilateral language representation, the second one was left-handed with a left language representation and the third one was right-handed with a left language representation. The 12 right insular sites where painful sensations occurred were stimulated in 10 different patients of whom nine were right-handed (eight with a left language representation and one with a bilateral language repre-sentation) and one was left-handed with a bilateral language representation. Therefore, only once was a painful sensation elicited during stimulation of the insular cortex on the side of a hand and language dominant hemisphere.

Non-painful Somaesthetic Sensations Evoked by Insular Cortex Stimulation

Non-painful somaesthetic sensations were elicited in 21 insular sites (21/93; 22.5%) in 16 patients (16/43; 37.2%). Nine sites were located in the left and 12 in the right insular cortex. Three of these 16 patients also reported painful evoked sensations described in the preceding section. Painful and non-painful somaesthetic sensations were evoked on different sites except for one site where a first LFS induced a non-painful cold somaes-thetic sensation in the contralateral arm and a second HFS stimulation induced an extremely painful electric shock in the contralateral half of the body. All non-painful somaesthetic-evoked responses but one were obtained using HFS. Sensations were elicited in the absence of AD in 18 sites (13 patients) and with a localized AD in three sites (in three patients). No diffusing AD was observed during non-painful somaesthetic sensations. Mean intensity of stimulation was 2.4 ± 1.8 mA.

Most reactive sites were located in the posterior part of the insular cortex (Figs 3A and 4A). These non-painful sensations were felt contralateral to the stimulation or bilaterally when midline parts of the body were concerned. They involved the face, the ear and/or the throat when stimulated sites were located in the upper part of the posterior section of the insular cortex and the limbs when sites were distributed in its inferior part (Fig. 3B). They were reported as warmth, cold or tingling sensations. Tingling and warmth sensations were evoked throughout the posterior part of the insular cortex. The only cold sensation was obtained by stimulating the upper posterior part of the insula (Fig. 3C). Since all sensations comprised at least a part of the contralateral body and since anatomical locations of right versus left reactive sites within the insular cortex did not differ significantly (Mann–Whitney test; two-tailed P > 0.23), coordinates of insular sites of both hemispheres were pooled to calculate mean coordinates of the non-painful insular somaes-thetic representation. These mean coordinates were x = 32 ± 3.5 mm, y = –13.1 ± 10.6 mm, z = 7.8 ± 8.3 mm (see Table 2), values that did not differ significantly from those of the insular pain representation (Student's t-test; two-tailed P > 0.05; Fig. 4B).

The proportion of the reactive sites found in the left versus the right hemisphere were not found statistically significant (Fisher's exact test; two-tailed P = 0.8).

Discussion

The major finding of this study is the possibility to induce painful and non-painful somaesthetic sensations by direct stimulation of the posterior part of the human insular cortex. We further demonstrate that both painful and non-painful somaesthetic insular representations overlap and tend to be somatotopically organized. Finally, an unexpected lateralization of painful sensation induced by stimulating the right hemisphere is disclosed.

Insular stimulations in our study were performed in epileptic patients. One could therefore argue that the responses we obtained are related to an abnormal (epileptogenic) cortex, or a cortex that has been submitted to repetitive fits leading to plastic changes (Gloor, 1990; Fish et al., 1993). Nonetheless, and concerning the painful responses, these were not reported in previous studies which were also performed in epileptic patients.

The proximity of the second somatosensory area (SII) with respect to the insula also raises the question as to the specificity of our responses. Even considering the current density spatial distribution during bipolar stimulation (Nathan et al., 1993), the threshold value leading to an effective stimulation of SII would have to be known in order to evaluate a possible involvement of SII by stimulations of the neighbouring insular cortex. The main argument favouring the specificity of our insular stimulations is that painful responses could never be obtained when stimu-lating directly SII using identical stimulation parameters (personal observations). This is coherent with the fact that Penfield and Jasper also never reported such painful responses when stimulating SII (Penfield and Jasper, 1954). Furthermore, the recording of the intracerebral activity during the stimulus paradigm gives us confidence in the specificity of our responses by providing the possibility to detect diffusing AD. It has been demonstrated that no significant cerebral blood flow change occurs during direct electrical cortical stimulation carried out with our paradigm of stimulation in the absence of AD (Kahane et al., 1999). As most of the responses reported in this study were obtained in this condition, one can assume that the spatial resolution of our functional mapping is not blurred by some remote effects of the stimulation on neighbouring structures. This assumption seems also acceptable for responses associated with a localized AD which remained confined to the stimulation site. Conversely, responses associated with a diffusing AD were discarded from our functional mapping study. In summary, according to the above considerations, pain and non-painful somaesthetic evoked responses obtained in this study can be reliably considered as specific of the stimulation of the posterior part of the insular cortex.

Pain Representation

Apart from two patients we already reported (Ostrowsky et al., 2000), truly painful responses have not been previously described during stimulation of any area of the cerebral cortex in humans. This can be attributed to different reasons: (i) studies of insular stimulation in monkeys and dogs were performed under anaesthesia (Kaada et al., 1949; Babkin and Van Buren, 1951; Hoffman and Rasmussen, 1953) thus precluding any study of pain-evoked responses. (ii) In humans, Penfield and co-workers extensively stimulated all cortical areas of the human brain, using surface electrodes, during cortectomies under local anaesthesia (Penfield and Brodley, 1937; Penfield and Jasper, 1954), including the insular cortex (Penfield and Faulk, 1955). However, they only explored the upper posterior part of the insula in rare cases after removal of the fronto-parietal oper-culum. As can be seen in our results, painful sensations could not be elicited from all sites situated in the upper posterior part of the insular cortex and a large number of stimulation sites in this area must therefore be sampled in order to disclose the insular pain representation.

The existence of a pain representation in the posterior insular cortex is coherent with anatomical connectivity and electro-physiological properties of this area in monkeys. Indeed, a major source of insular inputs originates from thalamic nuclei receiving spino-thalamic-tract fibres (STT) which carry nociceptive information (Burton and Jones, 1976; Mufson and Mesulam, 1984; Friedman and Murray, 1986; Apkarian and Hodge, 1989; Hodge and Apkarian, 1990). Furthermore, Burton and Jones showed a preferential thalamic input to the posterior part of the insular cortex from the suprageniculate-limitans nucleus (SG-Li) and the posterior complex (Po) in monkeys (Burton and Jones, 1976). In humans, Craig et al. described a thalamic structure specifically relaying painful and thermal stimuli: the posterior portion of the ventral medial nucleus, i.e. the VMpo (Craig et al., 1994). As the latter nucleus is embedded in the Po, one can assume that its efferent fibres terminate in a similar insular area. These connections have moreover been described by Craig (Craig, 1995), who showed tracing studies indicating the ana-tomical projection of the VMpo to the same part of the insular cortex associated in this study with pain and cold sensations. The ventral posterior inferior nucleus (VPI) is another relevant structure since it contains a high proportion of nociceptive neurons (Apkarian and Shi, 1994) and also projects to the posterior part of the insula (Burton and Jones, 1976; Mufson and Mesulam, 1984). Finally, and in agreement with all these possible inputs, single-unit recordings performed in monkeys have revealed nociceptive responses in neurons of the posterior part of the insular cortex (Robinson and Burton, 1980).

In humans, our results can be compared with previous clinical observations and studies using different techniques: evoked potential, positron emission tomography (PET) and functional magnetic resonance imaging (fMRI). Brain injuries located in the superior/posterior portion of the insula have led to im-pairment in pain perception (Greenspan and Winfield, 1992; Schmahmann and Leifer, 1992; Greenspan et al., 1999). In healthy volunteers, all dipolar source models of the earliest lateralized evoked potential to CO2 laser stimulation include bilateral sources in SII or the insular cortex (Tarkka and Treede, 1993; Bromm and Chen, 1995; Kakigi et al., 1995; Valeriani et al., 2000; Opsommer et al., 2001). Although these results are in rough agreement with ours, further more accurate comparisons are impeded by the limited spatial resolution of this technique.

Most PET studies reveal a bilateral activation of the insular cortex during peripheral painful stimuli. Since most of our responses were contralateral to the cortical stimulation we will only consider contralateral insular PET activation. When the mean coordinates of the insular sites eliciting painful sensations are compared with those obtained in PET studies, they are significantly different in each spatial dimension (Student's t-test; two-tailed P < 0.006). In the x (lateral–medial) and z (vertical) dimensions, the mean PET coordinates depart from ours by 5.8 mm and 6.7 mm, respectively (see Tables 1 and 3, and Fig. 4B). This could be due to the simultaneous activation of SII which is difficult to discriminate from that of the insula using PET technique allowing no more than a 4 mm spatial resolution with the most advanced tomographs. Thus, the significance of the statistical difference obtained for these two dimensions remains highly questionable. In the y (rostro-caudal) dimension, the difference increases to 15.2 mm (see Tables 1 and 3, and Fig. 4B). Even if some of the earliest PET studies were performed with a spatial resolution of ~8 mm, this difference could be considered as truly significant. However, recent PET studies (Craig et al., 1996; Xu et al., 1997; Paulson et al., 1998 ; Casey et al., 2001) are more and more showing two separate foci of maximal activation within the insular cortex: one located in its antero–inferior part, the other in its postero-superior extent (Table 4). Thus, not discriminating these two peaks of activation would lead to a biased mean y value shifted anteriorly with respect to our data. To overcome this bias, one can consider only the mean coordinates of the posterior peak of activation of these recent PET studies (x = 36.2 ± 2.5 mm, y = –13 ± 7 mm, z = 7 ± 7 mm, Table 4). Doing this, the coordinate differences between the posterior peak of PET activation and our sites of stimulations for x, y and z are 5, 3.5 and 6.2 mm, respectively. According to the technical limitations already mentioned, one can consider that these differences are not significant. Furthermore, a recent study (Casey et al., 2001) has demonstrated that the pattern of structures activated by a persistent thermal nociceptive stimulus could evolve in time with respect to the beginning of the stimulation. In the early period (0–60 s), an anterior insular activation was observed (y = 23 mm) while two more caudal peaks of activation appear during following scans (y = 1 mm and –15 mm, respectively). This phenomenon could also explain the discrepancies along the rostro-caudal axis of data reported in previous PET studies. Also worth noting is the comparison made by Casey's group between their results obtained with a conventional protocol in 1994 and those obtained in the late phase of the nociceptive stimulation in their more recent study. In very similar stimulus conditions, a difference of 18 mm in the antero-posterior dimension stands between the two activated peaks in the contralateral insula [see table 3 in Casey et al. (Casey et al., 2001)]. For all the above reasons, one can consider that comparisons between our data and those obtained by the PET technique is at this time very questionable.

Nonetheless, it remains to be determined the functional role of the antero-inferior pain representation disclosed in PET studies which is located in an area that was sampled by our electrodes but in which we could not evoke any pain responses (mean coordinates of the antero-inferior peak in recent PET studies showing two peaks of activation: x = 34.8 ± 3 mm, y = 9.2 ± 9.8 mm, z = 3.4 ± 5.8 mm, Table 4). The presence of such an antero-inferior peak of activation is not surprising per se since a projection from the VPI to the anterior part of the insula has been described (Burton and Jones, 1976). One can thus postulate that although both anterior and posterior parts of the insular cortex receive nociceptive inputs via the VPI, the anterior part is not involved in exteroceptive pain perception itself but in other related functions. This hypothesis would fit with the viscerosensitive, visceromotor and autonomic reactions previously observed when stimulating this particular anterior insular area (Penfield and Faulk, 1955; Oppenheimer et al., 1992; Ostrowsky et al., 2000).

Finally, the more recent introduction of fMRI with its improved spatial resolution could have led to a more precise topographic location of activated brain structures. To this date, only a few studies in the field of pain processing are available (Antognini et al., 1997; Davis et al., 1998a,b; Disbrow et al., 1998; Oshiro et al., 1998; Bushnell et al., 1999; Gelnar et al., 1999; Ploghaus et al., 1999; Kwan et al., 2000; Moulton et al., 2000). Three of them deserve special attention since they directly addressed the question we are interested in. Gelnar et al. who, unfortunately, did not explore the anterior part of the insula, reported an activated peak in its posterior part, mean coordinates of which (x = 42 mm, y = –10 mm, z = 14 mm) are in fairly good agreement with our data (Gelnar et al., 1999). Results obtained by Davis et al. are more puzzling, showing a large interindividual variability in activated insular areas which are distributed throughout its rostro-caudal axis with some preponderance in its anterior part (Davis et al., 1998b). A similar result is obtained by Ploghaus et al. who, in addition, found a distinct peak of activation in the anterior insular cortex spe-cifically related to the expectation of painful stimuli (Ploghaus et al., 1999). Together with our data, this last observation tends to attribute to the anterior insula a functional role not directly related to pain perception.

Non-painful Sensation Representation

It is generally accepted that tactile sensation is supported by the medial lemniscal system. However, according to its known connectivity, the ventral posterior lateral nucleus which receives its major input from the medial lemniscal tract does not project to the insular cortex in monkeys (Burton and Jones, 1976). The sensations we obtained in the posterior part of the insular cortex can nonetheless be divided into two types. Thermal ones (mainly warm) can be explained in the light of recent studies describing a STT component coding specifically not only for painful but also for thermal stimuli. It originates in the superficial lamina I of the spinal dorsal horn and terminates in a subdivision of the Po, the VMpo (Craig et al., 1994; Dostrovsky and Craig, 1996; Blomqvist et al., 2000). The projections of the latter nucleus on the posterior part of the insula (Burton and Jones, 1976) should thus ensure a thermal representation within the insula which can be activated by direct electrical stimulation. Similar con-siderations can account for the tingling sensations we observed. It has been demonstrated that wide dynamic range neurons located deeper in the dorsal horn (laminae III–IV) are able to carry tactile information beside pain related signals (Zhang et al., 1991). The STT recipient nuclei (Po, SG-Li, VPI) are thus able to supply the insula with non-nociceptive somaesthetic information through the limited but consistent proportion of wide-dynamic neurons they contain (Apkarian and Shi, 1994). This is supported by single-cell studies in monkeys showing a high percentage (nearly two-thirds) of neurons in the posterior insula responding to non-nociceptive stimulations (Robinson and Burton, 1980; Schneider et al., 1993). The above consider-ations refer to the most direct route for non-painful stimuli to the insular cortex. However we cannot exclude that somaesthetic inputs reach the insular cortex via area SII, which sends direct projections to the posterior insular cortex (Mufson and Mesulam, 1982) and responds to non-painful stimulation of the median nerve in humans (Frot and Mauguière, 1999).

The non-painful representation we obtained can also be compared in terms of location with previous studies in humans using different techniques. Dipolar source modelling of somato-sensory evoked potentials in humans have revealed dipoles, deep in the upper bank of the Sylvian fissure, posterior to the central fissure, compatible with generators in SII or the insular cortex (Baumgartner et al., 1991; Hari et al., 1993; Forss et al., 1996; Mauguière et al., 1997). Unfortunately, the limited spatial resolu-tion of this technique does not allow more precise comparisons. Cerebral blood flow increases in the insular cortex contralateral to innocuous somaesthetic stimulation have been reported in PET studies (Burton et al., 1993; Coghill et al., 1994; Craig et al., 1996, 2000; Derbyshire et al., 1997; Iadarola et al., 1998) but with mean coordinates of activation peaks significantly more rostral (16.9 mm; Student's t-test: two-tailed P = 0.0007) and lateral (4 mm; Student's t-test: two-tailed P = 0.003) than mean coordinates of sites where non-painful somaesthetic sensations were evoked in our patients (see Tables 2 and 5, and Fig. 4B). Although statistically significant, these differences remain highly questionable for the technical reasons already evoked for the insular pain representation. More puzzling is the fact that the non-painful somaesthetic area activated in PET studies was sampled in our study and no corresponding sensations were herein evoked. Looking carefully at the data previously obtained with the PET technique [see fig. 1 of Craig et al. (Craig et al., 1996) for 20 and 40°C thermal stimuli], it seems that the global insular activation can be constituted of two distinct peaks of activation, similar to the ones induced by painful stimuli. However, the same group, in a later study carried out in very similar stimulation conditions (Craig et al., 2000), did not confirm this pattern at least for the 20°C stimulus, mentioning only a posterior insular peak of activation, coordinates of which are in perfect agreement with our results. On the ipsilateral side of stimulation, an anterior peak of activation that was not present in the earlier study (Craig et al., 1996) was this time observed. Facing such data variability implies that we should show caution in more precise comparisons with our results. Finally, fMRI studies (Davis et al., 1998b; Gelnar et al., 1999; Disbrow et al., 2000; Sawamoto et al., 2000) do not allow direct comparisons since none of them gives the mean coordinates of an insular area activated by non-painful somaesthetic stimuli. Nevertheless, it is worth noting the distribution of responses following tactile stimuli reported by Davis et al. and found selectively in the posterior part of the insula, a location which tightly fits with our results (Davis et al., 1998b). However, in the same study, innocuous temperature-related activations in the insula were not segregated consistently in any particular region.

Somatotopic Organization of the Responses

Another original contribution of this study concerns a possible somatotopic organization of pain representation in the human insula. To our knowledge, this question has not been addressed before in single-cell studies in monkeys (Robinson and Burton, 1980), and only once in a PET study which failed to demonstrate distinct representations for hand and foot painful stimulation (Xu et al., 1997). Concerning the non-painful somaesthetic representation, no precise topography within the insular cortex can be found in Wieser's data (Wieser, 1983) whereas Penfield and Faulk (Penfield and Faulk, 1955) already noticed some somatotopic organization within the insular cortex with responses located to the limbs (and usually better lateralized) when stimulating insular sites located posteriorly and to the midline structures (lips, tongue, mouth and throat) when stimulating insular sites located more anteriorly. A similar trend is observed in our study.

Overlap Between Painful and Non-painful Representations

As can be seen in Figure 4B, insular areas where painful and non-painful somaesthetic responses were obtained, overlap. This is in agreement with the connectivity described above allowing somaesthetic inputs, whether painful or not, to converge to the posterior part of the insular cortex via direct STT-thalamo-cortical projections. At the cellular level, Robinson and Burton showed that the same insular neuron in the posterior insula was able to respond to both painful and non-painful somaesthetic stimuli (Robinson and Burton, 1980). The receptive fields of these insular neurons were found to be large, in agreement with our observations that responses obtained when stimulating the posterior part of the insular cortex included extensive parts of the body. This is an additional factor which could participate in the overlap between the representation of the two somaesthetic modalities. Finally, PET data showing similar activated regions in the insular cortex during painful and non-painful somaesthetic stimuli (Coghill et al., 1994; Craig et al., 1996; Derbyshire et al., 1997; Iadarola et al., 1998) are coherent with this overlapping location of the pain and non-painful somaesthetic representa-tions within the human insular cortex.

Lateralization of the Painful Representation

Whereas non-painful somaesthetic sensations were evoked in similar proportion by stimulation of right or left insular sites, pain was mostly elicited when stimulating the insula of the non-dominant language and/or manual hemisphere, i.e. the right insular cortex. Previous authors have also mentioned some asym-metrical features concerning the human insula. Thus, Clarke found that the volume of the left insula was larger than that of the right (Clarke, 1896). Later, Oppenheimer et al. demonstrated, a right-sided dominance of sympathetic effects (increased cardiac frequency and blood pressure) after insular cortex stimulation as opposed to a left-sided dominance for parasympathetic effects (Oppenheimer et al., 1992). The right-sided dominance of sites where painful sensations were evoked in our study would be therefore physiologically coherent, a sympathetic tone increase being expected during painful sensations. A right hemispheric dominance has also been described in PET studies for the affective–emotional component of pain with a preferential increased cerebral blood flow in the right anterior cingulate gyrus (Hsieh et al., 1995a, 1996, 1999). Furthermore, studies on cerebral lateralization have emphasized the fact that the right hemisphere is more involved in spatial as well as in atten-tional process and many aspects of emotion (Mesulam, 1981; Silberman and Weingartner, 1986; Gainotti, 1996). Attentional process and visuo-spatial integration for directing the indi-vidual's attention towards threatening external stimuli are of great importance for survival. Finally, the precocious and faster development of the right hemisphere during fetal life accounts for what is referred to as the right hemisphere ‘conservatism’ (Geschwind and Galaburda, 1985). Our data are in line with this theory, suggesting that pain integration, which is a determinant factor in species survival, could be preferentially dependent on the right hemisphere.

Table 1

Stereotactic coordinates of the 15 insular sites where painful sensations were elicited in absence of AD or with a localized AD in our study (*site located in left hemisphere)

 x y z 
 26 –22 
 30  –8 17 
 27 –15 20 
 40  –5 –1 
 31  –7 14 
 30 –24 
 28 –16 22 
 40 –25 22 
 32 –15 15 
 31* –14 
 29 –25 
 30 –20 18 
 27* –20 20 
 40* –15 18 
 27 –16 14 
Mean 31.2 –16.4 13.2 
SD  4.9  6.3  7.6 
 x y z 
 26 –22 
 30  –8 17 
 27 –15 20 
 40  –5 –1 
 31  –7 14 
 30 –24 
 28 –16 22 
 40 –25 22 
 32 –15 15 
 31* –14 
 29 –25 
 30 –20 18 
 27* –20 20 
 40* –15 18 
 27 –16 14 
Mean 31.2 –16.4 13.2 
SD  4.9  6.3  7.6 
Table 2

Stereotactic coordinates of the 21 sites where non-painful somaesthetic sensations were elicited in absence of AD or with a localized AD in our study (*site located in left hemisphere)

 x y z 
 40 –25 22 
 31 
 33* 10 
 36* –15 
 33  –9 
 32  –9 23 
 32* –27 
 30*  –9 16 
 34  –9 
 33 –15 14 
 36 –6 
 31 –25 
 33*  –9 
 29* –18 
 29  –9 13 
 37* –4 
 27* –27 
 31 –20 10 
 28 –18 
 25 –24 23 
 32* –23 
Mean 32 –13.1  7.8 
SD  3.5  10.6  8.3 
 x y z 
 40 –25 22 
 31 
 33* 10 
 36* –15 
 33  –9 
 32  –9 23 
 32* –27 
 30*  –9 16 
 34  –9 
 33 –15 14 
 36 –6 
 31 –25 
 33*  –9 
 29* –18 
 29  –9 13 
 37* –4 
 27* –27 
 31 –20 10 
 28 –18 
 25 –24 23 
 32* –23 
Mean 32 –13.1  7.8 
SD  3.5  10.6  8.3 
Table 3

Stereotactic coordinates of mean peaks of activation found in the insular cortex contralateral to the noxious peripheral painful stimuli described in previous PET studies performed in humans (*site located in left hemisphere)

Reference x y z 
The mean y coordinate of these PET studies (–1.2 mm) departs from ours (–16.4 mm) by 15.2 mm. 
(Caseyet al., 199437 –19 14 
(Coghill et al., 199438 
(Hsieh et al.,1995b33* 
(Casey et al.,199637 –19 14 
(Vogt et al.,199636 
(Craig et al.,199637*  –9 
(Craig et al., 1996 37* 
(Rainville et al., 199742  –6 12 
(Andersson et al.,199736* 
(Svenson et al.,1997) 44 
(Xu et al., 199736  –6 
(Xu et al., 1997 32  16 –4 
(Iadarola et al., 199838 
(Derbyshire et al., 199730*  –2 
(May et al., 199836* 
(Svensson et al., 199838 11 
(Svensson et al., 1998 43 
(Svensson et al., 199846 –2 
(Derbyshire and Jones, 199840* –20 16 
(Paulson et al., 199839 –22 16 
(Paulson et al., 199833 
(Tölle et al., 199935*  –6 
(Peyron et al., 199934*  –2 
(Casey et al., 200133  23 
(Casey et al., 200133 –15 
(Casey et al., 200139 11 
Mean 37  –1.2  6.5 
SD  3.8  11.1  5.5 
Reference x y z 
The mean y coordinate of these PET studies (–1.2 mm) departs from ours (–16.4 mm) by 15.2 mm. 
(Caseyet al., 199437 –19 14 
(Coghill et al., 199438 
(Hsieh et al.,1995b33* 
(Casey et al.,199637 –19 14 
(Vogt et al.,199636 
(Craig et al.,199637*  –9 
(Craig et al., 1996 37* 
(Rainville et al., 199742  –6 12 
(Andersson et al.,199736* 
(Svenson et al.,1997) 44 
(Xu et al., 199736  –6 
(Xu et al., 1997 32  16 –4 
(Iadarola et al., 199838 
(Derbyshire et al., 199730*  –2 
(May et al., 199836* 
(Svensson et al., 199838 11 
(Svensson et al., 1998 43 
(Svensson et al., 199846 –2 
(Derbyshire and Jones, 199840* –20 16 
(Paulson et al., 199839 –22 16 
(Paulson et al., 199833 
(Tölle et al., 199935*  –6 
(Peyron et al., 199934*  –2 
(Casey et al., 200133  23 
(Casey et al., 200133 –15 
(Casey et al., 200139 11 
Mean 37  –1.2  6.5 
SD  3.8  11.1  5.5 
Table 4

Separate stereotactic coordinates of the anterior and posterior peaks of activation reported in recent PET studies showing two distinct peaks of activation within the insular cortex contralateral to the peripheral noxious stimulation (*site located in left hemisphere)

Reference x y z 
Anterior peaks 
(Craig et al., 199637* 
(Xu et al., 199732  16 –4 
(Paulson et al., 199833 
(Casey et al., 200133  23 
(Casey et al., 200139 11 
Mean 34.8  9.2  3.4 
SD  9.8  5.8 
Posterior peaks 
(Craig et al., 199637*  –9 
(Xu et al., 199736  –6 
(Paulson et al., 199839 –22 16 
(Casey et al., 200133 –15 
Mean 36.2 –13 
SD  2.5 
Reference x y z 
Anterior peaks 
(Craig et al., 199637* 
(Xu et al., 199732  16 –4 
(Paulson et al., 199833 
(Casey et al., 200133  23 
(Casey et al., 200139 11 
Mean 34.8  9.2  3.4 
SD  9.8  5.8 
Posterior peaks 
(Craig et al., 199637*  –9 
(Xu et al., 199736  –6 
(Paulson et al., 199839 –22 16 
(Casey et al., 200133 –15 
Mean 36.2 –13 
SD  2.5 
Table 5

Stereotactic coordinates of mean peaks of activation in the insular cortex contralateral to peripheral non-painful stimuli disclosed in previous PET studies in humans (*site located in left hemisphere)

Reference x y z 
The mean y coordinate of these PET studies (3.8 mm) departs from ours (–13.1 mm) by 16.9 mm 
(Burton et al., 199339.9  5.9 
(Burton et al., 1993 30.4*  25.6  2.2 
(Burton et al., 1993 39.2  16.1  6.2 
(Burton et al., 199338.2*  7.1 11.1 
(Coghill et al., 199438 –13 –5 
(Craig et al., 199637*  –9 
(Craig et al., 199637* 
(Derbyshire et al., 199730*  –2 
(Iadarola et al., 199838  20 –8 
(Iadarola et al., 199834 –8 
(Craig et al., 200036* –22 24 
Mean 36.1  3.8  3.5 
SD  3.3  14.4  9.1 
Reference x y z 
The mean y coordinate of these PET studies (3.8 mm) departs from ours (–13.1 mm) by 16.9 mm 
(Burton et al., 199339.9  5.9 
(Burton et al., 1993 30.4*  25.6  2.2 
(Burton et al., 1993 39.2  16.1  6.2 
(Burton et al., 199338.2*  7.1 11.1 
(Coghill et al., 199438 –13 –5 
(Craig et al., 199637*  –9 
(Craig et al., 199637* 
(Derbyshire et al., 199730*  –2 
(Iadarola et al., 199838  20 –8 
(Iadarola et al., 199834 –8 
(Craig et al., 200036* –22 24 
Mean 36.1  3.8  3.5 
SD  3.3  14.4  9.1 
Figure 1.

A) Immediate post-implantation frontal X-ray showing the trajectory of the electrodes superimposed on the corresponding coronal section of the MRI. As can be seen, the first and second deepest contacts of the trans-opercular electrodes A and B (indicated by the arrows) are in the depth of the insular cortex. (B) A plot of all (93) stimulated insular sites on a sagittal section (x = 33 mm) of the insular cortex of the MNI standard brain. (C) A plot of all (93) stimulated insular sites on the corresponding sagittal section of the insular cortex of one of our patient's cerebral MRI, showing a similar sampling. Circles represent sites of stimulation, thick contours indicate sites where stimulation evoked a localized AD. AC–PC: horizontal plane passing through the anterior and posterior commissure; VAC: frontal plane passing trough the anterior commissure, perpendicular to the AC–PC plane; VPC: frontal plane passing through the posterior commissure, perpendicular to the AC–PC plane. The white line is the contour of the stimulated insular area. Open circles: insular sites in the right hemisphere; hatched circles: insular sites in the left hemisphere.

Figure 1.

A) Immediate post-implantation frontal X-ray showing the trajectory of the electrodes superimposed on the corresponding coronal section of the MRI. As can be seen, the first and second deepest contacts of the trans-opercular electrodes A and B (indicated by the arrows) are in the depth of the insular cortex. (B) A plot of all (93) stimulated insular sites on a sagittal section (x = 33 mm) of the insular cortex of the MNI standard brain. (C) A plot of all (93) stimulated insular sites on the corresponding sagittal section of the insular cortex of one of our patient's cerebral MRI, showing a similar sampling. Circles represent sites of stimulation, thick contours indicate sites where stimulation evoked a localized AD. AC–PC: horizontal plane passing through the anterior and posterior commissure; VAC: frontal plane passing trough the anterior commissure, perpendicular to the AC–PC plane; VPC: frontal plane passing through the posterior commissure, perpendicular to the AC–PC plane. The white line is the contour of the stimulated insular area. Open circles: insular sites in the right hemisphere; hatched circles: insular sites in the left hemisphere.

Figure 2.

A plot of sites where painful sensations occurred in 14 patients; (A) most sites are located in the upper posterior part of the insular cortex, in the right hemisphere; (B) somatotopic organization of the responses: the cephalic representation is more rostral than that of the limb and hand representation; (C) location of different pain qualities.

Figure 2.

A plot of sites where painful sensations occurred in 14 patients; (A) most sites are located in the upper posterior part of the insular cortex, in the right hemisphere; (B) somatotopic organization of the responses: the cephalic representation is more rostral than that of the limb and hand representation; (C) location of different pain qualities.

Figure 3.

Plotting of sites where non-painful sensations occurred in 16 patients; (A) most sites are located in the posterior part of the insular cortex, in both hemisphere; (B) somatotopic organization of the responses: the cephalic representation is similar to that of pain representation whereas limb and hand representations are more scattered across the posterior part of the insular cortex; (C) location of warmth, cold and tingling sensations.

Figure 3.

Plotting of sites where non-painful sensations occurred in 16 patients; (A) most sites are located in the posterior part of the insular cortex, in both hemisphere; (B) somatotopic organization of the responses: the cephalic representation is similar to that of pain representation whereas limb and hand representations are more scattered across the posterior part of the insular cortex; (C) location of warmth, cold and tingling sensations.

Figure 4.

(A) Projections of all insular stimulation sites on selected coronal sections of the MNI standard brain. The x-axis 5.5 mm dimension of each ellipse has been traced between the two contacts used for stimulation using as limits the inner side of the deepest insular contact and the outer side of the neighbouring contact. This representation illustrates the coronal projection of the estimated stimulation current dipole. Even using an average standard brain representation, some of these ellipses appear located deeper, in the external capsule, or overlap with the external capsule. This is due to rather large interindividual anatomical differences and the fact that some of the insular stimulation sites are located on coronal sections adjacent to those illustrated. However, note that for each stimulated site, the location of the corresponding two contacts of the trans-opercular electrode was checked to be strictly located in the thickness of the insular cortex on superimposed MRI and X-ray images in each patient. Non-painful sites are in yellow, painful sites are in red. (B) Respective location of painful and non-painful representations within the insula defined according to direct electrical stimulations (this study) or resulting from previous PET studies (see Tables 3 and 5). Centres of these areas correspond to mean y and z coordinates and their extents are defined from the associated y and z standard deviations.

(A) Projections of all insular stimulation sites on selected coronal sections of the MNI standard brain. The x-axis 5.5 mm dimension of each ellipse has been traced between the two contacts used for stimulation using as limits the inner side of the deepest insular contact and the outer side of the neighbouring contact. This representation illustrates the coronal projection of the estimated stimulation current dipole. Even using an average standard brain representation, some of these ellipses appear located deeper, in the external capsule, or overlap with the external capsule. This is due to rather large interindividual anatomical differences and the fact that some of the insular stimulation sites are located on coronal sections adjacent to those illustrated. However, note that for each stimulated site, the location of the corresponding two contacts of the trans-opercular electrode was checked to be strictly located in the thickness of the insular cortex on superimposed MRI and X-ray images in each patient. Non-painful sites are in yellow, painful sites are in red. (B) Respective location of painful and non-painful representations within the insula defined according to direct electrical stimulations (this study) or resulting from previous PET studies (see Tables 3 and 5). Centres of these areas correspond to mean y and z coordinates and their extents are defined from the associated y and z standard deviations.

Grateful thanks to all staff members of the video-EEG and SEEG unit for their constant help and patience, especially during stimulation sessions.

References

Andersson JL, Lilja A, Hartvig P, Langstrom B, Gordh T, Handwerker H, Torebjork E (
1997
) Somatotopic organization along the central sulcus, for pain localization in humans, as revealed by positron emission tomography.
Exp Brain Res
 
117
:
192
–199.
Antognini JF, Buonocore MH, Disbrow EA, Carstens E (
1997
) Isoflurane anesthesia blunts cerebral responses to noxious and innocuous stimuli: a fMRI study.
Life Sci
 
61
:
349
–354.
Apkarian AV, Hodge CJ (
1989
) Primate spinothalamic pathways: III. Thalamic terminations of the dorsolateral and ventral spinothalamic pathways.
J Comp Neurol
 
288
:
493
–511.
Apkarian AV, Shi T (
1994
) Squirrel monkey lateral thalamus. I. Somatic nociresponsive neurons and their relation to spinothalamic terminals.
J Neurosci
 
14
:
6779
–6795.
Babb TL, Mariani E, Seidner KA, Mutafyan G, Halgren E, Wilson CL, Crandall PH (
1980
) A circuit for safe diagnostic electrical stimulation of the human brain.
Neurol Res
 
2
:
181
–197.
Babkin B, Van Buren J (
1951
) Mechanism and cortical representation of the feeding pattern.
Arch Neurol Psychiat
 
66
:
1
–19.
Baumgartner C, Doppelbauer A, Deecke L, Barth DS, Zeitlhofer J, Lindinger G, Sutherling WW (
1991
) Neuromagnetic investigation of somatotopy of human hand somatosensory cortex.
Exp Brain Res
 
87
:
641
–648.
Blomqvist A, Zhang ET, Craig AD (
2000
) Cytoarchitectonic and immunohistochemical characterization of a specific pain and temperature relay, the posterior portion of the ventral medial nucleus, in the human thalamus.
Brain
 
123
:
601
–619.
Bromm B, Chen AC (
1995
) Brain electrical source analysis of laser evoked potentials in response to painful trigeminal nerve stimulation.
Electroencephalogr Clin Neurophysiol
 
95
:
14
–26.
Burton H, Jones EG (
1976
) The posterior thalamic region and its cortical projection in New World and Old World monkeys.
J Comp Neurol
 
168
:
249
–301.
Burton H, Videen TO, Raichle ME (
1993
) Tactile-vibration-activated foci in insular and parietal-opercular cortex studied with positron emission tomography: mapping the second somatosensory area in humans.
Somatosens Mot Res
 
10
:
297
–308.
Bushnell MC, Duncan GH, Hofbauer RK, Ha B, Chen JI, Carrier B (
1999
) Pain perception: is there a role for primary somatosensory cortex?
Proc Natl Acad Sci USA
 
96
:
7705
–7709.
Casey KL, Minoshima S, Berger KL, Koeppe RA, Morrow TJ, Frey KA (
1994
) Positron emission tomographic analysis of cerebral struc-tures activated specifically by repetitive noxious heat stimuli.
J Neurophysiol
 
71
:
802
–807.
Casey KL, Minoshima S, Morrow TJ, Koeppe RA (
1996
) Comparison of human cerebral activation pattern during cutaneous warmth, heat pain, and deep cold pain.
J Neurophysiol
 
76
:
571
–581.
Casey KL, Morrow TJ, Lorenz J, Minoshima S (
2001
) Temporal and spatial dynamics of human forebrain activity during heat pain: analysis by positron emission tomography.
J Neurophysiol
 
85
:
951
–959.
Clark TE (
1896
) The comparative anatomy of the insula.
J Comp Neurol
 
6
:
59
–100.
Coghill RC, Talbot JD, Evans AC, Meyer E, Gjedde A, Bushnell MC, Duncan GH (
1994
) Distributed processing of pain and vibration by the human brain.
J Neurosci
 
14
:
4095
–4108.
Craig AD (1995) In: Forebrain areas involved in pain processing (Besson JM, Guilbaud G, Ollat H, eds), pp. 13–26. Paris: John Libbey.
Craig AD, Bushnell MC, Zhang ET, Blomqvist A (
1994
) A thalamic nucleus specific for pain and temperature sensation.
Nature
 
372
:
770
–773.
Craig AD, Reiman EM, Evans A, Bushnell MC (
1996
) Functional imaging of an illusion of pain.
Nature
 
384
:
258
–260.
Craig AD, Chen K, Bandy D, Reiman EM (
2000
) Thermosensory activation of insular cortex.
Nat Neurosci
 
3
:
184
–190.
Davis KD, Kwan CL, Crawley AP, Mikulis DJ (
1998
) Event-related fMRI of pain: entering a new era in imaging pain.
NeuroReport
 
9
:
3019
–3023.
Davis KD, Kwan CL, Crawley AP, Mikulis DJ (
1998
) Functional MRI study of thalamic and cortical activations evoked by cutaneous heat, cold, and tactile stimuli.
J Neurophysiol
 
80
:
1533
–1546.
Derbyshire SW, Jones AK (
1998
) Cerebral responses to a continual tonic pain stimulus measured using positron emission tomography.
Pain
 
76
:
127
–135.
Derbyshire SW, Jones AK, Gyulai F, Clark S, Townsend D, Firestone LL (
1997
) Pain processing during three levels of noxious stimulation produces differential patterns of central activity.
Pain
 
73
:
431
–445.
Disbrow E, Buonocore M, Antognini J, Carstens E, Rowley HA (
1998
) Somatosensory cortex: a comparison of the response to noxious thermal, mechanical, and electrical stimuli using functional magnetic resonance imaging.
Hum Brain Mapp
 
6
:
150
–159.
Disbrow E, Roberts T, Krubitzer L (
2000
) Somatotopic organization of cortical fields in the lateral sulcus of Homo sapiens: evidence for SII and PV.
J Comp Neurol
 
418
:
1
–21.
Dostrovsky JO, Craig AD (
1996
) Cooling-specific spinothalamic neurons in the monkey.
J Neurophysiol
 
76
:
3656
–3665.
Fish DR, Gloor P, Quesney FL, Olivier A (
1993
) Clinical responses to electrical brain stimulation of the temporal and frontal lobes in patients with epilepsy. Pathophysiological implications.
Brain
 
116
:
397
–414.
Forss N, Merlet I, Vanni S, Hamalainen M, Mauguière F, Hari R (
1996
) Activation of human mesial cortex during somatosensory target detection task.
Brain Res
 
734
:
229
–235.
Friedman DP, Murray EA (
1986
) Thalamic connectivity of the second somatosensory area and neighboring somatosensory fields of the lateral sulcus of the macaque.
J Comp Neurol
 
252
:
348
–373.
Frot M, Mauguière F (
1999
) Timing and spatial distribution of somato-sensory responses recorded in the upper bank of the sylvian fissure (SII area) in humans.
Cereb Cortex
 
9
:
854
–863.
Gainotti G (
1996
) Lateralization of brain mechanisms underlying automatic and controlled forms of spatial orienting of attention.
Neurosci Biobehav Rev
 
20
:
617
–622.
Gelnar PA, Krauss BR, Sheehe PR, Szeverenyi NM, Apkarian AV (
1999
) A comparative fMRI study of cortical representations for thermal painful, vibrotactile, and motor performance tasks.
NeuroImage
 
10
:
460
–482.
Geschwind N, Galaburda AM (
1985
) Cerebral lateralization. Biological mechanisms, associations, and pathology: I. A hypothesis and a program for research.
Arch Neurol
 
42
:
428
–459.
Gloor P (
1990
) Experiential phenomena of temporal lobe epilepsy. Facts and hypotheses.
Brain
 
113
:
1673
–1694.
Gordon B, Lesser RP, Rance NE, Hart J, Jr, Webber R, Uematsu S, Fisher RS (
1990
) Parameters for direct cortical electrical stimulation in the human: histopathologic confirmation.
Electroencephalogr Clin Neurophysiol
 
75
:
371
–377.
Greenspan JD, Winfield JA (
1992
) Reversible pain and tactile deficits associated with a cerebral tumor compressing the posterior insula and parietal operculum.
Pain
 
50
:
29
–39.
Greenspan JD, Lee RR, Lenz FA (
1999
) Pain sensitivity alterations as a function of lesion location in the parasylvian cortex.
Pain
 
81
:
273
–282.
Hari R, Karhu J, Hamalainen M, Knuutila J, Salonen O, Sams M, Vilkman V (
1993
) Functional organization of the human first and second somatosensory cortices: a neuromagnetic study.
Eur J Neurosci
 
5
:
724
–734.
Hodge CJ, Apkarian AV (
1990
) The spinothalamic tract.
Crit Rev Neuro-biol
 
5
:
369
–397.
Hoffman B, Rasmussen T (
1953
) Stimulation studies of insular cortex in Macaca mulatta.
J Neurophysiol
 
16
:
343
–351.
Hsieh JC, Belfrage M, Stone-Elander S, Hansson P, Ingvar M (
1995
) Central representation of chronic ongoing neuropathic pain studied by positron emission tomography.
Pain
 
63
:
225
–236.
Hsieh JC, Stahle-Backdahl M, Hagermark O, Stone-Elander S, Rosenquist G, Ingvar M (
1995
) Traumatic nociceptive pain activates the hypothalamus and the periaqueductal gray: a positron emission tomography study.
Pain
 
64
:
303
–314.
Hsieh JC, Hannerz J, Ingvar M (
1996
) Right-lateralised central processing for pain of nitroglycerin-induced cluster headache.
Pain
 
67
:
59
–68.
Hsieh JC, Meyerson BA, Ingvar M (
1999
) PET study on central processing of pain in trigeminal neuropathy.
Eur J Pain
 
3
:
51
–65.
Iadarola MJ, Berman KF, Zeffiro TA, Byas-Smith MG, Gracely RH, Max MB, Bennett GJ (
1998
) Neural activation during acute capsaicin-evoked pain and allodynia assessed with PET.
Brain
 
121
:
931
–947.
Isnard J, Guenot M, Ostrowsky K, Sindou M, Mauguière F (
2000
) The role of the insular cortex in temporal lobe epilepsy.
Ann Neurol
 
48
:
614
–623.
Kaada B, Pribam K, Epstein J (
1949
) Respiratory and vascular responses in monkeys from temporal pole, insula, orbital surface and cingulate gyrus.
J Neurophysiol
 
12
:
347
–356.
Kahane P, Merlet I, Gregoire MC, Munari C, Perret J, Mauguière F (
1999
) An H(2) (15)O-PET study of cerebral blood flow changes during focal epileptic discharges induced by intracerebral electrical stimulation.
Brain
 
122
:
1851
–1865.
Kakigi R, Koyama S, Hoshiyama M, Kitamura Y, Shimojo M, Watanabe S (
1995
) Pain-related magnetic fields following painful CO2 laser stimulation in man.
Neurosci Lett
 
192
:
45
–48.
Kwan CL, Crawley AP, Mikulis DJ, Davis KD (
2000
) An fMRI study of the anterior cingulate cortex and surrounding medial wall activations evoked by noxious cutaneous heat and cold stimuli.
Pain
 
85
:
359
–374.
Mauguière F, Merlet I, Forss N, Vanni S, Jousmaki V, Adeleine P, Hari R (
1997
) Activation of a distributed somatosensory cortical network in the human brain. A dipole modelling study of magnetic fields evoked by median nerve stimulation. Part I: Location and activation timing of SEF sources.
Electroencephalogr Clin Neurophysiol
 
104
:
281
–289.
May A, Kaube H, Buchel C, Eichten C, Rijntjes M, Juptner M, Weiller C, Diener HC (
1998
) Experimental cranial pain elicited by capsaicin: a PET study.
Pain
 
74
:
61
–66.
Mesulam MM (
1981
) A cortical network for directed attention and unilateral neglect.
Ann Neurol
 
10
:
309
–325.
Moulton EA, Gullapalli RP, Emge DK, Gibbs KL, Greenspan JD (
2000
) Encoding of mechanical pain intensity within multiple regions of the posterior parietal operculum.
Soc Neurosci Abstr
 
26
:
440
.
Mufson EJ, Mesulam MM (
1982
) Insula of the old world monkey. II: Afferent cortical input and comments on the claustrum.
J Comp Neurol
 
212
:
23
–37.
Mufson EJ, Mesulam MM (
1984
) Thalamic connections of the insula in the rhesus monkey and comments on the paralimbic connectivity of the medial pulvinar nucleus.
J Comp Neurol
 
227
:
109
–120.
Nathan SS, Sinha SR, Gordon B, Lesser RP, Thakor NV (
1993
) Deter-mination of current density distributions generated by electrical stimulation of the human cerebral cortex.
Electroencephalogr Clin Neurophysiol
 
86
:
183
–192.
Oppenheimer SM, Gelb A, Girvin JP, Hachinski VC (
1992
) Cardiovascular effects of human insular cortex stimulation.
Neurology
 
42
:
1727
–1732.
Opsommer E, Weiss T, Plaghki L, Miltner WHR (
2001
) Dipole analysis of ultralate (C-fibers) evoked potentials after laser stimulation of tiny cutaneous surface areas in humans.
Neurosci Lett
 
298
:
41
–44.
Oshiro Y, Fuijita N, Tanaka H, Hirabuki N, Nakamura H, Yoshiya I (
1998
) Functional mapping of pain-related activation with echo-planar MRI: significance of the SII-insular region.
NeuroReport
 
9
:
2285
–2289.
Ostrowsky K, Isnard J, Ryvlin P, Guenot M, Fisher C, Mauguière F (
2000
) Functional mapping of the insular cortex: clinical implication in temporal lobe epilepsy.
Epilepsia
 
41
:
681
–686.
Paulson PE, Minoshima S, Morrow TJ, Casey KL (
1998
) Gender differences in pain perception and patterns of cerebral activation during noxious heat stimulation in humans.
Pain
 
76
:
223
–229.
Penfield W, Brodley E (
1937
) Somatic motor and sensory representation in the cerebral cortex of man as studied by electrical stimulation.
Brain
 
60
:
389
–443.
Penfield W, Faulk M (
1955
) The insula, further observation on its function.
Brain
 
78
:
445
–470.
Penfield W, Jasper H (
1954
) Epilepsy and the functional anatomy of the human brain.
Boston, MA: Little Brown
 .
Peyron R, Garcia-Larrea L, Gregoire MC, Costes N, Convers P, Lavenne F, Mauguière F, Michel D, Laurent B (
1999
) Haemodynamic brain responses to acute pain in humans: sensory and attentional networks.
Brain
 
122
:
1765
–1780.
Peyron R, Laurent B, Garcia-Larrea L (
2000
) Functional imaging of brain responses to pain. A review and meta-analysis (2000).
Neurophysiol Clin
 
30
:
263
–288.
Ploghaus A, Tracey I, Gati JS, Clare S, Menon RS, Matthews PM, Rawlins JN (
1999
) Dissociating pain from its anticipation in the human brain.
Science
 
284
:
1979
–1981.
Rainville P, Duncan GH, Price DD, Carrier B, Bushnell MC (
1997
) Pain affect encoded in human anterior cingulate but not somatosensory cortex.
Science
 
277
:
968
–971.
Robinson CJ, Burton H (
1980
) Somatic submodality distribution within the second somatosensory (SII), 7b, retroinsular, postauditory, and granular insular cortical areas of M. fascicularis.
J Comp Neurol
 
192
:
93
–108.
Sawamoto N, Honda M, Okada T, Hanakawa T, Kanda M, Fukuyama H, Konishi J, Shibasaki H (
2000
) Expectation of pain enhances responses to nonpainful somatosensory stimulation in the anterior cingulate cortex and parietal operculum/posterior insula: an event-related functional magnetic resonance imaging study.
J Neurosci
 
20
:
7438
–7445.
Schmahmann JD, Leifer D (
1992
) Parietal pseudothalamic pain syndrome. Clinical features and anatomical correlates.
Arch Neurol
 
49
:
1032
–1037.
Schneider R, Friedman D, Mishkin M (
1993
) A modality-specific somatosensory area within the insula of the rhesus monkeys.
Brain Res
 
621
:
116
–120.
Silberman EK, Weingartner H (
1986
) Hemispheric lateralization of functions related to emotion.
Brain Cogn
 
5
:
322
–353.
Svensson P, Beydoun A, Morrow TJ, Casey KL (
1997
) Human intra-muscular and cutaneous pain: psychophysical comparisons.
Exp Brain Res
 
114
:
390
–392.
Svensson P, Johannsen P, Jensen TS, Arendt-Nielsen L, Nielsen J, Stodkilde-Jorgensen H, Gee AD, Baarsgaard Hansen S, Gjedde A (
1998
) Cerebral blood-flow changes evoked by two levels of painful heat stimulation: a positron emission tomography study in humans.
Eur J Pain
 
2
:
95
–107.
Talairach J, Bancaud J (
1973
) Stereotactic approach to epilepsy: methodology of anatomo-functional stereotaxic investigations.
Prog Neurol Surg
 
5
:
297
–354.
Talairach J, Tournoux P (
1988
) Co-planar stereotaxic atlas of the human brain, 3-dimensional proportional system: an approach to cerebral imaging.
Stuttgart: Georg Thieme Verlag
 .
Tarkka IM, Treede RD (
1993
) Equivalent electrical source analysis of pain-related somatosensory evoked potentials elicited by a CO2 laser.
J Clin Neurophysiol
 
10
:
513
–519.
Tölle TR, Kaufmann T, Siessmeier T, Lautenbacher S, Berthele A, Munz F, Zieglgansberger W, Willoch F, Schwaiger M, Conrad B, Bartenstein P (
1999
) Region-specific encoding of sensory and affective components of pain in the human brain: a positron emission tomography correla-tion analysis.
Ann Neurol
 
45
:
40
–47.
Valeriani M, Restuccia D, Barba C, Le Pera D, Tonali P, Mauguière F (
2000
) Sources of cortical responses to painful CO(2) laser skin stimulation of the hand and foot in the human brain.
Clin Neurophysiol
 
111
:
1103
–1112.
Vogt BA, Derbyshire S, Jones AK (
1996
) Pain processing in four regions of human cingulate cortex localized with co-registered PET and MR imaging.
Eur J Neurosci
 
8
:
1461
–1473.
Wieser H (
1983
) Electroclinical features of the psychomotor seizure.
Stuttgart: Gustav Fisher
 .
Xu X, Fukuyama H, Yazawa S, Mima T, Hanakawa T, Magata Y, Kanda M, Fujiwara N, Shindo K, Nagamine T, Shibasaki H (
1997
) Functional localization of pain perception in the human brain studied by PET.
NeuroReport
 
8
:
555
–559.
Zhang DX, Owens CM, Willis WD (
1991
) Intracellular study of electro-physiological features of primate spinothalamic tract neurons and their responses to afferent inputs.
J Neurophysiol
 
65
:
1554
–1566.