Epilepsy & Behavior 19 (2010) 591–595

Contents lists available at ScienceDirect

Epilepsy & Behavior

j ourna l homepage: www.e lsev ie r.com/ locate /yebeh

Awareness of deficits during intracarotid anesthetic procedures in epilepsy:Comparisons of motor, naming, and comprehension awareness underamobarbital versus under etomidate

Sarah Banks a,⁎, Viviane Sziklas a, Jelena Djordjevic a, Rose Golinski b, Marilyn Jones-Gotman a

a Montreal Neurological Institute, Montreal, Canadab McGill University, Montreal, Canada

⁎ Corresponding author. Montreal Neurological InstituMontreal, QC, Canada, H3A 2B4. Fax: +1 514 398 1338

E-mail address: sarah.banks@mcgill.ca (S. Banks).

1525-5050/$ – see front matter © 2010 Elsevier Inc. Aldoi:10.1016/j.yebeh.2010.09.020

a b s t r a c t

a r t i c l e i n f o

Article history:Received 9 September 2010Revised 13 September 2010Accepted 18 September 2010Available online 5 November 2010

Keywords:AnosognosiaIntracarotid anesthetic procedureMemoryLanguageHemispheric specialization

Awareness of deficits is often impaired following disruption of the right hemisphere. Intracarotid anestheticprocedures (IAPs) represent a unique method by which we can assess functioning of each hemisphere inisolation. We used this technique to explore deficits of awareness of specific functions—motor ability, naming,and comprehension—in patients with temporal lobe epilepsy. Some patients were injected with amobarbital,whereas others were injected with etomidate. We found that injection into the right hemisphere, orepileptogenic focus in the right hemisphere following injection in the left, resulted in the lowest levels ofmotor awareness. We also found a higher level of awareness for expressive language deficits and lessawareness for receptive language deficits. Comparison of etomidate and amobarbital suggested moreawareness following injection of etomidate. We discuss how these findings contribute to our understandingof the right hemisphere's special role in awareness, and how research in other disorders and in comparativeneurology has shaped our conceptualization of the neuroanatomy of insight.

te, 3801 University, Room 276,.

l rights reserved.

© 2010 Elsevier Inc. All rights reserved.

1. Introduction

Anosognosia, or lack of insight into symptoms, is a commonoutcome of right hemisphere stroke [1,2]. It is also seen in othersyndromes, including behavioral change caused by right frontal tumor[3], and in cognitive symptoms of memory disorders, includingAlzheimer's, especially when more right-sided disruption is evident[4,5]. However, loss of awareness is also sometimes seen in bilateralconditions such as the cortical blindness associated with Anton'ssyndrome [6] and even in left hemisphere phenomena such as aphasia[7] and motor deficits following left hemisphere injury [8].

Different aphasia syndromes are associated with varying degrees ofawareness: Broca's aphasia, which is characterized by expressivelanguagedeficits, is generally associatedwith intact awareness,whereasWernicke's aphasia is characterized by incoherent, fluent speech inpatientswho act as if theymakeperfect sense [9]. This lack of awarenessof a language symptom is at odds with the concept of anosognosia as aconsequence only of right hemisphere injury. Indeed, some earlyresearch suggested that the jargon characteristic of Wernicke's aphasiaoccurs uniquely in patients with bilateral lesions [10]. However, morerecent research describes reduced insight in patients with evidence ofdisruption confined to the left hemisphere [11], although among such

patients this is true only of those who exhibit more comprehensiondeficit and less meaning in their expressed language [12].

Other than the apparent prevalence of anosognosia syndromes inright hemisphere brain injury, the anatomy of anosognosia is quitepoorly understood. It is largely agreed that there is a widespreadnetwork underlying awareness and that this includes both subcortical(e.g., thalamus [1]) and cortical regions. Cortical areas particularlyimportant in awareness appear to be those with more heteromodalassociation cortex, that is, the frontal and parietal lobes [3,13]. Thus,damage to a wide range of brain regions can lead to anosognosia forvarious symptoms. The propensity of right hemisphere damage tocause anosognosia remains somewhat unexplained.

Intracarotid anesthetic procedures (IAPs), a component of theworkup for surgery for medically intractable temporal lobe epilepsy,involve the injection of an anesthetic agent into one cerebralhemisphere, resulting in a temporary anesthetization of that hemi-sphere during which the language and memory skills of the contralat-eral hemisphere can be tested in isolation [14,15]. Following injection ofeach hemisphere there is a predictable loss of strength in thecontralateral hemibody, allowing for exploration of motor awareness.In addition, injection of the language-dominant hemisphere will resultin aphasia. This always involves anexpressive aphasia andnearly alwaysa receptive aphasia. The procedure offers a unique tool for assessingeach hemisphere's contribution to a task in isolation. Extant researchusing IAP to studyawareness (summarized inTable 1) demonstrates theexpected relationship between lack of awareness for motor weakness

Table 1Summary of IAP awareness literature.

Authors (year) [Ref.] Drug Dose N % Unawareness

Dominant injection Nondominant injection

Gilmore et al. (1992) [20] Methohexital Right 3.3±1 mg 8 0% motor 100% motorLeft 3.5±1.4 mg

Durkin et al. (1994) [32] Amobarbital Not reported 150 82% motor after either injection 82% motor after either injection4% motor after left injection only 12% motor after right injection only, difference

not explained by lateralization of epileptogenicfocus

Adair et al. (1995) [31] Methohexital 1–7 mg of 1% solution 69 48% motor 97% motorCarpenter et al. (1995) [17] Amobarbital 200 mg in 4 mL water

at 1 mL/s31 13% motor when focus ipsilateral 78% motor when focus ipsilateral, 82% when focus

contralateral82% when focus contralateral23% speech arrest when focus ipsilateral100% when focus contralateral

Lu et al. (1997) [33] Methohexital 3–5 mg in either 1 or2 injections

17 Not tested 55.6% motor, of which 60% had right pathology,20% left, and 20% unknown

Dodrill andOjemann (1997) [18] Amobarbital 112 mg men 172 Combined language, memory, and motorawareness: 43% when contralateral, 8%when ipsilateral

Combined language, memory, and motor awareness:73% when focus contralateral, 56% when ipsilateral100 mg women

Meador et al. (2000) [16] Amobarbital 100 mg 62 Not tested 88% motor, not analyzed by epileptogenic focusJones-Gotman et al.(2005) [19]

Etomidate 2-mg bolus, then6 mL/h infusion

16 38% motor 62% motor38% all language symptoms+31%comprehension only

592 S. Banks et al. / Epilepsy & Behavior 19 (2010) 591–595

and nondominant hemisphere injection [16]. However, this is compli-cated by the hemispheric location of the epileptogenic focus. Carpenterand colleagues found high levels of unawareness for motor weaknessfollowing nondominant injection, but similarly high levels followingdominant injections in patients with nondominant seizure foci [17].Dodrill and colleagues reported similar findings using an index scorethat combined levels of awareness for variousdeficits during anesthesia.They described the lowest levels of awareness following nondominantinjection in patients with a dominant focus, and the highest levels inpatients with a dominant focus following ipsilateral injection [18].Overall, these results converge to suggest that any reduced functioningof the right hemisphere increases vulnerability to anosognosia, at leastfor motor deficits. Although few studies have assessed insight forlanguage, Carpenter et al. also found that a nondominant epileptogenicfocus was associated with anosognosia for speech arrest followingdominant injection in all patients, whereas only 23% of patients with adominant focus were unaware of their speech arrest [17]. Given thatIAPs often produce both expressive and receptive language disruption,the dissociation seen between awareness for these two types of deficitcan also be investigated. Indeed, Jones-Gotman et al. reported thatfollowing dominant injection, 38% of patients denied any languagesymptoms, whereas an additional 14% of patients recalled expressive,but not receptive deficits [19].

Until fairly recently, the IAP was nearly always carried out usingamobarbital. The lack of availability of this barbiturate anesthetic inrecent years has left epilepsy centers searching for alternatives. At theMontreal Neurological Institute, we use a nonbarbiturate anesthetic,etomidate, which has properties that allow for a modified procedure.The Etomidate Speech and Memory test (or eSAM), as we call our IAP,involves an initial bolus of the drug, followed by an infusion. Thisprovides a great advantage over the sodium amobarbital procedure inthatwehavecontrol over the lengthof time forwhich thehemisphere ofinterest receives the anesthetic, avoiding a rushed period of evaluationof speech and memory. In an earlier study, our group reported lessinsight for contralateral weakness following nondominant than dom-inant injections of etomidate [19]; nevertheless, a greater proportionof patients showed some insight following dominant injection than isoften reported in the literature for the amobarbital IAP [16,17,20].

Here we report a study using an IAP to study each hemisphere'sisolated role in awareness of motor behavior and two componentsof language: naming and comprehension. The central aim of our studywas to compare awareness for motor deficits in each hemisphere inpatients during the eSAM test. We expected to find less awareness

following right hemisphere injection, but also expected that suchreduction in awareness would be more severe following injection intothe hemisphere contralateral to the lesion. We also investigated insightinto language and expected to find less awareness for comprehensiondeficits than for expressive difficulty. An additional aim of our studywas to compare awareness using etomidate versus using amobarbital.On the basis of our earlier findings we also expected that patientsadministered amobarbital would have less awareness during theprocedure than would those who were administered etomidate.

2. Methods

2.1. Subjects

Sixty-nine patients with intractable unilateral temporal lobeepilepsy being evaluated for surgery were recruited for this study.Theyunderwent an intracarotid anesthetic procedure as a component ofthis evaluation. At our center, the IAP is recommended when there is aquery regarding lateralization of language or if there is a concern thatsurgery may result in a severe loss in memory, in which case the IAP isused to estimate the potential for postoperative memory decline. Allsubjects in our study were left hemisphere dominant for languageaccording to their IAP results. Until relatively recently our center usedamobarbital for this procedure.More recently, amobarbitalwas replacedwith etomidate. Thirty-seven patients in this study received etomidate,and the remaining 32 received amobarbital. Demographics and detailsof lateralization of the epileptogenic focus are summarized in Table 2.This project was approved by the Montreal Neurological Institute'sResearch Ethics Board, and all patients gave written informed consent.

2.2. Procedure

All patients underwent a practice IAP to familiarize them with theprocedure the day before thefirst hemispherewas tested. The IAP itselftook place over 2 days, with one hemisphere being anesthetized eachday; each patient had an angiogram prior to injection to ascertainwhether there were any structural anomalies in the vasculature.Continuous EEG recordings were made during the testing to monitorthe effectiveness of the drug and also to detect seizure activity, whichrarely occurs during this procedure but could invalidate the results ofthe test.

Our test procedure has been detailed elsewhere [19,21]. Tosummarize, our procedure includes memory testing and a language

Table 2Demographic details, type of anesthesia, and epileptogenic focus for all subjects.

Etomidate Amobarbital

Left focus Right focus Left focus Right focus

SexFemale 8 10 9 6Male 12 7 11 6

HandednessRight 15 11 13 9Left 4 6 4 2Mixed 1 0 3 1

Age (years)Range 17–55 22–49 15–57 22–61Mean 39 39 34 43

Language (mother tongue)English 14 5 4 3French 6 12 15 9Other 0 0 1 0

Total 20 17 20 12

Fig. 1. Mean motor awareness following etomidate injection into each hemisphere, byepileptogenic focus. Error bars represent SEM. Higher bars refer to more awareness(1=no awareness, 2=some awareness, 3=full awareness).

Fig. 2. Mean awareness for comprehension and naming deficits following lefthemisphere injection, by side of focus. Error bars represent SEM. Higher bars refer togreater awareness (1=no awareness, 2=some awareness, 3=full awareness).

593S. Banks et al. / Epilepsy & Behavior 19 (2010) 591–595

component that tests receptive (commands) and expressive (namingobjects, counting, reading, spelling, repetition) language. Thememorytest comprises encoding of eight common objects following injectionwhen the drug is active, and testing of recognition memory using 24objects (8 objects that were shown prior to injection, the 8 targets,and 8 completely new objects as foils) following return to baseline onthe EEG. Slight differences exist in the test procedure depending onthe anesthetic administered because with etomidate we administer abolus followed by a continuous infusion and, hence, have control ofthe length of time duringwhichwe can test, which is not the casewiththe standard amobarbital bolus administration, for which the lengthof time during which the drug is active varies from patient to patient.However, the essential elements of the testing remain the same.

The final component of our IAP constitutes the core of the currentstudy: Following the recognitionmemory testwe ask the patients a setof questions regarding their awareness of (and, arguably, memory for)the various components of the test and of their experience during thetest.We askwhether they felt anyweakness during the test, if they hadany problems speaking, and if they had difficulty understanding theexaminer. Examiners responded to vague answers with follow-upquestions so that the patient's experience was clearly understood.

We coded responses to the awareness questions into threecategories: no awareness, some awareness, and full awareness. Thesecodes were given ranks of 1, 2, and 3 respectively. Only responses fromparticipants who actually demonstrated a deficit in a given area wereincluded (i.e., not all patients showed a comprehension deficit, so thoseparticipants were not included in analysis of awareness of comprehen-sion deficits, but contralateral weakness followed injection in allpatients and naming deficits followed every left hemisphere injection).

2.3. Data analysis

Data were analyzed using nonparametric techniques; specifically,Wilcoxon signed rank tests were used to make paired samplecomparisons between left and right injections. Between-subjecttests were made using Mann–Whitney U tests. We initially analyzedthe data from patients who received etomidate, and then comparedthese results with those of patients who received amobarbital.

3. Results

3.1. Patients who received etomidate

3.1.1. Motor awarenessAcross all patients who received etomidate, awareness for motor

deficit was significantly less following right hemisphere injection(57.5% no awareness vs 40% no awareness following left hemisphere

injection; Z=–2.041, Pb0.05). However, comparison of left and rightinjections in patients with left and right hemisphere TLE separately(Fig. 1) indicated that this difference reached significance only whenthe seizure focus was in the left hemisphere, with less awarenesswhen the right hemisphere was injected, that is, when a “bad” lefthemisphere was being tested (left hemisphere focus: Z=–2.18,Pb0.05). In patients with right TLE, there was no significant differencein awareness for motor deficits between patients who received leftand those who received right hemisphere injections (right hemi-sphere focus: Z=20, Pb0.527).

3.1.2. Awareness for language deficitsFollowing left (dominant) hemisphere injection, all patients

demonstrated speech arrest and 80% showed some degree ofcomprehension deficit. Seventy-five percent of patients who showedcomprehension deficit were completely unaware of their deficit,whereas only 10% of patients lacked awareness of their naming deficit.There was significantly less awareness of comprehension deficit thanof naming deficit (Z=–5.01, Pb0.0005), and this was true for patientswith a left (Z=–3.40, Pb0.005) or a right (Z=–3.76, Pb0.0005)epileptogenic focus (Fig. 2).

Fig. 3. Mean motor awareness following injection of each anesthetic, by hemisphere.Error bars represent SEM. Higher bars refer to greater awareness (1=no awareness,2=some awareness, 3=full awareness).

594 S. Banks et al. / Epilepsy & Behavior 19 (2010) 591–595

3.2. Comparison of anesthetics

3.2.1. Motor awarenessFollowing left injection, awareness of motor weakness was

greater with etomidate than with amobarbital (Amytal) (Z=–2.96,Pb0.005). Following right injection, this pattern approached signifi-cance (Z=–1.91, P=0.056) (Fig. 3).

3.2.2. Awareness for language deficitsComparison of anesthetics suggested considerably more awareness

of naming deficits with etomidate than with amobarbital (Z=–3.35,Pb0.005) (Fig. 4). Awareness of comprehension deficits, on the otherhand, was minimal for all patients, and no significant difference wasfound between etomidate and amobarbital.

4. Discussion

This study had three purposes: to compare motor awarenessduring left versus right IAP as a function of affected hemisphere; tocompare awareness of two features of language, expression and

Fig. 4. Mean awareness for comprehension and naming deficits following lefthemisphere injection, by anesthetic type. Error bars represent SEM. Higher bars referto greater awareness (1=no awareness, 2=some awareness, 3=full awareness).

comprehension; and to compare awareness under the newly intro-duced IAP anesthetic etomidate with that found under the moretraditional IAP anesthetic amobarbital. We found significantly lessawareness of motor deficits when the right hemisphere was injectedthan when the left hemisphere was injected, but awareness was alsoreduced when the right hemisphere harbored the epileptogenicfocus and the left hemisphere was injected. As expected, we foundpoor overall awareness for comprehension deficits compared withpreserved awareness for impaired naming, and this was independentof the side of the epileptogenic focus. Finally, comparing results underetomidate with those under amobarbital, we found that awarenessfor motor and naming deficits was greater under etomidate. Nodifference was seen for comprehension.

Consistent with earlier reports [17,18], we found that eitherinjecting the right hemisphere or having an epileptogenic focus inthe right hemisphere with a left hemisphere injection resulted inpoor motor awareness. This phenomenon provides further evidenceof the importance of the right hemisphere in body awareness. Thefinding that patients with right TLE showed reducedmotor awarenessfollowing left injection is interesting, as the area of the righthemisphere affected by epilepsy is mostly restricted to the medialtemporal structures, which are not usually associated with reducedawareness. However, one might speculate that the potential disrup-tion of a normal limbic system or its connections could result in a lackof normal emotional “tagging” of information. This could lead to lessappreciation of abnormal behavior in patients with right hemisphereTLE during left hemisphere injection.

Whereas there are many differences between the right and lefthemispheres that could contribute to their different roles inanosognosia, an intriguing avenue of current research that warrantsmention is the role of von Economo neurons in awareness. Thesespecial neurons are present only in apes and certain higher mammalssuch as elephants and some whales [22,23] and, in humans, are 30%more prevalent in the right than left hemisphere [24]. They developpostnatally [25], are specific to the frontoinsular cortex and anteriorcingulate cortex, and have a suggested role in error recognition[26,27]. Furthermore they are differentially targeted in a syndromecharacterized by early loss of insight: frontotemporal dementia [28].Evidence is converging on a key role for these neurons in awareness.Our finding of reduced awareness in patients with unilateral right TLEduring left hemispheric anesthesia may in part be explained in termsof disruption in the medial temporal–insula and –cingulate networksin these patients.

The reported dissociation between awareness for expressive andreceptive language within the same individuals (only previouslyreported by our group [19]) suggests a role for feedback in anosognosiawith insight lacking when comprehension deficits prevent accurateerror monitoring. Previous feedback-related models have focused onexplaining deficits in motor awareness [29], but have also beenextended to awareness of aphasia [9]. Importantly, the occurrence ofanosognosia for comprehension difficulties during left hemisphereanesthesia clearly indicates that anosognosia is not exclusively a right(or nondominant) hemisphere phenomenon.

We found less awareness following amobarbital than etomidateinjection. There are several possible explanations for this findingincluding some differences in the protocol, resulting from the bolus–perfusion technique used when administering etomidate. Thistechnique allows more control over the testing process. The twoanesthetics are also different in their mechanism.Whereas both drugsinfluence GABAA receptors [30] and both are hypnotics with amnesticeffects, their properties are different: for example, amobarbital hasan analgesic effect, whereas etomidate does not. However, EEGrecordings during each of the procedures produced similar resultsin an earlier study [19], suggesting that similar levels of anesthesiaare achieved in the two cases. The exact cause of the difference inlevels of awareness between the two anesthetics remains elusive, and

595S. Banks et al. / Epilepsy & Behavior 19 (2010) 591–595

warrants further research. Clinically, the increased level of insightmay mean that preparing patients well before the procedure isespecially important when using etomidate, because awareness ofevents experienced during the period of anesthesia may be unsettlingfor the patient.

An important limitation of this study is that awarenesswas queriedafter the anesthetic was no longer effective. However, earlier studiescomparing testing during and after the drug's active period showessentially the same results [16,31], so it is likely that questioningwhen the patient is back to baseline informs us of his or her awarenessduring the active period of the anesthetic, not merely the memoryof what happened. Although awareness is not always complete orentirely absent, our “some awareness” category may not have beenspecific enough; future studies may break this down further. Lastly,because all participants in our study had epilepsy, it is not clear towhatdegree the results of our study would generalize to individuals withother neurological disorders.

In summary, this study adds to our understanding of various aspectsof awareness including the role of each hemisphere in awareness ofmotor functioning and the selective vulnerability of different compo-nents of the language system to anosognosia.

References

[1] Bisiach E, Vallar G, Perani D, Papagno C, Berti A. Unawareness of disease followinglesions of the right hemisphere: anosognosia for hemiplegia and anosognosia forhemianopia. Neuropsychologia 1986;24:471–82.

[2] Hier DB, Mondlock J, Caplan LR. Behavioral abnormalities after right hemispherestroke. Neurology 1983;33:337–44.

[3] Prigatano GP. Disturbances of self-awareness of deficit following traumatic braininjury. In: Prigatano GP, Schacter DL, editors. Awareness of deficit after brain injury:clinical and theoretical issues. New York: Oxford Univ. Press; 1991. p. 111–26.

[4] Starkstein SE, Vazquez S, Migliorelli R, Teson A, Sabe L, Leiguarda R. A single-photon emission computed tomographic study of anosognosia in Alzheimer'sdisease. Arch Neurol 1995;52:415–20.

[5] Vogel A, Hasselbalch SG, Gade A, Ziebell M, Waldemar G. Cognitive and functionalneuroimaging correlate for anosognosia in mild cognitive impairment andAlzheimer's disease. Int J Geriatr Psychiatry 2005;20:238–46.

[6] Sugar HS, Goldstein AS. Visual anosognosia in cortical blindness: Anton'ssymptom. Am J Ophthalmol 1965;60:335–7.

[7] Alajouanine T. Verbal realization in aphasia. Brain 1956;79:1–28.[8] Cocchini G, Beschin N, Cameron A, Fotopoulou A, Della Sala S. Anosognosia for motor

impairment following left brain damage. Neuropsychology 2009;23:223–30.[9] Lebrun Y. Anosognosia in aphasics. Cortex 1987;23:251–63.[10] Weinstein EA, Lyerly OG, Cole M. Meaning in jargon aphasia. Trans Am Neurol

Assoc 1964;89:266–8.

[11] Kertesz A, Benson DF. Neologistic jargon: a clinicopathological study. Cortex1970;6:362–86.

[12] Banks SJ, Weintraub S. Cognitive deficits and reduced insight in primaryprogressive aphasia. Am J Alzheimers Dis Other Demen 2008;23:363–71.

[13] Pia L, Neppi-Modona M, Ricci R, Berti A. The anatomy of anosognosia forhemiplegia: a meta-analysis. Cortex 2004;40:367–77.

[14] Branch C, Milner B, Rasmussen T. Intracarotid sodium amytal for the lateralizationof cerebral speech dominance: observations in 123 patients. J Neurosurg 1964;21:399–405.

[15] Wada JA. Clinical experimental observations of carotid artery injections of sodiumamytal. Brain Cogn 1997;33:11–3.

[16] Meador KJ, Loring DW, Feinberg TE, Lee GP, Nichols ME. Anosognosia andasomatognosia during intracarotid amobarbital inactivation. Neurology 2000;55:816–20.

[17] Carpenter K, Berti A, Oxbury S, Molyneux AJ, Bisiach E, Oxbury JM. Awareness ofand memory for arm weakness during intracarotid sodium amytal testing. Brain1995;118(Pt 1):243–51.

[18] Dodrill CB, Ojemann GA. An exploratory comparison of three methods of memoryassessment with the intracarotid amobarbital procedure. Brain Cogn 1997;33:210–23.

[19] Jones-Gotman M, Sziklas V, Djordjevic J, et al. Etomidate speech and memory test(eSAM): a new drug and improved intracarotid procedure. Neurology 2005;65:1723–9.

[20] Gilmore RL, Heilman KM, Schmidt RP, Fennell EM, Quisling R. Anosognosia duringWada testing. Neurology 1992;42:925–7.

[21] Jones-Gotman M, Sziklas V. Intracarotid amobarbital procedure: a redesign.Epilepsia 1999;40(Suppl 2):78.

[22] Butti C, Sherwood CC, Hakeem AY, Allman JM, Hof PR. Total number and volume ofVon Economo neurons in the cerebral cortex of cetaceans. J Comp Neurol2009;515:243–59.

[23] Hakeem AY, Sherwood CC, Bonar CJ, Butti C, Hof PR, Allman JM. Von Economoneurons in the elephant brain. Anat Rec (Hoboken) 2009;292:242–8.

[24] Allman JM, Watson KK, Tetreault NA, Hakeem AY. Intuition and autism: a possiblerole for Von Economo neurons. Trends Cogn Sci 2005;9:367–73.

[25] Shankle WR, Rafii MS, Landing BH, Fallon JH. Approximate doubling of numbers ofneurons in postnatal human cerebral cortex and in 35 specific cytoarchitecturalareas from birth to 72 months. Pediatr Dev Pathol 1999;2:244–59.

[26] Klein TA, Endrass T, Kathmann N, Neumann J, von Cramon DY, Ullsperger M.Neural correlates of error awareness. Neuroimage 2007;34:1774–81.

[27] Craig AD. How do you feel—now? The anterior insula and human awareness. NatRev Neurosci 2009;10:59–70.

[28] Seeley WW. Selective functional, regional, and neuronal vulnerability infrontotemporal dementia. Curr Opin Neurol 2008;21:701–7.

[29] Heilman KM, Barrett AM, Adair JC. Possible mechanisms of anosognosia: a defectin self-awareness. Philos Trans R Soc Lond B 1998;353:1903–9.

[30] Olsen RW. Drug interactions at the GABA receptor–ionophore complex. Annu RevPharmacol Toxicol 1982;22:245–77.

[31] Adair JC, Gilmore RL, Fennell EB, Gold M, Heilman KM. Anosognosia duringintracarotid barbiturate anesthesia: unawareness or amnesia for weakness.Neurology 1995;45:241–3.

[32] Durkin MW, Meador KJ, Nichols ME, Lee GP, Loring BW. Anosognosia and theintracarotid amobarbital procedure (Wada test). Neurology 1994;44:978.

[33] Lu LH, Barrett AM, Schwartz RL, et al. Anosognosia and confabulation during theWada test. Neurology 1997;49:1316–22.