Psychosis and status epilepticus: Borderland or

hidden cause?Brent Elliott

Department of Neuropsychiatry, The National Hospital for Neurology and Neurosurgery,

Queen Square, London, United Kingdom

Psychiatry, over the past century, has tended to echo theview of Hughlings-Jackson that ‘‘compound mental statescannot be owing to an epileptic discharge’’ (Hughlings-Jackson, 1958). The approach to investigating psychosisin patients with epilepsy has been to concentrate on theempirical use of descriptive psychopathology in order todelineate the various psychiatric syndromes and deter-mine to what extent the various abnormalities of affect,thought (delusions), and perception (hallucinations) aresimilar to or differ from those seen in the ‘‘functional’’psychoses in patients without epilepsy.

An almost entirely separate ‘‘positivist’’ tradition hasevolved within neurology. The central theorem of thisschool is that hallucinations occur as a consequence of theactivation of a localized group of neurons, which can beinvestigated by cerebral recording and cerebral stimula-tion. The ‘‘gold standard’’ investigation has been intracra-nial stereoelectroencephalography (SEEG).

Key differences between these two approaches lie inthe definition of an hallucination and in the concept of a‘‘compound mental state.’’ David defines ‘‘hallucination’’from the psychiatric perspective as ‘‘a sensory experiencewhich occurs in the absence of corresponding externalstimulation of the relevant sensory organ, has sufficientsense of reality to resemble a veridical perception, overwhich the subject does not feel s/he has direct and volun-tary control, and which occurs in the awake state’’ (David,2004); usually these may be auditory, but tactile and olfac-tory hallucinations are also seen in functional psychosessuch as schizophrenia. The neurologic definition of a hal-lucination is undoubtedly broader; according to DenisWilliams ‘‘an hallucination, which is a percept without astimulus may be organic or psychotic…Local disturbanceof the brain has evoked a perceptual response…(yet) therecan be no fundamental difference between sensations felt

in a limb or through the eye as a result of a local epilepticdischarge…(crucially) the feelings called fear, depressionor pleasure arising during the attack (also) have no localreference and for physiological purposes can be consid-ered to be organic hallucinations’’ (Williams, 1959).

In psychiatry the presence of hallucinations is usuallytaken to signify psychosis (i.e., complex mental states),yet although most psychiatrists would consider a patientexperiencing complex auditory hallucinations or persecu-tory delusions to be ‘‘psychotic,’’ probably none wouldextend this term to include those experiencing ictal affectssuch as fear, rage, or depression. At the phenomenologiclevel the reduction of all these symptoms into the singlecategory ‘‘hallucination’’ is both confusing and unhelpful.At the electrophysiologic level, however, distinctionbetween many of these diverse psychopathologic catego-ries is often not possible. It can be demonstrated thatrepeated seizures of a single area within the same patientcan produce different psychic responses, whereas stimula-tion of widely distinct areas within the same individualcan produce remarkable similar phenomena (Fish et al.,1993).

Penfield noted that auras could be reproduced by elec-trical stimulation of the temporal lobe (Penfield, 1938).He was interested in the evocation of memory and distin-guished between what he called experiential (mentalevents from the patients personal past) and interpretivephenomena (to do with the present circumstances of thepatient). A key element of his later thinking was that expe-riential responses occur virtually only in seizures arisingfrom the lateral temporal isocortex, where the stimuluswas strong enough to provoke an afterdischarge (Penfield& Perot, 1963).

The anatomic basis was revised further by Gloor, whoimputed a key role for limbic structures rather than tempo-ral neocortex (Gloor et al., 1982). He considered that theexperiential phenomena were not the result of a loss ofinhibitory control, as Hughlings-Jackson believed, butwere the result of positive activation of limbic structures.In his matrix theory, he proposed that ‘‘an epileptic dis-charge within the temporal lobe at the onset of a seizure,

Address correspondence to Dr. Brent Elliott, Department of Neuro-psychiatry, The National Hospital for Neurology and Neuropsychiatry,Queen Square, London WC1N 3BG, U.K. E-mail: brent.elliott@btinternet.com

Wiley Periodicals, Inc.ª 2009 International League Against Epilepsy

Epilepsia, 50(Suppl. 12): 70–72, 2009doi: 10.1111/j.1528-1167.2009.02355.x

PROCEEDINGS: THE INNSBRUCK COLLOQUIUM ON STATUS EPILEPTICUS

70

when it has not yet become too diffuse or too intense, maybe able to recreate a specific matrix that may be similar oridentical to that normally encoding a natural experience.Repeated discharge through mechanisms of synaptic plas-ticity… may have strengthened the interconnectivity ofthe neurons constituting such a matrix’’ (Gloor, 1990).

It is clear that ‘‘experiential’’ phenomena differ mark-edly from the complex psychotic symptoms seen in thefunctional psychoses or in the psychoses of epilepsy.However, the development of intracranial SEEG hasallowed a variety of psychological symptoms (both affec-tive and hallucinatory) to be recorded in patients under-going surgical assessment (using arrays of recordingelectrodes typically placed in the hippocampus, parahip-pocampal gyrus, amygdalae, and cingulate gyrus) duringperiods of ongoing seizure activity, particularly if thisactivity was prolonged, therefore, amounting to limbicstatus epilepticus (Gloor et al., 1982). These symptomsoccupy a separate category between experiential and inter-pretive phenomena, and the ‘‘complex mental states’’ ofpsychosis and some such states have been recorded onSEEG (e.g., Wieser, 1983).

The question remains, however, can complex psychoticstates occur as a consequence of an epileptic discharge?The best evidence to support this hypothesis can be foundin SEEG recordings during episodes of complex partialstatus epilepticus and during certain cases of apparent‘‘postictal’’ psychosis. Any review of this literature relieson a limited number of individual case reports:

Trimble for example, provides a case of nonconvulsivestatus producing a state resembling schizophrenia in a22-year-old man with complex partial and generalizedtonic–clonic seizures since the age of 3½. He complainedof the sudden onset feeling that rays were being passedthrough his body to sterilize him as well as voices in thesecond and third persons criticizing him. He was admittedto a psychiatric hospital and diagnosed with schizo-phrenia. EEG, which included sphenoidal electrodes, wascarried out. The main features included frequent sharpwaves on the right side with phase reversals in the rightsphenoidal leads, which occurred during a period of floridparanoid psychosis. The mental state and EEG bothdemonstrated a quick response to intravenous diazepam(Trimble, 1991).

Takeda et al. (2001) provide a further example, withSEEG data recorded in a 25-year-old woman with intrac-table temporal lobe epilepsy who developed a postictalpsychotic episode 2 days after a cluster of 18 seizures. Herintracranial EEG during the psychotic episode demon-strated frequent discharges of the left amygdala, whichwere absent prior to the episode. Kanemoto (1997) reportsthe case of a 24-year-old woman with complex partial sei-zures since the age of 4 years, who developed a periictalCapgras syndrome (the delusion that a person close to thepatient has been replaced by one or more impostors).

During her presurgical assessment one episode of postictalCapgras syndrome occurred in association with 15 iso-lated ictal fears and two complex partial seizures; all 17 ofthese seizures showed clear-cut epileptiform discharge inthe left amygdalohippocampal region. There have, how-ever, been other reports, which demonstrate no convincinglink between psychotic states and EEG abnormalitiesas recorded by intracranial electrodes (So et al., 1990;Mathern et al., 1995). Oshima et al. (2006) suggest thatpostictal psychosis should, therefore, be subdivided intotwo types; first, a nuclear type representing the establishedclinical picture as described by Kanner & Barry (2001;Table 1) and occurring as an indirect after-effect of seizureactivity, and second an atypical periictal type, occurringas a direct manifestation of limbic epileptic discharge.

It is, however, the psychopathology of interictal psy-chosis that is closest in nature to the complex psychoticstates seen in the functional psychoses. Kristensen & HeinSindrup (1978) reviewed sphenoidal electrode recordingsin 96 patients with partial seizures who developed a para-noid/hallucinatory psychosis after a median of 18 years,and compared these with an epileptic control group. Theyfound that the psychotic patients had a significantly largernumber of temporal mediobasal independent spike focithan controls. Patients with psychosis also had signifi-cantly more frequent bilateral rather than unilateral medi-obasal spike foci than controls. Interestingly, Hughes(1985) found that the incidence of bilateral as opposed tounilateral foci in temporal lobe epilepsy increased withage at a rate of almost 1% per year.

The possibility that at least some of these ‘‘complex’’psychotic states may, at least in part, be ‘‘driven’’ by ongo-ing discharges in limbic structures is an interesting one.Unfortunately, the absence of less invasive means foridentifying limbic ictal activity in psychotic patients withepilepsy impedes our understanding in this area. However,as Wieser (1983) notes ‘‘paralleled by clear-cut timerelated epileptic discharges in hidden brain areas the pha-sic psychical disturbance of some patients would not havebeen accepted as an epileptic dysfunction without depth

Table 1. Common findings among different

case series of the features of postictal

psychosis

Delay between the onset of psychiatric symptoms and the time

of last seizure

Relatively short duration of psychosis

Affect-laden symptomatology

The clustering of symptoms into delusional and affective-like

psychoses

An increase in the frequency of secondary generalized tonic–

clonic seizures preceding the onset of postictal psychosis

The onset of postictal psychosis after a long duration of epilepsy

(for a mean period of >10 years)

A prompt response to low-dose antipsychotics or benzodiazepines

71

Psychosis and Status Epilepticus

Epilepsia, 50(Suppl. 12):70–72, 2009doi: 10.1111/j.1528-1167.2009.02355.x

recordings. Such observations are perhaps first steps inbuilding a long-desired bridge to psychiatry.’’

Acknowledgment

I confirm that I have read the Journal’s position on issues involved inethical publication and affirm that this report is consistent with thoseguidelines.

Disclosure: The author declares no conflicts of interest.

References

David AS. (2004) The cognitive neuropsychiatry of auditory verbal hal-lucinations: an overview. Cogn Neuropsychiatry 9:107–123.

Fish DR, Gloor P, Quesney FL, Olivier A. (1993) Clinical responses toelectrical brain stimulation of the temporal and frontal lobes inpatients with epilepsy: pathophysiological implications. Brain116:397–414.

Gloor P, Olivier A, Quesney LF, Andermann F, Horowitz S. (1982) Therole of the limbic system in experiential phenomena of temporal lobeepilepsy. Ann Neurol 12:129–144.

Gloor P. (1990) Experiential phenomena of temporal lobe epilepsy: factsand hypotheses. Brain 113:1673–1694.

Hughes JR. (1985) Long-term clinical and EEG changes in patients withepilepsy. Arch Neurol 42:213–223.

Hughlings-Jackson J. (1958) Lectures on the diagnosis of epilepsy(Harveian Society). In Taylor J (Ed.) The selected writings of

John Hughlings Jackson. Vol. 1. Basic Books, New York, pp. 276–307.

Kanemoto K. (1997) Peri ictal Capgras syndrome after clustered ictalfear: depth-electroencephalogram study. Epilepsia 38:847–850.

Kanner AM, Barry JJ. (2001) Controversies in Epilepsy and Behaviour:Is the psychopathology of epilepsy different from that of nonepilepticpatients? Epilepsy Behav 2:170–186.

Kristensen O, Hein Sindrup E. (1978) Psychomotor epilepsy and psycho-sis: II. Electroencephalographic findings (Sphenoidal electroderecordings). Acta Neurol Scand 57:370–379.

Mathern GW, Pretorius JK, Babb TL, Quinn B. (1995) Unilateral hippo-campal mossy fiber sprouting and bilateral asymmetric neuron losswith episodic postictal psychosis. J Neurosurg 82:228–233.

Oshima T, Tadokoro Y, Kanemoto K. (2006) A prospective study of po-stictal psychoses with emphasis on the periictal type. Epilepsia47:2131–2134.

Penfield W. (1938) The cerebral cortex in man I. The cerebral cortex andconsciousness. Arch Neurol Psychiatry 40:417–442.

Penfield W, Perot P. (1963) The brains record of auditory and visualexperience: a final summary and discussion. Brain 86:595–696.

So NK, Savard G, Andermann F, Olivier A, Quesney LF. (1990) Acutepostictal psychosis: a stereo EEG study. Epilepsia 31:188–193.

Takeda Y, Inoue Y, Tottori T, Mihara T. (2001) Acute psychosisduring intracranial EEG monitoring: close relationship betweenpsychotic symptoms and discharges in the amygdala. Epilepsia42:719–724.

Trimble MR. (1991) The psychoses of epilepsy. Raven Press, New York.Wieser HG. (1983) Depth recorded limbic seizures and psychopathology.

Neurosci Biobehav Rev 7:427–440.Williams D. (1959) The structure of emotions reflected in epileptic expe-

riences. Brain 79:29–57.

72

B. Elliott

Epilepsia, 50(Suppl. 12):70–72, 2009doi: 10.1111/j.1528-1167.2009.02355.x