The Psychoses of Epilepsy
The Psychoses of Epilepsy
Michael R. Trimble
New York, NY: Raven Press, 224 pp, 1991
Michael Trimble has had a longstanding interest in epilepsy and its relationship to psychiatric illness. In this monograph, he analyses and interprets the literature from a perspective gained over 15 years. Anyone who has ventured into the field will already be aware of the complex relationship between psychiatric illness and epilepsy and will have been confronted with the need to have a great deal of background information on limbic neuroanatomy, epilepsy, psychiatric illness, contemporary neurodiagnostic technology and correlations between the brain and behavior. Only then can one turn to the confusing and often contradictory literature on psychiatric illness and epilepsy. The author’s task was to guide the reader through these areas. Trimble does this, although in the end the reader is still left with many unanswered questions and an appreciation that much more work is needed.
This book is not easy to digest. The information is complex, and several chapters consist of an analysis of the published literature, which Trimble summarizes before drawing his conclusions. He begins with the history of the association between epilepsy and mental illness and states that all the current controversies about the nature of the association were foreshadowed by the French, German and English physicians of the nineteenth century.
To aid the reader, a chapter is devoted to epilepsy classification, classification of psychiatric illness and relevant neuroanatomy. Concepts of temporal lobe epilepsy, psycho-motor epilepsy and complex partial seizures are clarified, ending up with the current concept of limbic epilepsy, in which a distinction is made between seizures arising in the temporal neocortex and seizures arising in the mediobasal region of the temporal lobe. For non-neurologists, a clear definition of partial complex seizures is lacking on the issue of impaired consciousness. An operational definition would be helpful, since this criterion is frequently misunderstood. In partial complex seizures, impaired consciousness refers to an inability to respond normally to exogenous stimuli by virtue of an altered awareness or responsiveness. Responsiveness refers to the ability of the patient to carry out simple commands or willed movements, while awareness refers to a patient’s contact with events during the period in question and its recall. Trimble goes on to contrast the ICD-9 classification with the DSM-III-R. He provides a clear review of the conceptual issues relating to the distinction between organic mental disorders and functional mental disorders. Organic disorders classically referred to mental disorders caused by structural brain disease. “Functional” refers to mental illness caused by an alteration in brain functioning. This distinction has become increasingly uncertain with developments in neuroscience. Epilepsy is a classic example of organic/functional blurring, since it is predominantly a disorder in which there is malfunction at an electrical level associated with variable structural pathology.
Before plunging the reader into the clinical controversies, Trimble provides a selected review of the relevant neuroanatomical structures, neurotransmitters and the brain-behavior correlations of aggression, anxiety and sexual behavior. This is a difficult chapter, especially for those who have no background in neuroanatomy. However, it is an excellent chapter and is well worth the reader’s effort. The information provided is a foundation for all of neuropsychiatry and not simply the understanding of the relationship between epilepsy and mental illness.
“Forced normalization” stands at the center of the controversies surrounding the relationship between epilepsy and mental illness. Forced normalization is a concept defined by Landolt and refers to the phenomenon in which epileptics become psychotic when their EEGs become more normal or entirely normal The term “alternative psychosis” is used when there is an inverse relationship between clinical seizures and psychosis rather than the EEG and psychosis, simply put as “fewer seizures, more psychosis.” The documentation of this phenomenon is poor with respect to the clinical states and the EEG profiles. Trimble reviews recent studies which support the existence of forced normalization or alternative psychosis. The most compelling evidence comes from the study by Ramani and Gumnit, who during prolonged but not continuous videoEEG monitoring showed forced normalization in the case of a 21 year old woman. She had complex partial seizures, and there was a consistent inverse relationship between psychosis and both EEG spike activity and the frequency of her seizures.
The mechanism of forced normalization/alternative psychosis is unknown. It may involve a true biological antagonism (linked to activation of the reticular activating system or abnormal neurotransmission), a limbic status epilepticus not reflected on the surface EEG, altered brain metabolism resulting from anticonvulsant medication, insomnia or a psychological reaction in response to seizure control. Trimble has not mentioned the possibility that forced normalization is an artifact of EEG sampling. A routine EEG samples approximately 20 to 30 minutes of brain electrical activity, and it is not unusual for the EEG of a patient with epilepsy to be normal at one time and abnormal at another. Almost all studies that purport to show forced normalization rely upon showing a correlation between psychosis and changes on routine EEGs. Nonetheless, the evidence suggests that forced normalization does exist, albeit infrequently, and with it the controversy that seizures can both provoke and alleviate psychosis (more seizures, more psychosis; more seizures, less psychosis).
The psychopathology associated with seizures includes auras, ictal psychoses, postictal psychoses, post-operative psychoses and interictal psychoses. Auras are the subjective psychical experiences at the onset of seizures. Modern studies have located the origin of these psychical experiences on limbic structures, particularly the amygdala and hippocampus The auras are a positive expression of the function of the temporal lobes elicited by the electrical discharge. They are not the result of inactivation between the limbic structures and the release of unaffected brain tissue. Patient-specific variables, however, contribute to the content of any given psychical experience. Ictal psychoses are caused by either petit mal or partial complex status. Typically, the psychotic symptoms are part of a delirium, although ictal psychosis may occur during unimpaired consciousness.
Postictal psychopathology often merges with ictal psychopathology. Postictal psychoses are either deliria or psychoses occurring in the setting of clear consciousness. The latter are indistinguishable from functional psychoses other than their relationship to recent seizures. The mechanism of the postictal psychosis is uncertain but may be related to a Todd’s paralysis — the temporary inactivation of brain tissue by the preceding seizure. Postictal psychosis is the opposite of forced normalizaton in that psychosis is provoked by seizures (more seizures, more psychosis). What is one to believe? Forced normalization (fewer seizures, more psychosis) or postictal psychosis (more seizures, more psychosis). The answer lies in studies using prolonged and continuous EEG monitoring, where accurate correlations between EEG and behavior can be found. As noted, almost all the studies to date have relied on the routine EEG to show these correlations. The routine EEG may miss the seizures or interictal electrical activity. Partial complex seizures may be not be obvious because they mostly do not produce motor phenomena or collapse. Moreover, the seizures may be brief lasting a minute or less and may occur without the patient being aware of them because of associated amnesia. These principles are illustrated in a case study that we conducted using continuous EEG recording over a period of three weeks. During this study, it was clear that the patient’s psychosis was related to the frequency of the seizures, which settled when the seizures were controlled (more seizures, more psychosis).
To muddy the waters further, patients with epilepsy may become psychotic after temporal lobectomy when the majority if not all the pathological nervous tissue (most commonly medial temporal sclerosis), presumably with the discharging electrical focus, have been removed. While most of these patients have been psychotic pre-operatively, 3.8% to 28.5% of post-operative paranoid or schizophreniform psychosis occurs de novo and is more likely to occur after right than left temporal lobectomy. Temporal lobectomy usually alleviates or significantly controls the frequency of seizures. Thus, neither the pathological brain tissue nor seizures are a necessary or sufficient cause of psychosis. Rather, disruption of brain tissue removed from the site of pathology and seizures must occur to produce psychosis. Since temporal lobectomy decreases or stops seizures, forced normalization may be responsible for the disruption of brain tissue remote from the seizure focus. How this squares with the phenomenon of postictal psychosis (more seizures, more psychosis) is unknown and awaits further study.
Is nothing settled in this field? Apparently not. Trimble looks at the interictcal personality disorder. He summarizes the studies using the Minnesota Multiphasic Personality Inventory (MMPI) and the Bear-Fedio Inventory. This yields suggestive but inconclusive evidence of higher levels of non-specific, non-psychotic psychopathology among patients with epilepsy than among normal patients. As for a specific interictal behavioral syndrome in patients with temporal lobe epilepsy, this, too, has not been unequivocally proven. However, in some patients with temporal lobe epilepsy, behavioral changes are striking and include hypergraphia, religiosity, viscosity and altered sexual behavior.
Interictal psychosis is less controversial and develops in patients with longstanding epilepsy, particularly temporal lobe epilepsy. The interictal psychosis associated with temporal lobe epilepsy is typically schizophreniform, although paranoid and affective psychoses may occur. The interictal psychosis associated with generalized epilepsy is more likely to be a confusional psychosis. Patients with left-sided temporal lobe seizures, especially those with a mediobasal limbic focus, are at greatest risk of developing the interictal schizophreniform psychosis. This psychosis typically has a latent interval of 11 to 15 years. In a subgroup of these patients, the phenomenon of forced normalization appears to apply in that the psychosis materializes as the seizure frequency diminishes.
The chapter on the treatment of the epileptic psychoses provides a brief background on neuroleptics, lithium and anticonvulsants, focusing on what is known about the mechanisms of action, essential pharmokinetics and interactions with each other. Carbamazepine is identified as a very useful compound because of its anticonvulsant, antidepressant and antimanic properties. Of note is that all neuroleptics lower the seizure threshold, chlorpromazine being the most epileptogenic and fluphenazine and the butyrophenones, the least epileptogenic. The management of the ictal and interictal psychoses is based upon what is felt to be the underlying relationship of the psychosis to the ictus in a given patient. Psychosis associated with seizures is treated by suppressing the seizures with carbamapzeine in the event of complex partial seizures. Sodium valproic acid and ethosuximide are used in the event of generalized absence status. Haloperidol or pimozide are preferred if antipsychotics are needed. Surgery to treat the epilepsy should be considered if it will lead to enhanced seizure control, and with it, control of the psychosis.
Interictal psychoses are managed as regular psychiatric disorders. Lithium or carbamazepine are used for mania with neuroleptics, if necessary. Antidepressants or ECT are used to treat depression. Heterocyclic antidepressants and the selective serotonin reuptake inhibitors may all lower the seizure threshold, while antidepressant-anticonvulsant drug interactions may lead to higher serum anticonvulsant levels and lower serum antidepressant levels. ECT has its place especially if depression with psychosis is preceded by a reduction in the frequency of seizures. For patients in whom psychosis materializes as the frequency of seizures declines (forced normalisation), chlorpromazine is the drug of choice because of its capacity to lower the seizure threshold. Trimble suggests that in this situation it may be necessary to reduce anticonvulsant levels to allow seizures to occur.
In contrast, when psychosis bears no relationship to seizure frequency or occurswith greater seizure frequency, haloperidol or pimozide, neuroleptics that are least likely to lower seizure threshold should be used. In all situations, psychopharmocological intervention should combined with appropriate psychotherapy and environmental manipulation.
In the final chapter, Trimble looks at the links between electrical discharges, natural (seizures) or iatrotrogenic (ECT) and psychopathology. As a model of psychosis, Trimble initially reviews the neuropathological, electroencephalographic and neuroradiological data on patients with schizophrenia and concludes that they point to a limbic system disturbance involving the hippocampus, parahippocampal gyrus and amygdala, especially on the left side.
ECT is an established treatment for depression, particularly depressive psychosis, providing significant evidence of an antagonistic relationship between seizures and psychosis. The possible mechanisms of action of ECT are reviewed, but none are able to explain its antipsychotic effect.
The interictal psychosis of epilepsy is schizophreniform. Phenomenologically, this psychosis differs from schizophrenia by virtue of retained affective warmth, lack of personality disintegration and infrequency of catatonic symptoms. There is no good evidence that genetic or psychosocial factors are responsible for the interictal psychosis. Organic etiologies are more likely with two competing theories. The first is that seizures and psychosis are both symptoms of a common underlying organic lesion of the temporal lobes. Links to a specific pathology have not, however, been established. The second is that the seizures themselves are responsible for the psychosis. Here, difficulties are encountered since seizures can both provoke and alleviate psychosis. Moreover, psychosis may occur between seizures with a latent interval of 11 to 15 years (the interical psychosis) and after temporal lobectomy; surgery in which the damaged nervous tissue together with the seizure focus is removed. How seizures may do all this remains uncertain. Kindling or an alternation in dopamine metabolism have all been considered. The evidence to support kindling in humans is sparse, if it exists at all, while altered dopamine metabolism has not been shown, other than that the psychoses of epilepsy respond, like all other psychoses, to dopamine blockers.
This is a subject that lies at the heart of neuropsychiatry and traverses the terrain of brain-behavior correlations using seizures as an organic model of psychopathology. Trimble has done an admirable job in bringing together the available literature in this field with the continuing controversies. The book is a valuable reference for all psychiatrists and neurologists who manage patients with epilepsy and must grapple with the theoretical and practical implications of epileptic psychosis.