Abstract
Epilepsy is a disorder that has evoked much interest and controversy for centuries. It is an extremely important disorder in psychiatric practice because of the raised prevalence of most forms of psychiatric illness in patients with epilepsy. Psychoses are the specific psychiatric disorder most clearly associated with epilepsy [1]. Most research into epileptic psychosis, however, has concentrated on chronic, interictal psychosis so observations on post-ictal psychotic (PIP) states have been largely ignored. The following case is classical of PIP as it is described in the literature. Discussion on what it known about PIP and its management will follow.
Case report
CC is a 32-year-old single woman with a 10-year history of epilepsy that developed after a ventriculo-peritoneal shunt was inserted for idiopathic hydrocephalus. Serial electroencephalograms (EEG) showed asymmetrical delta slowing in the right frontal region. Cerebral computerised tomography and magnetic resonance imaging scans showed a small low density lesion in the cingulate gyrus in the right hemisphere. CC initially had complex partial seizures approximately 15 times per month and was treated with carbamezapine and acetazolamide. In the last 2 years, however, CC had two clusters of secondarily generalised tonic-clonic seizures and she had become acutely psychotic after both these episodes. On a third occasion, CC was referred to the liaison psychiatry team for review of bizarre behaviour that had begun 48 h after another cluster of seizures.
On assessment CC was alert and orientated but clearly acutely psychotic. She described a complicated delusional system which involved delusions of reference from medical equipment and incorporated visual and auditory hallucinations and illusions. She had concluded that she was being poisoned via her drip as part of a medical experiment because she was an invalid. A diagnosis of delirium was unlikely because of the absence of altered consciousness, the lack of a disturbed sleep-wake cycle, the absence of significant fluctuation in symptoms and her good orientation and attention on bedside cognitive testing.
CC was treated with haloperidol, initially 3 mg/day and increased to 7 mg/day after 1 week because of poor response. An EEG performed 5 days postictally showed no epileptiform activity. Within 14 days, the psychosis had resolved and CC was able to recall her symptoms in detail with good insight into their nature. Haloperidol was ceased and CC was able to return to her usual level of functioning at home.
Discussion
Approximately 70 cases of PIP have been found in the current review of the literature since 1988 [2–7]. Post-ictal psychosis accounts for approximately 25% of psychoses in epilepsy [8]. All researchers agreed that PIP was a diagnostic entity and most cases would have fulfilled diagnostic criteria proposed by Logsdail et al. [2] in their study of 14 cases of postictal psychosis: (i) the episode of confusion or psychosis manifested within 1 week of a seizure; (ii) the psychosis had a minimum length of 24 h and a maximum of 3 months; (iii) the mental state was characterised by delirium, or by delusions and hallucinations in clear consciousness, or by a mixture of these; (iv) the diagnosis excluded patients with anti-convulsant toxicity, recent head injury, previous inter-ictal psychosis or minor status on EEG.
It is interesting to note that Logsdail's definition combines confusional states and psychosis, when a conventional understanding of psychosis is psychotic symptoms in the absence of disturbed consciousness.
CC's case appeared to be classical of post-ictal psychosis as described in several studies [2–7] and highlights the following points.
(1) Psychosis occurs after an exacerbation of seizure frequency and appears after a lucid interval. This suggests a possible direct causal link between seizure activity and the development of psychosis. This is supported by the fact that psychotic symptoms usually appear at least a decade after the onset of seizure activity and by the further observation that psychosis often seems to be related to either an increase (as in this case) or a decrease (‘forced normalisation’) in seizure frequency [3,[9], [10]].
(2) The episodes are characterised by psychotic phenomena occurring in either clear or clouded consciousness. Patients frequently have no evidence of clouded consciousness [4]. This disputes Lishman's commonly held view [1] that PIP is an extension of a post-ictal confusional state with clouding of consciousness or at least amnesia for the episode a key characteristic. This raises many questions about the significance of clouding post-ictally and fundamental questions about the difference between delirium and psychosis.
(3) The psychotic symptoms were pleomorphic with marked mood changes as prominent symptoms [2, [3], [4], [6]]. The exact characteristics of the psychosis have varied between studies but all agreed that the mental state may be identical to functional psychosis in some patients. Hence, there is a significant possibility of misdiagnosis, especially in patients with clear consciousness.
(4) Continuous stereotactic depth and epidural EEG monitoring confirmed that post-ictal psychosis was likely to be an event distinct from peri-ictal confusion that may be caused by persistent seizure activity from deep brain structures such as the limbic system [5]. CC's EEG during psychosis showed no evidence of epileptic activity.
(5) Spontaneous resolution is usual but there is a tendency to recur. This clearly has treatment implications and thus makes therapeutic interventions difficult to assess because it is a self-limiting disorder. Anti-psychotic medication may shorten the period of psychosis and relieve distress in the acute phase. There is currently no evidence, however, that antipsychotic medication is required for long-term prophylaxis after symptoms resolve, especially given the risk of lowering the seizure threshold. Optimal seizure control may be the best management [6] but this remains speculative until definitions of PIP are clearer.
(6) The relationship between PIP and chronic psychosis (CP) is unclear but this has clinical importance. Some studies found differences in phenomenology between the two suggesting that they were quite different entities [4]. Others found that the two conditions had similar profiles for clinical and seizure variables suggesting shared aetiological factors [7]. Little is known of the longitudinal course of PIP but progression to CP is uncommon: Logsdail [2] found that only two out of 14 patients developed CP within 8 years of follow-up.
Theories about the aetiology of psychosis in epilepsy are far from conclusive. It is generally accepted that there is some as yet not fully clarified pathogenetic association between the two disorders. It may be that psychosis develops as a result of an antiepileptic drug or due to the psychosocial effects of living with epilepsy; psychosis and epilepsy may have a common cause or chronic seizure activity may in some cases cause psychosis [10]. It was particularly interesting in this case that CC's psychosis seemed to be entirely seizure specific. She was a woman who had suffered many traumatic life-events but these stressors had never precipitated psychiatric illness.
One possible hypothesis for psychosis as a result of seizure activity is kindling (repeated subthreshold electrical stimulation eventually leading to seizures). Psychosis may develop as seizures become more generalised (secondary epileptogenesis) and the chronic hyperexcitability leads to structural brain damage [9, [10]].
It is also speculated that kindling in certain areas of the brain may result in psychosis via a potentiation of dopamine or other neurotransmitter transmission [11]. Clustering of seizures may also result in more extensive brain damage and thus an increased vulnerability to psychosis [7]. Overall, however, the evidence is insufficient to fully elucidate the nature of the relationship. It is possible that the epileptic seizures themselves tend to cause the psychosis, or that the cause of the epilepsy (in this case possibly a lesion in the cingulate gyrus) is also the cause of the psychosis [1, [9]]. It may be that there is some validity in both of these possibilities and further research may clarify this issue.
Conclusions
Post-ictal psychosis appears to be a relatively rare condition that tends to present after an exacerbation in seizure frequency with pleomorphic psychotic phenomena and spontaneous resolution within days to weeks.
Post-ictal psychosis is important in psychiatry as it is an unusual clinical entity that may be misdiagnosed and therefore treated inappropriately. It is also interesting as it provides a model for organic psychoses in general. It is an area for fascinating further research.