Peri-ictal psychoses

Among peri-ictal psychoses, pre-ictal ones are the least common and least studied. These present with various unspecific symptoms during the hours (rarely up to 3 days) before a seizure, including derealisation and depersonalisation experiences, forced thinking, ideomotor aura, déjà vu, jamais vu, anxiety, euphoria and perceptual experiences such as hallucinations or illusions. ^{Reference Agrawal and Mula44} These symptoms usually end with the seizure but are not associated with detectable electroencephalogram changes. ^{Reference Gold, Sher and Maldonado45}

Ictal psychoses are episodes of non-convulsive status epilepticus, mostly of temporal lobe origin, but they represent an infrequent occurrence. ^{Reference Shorvon and Trinka46,Reference Subota, Khan, Josephson, Manji, Lukmanji and Roach47}

Post-ictal psychoses

Among all post-ictal symptoms, psychotic ones occur in only 4% of cases while headaches and migraines are much more frequent. ^{Reference Nadkarni, Arnedo and Devinsky48} However, when looking at peri-ictal psychotic symptoms, post-ictal ones are the most frequent, representing 60% of peri-ictal psychotic symptoms. ^{Reference Gold, Sher and Maldonado45,Reference Laxer, Trinka, Hirsch, Cendes, Langfitt and Delanty49,Reference Oshima, Tadokoro and Kanemoto50} The classic clinical presentation is characterised by a period of normal mental state, known as the lucid interval, following a seizure or cluster of seizures and lasting between 24 and 48 h. ^{Reference Devinsky, Abramson, Alper, FitzGerald, Perrine and Calderon51–Reference Tarrada, Hingray, Aron, Dupont, Maillard and de Toffol53} The psychotic episode is characterised by mixed moods with psychomotor excitation, ecstatic moods and mystic or religious delusions, ^{Reference Kanemoto, Kawasaki and Kawai26} resembling more an affective psychosis than a pure psychotic disorder. ^{Reference Logsdail and Toone54} However, post-ictal psychoses differ from depressive disorders with psychotic symptoms. While both post-ictal psychosis and psychotic symptoms in depression involve experiencing psychosis, the former occurs directly following a seizure, the psychotic symptoms do not necessarily follow the mood theme and, while post-ictal psychosis is usually temporary, resolving within days or weeks, psychotic depression may last longer depending on treatment.

Post-ictal psychoses represent psychiatric emergencies, given the high prevalence of aggressivity and mortality rates for suicide or other accidents historically reported in up to 28.5% of cases. ^{Reference Umbricht, Degreef, Barr, Lieberman, Pollack and Schaul55} Recent data on the prognosis of post-ictal psychoses are still lacking. In terms of long-term prognosis, historical data suggest the development of chronically unremitting psychotic symptoms in 21–25% of individuals. ^{Reference Slater, Beard and Glithero5,Reference Umbricht, Degreef, Barr, Lieberman, Pollack and Schaul55} More recent data indicate the possibility of bimodal psychosis characterised by relapses, against a background of chronic symptoms. ^{Reference Adachi and Ito18}

Interictal psychoses

Historical descriptions of chronic psychosis in epilepsy include the so-called interictal schizophrenia-like psychosis of epilepsy, which various authors have identified as distinct from sporadic schizophrenia in terms of clinical presentation and long-term outcomes. ^{Reference Trimble56,Reference Adachi, Kato, Onuma, Ito, Okazaki and Hara57} In fact, according to historical descriptions, psychoses in epilepsy seem to be characterised by low rates of unfavourable symptoms, low rates of cognitive impairment, better response to treatment and lower rates of hospitalisation. ^{Reference Laxer, Trinka, Hirsch, Cendes, Langfitt and Delanty49,Reference Adachi, Kato, Onuma, Ito, Okazaki and Hara57} Risk factors for the development of chronic psychosis in epilepsy include temporal lobe epilepsy, at least 15 years’ duration of active disease and a family history of psychosis. ^{Reference Adachi, Matsuura, Okubo, Oana, Takei and Kato30} Adachi et al have pointed out that the course of the disease differs between patients with interictal psychoses as compared with those with schizophrenia, with a progressive deterioration occurring only in the latter. ^{Reference Chen, Lusicic, O’Brien, Velakoulis, Adams and Kwan58}

Treatment-related psychoses

A retrospective study involving 2630 people with epilepsy between 1993 and 2015 looked specifically at the aetiology of psychotic symptoms; only one out of seven cases showed a clear temporal relationship with the treatment. All individuals showed complete remission following drug discontinuation, with risk factors including a diagnosis of temporal lobe epilepsy, female gender and treatment with levetiracetam. ^{Reference Chen, Choi, Hirsch, Katz, Legge and Buchsbaum59} One US retrospective study of 4085 adults newly started on antiseizure medication (ASM) found that around 17.2% developed psychiatric side-effects with any drug, with risk factors including a diagnosis of drug-resistant epilepsy and a previous psychiatric history. ^{Reference Mula, Kanner, Jetté and Sander60} Psychotic symptoms have been described with most ASMs; sodium channel blockers are those with the lowest risk of psychiatric side-effects. ^{Reference Löscher and Klein61} Cenobamate has a dual mechanism of action, blocking voltage-gated sodium channels through a pronounced action on persistent rather than transient currents and acting as a positive allosteric modulator of GABA_A receptors independently from the benzodiazepine binding site. ^{Reference Roberti, De Caro, Iannone, Zaccara, Lattanzi and Russo62–Reference Klein, Krauss, Aboumatar and Kamin64} Cenobamate has shown a low prevalence of psychiatric side-effects and a good overall tolerability profile. ^{Reference Calle-López, Ladino, Benjumea-Cuartas, Castrillón-Velilla, Téllez-Zenteno and Wolf65}

With regard to ASM-related psychotic symptoms, forced normalisation plays a unique role. This is an intriguing phenomenon characterised by the emergence of psychiatric disturbances following either the establishment of seizure control or reduction in epileptic activity in an individual with previous uncontrolled epilepsy. ^{Reference Lee, Denton, Ladino, Waterhouse, Vitali and Tellez-Zenteno66} The mechanism is still unknown, but has been described with both ASM and vagus nerve stimulation, raising the hypothesis that the phenomenon is linked to the neurobiology of seizure control. ^{Reference Weber, Dill and Datta67,Reference Macrodimitris, Sherman, Forde, Tellez-Zenteno, Metcalfe and Hernandez-Ronquillo68} Finally, one recent systematic review clarified that forced normalisation can be seen primarily in focal epilepsies (80%) and symptomatic aetiology (44%). ^{Reference Lee, Denton, Ladino, Waterhouse, Vitali and Tellez-Zenteno66} Studies looking at the neurobiological mechanisms of forced normalisation would be able to elucidate the neurobiological links between epilepsy and psychosis.

Postoperative de novo psychoses represent specific conditions with hitherto minimal data; epidemiological data remain sparse, but these represent a serious complication of epilepsy surgery. ^{Reference Devinsky, Barr, Vickrey, Berg, Bazil and Pacia69–Reference Kanemoto, Tsuda, Goji, Tadokoro, Oshima and Tachimori71}

Diagnosing psychotic disorders in epilepsy

Given the complexities of clinical presentations, it can be challenging for neurologists to identify early psychotic problems, especially for those with limited psychiatric training (Table 3).

Table 3 Clinical elements requiring consideration when diagnosing interictal psychosis in epilepsy

The prompt identification of symptoms represents the foremost step for a prompt referral to psychiatric services. There is only one screening instrument proposed for the identification of psychotic symptoms in epilepsy, namely the Emotions with Persecutory Delusions Scale (EPDS). ^{Reference Akrout Brizard, Limbu, Baeza-Velasco and Deb72} This self-reported questionnaire has shown good psychometric properties against clinical diagnosis. Nevertheless, the lack of a clear cut-off score and the complex scoring system significantly affect implementation in clinical practice. At present, there are no recommended screening instruments for psychotic symptoms in epilepsy. Clinicians, invividuals and their families need to be aware of the increased risk and potential clinical scenario, and neurologists should consider exploring with simple questions the presence of hallucinations or thought disorders for prompt referral of patients to mental health services. Patients with intellectual disabilities and epilepsy represent a unique population posing several challenges in terms of diagnosis and management of psychotic symptoms and, to date, data are still lacking. ^{Reference Salpekar, Mula, Agrawal and Kaufman73} From a psychiatric perspective, it can be challenging for psychiatrists not trained in the neuropsychiatry of epilepsy to immediately distinguish epilepsy-related presentations (e.g. post-ictal psychosis or forced normalisation) from comorbid schizophrenia or schizophreniform disorders, and to distinguish between episodic versus chronic conditions. Understanding the nature of these conditions is vital to inform optimal treatments. The field of neuropsychiatry, as a subspecialty, clearly represents the opportunity to provide a comprehensive approach to clinical problems such as this, bridging between neurology and psychiatry. ^{Reference Arora, Prakash, Rizzo, Blackman, David and Rogers74}

Interactions and seizure risk

Regarding kinetic interactions, ASMs with enzyme induction properties such as phenytoin, carbamazepine and barbiturates reduce the blood levels of almost all antipsychotic drugs. ^{Reference Löscher and Klein61} However, this interaction is particularly relevant for quetiapine because it is metabolised only by the enzyme CYP3A4, and concomitant use with any inducer would lead to almost undetectable blood levels below 700 mg. ⁷⁸ Individual differences in treatment response should be carefully considered in regard to olanzapine and clozapine, which have complex metabolisms with multiple enzymatic pathways involved. Conversely, not all antipsychotics appear to influence the enzymatic pathways of ASMs significantly and, consequently, do not appear to affect ASM blood levels. ^{Reference Gold, Sher and Maldonado45}

Data on pharmacodynamic interactions are generally limited, but it is essential to consider the implications of combining antipsychotics and ASMs with a similar spectrum of side-effects, especially for sedation, weight gain and cardiotoxicity; and the apparent risk of bone marrow suppression in regard to the combination of carbamazepine and clozapine. ^{Reference Gold, Sher and Maldonado45}

Antipsychotic drugs have always been considered as being associated with an increased risk of seizure worsening. In psychiatric practice, antipsychotic medications are mainly chosen for their side-effect profile, in particular the propensity for extrapyramidal side-effects, weight gain or arrhythmias, ^{Reference Alper, Schwartz, Kolts and Khan79} but it is obvious that the issue of seizures is particularly evident in regard to people with epilepsy.

A systematic review of Food and Drug Administration (FDA)-controlled trials of antipsychotic drugs showed that clozapine is associated with the highest risk of seizures, with a standardised incident ratio of 9.5 (7.27–12.20) as compared with placebo, and such a risk seems to be dose and titration dependent. ^{Reference Reichelt, Efthimiou, Leucht and Schneider-Thoma80} Olanzapine and quetiapine were also considered to carry some risk, with an incident ratio in the region of 2.50 (1.58–3.74) for olanzapine and 2.05 (1.21–3.23) for quetiapine. ^{Reference Reichelt, Efthimiou, Leucht and Schneider-Thoma80} However, a very recent network meta-analysis showed that all second-generation antipsychotics, including even clozapine, are associated with no significantly increased risk of seizure occurrence as compared with placebo. ^{Reference Ishibashi, Horisawa, Tokuda, Ishiyama, Ogasa and Tagashira81} This discrepancy with previous studies may be related to the adoption of high dosages and rapid titrations, which are not typically current standard of care. It is important to emphasise that all the above data are derived from psychiatric samples rather than from individuals with epilepsy. It is, therefore, evident that the issue of seizures as a side-effect of antipsychotics still deserves further clarification, and careful clinical monitoring is crucial in regard to people with epilepsy and psychosis. Among recent antipsychotics, lurasidone is a dopamine type 2 (D2), serotonin type 2 (5-HT2A) and 5-HT7 receptor antagonist ^{Reference Solmi, Fornaro, Ostinelli, Zangani, Croatto and Monaco82} and showed a favourable side-effect profile with a very low rate of side compared with placebo. ^{Reference Trimble, Kanner and Schmitz83}

Treatment duration

Regarding psychosis in epilepsy, treatment requirements and duration represent a challenge because the time course of psychosis in epilepsy differs from that of new-onset psychosis and depends on the type of psychosis (Table 2). Post-ictal psychotic episodes in epilepsy are more likely to be recurrent, while interictal psychoses are more likely to be chronic.

Regarding post-ictal psychosis, an international Delphi survey among a group of experts suggested that symptomatic antipsychotic treatment is generally warranted but should subsequently be tapered off (e.g. after some weeks), given the episodic nature of the condition. ^{Reference de Toffol, Trimble, Hesdorffer, Taylor, Sachdev and Clancy76} Moreover, given the condition’s self-limiting time course, it must be acknowledged that, in many cases, minor episodes can also be managed by observation, nursing or carer supervision. ^{Reference Adachi, Ito, Kanemoto, Akanuma, Okazaki and Ishida84} In many instances there is no need for long-term antipsychotic treatment, which is mainly prescribed to reduce mortality and morbidity; ^{Reference Gold, Sher and Maldonado45} however, in one out of four cases, post-ictal psychosis may progress to a chronic psychosis, ^{Reference Vivekananda, Cock and Mula85} and this would pose the question of long-term treatment with antipsychotics. Complete seizure control would represent the best preventive treatment for post-ictal psychoses and the long-term development of chronic psychoses. However, there are isolated case reports of individuals who developed psychosis even years following successful epilepsy surgery. ^{Reference Vivekananda, Cock and Mula85}

For other peri-ictal psychoses, psychotropic medications are not generally indicated ^{Reference Gold, Sher and Maldonado45} and resolve with effective management of the epilepsy, with no need to treat the psychosis directly.

Finally, the treatment of chronic interictal psychosis tends to be similar to that of primary schizophrenia, i.e. long term, with preferential use of antipsychotic medications, paying attention to the risk for interactions and seizure relapse. ^{Reference de Toffol, Trimble, Hesdorffer, Taylor, Sachdev and Clancy76,Reference Mula77} To date, no studies have specifically investigated depot or long-acting injectable antipsychotics in people with epilepsy, or whether these are associated with an increased risk of seizure deterioration as compared with oral formulations.

In conclusion, psychosis in epilepsy still represents a clinically relevant comorbidity deserving clinical attention. The relationship between these two conditions is complex and multifaceted, involving both shared neurobiological mechanisms and the impact of seizures on mental health. Epilepsy, in particular temporal lobe epilepsy, has been linked to an increased risk of psychotic disorders, but such a relationship is bidirectional. The precise mechanisms involved remain incompletely understood, but multiple factors are likely to be in play, from the involvement of specific brain structures or networks to neurochemical imbalances and individual genetic predisposition, all contributing to this co-occurrence. Treatment approaches often involve antipsychotic medications, although managing both conditions simultaneously can be challenging due to potential drug interactions and side-effects. Early identification and a holistic, individualised treatment plan are crucial for improving outcomes and quality of life for individuals affected by both epilepsy and psychosis.