Les psychoses interictales (PI) regroupent tous les troubles psychotiques évoluant en pleine conscience chez un sujet préalablement diagnostiqué comme porteur d’une épilepsie, qui ne suivent pas immédiatement la survenue d’une crise. Les données épidémiologiques démontrent l’existence d’une surreprésentation des troubles psychotiques dans l’épilepsie (taux de prévalence global de 5,6 %), situation particulièrement à risque de surmortalité prématurée. Le tableau clinique des PI correspond à celui d’une schizophrénie. Les facteurs de risque les plus importants sont l’existence d’une épilepsie temporale réfractaire, d’un retard cognitif et d’antécédents personnels et familiaux de psychose et/ou d’épilepsie. Il n’existe pas de modèle neurobiologique explicatif satisfaisant à l’heure actuelle mais des anomalies structurelles étendues de l’encéphale et des anomalies neurochimiques ont été retrouvées dans les PI, sans marqueur neuropathologique spécifique. Certains tableaux cliniques témoignent d’une affinité entre épilepsie et psychose (les crises induisent le trouble psychotique), d’autres d’un antagonisme entre les deux conditions (la disparition des crises induit le trouble psychotique), à travers les concepts de normalisation forcée et de psychose alternante. Le traitement antiépileptique peut participer au déclenchement d’un trouble psychotique. La stratégie thérapeutique n’est pas codifiée mais repose empiriquement sur la prescription d’un antipsychotique atypique et sur une réévaluation du traitement antiépileptique. Certains antipsychotiques (clozapine, olanzapine, quetiapine) peuvent abaisser le seuil épileptogène mais la prescription d’antipsychotiques chez une personne épileptique traitée par des antiépileptiques est possible en pratique en toute sécurité. L’analyse des interactions médicamenteuses et l’anticipation du risque d’effets indésirables cumulatifs entre antipsychotiques et antiépileptiques constituent des éléments cruciaux de la prise en charge pratique. Les PI sont encore à l’heure actuelle sous-diagnostiquées et sous-traitées.

Interictal psychosis (IIP) refers to psychosis that occurs in clear consciousness in persons with epilepsy (PWE) with temporal onset not during or immediately following a seizure. The pooled prevalence estimate of psychosis in PWE is 5.6%. PWE and schizophrenia have very high mortality, and more than one in four persons with both disorders die between the age of 25 and 50years. IIP can manifest in brief or chronic forms. The chronic forms of IIP may closely resemble schizophrenia. However, some authors have described the typical presence of persecutory and religious delusions, sudden mood swings and the preservation of affect, as well as rarity of negative symptoms and catatonic states, but these differences remain controversial. Typically, IIP starts after many years of active temporal lobe epilepsy. Several epilepsy-related variables are considered pathogenically relevant in IIP including epilepsy type and seizure characteristics. Risk factors for developing IIP are family history of psychosis, learning disability, early age of onset of epilepsy, unilateral or bilateral hippocampal sclerosis, history of status epilepticus, history of febrile seizures, and poorly controlled temporal lobe epilepsy. In patients with epilepsy and psychosis, structural imaging studies have shown several relevant changes leading to conflicting findings. Altered neuronal plasticity and excitability have been described in epilepsy and psychotic disorders. Neuropathological data suggest that IIP are not the result of classic epileptic pathology of the temporal lobe. Forced normalization (FN) and alternating psychosis refer to patients with poorly controlled epilepsy (focal or generalized) who have had psychotic episodes associated with remission of their seizures and disappearance of epileptiform activity on their EEGs. FN mainly occurs in temporal lobe epilepsy when patients have frequent seizures that are abruptly terminated triggered by an antiepileptic drug, vagus nerve stimulation or epilepsy surgery. Treatment is based on withdrawal of the responsible drug, and by transient use of antipsychotics for acute symptomatic control on a case-by-case basis. FN is an entity whose pathophysiology remains uncertain. Antiepileptic drugs (AEDs) may sometimes induce psychotic symptoms and psychosis could be a direct effect of the AEDs. IIP has been reported more frequently following the initiation of zonisamide, topiramate, and levetiracetam when compared with other antiepileptic drugs. However, AEDs do not appear to be the only determinant of IIP. The management of IIP requires a multidisciplinary approach with early involvement of a liaison psychiatrist associated with a neurologist. IIP are underdiagnosed and mistreated. Existing recommendations are extrapolated from those established for the treatment of schizophrenia with some additional guidance from expert opinions. A two-step procedure, not necessarily consecutive, is suggested. The first step requires reevaluation of the antiepileptic treatment. The second step requires initiation of atypical neuroleptics. Antipsychotic drugs should be selected with consideration of the balance between pharmacological profiles, efficacy, and adverse effects. Regarding pharmacokinetic interactions, AEDs with inducing properties reduce the blood levels of all antipsychotics. It is important to consider implications of combining neuroleptics and AEDs with a similar spectrum of side effects. Regarding the duration of treatment, IIP episodes are more likely to be recurrent than in primary schizophrenia. In practice, atypical neuroleptics with few motor side effects such as risperidone can be used as first choice, given the low propensity for drug–drug interactions and the low seizure risk, with the added suggestion to start low and go slow. Clozapine could be prescribed in selected cases.