MRI plays an increasingly important role in the diagnosis of status epilepticus (SE). Approximately half of patients with SE do not have pre-existing epilepsy, and the cause of de novo SE is frequently unknown. The role of MRI in the identification of causes of SE is invaluable. MRI is often helpful as a diagnostic tool in cases of non-convulsive status epilepticus (NCSE) with ambiguous EEG findings. Thus, MRI is recommended for all patients presenting with de novo SE, patients with known epilepsy with the first episode of SE and NCSE with equivocal EEG. Different peri-ictal MRI (PMA) alterations may be seen during ongoing SE or briefly after its cessation. They commonly present as peri-ictal hyper-perfusion, diffusion restriction and/or FLAIR-hyperintensity affecting specific brain areas such as the cortex, hippocampus, pulvinar of the thalamus, splenium of the corpus callosum, claustrum or cerebellum, frequently in combination, suggesting the existence of a “status epilepticus network”. MRI sequences, which are necessary for detecting PMA, include diffusion-weighted imaging, fluid attenuated inversion recovery, T1-weighted imaging with and without contrast application, as well as perfusion sequences such as arterial spin labeling. Recent research suggests that they may serve as biomarkers for predicting an outcome in SE. Patients with PMA seem to have a higher mortality rate compared to those without PMA. However, there is still a substantial knowledge gap and there are many open questions related to imaging in SE. Further prospective quantitative MRI studies with uniform protocols, timing and follow-up periods are needed to answer these important and clinically relevant questions.

Status epilepticus (SE) can be regarded as the most severe expression of seizure activity characterized by a low probability of spontaneous cessation and mechanisms leading to metabolic and inflammatory derangements with increased risk of brain damage, alterations of neural networks, and potentially life‐threatening systemic complications. Time‐based criteria are fundamental in diagnosing SE as response to treatment and outcomes worsen with increasing duration in terms of neurological impairment and mortality. Classification of status epilepticus includes four axes: semiology, EEG correlates, etiology, and age. Semiology, especially when evolving from SE with prominent motor phenomena (SE‐PM) to nonconvulsive status epilepticus (NCSE) with impaired consciousness, is associated with drug resistance and poor prognosis. The Salzburg EEG criteria define four subcategories of NCSE. The umbrella term ictal‐interictal continuum (IIC) includes a wide spectrum of EEG changes ranging from almost non‐epileptic situations to conditions that just fail to fulfill the criteria of NCSE. From a pathophysiologic point of view, EEG patterns on the IIC might be generated by both the underlying etiology and the superimposed hypersynchronous epileptic activity. It is impossible to disentangle their relative contributions by visual inspection of the EEG. However, it is essential to identify the contribution of ictal activity in an individual patient as it may be amenable to treatment with antiseizure medication (ASM). The main approach is to perform a diagnostic intravenous ASM trial (diagnostic IV‐ASM trial) during EEG recording and to assess for both EEG and clinical response. Furthermore, CT‐ and MRI‐perfusion studies help to clarify the diagnosis. Early and consequent treatment is necessary to minimize the total time spent in status, also called seizure burden. This educational review focuses on the diagnosis of status epilepticus in children and adults, excluding neonatal status.

Objective: Status epilepticus (SE) is the most severe expression of seizures, encompassing both SE with prominent motor symptoms and nonconvulsive SE (NCSE). Ictal–interictal continuum (IIC), an electroencephalographic phenomenon, is characterized by periodic discharges (PD), spike‐and‐waves or sharp‐and‐waves (SW), or lateralized rhythmic delta activity (LRDA). Peri‐ictal magnetic resonance imaging (MRI) abnormalities (PMA) may offer a potential surrogate marker for ictal activity, yet their association with IIC remains unclear. We aimed to investigate the occurrence of PMA in patients with SE and IIC, and to determine the relationship between IIC patterns and PMA through a latent cluster analysis (LCA). Methods: In a prospective cohort study, 223 adult patients diagnosed with SE or IIC underwent electroencephalography (EEG) and MRI within 48 h of diagnosis. Patients were stratified into two groups: the IIC group and SE group. PMA were assessed using the following MRI sequences: diffusion‐weighted imaging, fluid‐attenuated inversion recovery, and arterial spin labeling. LCA was performed to identify classes based on etiology, EEG patterns, and their localization. Results: PMA were as frequent in patients of the IIC group (23/49, 47%) as in patients of the SE group (64/149, 43%, p = .37). In the IIC group, peri‐ictal hyperperfusion was more frequently associated with lower frequency PD/SW (.5–1 Hz; 12/19, 63%), followed by LRDA (4/13, 31%) and higher frequency PD/SW (>1–2.5 Hz; 4/17, 24%). LCA revealed two classes; Class 1, characterized by nonunilateral high‐frequency PD/SW and triggering factors in epilepsy, had fewer PMA (18%) as compared to Class 2, characterized by predominantly unilateral low‐frequency PD/SW and diverse etiologies (50%; odds ratio = 5.79, p = .02). Significance: PMA occurrence in IIC aligned closely with that in SE, suggesting an overlap between IIC and SE and raising the critical question of whether patients with IIC may have NCSE. We propose an etiology‐driven approach for EEG interpretation in IIC, which may enhance diagnostic accuracy and treatment strategies.

Status epilepticus (SE) is a neurologic emergency with possible long-term sequelae, including the development of epilepsy. We aimed to determine whether peri-ictal MRI abnormalities (PMA) are associated with unprovoked seizures after de novo SE. Adults without epilepsy and a first nonhypoxic SE were included from a prospective database of the Christian Doppler University Hospital, Salzburg, Austria. MRI scans performed within 48 hours of SE diagnosis were evaluated for PMA on diffusion-weighted imaging (DWI), T2-weighted fluid-attenuated inversion recovery (FLAIR), and arterial spin labeling (ASL). The outcome was unprovoked seizures, assessed by retrospective review of hospital records and telephone interviews. Cumulative risk of seizures stratified by PMA occurrence (PMA on DWI/FLAIR, hyperperfusion on ASL, no PMA) was evaluated through survival analysis. Multivariable analysis using logistic regression was conducted to assess the influence of PMA, etiology, semiology, and ictal EEG patterns on seizure occurrence. Among 135 patients included (median age 70 [interquartile range (IQR) 58-80] years and 55% female), 43 (32%) experienced seizures during a median follow-up of 23 (IQR 9-39) months. The cumulative seizure probability at 1 and 4 years was 34% (95% CI 17-47) and 61% (95% CI 37-75) in patients with PMA on DWI/FLAIR; 0% and 13% (95% CI 0-29) in patients with hyperperfusion on ASL; and 25% (95% CI 14-34) and 36% (95% CI 19-49) in those without PMA. Other SE features associated with higher seizure risk were treatment refractoriness (odds ratio [OR] 2.93, 95% CI 1.23-6.98, p = 0.02), longer SE duration (OR 1.003, 95% CI 1.001-1.004, p = 0.02), and lateralized periodic discharges on ictal EEG (OR 3.59, 95% CI 1.37-9.83, p = 0.005). In multivariable analysis, PMA on DWI/FLAIR (log-odds 1.23, 95% CI 0.13-2.33, p = 0.03) and remote etiology (log-odds 1.85, 95% CI 0.45-3.25, p = 0.009) were independently associated with increased seizure probability. PMA on DWI or FLAIR were independently linked to a cumulative seizure risk exceeding 60% after de novo SE, consistent with a diagnosis of epilepsy. The findings of our study may help guide long-term treatment decisions in patients with de novo SE.

Background and ObjectivesReliable prediction of short-term mortality in status epilepticus (SE) can contribute to guide clinical decisions. Current prognostic systems achieve only acceptable predic...