Abstracts

Rationale: Acute alterations in cerebral MRI related to ongoing seizure activity present in various brain areas and they are usually completely reversible. Reports on acute MRI changes in status epilepticus (SE) are scarce and based mainly on small selected populations. We aimed to analyze these changes in a large patient cohort, identify risk factors for their occurrence and correlate them with EEG. Methods: In a retrospective monocentric study, we analyzed brain MRI changes in acute/subacute phases of SE in a large patient cohort with initial clinical suspicion of SE (n=1449) at the Department of Neurology, Christian Doppler Klinik, Paracelsus Medical University, Salzburg, Austria, covering a 5-year period (01.01.2011 – 31.12.2015). Inclusion criteria were: 1) final electro-clinical diagnosis of SE at discharge from the hospital, 2) MRI performed within 30 days from begin of SE.All patients underwent a high-resolution MRI (3-Tesla, Philips Achieva Stream) with a standard protocol for suspected SE (T1-weighted three-dimensional isovoxel TFE with and without intravenous contrast, axial and coronal T2-weighted turbo spin echo, T2-weighted FLAIR and DWI sequences). Two raters independently reviewed MRI scans; in case of disagreement, a third rater was consulted. Results: We identified 277 patients fulfilling selection criteria with a mean age of 63 years (range 13-90 years), 42% were women. The most common causes of SE were brain tumors (n=58, 21%), incompliance in previously diagnosed epilepsy (n=45, 16%) and chronic cerebrovascular disease (n=30, 11%). Acute/subacute MRI changes related to SE were observed in 32 (12%) patients. The most important risk factor for SE-related MRI alterations was SE duration: the mean duration of SE was 6 days in the group with MRI changes, compared to 2 days in the group without MRI alterations (p = 0.011). In the univariate analysis, only inflammatory/autoimmune disease (OR 4.23, CI 1.20-14.96) and sepsis (OR 8.1, CI 1.1-59.63) were associated with an increased risk of acute MRI changes.The majority of all patients had a non-convulsive SE (NCSE, n=196, 71%), which either evolved from previous motor manifestations or was the only semiology of the SE episode. Generalized convulsive status epilepticus evolving into a NCSE was highly represented in the group with acute MRI alterations (OR 5.00, CI 1.83-13.68).Lateralized periodic discharges on EEG were strongly associated with SE-related MRI alterations (p < 0.0001). MRI changes due to SE were unilateral (n=23, 72%), multifocal (n=19, 59%), involving mesio-temporal structures (n=17, 53%). In 16 patients (50%) MRI alterations well co-localized with focal pattern on EEG. In 10 patients out of 14 (71%) in whom the follow-up MRI was performed, the MRI alterations disappeared completely. Conclusions: MRI changes, related to ongoing seizure activity, occurred in 12% of patients with SE especially in refractory cases. In half of patients, MRI alterations corresponded to focal EEG abnormality. In the majority of cases, MRI changes were fully reversible. Funding: None