Abstracts

Rationale: Differentiation of dissociative attacks from epileptic seizures has long been a cause of diagnostic difficulty. Specific clinical features such as complex motor, rapid limb movements, ictal vocalizations have been reported as being reliable semiological features to aid in recognition of dissociative attacks [1]. However, increasing knowledge of seizure semiology and underlying functional neuroanatomical networks, has shown that these clinical features also exist in epileptic seizures. Therefore, clinical pitfalls exist if the diagnosis of dissociative attacks is made based on a set of clinical cues without recording the event on EEG. The aim of this study is to identify the common semiology captured on video-electroencephalography (EEG) monitoring (VEM) in patients who have epilepsy but were misdiagnosed as having dissociative attacks, and categorize the semiology and neuroanatomical networks involved.

Methods: A total of 233 patients evaluated at an advanced epilepsy unit within Brisbane, Australia from January 2015 to June 2019 were reviewed. Cases were included if a diagnostic re-classification (from dissociative attacks to epilepsy) was made after VEM and these were compared to control epilepsy patients with an established diagnosis. Semiological features were recorded, classified according to neuroanatomical networks and compared between the groups.

Results: There were 16 patients (6.9%) that had been diagnosed as dissociative attacks prior to presentation, these patients were subsequently re-diagnosed as having epileptic seizures after inpatient VEM. The mean age was 33.2 years with a mean diagnosis duration of 12.8 years. There were 8 males and 8 females. Semiological features that were significantly correlated with misdiagnosis included: complex motor, facial modification, fear, screaming, hyperkinetic, autonomic symptoms and chapeau de gendarme. These features were commonly attributed to mesial networks such as the pre-frontal region(4), perisylvian(4), temporo-perisylvian(3) and pre-motor(2) networks.

Conclusions: Physicians who encounter patients with the semiologies identified within this study should consider referral for further appropriate investigations to ensure diagnostic accuracy.

This study highlights the importance of capturing a clinical event on VEM as this provides a more objective evaluation, rather than relying on clinical descriptive features in formulating diagnoses. Anatomo-electro-clinical correlation and an understand of functional neuroanatomical networks and its implications on seizure semiology is vital, as the regions implicated in these epilepsies often produce subtle EEG change due to their mesial location. Consequently, stereotypy and integration of temporo-spatial features during an attack are essential components to seizure diagnosis.

Funding: Please list any funding that was received in support of this abstract.: No funding applicable.