William O. Tatum IV, DO
Mayo Clinic Florida, Jacksonville, FL
A 38-year-old female with ADHD, anxiety-depression, sexual abuse (accused), and refractory seizures presented with seizure onset at 4 years of age. No prior risk factors for epilepsy were present. Early seizures were described as “grand mal” seizures. She reportedly had an EEG which was abnormal and she was treated and failed multiple antiseizure drugs (ASDs) and seizures continued. Initially they occurred daily but over the years several semiologies became noted and occurred 1-2 times per week upon evaluation. Over the years her mother served as her primary observer. She was noted to have “spacey spells”, seizures with choking and gasping, others with violent shaking, and some events where she would “quite breathing”. Several hospitalizations were performed with video-EEG monitoring (VEM). Each one captured multiple spontaneous habitual events that reflected psychogenic non-epileptic attacks (PNEA). Brain MRI was repeatedly normal. An EEG was reportedly abnormal. She was subsequently re-evaluated at our institution on carbamazepine and clonazepam.
Her neurological examination was normal. A high-resolution brain MRI with an epilepsy protocol was normal. A routine scalp EEG demonstrated left mid-temporal epileptiform discharges (Figure 1). She was admitted to Mayo Clinic Florida’s Epilepsy Monitoring Unit for differential diagnosis. Spontaneous events on day 1 and triggered events using hyperventilation and photic stimulation on day 2 reproduced clear PNEA typical of her outpatient events.
Figure 1: Routine scalp EEG with a 3 second burst of repetitive left temporal-parietal spike-and-wave complexes repeating at 1 Hz in second 6-8.
Video-EEG recording of the habitual event is the gold standard for the diagnosis of PNEA. However, the scalp interictal EEG demonstrated epileptiform discharges producing discordant features. One possible explanation is incorrect interpretation of a normal variant such as wicket spike on EEG in a patient with PNEA. Wickets consist of intermittent asymmetric independent bitemporal synchronous and asynchronous bursts of monophasic arciform potentials. They bear no relationship to epilepsy though are commonly the subject of misinterpretation and treatment (1). Discrete isolated spike-and-slow waves in our patient suggested the specificity for epilepsy was >90% with only 1-2% of people not manifesting seizures. Yet all of her reported clinical semiologies were identified as PNEA during VEM suggested controlled epilepsy at the time of evaluation. An additional possibility unfolded with VEM. With her last admission for VEM an aggressive taper of her ASDs was performed and disclosed a semiology that was unrecognized by the patient and unidentified by observers as epileptic. An electroclinical subtle focal seizure without awareness (SWA) was captured on day 3 (Figure 2). This was misidentified as a non-epileptic behavioral event when the patient manifested subtle brief manual automatisms. The event occurred with mild brief unresponsiveness and without patient awareness. Unless ASD taper had reproduced the subtle electroclinical seizures without awareness the presence of “active” epilepsy would otherwise have evaded detection.
Figure 2: Ictal scalp EEG recording of a left temporal subtle focal seizure with subtle looking and brief unresponsiveness and manual automatisms though without awareness of the event.
A trap involved in VEM is when all seizure types are not captured. Focal seizures have variable impairment of consciousness and may retain partial awareness during the ictus. However, seizures without awareness (SWA) that go unrecognized by patients are common with 30% of epilepsy patients who are never aware of their seizures identified during VEM (2). Nonclinical seizures are subclinical and only identifiable on the EEG without an associated behavioral or subjective clinical correlate and differ from SWA. Our patient with SWA was having SWA weekly after retrospective review with her mother of the captured event. Because our patient and her observers did not recognize them, and because PNEA continued to be uncontrolled, recognition of her subtle epileptic seizures was masked. Seizure awareness has a neuroanatomical foundation in the function of memory encoding and recall and therefore not unexpectedly occurs commonly in temporal lobe epilepsy (2). When seizures have a subtle semiology that is combined with patient unawareness, a definitive diagnosis of epilepsy may evade detection for years (3). Video-EEG is the means not only to diagnosis in this case but also critical in following the response to therapy.