F80. Seizures and hyperexcitable EEG patterns in spontaneous deep intraparenchymal hemorrhage

Abstract

Introduction The incidence of electrographic seizures after non-traumatic intraparenchymal hemorrhage (IPH) ranges from 1.7% to 31%. It is unclear if the risk of seizures or hyperexcitable patterns (HEPs: any rhythmic delta activity except generalized, any periodic discharges, or any spike-wave pattern, using ACNS criteria) for deep IPH is similar to lobar, subarachnoid and subdural hemorrhages. We hypothesized that the incidence of seizures or hyperexcitable patterns (SZ or HEPs) in deep IPH is lower than reported, but thalamic involvement may confer higher risk due to its role in physiologic and pathologic rhythmic activity, such as sleep spindles and generalized spike wave discharges. Methods On retrospective review, 45 patients had deep IPH (defined as IPH not involving cortex/juxtacortical regions other than the insula (analyzed separately), with or without intraventricular hemorrhage IVH) and underwent continuous EEG (cEEG) between 1/2013 and 12/2016. Patients with involvement of cortex, subarachnoid or subdural areas were excluded. Age, sex, prior history of epilepsy, clinical seizure at ictus, anti-seizure drugs (ASDs), sedative infusions during EEG and surgical interventions such as ventriculostomy (EVD) or decompressive hemicraniectomy (DHC) were reviewed. The cEEG closest to admission was reviewed for seizures, HEPs, and generalized rhythmic delta activity (GRDA). The head CT in closest proximity to the index cEEG was reviewed for hematoma volume, thalamic involvement and insular involvement. Results Two of 45 patients had SZ, 7/45 had HEPs, 1/45 had both and a total of 8/45 had SZ or HEPs. Of the HEPs , LRDA was seen in 4/8, LPDs in 2/8, both LPDs and LRDA in 1/8. Of those with the lateralized HEPs, 4/7 were contralateral to IPH, 1/7 (14.2%) were bilateral and 2/7 were ipsilateral. Of the 5/7 with HEPs contralateral to the IPH, 2 had an EVD ipsilateral to the HEPs. Patients with SZ or HEPs had significantly higher hematoma volume compared to patients who did not have SZ or HEPs (32.7 ± 19.4 vs 15.4 ± 17.7, p = 0.02). Patients who underwent surgery (DHC/EVD or both) had higher incidence of HEPs or SZ compared to those who did not (7/23, 30.4% vs 1/22, 4.5%, p = 0.047). DHC was associated with significantly higher risk of SZ or HEPs (4/4, 100% vs 4/41, 9.8%, p Conclusion Risk of SZ or HEPs with deep IPH ± IVH was 17.7% and correlated with mean hematoma volume and DHC. The, risk in the absence of a surgical intervention was low (1/22, 4.5%, vs 7/23, 30.4% p = 0.047). Thalamic involvement did not correlate with SZ or HEPs in this small cohort of patients with deep IPH.