Abstracts

Rationale: Post-traumatic epilepsy (PTE) is a common cause of morbidity in children after a traumatic brain injury (TBI), occurring in 10-20% of children following severe TBI.^{1, 2} In the present study, we describe 47 patients who were diagnosed with PTE at a tertiary pediatric referral center. Herein, we aim to understand seizure semiology in children following TBI, while analyzing the clinical variables in this group. Methods: A retrospective electronic chart review of patients who had suffered TBI and subsequently referred to any one of the seven child neurologists at Children's Hospital of Michigan from years 2002 to 2012 was performed. A total of 321 charts were reviewed and, of these, 47 patients (15%) were diagnosed with PTE. Multiple clinical variables were reviewed. Seizures were categorized using the ictal semiological classification (Lüders et al. 1998), based on the history and/or available video-EEG data.³ Results: Of the 47 children, age 1 day to 15 years, 31 (66%) were males and 16 (34%) females. Eight (17%) had mild and 39 (83%) had severe TBI. No patients with mild TBI had neurosurgical intervention or were placed on any empiric therapy for seizures within 24 hours of injury. Seven patients with mild TBI had normal head computed tomography (CT) on the day of TBI, and 1 had no imaging. All patients with severe TBI had abnormal CT. Twenty-one (54%) of the 39 patients with severe TBI had non-accidental trauma (NAT) whereas, all of the 8 patients with mild TBI had accidental trauma (AT). Age at the time of NAT ranged from19 days to 15 years. Seven (88%) of 8 patients with mild TBI were on one AED. In the severe TBI group of 39 patients, fifteen (38%) were on one AED, 19 (49%) were on two or more AEDs, and 3 (8%) were off AED. Nineteen (59%) of the 32 patients with severe TBI had one seizure type, 8 (25%) had two, and 3 (9%) had three seizures types. Six (19%) patients had infantile spasms and 1 of them showed evolution to Lennox-Gastaut Syndrome. Epileptic spasms were not seen in the patients with mild TBI. Conclusions: The 15% prevalence of PTE in our TBI population is in good agreement with previous studies, as is the frequency of posttraumatic seizures after inflicted versus noninflicted TBI. Patients with severe TBI had increased number of seizure types, including epileptic spasms, and hypomotor seizures (ILAE: complex partial seizures). ⁴ The latter were the commonest in both mild and severe TBI subgroups. In the severe TBI subgroup, there was a high prevalence of patients who suffered NAT; epileptic spasms were common in this subgroup which tended to be younger. This association requires further investigation. No patient with mild TBI had more than two seizure types, and none had epileptic spams. Limitations of our study include the small sample size and retrospective design. The current findings add to the paucity of knowledge of seizure semiology in this group of patients, and emphasize the need for biomarkers to predict the occurrence of posttraumatic epilepsy.