Abstracts

Rationale: Children with infantile spasms (IS) have a high risk of mortality. We evaluated the factors leading to mortality and causes of early death (< 5 years) in children with IS. Methods: We identified and retrospectively reviewed cases of IS (active or inactive) followed at our tertiary pediatric hospital between 2000-2010. Inclusion required an EEG with hypsarrhythmia or a pattern compatible with IS. Subjects with no follow-up after initial care were excluded. We collected demographics, history, etiology, seizure history prior to IS, details and persistence of IS, and need for a g-tube. Developmental outcome was estimated based on chart review, classified as either "normal/mildly delayed" or "severely affected." For those who died before 5 years of age, causes of death were assigned by author consensus from death certificates/autopsy reports as available, medical records, or from the treating neurologist. Comorbidities were categorized as "significant cardiorespiratory" (including patients with tracheostomy), "other systemic" (including significant sleep apnea), or "less severe systemic involvement" (e.g., renal problems, mild cardiac abnormalities). Clinical features associated with death in children with IS (active or inactive) were examined. Results: Among 355 patients screened for IS, 184 patients (108 males) met inclusion criteria (among excluded patients, 63 lacked adequate follow-up and 108 lacked confirmation of IS). Of the 184, 11 children (6.0%) died before 5 years of age and 7 died after 5 years. Mean age of survivors at last follow-up was 7.2 ± 2.6 years. All children who died had severe developmental delay. Those who died were more likely to have persistence of IS despite treatment compared to survivors (47.4 vs. 20.6%, p<0.018), presence of a g-tube (64.3 vs. 29.8%, p<0.014), and significant cardiac/respiratory and or other system comorbidity (66.7 vs. 8.2%, p<0.0001). Among the 11 children with death before 5 years, respiratory failure was the immediate cause of death in 8 (73%) cases (4 who died in hospital, 1 with mention on the death certificate and 3 in whom care was directed to comfort measures only). Of the remaining 3 patients, one patient each died from status epilepticus, gastroenteritis with cardiac arrest, and unknown cause. Two patients had epilepsy-related causes of death, one with status epilepticus and another with seizure-induced worsening of respiratory status. Two of the 11 deaths had progressive metabolic disorders, (Menkes disease in one and an undiagnosed metabolic disorder in another). In 8/11 cases (73%) with early death, the families had requested "comfort care only" measures or withdrawal of life support. Conclusions: Mortality in IS is related to significant underlying comorbidities and severity of neurological impairment. In children with early mortality, terminal cause of death in many was often respiratory failure, which in most cases was related to the family's decision to redirect toward comfort care.