Late Post-traumatic Epilepsy in Children and Young Adults: Impropriety of Long-Term Antiepileptic Prophylaxis and Risks in Tapering.

Background: After traumatic brain injury, epilepsy affects up to 20 % of children. It is a risk factor, for both clinical recovery and cognitive performance; therefore pharmacological therapy is advisable. Current guidelines recommend prophylaxis to be initiated as soon as possible and tapered 1 week after trauma. However, no guideline exists for paediatric patients and the clinical practice is heterogeneous.
Objective: In our institute, prophylaxis was routinely tapered 6 months after trauma. Therefore we investigated whether this prophylaxis or its tapering influenced the development of post-traumatic epilepsy, together with several clinical-demographic factors.
Methods: The study population comprised all patients with post-traumatic brain injury referred to this institute between 2002 and 2009 who consented to participate. Clinical, epileptological and pharmacological data were collected. The role of prophylaxis and several other predictors on occurrence of post-traumatic epilepsy was analysed through logistic regressions.
Results: Two hundred and three patients (145 paediatric) were followed for 57 months on average. Risk factors for epilepsy were past neurosurgery [odds ratio (OR) = 2.61, 95 % confidence interval (CI) 1.15-5.96], presence of epileptiform anomalies (OR = 6.92, 95 % CI 3.02-15.86) and the presence of prophylaxis (OR = 2.49, 95 % CI 1.12-5.52), while higher intelligence quotient (IQ) was protective (OR = 0.96, 95 % CI 0.95-0.98). While evaluating possible different effects within and after 6 months (tapering, for those under prophylaxis), we found that epileptiform anomalies (OR = 7.61, 95 % CI 2.33-24.93, and OR = 8.21, 95 % CI 3.00-22.44) and IQ (OR = 0.96, 95 % CI 0.94-0.98, and OR = 0.97, 95 % CI 0.95-0.98) were always significant predictors of epilepsy, while neurosurgery (OR = 4.38, 95 % CI 1.10-17.45) was significant only within 6 months from trauma, and prophylaxis (OR = 3.98, 95 % CI 1.62-9.75) only afterwards.
Conclusions: These results suggest that prophylaxis was irrelevant when present; furthermore its tapering increased the risk of epilepsy. Since the presence of epileptiform anomalies was the main predictor of post-traumatic epilepsy, such anomalies may be useful to better direct the choice of prophylaxis.