Abstracts

Rationale: After an episode of Convulsive Status Epilepticus (CSE) there is understandable concern about the harm that prolonged seizures may do to the developing brain. There is a known association between prolonged febrile seizures (PFS) and the subsequent development of mesial temporal sclerosis (MTS). Brain imaging following PFS has shown signs suggestive of acute hippocampal injury. Our longitudinal, prospective study of childhood CSE is the first to investigate the evolution of hippocampal changes and whether they are specific to PFS. Methods: Children with CSE were recruited from London hospitals and detailed clinical and demographic data collected. MRI investigations were performed in a Seimens Avanto 1.5T scanner at 1, 4 and 12 months after the episode of CSE, including a T1 weighted three-dimensional fast low angle shot (3D-FLASH), which was used to derive the hippocampal volume (HV). HV was measured by tracing consecutive coronal slices with simultaneous reference to a visualisation in 3 orthogonal planes using the 3D-FLASH dataset. Brain volume (BV) was measured by using automated software (BET - FSL) to strip the skull and surrounding soft tissue with manual correction as necessary. Each patient was assigned to an aetiological group based on their initial clinical history and examination. Controls with no previous seizures and no neurological abnormalities were also enrolled. Data was analysed in PASW 18.0 (Chicago, Illinois) for Windows using univariate ANOVA to compare groups and linear regression to model hippocampal growth against age and BV. Results: 67 patient and 19 controls were available for analysis with at least one MRI brain scan. Full demographic details are available in Table 1. None of the children developed clinical MTS during this study. After adjusting for age and BV, mean hippocampal volume 1 month after CSE was significantly related to brain volume (p < 0.001), age (p=0.035), and aetiology (p =0.011). Children with symptomatic CSE had a 206 mm3 (43-369) smaller mean HV (p=0.014) than controls. All other groups did not differ significantly from controls. In order to compare hippocampal growth over time, a model of predicted hippocampal volume was constructed. There was no significant difference in growth rates between the groups at 4 months or at 1 year. Conclusions: We found no evidence of long term hippocampal injury in children following an episode of CSE. In particular, any hippocampal swelling that occurs immediately following a PFS has resolved at 1 month and does not have a long term effect on hippocampal growth. This suggests that the risk of progression to MTS is low. Mean HV at 1 month is reduced in children with symptomatic CSE. As these children also had a range of other structural brain abnormalities, this is likely to be only one element of a wider brain disorder. Our study suggests that long term hippocampal injury in previously normal children after CSE is uncommon. Further analysis of our cohort will enable us to ascertain whether there is a subgroup of children who may show signs of long term hippocampal injury and any particular risk factors for this.