Abstracts

Rationale:
Focal cortical dysplasias (FCD) are a common cause of pediatric epilepsy. Yet, there are children with FCD who do not develop epilepsy. Location may play a role in epileptogenesis- a small lesion may disrupt normal network function leading to seizures. The purpose of this study is to evaluate differences in network distribution of FCD in children with and without epilepsy, and to see if there is preferential distribution to one of seven established distributed cognitive networks.
Method:
61 FCD epilepsy patients (mean age at abnormal scan +/- SD (range) 9.57+/-5.6 yr (0.24-21), 29 female, mean age onset 5.7+/- 4.6 yr (0.002-16), 28 left hemisphere) and 14 FCD patients without epilepsy (NonEPI, mean age 10.66+/-5.84 yr (1.38-17.75), 7 female, 11 left hemisphere) were included in this study. Epilepsy patients are from a retrospective cohort who underwent resective epilepsy surgery between January 1, 2009 – December 31, 2019 at Children’s National Hospital. All had pathological confirmation and categorization of FCD. NonEPI FCD patients were identified from a neuroradiology MRI database of children found to have FCD from January 2011 to January 2019. All patients had a high-resolution MRI (1.5 Tesla before 2011 and 3 Tesla afterwards), including specialized infant sequences for those < 24 months.  FCD 3D volumes were hand-drawn from either T2 or SPGR images in patients’ native space using Mricrogl software. Selection was based on whichever MRI sequence best delineated the abnormality. ROIs were then projected onto standard Montreal Neurological Institute (MNI) space in SPM 12 by applying the deformation field generated through the normalization process of T1. We examined the pattern of each individual patient’s FCD in relation to 7 distributed cortical networks by assigning each subject into one of the 7 networks (FCD Dominant Residing Network), based on their maximal extent of FCD overlap with the networks. We separated FCD epilepsy patents into three age onset groups (< 1 yr N=28, >1 and < 5 yr N=21, >5 yr N=12). We constructed a crosstab table of Groups x Networks and tested whether the proportion/distribution of patients are different through Chi-square or Fisher exact test where appropriate.
Results:
Within epilepsy patients, most Dominant Residing Networks are equally represented across age onset groups (Table 1). For those with epilepsy, FCD in Default Mode Network (DMN), sensorimotor, and visual networks are associated with younger age onset (χ2=20.2, p=0.06, df=12). Patients with older age onset FCD had predilection for the limbic network, which includes anterior inferior temporal lobe, anterior parahippocampus and medial orbital frontal lobe. Those without epilepsy had no FCD in the limbic network (Table 2 Fisher exact test, p=0.03), but had similar distributions as those with younger age onset of epilepsy.
Conclusion:
FCD in younger epilepsy onset patients preferentially occur in visual, sensorimotor and DMN networks that may reflect earlier structural maturation of these networks. FCD in older patients are more likely to reside in the limbic network that may reflect later maturation. FCD in those without epilepsy may be located in DMN, but are less likely to occur in visual or limbic networks.
Funding:
:Hess Foundation