Distinctive MRI features of the epileptogenic zone in children with tuberous sclerosis

• Published in: European journal of radiology Vol. 83; no. 4; pp. 703 - 709
• Main Authors: Jahodova, A, Krsek, P, Kyncl, M, Jezdik, P, Kudr, M, Komarek, V, Jayakar, P, Miller, I, Resnick, T, Duchowny, M
• Format: Journal Article
• Language: English
• Published: Ireland Elsevier Ireland Ltd 01.04.2014
• Subjects: Adolescent; Brain - pathology; Child; Child, Preschool; Cortical dysplasia; Epilepsy - etiology; Epilepsy - pathology; Epilepsy surgery; Female; Humans; Infant; Infant, Newborn; Magnetic resonance; Magnetic Resonance Imaging - methods; Male; Nerve Net; Radiology; Reproducibility of Results; Sensitivity and Specificity; Tuberous Sclerosis - complications; Tuberous Sclerosis - pathology; Tuberous sclerosis complex; Cortical dysplasia; Epilepsy surgery; Magnetic resonance; Tuberous sclerosis complex
• ISSN: 0720-048X; 1872-7727
• DOI: 10.1016/j.ejrad.2013.12.024

Abstract Objective Localization of the epileptogenic zone (EZ) is challenging in children with tuberous sclerosis complex (TSC). We sought to ascertain whether brain MRI could identify the EZ in TSC patients independent of the clinical and diagnostic data. Methods Presurgical MRI's of 34 children with TSC who underwent epilepsy surgery at Miami Children's Hospital were retrospectively reevaluated by experts blinded to all other data. Changes typical of TSC (tubers, calcifications, cystic changes) and abnormalities of the perituberal cortex typical of focal cortical dysplasia (FCD) (increased cortical thickness, abnormal gyration, transmantle change, gray/white matter junction blurring) were identified and their localization was compared with the resection site. Sensitivity, specificity and accuracy of individual MRI features to localize the EZ were determined and statistically compared between postoperatively seizure-free and non-seizure-free patients as well as clusters of features typical of FCD and TSC. Results MRI alone correctly localized the resection cavity in all 19 postoperatively seizure-free patients and 12 of 15 non-seizure-free subjects. Sensitivity, specificity and accuracy of MRI features typical of FCD to localize EZ (90%, 96% and 96%, respectively) were superior to those typical of TCS (79%, 75% and 75%, p < 0.0001). Increased cortical thickness and abnormal gyral formation outside tubers occurred only in the resection site. Resection sites were better predicted by MRI in seizure-free than in non-seizure-free patients. Conclusion Thorough MRI evaluation identifies the EZ in a significant proportion of TSC patients. Epileptogenic regions were mostly characterized by “FCD-like” changes outside cortical tubers. The findings may have important practical consequences for surgical planning in TSC.