Ebel Matyas, Stevering Carmen H, Holubova Zuzana, Lequin Maarten H, Kyncl Martin, Straka Barbora, Hulshof Hanna M, Jahodova Alena, Kudr Martin, Braun Kees P J, Jansen Floor E, Krsek Pavel
Department of Paediatric Neurology, Motol Epilepsy Center, Second Faculty of Medicine, Charles University and Motol University Hospital, Prague 150 06, Czech Republic.
Department of Paediatric Neurology, University Medical Center Utrecht Brain Center, PO Box 85090 AB 3508, Utrecht, The Netherlands.
Brain Commun. 2025 Jun 26;7(4):fcaf241. doi: 10.1093/braincomms/fcaf241. eCollection 2025.
Identification of the epileptogenic lesion is challenging in tuberous sclerosis complex as multiple lesions might represent the seizure onset zone. A combination of dysplastic MRI features has diagnostic value in pre-surgical evaluation. However, these radiological characteristics may be difficult to identify and have not been studied on early unmyelinated brain MRI in tuberous sclerosis complex infants. Our study aimed to assess the diagnostic accuracy of T2-hypointense lesions on unmyelinated MRI in identifying the epileptogenic lesion. We included children with tuberous sclerosis complex who underwent resective or disconnective epilepsy surgery in the Motol University Hospital Prague and the University Medical Center Utrecht with available (i) unmyelinated MRI (before the age of 9 months), (ii) pre- and post-operative brain MRI and (iii) at least 2 years follow-up post-surgery. We identified T2-hypointense lesions and highly dysplastic lesions on unmyelinated or myelinated MRI, assessing their diagnostic accuracy in epileptogenic lesion identification by comparing seizure free to non-seizure free patients. Twenty-seven patients met inclusion criteria. We identified 54 T2-hypointense lesions in 24 patients, 30 were already highly dysplastic on unmyelinated MRI, showing cortical thickening and transmantle sign in most cases, while calcifications appeared later. Diagnostic accuracy of T2-hypointense (70.8%) was superior to the presence of the most dysplastic features (55.6%) in epileptogenic lesion identification. Positive predictive value for complete resection of all T2-hypointense lesions was 63.6%, compared to 50.0% for highly dysplastic lesions. Seizure recurrence was high (negative predictive value 76.9%) when T2-hypointense lesions remained outside the resected area. Assessing T2-hypointense lesions on unmyelinated brain MRI has important diagnostic value in identifying the epileptogenic lesion in pre-surgical work-up in infants with tuberous sclerosis complex and drug-resistant epilepsy. Unmyelinated brain MRI deserves a more important position in pre-surgical evaluation in infants with tuberous sclerosis complex and drug-resistant epilepsy.
在结节性硬化症中，确定致痫性病变具有挑战性，因为多个病变可能代表癫痫发作起始区。发育异常的MRI特征组合在术前评估中具有诊断价值。然而，这些放射学特征可能难以识别，并且尚未在结节性硬化症婴儿的早期未髓鞘化脑MRI上进行研究。我们的研究旨在评估未髓鞘化MRI上T2低信号病变在识别致痫性病变方面的诊断准确性。我们纳入了在布拉格Motol大学医院和乌得勒支大学医学中心接受切除性或离断性癫痫手术的结节性硬化症儿童，这些儿童具备以下条件：(i)未髓鞘化MRI（9个月龄之前），(ii)术前和术后脑MRI，以及(iii)术后至少2年的随访。我们在未髓鞘化或髓鞘化MRI上识别T2低信号病变和高度发育异常的病变，通过比较无癫痫发作患者和有癫痫发作患者来评估它们在识别致痫性病变方面的诊断准确性。27名患者符合纳入标准。我们在24名患者中识别出54个T2低信号病变，其中30个在未髓鞘化MRI上已经高度发育异常，大多数病例表现为皮质增厚和跨脑叶征，而钙化出现较晚。在识别致痫性病变方面，T2低信号病变的诊断准确性（70.8%）优于最发育异常特征的存在情况（55.6%）。完全切除所有T2低信号病变的阳性预测值为63.6%，而高度发育异常病变的阳性预测值为50.0%。当T2低信号病变留在切除区域之外时，癫痫复发率很高（阴性预测值76.9%）。在结节性硬化症合并耐药性癫痫的婴儿术前检查中，评估未髓鞘化脑MRI上的T2低信号病变在识别致痫性病变方面具有重要诊断价值。未髓鞘化脑MRI在结节性硬化症合并耐药性癫痫的婴儿术前评估中应占据更重要的地位。
