Seizure outcome in surgically treated pediatric gangliogliomas and dysembryoplastic neuroepitheliomas according to imaging and resection strategies

Julia Shawarba,Karl Roessler,Matthias Tomschik,Jonathan Wais,Fabian Winter,Florian Mayer,Gregor Kasprian,Christine Haberler,Tatjana Traub-Weidinger,Martin Niederle,Thomas Czech,Johannes Herta,Christian Dorfer,Martha Feucht
DOI: https://doi.org/10.1016/j.seizure.2024.09.002
IF: 3.414
Seizure
Abstract:Purpose: Imaging and resection strategies for pediatric gangliogliomas (GG) and dysembryoplastic neuroepitheliomas (DNET) presenting with epilepsy were retrospectively analyzed in a consecutive institutional series of surgically treated patients. Methods: Twenty-two children (median 8 years, 3-18 years) presented with seizures for 30 months median (14-55.2 months) due to a histologically verified GG/DNET. Results: There were 20 GG and 2 DNT, 68 % located temporal, 32 % extra-temporal. Seizure history was significantly longer in temporal cases (38 versus 14 months median, p < 0.01). MRI contrast enhancement was present in 50 % and methionine (MET) positron emission tomography (PET) uptake in 70 % (standard uptake values (SUVs) 2.92 mean, from 1.6 to 6.4). 27 % had glucose PET hypometabolism. Primarily, in temporal GG, ECoG (electrocorticography) -guided lesionectomies were performed in 87 % and antero-mesial temporal lobe resections (AMTLR) in 13 %, whereas in extra-temporal GG/DNETs, lesionectomies were performed in 100 %. ILAE Class 1 seizure outcome was primarily achieved in 73 % of the temporal cases, and was increased to 93 % by performing six repeat surgeries using AMTLR. Extratemporal patients experienced ILAE Class 1 seizure outcomes in 86 % without additional surgeries, although harboring significantly more residual tumor (p < 0.005, mean follow-up 28 months). Conclusion: In children, MET PET imaging for suspected GG is proposed preoperatively showing a high diagnostic sensitivity and an option to delineate the lesions for navigated resection, whereas MRI contrast behavior was of no differential diagnostic use. As a surgical strategy we propose primarily lesionectomies for extratemporal but AMTLR for temporal GG respecting eloquent brain areas.
What problem does this paper attempt to address?