Language network reorganization before and after temporal lobe epilepsy surgery

Abstract

OBJECTIVEEpilepsy surgery is the recommended treatment option for patients with drug-resistant temporal lobe epilepsy (TLE). This method offers a good chance of seizure freedom but carries a considerable risk of postoperative language impairment. The extremely variable neurocognitive profiles in surgical epilepsy patients cannot be fully explained by extent of resection, fiber integrity, or current task-based functional MRI (fMRI). In this study, the authors aimed to investigate pathology- and surgery-triggered language organization in TLE by using fMRI activation and network analysis as well as considering structural and neuropsychological measures.METHODSTwenty-eight patients with unilateral TLE (16 right, 12 left) underwent T1-weighted imaging, diffusion tensor imaging, and task-based language fMRI pre- and postoperatively (n = 15 anterior temporal lobectomy, n = 11 selective amygdalohippocampectomy, n = 2 focal resection). Twenty-two healthy subjects served as the control cohort. Functional connectivity, activation maps, and laterality indices for language dominance were analyzed from fMRI data. Postoperative fractional anisotropy values of 7 major tracts were calculated. Naming, semantic, and phonematic verbal fluency scores before and after surgery were correlated with imaging parameters.RESULTSfMRI network analysis revealed widespread, bihemispheric alterations in language architecture that were not captured by activation analysis. These network changes were found preoperatively and proceeded after surgery with characteristic patterns in the left and right TLEs. Ipsilesional fronto-temporal connectivity decreased in both left and right TLE. In left TLE specifically, preoperative atypical language dominance predicted better postoperative verbal fluency and naming function. In right TLE, left frontal language dominance correlated with good semantic verbal fluency before and after surgery, and left fronto-temporal language laterality predicted good naming outcome. Ongoing seizures after surgery (Engel classes ID–IV) were associated with naming deterioration irrespective of seizure side. Functional findings were not explained by the extent of resection or integrity of major white matter tracts.CONCLUSIONSFunctional connectivity analysis contributes unique insight into bihemispheric remodeling processes of language networks after epilepsy surgery, with characteristic findings in left and right TLE. Presurgical contralateral language recruitment is associated with better postsurgical language outcome in left and right TLE.

Mapping of Language-and-Memory Networks in Patients With Temporal Lobe Epilepsy by Using the GE2REC Protocol

“…This mesial temporal implication aligns with previous studies showing the mesial temporal contribution to word retrieval during picture naming and word fluency tasks (Bonelli et al, 2011;Hamamé et al, 2014). On the other hand, P2 showed greater righthemispheric involvement after surgery, as revealed by sentence generation and recall tasks, in line with previous findings on reorganization patterns (Bonelli et al, 2012;Foesleitner et al, 2021). In the P2 case, the reorganization was not cognitively efficient, possibly related to weaker coupling between left and right frontal regions (Bonelli et al, 2012;Foesleitner et al, 2021) probably induced by late ASO.…”

“…Although inter-hemispheric network reorganization is a common finding in LTLE (Goldmann and Golby, 2005;Powell et al, 2007;Cousin et al, 2008;Hamberger and Cole, 2011;Bonelli et al, 2012;Sidhu et al, 2013;Baciu and Perrone-Bertolotti, 2015;Torlay et al, 2017;Foesleitner et al, 2021), our patients did not show an evident inter-hemispheric reorganization as revealed by group-level analyses. This result can be explained by the fact that our patients had late age of seizures onset (ASO, see Table 1) generally associated with intrahemispheric reorganization, compared to patients with early ASO who more frequently show inter-hemispheric reorganization (Baciu and Perrone-Bertolotti, 2015).…”

“…Since the clinical utility of an fMRI protocol is highly associated with its ability to identify the dominant hemisphere for a specific function and the network of regions engaged in that function ( Benjamin et al, 2020 ), the analyses employed in this study aimed to determine the regions active during interactive language and memory GE2REC tasks. Nevertheless, other analyses could also be applied to the fMRI data obtained with this protocol, such as functional connectivity ( Banjac et al, 2021 ) that can also contribute to the presurgical evaluation of epilepsy patients ( Foesleitner et al, 2021 ).…”

“…Nevertheless, other mapping methods using fMRI can also be used in the presurgical evaluation. For instance, studies show clinical application of functional connectivity ( Bettus et al, 2010 ; Pravatà et al, 2014 ; Tracy and Doucet, 2015 ; DeSalvo et al, 2020 ; Foesleitner et al, 2021 ) and effective connectivity ( Vaudano et al, 2021 ).…”

Can intraoperative electrocorticography be used to minimize the extent of resection in patients with temporal lobe epilepsy associated with hippocampal sclerosis?

“…We should consider an epileptic mechanism with functional and/or pathological networking. [25][26][27][28] When considering the seizure mechanism based on the epilepsy network, cases in which seizures ceased after SA are those in which the epileptic network was limited to the resected structures, whereas cases that require ATL are those with a wider network. We should also consider the running-down phenomenon after epileptic focus resection.…”