From the Department of Biomedical Engineering (B.A., L.E.S., G.J., D.J.D., B.P.R., D.J.E., V.L.M.), Vanderbilt University; and Vanderbilt University Institute of Imaging Science (B.A., L.E.S., A.J., G.J., D.J.D., B.P.R., D.J.E., V.L.M.), Department of Radiology and Radiological Sciences, and Department of Neurological Surgery (D.J.E., V.L.M.), Vanderbilt University Medical Center, Nashville, TN.
Neurology. 2024 Oct 8;103(7):e209816. doi: 10.1212/WNL.0000000000209816. Epub 2024 Sep 3.
Despite the success of presurgical network connectivity studies in predicting short-term (1-year) seizure outcomes, later seizure recurrence occurs in some patients with temporal lobe epilepsy (TLE). To uncover contributors to this recurrence, we investigated the relationship between functional connectivity and seizure outcomes at different time points after surgery in these patients.
Patients included were clinically diagnosed with unilateral mesial TLE after a standard clinical evaluation and underwent selective amygdalohippocampectomy. Healthy controls had no history of seizures or head injury. Using resting-state fMRI, we assessed the postsurgical functional connectivity node strength, computed as the node's total strength to all other nodes, between seizure-free (Engel Ia-Ib) and nonseizure-free (Engel Ic-IV) acquisitions. The change over time after surgery in different outcome groups in these nodes was also characterized.
Patients with TLE (n = 32, mean age: 43.1 ± 11.9 years; 46.8% female) and 85 healthy controls (mean age: 37.7 ± 13.5 years; 48.2% female) were included. Resting fMRI was acquired before surgery and at least once after surgery in each patient (range 1-4 scans, 5-60 months). Differences between patients with (n = 30) and without (n = 18) seizure freedom were detected in the posterior insula ipsilateral to the resection (I-PIns: 95% CI -154.8 to -50.1, = 2.8 × 10) and the bilateral central operculum (I-CO: 95% CI -163.2 to -65.1, = 2.6 × 10, C-CO: 95% CI -172.7 to -55.8, = 2.8 × 10). In these nodes, only those who were seizure-free had increased node strength after surgery that increased linearly over time (I-CO: 95% CI 1.0-5.2, = 4.2 × 10, C-CO: 95% CI 1.0-5.2, = 5.5 × 10, I-PIns: 95% CI 1.6-5.5, = 0.9 × 10). Different outcome groups were not distinguished by node strength before surgery.
The findings suggest that network evolution in the first 5 years after selective amygdalohippocampectomy surgery is related to seizure outcomes in TLE. This highlights the need to identify presurgical and surgical conditions that lead to disparate postsurgical trajectories between seizure-free and nonseizure-free patients to identify potential contributors to long-term seizure outcomes. However, the lack of including other surgical approaches may affect the generalizability of the results.
尽管术前网络连通性研究在预测短期（1 年）癫痫发作结局方面取得了成功，但一些颞叶癫痫（TLE）患者仍会出现后期癫痫复发。为了揭示这种复发的原因，我们研究了这些患者手术后不同时间点的功能连通性与癫痫发作结局之间的关系。
纳入的患者经标准临床评估后被临床诊断为单侧内侧 TLE，并接受了选择性杏仁核海马切除术。健康对照组无癫痫发作或头部外伤史。使用静息态 fMRI，我们评估了无癫痫发作（Engel Ia-Ib）和有癫痫发作（Engel Ic-IV）采集之间手术后不同时间点无癫痫发作患者的功能连通性节点强度，计算方法为节点与所有其他节点的总强度。还描述了不同结局组中这些节点随时间的变化。
共纳入 32 例 TLE 患者（平均年龄：43.1±11.9 岁；46.8%为女性）和 85 名健康对照者（平均年龄：37.7±13.5 岁；48.2%为女性）。每位患者在术前和术后至少进行了一次静息 fMRI 采集（范围 1-4 次扫描，5-60 个月）。在有（n=30）和无（n=18）癫痫发作患者之间检测到与切除术同侧的后岛叶（I-PIns：95%CI-154.8 至-50.1，=2.8×10）和双侧中央脑回（I-CO：95%CI-163.2 至-65.1，=2.6×10，C-CO：95%CI-172.7 至-55.8，=2.8×10）存在差异。在这些节点中，只有那些无癫痫发作的患者在手术后节点强度增加，且随时间呈线性增加（I-CO：95%CI1.0-5.2，=4.2×10，C-CO：95%CI1.0-5.2，=5.5×10，I-PIns：95%CI1.6-5.5，=0.9×10）。术前不同结局组之间的节点强度没有差异。
研究结果表明，选择性杏仁核海马切除术 5 年内的网络演变与 TLE 的癫痫发作结局有关。这突出表明需要识别导致无癫痫发作和有癫痫发作患者术后轨迹不同的术前和手术条件，以确定长期癫痫发作结局的潜在原因。然而，缺乏其他手术方法的纳入可能会影响结果的普遍性。
