High-frequency oscillations in epilepsy and surgical outcome. A meta-analysis.
High-frequency oscillations in epilepsy and surgical outcome. A meta-analysis.
复制标题
DOI：
10.3389/fnhum.2015.00574
复制
发表时间：
2015
影响因子：
2.9
通讯作者：
Trinka E
中科院分区：
文献类型：
--
作者：
Höller Y;Kutil R;Klaffenböck L;Thomschewski A;Höller PM;Bathke AC;Jacobs J;Taylor AC;Nardone R;Trinka E
高频振荡（hfo）被认为是一种潜在的致痫性标志物。目前的研究努力寻找有效的证据，证明这些hfo可以帮助划定难治性癫痫患者的待切除区域，并改善手术结果。在当前的荟萃分析中，我们评估了切除可检测到hfo的区域与癫痫手术后预后之间的关系。我们对所有与hfo发生区域切除与术后结果相关的研究进行了系统回顾。我们将结果（癫痫发作自由）与切除比联系起来，即检测到hfo的通道数量与其中切除区域内通道数量之间的比率。我们比较了无癫痫发作和非癫痫发作患者的切除率。总共纳入了11项研究。在10项研究中，波纹（80-200赫兹）被分析，在7项研究中，快速波纹（>200赫兹）被研究。我们发现波纹（dif = 0.18; CI: 0.10-0.27）和快速波纹（dif = 0.17; CI: 0.01-0.33）的结果组之间的切比具有可比性差异（dif）和大部分重叠的置信区间（CI）。亚组分析显示，自动检测（dif = 0.22; CI: 0.03-0.41）与目视检测（dif = 0.17; CI: 0.08-0.27）相当。考虑hfo的频率（dif = 0.24; CI: 0.09-0.38）与结果的相关性比考虑每个电极显示hfo （dif = 0.15; CI = 0.03-0.27）更强。在荟萃分析中发现的效应量很小但很重要。自动检测和应用检测阈值以检测频繁发生hfo的通道对于产生可用于手术前评估的标记物非常重要。为了比较采用不同方法的研究，需要详细和标准化的报告。
High frequency oscillations (HFOs) are estimated as a potential marker for epileptogenicity. Current research strives for valid evidence that these HFOs could aid the delineation of the to-be resected area in patients with refractory epilepsy and improve surgical outcomes. In the present meta-analysis, we evaluated the relation between resection of regions from which HFOs can be detected and outcome after epilepsy surgery. We conducted a systematic review of all studies that related the resection of HFO-generating areas to postsurgical outcome. We related the outcome (seizure freedom) to resection ratio, that is, the ratio between the number of channels on which HFOs were detected and, among these, the number of channels that were inside the resected area. We compared the resection ratio between seizure free and not seizure free patients. In total, 11 studies were included. In 10 studies, ripples (80–200 Hz) were analyzed, and in 7 studies, fast ripples (>200 Hz) were studied. We found comparable differences (dif) and largely overlapping confidence intervals (CI) in resection ratios between outcome groups for ripples (dif = 0.18; CI: 0.10–0.27) and fast ripples (dif = 0.17; CI: 0.01–0.33). Subgroup analysis showed that automated detection (dif = 0.22; CI: 0.03–0.41) was comparable to visual detection (dif = 0.17; CI: 0.08–0.27). Considering frequency of HFOs (dif = 0.24; CI: 0.09–0.38) was related more strongly to outcome than considering each electrode that was showing HFOs (dif = 0.15; CI = 0.03–0.27). The effect sizes found in the meta-analysis are small but significant. Automated detection and application of a detection threshold in order to detect channels with a frequent occurrence of HFOs is important to yield a marker that could be useful in presurgical evaluation. In order to compare studies with different methodological approaches, detailed and standardized reporting is warranted.