Mandonnet Valéry, Rheault François, Barberis Marion, Prevost Cécile, Letrange Sophie, Poisson Isabelle, Froelich Sébastien, Mandonnet Emmanuel
Frontlab, Paris Brain Institute, CNRS UMR 7225, INSERM U1127, Paris, France.
Department of Neurosurgery, Lariboisière Hospital, Paris, France.
Acta Neurochir (Wien). 2024 Dec 18;166(1):507. doi: 10.1007/s00701-024-06374-7.
To provide an explanation for the intraoperative onset of severe naming deficits in the course of awake resection of left insular glioma.
We retrospectively reviewed a series of 14 patients operated on in awake conditions for a left insular IDH-mutated glioma. Preoperative MRI included high-resolution diffusion sequences, to which constrained spherical deconvolution pipeline was applied, to obtain a whole brain tractogram. Whole brain T1 parcellation was obtained by Freesurfer, allowing to dissect the tractogram and identify the connections between the caudate nucleus and Broca's area. Postoperative MRI standard diffusion the day after surgery was performed, allowing to delineate and register to the preoperative MRI any area of hyperintense diffusion with low apparent diffusion coefficient. The pathway between pars triangularis (resp. opercularis) and Broca's area were considered as damaged whenever more than 50% of streamlines were passing through the mini-strokes. Patients' language abilities (including a picture naming task) were assessed and reported before, during and after surgery by certified speech therapists. Severe postoperative naming deficits were defined as a score lower than 40/80 items. Contingency tables were analyzed with Fisher exact test (statistical significance set at 0.05).
Out the 14 patients, 8 patients had a mini-stroke on the immediate postoperative MRI. None of the 6 patients without any stroke had postoperative naming severe deficits. Five out the 8 patients with a mini-stroke had a severe postoperative naming deficit, characterized by strong verbal perseverations. This difference was statistically significant (exact Fisher test, p = 0.03). For the five patients with a deficit, the mini-stroke damaged either the pars triangularis - caudate pathway or the pars opercularis - caudate pathway, whenever the pars triangularis was resected after negative cortical mapping. For the three patients without severe postoperative naming deficit, the mini-stroke spared the Broca-caudate pathway. All patients recovered quasi-normal naming abilities at the 4-month postoperative evaluation.
The occurrence of mini-strokes within the connections between Broca's area and the caudate nucleus explains the sudden naming deficits observed intraoperatively in some patients during awake resection of IDH-mutated insular glioma. Further studies are needed to better predict such event and to assess its impact on other cognitive functions.
为左岛叶胶质瘤清醒切除术中严重命名障碍的术中发作提供一种解释。
我们回顾性分析了14例在清醒状态下接受手术的左岛叶异柠檬酸脱氢酶（IDH）突变型胶质瘤患者。术前磁共振成像（MRI）包括高分辨率扩散序列，并应用受限球形去卷积管道，以获得全脑纤维束成像。通过Freesurfer获得全脑T1分割图像，以便剖析纤维束成像并识别尾状核与布洛卡区之间的连接。术后第一天进行标准扩散MRI，以描绘并将任何表观扩散系数低的高扩散区域与术前MRI进行配准。当超过50%的纤维束穿过微小梗死灶时，三角部（或盖部）与布洛卡区之间的通路被视为受损。由专业言语治疗师在手术前、手术期间和手术后评估并报告患者的语言能力（包括图片命名任务）。术后严重命名障碍定义为得分低于40/80项。列联表采用Fisher精确检验进行分析（统计学显著性设定为0.05）。
在14例患者中，8例患者术后即刻MRI出现微小梗死灶。6例无任何梗死灶的患者均无术后严重命名障碍。8例有微小梗死灶的患者中有5例出现严重的术后命名障碍，其特征为强烈的言语持续性。这种差异具有统计学显著性（Fisher精确检验，p = 0.03）。对于5例有命名障碍的患者，当三角部在皮质映射阴性后被切除时，微小梗死灶损害了三角部-尾状核通路或盖部-尾状核通路。对于3例无严重术后命名障碍的患者，微小梗死灶未累及布洛卡-尾状核通路。所有患者在术后4个月评估时恢复了近似正常的命名能力。
布洛卡区与尾状核之间连接内微小梗死灶的出现解释了在IDH突变型岛叶胶质瘤清醒切除术中一些患者术中观察到的突然命名障碍。需要进一步研究以更好地预测此类事件并评估其对其他认知功能的影响。
