大型（Koos 4级）前庭神经鞘瘤的形态学异质性：手术后果及面神经结局的预测因素

文献信息

中文摘要

背景/目的： 显微手术是导致症状性脑干受压的大型前庭神经鞘瘤（VS）的主要治疗方法。Koos 4级肿瘤在解剖学上仍具有异质性。我们评估了手术和功能结局，并探讨一种研究特定的细分方法（改编自Hannover T4a/T4b区分）是否能提供超出既有临床和手术变量的预后信息。
方法： 这项回顾性单中心研究纳入了2013年1月至2024年12月期间接受治疗的单侧Koos 4级VS的139例成人。肿瘤被分类为Koos 4a（脑干受压但无第四脑室受压）或Koos 4b（脑干移位伴第四脑室受压）。主要终点为最后一次有记录的临床随访时的面神经功能。切除范围由术前和术后早期增强MRI确定。
结果： 75例患者（54.0%）为Koos 4a肿瘤，64例（46.0%）为Koos 4b肿瘤。Koos 4b肿瘤更大（中位数38 vs. 30 mm；p < 0.001）。73.5%实现了全切除，13.7%发生了并发症。良好的面神经功能（House-Brackmann I-II级）从术前的98.5%下降至术后即刻的36.0%，在最后一次可用随访时记录为58.7%。在探索性完整病例多变量分析（n = 104）中，较长的手术持续时间与良好面神经功能的较低几率相关（每30分钟调整OR，0.84；95% CI，0.74-0.95；p = 0.005）。这一关联不应被解释为因果关系，因为手术持续时间可能反映肿瘤复杂性。研究特定亚型在调整后与结局无关，但与肿瘤直径的共线性及模型精度有限妨碍了得出确定性结论。
结论： 这种研究特定细分描述了解剖范围和术前临床负担的差异。其独立预后价值在这项单中心队列中未得到证实，需要外部验证。手术持续时间似乎是手术复杂性的标志，而非可干预的因果危险因素。

英文摘要

Background/Objectives: Microsurgery is the principal treatment for large vestibular schwannomas (VS) causing symptomatic brainstem compression. Koos grade 4 tumors remain anatomically heterogeneous. We evaluated surgical and functional outcomes and explored whether a study-specific subdivision, adapted from the Hannover T4a/T4b distinction, provides prognostic information beyond established clinical and surgical variables. Methods: This retrospective single-center study included 139 adults treated for unilateral Koos grade 4 VS between January 2013 and December 2024. Tumors were categorized as Koos 4a (brainstem compression without fourth ventricular compression) or Koos 4b (brainstem displacement with fourth ventricular compression). The primary endpoint was facial nerve function at the last documented clinical follow-up. Resection extent was determined by preoperative and early postoperative contrast-enhanced MRI. Results: Seventy-five patients (54.0%) had Koos 4a and 64 (46.0%) Koos 4b tumors. Koos 4b tumors were larger (median 38 vs. 30 mm; p < 0.001). Gross-total resection was achieved in 73.5%, and complications occurred in 13.7%. Favorable facial nerve function (House-Brackmann grades I-II) decreased from 98.5% preoperatively to 36.0% immediately postoperatively and was documented in 58.7% at the last available follow-up. In exploratory complete-case multivariable analysis (n = 104), longer operative duration was associated with lower odds of favorable facial nerve function (adjusted OR per 30 min, 0.84; 95% CI, 0.74-0.95; p = 0.005). This association should not be interpreted causally because operative duration likely reflects tumor complexity. The study-specific subtype was not associated with outcome after adjustment, but collinearity with tumor diameter and limited model precision preclude a definitive conclusion. Conclusions: This study-specific subdivision describes differences in anatomical extent and preoperative clinical burden. Its independent prognostic value was not demonstrated in this single-center cohort and requires external validation. Operative duration appears to be a marker of surgical complexity rather than an actionable causal risk factor.