关于听神经瘤切除最适当的入路问题至今仍有争论。一般的观点认为:较小的肿瘤,患侧听力损失严重者可经迷路切除肿瘤,而较大的病变最好经枕下入路切除。本文作者100例经迷路手术的经验却与此相反。认为对较大的肿瘤,经迷路手术面神经保留的机会大且病残率低,故喜欢选用此入路。在文中作者介绍了100例经迷路手术的效果,术中防止脑脊液漏的方法,并就保留面神经和挽救听力的有关问题进行了讨论。100例肿瘤中,小型肿瘤(内听道内)4例,中型:(<2.5cm,达内听道外)33例,大型(>2.5cm)66例。复发性肿瘤3例。手术按House 所述的经迷路手术进行。术前不作脑室引流及腰穿引流,不常规使用类固醇。97例肿瘤全切,追踪1～7年无1例复发。手术死亡3例,其肿瘤直径均大于3.5cm,死因分别为脑干梗塞、桥小脑角血肿及肺栓塞。本组早期13例发生脑脊液漏,其有效防止方法为:用骨尘与血混合充填中耳腔及其入口,用阔筋膜和纤维蛋白胶密封,再用脂肪填塞颞骨缺损。5例大型肿瘤病人发生短暂性球麻痹,均完全恢复。面神经保留率82%。58例需行暂时性睑缘缝合。术后一年面神经功能正常14例(House Ⅰ级),39例功能较好 The debate on the most appropriate approach to the removal of acoustic neuromas remains controversial. The general point of view is that in smaller tumors, those with severe hearing loss on the affected side may be treated to get rid of the tumor through a labyrinth, while larger lesions are best removed through the suboccipital approach. The author of this article has 100 patients with laparotomy who have the opposite experience. It is considered that for larger tumors, the chance of facial nerve retention through labyrinth surgery is large and the morbidity is low, so we prefer to use this approach. In the article, the author introduced the results of 100 cases of labyrinth surgery, prevention of cerebrospinal fluid leakage during surgery, and discussed related issues of facial nerve preservation and hearing rescue. Among the 100 tumors, 4 were small tumors (internal auditory canal), 33 were moderate (<2.5 cm, outside the auditory canal), and 66 were large (>2.5 cm). 3 cases of recurrent tumors. Surgery was performed as described by House’s labyrinth surgery. No ventricular drainage and lumbar puncture were performed before surgery, and steroids were not routinely used. Ninety-seven cases of tumors were completely resected and no recurrence was observed in 1 to 7 years. Three patients died of surgery and their tumor diameters were all greater than 3.5cm. The causes of death were brainstem infarction, hematomas of cerebellopontine angle, and pulmonary embolism. Cerebrospinal fluid leakage occurred in 13 cases in the early stage of this group. The effective prevention method was: filling the middle ear cavity and its entrance with bone dust and blood, sealing with fascia lata and fibrin glue, and filling the tibial defect with fat. Five patients with large tumors developed transient palsy, and all recovered completely. The facial nerve retention rate is 82%. 58 patients required a temporary palatal suture. One year after surgery, normal facial nerve function was found in 14 cases (House I grade), and 39 cases had better function.