CLINICAL STUDY
Large Vestibular Schwannoma Resection Through the Suboccipital Retrosigmoid Keyhole Approach Cui Daming, MD,* Shen Yiwen, MD,Þ Zhou Bin, MD,* Xue Yajun, MD,* Yin Jia, MD,* Shen Rui, MD,* Shen Zhaoli, MD,* and Lou Meiqing, MD*
Objective: The objective of this study was to retrospectively study the outcomes of large vestibular schwannoma resection through the suboccipital retrosigmoid keyhole approach and emphasize technical details and advantages of surgical resection of large vestibular schwannomas via this approach. Methods: From January 2010 to September 2012, 37 consecutive patients (16 men and 21 women) with vestibular schwannoma, 4 cm or greater, received surgical resection through the suboccipital retrosigmoid keyhole approach in our department. Clinical records, radiographic findings, operative summaries, and follow-up data were analyzed retrospectively. Results: The mean age of these patients was 45.1 T 11.6 years. Thirty-six patients underwent primary keyhole surgical removal, and 1 underwent surgery for residual tumor after gamma knife. Gross total tumor removal was accomplished in 35 patients (94.6%), near total resection in 1 (2.7%), and subtotal resection in 1 patient (2.7%). The facial nerve was anatomically intact in all 37 patients (100%). Facial nerve function was assessed in 6 to 12 months after operation. Good function (House-Brackmann facial nerve grade IYII) was present in 81.1% of the patients, whereas acceptable function (grade III) was present in 11.1%. Cerebrospinal fluid (CSF) leakage that required surgical intervention occurred in only 5.4% of the patients, and meningitis occurred in 8.1%. In addition, 3 patients (8.1%) had hydrocephalus requiring a temporary ventricular diversion. There were no deaths. Conclusions: The suboccipital retrosigmoid keyhole approach is a valid choice for removing large vestibular schwannomas. Through this approach, cerebellopontine angle can be effectively exposed. Skills to protect facial nerve and extensive experience in microsurgical techniques can significantly improve the total resection rate and
postoperative facial nerve function. The authors recommend this approach for patients with vestibular schwannomas larger than 3 to 4 cm. Key Words: Vestibular schwannoma, cerebellopontine angle, internal acoustic meatus, facial nerve, suboccipital retrosigmoid keyhole approach (J Craniofac Surg 2014;25: 463Y468)
R
emoval of vestibular schwannomas (VSs, or acoustic schwannomas) remains one of the most challenging operations in neurosurgery. Larger VSs, larger than 3 to 4 cm in the greatest extracanalicular diameter, carry their own unique risks for resection.1 Surgical removal of large VSs has been associated with higher complication rates, poor facial nerve function, and high tumor recurrence rate.2 To resolve these problems, neurosurgeons and neurootologists have to innovate different surgical approaches and methods.3 However, it has been reported that the neurological and surgical complications occur in up to 20% of patients with large VSs.4,5 An excellent facial nerve function has been reported in only 50% of patients with large VSs.6 We operated on 37 patients with large VSs via the suboccipital retrosigmoid keyhole approach, and the postoperative facial nerve function was followed up for 6 to 12 months. We reviewed the surgical outcomes in these cases, including the extent of tumor removal, postoperative facial nerve function, and postoperative complications. Technical details and advantages of surgical resection of large VSs via the suboccipital retrosigmoid keyhole approach will be stressed on this article.
CLINICAL MATERIALS AND METHODS Patients and Presentations
From the *Department of Neurosurgery, Shanghai Tenth People’s Hospital, Tongji University; and †Department of Neurosurgery, Huashan Hospital, Fudan University, Shanghai, China. Received October 14, 2013. Accepted for publication November 11, 2013. Address correspondence and reprint requests to Lou Meiqing, MD, Department of Neurosurgery, Shanghai Tenth People’s Hospital, Tongji University, Shanghai 200072, China. E-mail: [email protected] C.D. and S.Y. contributed equally to this study. This work was supported by National Natural Science Foundation of China (no. 81201979) and Research Fund of Shanghai Municipal Health Bureau (no. 2010133). The authors report no conflicts of interest. Copyright * 2014 by Mutaz B. Habal, MD ISSN: 1049-2275 DOI: 10.1097/SCS.0000000000000528
The Journal of Craniofacial Surgery
From January 2010 to September 2012, 37 patients with large VSs (Q4 cm in largest diameter) underwent primary or secondary removal of their tumors via the suboccipital retrosigmoid keyhole approach. There were 16 males and 21 females with a mean age at the time of surgery of 45.1 T 11.6 years. Of the 37 patients, 1 patient presented with recurrent large VS after gamma knife surgery, and the remaining patients were newly diagnosed with large VSs. The typical clinical presentations of patients with large VSs included dizziness, headache, tinnitus, and progressive hearing loss. The main symptoms observed in the 37 patients were summarized as follows: hearing loss in 34 cases, dizziness in 28 cases, headache and vomiting in 12 cases, facial numbness in 11 cases, instability of gait in 10 cases, slight distortion of commissure in 8 cases, and bucking in 6 cases. Frequent preoperative cranial deficits included nerve deafness in 34 cases including 26 cases of complete hearing loss, slight facial palsy in 8 cases, ipsilateral facial hypalgesia in 17 cases, and decreased gag reflex in 9 cases.
& Volume 25, Number 2, March 2014
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
463
Daming et al
The Journal of Craniofacial Surgery
Imaging Studies Preoperative Imaging preparations should include magnetic resonance imaging (MRI) and computed tomography. Tumor anatomy (size and cystic, solid, or combined) and its relation to neurovascular structures was evaluated using MRI. Computed tomography scans were used for recognition of the extent of pneumatization of the temporal bone and the position of the venous anatomy. The tumor was located on the left cerebellopontine angle (CPA) in 15 patients and on the right CPA in 22 patients. The sizes of the tumors ranged from 4 to 6.1 cm, with a mean of 4.7 T 1.2 cm; 37.8% of the tumors were 5 cm or larger. Fourteen tumors were solid, and 23 tumors were partial cystic. Low signal ring between VS and
& Volume 25, Number 2, March 2014
brainstem was observed in 9 patients with intravenous gadolinium administration. Magnetic resonance imaging fluidYattenuated inversion recovery images revealed 19 patients with the pons and medulla oblongata edema (Fig. 1).
Surgical Technique Suboccipital Retrosigmoid Keyhole Craniotomy The patient was maintained in the lateral-prone position with the upper body slightly elevated at 15 degrees, the vertex title down, and the chin flexed. The tip of the digastric groove was used to localize the surgical incision. During the operation, the position was
FIGURE 1. Imaging features of a typical patient with large VS. Magnetic resonance imaging shows a large cystic and solid mass with heterogeneous enhancement in the right CPA (AYC). Computed tomography scan shows enlarged right IAM (D). Magnetic resonance imaging confirms large VS was totally removed after 3 months of follow-up (E). Computed tomography scan indicates the right CPA bone defect after suboccipitoretrosigmoid keyhole approach (F). Operative incision and facial nerve function test (GYI).
464
* 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery
& Volume 25, Number 2, March 2014
adjusted by the rotation of the operating bed to facilitate the maximal exposure and resection of the tumor. A 3- to 4-cm transverse (‘‘V’’sharp) incision in length was utilized behind the ear. A suboccipital craniectomy was performed to expose the inferior margin of transverse sinus and the posterior margin of sigmoid sinus. Exposure area was about 1.5 to 2.0 cm (in diameter). After the dura was opened, the CSF was drained by opening the lateral cerebellomedullary cistern and the arachnoid membranes along the inferior pole of the tumor to expose the CPA with minimal cerebellum retraction. The arachnoid membranes over the posterior aspect of the tumor capsule are excised (Figs. 2AYG).
Intraoperative Monitoring Intraoperative facial nerve monitoring (Medtronic’s NIMResponse motor nerve monitoring system; Medtronic XOMED, Jacksonville, FL) is used in 19 patients. Motor-evoked potential recordings of the orbicularis oris and oculi muscles were performed to monitor facial nerve function. A bipolar stimulus with an intensity of 1 mA and a duration of 0.1 ms was used to assess facial nerve response. During dissection, a facial nerve stimulator was used to obtain periodic real-time motor-evoked potential recordings and further identify facial nerve.
Management of VSs in the Internal Acoustic Meatus The inferior wall of the internal acoustic meatus (IAM) was drilled to the fundus after the dura is cut open around the posterior
Large Vestibular Schwannoma Resection
lip of IAM. High-speed drill was used to remove the bone under continuous irrigation to avoid heat injury to the nerves in the canal. Meanwhile, a piece of soaked gel-foam sponge was also utilized to protect underlying structures from the bone dust. Extension of bone removal allowed identification of the distal cranial nerve VII/VIII complex free of tumor and palpating the fundus of the IAM with a microYnerve hook (about 7Y9 mm deep). For patients with high jugular bulb, drilling the inferior wall of IAM must be more cautious to avoid venous sinus. Then, the dura within the IAM was cut, and the tumor was gently pulled out using the microYnerve hook along the long axis of the IAM from inside to outside. The vestibular and facial nerves were carefully identified before superior/inferior vestibular nerve adhered to the tumor was cut, to allow the tumor mobilized out of the canal. In this stage, blunt separation needs to be performed in combination with sharp dissection to avoid mechanical injury to intracanalicular segment of the facial nerve (Figs. 2HYI).
Microdissection of VSs Outside the IAM After the tumor located in the IAM was removed, the surgeon could get a general idea about the position of the facial nerve and once again confirmed no nerve fibers on the surface of the tumor. Coagulation of the posterior capsule was followed by aggressive internal tumor debulking. At this step, an intratumoral ‘‘peeling away an onion’’ strategy can be applied and make full use of Cavitron ultrasonic aspirator (Cavitron Corp., New York, NY) to reduce the volume of tumor from the inside. The peripheral tumor tissue gradually collapsed into the center of the visual field under the
FIGURE 2. Operative process of a typical patient with large VS. The patient body position in the process of operation (A). Body surface location of incision, asterion, transverse sinus, and retrosigmoid sinus (B). Incising the skin and suboccipital muscle (C). Using a detacher to mark the position of the digastric groove (D). Bone defect made via suboccipitoretrosigmoid keyhole approach (white disk as a yuan coin size) (E, F). The CSF of cerebellomedullary cistern was released under a microscope (G). Drilling the posterior wall of IAM (H). The most lateral tumor portion was carefully mobilized out of the IAM with a microdissector (I). The intracapsular tumor was debulked whenever possible with the Cavitron ultrasonic surgical aspirator (J). Abducens nerve (K). The facial nerve located immediately behind the abducens nerve at the lateral aspect of the pontomedullary sulcus (L).
* 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
465
The Journal of Craniofacial Surgery
Daming et al
influence of pressure. The internal decompression in combination with the extracapsular dissection was applied to resect the tumor in the following sequence: inferior pole, superior pole, brainstem, and ventral portion. While dissecting the inferior border of the tumor, the lower cranial nerves should be identified and preserved, and during mobilization of the upper pole of the tumor, attention should be devoted to protect the trigeminal nerve and petrous vein. When separating the interface between the tumor and the brainstem, the tumor capsule should be rolled laterally away from the middle cerebellar peduncle and brainstem toward the porus acusticus. The capsule is pulled on gently as dissection is performed using microforceps to detach the arachnoid membranes from the tumor. The soaked gelatin sponge was put into the space between the tumor and the brainstem and left behind to avoid the tumor back. The veins along the surface of the brainstem are often engorged and prone to avulsion. Gentle pressure using the gelatin sponge was a reasonable choice; bipolar coagulation should be strictly avoided. For the ventral, the facial nerve and cochlear nerve often become very attenuated and difficult to identify because of the tumor volume. At this time, the facial nerve monitoring might facilitate the nerve function preservation, but without facial nerve monitoring, the arachnoid membrane and tumor capsule must be reidentified along the superior or inferior pole of the tumor; microdissection was strictly performed along the capsule. Continuing to roll the capsule laterally, the facial nerve was dissected and peeled away from the tumor. Finally, the tumor was totally removed (Figs. 2JYL).
Postoperative Management and Follow-up After completing tumor resection, increasing the patient’s airway pressure to 30 mm Hg for 15 seconds checked out the effectiveness of the surgical hemostasis. Suboccipital muscle mixed glue and gel-foam sponge graft was placed within the defect of IAM to prevent formation of an intradural CSF leakage. After the retractor was removed and cerebellum re-expanded, the dura was reapproximated in a watertight fashion. Exposed mastoid air cells were then waxed, and a titanium mesh was placed over the cranial defect. In the first day after operation, a routine computed tomography was performed, and MRI was performed before discharge (Figs. 1E, F). The preservation of facial nerve function and hearing was assessed 6 to 12 months after surgery (Figs. 1GYI). The patients were divided into 3 subgroups: good facial nerve function (House-Brackmann [HB] grades I-II), regular facial nerve function (HB grades III), and poor facial nerve function (HB grades IVYV).
RESULTS The Extent of Tumor Resection Diagnosis of VS in 37 patients was confirmed by pathology. Gross total resection was performed in 35 (94.6%) of the 37 patients, near total resection in 1 (2.7%), and subtotal resection in 1 patient (2.7%). There was no death in the group. Most tumors were hypovascular and soft. However, in 1 patient, the tumor was hypovascular and relatively hard. Near the IAM, the tumor adhered to facial nerve closely. To prevent injury to facial nerve, subtotal tumor resection was attempted. Another patient, who had received gamma knife surgery 7 years before, was admitted to our department with recurrent VS. Regarding the intraoperative findings, the mass was tough and calcified, and there were severe adhesions with the surrounding vital structures, especially with the point of facial nerve origination on the brainstem. Upon dissection, there was a sharp rise in blood pressure, even reaching 198/124 mm Hg. Fluctuation of blood pressure was too high, and thus, to ensure the safety of operation, the tumor was only subtotal resected.
466
& Volume 25, Number 2, March 2014
The VS and Facial Nerve The facial nerve was anatomically intact at the end of surgery in 37 patients (100%). The facial nerve was compressed by the tumor without close adherence between the facial nerve and the tumor in 31 cases (83.8%). The facial nerve adhered closely to the tumor in 5 cases (13.5%). The facial nerve was infiltrated by the tumor in 1 patient (2.7%). The most common location of the facial nerve was on the anterior superior aspect of the tumor capsule. The facial nerve was on the anterior superior aspect of the tumor in 15 cases (40.5%), posterior inferior aspect in 13 cases (35.1%), and anterior middle third in 9 patients (24.4%). In our study, no case of the facial nerve running along the surface of the tumor was noted. But the facial nerve was compressed or infiltrated by the tumor; it lost the original shape and became very thin and spread out.
Complications and Postoperative Follow-up Postoperative care included on average 1 day’s stay in an intensive care unit. There were no postoperative deaths. The most common complication was transient headache and dizziness. Cerebrospinal fluid leakage occurred in 2 patients (5.4%) and was cured by lumbar CSF drainage and compression on the region of the operative wound. Three patients (8.1%) developed bacterial meningitis treated successfully with antibiotics and lumbar puncture. In this series, 3 patients (8.1%) showed some decreased function of the low cranial nerves manifesting as transient dysphagia and dysphonia and need gastric intubation; 5 patients (13.5%) had ipsilateral facial numbness; however, 3 months later, the sign previously mentioned recovered. In addition, 3 patients (8.1%) had hydrocephalus requiring a temporary ventricular diversion. Facial nerve function was assessed in 6 to 12 months after operation. Good function (HB grades IYII) was present in 81.1% of the patients, with acceptable function (HB grade III) in 11.1%, and with poor function (HB grade IV) in 5.4%. Three patients had varying degrees of hearing preservation; their hearing roughly returned to preoperative level (at 50Y60 dB) 6 months later after operation. Two of them underwent gross tumor resection, and 1 patient underwent partial tumor removal.
DISCUSSION Surgical resection is the most effective treatment for large VSs. Although advances in neuroimaging, intraoperative nerve monitoring, and microsurgical technique have transformed a once dangerous and complex operation into a ‘‘relatively simple’’ operation associated with low morbidity and mortality, preservation of normal nerve function still is the primary focus of VS surgery.2Y5,7 Many approaches are used for VS surgery. Hearing level and tumor size are the main factors in choosing the approach. Large VSs are commonly approached by the suboccipital retrosigmoid approach, but some studies have reported good results using the translabyrinthine or combined approach in the tumors of more than 4 cm.7Y9 We also used the suboccipital retrosigmoid approach in all cases. To reduce surgical injury, we conducted a series of improvement in view of the conventional suboccipital retrosigmoid approach. First, a transverse incision was adopted behind the ear, this incision could effectively prevent the scalp flap from interfering with the working zone of the surgeon and decrease the operator’s working distance to the tumor. Second, we emphasized that sigmoid sinus exposure must be adequate, and inversely requirements of transverse sinus exposure are not high. Actually, exposure of the sigmoid sinus was responsible for effective drainage of CSF in the CPA. Third, by means of the amplifying effect of keyhole, the deep tissue exposure is far greater than the bone defect area. By rotating the head, various views from different angles can be obtained through the suboccipital retrosigmoid keyhole approach during the operation. By the * 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery
& Volume 25, Number 2, March 2014
previously mentioned methods, a diameter of 1.5- to 2.0-cm bone defect may satisfy the needs of operation. This approach was sufficient for removing a 60-mm-diameter tumor that extended from IAM to CPA in our series. In most cases of large VSs, the brain tension is significant and the effective CSF drainage is relatively difficult. To solve this problem, we suggested that the outer edge of the bony window must reach the edge of the sigmoid sinus to prevent the residual bone from covering the cerebellar hemisphere. If significant obstructive hydrocephalus was present preoperatively, CSF drainage through a previously placed external ventricular drain was achieved before dural opening to decrease internal cerebral pressure. Dural opening was completed using a C-shaped incision based toward the sigmoid sinus. Latex finger cots might be used to protect the cerebellar hemispheres instead of a gelatin sponge to avoid swelling of the sponge and narrowing of the surgical space. In this procedure, the rational use of retractor system was also very important. In addition, we emphasized that the CSF was released by opening the lateral cerebellomedullary cistern and especially opening the arachnoid membranes caudal to the tumor to prevent aggressive cerebellar retractor, cerebellar contusions, and even the excision of the lateral portion of cerebellum. Vestibular schwannomas originate from schwann cells of the vestibular nerve within the IAM. The proper management of the tumor within IAM is a prerequisite for gross total resection of large VSs.2 Removal of the posterior wall of the IAM is an essential step in VS surgery to obtain effective exposure and minimize morbidity associated with loss of hearing and facial paralysis. The labyrinthine structures and pneumatization of the posterior wall of the IAM and bulb of the jugular vein must be evaluated, and opening of the IAM is performed based on the preoperative radiological findings and the experience of surgeon. Yokoyama et al10 attempted to estimate the preoperative risk associated with opening the inner ear structures via the retrosigmoid approach by introducing the S-F line. Stieglitz et al11 tried to predict the incidence of CSF leakage by analyzing the average length of the IAM and diameter of the IAM porus. As our experiences, drilling the desirable depth of the IAM was approximately 7 to 9 mm, and the microYnerve hook could palpate the fundus of the IAM as appropriate. Drilling deeper than this would result in damage to the semicircular canals, whereas superficial drilling would result in increasing the likelihood of tumor remnant. In addition, the superior and inferior walls of the IAM should be removal so that the IAM was open wide to improve exposure of the tumor and facilitate complete tumor resection. In our study, all 37 patients underwent drilling the posterior wall of the IAM, including 5 patients with the high location of the bulb of the jugular vein. No patients had the bulb of jugular vein injuryYrelated hemorrhage. The damage of mastoid air cells is the main cause of postoperative CSF rhinorrhea. It has been reported that in patients who are highly susceptible to mastoid air cell injury, a mixture of free fat graft and fibrin glue should be used to close breakage and reduce the incidence of CSF leakage.11,12 Our experience revealed that the sandwich structure of muscle mixed with the gelatin sponge and the fibrin gel had a better waterproofing effect. Management of the tumor within the IAM, under the premise of the facial nerve preservation, is a crucial step of gross total removal of the tumor. There is little variation of the anatomic position of the facial nerve in the IAM, and removal of the tumor within the IAM will convey additional information regarding the route of the facial nerve; subsequently, devascularization of the tumor by coagulating the feeders from the dura over the porus acusticus can greatly minimize the potential of intraoperative bleeding.3 In this procedure, the manipulation must be gentle. The operator had to use a careful combination of blunt and sharp microdissection techniques to avoid the facial nerve injury. With the help of facial nerve monitoring, the
Large Vestibular Schwannoma Resection
likelihood of its anatomic preservation respectively increased,13 but if no facial nerve monitoring could be utilized, more attention should be paid while the tumor capsules are microdissected. The surgeon might use the blunt-tipped stimulator as an instrument to separate the nerve from the capsule. In this step, we emphasized that the operation should be performed without tension. In the 1970s, Yasargil et al14 proposed VSs as an extraarachnoid lesion and considered that the VS originates epiarachnoidally in the IAM and pushes the arachnoid membrane of the cerebellopontine cistern medially causing a folding of the arachnoid between the tumor and the brain. And then, Tarlov15 published another view of the arachnoidal relation to the VS in the CPA. He believed that the tumor originates epiarachnoidally in the IAM, and it pushes the folded layer of arachnoid with enlargement, resulting in the formation of a second layer of arachnoid between the tumor and the brainstem. With this understanding, he recommended that the double layers of the arachnoid should be separated from the tumor capsule to allow dissection of the tumor and capsule within this arachnoid envelope. In 2002, Ohata et al,16 through the operative findings of their sequential cases, put forward the concept of subarachnoidal origin of VSs. He suggested that the VS originates in the subarachnoid within the IAM and grows epiarachnoidally in the CPA. Rearrangement of the arachnoid begins with its adhesion to the medial pole of the tumor along the porus, resulting in arachnoidal invagination into the cerebellopontine cistern with further growth of the tumor. With this understanding, we considered that VS resection should be performed along the capsule of the tumor, maintaining the integrity of the arachnoid, to avoid nerve and blood vessel damage. After maximal internal debulking, the tumor outside the central field of vision was gradually pushed back into the surgical area via the pulsatile effect of CSF and elastic recoil of brain tissue back. Making the best of the Cavitron ultrasonic aspirator would tremendously shorten the operation time and facilitate the operation process.17 After significantly reducing the volume, the tumor can be easily dissected in a sequence. The primary purpose of dissection of the inferior pole of the tumor was to preserve the lower cranial nerves and the posterior inferior cerebellar artery and its branches. In this region, lower cranial nerves usually did not significantly attach to the tumor capsule and were relatively simple to mobilize away from the tumor capsule. After finishing the step, we subsequently kept on internal debulking along the superior pole of the tumor and peeled away the capsule from the tentorium and the fifth cranial nerve. The petrosal vein was protected if possible. The facial nerve was often adherent along the superior pole of the tumor; more attention and patience must be paid. When separating the interface between the tumor and the brainstem, the tumor capsule was gently rolled laterally away from the middle cerebellar peduncle toward the porus acusticus. In the presence of preoperative brainstem edema, violation of the pial membranes was quite common. Once this event occurred, the surgeon should make full use of the microdissector to peel away the tumor from the brainstem. The soaked gelatin sponge was put into the space between the tumor and the brainstem and left behind to avoid the tumor back. For the ventral, identification of the facial nerve was mainly based on familiarity with anatomic landmarks and intraoperative facial nerve monitoring. The eighth cranial nerve locates immediately behind the facial nerve at the lateral aspect of the pontomedullary sulcus and joins the brainstem in front of the site of attachment to the flocculus. Meanwhile, the facial nerve is usually white, and the acoustic nerve is gray. The use of these anatomic landmarks will assist in the identification of the root exit zone of the facial nerve during tumor resection. Through combining this information with the route of the facial nerve over the capsule conveyed by removing the tumor within the IAM, the surgeon could generally learn the anatomic location of the facial nerve and continue
* 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
467
The Journal of Craniofacial Surgery
Daming et al
the end-to-end dissection of the nerve from the tumor. The IAM is the most common site of serious adhesion between the facial nerve and the tumor. Ohata et al16 reported that because of tumor growth, the arachnoid matter covering the tumor occasionally became incomplete or even disappeared, resulting in local adhesion between the tumor and the facial nerve. We encountered 1 patient with the similar conditions. To avoid damage to the facial nerve, a small piece of tumor was not removed. Common problems after the removal of a large VS are CSF leakage, meningitis, and other neurological complications.18 In our cases, most of patients were cured with conservative methods. Facial nerve function was assessed in 6 to 12 months after operation. House-Brackmann grades IYII were present in 81.1% of the patients, with HB grade III in 11.1%, and with HB grade IV in 5.4%. Three patients had varying degrees of hearing preservation; their hearing roughly returned to preoperative level (at 50Y60 dB) 6 months later after operation. Two of them underwent gross tumor resection, and 1 patient underwent partial tumor removal. Despite the small number of cases, which is insufficient to give a hypothesis, these findings suggested that dissection along the capsule of the tumor and maintaining the integrity of the arachnoid were prerequisites for preservation of normal nerve function. In summary, the suboccipital retrosigmoid keyhole approach, by means of the amplifying effect of keyhole, uses a smaller bone defect to optimize surgical vision and avoid unnecessary exposure or destruction of the skull structure. Our data illustrate that this approach combined with appropriate surgical strategies and operating techniques can facilitate effective and complete large VS resection.
REFERENCES 1. Huang X, Zhang J, Yang H, et al. Use of intraoperative ultrasonography to monitor surgery for large acoustic neuromas: a pilot study. J Med Ultrasonics 2010;37:15Y19 2. Samii M, Gerganov VM, Samii A. Functional outcome after complete surgical removal of giant vestibular schwannomas. J Neurosurg 2010;112:860Y867 3. Kulwin CG, Cohen-Gadol AA. Technical nuances of resection of giant (95 cm) vestibular schwannomas: pearls for success. Neurosurg Focus 2012;33:1Y8 4. Jung S, Kang SS, Kim TS, et al. Current surgical results of retrosigmoid approach in extra large vestibular schwannomas. Surg Neurol 2000;53:370Y378
468
& Volume 25, Number 2, March 2014
5. Kim WH, Park CK, Kim DG, et al. Management of recurrent vestibular schwannomas. J Korean Neurosurg Soc 2006;39:87Y91 6. Comey CH, Jannetta PJ, Sheptak PE, et al. Staged removal of acoustic tumors: techniques and lessons learned from a series of 83 patients. Neurosurgery 1995;37:915Y921 7. Yamakami I, Uchino Y, Kobayashi E, et al. Removal of large acoustic neurinomas (vestibular schwannomas) by the retrosigmoid approach with no mortality and minimal morbidity. J Neurol Neurosurg Psychiatry 2004;75:453Y458 8. Jamro´z B, Niemczyk K, Morawski K, et al. Extended middle fossa approach in treatment of vestibular schwannoma technique of surgery and postoperative complications. Otolaryngol Pol 2010;64:3Y9 9. Raslan AM, Liu JK, McMenomey SO, et al. Staged resection of large vestibular schwannomas. J Neurosurg 2012;116:1126Y1133 10. Yokoyama T, Uemura K, Ryu H, et al. Surgical approach to the internal auditory meatus in acoustic neuroma surgery: significance of preoperative high-resolution computed tomography. Neurosurgery 1996;39:965Y970 11. Stieglitz LH, Giordano M, Gerganov VM, et al. How obliteration of petrosal air cells by vestibular schwannoma influences the risk of postoperative CSF fistula. Clin Neurol Neurosurg 2011;113:746Y751 12. Lu¨demann WO, Stieglitz LH, Gerganov V, et al. Fat implant is superior to muscle implant in vestibular schwannoma surgery for the prevention of cerebrospinal fluid fistulae. Neurosurgery 2008;63(1 suppl 1): ONS38YONS42; discussion 42Y43 13. Yingling CD, Gardi JN. Intraoperative monitoring of facial and cochlear nerves during acoustic neuroma surgery. Otolaryngol Clin North Am 1992;25:413Y448 14. Yasargil MG, Smith RD, Gasser JC. Microsurgical approach to acoustic neurinoma. Adv Tech Neurosurg 1977;4:93Y129 15. Tarlov E. Total one-stage suboccipital microsurgical removal of acoustic neuromas of all size: with emphasis on arachnoid planes and on saving the facial nerve. Surg Clin N Am 1980;60:565Y591 16. Ohata K, Tsuyuguchi N, Morino M, et al. A hypothesis of epiarachnodal growth of vestibular schwannoma at the cerebellopontine angle: surgical importance. J Postgrad Med 2002;48:253Y259 17. Samii M, Gerganov V, Samii A. Improved preservation of hearing and facial nerve function in vestibular schwannoma surgery via the retrosigmoid approach in a series of 200 patients. J Neurosurg 2006;105:527Y535 18. Charpiot A, Tringali S, Zaouche S, et al. Perioperative complications after translabyrinthine removal of large or giant vestibular schwannoma: outcomes for 123 patients. Acta Otolaryngol 2010;130:1249Y1255
* 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
Large Vestibular Schwannoma Resection Through the Suboccipital Retrosigmoid Keyhole Approach Cui Daming, MD,* Shen Yiwen, MD,Þ Zhou Bin, MD,* Xue Yajun, MD,* Yin Jia, MD,* Shen Rui, MD,* Shen Zhaoli, MD,* and Lou Meiqing, MD*
Objective: The objective of this study was to retrospectively study the outcomes of large vestibular schwannoma resection through the suboccipital retrosigmoid keyhole approach and emphasize technical details and advantages of surgical resection of large vestibular schwannomas via this approach. Methods: From January 2010 to September 2012, 37 consecutive patients (16 men and 21 women) with vestibular schwannoma, 4 cm or greater, received surgical resection through the suboccipital retrosigmoid keyhole approach in our department. Clinical records, radiographic findings, operative summaries, and follow-up data were analyzed retrospectively. Results: The mean age of these patients was 45.1 T 11.6 years. Thirty-six patients underwent primary keyhole surgical removal, and 1 underwent surgery for residual tumor after gamma knife. Gross total tumor removal was accomplished in 35 patients (94.6%), near total resection in 1 (2.7%), and subtotal resection in 1 patient (2.7%). The facial nerve was anatomically intact in all 37 patients (100%). Facial nerve function was assessed in 6 to 12 months after operation. Good function (House-Brackmann facial nerve grade IYII) was present in 81.1% of the patients, whereas acceptable function (grade III) was present in 11.1%. Cerebrospinal fluid (CSF) leakage that required surgical intervention occurred in only 5.4% of the patients, and meningitis occurred in 8.1%. In addition, 3 patients (8.1%) had hydrocephalus requiring a temporary ventricular diversion. There were no deaths. Conclusions: The suboccipital retrosigmoid keyhole approach is a valid choice for removing large vestibular schwannomas. Through this approach, cerebellopontine angle can be effectively exposed. Skills to protect facial nerve and extensive experience in microsurgical techniques can significantly improve the total resection rate and
postoperative facial nerve function. The authors recommend this approach for patients with vestibular schwannomas larger than 3 to 4 cm. Key Words: Vestibular schwannoma, cerebellopontine angle, internal acoustic meatus, facial nerve, suboccipital retrosigmoid keyhole approach (J Craniofac Surg 2014;25: 463Y468)
R
emoval of vestibular schwannomas (VSs, or acoustic schwannomas) remains one of the most challenging operations in neurosurgery. Larger VSs, larger than 3 to 4 cm in the greatest extracanalicular diameter, carry their own unique risks for resection.1 Surgical removal of large VSs has been associated with higher complication rates, poor facial nerve function, and high tumor recurrence rate.2 To resolve these problems, neurosurgeons and neurootologists have to innovate different surgical approaches and methods.3 However, it has been reported that the neurological and surgical complications occur in up to 20% of patients with large VSs.4,5 An excellent facial nerve function has been reported in only 50% of patients with large VSs.6 We operated on 37 patients with large VSs via the suboccipital retrosigmoid keyhole approach, and the postoperative facial nerve function was followed up for 6 to 12 months. We reviewed the surgical outcomes in these cases, including the extent of tumor removal, postoperative facial nerve function, and postoperative complications. Technical details and advantages of surgical resection of large VSs via the suboccipital retrosigmoid keyhole approach will be stressed on this article.
CLINICAL MATERIALS AND METHODS Patients and Presentations
From the *Department of Neurosurgery, Shanghai Tenth People’s Hospital, Tongji University; and †Department of Neurosurgery, Huashan Hospital, Fudan University, Shanghai, China. Received October 14, 2013. Accepted for publication November 11, 2013. Address correspondence and reprint requests to Lou Meiqing, MD, Department of Neurosurgery, Shanghai Tenth People’s Hospital, Tongji University, Shanghai 200072, China. E-mail: [email protected] C.D. and S.Y. contributed equally to this study. This work was supported by National Natural Science Foundation of China (no. 81201979) and Research Fund of Shanghai Municipal Health Bureau (no. 2010133). The authors report no conflicts of interest. Copyright * 2014 by Mutaz B. Habal, MD ISSN: 1049-2275 DOI: 10.1097/SCS.0000000000000528
The Journal of Craniofacial Surgery
From January 2010 to September 2012, 37 patients with large VSs (Q4 cm in largest diameter) underwent primary or secondary removal of their tumors via the suboccipital retrosigmoid keyhole approach. There were 16 males and 21 females with a mean age at the time of surgery of 45.1 T 11.6 years. Of the 37 patients, 1 patient presented with recurrent large VS after gamma knife surgery, and the remaining patients were newly diagnosed with large VSs. The typical clinical presentations of patients with large VSs included dizziness, headache, tinnitus, and progressive hearing loss. The main symptoms observed in the 37 patients were summarized as follows: hearing loss in 34 cases, dizziness in 28 cases, headache and vomiting in 12 cases, facial numbness in 11 cases, instability of gait in 10 cases, slight distortion of commissure in 8 cases, and bucking in 6 cases. Frequent preoperative cranial deficits included nerve deafness in 34 cases including 26 cases of complete hearing loss, slight facial palsy in 8 cases, ipsilateral facial hypalgesia in 17 cases, and decreased gag reflex in 9 cases.
& Volume 25, Number 2, March 2014
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
463
Daming et al
The Journal of Craniofacial Surgery
Imaging Studies Preoperative Imaging preparations should include magnetic resonance imaging (MRI) and computed tomography. Tumor anatomy (size and cystic, solid, or combined) and its relation to neurovascular structures was evaluated using MRI. Computed tomography scans were used for recognition of the extent of pneumatization of the temporal bone and the position of the venous anatomy. The tumor was located on the left cerebellopontine angle (CPA) in 15 patients and on the right CPA in 22 patients. The sizes of the tumors ranged from 4 to 6.1 cm, with a mean of 4.7 T 1.2 cm; 37.8% of the tumors were 5 cm or larger. Fourteen tumors were solid, and 23 tumors were partial cystic. Low signal ring between VS and
& Volume 25, Number 2, March 2014
brainstem was observed in 9 patients with intravenous gadolinium administration. Magnetic resonance imaging fluidYattenuated inversion recovery images revealed 19 patients with the pons and medulla oblongata edema (Fig. 1).
Surgical Technique Suboccipital Retrosigmoid Keyhole Craniotomy The patient was maintained in the lateral-prone position with the upper body slightly elevated at 15 degrees, the vertex title down, and the chin flexed. The tip of the digastric groove was used to localize the surgical incision. During the operation, the position was
FIGURE 1. Imaging features of a typical patient with large VS. Magnetic resonance imaging shows a large cystic and solid mass with heterogeneous enhancement in the right CPA (AYC). Computed tomography scan shows enlarged right IAM (D). Magnetic resonance imaging confirms large VS was totally removed after 3 months of follow-up (E). Computed tomography scan indicates the right CPA bone defect after suboccipitoretrosigmoid keyhole approach (F). Operative incision and facial nerve function test (GYI).
464
* 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery
& Volume 25, Number 2, March 2014
adjusted by the rotation of the operating bed to facilitate the maximal exposure and resection of the tumor. A 3- to 4-cm transverse (‘‘V’’sharp) incision in length was utilized behind the ear. A suboccipital craniectomy was performed to expose the inferior margin of transverse sinus and the posterior margin of sigmoid sinus. Exposure area was about 1.5 to 2.0 cm (in diameter). After the dura was opened, the CSF was drained by opening the lateral cerebellomedullary cistern and the arachnoid membranes along the inferior pole of the tumor to expose the CPA with minimal cerebellum retraction. The arachnoid membranes over the posterior aspect of the tumor capsule are excised (Figs. 2AYG).
Intraoperative Monitoring Intraoperative facial nerve monitoring (Medtronic’s NIMResponse motor nerve monitoring system; Medtronic XOMED, Jacksonville, FL) is used in 19 patients. Motor-evoked potential recordings of the orbicularis oris and oculi muscles were performed to monitor facial nerve function. A bipolar stimulus with an intensity of 1 mA and a duration of 0.1 ms was used to assess facial nerve response. During dissection, a facial nerve stimulator was used to obtain periodic real-time motor-evoked potential recordings and further identify facial nerve.
Management of VSs in the Internal Acoustic Meatus The inferior wall of the internal acoustic meatus (IAM) was drilled to the fundus after the dura is cut open around the posterior
Large Vestibular Schwannoma Resection
lip of IAM. High-speed drill was used to remove the bone under continuous irrigation to avoid heat injury to the nerves in the canal. Meanwhile, a piece of soaked gel-foam sponge was also utilized to protect underlying structures from the bone dust. Extension of bone removal allowed identification of the distal cranial nerve VII/VIII complex free of tumor and palpating the fundus of the IAM with a microYnerve hook (about 7Y9 mm deep). For patients with high jugular bulb, drilling the inferior wall of IAM must be more cautious to avoid venous sinus. Then, the dura within the IAM was cut, and the tumor was gently pulled out using the microYnerve hook along the long axis of the IAM from inside to outside. The vestibular and facial nerves were carefully identified before superior/inferior vestibular nerve adhered to the tumor was cut, to allow the tumor mobilized out of the canal. In this stage, blunt separation needs to be performed in combination with sharp dissection to avoid mechanical injury to intracanalicular segment of the facial nerve (Figs. 2HYI).
Microdissection of VSs Outside the IAM After the tumor located in the IAM was removed, the surgeon could get a general idea about the position of the facial nerve and once again confirmed no nerve fibers on the surface of the tumor. Coagulation of the posterior capsule was followed by aggressive internal tumor debulking. At this step, an intratumoral ‘‘peeling away an onion’’ strategy can be applied and make full use of Cavitron ultrasonic aspirator (Cavitron Corp., New York, NY) to reduce the volume of tumor from the inside. The peripheral tumor tissue gradually collapsed into the center of the visual field under the
FIGURE 2. Operative process of a typical patient with large VS. The patient body position in the process of operation (A). Body surface location of incision, asterion, transverse sinus, and retrosigmoid sinus (B). Incising the skin and suboccipital muscle (C). Using a detacher to mark the position of the digastric groove (D). Bone defect made via suboccipitoretrosigmoid keyhole approach (white disk as a yuan coin size) (E, F). The CSF of cerebellomedullary cistern was released under a microscope (G). Drilling the posterior wall of IAM (H). The most lateral tumor portion was carefully mobilized out of the IAM with a microdissector (I). The intracapsular tumor was debulked whenever possible with the Cavitron ultrasonic surgical aspirator (J). Abducens nerve (K). The facial nerve located immediately behind the abducens nerve at the lateral aspect of the pontomedullary sulcus (L).
* 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
465
The Journal of Craniofacial Surgery
Daming et al
influence of pressure. The internal decompression in combination with the extracapsular dissection was applied to resect the tumor in the following sequence: inferior pole, superior pole, brainstem, and ventral portion. While dissecting the inferior border of the tumor, the lower cranial nerves should be identified and preserved, and during mobilization of the upper pole of the tumor, attention should be devoted to protect the trigeminal nerve and petrous vein. When separating the interface between the tumor and the brainstem, the tumor capsule should be rolled laterally away from the middle cerebellar peduncle and brainstem toward the porus acusticus. The capsule is pulled on gently as dissection is performed using microforceps to detach the arachnoid membranes from the tumor. The soaked gelatin sponge was put into the space between the tumor and the brainstem and left behind to avoid the tumor back. The veins along the surface of the brainstem are often engorged and prone to avulsion. Gentle pressure using the gelatin sponge was a reasonable choice; bipolar coagulation should be strictly avoided. For the ventral, the facial nerve and cochlear nerve often become very attenuated and difficult to identify because of the tumor volume. At this time, the facial nerve monitoring might facilitate the nerve function preservation, but without facial nerve monitoring, the arachnoid membrane and tumor capsule must be reidentified along the superior or inferior pole of the tumor; microdissection was strictly performed along the capsule. Continuing to roll the capsule laterally, the facial nerve was dissected and peeled away from the tumor. Finally, the tumor was totally removed (Figs. 2JYL).
Postoperative Management and Follow-up After completing tumor resection, increasing the patient’s airway pressure to 30 mm Hg for 15 seconds checked out the effectiveness of the surgical hemostasis. Suboccipital muscle mixed glue and gel-foam sponge graft was placed within the defect of IAM to prevent formation of an intradural CSF leakage. After the retractor was removed and cerebellum re-expanded, the dura was reapproximated in a watertight fashion. Exposed mastoid air cells were then waxed, and a titanium mesh was placed over the cranial defect. In the first day after operation, a routine computed tomography was performed, and MRI was performed before discharge (Figs. 1E, F). The preservation of facial nerve function and hearing was assessed 6 to 12 months after surgery (Figs. 1GYI). The patients were divided into 3 subgroups: good facial nerve function (House-Brackmann [HB] grades I-II), regular facial nerve function (HB grades III), and poor facial nerve function (HB grades IVYV).
RESULTS The Extent of Tumor Resection Diagnosis of VS in 37 patients was confirmed by pathology. Gross total resection was performed in 35 (94.6%) of the 37 patients, near total resection in 1 (2.7%), and subtotal resection in 1 patient (2.7%). There was no death in the group. Most tumors were hypovascular and soft. However, in 1 patient, the tumor was hypovascular and relatively hard. Near the IAM, the tumor adhered to facial nerve closely. To prevent injury to facial nerve, subtotal tumor resection was attempted. Another patient, who had received gamma knife surgery 7 years before, was admitted to our department with recurrent VS. Regarding the intraoperative findings, the mass was tough and calcified, and there were severe adhesions with the surrounding vital structures, especially with the point of facial nerve origination on the brainstem. Upon dissection, there was a sharp rise in blood pressure, even reaching 198/124 mm Hg. Fluctuation of blood pressure was too high, and thus, to ensure the safety of operation, the tumor was only subtotal resected.
466
& Volume 25, Number 2, March 2014
The VS and Facial Nerve The facial nerve was anatomically intact at the end of surgery in 37 patients (100%). The facial nerve was compressed by the tumor without close adherence between the facial nerve and the tumor in 31 cases (83.8%). The facial nerve adhered closely to the tumor in 5 cases (13.5%). The facial nerve was infiltrated by the tumor in 1 patient (2.7%). The most common location of the facial nerve was on the anterior superior aspect of the tumor capsule. The facial nerve was on the anterior superior aspect of the tumor in 15 cases (40.5%), posterior inferior aspect in 13 cases (35.1%), and anterior middle third in 9 patients (24.4%). In our study, no case of the facial nerve running along the surface of the tumor was noted. But the facial nerve was compressed or infiltrated by the tumor; it lost the original shape and became very thin and spread out.
Complications and Postoperative Follow-up Postoperative care included on average 1 day’s stay in an intensive care unit. There were no postoperative deaths. The most common complication was transient headache and dizziness. Cerebrospinal fluid leakage occurred in 2 patients (5.4%) and was cured by lumbar CSF drainage and compression on the region of the operative wound. Three patients (8.1%) developed bacterial meningitis treated successfully with antibiotics and lumbar puncture. In this series, 3 patients (8.1%) showed some decreased function of the low cranial nerves manifesting as transient dysphagia and dysphonia and need gastric intubation; 5 patients (13.5%) had ipsilateral facial numbness; however, 3 months later, the sign previously mentioned recovered. In addition, 3 patients (8.1%) had hydrocephalus requiring a temporary ventricular diversion. Facial nerve function was assessed in 6 to 12 months after operation. Good function (HB grades IYII) was present in 81.1% of the patients, with acceptable function (HB grade III) in 11.1%, and with poor function (HB grade IV) in 5.4%. Three patients had varying degrees of hearing preservation; their hearing roughly returned to preoperative level (at 50Y60 dB) 6 months later after operation. Two of them underwent gross tumor resection, and 1 patient underwent partial tumor removal.
DISCUSSION Surgical resection is the most effective treatment for large VSs. Although advances in neuroimaging, intraoperative nerve monitoring, and microsurgical technique have transformed a once dangerous and complex operation into a ‘‘relatively simple’’ operation associated with low morbidity and mortality, preservation of normal nerve function still is the primary focus of VS surgery.2Y5,7 Many approaches are used for VS surgery. Hearing level and tumor size are the main factors in choosing the approach. Large VSs are commonly approached by the suboccipital retrosigmoid approach, but some studies have reported good results using the translabyrinthine or combined approach in the tumors of more than 4 cm.7Y9 We also used the suboccipital retrosigmoid approach in all cases. To reduce surgical injury, we conducted a series of improvement in view of the conventional suboccipital retrosigmoid approach. First, a transverse incision was adopted behind the ear, this incision could effectively prevent the scalp flap from interfering with the working zone of the surgeon and decrease the operator’s working distance to the tumor. Second, we emphasized that sigmoid sinus exposure must be adequate, and inversely requirements of transverse sinus exposure are not high. Actually, exposure of the sigmoid sinus was responsible for effective drainage of CSF in the CPA. Third, by means of the amplifying effect of keyhole, the deep tissue exposure is far greater than the bone defect area. By rotating the head, various views from different angles can be obtained through the suboccipital retrosigmoid keyhole approach during the operation. By the * 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery
& Volume 25, Number 2, March 2014
previously mentioned methods, a diameter of 1.5- to 2.0-cm bone defect may satisfy the needs of operation. This approach was sufficient for removing a 60-mm-diameter tumor that extended from IAM to CPA in our series. In most cases of large VSs, the brain tension is significant and the effective CSF drainage is relatively difficult. To solve this problem, we suggested that the outer edge of the bony window must reach the edge of the sigmoid sinus to prevent the residual bone from covering the cerebellar hemisphere. If significant obstructive hydrocephalus was present preoperatively, CSF drainage through a previously placed external ventricular drain was achieved before dural opening to decrease internal cerebral pressure. Dural opening was completed using a C-shaped incision based toward the sigmoid sinus. Latex finger cots might be used to protect the cerebellar hemispheres instead of a gelatin sponge to avoid swelling of the sponge and narrowing of the surgical space. In this procedure, the rational use of retractor system was also very important. In addition, we emphasized that the CSF was released by opening the lateral cerebellomedullary cistern and especially opening the arachnoid membranes caudal to the tumor to prevent aggressive cerebellar retractor, cerebellar contusions, and even the excision of the lateral portion of cerebellum. Vestibular schwannomas originate from schwann cells of the vestibular nerve within the IAM. The proper management of the tumor within IAM is a prerequisite for gross total resection of large VSs.2 Removal of the posterior wall of the IAM is an essential step in VS surgery to obtain effective exposure and minimize morbidity associated with loss of hearing and facial paralysis. The labyrinthine structures and pneumatization of the posterior wall of the IAM and bulb of the jugular vein must be evaluated, and opening of the IAM is performed based on the preoperative radiological findings and the experience of surgeon. Yokoyama et al10 attempted to estimate the preoperative risk associated with opening the inner ear structures via the retrosigmoid approach by introducing the S-F line. Stieglitz et al11 tried to predict the incidence of CSF leakage by analyzing the average length of the IAM and diameter of the IAM porus. As our experiences, drilling the desirable depth of the IAM was approximately 7 to 9 mm, and the microYnerve hook could palpate the fundus of the IAM as appropriate. Drilling deeper than this would result in damage to the semicircular canals, whereas superficial drilling would result in increasing the likelihood of tumor remnant. In addition, the superior and inferior walls of the IAM should be removal so that the IAM was open wide to improve exposure of the tumor and facilitate complete tumor resection. In our study, all 37 patients underwent drilling the posterior wall of the IAM, including 5 patients with the high location of the bulb of the jugular vein. No patients had the bulb of jugular vein injuryYrelated hemorrhage. The damage of mastoid air cells is the main cause of postoperative CSF rhinorrhea. It has been reported that in patients who are highly susceptible to mastoid air cell injury, a mixture of free fat graft and fibrin glue should be used to close breakage and reduce the incidence of CSF leakage.11,12 Our experience revealed that the sandwich structure of muscle mixed with the gelatin sponge and the fibrin gel had a better waterproofing effect. Management of the tumor within the IAM, under the premise of the facial nerve preservation, is a crucial step of gross total removal of the tumor. There is little variation of the anatomic position of the facial nerve in the IAM, and removal of the tumor within the IAM will convey additional information regarding the route of the facial nerve; subsequently, devascularization of the tumor by coagulating the feeders from the dura over the porus acusticus can greatly minimize the potential of intraoperative bleeding.3 In this procedure, the manipulation must be gentle. The operator had to use a careful combination of blunt and sharp microdissection techniques to avoid the facial nerve injury. With the help of facial nerve monitoring, the
Large Vestibular Schwannoma Resection
likelihood of its anatomic preservation respectively increased,13 but if no facial nerve monitoring could be utilized, more attention should be paid while the tumor capsules are microdissected. The surgeon might use the blunt-tipped stimulator as an instrument to separate the nerve from the capsule. In this step, we emphasized that the operation should be performed without tension. In the 1970s, Yasargil et al14 proposed VSs as an extraarachnoid lesion and considered that the VS originates epiarachnoidally in the IAM and pushes the arachnoid membrane of the cerebellopontine cistern medially causing a folding of the arachnoid between the tumor and the brain. And then, Tarlov15 published another view of the arachnoidal relation to the VS in the CPA. He believed that the tumor originates epiarachnoidally in the IAM, and it pushes the folded layer of arachnoid with enlargement, resulting in the formation of a second layer of arachnoid between the tumor and the brainstem. With this understanding, he recommended that the double layers of the arachnoid should be separated from the tumor capsule to allow dissection of the tumor and capsule within this arachnoid envelope. In 2002, Ohata et al,16 through the operative findings of their sequential cases, put forward the concept of subarachnoidal origin of VSs. He suggested that the VS originates in the subarachnoid within the IAM and grows epiarachnoidally in the CPA. Rearrangement of the arachnoid begins with its adhesion to the medial pole of the tumor along the porus, resulting in arachnoidal invagination into the cerebellopontine cistern with further growth of the tumor. With this understanding, we considered that VS resection should be performed along the capsule of the tumor, maintaining the integrity of the arachnoid, to avoid nerve and blood vessel damage. After maximal internal debulking, the tumor outside the central field of vision was gradually pushed back into the surgical area via the pulsatile effect of CSF and elastic recoil of brain tissue back. Making the best of the Cavitron ultrasonic aspirator would tremendously shorten the operation time and facilitate the operation process.17 After significantly reducing the volume, the tumor can be easily dissected in a sequence. The primary purpose of dissection of the inferior pole of the tumor was to preserve the lower cranial nerves and the posterior inferior cerebellar artery and its branches. In this region, lower cranial nerves usually did not significantly attach to the tumor capsule and were relatively simple to mobilize away from the tumor capsule. After finishing the step, we subsequently kept on internal debulking along the superior pole of the tumor and peeled away the capsule from the tentorium and the fifth cranial nerve. The petrosal vein was protected if possible. The facial nerve was often adherent along the superior pole of the tumor; more attention and patience must be paid. When separating the interface between the tumor and the brainstem, the tumor capsule was gently rolled laterally away from the middle cerebellar peduncle toward the porus acusticus. In the presence of preoperative brainstem edema, violation of the pial membranes was quite common. Once this event occurred, the surgeon should make full use of the microdissector to peel away the tumor from the brainstem. The soaked gelatin sponge was put into the space between the tumor and the brainstem and left behind to avoid the tumor back. For the ventral, identification of the facial nerve was mainly based on familiarity with anatomic landmarks and intraoperative facial nerve monitoring. The eighth cranial nerve locates immediately behind the facial nerve at the lateral aspect of the pontomedullary sulcus and joins the brainstem in front of the site of attachment to the flocculus. Meanwhile, the facial nerve is usually white, and the acoustic nerve is gray. The use of these anatomic landmarks will assist in the identification of the root exit zone of the facial nerve during tumor resection. Through combining this information with the route of the facial nerve over the capsule conveyed by removing the tumor within the IAM, the surgeon could generally learn the anatomic location of the facial nerve and continue
* 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
467
The Journal of Craniofacial Surgery
Daming et al
the end-to-end dissection of the nerve from the tumor. The IAM is the most common site of serious adhesion between the facial nerve and the tumor. Ohata et al16 reported that because of tumor growth, the arachnoid matter covering the tumor occasionally became incomplete or even disappeared, resulting in local adhesion between the tumor and the facial nerve. We encountered 1 patient with the similar conditions. To avoid damage to the facial nerve, a small piece of tumor was not removed. Common problems after the removal of a large VS are CSF leakage, meningitis, and other neurological complications.18 In our cases, most of patients were cured with conservative methods. Facial nerve function was assessed in 6 to 12 months after operation. House-Brackmann grades IYII were present in 81.1% of the patients, with HB grade III in 11.1%, and with HB grade IV in 5.4%. Three patients had varying degrees of hearing preservation; their hearing roughly returned to preoperative level (at 50Y60 dB) 6 months later after operation. Two of them underwent gross tumor resection, and 1 patient underwent partial tumor removal. Despite the small number of cases, which is insufficient to give a hypothesis, these findings suggested that dissection along the capsule of the tumor and maintaining the integrity of the arachnoid were prerequisites for preservation of normal nerve function. In summary, the suboccipital retrosigmoid keyhole approach, by means of the amplifying effect of keyhole, uses a smaller bone defect to optimize surgical vision and avoid unnecessary exposure or destruction of the skull structure. Our data illustrate that this approach combined with appropriate surgical strategies and operating techniques can facilitate effective and complete large VS resection.
REFERENCES 1. Huang X, Zhang J, Yang H, et al. Use of intraoperative ultrasonography to monitor surgery for large acoustic neuromas: a pilot study. J Med Ultrasonics 2010;37:15Y19 2. Samii M, Gerganov VM, Samii A. Functional outcome after complete surgical removal of giant vestibular schwannomas. J Neurosurg 2010;112:860Y867 3. Kulwin CG, Cohen-Gadol AA. Technical nuances of resection of giant (95 cm) vestibular schwannomas: pearls for success. Neurosurg Focus 2012;33:1Y8 4. Jung S, Kang SS, Kim TS, et al. Current surgical results of retrosigmoid approach in extra large vestibular schwannomas. Surg Neurol 2000;53:370Y378
468
& Volume 25, Number 2, March 2014
5. Kim WH, Park CK, Kim DG, et al. Management of recurrent vestibular schwannomas. J Korean Neurosurg Soc 2006;39:87Y91 6. Comey CH, Jannetta PJ, Sheptak PE, et al. Staged removal of acoustic tumors: techniques and lessons learned from a series of 83 patients. Neurosurgery 1995;37:915Y921 7. Yamakami I, Uchino Y, Kobayashi E, et al. Removal of large acoustic neurinomas (vestibular schwannomas) by the retrosigmoid approach with no mortality and minimal morbidity. J Neurol Neurosurg Psychiatry 2004;75:453Y458 8. Jamro´z B, Niemczyk K, Morawski K, et al. Extended middle fossa approach in treatment of vestibular schwannoma technique of surgery and postoperative complications. Otolaryngol Pol 2010;64:3Y9 9. Raslan AM, Liu JK, McMenomey SO, et al. Staged resection of large vestibular schwannomas. J Neurosurg 2012;116:1126Y1133 10. Yokoyama T, Uemura K, Ryu H, et al. Surgical approach to the internal auditory meatus in acoustic neuroma surgery: significance of preoperative high-resolution computed tomography. Neurosurgery 1996;39:965Y970 11. Stieglitz LH, Giordano M, Gerganov VM, et al. How obliteration of petrosal air cells by vestibular schwannoma influences the risk of postoperative CSF fistula. Clin Neurol Neurosurg 2011;113:746Y751 12. Lu¨demann WO, Stieglitz LH, Gerganov V, et al. Fat implant is superior to muscle implant in vestibular schwannoma surgery for the prevention of cerebrospinal fluid fistulae. Neurosurgery 2008;63(1 suppl 1): ONS38YONS42; discussion 42Y43 13. Yingling CD, Gardi JN. Intraoperative monitoring of facial and cochlear nerves during acoustic neuroma surgery. Otolaryngol Clin North Am 1992;25:413Y448 14. Yasargil MG, Smith RD, Gasser JC. Microsurgical approach to acoustic neurinoma. Adv Tech Neurosurg 1977;4:93Y129 15. Tarlov E. Total one-stage suboccipital microsurgical removal of acoustic neuromas of all size: with emphasis on arachnoid planes and on saving the facial nerve. Surg Clin N Am 1980;60:565Y591 16. Ohata K, Tsuyuguchi N, Morino M, et al. A hypothesis of epiarachnodal growth of vestibular schwannoma at the cerebellopontine angle: surgical importance. J Postgrad Med 2002;48:253Y259 17. Samii M, Gerganov V, Samii A. Improved preservation of hearing and facial nerve function in vestibular schwannoma surgery via the retrosigmoid approach in a series of 200 patients. J Neurosurg 2006;105:527Y535 18. Charpiot A, Tringali S, Zaouche S, et al. Perioperative complications after translabyrinthine removal of large or giant vestibular schwannoma: outcomes for 123 patients. Acta Otolaryngol 2010;130:1249Y1255
* 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
