Endoscope-Assisted Retrosigmoid Approach for Vestibular Schwannomas With Intracanalicular Extensions: Facial Nerve Outcomes

Objective To explore the role of neuroendoscope assistance during surgical resection of the intracanalicular portion of vestibular schwannomas via the retrosigmoid approach and the subsequent early facial nerve outcomes. Methods Patients of vestibular schwannoma with intracanalicular extensions undergoing retrosigmoid dissection at a single institution were retrospectively analyzed in this study. Several surgical techniques were applied to ensure maximal and safe removal of tumors. Tumors extending less than 10 mm into the internal acoustic canal (IAC) were classified as Grade A, while those extending over 10 mm into IAC were taken as Grade B. Neuroendoscope was applied at the end of microscopic phase to search for potential remnants for Grade B tumors. Absolute tumor extension was defined and measured. House and Brackmann (HB) scale was used to evaluate immediate CN VII outcomes. Results Of the 61 patients, there were 38 females and 23 males. A total of 18 (29.51%) cases were Koos Grade II, 12 (19.67%) cases Koos Grade III, and 31 (50.82%) cases Koos Grade IV. There were 38 cases (62.30%) of Grade A and 23 cases (37.70%) of Grade B. Gross total resection was achieved in 60 cases (98.36%). Four cases of intracanalicular remnants were detected and completely removed under endoscopic visualizations. There was a significantly higher proportion (17%, p = 0.02) of intracanalicular remnants in Grade B than Grade A. CN VII and VIII were anatomically preserved in all cases. A total of 55 cases (90.16%) retained good (HB Grades 1 and 2) facial nerve outcomes. Conclusions In Grade B vestibular schwannomas, after maximal microsurgical removal, endoscopic evaluation of the intracanalicular portion revealed residual tumors in 17% of the patients. Hence endoscopic evaluation of the potential intracanalicular remnants for tumor extending over 10 mm within IAC (Grade B) is recommended.


INTRODUCTION
Vestibular schwannoma is the most common benign tumor in the cerebellopontine angle (CPA). Management goal is to achieve optimal resection while preserving local structures, especially for CN VII (1)(2)(3). The retrosigmoid approach is the workhorse for management of CPA lesions. Most vestibular schwannomas present with intracanalicular extensions (4). In retrosigmoid approach, drilling the posterior wall of IAC is one of the key steps to dissect the intracanalicular portion of the tumor. However, for tumors located at the lateral end of IAC, "blind spots" under the microscope may impede gross total resection. With advances of techniques and technologies, the application of an endoscope can potentially solve the problem. However, the role of an endoscope during microsurgery has not been fully investigated. Here we report a retrospective single-operator series of 61 cases of vestibular schwannoma with intracanalicular extensions, where an endoscope was applied at the end of the microscopic phase to confirm the presence of remnants if tumor extends over 10 mm into IAC. Early facial nerve outcomes were satisfactory, and the post-operative course was uneventful.

Sample Selection
After approval of the institutional review board, patients with pathologically confirmed vestibular schwannoma from November 1, 2019 to December 31, 2020 at the Department of Neurosurgery, Shanghai General Hospital, Shanghai Jiao Tong University were retrospectively reviewed. Medical records, imaging studies, and surgical footages were evaluated by senior authors (JY and YL) independently; any discrepancy was discussed and solved.

Defining the Absolute and Relative Tumor Extension Into IAC
Imaging studies using heavily T2-weighted fast imaging employing steady-state acquisition with cycle phase (FIESTA-C) MRI were reviewed by two authors independently. A straight line (labeled as L; Figure 1) was drawn medially from the most lateral point of the intracanalicular portion to the midpoint between the anterior and posterior lip of the ipsilateral porus acusticus on the same axial slice. The absolute extension of the tumor into IAC was graded by the length of L, where Grade A was defined as L <10 mm and Grade B was for L ≥10 mm.

Tumor Resection
This is a single-surgeon retrospective series in which all cases were performed by the senior author (ML) and the assistants. Since most of the procedural details have been well established in literature, they will not be exhaustive here. All patients were placed in the park bench position for retrosigmoid approach regardless of the tumor size. Neuromonitoring with a stimulus threshold of 0.1 mA was applied throughout the procedure to preserve the functions of CN VII. We started with initial debulking of the extracanalicular portion at its center and proceeded to early drill-out of the posterior wall of IAC (detailed below). After removal of the intracanalicular portion, we proceeded to decrease tumor burdens at the CPA cistern. In this paper, we focus on surgical nuances that are critical to ensure maximal tumor removal and preservation of CN VII functions.
After defining the cleavage plane, care should be taken to protect the perineurium as much as possible while performing subperineurial dissection. In practice, when the cleavage plane was initially "toward" the observer, microscopic forceps were used to peel off the perineurium. Conversely, when the cleavage plane was initially "away" from the observer, microdissectors should be used to "push away" the perineurium. In certain conditions, strict adhesions of the perineurium were sometimes encountered when neither "peeling" nor "pushing" was successful. It was then necessary to use sharp dissection by blunt-tipped microscissors. Strict adhesions were usually encountered at the porus acusticus, where CN VII was prone to inadvertent injuries. We would deal with this site at the end stage of the tumor removal procedure.
The drilling range depended largely on the local anatomy and intracanalicular extension of the tumor. The Tübingen line was first identified to locate the course of the posterior wall of IAC. For all Grade A and B tumors, an average of 8 mm of the posterior wall of IAC was first removed, with a drilling angle of approximately 43° (Figure 2). About 12-14 mm of dura was removed and incised with a No. 15 blade along the posterior petrous surface (Figure 2). These were not strict doctrines to follow but tangible rules to be individually adjusted. If a highriding jugular bulb (HJB) was identified as drilling came close to it, an artificial dural substitute was used to slightly displace the HJB. We initially chose a 4 mm diamond drill bit (Medtronic plc, Minnesota, United States), and switched to smaller calibers (e.g., 2.3 mm) with further drilling. A 180˚exposure of the posterior circumference of the porus acusticus was made. During drilling, tactile feedback on the instrument was important to indicate the proximity of semicircular canal (SCC) and the posterior wall of IAC.
Techniques of managing the intracanalicular portion were different from those of the CPA cistern. Bipolar forceps were not allowed. Following drilling of the posterior wall of IAC, dura was incised open, and we started with tumor debulking to lessen the tumor burden on nerves. After confirmation by stimulation, the vestibular nerve was sacrificed. Tumor removal was performed at the CN VIII (caudal) side, followed by the CN VII (cephalad) side. We prefer fine-tipped forceps, blunt-tipped scissors, finetipped suction tubes (1.2-1.5 mm) and round or flat-tipped dissectors to perform these manipulations to decrease the risk of nerve injuries. During the procedure, tumor resection was proceeded under microscopic view with pulsatile (set interval: about 2 s) irrigation of lukewarm (37°C) natural saline. Manipulating force was applied to the tumor only, and direct contact of instruments with the nerve should always be avoided.
After the microscopic phase, a 0-degree neuroendoscope (2.9 mm; Karl Storz Endoskope, SE & Co. KG, Tuttlingen, Germany) would be applied to search for potential remnants in Grade B tumors. If confirmed, angled nerve hooks should usually suffice to palpate and "undress" tumor remnants. In some cases, an extra 2 mm of the posterior wall of IAC (hence a total of 10 mm) would be drilled under endoscopic view to provide space for more laterally located tumor remnants. Angled instruments, including angled nerve hooks, angled dissectors ( Figure 3), and angled suction tubes (1.5 mm), were superior to other instruments at the endoscopic phase. These angled, personally curated instruments were modified from market-sold models (Symmetry Surgical GmbH, Tuttlingen, Germany), and were highly efficient in removing the remnants located at the fundus of IAC. To remove the tumor, we performed counter-clockwise maneuvers to separate the left-sided tumors and clockwise maneuvers for the right-sided ones. The tumor was separated from CN VII in a circular motion, with force mainly applied to the tumor body instead of CN VII.

Evaluation of Facial Nerve Functions
Immediate post-operative facial nerve functions were graded on the House and Brackmann (HB) scale after patient awakening from anesthesia in the recovery unit.

Statistical Analysis
Stata/SE 17.0 (StataCorp LLC, Texas, USA) was utilized for statistical analysis. Fisher's exact test was used to calculate differences of frequencies in categorical variables. A p-value <0.05 was deemed significant to reject null hypothesis.  FIGURE 2 | The ideal drilling can maximize tumor exposure while avoiding injuries to neighboring structures (i.e., the bony labyrinth, the endolymphatic sac). To date, neither a standardized surgical planning nor a precise drilling method has been introduced to clinical practice. However, drilling the posterior wall of IAC is not an improvised technique but has flexible rules to follow. The artist's (Hongchan Li) illustration here shows several essential parameters in defining the safe drilling range of the posterior wall of IAC. Lines AB and CD correspond to the "entry borders" of lateral drilling, which is approximately 12-14 mm in length. This means that approximately 12-14 mm of dura along the petrosal surface is incised, after identification of the Tübingen line. ∠ABE is the drilling angle, usually 43˚. Line GF denotes the length of the drilled posterior wall of IAC, which is about 8 mm in length. Line AE denotes the antero-superior border of the drilling range. Of note, these parameters are individualized to each patient.

Surgical Outcomes and Post-Operative HB Scale
Of the 61 cases, 60 cases (98.36%) achieved gross total resection and 1 case (1.64%) underwent sub-total resection ( Table 1). Intracanalicular remnants were found and completely removed in 4 cases after endoscopic inspection. The proportion of intracanalicular remnants in Grade B (4/23, 17%) was significantly higher (Fisher's exact; p = 0.02) than that of Grade A (0/38). HJB was identified in one of the four remnant cases (Figure 4). Another patient presented with recurrent tumor invading the adjacent bone proper. The fundus of IAC was eroded, which conferred an irregular shape on the tumor. The narrow corridors were navigated using neuroendoscope to facilitate resection ( Figure 5). CN VII and VIII were anatomically preserved in all patients. Regarding post-operative HB scale, 50 cases (81.97%) were Grade 1, 5 cases (8.20%) Grade 2, 4 cases (6.56%) Grade 3, and 2 cases (3.28%) Grade 4 ( Table 1). The results of hearing outcomes are beyond the scope of this article.

DISCUSSION
Gross total resection has been associated with improved quality of life and potentially long-term control of vestibular schwannomas (3,5,6). Most vestibular schwannomas have intracanalicular portions. One of the most common causes of vestibular schwannoma recurrence is remnants at the fundus of IAC (6,7). To achieve maximal resection for tumors with intracanalicular extension, care must be taken to prevent concomitant injuries to the posterior SCC, the endolymphatic duct, and a potential HJB (8,9). Injuries to the former two structures may cause hearing loss, while injuries to the later may lead to hemorrhage and air embolisms (9). Retrosigmoid approach, a workhorse in surgical management of CPA lesions, is a versatile option in terms of preserving functions in most tumor sizes (10)(11)(12)(13). However, the preservation rate of CN VII functions using retrosigmoid approach vary from 80 to 92% in tumor sizes of 1-2 cm to 50-76% in those > 2cm (14). Many factors, namely, surgical skills, tumor sizes, tumor extensions (intracanalicular and/or extracanalicular), cystic characteristics, and the topography of IAC may all or in part contribute to worse CN VII outcomes (15)(16)(17)(18)(19)(20)(21)(22)(23). Due to a relatively smaller sample size, we did not find any pre-operative CN VII dysfunctions in our sample. This would in effect diminish the impact of preoperative HB scale on post-operative HB scale.
To reach the intracanalicular portion, drilling the posterior wall of IAC is one of the key steps in retrosigmoid approach (8,24). The surgical nuance here is to ensure maximal exposure of the posterior wall of IAC with protection of neighboring structures. In a healthy individual, the shape of porus acusticus is oval or elliptical (25). The average length of IAC ranges from 5.5 to 12.3 mm, mean diameter 4 mm, according to the present literature (26,27). To date, the standard of optimal drilling range has not been established. At least three parameters are required to define a drilling range: the length of lateral drilling, the length of the drilled posterior wall of IAC, and the drilling angle. The length of lateral drilling was proposed by some experts to be 10 ± 2 mm (28). The length of drilled posterior wall of IAC ranged from 5 to 10 mm, according to the present literature (28)(29)(30). Using frameless navigation on 10 cadaveric heads, Pillai et al. concluded the angle of drilling as 43.3 ± 6.0˚ (24). In our experience, we made a "flexible rule", in which the length of lateral drilling was approximately 12-14 mm, drilling angle 43å nd the length of drilled posterior wall of IAC 8 mm (Figure 2). After about 8 mm of the posterior wall of IAC was drilled, there were still blind spots at the lateral end of IAC under microscopic view. If tumor remnants were located there, the risk of nerve injuries under "blind" manipulations would increase. Therefore, the application of an endoscope should theoretically minimize the risks of nerve injuries caused by "blind dissection".
Many reports of endoscopic usage in vestibular schwannomas and/or other CPA lesions were seen in literature (24,(31)(32)(33). The purely endoscopic approach requires fewer brain retractions, and the advantage of increased visualization at IAC by navigating through narrow corridors would minimize the risks of complications (34). However, the usefulness of purely endoscopic approach for vestibular schwannomas is still in debate (34)(35)(36). Having an assistant to hold the endoscope with both hands may overcome the single-handed technique by the operator, but the dynamic movement of the endoscope requires constant searching for a pivoting point for balance. The bulky manipulation by two hands and the time-consuming nature of endoscopic surgery may compromise visual stabilization. A pneumatic endoscope holder is surely more versatile but less flexible than freehand manipulation. In addition, navigating solely by endoscope is prone to disorientation and concomitant injuries to the vital and complex neurovasculature of CPA (37). Therefore, microscopic resection is still the mainstay of treating vestibular schwannomas.
Using a neuroendoscope after microscopic phase is the optimal tactic to circumvent these problems. The surgery should always start with microscopic dissection, which is familiar and less time-consuming for many neurosurgeons. For tumors with less extension into IAC, microscopic dissection would be adequate. While in tumors with ≥10 mm extension into IAC, we would opt for an extra endoscopic phase. The optimal visualization ( Figure 6) brought by the neuroendoscope would identify potential tumor remnants and facilitate gross total resection using angled instruments (35,36). If a routinely performed 8 mm of drilling is not enough for more laterally located tumors, an extra 2 mm could be drilled under endoscopic visualization to manage more laterally located tumor remnants.
In practice, neuroendoscopy involves a significant learning curve that demands extensive training. During the surgery, the assistant needs to be familiar with not only the procedure, but also the preferences of the operating surgeon. It is also important for the assistant to irrigate the neuroendoscope to avoid debris accumulation on the tip and thermal injuries to nerves. Above all, in the hands of an experienced neurosurgical team, endoscopeassisted microsurgery of vestibular schwannomas should be

CONCLUSIONS
In Grade B vestibular schwannomas, after maximal microsurgical removal, endoscopic evaluation of the intracanalicular portion revealed residual tumors in 17% of the patients. Hence endoscopic evaluation of the potential intracanalicular remnants for tumor extending over 10 mm within IAC is recommended.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the Institutional Review Board of Shanghai General Hospital. Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin.

AUTHOR CONTRIBUTIONS
YB, YN, DG, and ML conceived of and directed the study. QinZ, QiaZ, and JT provided clinical information and analysis the data. JL and FS participated in data acquisition. HL was responsible for data acquisition and figure productions. JY and YL reviewed the manuscript. All authors contributed to the article and approved the submitted version.