Surgery After Surgery for Vestibular Schwannoma: A Case Series

Objective We retrospectively evaluated the oncological and functional effectiveness of revision surgery for recurrent or remnant vestibular schwannoma (rVS). Methods We included 29 consecutive patients with unilateral hearing loss (16 women; mean age: 42.2 years) that underwent surgery for rVS. Previous surgeries included gross total resections (GTRs, n=11) or subtotal resections (n=18); mean times to recurrence were 9.45 and 4.15 years, respectively. House–Brackmann (HB) grading of facial nerve (FN) weakness (grades II-IV) indicated that 22 (75.9%) patients had deep, long-lasting FN paresis (HB grades: IV-VI). The mean recurrent tumor size was 23.3 mm (range: 6 to 51). Seven patients had neurofibromatosis type 2. Results All patients received revision GTRs. Fourteen small- to medium-sized tumors located at the bottom of the internal acoustic canal required the translabyrinthine approach (TLA); 12 large and small tumors, predominantly in the cerebellopontine angle, required the retrosigmoid approach (RSA); and 2 required both TLA and RSA. One tumor that progressed to the petrous apex required the middle fossa approach. Fifteen patients underwent facial neurorrhaphy. Of these, 11 received hemihypoglossal–facial neurorrhaphies (HHFNs); nine with simultaneous revision surgery. In follow-up, 10 patients (34.48%) experienced persistent deep FN paresis (HB grades IV-VI). After HHFN, all patients improved from HB grade VI to III (n=10) or IV (n=1). No tumors recurred during follow-up (mean, 3.46 years). Conclusions Aggressive microsurgical rVS treatment combined with FN reconstruction provided durable oncological and neurological effects. Surgery was a reasonable alternative to radiosurgery, particularly in facial neurorrhaphy, where it provided a one-step treatment.


INTRODUCTION
Vestibular schwannoma (VS) surgery has changed beyond recognition since the early 20 th century, when it was associated with mortality rates as high as 86% (1). The subsequent drop to 15.4% mortality, due to contributions from Harvey Cushing, was considered a milestone in VS surgery (2). In contrast, contemporary skull base surgery includes gross total resections (GTRs), but also aims for a good functional outcome, mostly by preserving facial nerve (FN) and other cranial nerve functions (3)(4)(5)(6) in addition to reducing the mortality rate (7,8).
Tumor recurrence or progression management remains controversial. In our department, we follow the VS management principles described by Samii, namely a GTR with the preservation of facial and cochlear nerve functions (23). A second microsurgery is much more difficult to perform than an initial GTR, due to adhesions and scarring, and it is even more difficult after a radiosurgery (24). In this retrospective case study, we described a series of patients that underwent revision surgery for VS and, in some cases, they received different methods of FN reanimation (25)(26)(27). In this study, we aimed to evaluate the oncological and functional effectiveness of revision surgery for recurrent and residual VS (together referred to as rVS).

MATERIALS AND METHODS
This study was designed as a retrospective, single-center, consecutive case series, undertaken in an academic setting. All participants met the following criteria: • Previous history of surgery for a VS, with or without radiation therapy; • Histopathologic diagnosis of a VS, during the first and second surgeries; • Regrowth, diagnosed in radiological imaging, or a residual tumor in patients with FN palsy; • Revision surgery required for VS.
No specific exclusion criteria were applied. Patients underwent reoperations from 2002 to 2018. Data were collected between 2018 and 2019. We analyzed 29 patients (16 women, 13 men; mean age: 42.2 years) that underwent surgery for rVS. The mean time to revision surgery was 6.16 years ( Table 1).
Eleven patients (37.9%) had undergone previous GTR surgeries in our department with the retrosigmoid approach (RSA). In this group, the mean time to revision surgery was 9.45 years (range 4 to 19). Eighteen other patients (62.1%) were referred from other centers, due to tumor progression after subtotal resections (all RSA); among these, 4 patients had undergone additional radiosurgery. The mean time to revision surgery was 4.15 years (range 0.5 to 16). Over one third (37.9%; 11/29) of the initial surgeries were GTRs; 55.2% (16/29) were subtotal resections; and 2 were unknown, due to the lack of available data.
All patients presented with unilateral hearing loss (American Academy of Otolaryngology-Head and Neck Surgery class D  : IV-VI; Tables 1, 2). Recurrent tumor sizes ranged from 6 to 51 mm (mean: 23.3 mm). Seven patients (24.1%) had neurofibromatosis type 2 (NF2). In those cases, surgery notes were reviewed to ensure that the tumor did not originate from a location other than the eighth cranial nerve.
The following pre-intervention procedures were included specifically for this study: when possible, descriptions of previous surgeries were acquired, and current scans were performed with magnetic resonance imaging (MRI), electromyography for the FN, and a bone window computed tomography (CT) for the temporal bone. Otherwise, all patients were prepared for elective surgery in a routine fashion. Hearing was not tested, because all patients were unilaterally deaf.

Surgery Details
Each surgery was performed under general anesthesia, with the patient positioned on the back, with the head rotated contralaterally. The TLA was used for 14 small-to mediumsized tumors (median maximal size: 15.63 mm) that arose from the bottom of the internal acoustic canal (IAC). The RSA was used for larger tumors (median maximal size: 28.33 mm) and for smaller tumors (n=12) that were predominantly located in the CPA. In two cases, a combination of TLA and RSA was performed. The middle fossa approach was employed for one tumor that had progressed to the petrous apex ( Figure 1).
Intraoperative neurophysiological monitoring was performed only in selected cases. Hearing was not monitored, because all patients were unilaterally deaf. FN was monitored in patients that had at least partially preserved FN function after a previous surgery.
Patients that underwent facial neurorrhaphy were instructed by a neurophysiotherapist to perform self-massaging of the facial muscles postoperatively. Then, when the first signs of muscle reinnervation appeared, patients were to perform exercises in front of the mirror. Patients were monitored postoperatively, as follows: outpatient clinic visits at ½, 1, 1.5, and 3.5 years after surgery; and MRIs at ½, 1,5, and 3.5, years after surgery.

General and Oncological Results
There were no deaths. Postoperative complications occurred in four patients (13.8%). These included one subcutaneous hematoma, after an abdominal fat harvest (treated with a wound revision); one otorrhea (treated with a temporary lumbar drainage); one rhinorrhea (treated with duraplasty); and one otorrhea combined with chronic otitis media (treated with temporary lumbar drainage and a petrosectomy). The cerebrospinal fluid (CSF) leak rate was 10.3% (3/29).
In all cases but one, GTR was accomplished during the second surgery. The one unsuccessful GTR was performed in a patient admitted in poor general condition (Modified Rankin Scale 5), and revision of a previous partial VS resection improved his neurological deficits, but not significantly. Among the other 28 patients, no subsequent tumor recurrence was noted during amen follow-up of 3.46 years (range: 0.25-11).

Facial Nerve Functional Results
The FN function improved during follow-up in 69.0% (20/29) of patients, and it remained unchanged in 24.1% of patients (7/29). Additionally, two patients (6.9%, 2/29) experienced worse FN function after revision surgery. Both of those patients had been initially treated in our department, 8 and 9 years prior to revision surgery. They initially had HB grades of II and IV, and after revision surgery, their HB grades were IV and VI, respectively. Neither of these patients was eligible for facial neurorrhaphy, due to HB grade IV, in the former patient, and long-lasting FN paralysis, in the latter patient. When analyzed separately, 71.4% of patients (5/7) with NF2 experienced better FN function after revision surgery, and in 28.6% (2/7) it remained unchanged.
During the last 5 years of the study (revision surgeries from2013 to 2018), 14 patients had available data. Of these, 9 received the HHFN technique, and the results were satisfactory results in each case. Due to our complex surgical strategy, 11 (78.6%) patients moved from an unsatisfactory functional grade (HB grades VI-VI) to a satisfactory functional grade (HB grades I-III, Table 4).

Illustrative Case 1-Middle Fossa Revision and Classic Facial Neurorrhaphy
An 18-year old male underwent a GTR with an RSA for a sporadic VS in our department. Then, he underwent a classic hypoglossalfacial neurorrhaphy. His FN function improved to Values are n (%) or the mean (range), as indicated. *Twice for one patient. CPA, cerebellopontine angle; IAC, internal acoustic canal; GTR, gross total resection; MFA, middle fossa approach; N/A, not available; RSA, retrosigmoid approach; TLA, translabyrinthine approach. HB grade IV. Eight years later, he was diagnosed with a recurrent tumor, with a maximal size of 33 mm, which had progressed toward the petrous apex. The patient underwent revision surgery via a middle fossa approach, and the tumor was totally resected. The postoperative course was uneventful. No subsequent tumor regrowth was noted in a follow-up of 11 years (Figure 1).

Illustrative Case 2-Retrosigmoid Revision and Hemihypoglossal-Facial Neurorrhaphy
A 54-year old female presented with FN paresis after a subtotal VS resection in another center one year earlier. The neurological examination at admission revealed deafness, right sided dysmmetry, dizziness, subjective diplopia, hypoaesthesia on the right side of the face, and paresis of the soft palate, but no difficulties in swallowing. Her follow-up MRI revealed a tumor remnant, located in the right CPA, intrameatally. She was eligible for revision surgery via RSA with a simultaneous HHFN. After revision surgery, the patient developed hypoglossal paresis, which subsequently resolved. Two months later, the patient developed otorrhea with chronic otitis media. A CT revealed a fistula between CPA, mastoid air cells and the external acoustic canal. The patient underwent petrosectomy. The subsequent postoperative course was uneventful; her follow-up MRI confirmed the completeness of the VS resection. In a 2-year follow-up, FN function improved to HB grade III (Figure 2).

Illustrative Case 3-Translabyrinthine Revision and Hemihypoglossal-Facial Neurorrhaphy
A 53-year old male presented with right facial and abducens nerve paresis and right-sided deafness after a partial VS removal 6 months earlier in another center. A tumor remnant was located close to the porus acusticus. He was eligible for a tumor remnant resection with a TLA and simultaneous HHFN. The postoperative course was uneventful. He was discharged one week after surgery. At a 1-year follow-up, his FN function improved to HB grade III, and a postoperative MRI revealed no tumor remnant (Figure 3).

Vestibular Schwannoma Recurrences and Remnants
The source of a VS recurrence might be a microscopic tumor remnant on the cranial nerves or in the fundus of the internal auditory canal (24). To address these possibilities, various attempts have been made to maximize the VS resection via a retrosigmoid or middle fossa approach; i.e., with endoscopy (28)(29)(30). However, in a previous retrospective study, Panigrahi et al. found that the Ki-67 labeling index, rather than the extent of resection, was associated with VS recurrences (31). Additionally, Freeman et al. hypothesized that a recurrence or regrowth after a GTR was more likely to reflect the biological behavior of the individual tumor, rather than the size of a tumor residue (32). On the other hand, the number and relative contribution of revision surgeries for growing tumor remnants will probably increase, due to the current popularity of less invasive approaches, like non-total "functional" resections, with or without radiosurgery (20,33). Therefore, complex solutions should be developed for the most demanding resections.

Management Options
There are three therapeutic options for treating a VS, including observation, revision surgery, and radiosurgery. According to Tomita et al., a small remnant VS after microsurgery could be managed conservatively without additional treatment (34). Similarly, Troude et al. preferred a "wait and scan policy" over a postoperative SRS (35). Although the "wait and scan" policy is advantageously non-invasive, it was previously observed that, in certain patients with VS, the quality of life was mostly improved after microsurgery. However, one potential explanation for that result might be that the patient was not told that he/she had an intracranial tumor; this knowledge appeared to cause an intolerable burden, regardless of the tumor size, even though tumors often remained stable in observation groups (36,37). Revision surgery is the only method that actually removes a tumor, but the trade-off is the high risk of postoperative complications, compared to other alternatives. Stereotactic radiosurgery is a less invasive alternative to microsurgery, but it does not actually eliminate the tumor. Moreover, large or giant tumors usually cannot be treated with SRS; therefore, they require microsurgical resection (38,39). Furthermore, it should not be forgotten that, although rare and controversial, a VS can be malignantly transformed after a SRS, and radiation can induce a VS (40)(41)(42)(43)(44).

Aim of the Revision Surgery
The aim of the second surgery should be an oncological cure and preserving FN function. Alternatively, in cases with FN paralysis, neurorrhaphy techniques should be carried out alongside the tumor resection, when the time period after paralysis onset allows. In contrast to the initial VS treatment, a second surgery does not usually aim to preserve hearing, because the patients are unilaterally deaf.

Complications
Several previous investigations of microsurgery for rVS showed satisfactory results and low complication rates (45). The rate of postoperative complications in our series was 10.3%, and the CSF leak rate was 13.8%. Three patients required different surgical revisions without intradural CPA inspections. Perry et al. described two cases in a 6-patient series where a CSF leak (33%) was successfully treated with a lumbar drain and surgical revision. Their summary of a previous series indicated a 4% CSF leak rate (45). Apart from that study, no other study reported major complications (24,32,45,46).

Approach Selection
When considering the rVS treatment approach, the surgeon should bear in mind both the location of the tumor and its size. Typically, the approach should be different than that previously used to avoid scar tissue, which can adhere tightly to the cranial nerves, brainstem, or cerebellum (32). In our experience, the TLA was most suitable for resecting minor remnants, when the primary aim was an HHFN, because, after skeletonizing the FN in the Fallopian canal, limited expansion of the bony drilling was required to access the tumor. On the other hand, for major tumor progressions toward the brain stem or petrous apex, the RSA and middle fossa approach are also reasonable choices.

Oncological Results
In

Functional Results
In our series, the rate of deep FN deficits (HB grades IV-VI) was reduced from 75.9% to 34.5% in the long-term follow-up (p <0.05, Fisher's exact test, Table 2). Considering only the results from the last 5 years of our study, 82% (9/11) of patients with complete FN paralysis recovered satisfactory function. According to Samii et al., the main predictor of FN outcome in a second VS surgery was the level of facial function before revision surgery (24). Postsurgical results can be further improved by combining revision surgery with facial neurorrhaphy. In our series, after an HHFN, all patients improved from HB grade VI to HB grade III, except one, who improved to grade IV. Fourteen of our patients did not undergo facial neurorrhaphy, partly because the timeframe was exceeded for performing this procedure (n=7). Prolonged times between the FN paralysis onset and neurorrhaphy was correlated with worse FN outcomes (27). The remaining 7 patients had initial HB grades of II to IV, which excluded facial neurorrhaphy during revision surgery. In follow-up, the HB grades of this subgroup improved, except in 3 patients: the HB grade remained unchanged in 1 patient, and it worsened in 2 patients. For patients with unsatisfactory long-term FN function, plastic surgery was advised.
A substantial number of our patients had unsatisfactory FN function, particularly during the early part of the study period. The two main explanations were that some patients did not consent to receive face reanimation procedures and some surgeons underestimated the significance of postoperative FN weakness. Nineteen patients presented with HB grade VI after the first surgery; of these, 13 (68.4%) were treated initially in

Our Strategy for Recurrent and Residual VS
We performed GTRs during revision surgery in 97%of patients, and we observed FN functional improvements in 69% of patients. Based on these results, we suggest that revision surgery should be considered when: (1) a patient requires facial neurorrhaphy and has a stable residual tumor; or (2) the tumor has regrown (here, the HHFN should be performed, when necessary). Moreover, we prefer neurorrhaphy techniques, rather than plastic surgery, up to the end of the 3 rd year after paresis onset (27). Additionally, it is important to provide a multidisciplinary, complex approach, from the diagnosis to late follow-up, with close surveillance of postoperative FN function.

Study Limitations
This study was limited by its retrospective, single-center design. Additionally, patients with and without NF2 were analyzed together, despite differences in the management of these two groups.

CONCLUSIONS
This study demonstrated that aggressive microsurgical rVS treatment, together with modern FN reconstruction techniques provided an acceptable risk profile, yielded durable oncological effects, and could restore satisfactory FN function. We found that surgery was a reasonable alternative to SRS, particularly in patients that required facial neurorrhaphy, because it offered a one-step treatment.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by The Bioethics Committee of the Medical University of Warsaw. The patients/participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.

AUTHOR CONTRIBUTIONS
PK designed the project and critically reviewed it. ŁP prepared the manuscript, collected and reviewed data. WR and TD collected case data. AM did final approval of the version to be published. All authors have participated sufficiently in the conception and design of this work or the analysis and interpretation of the data, as well as the writing of the manuscript, to take public responsibility for it. All authors contributed to the article and approved the submitted version.