Surgical Outcomes and Predictive Factor Analysis for Facial Nerve Preservation in Patients With Cerebellopontine Angle (CPA) Tumors: A Ten-Year Single Institutional Study

Objective To analyze the surgical outcome and predictive factors for facial nerve preservation in patients with surgically operated cerebellopontine angle (CPA) tumors. Methodology Methodology Data were retrospectively retrieved from inpatient medical records of patients admitted with CPA tumors from January 1, 2011, to December 31, 2020, at our institute. Epidemiological, clinical and radiological findings, histopathological types, surgical outcomes, and facial nerve function of these patients were recorded using a data-gathering tool. Results Out of 230 patients, 188 (81.7%) were diagnosed histopathologically with vestibular schwannoma (VS), 20 (8.7%) with meningioma, 15 (6.5%) with epidermoid, and 7 (3.1%) with other conditions. The most common clinical features were hearing loss in VS and headaches in meningioma and epidermoid. Preoperatively, 103 (44.8%) had grade 2, 68 (29.6%) had grades 3 or 4, and 8 (3.5%) had grade 5 facial nerve palsy, while post-operatively, 93 (40.9%) patients had grade 2, 83 (36.6%) had grades 3 or 4, and 6 (2.6%) had grade 5 facial palsy. Greater facial nerve preservation was observed in patients with tumor sizes <4 cm (p=0.0041) and in those who underwent near-total (NTR) or subtotal resection (STR) (p=0.0442). Excellent facial nerve outcomes (HB grades 1 or 2) were noted in patients who underwent intraoperative facial nerve monitoring (p<0.0001). CSF leak and meningitis were present in 3.5% and 2.2% of patients, respectively. The mortality rate was 6.1%, with a recurrence rate of 4.8%. Conclusion Intraoperative facial nerve monitoring, tumor size less than 4 cm, and extent of resection (NTR/STR) are predictive factors that significantly affect facial nerve outcomes.

The most frequently witnessed symptoms of CPA tumors include hearing loss, headache, dizziness, ataxia, tinnitus, vertigo, and dysfunction of the cranial nerves [4].Surgery is considered the optimal treatment for CPA tumors after proper preoperative planning, except for small tumors (<2.5 cm), which can be managed by radiosurgery [5].
Surgical outcomes and factors predicting facial nerve morbidity and severity have rarely been specifically studied in North India, especially in our region.Preoperative facial nerve status, tumor size, and extent of resection majorly affect facial nerve morbidity in the immediate postoperative period.To the best of our knowledge, only a few studies [6][7][8] in North India have correlated these factors with facial nerve outcomes.Our center, being a tertiary care center, caters to a large population of Uttar Pradesh, including parts of Uttarakhand, Bihar, and Nepal.Therefore, this study aims to analyze the epidemiological and clinicoradiological characteristics, extent of resection, surgical outcomes, and predictive factors for facial nerve preservation in CPA tumors.

Materials And Methods
A retrospective observational study was conducted in the Department of Neurosurgery at King George's Medical University, Lucknow, between January 1, 2011, and December 31, 2020.The King George's Medical University, Uttar Pradesh (UP), Institutional Ethics Committee reviewed the research article in its meeting held on April 2, 2022, and granted approval with number VII-PGTSC-IIA/P4.The study examined the inpatient medical records of 345 patients.Cases with incomplete records and missing data were excluded.Two hundred thirty patients underwent final surgical intervention for their CPA tumors, and 115 patients were lost after CSF diversion.Ultimately, 230 patients were included in the study.The collected data were analyzed based on epidemiological, clinical, radiological, surgical outcomes, histopathological types, facial nerve dysfunction, and perioperative complications.Patients recruited for the study were followed up, and their clinical data, including facial nerve function, were recorded.Nineteen of these patients underwent intraoperative facial nerve monitoring.Thus, patients were also divided based on whether they were monitored intraoperatively or did not undergo intraoperative monitoring.Data were also evaluated for these two groups.The clinical information of the study subjects, including facial nerve function assessed by House and Brackmann grading (HB grading), was evaluated.These individuals underwent brain contrast-enhanced computed tomography (CECT) and MRI with contrast enhancement (CEMRI) in follow-up.The Chi-square test was used to analyze categorical variables, and a student t-test was used to compute a continuous variable.To assess statistical significance at the 5% level of significance, probability (p) was determined.

Results
Out of 230 patients, the predominant histopathology was VS, accounting for 188 (81.7%) patients, followed by 20 (8.7%) with meningioma, 15 (6.5%) with epidermoid, and 7 (3.1%) with other conditions (Figure 1).A majority, 94.8%, were sporadic cases, with only 5.2% being NF-2 associated.The mean age of the patients was 36.3 ± 14.2 years.Most patients with VS and epidermoid were males, while females predominated in meningiomas (Table 1).VS patients were more common in the second to third decades of life, followed by the third to fourth decades.Meningiomas and epidermoids were more commonly found in the third to fourth decades of life (Table 1).
One hundred fourteen patients were followed up with a median follow-up period of 6 years.Facial nerve function improved in 32 patients (28.1%), remained the same in 77 (67.6%), and deteriorated in 5 (4.4%) during the follow-up.
Epidermoids were 100% hypointense on T1-weighted and hyperintense on T2-weighted.Post-contrast, all patients with VS showed heterogeneous contrast enhancement, all meningiomas were broad-based and had homogenous contrast enhancement, while epidermoids did not enhance and showed restriction on diffusion-weighted imaging (DWI) images (Table 2).One hundred fifty-four (66.9%) patients with VS had internal acoustic meatus (IAM) widening, while all patients with meningiomas had broad-based tumors with dural enhancement.

Complications and outcomes
Postoperatively, wound infection was observed in 11 (4.8%) patients.Of these, a CSF leak was present in 8 (3.5%) patients with tumor size > 4 cm.Five patients were managed conservatively with compression bandages and medical therapy (Acetazolamide), while three required lumbar drain placement.Meningitis developed in 5 (2.2%) patients, of which 3 died from it.Postoperative hematoma developed in 5 (2.2%) patients, four of whom required re-exploration.Twenty-six (11.5%) patients developed post-operative lower cranial nerve palsy; of these, 10 were tracheostomized, and in 3 patients, the palsy improved (Table 3).Fourteen patients died during this study (a mortality rate of 6.1%).Of these, 10 died during admission and 4 during follow-up.Three patients died in the immediate postoperative period due to cardiorespiratory arrest, three from meningitis, three from respiratory failure, and one from hydrocephalus.Four patients died during follow-up due to comorbidities.Eleven patients (4.8%) experienced a recurrence in our study; four were reoperated, three were sent for gamma knife surgery, and four were followed up with no clinical deterioration (Table 3).

Tumor size, resection, and facial nerve preservation
The mean tumor size of CPA tumors in this study was 2.7 cm, with a range of 1.2-7.5 cm.It was observed that the majority of tumors were either large (2.6-4 cm) or giant (>4 cm), accounting for 118 (51.3%) and 107 (46.5%) patients, respectively.Facial nerve preservation was greater in patients with tumors smaller than 4 cm, which was statistically significant (p-value = 0.0041) (Table 4).All patients with CPA tumors underwent surgery via the retromastoid suboccipital approach, except for one patient with VS, who was operated through the translabyrinthine approach.Two hundred fifteen (93.5%) patients underwent gross total resection (GTR) and near-total resection (NTR), while 15 (6.5%) patients underwent subtotal resection (STR).

Complications
No statistically significant correlation between the type of resection and tumor size was noticed (p-value = 0.251).However, a statistically significant association of facial nerve preservation with the type of resection was observed (p-value = 0.0442) (Table 4).
It was observed that preoperatively, the majority of patients who underwent facial nerve monitoring had excellent facial nerve status (HB grade 1 or 2), with 16 out of 19 patients.Postoperatively, the facial nerve grade deteriorated in only one patient from an excellent grade (HB grade 1 or 2) to an intermediate grade (HB grade 3 or 4).
Nineteen patients with CPA tumors underwent facial nerve monitoring, of which 18 (94.7%)achieved facial nerve preservation in the postoperative period, compared to 165 (79.3%) patients who were not monitored (p-value = 0.104).It was observed during the study that excellent facial nerve outcomes (HB grade 1 or 2) were higher in patients who underwent intraoperative facial nerve monitoring compared to those who did not (78.9%vs 59.1%), which was statistically significant (p-value < 0.0001).Intermediate (HB grade 3 or 4) and poor (HB grade 5 or 6) facial nerve outcomes were higher in non-monitored patients compared to monitored patients (37.9% and 21.9% versus 2.9% and 0%) (

Discussion
There is pathological diversity in CPA tumors as they originate from various sources and abut different adjacent nerves and vascular structures.As previously discussed, CPA tumors can be categorized into two types: vestibular schwannomas (VS) and non-vestibular tumors (meningioma, epidermoid, other cranial nerve schwannoma, and cholesteatoma) [2].Of all CPA tumors, VS accounts for 80-94%, while the remainder are non-vestibular tumors [1][2][3].
In this study, VS accounted for 81.7%, while the non-vestibular tumors included 8.7% meningioma, 6.5% epidermoid, and 3.1% others.Hearing loss and headache were the most common clinical features, with hearing loss observed in 100% of VS patients, 60% of meningioma patients, and 40% of epidermoid patients.Headache was reported in 90.9% of VS patients, 90% of meningioma patients, and 53.3% of epidermoid patients.Facial nerve involvement was also a prominent clinical feature, observed in 87.8% of VS patients, 40% of meningioma patients, and 26.7% of epidermoid patients.The majority of tumors in our study were either large (2.6-4 cm) or giant (> 4 cm), reflecting late presentation at the outpatient department.We achieved GTR and NTR in 93.5% of patients.CSF diversion was performed in 163 (70.86%) patients.There was a significant correlation between tumor size and facial nerve outcome, with a p-value of 0.0041.Our study found that tumor size and extent of resection were major predictive factors for facial nerve preservation (p-values of 0.0041 and 0.0442, respectively).Monitoring significantly aided facial nerve outcomes, with no poor outcomes in those monitored intra-operatively (p-value = 0.0143).CSF leaks were more common in patients with tumors larger than 4 cm.The mortality rate was 6.1% in this study, with 10 out of 14 patients who died during admission due to meningitis, respiratory failure, and hydrocephalus.
Similar histopathological distributions have been observed in other studies [2,6,[9][10].Consistent with findings by Kennedy GRS et al. [2], non-vestibular tumors were more likely to occur in the third to fourth decade of life.Unlike other studies, our study had slightly more males than females, with only meningioma showing a female predominance [2,[11][12][13][14][15][16].Kennedy GS et al., in their prospective study of 50 patients with CPA tumors, reported hearing loss as the most prominent clinical feature in 89% of cases, followed by cerebellar ataxia in 81% and headache in 74% of patients.Facial nerve and lower cranial nerve dysfunction were present in 65% and 20% of patients, respectively, in their study [2].In comparison, Memari F et al. found that hearing loss was the most common presenting symptom in 85% of patients with CPA tumors, while 50% of patients experienced headaches and 53% had cerebellar ataxia.They reported lower percentages of facial nerve involvement (53%) and lower cranial nerve involvement (6%) in their study [15].Similar lower incidences of facial nerve involvement in CPA tumors have been observed in other studies as well [17,18].This difference could be attributed to the inclusion of non-vestibular tumors in these studies, whereas our study primarily involved VS patients.Similar to our study, Kennedy GS et al.'s study found that most CPA tumors (96%) were either large (2.6-4 cm) or giant (> 4 cm) [2].Dhar S et al., in their study of 126 CPA tumors, reported an average tumor size of 3.53 cm [6].
The key point to focus on while operating on a CPA tumor is the complete removal of the tumor lesion while simultaneously preserving the functionality of the cranial nerves present in the surrounding area.It has been observed that up to 80-90% of tumor resections are possible, depending on the surgeon's level of experience [7].Due to the tumor's intimate relationship with important structures of the CPA, including cranial nerves, the brainstem, and vessels, complete tumor removal is challenging without compromising these structures, especially the cranial nerves.In our study, we achieved GTR and NTR in 93.5% of patients, which was higher compared to the study by Nayak P et al. [7].Using the retromastoid approach, Samii M and Matthies C completely removed 979 tumors; anatomical preservation of the facial nerve was achieved in 93% of patients, and of the cochlear nerve in 68% [11].CSF diversion (VP shunt/ETV) was performed in 163 (70.86%) patients in the current study, much higher than the rates reported by Ramamurthi B [19], Kennedy GS et al. [2], and Jain VK et al. [10], which were 66%, 36%, and 8.5%, respectively.
Similar to the facial nerve preservation rate in our study, which was 80.6%, Samii M and Matthies C [11], Jain VK et al. [10], and Kennedy GS et al. [2] reported facial nerve preservation rates of 93%, 84.3%, and 67%, respectively, in their studies.Preoperatively, 44.8% of our patients had grade 2, 29.6% had grade 3 or 4, and 3.4% had grade 5 facial nerve palsy, while postoperatively, 40.9% had grade 2, 36.6% had grade 3 or 4, and 2.6% had grade 5 facial palsy, contrasting with the studies by Memari F et al. [15] and Joarder MA et al. [12].Dhar S et al., in their study of 126 CPA tumor patients, found that the majority (71.4%) had grade 3 facial palsy preoperatively.Out of these, 42 patients experienced poor HB-grade facial nerve dysfunction immediately postoperatively, while the rest maintained similar facial nerve status [6].In another study by Nair S et al., facial nerve function was intact in 78% of cases before surgery, but at the 6-month follow-up, 90% of cases had intermediate to poor preservation of the facial nerve.Among those with anatomically preserved facial nerves, more than half had HB grade 4 facial palsy [20].
A significant correlation was observed between the size of the tumors and the outcome of the facial nerve in this study, with poorer outcomes in patients who had tumors larger than 4 cm (88.5% vs. 71.4%,p-value = 0.0041), which was higher compared to the studies by Kennedy GS et al. (68% vs. 58%) [2] and Joarder MA et al. (74% vs. 62%) [12].In the study by Jain VK et al., large tumors had a lower percentage of facial nerve preservation; among those with tumors larger than 4 cm, the facial nerve could be preserved in only 34.6% of patients [10].Facial nerve preservation was higher in patients who underwent NTE and STE of tumors compared to GTE, which was statistically significant (p-value = 0.0442) in our study, consistent with previous studies [21][22][23].Samii M and Matthies C reported tumor size, previous surgery or radiosurgery, and a surgeon's operative experience as major predictive factors for facial nerve preservation [11].
CSF leaks and meningitis developed in 3.5% and 2.2% of patients in our study, respectively, far less than in the study by Kennedy GS et al. [2].Our study's mortality rates were high compared to other studies [15,18,[24][25][26].Late presentation after being treated at various clinics and the large size of tumors with hydrocephalus were major factors responsible for higher mortality in our study.The limitation of this study was that intra-operative facial nerve monitoring was not performed on all patients as indicated due to a lack of the required expertise and equipment in the initial years.

FIGURE 1 :
FIGURE 1: Distribution of tumors according to the histological types.A: Vestibular schwannoma B: Meningioma C: Epidermoid D: Others

TABLE 1 : Baseline characteristics of patients based on histopathological types of CPA tumors.
CPA: Cerebellopontine angle; ICP: Intracranial pressure.

TABLE 2 : Radiological MRI imaging of histological types of CPA tumors.
CPA: Cerebellopontine angle.

TABLE 3 : Distribution of study participants based on the type of resection, tumor size, and facial nerve preservation.
*Three patients expired on post-operative day 0. **Extent of resection is defined based on post-operative CECT/CEMRI brain.**Gross total resection (GTR) is defined as no residual tumor on imaging.**Near total resection (NTR) is defined as a tumor remnant less than 2.5 cm².**Subtotal resection (STR) is defined as any tumor larger than that defined for NTR.**A p-value < 0.05 is considered significant.2024 Verma et al.Cureus 16(6): e61756.DOI 10.7759/cureus.61756