Facial Nerve Adherence in Vestibular Schwannomas: Classification and Radiological Predictors

Abstract

Background  Facial nerve palsy is one of the most frequent complications after resection of vestibular schwannomas (VS). Several mechanisms have been implicated in the poor postoperative facial nerve outcome. Adherence between the facial nerve and tumor capsule is one of the most relevant factors. There is no clear preoperative parameter permitting identification of these adhesions.

Objective  The aim of this study was to identify the correlation between the grade of adherence of the facial nerve to the tumor capsule and its functional outcome after VS resection.

Methods  A total of 26 patients with sporadic VS (tumor sized T3, T4A, and T4B according to Hannover classification) were evaluated. Grade of adherence of the facial nerve to the tumor capsule was checked during surgery and graduated according to a proposed scale into 1 to 3 different grades. Facial nerve function was assessed postoperatively and after 1-year follow-up. Size of tumor according to Hannover classification, presence of cystic components, “cerebrospinal fluid (CSF) cleft sign,” and the contour of tumor capsule were tested as radiological predictors of grade of adherence to the facial nerve.

Results and Conclusion  Only Grade 2 (11 cases) and 3 (15 cases) of adherence were seen in large VS. Lower grade of adherence was associated with good facial nerve outcome after 1-year follow-up ( p  = 0.029). Presence of the “CSF cleft sign” and regular contour of tumor capsule were independent predictors of adherence. When both factors were associated, sensitivity and specificity of this method were 83 and 80%, respectively.

Introduction

Facial nerve palsy is one of the most frequent complications following vestibular schwannoma (VS) resection and affects, even transiently, up to 53% of the patients. ^{1 2 3} Several mechanisms have been implicated in postoperative facial nerve palsy after VS surgery. Ischemia of the nerve and direct rupture of axonal fibers are the more frequent. ^{4 5}

Grade of facial nerve adherence to the tumor capsule is presumed to direct influence in the postoperative facial nerve function.

In the literature, all series targeting this matter are purely based on the surgeon's perception, lacking objective analysis, and conclusions. ^{6 7}

The objectives of this study were: (1) propose a classification method for facial nerve adherence to the VS capsule, (2) correlate this classification with the functional outcome of the facial nerve, and (3) identify preoperative radiological features associated with higher grades of adherence.

Methods

Population

Between April 2016 to May 2018, 50 patients harboring unilateral sporadic VS staged according to Hannover scale ⁸ as T3 (tumor touching the brainstem), T4A (tumor compressing the brainstem), and T4B (hydrocephalus secondary to compression of the fourth ventricle) were submitted to complete surgical resection at the Neurological Institute of Curitiba (INC), Brazil. Tumors classified as T1 and T2 were excluded from this study. Surgery was performed via retrosigmoid/transmeatal approach, in supine position, by the senior author (R.R.) as previously described. ⁹ Patients were included in the study if they had diagnosis of VS confirmed by pathological examination and the surgical video was available for analysis. Precise intraoperative facial nerve identification at the brainstem and internal auditory canal by electro-stimulation with 2.0 mA and a minimum of 1-year follow-up were requirements of inclusion in this study. Based on this criteria, 26 patients were eligible for inclusion ( Fig. 1 ). The local institutional ethical committee approved this retrospective review, and the signed consent was waived.

Fig. 1.

Flowchart of patient sample and exclusion criteria.

Clinical and Radiological Assessment

All patients had facial nerve function assessed according to the House–Brackmann (HB) scale ¹⁰ preoperatively and postoperatively on day 3 and 1-year after surgery. HB grade I and II were considered as good facial nerve outcome. All patients were preoperatively evaluated with brain MRI. On 3D, balanced steady-state gradient echo sequence the presence of cystic component and “cerebrospinal fluid (CSF) cleft sign” on the interface tumor/brainstem were evaluated. The tumor was considered a cystic VS if it showed presence of hyperintense areas on T2 weighted images and cystic elements were intraoperatively identified. The “CSF cleft sign” was considered present when evidenced on the interface between the tumor and the brainstem around the emergence of the facial nerve at the pontomedullary junction ( Fig. 2 ). Tumor capsule contour was considered regular if no lobulations were seen on its interface with the brainstem.

Fig. 2.

Three-dimensional balanced steady-state gradient echo sequence demonstrating the presence of the “CSF cleft sign” between the tumor and the brainstem. ( A ) The sign is evidenced in a patient harboring a left Hannover T3 Vestibular schwannoma (black arrow). ( B ) Small Hannover T3 schwannoma with regular capsule demonstrates the absence of the sign (white arrow) suggesting that its presence is not related with the tumor size. ( C ) “CSF cleft sign” is demonstrated in a large left Hannover T4A vestibular schwannoma (black arrow). ( D ) Sign is absent in a similar T4A patient (white arrow). CSF, cerebrospinal fluid.

Surgical Videos Analysis

The surgical videos were blindly analyzed by an experienced skull base surgeon (R.R.) and adherence of the tumor capsule to the facial nerve was classified into three different types. Type 1 adherence: displacement of the facial nerve by the tumor, but the arachnoid membrane between the tumor and the facial nerve was not infiltrated. Type 2 adherence: the tumor displaces the facial nerve and infiltrates the arachnoid layer but pia mater and the perineurium of the facial nerve were not infiltrated; dissection plane is precisely identified ( Fig. 3A ). Type 3 adherence: the tumor displaces the facial nerve and infiltrates the arachnoid layer, and the dissection plane is not visible by the infiltration of the pia mater and the perineurium of the facial nerve ( Fig. 3B ).

Fig. 3.

( A ) Left T4A vestibular schwannoma. A partially removed tumor is being pulled off and no arachnoid layer is seen. The pia mater of the facial nerve is preserved, and the dissection plane is clearly visible (white arrow). ( B ) Left T4B vestibular schwannoma. A bimanual dissection technique is demonstrated. The tumor capsule (vestibular schwannomas) is being pulled off with forceps and the facial nerve is visualized with no arachnoid layer interposed between the tumor and the facial nerve. The pia mater of the facial nerve is infiltrated, distorting the anatomy of the facial nerve. No clear dissection plane is visible (white arrow).

Statistical Analysis

Fisher's exact test was used to compare the facial nerve function in different groups. The occurrence of the “CSF cleft sign” and the features of the tumor capsule contour among the different grades of adherence were compared using Chi-square test. A value of p <0.05 was considered statistically significant.

Results

Patients

Demographic and clinical data are shown in the Table 1 . Total 14 male and 12 female patients with mean age of 47 ± 14 years (mean ± standard deviation) were studied. Facial nerve was anatomically preserved in all cases. Seven patients (26.9%) presented tumor size T3, 9 (34.6%) T4A, and 10 (38.4%) T4B. Preoperatively, all patients had normal facial function and 16 patients developed facial palsy (HB ≥ III) on day 3 after surgery. Total 55% of the patients with HB III and 50% HB IV improved to HB ≤ II in 1-year follow-up.

Table 1. Demographic and clinical data.

Patient	Age	Gender	Side	Hannover	Facial nerve 3rd PO	Facial nerve 1-Year	Grade of facial nerve adherence	Cystic	CSF cleft sign	Tumor capsule
1	37	M	L	T4A	4	2	2	No	No	Regular
2	67	M	R	T4A	4	3	3	No	Yes	Regular
3	41	F	R	T3	4	3	2	No	No	Regular
4	24	M	L	T4B	2	2	2	Yes	Yes	Regular
5	40	F	R	T4B	1	5	3	Yes	No	Irregular
6	41	M	R	T3	3	2	2	No	Yes	Irregular
7	59	M	R	T3	3	3	3	No	No	Irregular
8	45	F	R	T4A	1	3	3	Yes	No	Irregular
9	31	M	R	T4B	5	3	3	No	No	Irregular
10	54	M	R	T4A	4	2	3	Yes	No	Irregular
11	49	M	R	T4B	2	1	2	Yes	Yes	Regular
12	18	F	R	T4B	3	4	3	Yes	No	Irregular
13	52	M	L	T4B	1	1	3	No	No	Irregular
14	78	F	R	T4B	5	5	3	Yes	No	Irregular
15	39	F	L	T4B	3	3	3	Yes	Yes	Irregular
16	52	F	L	T4A	2	3	2	Yes	Yes	Regular
17	41	F	L	T4B	1	1	2	Yes	No	Irregular
18	69	M	L	T3	3	2	2	No	Yes	Irregular
19	42	F	L	T3	2	2	3	No	No	Regular
20	67	M	R	T4B	5	3	3	Yes	No	Irregular
21	54	F	R	T4A	3	1	3	Yes	No	Irregular
22	67	F	L	T3	3	2	2	No	Yes	Regular
23	42	F	R	T4A	2	2	2	No	No	Regular
24	50	M	R	T4A	3	3	3	No	No	Irregular
25	35	M	L	T3	3	2	2	No	Yes	Regular
26	52	M	R	T4A	2	3	3	No	No	Irregular

Abbreviations: CSF, cerebrospinal fluid; F, female; L, left; M, male; PO, postoperation; R, right.

Hannover Scale and the Grade of Adherence

Total 15 (57.7%) and 11 (42.3%) patients were respectively classified as Grade 3 and 2 of adherence. No Grade 1 of adherence was seen. Additional review of our VS series demonstrated that Grade 1 of adherence was only seen among patients with tumor size T1 and T2. This finding is beyond the scope of this paper (data are not shown). No statistical differences in grade of adherence were found between tumor sizes T3 and T4 ( p 0.45) or T4A and T4B ( p 0.8).

Grade of Adherence and Facial Nerve Function

Grade of adherence did not influence the early facial nerve function (day 3 postoperation; p 0.5), but strongly influenced the late facial nerve function (1-year follow-up). Nine patients (81%) from the group Grade 2 of adherence demonstrated good late facial nerve function compared with 4 patients (26.6%) on the Grade 3 ( p  = 0.029).

Radiological Characteristic and the Grade of Adherence

Cystic components and the “CSF cleft sign” were observed in 12 patients (46.1%) and 9 patients (34.6%), respectively. Tumor contour was regular in 36.4% of the cases. Presence of cystic component was not associated with the grade of adherence. On the other hand, the presence of “CSF cleft sign” ( p 0.002) and tumor capsule with regular contour ( p 0.004) were associated with lower grade of adherence (Grade 2). A sensitivity of 83% and specificity of 80% in predicting the grade of adherence were found when associating these both characteristics.

Discussion

Degenerative alterations in the facial nerve nucleus and loss of plasticity in the facial motor cortex after long-standing facial nerve injuries causing muscle atrophy can significantly compromise the surgical results of reanimation procedures. ^{11 12}

The use of intraoperative neurophysiological monitoring is an essential tool for preservation of the facial nerve in VS surgery and additionally, based on the stimulation thresholds, may predict facial nerve function outcome. ¹³

Ischemia and direct rupture of axonal fibers are significant mechanisms of facial nerve injury during VS resection. Grade of adherence of the facial nerve to the tumor capsule is an independent factor in facial nerve injury during VS surgery. ⁶

The structural anatomy of the intracranial portion of the facial nerve truly resembles a nerve root of the spinal cord, lacking epineurium but covered with pia mater and arachnoid. ^{14 15} With a hypothetical epiarachnoidal growth of VS, these structures could be affected in different degrees. ¹⁶

In our series, three distinct grades of facial nerve adherence could be identified. Grade 1 was marked by the preservation of the arachnoidal layer between VS and the facial nerve. Grade 2, the arachnoid around the facial nerve was infiltrated, but the dissection plane was easily identified by preservation of the pia mater. Grade 3 was characterized by infiltration of the pia mater and the perineurium of the facial nerve making the identification of the dissection plane very arduous. In our surgical series, Grade I of adherence was exclusive to patients with tumor size T1 and T2 that were beyond of scope. The facial nerve adherence classification proposed in this study remains subjective as it is not possible to demonstrate these degrees of facial nerve involvement with anatomopathological examinations.

In patients with large VS (T3 and T4), this classification was strongly associated with the late but not with the early facial nerve function. The transient edema and neuropraxis that follows facial nerve manipulation is possibly associated with this occurrence. ^{1 2 17 18}

In our series, grade of adherence was not influenced by either tumor size or presence of cystic component, opposing to that presented by Piccirillo et al. ¹⁸ These results, however, must be carefully interpreted since cystic VS were not subclassified in our analysis.

We could not find in the literature any previous report discussing the preoperative predictability of the facial nerve adherence in VS patients. The presence of the “CSF cleft sign,” a known characteristic of extra axial tumors easily identified in T2-sequences, ¹⁹ in association with the regular contour of the tumor capsule, predicted in this study the grade of adherence with a sensitivity of 83%.

Some authors have demonstrated the presence of inflammatory antigen (CD45, CD68, M-CSF, and IL-4) in fast growing VS. ^{20 21} This inflammatory environment created by fast growing tumors could play some role in different types of adhesion between the tumor and the brainstem, possibly accounting for presence or absence of the “CSF cleft sign.”

Conclusion

Grade of adherence of VS to the facial nerve is one of the most important factors in late outcome of facial nerve function after surgical resection. Careful analysis of the “CSF cleft sign” and the contour of the tumor may predict the grade of adherence of the facial nerve to the tumor capsule. These findings could anticipate the difficulties in dissecting and preserving the facial nerve.