Our Experience With Facial Nerve Monitoring in Vestibular Schwannoma Surgery Under Partial Neuromuscular Blockade

Acta Otorrinolaringol Esp. 2015;66(4):192---198

www.elsevier.es/otorrino

ORIGINAL ARTICLE

Our Experience With Facial Nerve Monitoring in Vestibular Schwannoma Surgery Under Partial Neuromuscular Blockade夽 Jorge Vega-Céliz,a,∗ Emili Amilibia-Cabeza,b José Prades-Martí,b Nuria Miró-Castillo,b Marta Pérez-Grau,b Teresa Pintanel Rius,c Francesc Roca-Ribas Serdàb a

Servicio de Otorrinolaringología, Hospital Universitari de Bellvitge, L’Hospitalet de Llobregat, Barcelona, Spain Servicio de Otorrinolaringología, Hospital Universitari Germans Trias i Pujol, Badalona, Barcelona, Spain c Servicio de Anestesiología y Reanimación, Hospital Universitari Germans Trias i Pujol, Badalona, Barcelona, Spain b

Received 3 June 2014; accepted 8 August 2014

KEYWORDS Vestibular schwannoma; Facial nerve; Intraoperative monitoring; Neuromuscular block

Abstract Introduction and objectives: Facial nerve monitoring is fundamental in the preservation of the facial nerve in vestibular schwannoma surgery. Our objective was to analyse the usefulness of facial nerve monitoring under partial neuromuscular blockade. Methods: This was a retrospective analysis of 69 patients operated in a tertiary hospital. Results: We monitored 100% of the cases. In 75% of the cases, we could measure an electromyographic response after tumour resection. In 17 cases, there was an absence of electromyographic response. Fifteen of them had an anatomic lesion with loss of continuity of the facial nerve and, in 2 cases, there was a lesion with preservation of the nerve. Preoperative facial palsy (29% 7%; P=.0349), large tumour size (88 vs 38%; P=.0276), and a non-functional audition (88 vs 51%; P=.0276) were significantly related with an absence of electromyographic response. Conclusions: Facial nerve monitoring under neuromuscular blockade is possible and safe in patients without previous facial palsy. If the patient had an electromyographic response after tumour excision, they developed better facial function in the postoperative period and after a year of follow up. © 2014 Elsevier Espa˜ na, S.L.U. and Sociedad Espa˜ nola de Otorrinolaringología y Patología Cérvico-Facial. All rights reserved.

夽 Please cite this article as: Vega-Céliz J, Amilibia-Cabeza E, Prades-Martí J, Miró-Castillo N, Pérez-Grau M, Pintanel Rius T, et al. Nuestra experiencia con la monitorización del nervio facial en cirugía del schwannoma del vestibular bajo bloqueo neuromuscular parcial. Acta Otorrinolaringol Esp. 2015;66:192---198. ∗ Corresponding author. E-mail address: [email protected] (J. Vega-Céliz).

2173-5735/© 2014 Elsevier Espa˜ na, S.L.U. and Sociedad Espa˜ nola de Otorrinolaringología y Patología Cérvico-Facial. All rights reserved.

Monitoring and Facial Vestibular Schwannoma Surgery Under Neuromuscular Blockade

PALABRAS CLAVE Schwannoma vestibular; Nervio facial; Monitorización intraoperatoria; Bloqueo neuromuscular

193

Nuestra experiencia con la monitorización del nervio facial en cirugía del schwannoma del vestibular bajo bloqueo neuromuscular parcial Resumen Introducción y objetivos: La monitorización del nervio facial es fundamental en la preservación del nervio facial en la cirugía del schwannoma del nervio vestibular. Nuestro objetivo es analizar la utilidad de la monitorización facial bajo bloqueo neuromuscular parcial. Métodos: Análisis retrospectivo de 69 pacientes operados en un hospital de nivel terciario. Resultados: En el 100% de los casos se pudo realizar monitorización electromiográfica. Se obtuvo respuesta electromiográfica tras la escisión tumoral en el 75% de los casos. En 17 casos no hubo respuesta electromiográfica. En 15 hubo lesión de continuidad del nervio facial y en 2 hubo lesión pero se logró mantener la continuidad anatómica del nervio. La presencia de parálisis facial preoperatoria (29 vs. 7%; p = 0,0349), el mayor tama˜ no tumoral (88 vs. 38%; p = 0,0276) y una audición no funcional (88 vs. 51%; p = 0,0276) son factores significativamente más prevalentes en el grupo en que no se pudo obtener una respuesta electromiográfica. Conclusiones: La monitorización del nervio facial bajo bloqueo neuromuscular moderado es factible y segura en pacientes sin lesión facial preoperatoria. Los pacientes que obtienen respuesta electromiográfica tras la extirpación del tumour tienen mejor función facial en el postoperatorio y al a˜ no de seguimiento. na, S.L.U. and Sociedad Espa˜ nola de Otorrinolaringología y Patología © 2014 Elsevier Espa˜ Cérvico-Facial. Todos los derechos reservados.

Introduction From the introduction of electromyographic (EMG) monitoring of the facial nerve by Delgado1 in 1979, the percentage of severe postoperative facial dysfunction has fallen from some 15%---59% in the era before monitoring to approximately 10%---33% with the use of facial nerve monitoring. Facial nerve monitoring has been established as an essential part of cranial base surgery. However, it has been impossible to standardise clinical application of EMG recording.2 There is a general acceptance of EMG criteria that permit predicting facial function; however, the lack standards for electrode mounting and stimulation parameters has prevented establishing which method is the best.3 To perform EMG monitoring of the facial nerve, avoiding the use of neuromuscular blockade (NMB) has been recommended. However, it is considered admissible to use intraoperative blockade drugs if their dosage is controlled by adequately monitoring the degree of peripheral NMB.4 Among the advantages involved with the use of partial NMB are: (1) facilitating surgical exposure, (2) eliminating the need for the surgeon to interrupt the procedure to monitor the muscle potentials evoked, (3) reducing the risk of unexpected movements (especially in patients with tolerance to opioid anaesthetics) and (4) reducing the excessive noise in the EMG register, improving the signal/noise ratio by reducing the time required to acquire the signals.5 The present study investigated the results of EMG facial nerve monitoring in patients operated on for vestibular schwannoma (VS) under partial NMB.

Methods This was a retrospective analysis of the database of patients with VS in follow-up by the Ear, Nose and Throat Service at

our hospital. Patients who had surgery under partial NMB and with a register of intraoperative facial EMG monitoring were studied. Tumour size was evaluated using the classification of Tos and Thomsen,6 facial function was recorded using the House-Brackmann (HB) scale7 and hearing was classified according to the criteria of the American Academy of Otolaryngology-Head and Neck Surgery (AAO-HNS).8 The preoperative clinical factors and the intraoperative variables of interest were analysed: type of approach, tumour resection, cranial nerve VII lesion, intraoperative complications and operating time. These variables were compared based on type of EMG response and evolution of facial function in the immediate postoperative period and at 1 year’s follow-up. The intraoperative facial nerve lesion with loss of continuity is defined as the transection of the nerve; it is accompanied by the loss of the EMG register upon stimulating the area of the pontomedullary junction after tumour removal. Facial nerve lesion without loss of nerve continuity is defined as lesion of the nerve without its transection and accompanied by A-train in the EMG register. (Sinusoidaltype EMG waveform that produces a high frequency sound; it has a sudden initiation and presents ranges of maximum amplitude of 100---200 ␮V, never exceeding 500 ␮V.)

General Anaesthesia Protocol In induction, the following was used: propofol: 2---3 mg/kg bolus, rocuronium: 0.6 mg/kg bolus and remifentanil: 0.5 mcg/kg to pass in 3 min. During maintenance, the following was used: rocuronium, in continuous perfusion of 0.15 mg/kg/h (half the dosage required for complete NMB); remifentanil, 0.1---0.3 mcg/kg/min and desflurane 4%---5% (<1 of minimum alveolar concentration). If there was intracranial hypertension, propofol was substituted for

194 desflurane: 5---10 mg/kg/h. Standard evaluation with electrocardiograph, pulse oximetry and non-invasive monitoring of the blood pressure was performed. The depth of anaesthesia was monitored using the bispectral index, keeping it in a range of 40---60. Ventilation used oxygen at 40% to maintain a carbon dioxide concentration at the end of expiration of 35---40 mmHg. Body temperature was kept at 36---37 ◦ C using a hot air system. To evaluate the degree of NMB, the cubital nerve was monitored using the technique of 4-intrain, in which 4 supramaximal stimuli every 0.5 s (2 Hz) were used to provoke muscle contraction. The NMB level during maintenance was 80%---85% with a train-of-4 at 2. This NMB level has been identified as capable of maintaining an EMG response at different NMB levels.9

J. Vega-Céliz et al.

Intracanalicular 9% (6) Small (≤1 cm) 7% (5) Medium (1-2.5 cm) 38% (26) Large (2-5.4 cm) 39% (27) Giant (>4 cm) 7% (5)

Total=69

Figure 1 Classification of Tos and Thomsen. General description of the series.

Facial Monitoring EMG monitoring of 2 intraoperative canales was performed with the NIM-Response 2.0 system (Medtronic Xomed Inc., Jacksonville, FL, USA) in all the patients. Two paired bipolar subdermal electrodes were placed in the orbicularis oculi and in the oris. The impedance differences between electrodes were kept to 1 k during the 2 channel registers. The facial nerve was simulated electrically with a monopolar probe with a 0.5 mm point. The stimulation applied was of squared waves lasting 100 ms with a frequency of 4 Hz. The monitor signalled when there was a response greater than 100 ␮V with a signal filter of 3.10 ms. After tumour resection, the facial nerve was stimulated in the pontomedullary junction proximal to the area where the tumour was adhered to the facial nerve. The minimum stimulus capable of producing an EMG response of 100 ␮V in at least 1 channel was registered. The stimulation protocol began with a stimulus of 0.01 mA and then continued with stimuli with increments of 0.01 mA up to a maximum stimulation of 2.0 mA.

Statistical Analysis The values were expressed as average±mean standard deviation (SD) error or in percentages. The software GraphPad® , version Prism 6 (GraphPad Software, Inc., La Jolla, CA, USA) for MAC® was used. The differences between the various groups were studied with Student’s t-test or the test of Mann---Whitney U to compare continuous variables, while the 2 test or Fisher’s exact test were used to compare categorical variables. The area under the curve of the receiver operating characteristic (ROC) was analysed to determine the diagnostic performance of the EMG response in predicting facial function. The following intervals were used: 0.5---0.6: poor test; 0.6---0.75: regular test; 0.75---0.9: good test; 0.9---0.97: very good test; 0.97---1: excellent test. Statistical tests were considered significant when P<.05 was obtained after correcting for multiple comparisons.

Results The general description of the series is as follows: 69 patients were identified that fulfilled the inclusion criteria (patients with VS, intervention under NMB, registry of auditory function and of preoperative facial function, in the immediate postoperative period and at 1 year’s follow-up)

No lesion 71% (49) VII lesion with anatomical preservation 7% (5) VII lesion without anatomical preservation 22% (15)

Figure 2 Surgical result for the facial nerve. General description of the series.

in the period from April 1999 to February 2013. The mean age was 48 years (range, 12---77 years). There were 28 men (40%) and 41 women (60%). From the patients, 2 had type 2 neurofibromatosis, 2 had received prior treatment: 1 with surgery and radiotherapy and the other, only radiotherapy. At the time of indication of surgery, the patients presented deafness in 61 cases (88%). Of these cases, 7 were of sudden type (10%), with tinnitus in 46 cases (67%), instability/vertigo in 29 cases (42%), facial paralysis in 9 cases (13%), 2 cases with facial hemispasm and with affectation of the cranial nerve V (hyperalgesia and facial or corneal hypoesthesia) in 10 cases (14%). According to Tos and Thomsen’s classification, the majority of the tumours in our series were medium or large (76%) (Fig. 1). Surgery was on 35 right sides (51%) and 34 left sides (49%). The most frequent approach used was the translabyrinthine with 65 cases (94%), followed by 3 cases in middle fossa (5%) and 1 case with transcochlear approach (1%). Total tumour resection was achieved in 60 cases (87%), almost total resection (tumour remains less than 2 mm) in 6 cases (9%) and partial (tumour remains less than 5 mm) in 3 cases (4%). In the majority of the cases there was no lesion to the facial nerve (71%) and in case of lesion to the cranial nerve VII, in 25% of the cases the nerve was preserved anatomically (Fig. 2). Other intraoperative complications were 1 arterial lesion, 1 venous lesion, 1 lesion of the cranial nerve V and 1 lesion of low cranial nerves. In the series, mortality was 0%. In respect to management of facial paralysis in our series, facial reconstruction surgery was performed in 15 cases (22%): 2 intraoperatively (1 case with end-to-end anastomosis of the VII par and 1 case of anastomosis that needed greater auricular nerve graft).

Monitoring and Facial Vestibular Schwannoma Surgery Under Neuromuscular Blockade

Percentage

1.0

50% 0.5 22% 13%

11%

2%

2%

H

B

VI

V B H

IV B H

H

B

III

II B H

H

B

I

0.0

Figure 3 Function of the facial nerve in the immediate postoperative period according to the House-Brackmann (HB) scale. General description of the series.

Percentage

1.0

61% 0.5

18%

14%

5%

1%

1%

VI H

B

V B H

IV H

B

III H B

B H

H

B

I

II

0.0

Figure 4 Facial nerve function at 1 year’s follow-up according to the House-Brackmann (HB) scale. General description of the series.

A total of 13 cases required surgery in the postoperative period, carrying out hypoglossal-facial anastomosis. In addition, a palpebral weight was placed in 6 of these patients. Mean hospital stay was 13 days (range, 7---40). As far as facial function at follow-up was concerned, in the immediate postoperative period 61% of the patients presented good facial function (HB I---II), while this percentage increased to 79% at 1 year’s follow-up (Figs. 3 and 4).

Analysis of the Usefulness of Facial Monitoring The protocol for EMG facial monitoring was carried out and was used as a guide in identifying the cranial nerve VII in 100% of the cases. When the tumour was excised in 20 patients, there was a lesion of the cranial nerve VII. The absence of anatomical continuity of the facial nerve was confirmed in 15 patients. The anatomical location of the VII lesion was, in 13 cases, at the proximal level, near its exit from the brainstem, without being able to identify the proximal end; this prevented intraoperative repair. In 1 case, the lesion was at the intracisternal level, requiring an end-to-end anastomosis, while in another case, it was at meatus level with the need for anastomosis with auricular nerve graft. In the 15 patients with VII lesion without preservation of the anatomical continuity of the facial nerve, no EMG response was obtained to incremental stimuli in the area of the pontomedullary junction. However, it was possible to obtain EMG responses upon stimulating the facial nerve at distal

195

level from the tumour. In 5 cases, there was facial nerve lesion without loss of anatomical continuity, associated with the registry of A-trains. Of these, in 2 cases it was impossible to obtain EMG responses in the area of the pontomedullary junction and at distal level to the tumour. EMG response after tumour removal was performed stimulating at the level of the pontomedullary junction in 52 patients (75%). The minimum stimulus capable of producing an EMG response of 100 ␮V in at least 1 channel presented a normal distribution when using the D’Agostino-Pearson omnibus test (P<.0001). The mean minimum stimulation threshold was 0.1648 mA; SD±0.1428 mA; range, 0.02---1 mA. A univariate analysis comparing the preoperative, intraoperative and postoperative characteristics between the patients that had a lack of EMG response and those with response to the incremental stimuli at the level of the pontomedullary junction was carried out. From this analysis, the following were identified as independent factors associated with the absence of EMG response: the presence of preoperative facial paralysis, patients with large or giant tumours, and patients with non-functional hearing (AAO-HNS C and D). Table 1 presents a summary of the variables significantly associated with the absence of EMG response after tumour removal. Identifying the diagnostic performance of the minimum stimulus capable of producing an EMG response of 100 ␮V in at least 1 canal was attempted. The area under the curve of the patients with good facial nerve function (HB I---II) was compared with that of the patients with poor facial function (HB>III), using ROC curves. The diagnostic value of the test in the immediate postoperative period was regular (area under the curve=0.6851; confidence interval [CI] 95%: 0.5048---0.8655; P=.06142) and was still regular at 1 year’s follow-up (area under the curve=0.6973; CI 95%: 0.3595---1.035; P=.2551). Three thresholds were compared with Fisher’s exact test: 0.05 mA, 0.1 mA and 0.2 mA in the diagnostic capacity of good facial nerve function (HB I---II) in the immediate postoperative period and at 1 year of followup. However, no cut-off value could be identified that made it possible to differentiate good facial function (0.05 mA in the immediate postoperative period, P=.5669; 0.05 mA at 1 year of follow-up, P=1.000; 0.1 mA in the immediate postoperative period, P=.1693; 0.1 mA at 1 year of followup, P=.567; 0.2 mA in the immediate postoperative period, P=.1006; 0.2 mA at 1 year of follow-up, P=.3131).

Analysis of Subgroups Based on Facial Function Results A univariate analysis was performed on the clinical, intraoperative and follow-up variables between patients with good facial nerve function (defined as HB I---IIresult) and those with poor facial function (HB>III) in the immediate postoperative period and at 1 year of follow-up. Analysing facial function in the immediate postoperative period, we found that 41 patients presented good facial nerve function and 28 presented poor facial function. Neither age nor sex was a variable that was different between these groups. Greater tumour size was associated with poor facial function. The fact that large and giant tumours were associated with worse facial

196

J. Vega-Céliz et al.

Table 1 Cranial Nerve VII EMG Monitoring. Comparison Between Patients With Loss of Electromyographic Response After Tumour Removal (nr) and Patients Maintaining the Electromyographic Response (EMG-r). Variables

nr % (n) n=17

EMG-r % (n) n=52

P

Pre-S facial paralysis Large+giant size Non-functional AAO-HNS C+D Surgery time >6 h Intraoperative complications Facial reconstruction PO complications CSF fistula HB III---VI PO HB III--VI 1 year

29 (5) 88 (15) 88 (15) 70 (12) 23 (4) 82 (14) 58 (10) 41 (7) 100 (17) 76 (13)

7 (4) 38 (20) 51 (27) 34 (18) 0 (0) 1 (1) 13 (7) 9 (5) 21 (11) 6 (3)

.0349 .0276 .0095 .0123 .0028 <.0001 .0005 .0120 <.0001 <.0001

AAO-HNS: American Academy of Otolaryngology-Head and Neck Surgery; CSF: cerebrospinal fluid; EMG-r: EMG response after tumour removal; HB: House-Brackmann; S: surgical intervention; nr: no EMG response after tumour excision; PO: postoperative.

Table 2 Analysis of Intraoperative Variables Based on Facial Function in the Immediate Postoperative Period. Facial Result in the Immediate Postoperative Period Comparing Patients With Good Facial Function (HB I---II: House-Brackmann I---II) and Poor Facial Function (HB III---VI: House-Brackmann III---VI). HB I---II % (n) or X±SEM n=41

HB III---VI % (n) or X±SEM n=28

P

Surgical approach Transcochlear Fossa media Translabyrinthine Surgery time >6 h

0 3 38 66 (11)

1 0 27 77 (18)

Ns Ns Ns .0029

Tumour resection Complete Almost complete (2%) Partial Intra-surgery complications CPA arterial lesion CPA venous lesion V lesion V VII lesion with anatomical preservation VII lesion without anatomical preservation Low cranial nerve lesion Nr monitoring EMG-r

25 (38) 3 (2) 35 (1) 1 0 0 0 1 0 0 0 (0) 0.1407±0.01

66 (22) 44 (4) 77 (2) 20 1 1 1 4 15 1 60 (17) 0.2545±0.07

Ns Ns Ns <.0001 Ns Ns Ns Ns <.0001 Ns <.0001 .1782

Variables

CPA: cerebellopontine angle; EMG-r: EMG response after tumour excision; HB: House-Brackmann; Nr: no EMG response after tumour excision; Ns: not significant; PO: postoperative; X±SEM: mean±standard error of the mean.

function (P=.0038) was identified (Fig. 5). On analysing preoperative auditory function according to the AAO-HNS classification, State A was associated with good facial function (21% vs 0%; P=.0086), while useless auditory function (C+D) was associated with poor facial nerve function (75% vs 51%; P=.0469). With respect to the analysis of the intraoperative and follow-up variables, the patients that presented poor facial function required longer surgeries (>6 h) in comparison with the patients with good facial function (77% vs 66%; P=.0029). There were no differences in type of approach or type of tumour resection. However, patients with poor facial function presented a greater number of intraoperative complications in

general (P=.0001) and, in particular, lesion of cranial nerve VII without preservation of the anatomical continuity of the facial nerve (P<.0001); this was related to an absence of EMG response. Analysing the patients that presented an EMG response, we identified greater response thresholds in the patients with poor facial function, but this did not reach statistical significance (0.1407±0.01 vs 0.2545±0.07 mA; P=.1782) (Table 2). Analysing facial function at 1 year’s follow-up, large size tumours and affectation of the cranial nerve V were identified as factors that predicted poor facial function. Both prolonged surgery time and development of intraoperative complications were significantly associated with

Monitoring and Facial Vestibular Schwannoma Surgery Under Neuromuscular Blockade

197

Table 3 Variables Associated Significantly With Facial Function at 1 Year’s Follow-up. Facial Result at 1 Year’s Follow-up Comparing Patients With Good Facial Function (HB I II: House-Brackmann I---II) and Poor Facial Function (HB III---VI: House-Brackmann III---VI). Variables

HB I-II % (n) or X±SEM n=53

HB III-VI % (n) or X±SEM n=16

P

V affectation Large size Surgery time >6 h Intra-surgery complications VII lesion with anatomical preservation VII lesion without anatomical preservation Nr monitoring EMG-r

9 (5) 32 (17) 33 (18) 11 (6) 1 (1) 7 (4) 7 (4) 0.1484±0.01

31 (5) 62 (10) 68 (11) 93 (15) 25 (4) 68 (11) 81 (13) 0.4333±0.28

.0446 .0413 .0203 <.0001 .0090 <.0001 <.0001 .4224

EMG-r: EMG response after tumour removal; HB: House-Brackmann; Nr: no EMG response after tumour removal; PO: postoperative; X±SEM: mean±standard error of the mean.

0.0038

No. patients HB I-II po

30

20

0.0493

20 10 10 0

cm

cm (>

4

.4 -5 G

ia

nt

(2 e rg La

m iu M

ed

)

)

cm ) .5 -2

(1

l( al Sm

In

tra

ca

ni

≤1

cu

)c

la

m

r

0

No. patients HB III-IV po

40

HB I-II po

HB III-IV po

Figure 5 Clinical data before the surgery. According to facial function in the immediate postoperative (po) period.

poor facial function. As for the analysis of facial monitoring, the absence of EMG response persisted at 1 year’s follow-up as a factor associated with worse facial function. Analysing the patients with EMG response, we again found greater response thresholds in patients with poor facial function. However, just as in the immediate postoperative period, this difference was not significant (0.1484±0.01 vs 0.4333±0.28 mA; P=.4224). These differences are presented in Table 3.

Discussion Intraoperative monitoring of the facial nerve is considered an essential part of cerebellopontine angle surgery, and especially in VS surgery. Monitoring permits us to locate the facial nerve and evaluate facial function after tumour removal. The effects of NMB on EMG monitoring, both active and passive, of the cranial nerves and the roots of spinal nerves has not been studied appropriately.2---4 Our series is the largest one reported of VS surgeries under NMB. It was possible to perform EMG monitoring of

the facial nerve in 100% of our patients. In our series, the register of intraoperative facial EMG activity was not affected, which relates to various reports in the literature. In 1992, Lennon et al10 were the first to analyse facial EMG response under NMB. Their series consisted of 10 patients treated using VS surgery with retrosigmoid approach and they were able to demonstrate that, under NMB of 50% measured by EMG response of the hypothenar muscles, facial EMG response was not affected. After the study of Lennon came that of Blair et al,11 which demonstrated in 8 patients that it was possible to obtain facial EMG responses even at level of 75% blockade at the level of the 4-train receptor. It also showed that, in patients with a chronic lesion of the facial nerve, it is much more sensitive at NMB level. They concluded that lower levels of NMB or abstaining from its use were necessary if one wished to avoid lesion of the facial nerve in these circumstances. Brauer et al12 studied, in 11 patients with VS surgery, facial EMG response under moderate or deep NMB (70%---100%) measured by EMG response at the level of hypothenar muscles. They showed that it was possible to obtain an EMG response from the facial nerve even under a level of complete NMB. In our series, we have identified that prior facial paralysis, larger size tumours and auditory affectation are related to loss of EMG response and worse facial function. The lack of EMG response in our study is not only associated with poor facial function, it correlates with the need for nerve reconstruction or rehabilitation surgeries of facial function. In the cases in which an EMG response was obtained following tumour excision, we found a tendency (not significant) for the patients with worse facial function to present more intense levels of facial stimulus. This may be due to the system of EMG register. To correct this problem, the sensitivity of the test could be increased adding channels to analyse the EMG response or using lower levels of NMB or abstaining from its use in patients with poor prior facial function, greater tumour size or hearing loss. Upon analysing the diagnostic performance of the EMG response capable of producing a response of 100 ␮V in at least 1 channel, we found that the analysis of ROC curves showed that its usefulness is medium. When attempting to identify a threshold value that would let us predict postoperative facial function (both in the immediate postoperative

198 period as well as at 1 year’s follow-up), it was impossible to establish a prognostic threshold with intensities of 0.05 mA and 0.1 mA, as well as with 0.2 mA. This could be because, under NMB, the technique capable of detecting a minimum threshold is not sufficiently sensitive. An alternative solution to this problem could be the use of subthreshold EMG register following tumour removal. The use of this method in EMG register of the facial nerve is interesting, given that, on the one hand, it obviates the NMB level by stimulating all the motor end-plates capable of responding to the stimulus; but on the other hand, it has the disadvantage of using high intensities that might damage the nerve fibres.

Conclusions Facial monitoring under moderate NMB is feasible and safe in patients without preoperative facial lesions in VS surgery. Tumour size and poor auditory function are other factors that should be taken into account when using NMB in VS surgery. The minimum stimulus capable of producing a response of 100 mV possesses medium diagnostic usefulness; consequently, new protocols of facial nerve monitoring under NMB should be developed. Postoperative facial function is poor in patients in which the EMG response is lost after tumour removal in VS surgery under partial NMB.

Conflict of Interests The authors have no conflicts of interest to declare.

References 1. Delgado T, Buchheit W, Rosenholtz H, Chrissian S. Intraoperative monitoring of facial nerve muscle evoked responses obtained by intracranial stimulation of the facial nerve: a more accurate technique for facial nerve dissection. Neurosurgery. 1979;4:418---20.

J. Vega-Céliz et al. 2. Arnoldner C, Mick P, Pirouzmand F, Houlden D, Lin VYM, Nedzelski JM, et al. Facial nerve prognostication in vestibular schwannoma surgery: the concept of percent maximum and its predictability. Laryngoscope. 2013;123:2533---8. 3. Acioly MA, Liebsch M, de Aguiar PH, Tatagiba M. Facial nerve monitoring during cerebellopontine angle and skull base tumor surgery: a systematic review from description to current success on function prediction. World Neurosurg. 2013;80:271---300. 4. Ingelmo I, Trapero JG, Puig A, de Blas G, Regidor I, León JM. Intraoperative facial nerve monitoring: anesthetics and neurophysiologic aspects. Rev Esp Anestesiol Reanim. 2003;50: 460---71. 5. Sloan TB. Muscle relaxant use during intraoperative neurophysiologic monitoring. J Clin Monit Comput. 2013;27:35---46. 6. Tos M, Thomsen J. Proposal for reporting size of vestibular schwannoma. In: Kanzaki J, Tos M, Sanna M, Moffat DA, Kunihiro T, Inoue Y, editors. Acoustic Neuroma. Consensus on Systems for Reporting Results Series. Keio University International Symposia for Life Sciencies and Medicine, 10. Tokyo: Springer; 2003. p. 3---7. 7. House JW, Brackmann DE. Facial nerve grading system. Otolaryngol Head Neck Surg. 1985;93:146---7. 8. American Academy of Otolaryngology --- Head and Neck Surgery Foundation, INC. Committee on Hearing and Equilibrium guidelines for the evaluation of hearing preservation in acoustic neuroma (vestibular, schwannoma). Otolaryngol Head Neck Surg. 1995;113:179---80. 9. Choe WJ, Kim JH, Park SY, Kim J. Electromyographic response of facial nerve stimulation under different levels of neuromuscular blockade during middle ear surgery. J Int Med Res. 2013;41:762---70. 10. Lennon RL, Hosking MP, Daube JR, Welna JO. Effect of partial neuromuscular blockade on intraoperative electromyography in patients undergoing resection of acoustic neuromas. Anesth Analg. 1992;75:729---33. 11. Blair EA, Teeple E Jr, Sutherland RM, Shih T, Chen D. Effect of neuromuscular blockade on facial nerve monitoring. Am J Otol. 1994;15:161---7. 12. Brauer M, Knuettgen D, Quester R, Doehn M. Electromyographic facial nerve monitoring during ressection for acoustic neurinoma under moderate to profound levels of peripheral neuromuscular blockade. Eur J Anaesthesiol. 1996;13:612---5.