- Email: [email protected]
An optimized abnormal muscle response recording method for intraoperative monitoring of hemifacial spasm and its long-term prognostic value
International Journal of Surgery 38 (2017) 67e73
Contents lists available at ScienceDirect
International Journal of Surgery journal homepage: www.journal-surgery.net
Original research
An optimized abnormal muscle response recording method for intraoperative monitoring of hemifacial spasm and its long-term prognostic value Chuyi Huang a, b, 1, Suhua Miao c, 1, Heling Chu d, Chuanfu Dai e, Jinting Wu a, c, Junhua Wang c, Huancong Zuo c, *, Yu Ma c, ** a
Clinical Neuroscience Institute, Yuquan Hospital, Medical Center, Tsinghua University, No. 5 Shijingshan Road, Shijingshan District, Beijing, 100040, China Department of Neurology, Shanghai Jiaotong University Affiliated Sixth People's Hospital, No. 600 Yishan Road, Shanghai, 200030, China Department of Neurosurgery, Tsinghua University Yuquan Hospital, No. 5 Shijingshan Road, Shijingshan District, Beijing, 100049, China d Department of Neurology, Huashan Hospital, Fudan University, No.12 Mid. Wulumuqi Road, Shanghai, 200040, China e Department of Otolaryngology, Head and Neck Surgery, Beijing Tsinghua Changgung Hospital, Tsinghua University, Beijing, 102218, China b c
h i g h l i g h t s Providing an optimized AMR recording method during MVD for HFS. Optimized method improves the positive detection rate of AMR. Optimized method increases accuracy of decompression effect prediction. Increasing the immediate remission rate and reducing the delayed recovery rate.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 12 September 2016 Received in revised form 15 November 2016 Accepted 20 December 2016 Available online 25 December 2016
Background: Intraoperative electrophysiological monitoring is used to determine whether decompression is sufficient during microvascular decompression (MVD) for hemifacial spasm (HFS). However, the real offending vessel is sometimes neglected by the neurosurgeons. Here, we reported our experience in using optimized abnormal muscle response (AMR) monitoring and continuous intraoperative monitoring for MVD. Methods: This study included 2161 HFS patients who underwent MVD using traditional (1023 patients) and optimized (1138 patients) methods. Modified AMR monitoring was adopted in our study, with the zygomatic branch of the facial nerve stimulated and the temporal branch, buccal branch, marginal mandibular branch and cervical branch of the facial nerve detected for AMR. These cases were analyzed retrospectively with emphasis on the postoperative outcomes and intraoperative findings. The therapeutic effect was evaluated at day 1, month 3 and year 1 after operation. Results: The relief rate at day 1, month 3 and year 1 after operation for patients who employed optimized AMR recording method was 95.1%, 97.4% and 99.3%, comparing with 92.2%, 95.0% and 97.8% in traditional method. There was significant difference in achieved immediate remission and recovery rate during 12month follow-up between the two groups (P < 0.05). The modified intraoperative monitoring showed the sensitivity of AMR disappearance to judge the relief at day 1, month 3 and year 1 after HFS operation was 95.7%, 96.3% and 97.3%, respectively; the specificity was 44.6%, 43.3% and 50.0%, respectively; the accuracy was 93.1%, 94.9% and 97.4%, respectively.
Keywords: Abnormal muscle response Hemifacial spasm Microvascular decompression surgery
* Corresponding author. Clinical Neuroscience Institute, Yuquan Hospital, Medical Center, Tsinghua University, No. 5 Shijingshan Road, Shijingshan district, Beijing, 100040, China. ** Corresponding author. Clinical Neuroscience Institute, Yuquan Hospital, Medical Center, Tsinghua University, No. 5 Shijingshan Road, Shijingshan district, Beijing, 100040, China. E-mail addresses: [email protected] (H. Zuo), [email protected] (Y. Ma). 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.ijsu.2016.12.032 1743-9191/© 2016 IJS Publishing Group Ltd. Published by Elsevier Ltd. All rights reserved.
68
C. Huang et al. / International Journal of Surgery 38 (2017) 67e73
Conclusions: Our findings demonstrated that the optimized method could improve the positive detection rate of AMR and accuracy of decompression effect prediction. The evaluation for the decompression effect by optimized intraoperative monitoring can increase the immediate remission rate and reduce the delayed recovery rate. © 2016 IJS Publishing Group Ltd. Published by Elsevier Ltd. All rights reserved.
1. Introduction Hemifacial spasm (HFS) is a syndrome of the unilateral facial nerve hyperexcitability, resulting in an involuntary contraction or twitching of the facial muscles, which is thought to be caused by neurovascular compression of the facial nerve at its root exit zone (REZ) [1,2]. Due to the anatomical configuration of the posterior fossa, orientals are more susceptible to the attack [3]. Microvascular decompression (MVD), initially developed by Jannetta [4,5] and Gardner [6], is widely accepted as a safe and effective treatment modality for HFS owing to its high curative and low recurrence ratio [1]. However, the postoperative curative rate varies between 84% and 97%, and there is a 2.6%e18.3% chance of delayed relief and a 1%e5.3% chance of recurrence [3,7,8]. Delayed improvement or recurrence may be directly related to incomplete decompression. Furthermore, small arterioles compressing the branches of the facial nerve have often been ignored. Abnormal muscle response (AMR), as an objective indicator in the MVD, is an abnormal facial muscle EMG activity, which is elicited by stimulating a branch of the facial nerve and recorded from facial muscles innervated by the other branches. Besides, AMR is an abnormal electromyogram response observed exclusively in individuals who suffer from HFS [9] with the latency of approximately 9e10 ms and the amplitude of 0.1e0.2 mV. In most HFS patients, intraoperative facial nerve EMG is instantaneously eliminated when the involving vessels are isolated from the facial nerve [10]. Accordingly, AMR is used for identifying offending vessels and confirming if the decompression of the facial nerve is satisfactory [11]. Occasionally, multiple compressing vessels are observed during the surgery and the traditional method for facial nerve EMG monitoring cannot determine which one is the major culprit and multiple neurovascular compression is observed in about 38% cases of HFS [12]. In addition, controversy remains regarding the value of intraoperative AMR monitoring as a reliable and significant indicator of post-operative situation [13,14]. Failure to identify the offending vessels is the primary cause for MVD failure. Intraoperative electrophysiology is a favorable adjuvant to navigate the surgery and a new monitoring method is therefore needed when AMR derived from only one branch of the facial nerve is occasionally not available or unreliable. This study was an observational clinical trial and retrospective study exploring a new method which recorded the AMR when the intraoperative facial nerve EMG was absent before sufficient decompression, hoping to investigate if the improved AMR findings for MVD surgery adequately reflected the long-term outcomes of HFS patients and increased the efficiency of surgery. 2. Materials and methods 2.1. Patients This study involved 1138 classical HFS patients (455 men and 683 women) who underwent optimized AMR monitoring during MVD, from 1st January 2013 to 31 January 2015 in Yuquan Hospital, Tsinghua University. Also, the current study included 1023 HFS undergoing traditional intraoperative monitoring from 1st January
2010 to 31st December 2012 in the same center. All the patients were surgically treated by Huancong Zuo medical group. Clinical characteristics of all the HFS patients are shown in Table 1. The mean age of the patients at surgery was 53.6 years (range, 28e70 years), and the median duration of the symptom was 68.1 months (range, 5e192 months). The spasm affected the left side in 531 patients and the right side in 607 patients. The diagnosis of HFS was made based on the clinical history of typical symptoms and physical examination. Besides, preoperative magnetic resonance imaging was performed in all these HFS patients to exclude tumors or any other causes of HFS. To confirm the neurovascular compression before surgery and rule out other diseases, three-dimensional time of flight magnetic resonance angiography (3D-TOF MRA) was performed. All the patients had been medicated preoperatively but eventually in vain. All these patients gave their consent for inclusion in this submission. Their clinical features and surgical outcomes were analyzed. 2.2. Anesthesia General anesthesia was induced and maintained by intravenous infusion of propofol (1 mg/kg for induction and 7e8 mg/kg/h for maintenance) and fentanyl. Higher dose of muscle relaxant would cause neuromuscular junction dysfunction and affect electromyography responses of the facial nerve, accordingly affecting the sensitivity and accuracy of intraoperative abnormal facial muscle EMG activity monitoring. After anesthesia was stabilized, abnormal EMG activity was elicited with supramaximal stimulation. 2.3. Optimized intraoperative AMR monitoring method Monitoring equipment and electrodes: Eclipse Neurological Workstation was used for intraoperative monitoring (AXON, USA). The stimulating and recording electrodes were needle electrodes. All patients received operation under electrophysiological monitoring. General anesthesia was induced by using the muscle relaxant that had a shorter half-life. No more muscle relaxant was administered after anesthesia induction.
Table 1 Clinical characteristics of 1138 patients with HFS who employed the optimized methods. Variable Age (years) Sex Male, No. (%) Female, No. (%) HFS side Right, No. (%) Left, No. (%) Duration of HFS (months) Follow-up periods (months) Type of vessel compression Single, No. (%) Multiple, No. (%)
53.6 ± 12.1 455 (40.0) 683 (60.0) 607 (53.3) 531 (46.7) 68.1 ± 13.6 20.7 ± 9.3 849 (74.6) 289 (25.4)
C. Huang et al. / International Journal of Surgery 38 (2017) 67e73
Monitoring method: The temporal branch of the facial nerve on the affected side was stimulated (the anode and cathode were placed 1 cm lateral to the outer canthus on the two sides with a distance of 1 cm). AMR waves were recorded at frontal muscle on the affected side (1 cm above the mid-point of supercillary arch), orbicularis oris on the affected side (0.5 cm medial to the upper margin of the corner of mouth), mentalis (1 cm from the mid-point of the line connecting the center of the lower lip and the center of the mandible towards the affected side) and platysma (3 cm from the mid-point of anterior midline of the neck towards the affected side). Two recording electrodes were inserted into the muscle belly with a distance of 0.5 cm. The stimulation parameters: square wave, bandwidth 0.2 ms, frequency 4.7 Hz, intensity 15e45 mA. Recording parameters: frequency 20e3000 Hz, scan time 50 ms. Recording time of AMR waves: before opening dura mater, and during discharge of cerebrospinal fluid, dissection of arachnoid membrane and blood vessels, implanting Teflon Felt, and after washing with water and suturing the dura mater. After the anesthetic agent was totally metabolized and before the dura mater was opened, AMR waves were recorded every 3 min, for 20 times consecutively. AMR waves were continuously monitored during discharge of cerebrospinal fluid, dissection of arachnoid membrane and blood vessels and implantation of Teflon Felt. If AMR waves disappeared, the stimulation intensity was increased to as high as 60 mA. If AMR waves were still absent, then it was judged as disappearance of AMR. After washing with water and suturing the dura mater, AMR waves were further monitored every 3 min.
69
either an interview in the out-patient department or telephone interview at the end of the postoperative follow-up period in October 2015, and the mean duration of the postoperative followup was 20.7 ± 9.3 months (range, 5e38 months). It was judged that patients were “cured” when their symptoms of HFS were totally eliminated. The postoperative result of patients was considered “relief” if the symptom was improved by more than 75% in degree and frequency. In addition, “no relief” was judged when decrease in spasm was less than 25% or unchanged. Additionally, we also focused on possible operative complications.
2.6. Study approval and informed consent This study was approved by the ethics committee of Yuquan Hospital, Tsinghua University. All individuals gave their informed consent prior to their inclusion in this study.
2.7. Statistical analysis SAS (version V8) was used for data analysis. Data for categorical variables are expressed as a percentage and compared with c2 Test or Fisher exact test (2-tailed). Data for continuous variables are expressed as mean ± SD and compared with two-tailed Student t tests. Differences were considered significant at P < 0.05. All analyses were conducted by an observer blinded to the groups.
2.4. Decompression procedures 3. Results Patients were placed in the lateral oblique position. The body was rotated slightly towards the operator, and the head was fixed in a neutral position with frame fixation. A retro-mastoid craniectomy was created close to the sigmoid sinus laterally in the posterior fossa. After the edge of the sigmoid sinus was identified, the dura mater and the arachnoids were thoroughly opened, and the choroid plexus and flocculus were then gently retracted to expose the entire REZ. The latter was subsequently ascertained close to the brainstem via the infrafloccular route. The facial nerve was then explored circumferentially and thoroughly along its entire intracranial course, including the attached segment (AS), the root exit point (RExP), the REZ and distal cisternal portion (CP), from its REZ at the brainstem laterally to its entrance into the meatus [15,16]. In order to separate the offending vessels from the facial nerve, the offending arteries were moved away and dissected from the REZ by inserting a small piece of soft shredded polytetrafluoroethylene (Teflon) felt between the vessels and the brainstem or the flocculus. Any involved vessels, including arterioles, were detached from the facial nerve. The course of the offending vessel was transposed or thin pieces of Teflon Felt were interposed until the whole intracranial segment of the facial nerve was exposed and totally decompressed. Provided that the AMR was absent after decompression, we believed satisfactory surgical outcomes were obtained and the incision was closed. If the AMR did not cease instantaneously following the first decompression, or if AMR experienced recurrence, we would further inspect the distal region of the facial nerve to determine whether the offending arteries were completely eliminated. If AMR was found to be persistent following adequate decompression, the microsurgical procedures should be completed. 2.5. Surgical outcome evaluation and postoperative follow-up All these patients included were evaluated for HFS on the next day after operation and at an interval of 6 months after surgery, by
3.1. Neurophysiological recordings The basic information of patients receiving modified MVD is shown in Table 1. AMR changes at frontal muscle, orbicularis oris, mentalis and platysma during MVD are shown in Fig. 1. The improved MVD group consisted of 1138 patients. According to electromyography at year 1 after surgery, 1130 cases had no AMR (99.3%), while AMR was still present in 8 cases (0.7%). The two groups showed no significant difference in age, gender, course of disease of HFS, the affected site, follow-up period and the number of compressed vessels (single or multiple) (Table 2). The detection rates of AMR at different stages of surgery were compared between the two different neurophysiological approaches. It was found that the number of cases presenting with no AMR during dura mater opening, vessel dissection and Teflon Felt implantation was not significantly different (Table 3). The response rate was also compared. At day 1, month 3 and year 1 after surgery, the response rate of the modified MVD group was 95.1%, 97.4% and 99.3% compared to 92.2%, 95.0% and 97.8% in the traditional MVD group, respectively (P < 0.05) (Table 4). The relationship between AMR and symptoms of HFS was analyzed in the modified MVD group. At day 1 after surgery, AMR was absent in 1066 cases who received modified MVD, and 1035 cases were responsive to the treatment. Of 72 cases who still presented with AMR, 25 cases failed to achieve a response. There was a significant correlation between AMR and symptoms of HFS (P < 0.05). At month 3 and year 1 after surgery, 1084 and 1104 cases presented with no AMR, respectively; among these cases, 1067 and 1100 cases achieved a response, respectively, which also indicated significant correlation (P < 0.05). Therefore, at day 1, month 3 and year 1 after modified MVD surgery, the diagnostic sensitivity of AMR disappearance for HFS relief was 95.7%, 96.3% and 97.3%, respectively; the specificity was 44.6%, 43.3% and 50.0%, respectively; accuracy was 93.1%, 94.9% and 97.4%, respectively (Table 5).
70
C. Huang et al. / International Journal of Surgery 38 (2017) 67e73
Fig. 1. AMR waves detected in HFS patients. (A) AMR waves are detected intraoperatively at the temporal branch, buccal branch, marginal mandibular branch and cervical branch of the facial nerve for many HFS patients. (B) If only one branch is detected, AMR may be still present in other branches and insufficient decompression will affect these branches, which further leads to relapse.
Table 2 Demographic data of patients who employed the optimized methods with association of intraoperative AMR findings. Characters Age (years) Sex (Male/Female) Duration of HFS (years) HFS side (right/left) Follow-up periods (months) Type of vessel compression (single/multiple)
Non-AMR-disappeared (n ¼ 23)
AMR-disappeared (n ¼ 1115)
p-value
47.5 ± 10.4 10/13 72.2 ± 15.2 13/10 18.2 ± 8.1
53.7 ± 12.6 445/670 68.0 ± 13.1 594/521 20.8 ± 9.5
0.7453 0.7295 0.7414 0.7573 0.8110
19/4
987/128
0.3808
Table 4 Relief rates at day 1, month 3 and year 1 after surgery for patients who employed the traditional and optimized method. Groups
1 day HFS ()
Traditional Optimized p-value
3 months HFS (þ)
943 80 1082 56 0.0056
Groups
Traditional Optimized P value
HFS ()
HFS (þ)
92.2% 95.1%
972 51 1108 30 0.0041
1 year Relief rate
HFS ()
HFS (þ)
95.0% 97.4%
1000 23 1130 8 0.0026
Relief rate 97.8% 99.3%
Table 5 Time course of postoperative outcomes with association of intraoperative AMR recordings.
Table 3 The number of patients and timing of abnormal EMG disappearance. Groups
Relief rate
Timing of AMR disappearance Vessel dissection
Teflon patch implantation
108 139 0.2267
67 74 0.9433
786 863 0.5859
3 months
1 year
HFS () HFS (þ) HFS () HFS (þ) HFS () HFS (þ) Total
Dura opening
1 day
961/1023 1066/1138 0.7977
3.2. Intraoperative AMR monitoring findings In the current study, AMR waves are detected intraoperatively at the 4 branches of the facial nerve (temporal branch, buccal branch, marginal mandibular branch and cervical branch) for the patients suffered from HFS (Fig. 1). If only one branch is detected (Fig. 2),
AMR-disappeared Non-AMR-disappeared Total p-value
1035 47 1082 0.0000
31 25 56
1067 41 1108 0.0000
17 13 30
1100 30 1130 0.0000
4 4 8
AMR may be still present in other branches and insufficient decompression will affect these branches, which further leads to relapse. Fig. 3 shows the changes of AMR of the 4 branches of the facial nerve during every step of MVD and the intraoperative AMR monitoring depicted a gradual disappearance of AMR during craniotomy. Fig. 3 indicates that the root of the facial nerve may be compressed by several vessels, which leads to persistence of HFS or HFS relapse. Therefore, the vessels should be decompressed one by one to achieve the optimal efficacy.
C. Huang et al. / International Journal of Surgery 38 (2017) 67e73
71
Fig. 2. Changes of AMR of the 4 branches of the facial nerve during MVD. Intraoperative AMR monitoring depicted a gradual disappearance of AMR during craniotomy: (A) before anesthesia (baseline response); (B) craniotomy and after; (C) dura opening; (D) cerebellar retraction with removal of the cerebrospinal fluid; (E) dissociation of blood vessels and neuroanatomy; (F) interposition of polytetrafluoroethylene between the vessel and the brainstem after mobilization of the offending vessels; (G) dura mater suturing. AMR disappeared after microvascular decompression.
Fig. 3. Intraoperative findings of MVD. The root of the facial nerve may be compressed by several vessels, which leads to HFS. Therefore, the vessels should be decompressed one by one to achieve the best result. (A) Before MVD; (B) After MVD.
3.3. Recurrence rates and complications In our study, the recurrence rates at 1 year after surgery of traditional and optimized intraoperative monitoring were 0.39% and 0.18% respectively. No patient died from MVD surgery. The main complications at 1 year after surgery were tinnitus, mild facial palsy, hearing disorder and leakage of cerebrospinal fluid. The incidence rates were 0.98%, 88%, 0.78%, 0.49% (traditional monitoring) and 0.62%, 0.53%, 0.53%, 0.26% (optimized monitoring). 4. Discussion During the MVD, the real offending vessel is sometimes neglected by the operators. In the present research, a new experience using optimized intraoperative AMR monitoring and continuous intraoperative monitoring is adopted for MVD and finally proved a practical value. The current study demonstrated that the
optimized method could improve the positive detection rate of AMR and accuracy of decompression effect prediction, besides, evaluation for the decompression effect by optimized monitoring method can increase the immediate remission rate and reduce the delayed recovery rate. It has been assumed that AMR disappearance, being indicative of adequate decompression in MVD, correlates with resolution of HFS. However, the validity of intraoperative electrophysiological monitoring in predicting the postoperative situation in clinical practice remains controversial. Many investigators have studied the relationship between the persistence or disappearance of AMR findings after MVD and the prognosis of HFS [17,18]. A recent metaanalysis indicated that the chance of recovery when the AMR wave vanished after surgery was 4.2 times greater than that if the response persisted [19]. However, it was sometimes difficult to ascertain the culprit vessels and the real offending vessels were overlooked intraoperatively [20,21]. AMR usually vanishes after the
72
C. Huang et al. / International Journal of Surgery 38 (2017) 67e73
MVD surgery, but more frequently, the wave is stable at the beginning but becomes unstable when the neurovascular conflict is exposed. Under this situation, it is difficult to tell if the vessels decompressed is a real culprit or not, especially when an apparently offending artery is already found. Optimized AMR monitoring will thus help to identify the real culprit vessels, especially in such atypical compression cases. Under normal conditions, the stimulation of one branch of the facial nerve only causes contraction of the muscles innervated by this branch. But in HFS patients, electrical stimulation of one branch of the facial nerve can also cause contraction of muscles innervated by other branches of the facial nerve. Moreover, the evoked potentials are slightly delayed during electromyography (EMG) on the muscles innervated by other ipsilateral facial nerve branches. This response is known as AMR or lateral spread response. This is followed by subsequent conduction to the facial nerve branches. However, no standards have been established as to which facial nerve branches to be stimulated and the position of AMR recording. Generally two facial nerve branches that are distant from each other are stimulated and recorded. Typically the marginal mandibular branch is stimulated and the recording is performed at the orbicularis oculi; or the frontal branch is stimulated and the recording is performed at the orbicularis oris or the mentalis. Through comparison, Tamura [22] found that it was easier to stimulate the zygomatic branch and record AMR at the mentalis (orbicularis oris) than to stimulate the marginal mandibular branch and record AMR at the orbicularis oculi. This method may preclude waveform disturbance caused by contraction of the surrounding muscles. But according to our clinical observation, for some patients, when one facial nerve branch was stimulated, AMR was not recorded on all branches. The wave detection rate was low and it was difficult to decide whether the facial nerve was fully decompressed. Optimized AMR monitoring was adopted in our center, with the zygomatic branch of the facial nerve stimulated and the temporal branch, buccal branch, marginal mandibular branch and cervical branch of the facial nerve detected for AMR. This method can improve the wave detection rate, the positive detection rate and accuracy of decompression effect prediction. AMR monitoring using this method can provide guidance for MVD and help prevent post-operative complications due to excess operation. Even though AMR disappearance is regarded as an indicator of adequate decompression, the reliability of AMR monitoring as an excellent indicator of long-term outcome remains a debate, which focuses on whether the persistent AMR following decompression is correlated with improved clinical symptoms or favorable prognosis. Our studies indicated that HFS patients with AMR reduction to 25% or less were both correlated with satisfactory outcomes at the last follow-up. On the other hand, several studies also found that patients with reduction in the AMR amplitude frequently presented with good efficacy and experienced complete resolution compared to those with remained and unchanged AMR [23,24]. However, our findings indicated that even though persistent AMR was obtained after the initial decompression, patients can still achieve excellent outcomes if a thorough and effective decompression was performed, which accorded with the previous studies [10,17,18,25,26]. lntraoperative facial nerve EMG is thus useful for predicting the symptom-free status of patients after surgery and yields very few false-negative findings, if AMR amplitude reduction and cessation are both considered as reliable indicators of MVD completion. The cessation of AMR in the completion of MVD underlies a high likelihood of HFS cure in the follow-up period. According to the previous researches, the proportion of patients with persistent
spasm despite absent AMR has been consistently low, ranging from 0 to 11.6% [23]. Several large clinical series published in recent years showed that the overall cure rate of MVD was approximately 92%e 98.7% (immediate plus delayed resolution) [20, 23, 27], and our findings demonstrated better outcomes than other major centers around the world. Although the cure rate overall seems satisfactory, MVD is not effective in the long term in approximately 10%e30% of patients with HFS [20], which means there is still much space for improvement. In traditional AMR monitoring, AMR was detected in only one branch of the facial nerve. However, there was the possibility that AMR was present in other branches and insufficient decompression of these branches may result in failed surgery or relapse. This was an important reason for poor short-term or longterm prognosis. It can be seen that AMR waves were present in the temporal branch, buccal branch, marginal mandibular branch and cervical branch of the facial nerve simultaneously. If only one branch was detected, compression of other affected branches would still lead to HFS. Thus, a new method is needed to further improve the long-term efficacy. Our investigation suggested that the current optimized AMR recording method provided more useful information than the traditional one, and helped to confirm whether the compression site was sufficiently decompressed as well. Moreover, the current findings demonstrated that the AMR vanished after the offending vessels were properly treated in 99.3% of the relieved HFS patients, which was in accordance with the previous researches [28]. The recurrent spasm may be sometimes attributed to the movement of the implant or development of new vessels [19]. However, misidentification of the involved vessels or unidentified small arterioles which are accounted for the failure operation is often avoidable. On account that the facial nerve REZ is the most common area to discover the neurovascular conflict, the surgeons are inclined to finish further exposure when a “typical” REZ compression is found in this area, especially when a dent is visualized on the depressed nerve when the vessel is removed and AMR has also abolished. The utility of the current optimized method is particularly helpful for individuals with HFS who experience facial nerve compression by multiple vessels or neurovascular compression in the distal portion of the facial nerve, and plays a crucial role to identify whether there is sufficient decompression. Consequently, to obtain excellent postoperative outcomes, we strongly recommend monitoring all the facial nerve branches, as well as exposing the whole intracranial course of the facial nerve root during MVD as possible. 5. Conclusion Intraoperative neurophysiology is a favorable navigator for the surgery. We now use a new method for intraoperative monitoring of HFS. The optimized abnormal muscle response (AMR) monitoring provides more useful information than the traditional methods, especially when AMR is absent before decompression or unreliable occasionally. Particularly, monitoring all the facial nerve branches for AMR may be an effective strategy for MVD surgery. When AMR recorded by only one branch of the facial nerve is absent, this can increase the wave detection rate, guide the surgical procedures and improve the success rate of operation. Being a reliable indicator of long-term post-operative outcomes, optimized intraoperative neural electrophysiological monitoring is expected to optimize the effectiveness of MVD and improve reliability of intraoperative AMR monitoring. And this study has determined the practical value of optimized intraoperative electrophysiological monitoring in MVD for HFS. We believe it is a good practice to use this technique in our current patients and it promotes the
C. Huang et al. / International Journal of Surgery 38 (2017) 67e73
immediate effect of the microvascular decompression surgery for HFS patients.
[9]
[10]
[11]
[12] [13]
Ethical approval
[14]
Not sought for and not necessary as this was a retrospective study.
[15]
[16]
Sources of funding None. Author contribution Study design: Huancong Zuo and Yu Ma. Data collection: Chuyi Huang, Suhua Miao and Chuanfu Dai. Data analysis: Heling Chu, Jinting Wu and Junhua Wang. Writing: Chuyi Huang, Suhua Miao and Heling Chu. Conflicts of interest None. Guarantor Huancong Zuo. References [1] A.R. Moller, The cranial nerve vascular compression syndrome: II. A review of pathophysiology, Acta. Neurochir. (Wien) 113 (1991) 24e30. [2] A.R. Moller, Vascular compression of cranial nerves: II: pathophysiology, Neurol. Res. 21 (1999) 439e443. [3] L. Xia, J. Zhong, J. Zhu, N.N. Dou, M.X. Liu, S.T. Li, Delayed relief of hemifacial spasm after microvascular decompression, J. Craniofac. Surg. 26 (2015) 408e410. [4] P.J. Jannetta, Microsurgical approach to trigeminal nerve for tic douloureux, Prog. Neurol. Surg. 7 (1976) 180e200. [5] P.J. Jannetta, Neurovascular compression in cranial nerve and systemic disease, Ann. Surg. 192 (1980) 518e525. [6] W.J. Gardner, Concerning the mechanism of trigeminal neuralgia and hemifacial spasm, J. Neurosurg. 19 (1962) 947e958. [7] J. Zhu, S.T. Li, J. Zhong, H.X. Guan, T.T. Ying, M. Yang, et al., Role of arterioles in management of microvascular decompression in patients with hemifacial spasm, J. Clin. Neurosci. 19 (2012) 375e379. [8] M. Yang, X. Zheng, T. Ying, J. Zhu, W. Zhang, X. Yang, et al., Combined intraoperative monitoring of abnormal muscle response and Z-L response for
[17]
[18]
[19]
[20]
[21] [22] [23]
[24]
[25]
[26]
[27]
[28]
73
hemifacial spasm with tandem compression type, Acta. Neurochir. (Wien) 156 (2014) 1161e1166. T.T. Ying, S.T. Li, J. Zhong, X.Y. Li, X.H. Wang, J. Zhu, The value of abnormal muscle response monitoring during microvascular decompression surgery for hemifacial spasm, Int. J. Surg. 9 (2011) 347e351. J. Li, Y. Zhang, H. Zhu, Y. Li, Prognostic value of intra-operative abnormal muscle response monitoring during microvascular decompression for longterm outcome of hemifacial spasm, J. Clin. Neurosci. 19 (2012) 44e48. K. Heuser, E. Kerty, P.K. Eide, M. Cvancarova, E. Dietrichs, Microvascular decompression for hemifacial spasm: postoperative neurologic follow-up and evaluation of life quality, Eur. J. Neurol. 14 (2007) 335e340. M. Campos-Benitez, A.M. Kaufmann, Neurovascular compression findings in hemifacial spasm, J. Neurosurg. 109 (2008) 416e420. S. Saito, A.R. Moller, Chronic electrical stimulation of the facial nerve causes signs of facial nucleus hyperactivity, Neurol. Res. 15 (1993) 225e231. D.B. Sanders, Ephaptic transmission in hemifacial spasm: a single-fiber EMG study, Muscle Nerve 12 (1989) 690e694. Y. Li, X. Zheng, X. Hua, T. Ying, J. Zhong, W. Zhang, et al., Surgical treatment of hemifacial spasm with zone-4 offending vessel, Acta. Neurochir. (Wien) 155 (2013) 849e853. M.C. Kang, Y.S. Choi, H.K. Choi, S.H. Lee, C.G. Ghang, C.H. Kim, Efficacy of the disappearance of lateral spread response before and after microvascular decompression for predicting the long-term results of hemifacial spasm over two years, J. Korean Neurosurg. Soc. 52 (2012) 372e376. N. Kiya, U. Bannur, A. Yamauchi, K. Yoshida, Y. Kato, T. Kanno, Monitoring of facial evoked EMG for hemifacial spasm: a critical analysis of its prognostic value, Acta. Neurochir. (Wien) 143 (2001) 365e368. D.S. Kong, K. Park, B.G. Shin, J.A. Lee, D.O. Eum, Prognostic value of the lateral spread response for intra-operative electromyography monitoring of the facial musculature during microvascular decompression for hemifacial spasm, J. Neurosurg. 106 (2007) 384e387. R.F. Sekula Jr., S. Bhatia, A.M. Frederickson, P.J. Jannetta, M.R. Quigley, G.A. Small, et al., Utility of intraoperative electromyography in microvascular decompression for hemifacial spasm: a meta-analysis, Neurosurg. Focus 27 (2009) E10. S. Li, W. Hong, Y. Tang, T. Ying, W. Zhang, X. Li, et al., Reoperation for persistent hemifacial spasm after microvascular decompression with the aid of intraoperative monitoring of abnormal muscle response, Acta. Neurochir. (Wien) 152 (2010) 2113e2118. J. Zhong, S.T. Li, J. Zhu, H.X. Guan, Is entire nerve root decompression necessary for hemifacial spasm? Int. J. Surg. 9 (2011) 254e257. K. Tamura, T. Nishimura, K. Mori, T. Ohira, Abnormal evoked EMG and blink reflex response in patients with HFS, No. To. Shinkei 46 (1994) 479e485. S. Yamashita, T. Kawaguchi, M. Fukuda, M. Watanabe, R. Tanaka, S. Kameyama, Abnormal muscle response monitoring during microvascular decompression for hemifacial spasm, Acta. Neurochir. (Wien) 147 (2005) 933e938. M. Fukuda, S. Yamashita, T. Kawaguchi, M. Watanabe, H. Murakami, T. Takao, et al., Abnormal muscle response monitoring during microvascular decompression for hemifacial spasm and long term results, No. Shinkei. Geka 34 (2006) 583e589. J. Hatem, M. Sindou, C. Vial, Intra-operative monitoring of facial EMG responses during microvascular decompression for hemifacial spasm. Prognostic value for long-term outcome: a study in a 33-patient series, Br. J. Neurosurg. 15 (2001) 496e499. W.I. Joo, K.J. Lee, H.K. Park, C.K. Chough, H.K. Rha, Prognostic value of intraoperative lateral spread response monitoring during microvascular decompression in patients with hemifacial spasm, J. Clin. Neurosci. 15 (2008) 1335e1339. R. Huh, I.B. Han, J.Y. Moon, J.W. Chang, S.S. Chung, Microvascular decompression for hemifacial spasm: analyses of operative complications in 1582 consecutive patients, Surg. Neurol. 69 (2008) 153e157. M. Fukuda, M. Oishi, T. Takao, T. Hiraishi, Y. Sato, Y. Fujii, Monitoring of abnormal muscle response and facial motor evoked potential during microvascular decompression for hemifacial spasm, Surg. Neurol. Int. 3 (2012) 118.
Contents lists available at ScienceDirect
International Journal of Surgery journal homepage: www.journal-surgery.net
Original research
An optimized abnormal muscle response recording method for intraoperative monitoring of hemifacial spasm and its long-term prognostic value Chuyi Huang a, b, 1, Suhua Miao c, 1, Heling Chu d, Chuanfu Dai e, Jinting Wu a, c, Junhua Wang c, Huancong Zuo c, *, Yu Ma c, ** a
Clinical Neuroscience Institute, Yuquan Hospital, Medical Center, Tsinghua University, No. 5 Shijingshan Road, Shijingshan District, Beijing, 100040, China Department of Neurology, Shanghai Jiaotong University Affiliated Sixth People's Hospital, No. 600 Yishan Road, Shanghai, 200030, China Department of Neurosurgery, Tsinghua University Yuquan Hospital, No. 5 Shijingshan Road, Shijingshan District, Beijing, 100049, China d Department of Neurology, Huashan Hospital, Fudan University, No.12 Mid. Wulumuqi Road, Shanghai, 200040, China e Department of Otolaryngology, Head and Neck Surgery, Beijing Tsinghua Changgung Hospital, Tsinghua University, Beijing, 102218, China b c
h i g h l i g h t s Providing an optimized AMR recording method during MVD for HFS. Optimized method improves the positive detection rate of AMR. Optimized method increases accuracy of decompression effect prediction. Increasing the immediate remission rate and reducing the delayed recovery rate.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 12 September 2016 Received in revised form 15 November 2016 Accepted 20 December 2016 Available online 25 December 2016
Background: Intraoperative electrophysiological monitoring is used to determine whether decompression is sufficient during microvascular decompression (MVD) for hemifacial spasm (HFS). However, the real offending vessel is sometimes neglected by the neurosurgeons. Here, we reported our experience in using optimized abnormal muscle response (AMR) monitoring and continuous intraoperative monitoring for MVD. Methods: This study included 2161 HFS patients who underwent MVD using traditional (1023 patients) and optimized (1138 patients) methods. Modified AMR monitoring was adopted in our study, with the zygomatic branch of the facial nerve stimulated and the temporal branch, buccal branch, marginal mandibular branch and cervical branch of the facial nerve detected for AMR. These cases were analyzed retrospectively with emphasis on the postoperative outcomes and intraoperative findings. The therapeutic effect was evaluated at day 1, month 3 and year 1 after operation. Results: The relief rate at day 1, month 3 and year 1 after operation for patients who employed optimized AMR recording method was 95.1%, 97.4% and 99.3%, comparing with 92.2%, 95.0% and 97.8% in traditional method. There was significant difference in achieved immediate remission and recovery rate during 12month follow-up between the two groups (P < 0.05). The modified intraoperative monitoring showed the sensitivity of AMR disappearance to judge the relief at day 1, month 3 and year 1 after HFS operation was 95.7%, 96.3% and 97.3%, respectively; the specificity was 44.6%, 43.3% and 50.0%, respectively; the accuracy was 93.1%, 94.9% and 97.4%, respectively.
Keywords: Abnormal muscle response Hemifacial spasm Microvascular decompression surgery
* Corresponding author. Clinical Neuroscience Institute, Yuquan Hospital, Medical Center, Tsinghua University, No. 5 Shijingshan Road, Shijingshan district, Beijing, 100040, China. ** Corresponding author. Clinical Neuroscience Institute, Yuquan Hospital, Medical Center, Tsinghua University, No. 5 Shijingshan Road, Shijingshan district, Beijing, 100040, China. E-mail addresses: [email protected] (H. Zuo), [email protected] (Y. Ma). 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.ijsu.2016.12.032 1743-9191/© 2016 IJS Publishing Group Ltd. Published by Elsevier Ltd. All rights reserved.
68
C. Huang et al. / International Journal of Surgery 38 (2017) 67e73
Conclusions: Our findings demonstrated that the optimized method could improve the positive detection rate of AMR and accuracy of decompression effect prediction. The evaluation for the decompression effect by optimized intraoperative monitoring can increase the immediate remission rate and reduce the delayed recovery rate. © 2016 IJS Publishing Group Ltd. Published by Elsevier Ltd. All rights reserved.
1. Introduction Hemifacial spasm (HFS) is a syndrome of the unilateral facial nerve hyperexcitability, resulting in an involuntary contraction or twitching of the facial muscles, which is thought to be caused by neurovascular compression of the facial nerve at its root exit zone (REZ) [1,2]. Due to the anatomical configuration of the posterior fossa, orientals are more susceptible to the attack [3]. Microvascular decompression (MVD), initially developed by Jannetta [4,5] and Gardner [6], is widely accepted as a safe and effective treatment modality for HFS owing to its high curative and low recurrence ratio [1]. However, the postoperative curative rate varies between 84% and 97%, and there is a 2.6%e18.3% chance of delayed relief and a 1%e5.3% chance of recurrence [3,7,8]. Delayed improvement or recurrence may be directly related to incomplete decompression. Furthermore, small arterioles compressing the branches of the facial nerve have often been ignored. Abnormal muscle response (AMR), as an objective indicator in the MVD, is an abnormal facial muscle EMG activity, which is elicited by stimulating a branch of the facial nerve and recorded from facial muscles innervated by the other branches. Besides, AMR is an abnormal electromyogram response observed exclusively in individuals who suffer from HFS [9] with the latency of approximately 9e10 ms and the amplitude of 0.1e0.2 mV. In most HFS patients, intraoperative facial nerve EMG is instantaneously eliminated when the involving vessels are isolated from the facial nerve [10]. Accordingly, AMR is used for identifying offending vessels and confirming if the decompression of the facial nerve is satisfactory [11]. Occasionally, multiple compressing vessels are observed during the surgery and the traditional method for facial nerve EMG monitoring cannot determine which one is the major culprit and multiple neurovascular compression is observed in about 38% cases of HFS [12]. In addition, controversy remains regarding the value of intraoperative AMR monitoring as a reliable and significant indicator of post-operative situation [13,14]. Failure to identify the offending vessels is the primary cause for MVD failure. Intraoperative electrophysiology is a favorable adjuvant to navigate the surgery and a new monitoring method is therefore needed when AMR derived from only one branch of the facial nerve is occasionally not available or unreliable. This study was an observational clinical trial and retrospective study exploring a new method which recorded the AMR when the intraoperative facial nerve EMG was absent before sufficient decompression, hoping to investigate if the improved AMR findings for MVD surgery adequately reflected the long-term outcomes of HFS patients and increased the efficiency of surgery. 2. Materials and methods 2.1. Patients This study involved 1138 classical HFS patients (455 men and 683 women) who underwent optimized AMR monitoring during MVD, from 1st January 2013 to 31 January 2015 in Yuquan Hospital, Tsinghua University. Also, the current study included 1023 HFS undergoing traditional intraoperative monitoring from 1st January
2010 to 31st December 2012 in the same center. All the patients were surgically treated by Huancong Zuo medical group. Clinical characteristics of all the HFS patients are shown in Table 1. The mean age of the patients at surgery was 53.6 years (range, 28e70 years), and the median duration of the symptom was 68.1 months (range, 5e192 months). The spasm affected the left side in 531 patients and the right side in 607 patients. The diagnosis of HFS was made based on the clinical history of typical symptoms and physical examination. Besides, preoperative magnetic resonance imaging was performed in all these HFS patients to exclude tumors or any other causes of HFS. To confirm the neurovascular compression before surgery and rule out other diseases, three-dimensional time of flight magnetic resonance angiography (3D-TOF MRA) was performed. All the patients had been medicated preoperatively but eventually in vain. All these patients gave their consent for inclusion in this submission. Their clinical features and surgical outcomes were analyzed. 2.2. Anesthesia General anesthesia was induced and maintained by intravenous infusion of propofol (1 mg/kg for induction and 7e8 mg/kg/h for maintenance) and fentanyl. Higher dose of muscle relaxant would cause neuromuscular junction dysfunction and affect electromyography responses of the facial nerve, accordingly affecting the sensitivity and accuracy of intraoperative abnormal facial muscle EMG activity monitoring. After anesthesia was stabilized, abnormal EMG activity was elicited with supramaximal stimulation. 2.3. Optimized intraoperative AMR monitoring method Monitoring equipment and electrodes: Eclipse Neurological Workstation was used for intraoperative monitoring (AXON, USA). The stimulating and recording electrodes were needle electrodes. All patients received operation under electrophysiological monitoring. General anesthesia was induced by using the muscle relaxant that had a shorter half-life. No more muscle relaxant was administered after anesthesia induction.
Table 1 Clinical characteristics of 1138 patients with HFS who employed the optimized methods. Variable Age (years) Sex Male, No. (%) Female, No. (%) HFS side Right, No. (%) Left, No. (%) Duration of HFS (months) Follow-up periods (months) Type of vessel compression Single, No. (%) Multiple, No. (%)
53.6 ± 12.1 455 (40.0) 683 (60.0) 607 (53.3) 531 (46.7) 68.1 ± 13.6 20.7 ± 9.3 849 (74.6) 289 (25.4)
C. Huang et al. / International Journal of Surgery 38 (2017) 67e73
Monitoring method: The temporal branch of the facial nerve on the affected side was stimulated (the anode and cathode were placed 1 cm lateral to the outer canthus on the two sides with a distance of 1 cm). AMR waves were recorded at frontal muscle on the affected side (1 cm above the mid-point of supercillary arch), orbicularis oris on the affected side (0.5 cm medial to the upper margin of the corner of mouth), mentalis (1 cm from the mid-point of the line connecting the center of the lower lip and the center of the mandible towards the affected side) and platysma (3 cm from the mid-point of anterior midline of the neck towards the affected side). Two recording electrodes were inserted into the muscle belly with a distance of 0.5 cm. The stimulation parameters: square wave, bandwidth 0.2 ms, frequency 4.7 Hz, intensity 15e45 mA. Recording parameters: frequency 20e3000 Hz, scan time 50 ms. Recording time of AMR waves: before opening dura mater, and during discharge of cerebrospinal fluid, dissection of arachnoid membrane and blood vessels, implanting Teflon Felt, and after washing with water and suturing the dura mater. After the anesthetic agent was totally metabolized and before the dura mater was opened, AMR waves were recorded every 3 min, for 20 times consecutively. AMR waves were continuously monitored during discharge of cerebrospinal fluid, dissection of arachnoid membrane and blood vessels and implantation of Teflon Felt. If AMR waves disappeared, the stimulation intensity was increased to as high as 60 mA. If AMR waves were still absent, then it was judged as disappearance of AMR. After washing with water and suturing the dura mater, AMR waves were further monitored every 3 min.
69
either an interview in the out-patient department or telephone interview at the end of the postoperative follow-up period in October 2015, and the mean duration of the postoperative followup was 20.7 ± 9.3 months (range, 5e38 months). It was judged that patients were “cured” when their symptoms of HFS were totally eliminated. The postoperative result of patients was considered “relief” if the symptom was improved by more than 75% in degree and frequency. In addition, “no relief” was judged when decrease in spasm was less than 25% or unchanged. Additionally, we also focused on possible operative complications.
2.6. Study approval and informed consent This study was approved by the ethics committee of Yuquan Hospital, Tsinghua University. All individuals gave their informed consent prior to their inclusion in this study.
2.7. Statistical analysis SAS (version V8) was used for data analysis. Data for categorical variables are expressed as a percentage and compared with c2 Test or Fisher exact test (2-tailed). Data for continuous variables are expressed as mean ± SD and compared with two-tailed Student t tests. Differences were considered significant at P < 0.05. All analyses were conducted by an observer blinded to the groups.
2.4. Decompression procedures 3. Results Patients were placed in the lateral oblique position. The body was rotated slightly towards the operator, and the head was fixed in a neutral position with frame fixation. A retro-mastoid craniectomy was created close to the sigmoid sinus laterally in the posterior fossa. After the edge of the sigmoid sinus was identified, the dura mater and the arachnoids were thoroughly opened, and the choroid plexus and flocculus were then gently retracted to expose the entire REZ. The latter was subsequently ascertained close to the brainstem via the infrafloccular route. The facial nerve was then explored circumferentially and thoroughly along its entire intracranial course, including the attached segment (AS), the root exit point (RExP), the REZ and distal cisternal portion (CP), from its REZ at the brainstem laterally to its entrance into the meatus [15,16]. In order to separate the offending vessels from the facial nerve, the offending arteries were moved away and dissected from the REZ by inserting a small piece of soft shredded polytetrafluoroethylene (Teflon) felt between the vessels and the brainstem or the flocculus. Any involved vessels, including arterioles, were detached from the facial nerve. The course of the offending vessel was transposed or thin pieces of Teflon Felt were interposed until the whole intracranial segment of the facial nerve was exposed and totally decompressed. Provided that the AMR was absent after decompression, we believed satisfactory surgical outcomes were obtained and the incision was closed. If the AMR did not cease instantaneously following the first decompression, or if AMR experienced recurrence, we would further inspect the distal region of the facial nerve to determine whether the offending arteries were completely eliminated. If AMR was found to be persistent following adequate decompression, the microsurgical procedures should be completed. 2.5. Surgical outcome evaluation and postoperative follow-up All these patients included were evaluated for HFS on the next day after operation and at an interval of 6 months after surgery, by
3.1. Neurophysiological recordings The basic information of patients receiving modified MVD is shown in Table 1. AMR changes at frontal muscle, orbicularis oris, mentalis and platysma during MVD are shown in Fig. 1. The improved MVD group consisted of 1138 patients. According to electromyography at year 1 after surgery, 1130 cases had no AMR (99.3%), while AMR was still present in 8 cases (0.7%). The two groups showed no significant difference in age, gender, course of disease of HFS, the affected site, follow-up period and the number of compressed vessels (single or multiple) (Table 2). The detection rates of AMR at different stages of surgery were compared between the two different neurophysiological approaches. It was found that the number of cases presenting with no AMR during dura mater opening, vessel dissection and Teflon Felt implantation was not significantly different (Table 3). The response rate was also compared. At day 1, month 3 and year 1 after surgery, the response rate of the modified MVD group was 95.1%, 97.4% and 99.3% compared to 92.2%, 95.0% and 97.8% in the traditional MVD group, respectively (P < 0.05) (Table 4). The relationship between AMR and symptoms of HFS was analyzed in the modified MVD group. At day 1 after surgery, AMR was absent in 1066 cases who received modified MVD, and 1035 cases were responsive to the treatment. Of 72 cases who still presented with AMR, 25 cases failed to achieve a response. There was a significant correlation between AMR and symptoms of HFS (P < 0.05). At month 3 and year 1 after surgery, 1084 and 1104 cases presented with no AMR, respectively; among these cases, 1067 and 1100 cases achieved a response, respectively, which also indicated significant correlation (P < 0.05). Therefore, at day 1, month 3 and year 1 after modified MVD surgery, the diagnostic sensitivity of AMR disappearance for HFS relief was 95.7%, 96.3% and 97.3%, respectively; the specificity was 44.6%, 43.3% and 50.0%, respectively; accuracy was 93.1%, 94.9% and 97.4%, respectively (Table 5).
70
C. Huang et al. / International Journal of Surgery 38 (2017) 67e73
Fig. 1. AMR waves detected in HFS patients. (A) AMR waves are detected intraoperatively at the temporal branch, buccal branch, marginal mandibular branch and cervical branch of the facial nerve for many HFS patients. (B) If only one branch is detected, AMR may be still present in other branches and insufficient decompression will affect these branches, which further leads to relapse.
Table 2 Demographic data of patients who employed the optimized methods with association of intraoperative AMR findings. Characters Age (years) Sex (Male/Female) Duration of HFS (years) HFS side (right/left) Follow-up periods (months) Type of vessel compression (single/multiple)
Non-AMR-disappeared (n ¼ 23)
AMR-disappeared (n ¼ 1115)
p-value
47.5 ± 10.4 10/13 72.2 ± 15.2 13/10 18.2 ± 8.1
53.7 ± 12.6 445/670 68.0 ± 13.1 594/521 20.8 ± 9.5
0.7453 0.7295 0.7414 0.7573 0.8110
19/4
987/128
0.3808
Table 4 Relief rates at day 1, month 3 and year 1 after surgery for patients who employed the traditional and optimized method. Groups
1 day HFS ()
Traditional Optimized p-value
3 months HFS (þ)
943 80 1082 56 0.0056
Groups
Traditional Optimized P value
HFS ()
HFS (þ)
92.2% 95.1%
972 51 1108 30 0.0041
1 year Relief rate
HFS ()
HFS (þ)
95.0% 97.4%
1000 23 1130 8 0.0026
Relief rate 97.8% 99.3%
Table 5 Time course of postoperative outcomes with association of intraoperative AMR recordings.
Table 3 The number of patients and timing of abnormal EMG disappearance. Groups
Relief rate
Timing of AMR disappearance Vessel dissection
Teflon patch implantation
108 139 0.2267
67 74 0.9433
786 863 0.5859
3 months
1 year
HFS () HFS (þ) HFS () HFS (þ) HFS () HFS (þ) Total
Dura opening
1 day
961/1023 1066/1138 0.7977
3.2. Intraoperative AMR monitoring findings In the current study, AMR waves are detected intraoperatively at the 4 branches of the facial nerve (temporal branch, buccal branch, marginal mandibular branch and cervical branch) for the patients suffered from HFS (Fig. 1). If only one branch is detected (Fig. 2),
AMR-disappeared Non-AMR-disappeared Total p-value
1035 47 1082 0.0000
31 25 56
1067 41 1108 0.0000
17 13 30
1100 30 1130 0.0000
4 4 8
AMR may be still present in other branches and insufficient decompression will affect these branches, which further leads to relapse. Fig. 3 shows the changes of AMR of the 4 branches of the facial nerve during every step of MVD and the intraoperative AMR monitoring depicted a gradual disappearance of AMR during craniotomy. Fig. 3 indicates that the root of the facial nerve may be compressed by several vessels, which leads to persistence of HFS or HFS relapse. Therefore, the vessels should be decompressed one by one to achieve the optimal efficacy.
C. Huang et al. / International Journal of Surgery 38 (2017) 67e73
71
Fig. 2. Changes of AMR of the 4 branches of the facial nerve during MVD. Intraoperative AMR monitoring depicted a gradual disappearance of AMR during craniotomy: (A) before anesthesia (baseline response); (B) craniotomy and after; (C) dura opening; (D) cerebellar retraction with removal of the cerebrospinal fluid; (E) dissociation of blood vessels and neuroanatomy; (F) interposition of polytetrafluoroethylene between the vessel and the brainstem after mobilization of the offending vessels; (G) dura mater suturing. AMR disappeared after microvascular decompression.
Fig. 3. Intraoperative findings of MVD. The root of the facial nerve may be compressed by several vessels, which leads to HFS. Therefore, the vessels should be decompressed one by one to achieve the best result. (A) Before MVD; (B) After MVD.
3.3. Recurrence rates and complications In our study, the recurrence rates at 1 year after surgery of traditional and optimized intraoperative monitoring were 0.39% and 0.18% respectively. No patient died from MVD surgery. The main complications at 1 year after surgery were tinnitus, mild facial palsy, hearing disorder and leakage of cerebrospinal fluid. The incidence rates were 0.98%, 88%, 0.78%, 0.49% (traditional monitoring) and 0.62%, 0.53%, 0.53%, 0.26% (optimized monitoring). 4. Discussion During the MVD, the real offending vessel is sometimes neglected by the operators. In the present research, a new experience using optimized intraoperative AMR monitoring and continuous intraoperative monitoring is adopted for MVD and finally proved a practical value. The current study demonstrated that the
optimized method could improve the positive detection rate of AMR and accuracy of decompression effect prediction, besides, evaluation for the decompression effect by optimized monitoring method can increase the immediate remission rate and reduce the delayed recovery rate. It has been assumed that AMR disappearance, being indicative of adequate decompression in MVD, correlates with resolution of HFS. However, the validity of intraoperative electrophysiological monitoring in predicting the postoperative situation in clinical practice remains controversial. Many investigators have studied the relationship between the persistence or disappearance of AMR findings after MVD and the prognosis of HFS [17,18]. A recent metaanalysis indicated that the chance of recovery when the AMR wave vanished after surgery was 4.2 times greater than that if the response persisted [19]. However, it was sometimes difficult to ascertain the culprit vessels and the real offending vessels were overlooked intraoperatively [20,21]. AMR usually vanishes after the
72
C. Huang et al. / International Journal of Surgery 38 (2017) 67e73
MVD surgery, but more frequently, the wave is stable at the beginning but becomes unstable when the neurovascular conflict is exposed. Under this situation, it is difficult to tell if the vessels decompressed is a real culprit or not, especially when an apparently offending artery is already found. Optimized AMR monitoring will thus help to identify the real culprit vessels, especially in such atypical compression cases. Under normal conditions, the stimulation of one branch of the facial nerve only causes contraction of the muscles innervated by this branch. But in HFS patients, electrical stimulation of one branch of the facial nerve can also cause contraction of muscles innervated by other branches of the facial nerve. Moreover, the evoked potentials are slightly delayed during electromyography (EMG) on the muscles innervated by other ipsilateral facial nerve branches. This response is known as AMR or lateral spread response. This is followed by subsequent conduction to the facial nerve branches. However, no standards have been established as to which facial nerve branches to be stimulated and the position of AMR recording. Generally two facial nerve branches that are distant from each other are stimulated and recorded. Typically the marginal mandibular branch is stimulated and the recording is performed at the orbicularis oculi; or the frontal branch is stimulated and the recording is performed at the orbicularis oris or the mentalis. Through comparison, Tamura [22] found that it was easier to stimulate the zygomatic branch and record AMR at the mentalis (orbicularis oris) than to stimulate the marginal mandibular branch and record AMR at the orbicularis oculi. This method may preclude waveform disturbance caused by contraction of the surrounding muscles. But according to our clinical observation, for some patients, when one facial nerve branch was stimulated, AMR was not recorded on all branches. The wave detection rate was low and it was difficult to decide whether the facial nerve was fully decompressed. Optimized AMR monitoring was adopted in our center, with the zygomatic branch of the facial nerve stimulated and the temporal branch, buccal branch, marginal mandibular branch and cervical branch of the facial nerve detected for AMR. This method can improve the wave detection rate, the positive detection rate and accuracy of decompression effect prediction. AMR monitoring using this method can provide guidance for MVD and help prevent post-operative complications due to excess operation. Even though AMR disappearance is regarded as an indicator of adequate decompression, the reliability of AMR monitoring as an excellent indicator of long-term outcome remains a debate, which focuses on whether the persistent AMR following decompression is correlated with improved clinical symptoms or favorable prognosis. Our studies indicated that HFS patients with AMR reduction to 25% or less were both correlated with satisfactory outcomes at the last follow-up. On the other hand, several studies also found that patients with reduction in the AMR amplitude frequently presented with good efficacy and experienced complete resolution compared to those with remained and unchanged AMR [23,24]. However, our findings indicated that even though persistent AMR was obtained after the initial decompression, patients can still achieve excellent outcomes if a thorough and effective decompression was performed, which accorded with the previous studies [10,17,18,25,26]. lntraoperative facial nerve EMG is thus useful for predicting the symptom-free status of patients after surgery and yields very few false-negative findings, if AMR amplitude reduction and cessation are both considered as reliable indicators of MVD completion. The cessation of AMR in the completion of MVD underlies a high likelihood of HFS cure in the follow-up period. According to the previous researches, the proportion of patients with persistent
spasm despite absent AMR has been consistently low, ranging from 0 to 11.6% [23]. Several large clinical series published in recent years showed that the overall cure rate of MVD was approximately 92%e 98.7% (immediate plus delayed resolution) [20, 23, 27], and our findings demonstrated better outcomes than other major centers around the world. Although the cure rate overall seems satisfactory, MVD is not effective in the long term in approximately 10%e30% of patients with HFS [20], which means there is still much space for improvement. In traditional AMR monitoring, AMR was detected in only one branch of the facial nerve. However, there was the possibility that AMR was present in other branches and insufficient decompression of these branches may result in failed surgery or relapse. This was an important reason for poor short-term or longterm prognosis. It can be seen that AMR waves were present in the temporal branch, buccal branch, marginal mandibular branch and cervical branch of the facial nerve simultaneously. If only one branch was detected, compression of other affected branches would still lead to HFS. Thus, a new method is needed to further improve the long-term efficacy. Our investigation suggested that the current optimized AMR recording method provided more useful information than the traditional one, and helped to confirm whether the compression site was sufficiently decompressed as well. Moreover, the current findings demonstrated that the AMR vanished after the offending vessels were properly treated in 99.3% of the relieved HFS patients, which was in accordance with the previous researches [28]. The recurrent spasm may be sometimes attributed to the movement of the implant or development of new vessels [19]. However, misidentification of the involved vessels or unidentified small arterioles which are accounted for the failure operation is often avoidable. On account that the facial nerve REZ is the most common area to discover the neurovascular conflict, the surgeons are inclined to finish further exposure when a “typical” REZ compression is found in this area, especially when a dent is visualized on the depressed nerve when the vessel is removed and AMR has also abolished. The utility of the current optimized method is particularly helpful for individuals with HFS who experience facial nerve compression by multiple vessels or neurovascular compression in the distal portion of the facial nerve, and plays a crucial role to identify whether there is sufficient decompression. Consequently, to obtain excellent postoperative outcomes, we strongly recommend monitoring all the facial nerve branches, as well as exposing the whole intracranial course of the facial nerve root during MVD as possible. 5. Conclusion Intraoperative neurophysiology is a favorable navigator for the surgery. We now use a new method for intraoperative monitoring of HFS. The optimized abnormal muscle response (AMR) monitoring provides more useful information than the traditional methods, especially when AMR is absent before decompression or unreliable occasionally. Particularly, monitoring all the facial nerve branches for AMR may be an effective strategy for MVD surgery. When AMR recorded by only one branch of the facial nerve is absent, this can increase the wave detection rate, guide the surgical procedures and improve the success rate of operation. Being a reliable indicator of long-term post-operative outcomes, optimized intraoperative neural electrophysiological monitoring is expected to optimize the effectiveness of MVD and improve reliability of intraoperative AMR monitoring. And this study has determined the practical value of optimized intraoperative electrophysiological monitoring in MVD for HFS. We believe it is a good practice to use this technique in our current patients and it promotes the
C. Huang et al. / International Journal of Surgery 38 (2017) 67e73
immediate effect of the microvascular decompression surgery for HFS patients.
[9]
[10]
[11]
[12] [13]
Ethical approval
[14]
Not sought for and not necessary as this was a retrospective study.
[15]
[16]
Sources of funding None. Author contribution Study design: Huancong Zuo and Yu Ma. Data collection: Chuyi Huang, Suhua Miao and Chuanfu Dai. Data analysis: Heling Chu, Jinting Wu and Junhua Wang. Writing: Chuyi Huang, Suhua Miao and Heling Chu. Conflicts of interest None. Guarantor Huancong Zuo. References [1] A.R. Moller, The cranial nerve vascular compression syndrome: II. A review of pathophysiology, Acta. Neurochir. (Wien) 113 (1991) 24e30. [2] A.R. Moller, Vascular compression of cranial nerves: II: pathophysiology, Neurol. Res. 21 (1999) 439e443. [3] L. Xia, J. Zhong, J. Zhu, N.N. Dou, M.X. Liu, S.T. Li, Delayed relief of hemifacial spasm after microvascular decompression, J. Craniofac. Surg. 26 (2015) 408e410. [4] P.J. Jannetta, Microsurgical approach to trigeminal nerve for tic douloureux, Prog. Neurol. Surg. 7 (1976) 180e200. [5] P.J. Jannetta, Neurovascular compression in cranial nerve and systemic disease, Ann. Surg. 192 (1980) 518e525. [6] W.J. Gardner, Concerning the mechanism of trigeminal neuralgia and hemifacial spasm, J. Neurosurg. 19 (1962) 947e958. [7] J. Zhu, S.T. Li, J. Zhong, H.X. Guan, T.T. Ying, M. Yang, et al., Role of arterioles in management of microvascular decompression in patients with hemifacial spasm, J. Clin. Neurosci. 19 (2012) 375e379. [8] M. Yang, X. Zheng, T. Ying, J. Zhu, W. Zhang, X. Yang, et al., Combined intraoperative monitoring of abnormal muscle response and Z-L response for
[17]
[18]
[19]
[20]
[21] [22] [23]
[24]
[25]
[26]
[27]
[28]
73
hemifacial spasm with tandem compression type, Acta. Neurochir. (Wien) 156 (2014) 1161e1166. T.T. Ying, S.T. Li, J. Zhong, X.Y. Li, X.H. Wang, J. Zhu, The value of abnormal muscle response monitoring during microvascular decompression surgery for hemifacial spasm, Int. J. Surg. 9 (2011) 347e351. J. Li, Y. Zhang, H. Zhu, Y. Li, Prognostic value of intra-operative abnormal muscle response monitoring during microvascular decompression for longterm outcome of hemifacial spasm, J. Clin. Neurosci. 19 (2012) 44e48. K. Heuser, E. Kerty, P.K. Eide, M. Cvancarova, E. Dietrichs, Microvascular decompression for hemifacial spasm: postoperative neurologic follow-up and evaluation of life quality, Eur. J. Neurol. 14 (2007) 335e340. M. Campos-Benitez, A.M. Kaufmann, Neurovascular compression findings in hemifacial spasm, J. Neurosurg. 109 (2008) 416e420. S. Saito, A.R. Moller, Chronic electrical stimulation of the facial nerve causes signs of facial nucleus hyperactivity, Neurol. Res. 15 (1993) 225e231. D.B. Sanders, Ephaptic transmission in hemifacial spasm: a single-fiber EMG study, Muscle Nerve 12 (1989) 690e694. Y. Li, X. Zheng, X. Hua, T. Ying, J. Zhong, W. Zhang, et al., Surgical treatment of hemifacial spasm with zone-4 offending vessel, Acta. Neurochir. (Wien) 155 (2013) 849e853. M.C. Kang, Y.S. Choi, H.K. Choi, S.H. Lee, C.G. Ghang, C.H. Kim, Efficacy of the disappearance of lateral spread response before and after microvascular decompression for predicting the long-term results of hemifacial spasm over two years, J. Korean Neurosurg. Soc. 52 (2012) 372e376. N. Kiya, U. Bannur, A. Yamauchi, K. Yoshida, Y. Kato, T. Kanno, Monitoring of facial evoked EMG for hemifacial spasm: a critical analysis of its prognostic value, Acta. Neurochir. (Wien) 143 (2001) 365e368. D.S. Kong, K. Park, B.G. Shin, J.A. Lee, D.O. Eum, Prognostic value of the lateral spread response for intra-operative electromyography monitoring of the facial musculature during microvascular decompression for hemifacial spasm, J. Neurosurg. 106 (2007) 384e387. R.F. Sekula Jr., S. Bhatia, A.M. Frederickson, P.J. Jannetta, M.R. Quigley, G.A. Small, et al., Utility of intraoperative electromyography in microvascular decompression for hemifacial spasm: a meta-analysis, Neurosurg. Focus 27 (2009) E10. S. Li, W. Hong, Y. Tang, T. Ying, W. Zhang, X. Li, et al., Reoperation for persistent hemifacial spasm after microvascular decompression with the aid of intraoperative monitoring of abnormal muscle response, Acta. Neurochir. (Wien) 152 (2010) 2113e2118. J. Zhong, S.T. Li, J. Zhu, H.X. Guan, Is entire nerve root decompression necessary for hemifacial spasm? Int. J. Surg. 9 (2011) 254e257. K. Tamura, T. Nishimura, K. Mori, T. Ohira, Abnormal evoked EMG and blink reflex response in patients with HFS, No. To. Shinkei 46 (1994) 479e485. S. Yamashita, T. Kawaguchi, M. Fukuda, M. Watanabe, R. Tanaka, S. Kameyama, Abnormal muscle response monitoring during microvascular decompression for hemifacial spasm, Acta. Neurochir. (Wien) 147 (2005) 933e938. M. Fukuda, S. Yamashita, T. Kawaguchi, M. Watanabe, H. Murakami, T. Takao, et al., Abnormal muscle response monitoring during microvascular decompression for hemifacial spasm and long term results, No. Shinkei. Geka 34 (2006) 583e589. J. Hatem, M. Sindou, C. Vial, Intra-operative monitoring of facial EMG responses during microvascular decompression for hemifacial spasm. Prognostic value for long-term outcome: a study in a 33-patient series, Br. J. Neurosurg. 15 (2001) 496e499. W.I. Joo, K.J. Lee, H.K. Park, C.K. Chough, H.K. Rha, Prognostic value of intraoperative lateral spread response monitoring during microvascular decompression in patients with hemifacial spasm, J. Clin. Neurosci. 15 (2008) 1335e1339. R. Huh, I.B. Han, J.Y. Moon, J.W. Chang, S.S. Chung, Microvascular decompression for hemifacial spasm: analyses of operative complications in 1582 consecutive patients, Surg. Neurol. 69 (2008) 153e157. M. Fukuda, M. Oishi, T. Takao, T. Hiraishi, Y. Sato, Y. Fujii, Monitoring of abnormal muscle response and facial motor evoked potential during microvascular decompression for hemifacial spasm, Surg. Neurol. Int. 3 (2012) 118.