*Article* **Operative Findings of over 5000 Microvascular Decompression Surgeries for Hemifacial Spasm: Our Perspective and Current Updates**

**Jae Sung Park <sup>1</sup> and Kwan Park 2,3,\***


**Abstract:** Hemifacial spasm (HFS) is a hyperactive cranial neuropathy, and it has been well established that the cause of primary HFS is compression on the root exit zone (REZ) of the facial– vestibulocochlear nerve complex (CN VII-VIII) by a vessel or vessels. MVD is the only curative treatment option for HFS with a high success rate and low incidence of recurrence and complications. We categorize six classical compressive patterns on the REZ as well as five challenging types. Knowledge of these patterns may help in achieving a better surgical outcome.

**Keywords:** hemifacial spasm; microvascular decompression; compressive patterns

#### **1. Introduction**

Hemifacial spasm (HFS) is a hyperactive cranial neuropathy, and it has been well established that the cause of primary HFS is compression on the root exit zone (REZ) of the facial–vestibulocochlear nerve complex (CN VII-VIII) by a vessel or vessels [1,2]. The modern understanding of its etiology has contributed to the development of microvascular decompression (MVD) which can offer a cure in a non-destructive way [3,4]. Accordingly, MVD has been accepted as the treatment of choice for medically intractable primary HFS [5]. An uncountable number of HFS patients have been benefited by MVD around the globe, but there still remains much to discover. We believe this report stands out in that it is used to describe detailed operative findings of more than 5000 MVD procedures performed by a single surgeon in a single institution. A brief overview and current updates of HFS from our experience and perspective are to be presented.

#### **2. Materials and Methods**

From January 2004 to March 2020, 5026 MVDs were performed for HFS by a single surgeon in a single institution. All patients had been diagnosed with medically intractable primary HFS and underwent an MVD via a lateral retrosigmoid suboccipital approach. Preoperative evaluation included computed tomography (CT), magnetic resonance imaging (MRI) along with T2 weighted sequences, and three-dimensional time of flight MR angiography (3D TOF MRA). Neurovascular conflict is better visualized by high resolution T2 weighted images than 3D TOF MRA, especially in cases of venous compression. Pure tone audiometry along with speech audiometry were carried out before and after the MVD.

Under general anesthesia, a lateral retrosigmoid suboccipital craniotomy was performed, followed by incision of the dura, careful dissection of the arachnoid layer, and gentle retraction of the flocculus. Upon exposure of the REZ of CN VII-VIII complex, the compressing vessel, or offending vessel, was identified. A Teflon sponge was inserted

**Citation:** Park, J.S.; Park, K. Operative Findings of over 5000 Microvascular Decompression Surgeries for Hemifacial Spasm: Our Perspective and Current Updates. *Life* **2023**, *13*, 1904. https://doi.org/ 10.3390/life13091904

Academic Editor: Katalin Prokai-Tatrai

Received: 6 July 2023 Revised: 7 August 2023 Accepted: 11 September 2023 Published: 13 September 2023

**Copyright:** © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

between the REZ and the offending vessel, which completed the decompression process. Throughout the surgery, electrophysiological evaluation was used to monitor the facial nerve, i.e., the disappearance of the lateral spread response (LSR), free running electromyography (EMG) and direct nerve stimulation, as well as brainstem auditory evoked potentials (BAEP), in all patients. The compression patterns were described and categorized by the surgeon (K.P.), and, according to the categorization, they were illustrated by the first author (J.P.).

Preoperative and postoperative evaluation for symptoms were described and collected by a single nurse practitioner to minimize response bias. The data processing was carried out using commercially available software (IBM SPSS Statistics, version 24). The Chisquare test and Fisher's exact test were employed when analyzing cross tables between compression patterns and clinical outcomes. Patient consent was not necessary because of the retrospective nature of the study, and the validity of the findings would not be affected by the absence of patient consent. Moreover, no additional risk to patient safety was expected from this study without patient consent. The first author, J.P, owns all copyright privileges regarding all illustrations.

#### **3. Results**

Over the past 16 years, operative findings were described and recorded by the surgeon (K.P.). When the REZ was inspected through a microscope, how it was compressed by a vessel or vessels was not uniform. After around 5–7 years in his career as a neurosurgeon specialized in MVD, the surgeon noticed that in the vast majority of cases, there was a contributing factor that made compression somewhat inevitable. A thickened arachnoid membrane was the first thing that inspired him to pursue this categorization process according to the contributing factors, which eventually led to the creation of the "compression patterns". When the thickened arachnoid membrane was found around the compressing vessels, dissection of the arachnoid membrane often led to the disappearance of the LSR, which could indicate that the thickened arachnoid membrane was the cause of the compression. The surgeon hypothesized that the vessel was "forced" to compress the REZ by the thickened arachnoid membrane pushing it to the REZ (Figure 1B). Indeed, the arachnoid type, the most frequently observed one, accounts for 27.9% of all cases [6]. Under the same hypothesis, other forms of compression were described. When short and tight perforating arteries from the offending vessel were tethering the vessel to the REZ, we named it the "perforator" type (Figure 1C). This perforator type was also grouped in the challenging ones because the short and tight perforating arteries limited the working space for decompression, and they must not be injured to avoid any irreversible sequelae such as brain stem infarction or intracranial hemorrhage (Figure 2A). The compressing vessels of overall HFS, in order of frequency, consisted of the anterior inferior cerebellar artery (AICA, 51.7%), the posterior inferior cerebellar artery (PICA, 21.6%), and the vertebral artery (VA) [6]. AICA was involved in the perforating type in 84.5% of instances, which was disproportionately higher than in other type (*p* < 0.005) [6]. The branch type (Figure 1D) referred to a compression where the REZ was caught between branches of the offending artery, while the sandwich type (Figure 1E) illustrated a compression of the REZ by two independent arteries on each side. The two arteries in the sandwich type were either AICA + PICA or AICA + another branch of AICA; the VA was not involved in any sandwich types. In the tandem type, on the contrary, the VA was one of the two arteries in 61.5% of instances [6]. Figure 1F depicts the tandem type where a larger artery, most commonly the vertebral artery, compresses a smaller one that is in contact with the REZ. The tandem type was also categorized as a challenging one (Figure 2B), since Teflon pieces inserted between the REZ and the smaller offending artery might not be sufficient for a complete decompression; the smaller artery must be released from the pressure by the larger one as well. When there was no contributing factor other than the vascular loop itself, they were categorized as the "loop type" (Figure 1A). PICA was responsible for 72.7% of the loop types.

**Figure 1.** Classic compression patterns. (**A**) Loop type, (**B**) arachnoid type, black arrow heads: thickened arachnoid membrane, (**C**) perforator type, hollow arrow heads: perforating arteries, (**D**) branch type, (**E**) sandwich type, and (**F**) tandem type. ① Cerebellum, ② brain retractor, ③ brain stem, ④ vestibulocochlear nerve, ⑤ facial nerve, ⑥ primary offending vessel, and ⑦ secondary offending vessel. **Figure 1.** Classic compression patterns. (**A**) Loop type, (**B**) arachnoid type, black arrow heads: thickened arachnoid membrane, (**C**) perforator type, hollow arrow heads: perforating arteries, (**D**) branch type, (**E**) sandwich type, and (**F**) tandem type. <sup>1</sup> Cerebellum, <sup>2</sup> brain retractor, <sup>3</sup> brain stem, <sup>4</sup> vestibulocochlear nerve, <sup>5</sup> facial nerve, <sup>6</sup> primary offending vessel, and <sup>7</sup> secondary offending vessel.

As an addendum to the original "compression patterns", we selected five types as

"challenging patterns", for they were found to pose an additional challenge during the

tandem, three unusual ones were added: cisternal, encircling, and penetrating (Figure 2). The combined number of these five types accounted for 50.4% of instances (1527 of 3028) [7]. Although these five types were named "challenging", the clinical outcome, in terms of improvement of symptoms and postoperative complications, did not differ from that of the overall cases. The success rate of the challenging group was 88.6% whereas that of

offending artery might not be sufficient for a complete decompression; the smaller artery must be released from the pressure by the larger one as well. When there was no contributing factor other than the vascular loop itself, they were categorized as the "loop

type" (Figure 1A). PICA was responsible for 72.7% of the loop types.

deafness (1, 7.7%) [7].

total cases was 90.1%. Likewise, the complication rate of the former was 0.71 while that of the latter was 0.89% [7]. Re-do surgeries (13 out of 1527), however, yielded a lower spasm-free rate (10 of 13, 76.9%) along with a substantially higher incidence of intraoperative BAEP change (7 of 13, 53.8%), postoperative facial palsy (6, 46.2%), and

**Figure 2.** Challenging patterns. (**A**) Perforator type, hollow arrow heads: perforating arteries, (**B**) tandem type, (**C**) cisternal type, (**D**) encircling type, and (**E**) penetrating type. ① Cerebellum, ② brain retractor, ③ brain stem, ④ vestibulocochlear nerve, ⑤ facial nerve, ⑥ primary offending vessel, and ⑦ secondary offending vessel. **Figure 2.** Challenging patterns. (**A**) Perforator type, hollow arrow heads: perforating arteries, (**B**) tandem type, (**C**) cisternal type, (**D**) encircling type, and (**E**) penetrating type. <sup>1</sup> Cerebellum, <sup>2</sup> brain retractor, <sup>3</sup> brain stem, <sup>4</sup> vestibulocochlear nerve, <sup>5</sup> facial nerve, <sup>6</sup> primary offending vessel, and <sup>7</sup> secondary offending vessel.

**4. Discussion**  A bibliometric analysis of hemifacial spasm (HFS) in 2022 reported that the second largest number (3.26%) of all HFS-related articles in the world were published from Sungkyunkwan university in South Korea where Prof. Kwan Park initiated the hemifacial clinic and personally performed over 5000 microvascular decompressions (MVD) for HFS [8]. We hereby present a concise overview with current updates of HFS. As an addendum to the original "compression patterns", we selected five types as "challenging patterns", for they were found to pose an additional challenge during the process of decompression. Besides the two aforementioned types, i.e., the perforator and tandem, three unusual ones were added: cisternal, encircling, and penetrating (Figure 2). The combined number of these five types accounted for 50.4% of instances (1527 of 3028) [7]. Although these five types were named "challenging", the clinical outcome, in terms of improvement of symptoms and postoperative complications, did not differ from that of the overall cases. The success rate of the challenging group was 88.6% whereas that of total cases was 90.1%. Likewise, the complication rate of the former was 0.71 while that of the latter was 0.89% [7]. Re-do surgeries (13 out of 1527), however, yielded a lower spasm-free rate (10 of 13, 76.9%) along with a substantially higher incidence of intraoperative BAEP change (7 of 13, 53.8%), postoperative facial palsy (6, 46.2%), and deafness (1, 7.7%) [7].

#### **4. Discussion**

A bibliometric analysis of hemifacial spasm (HFS) in 2022 reported that the second largest number (3.26%) of all HFS-related articles in the world were published from Sungkyunkwan university in South Korea where Prof. Kwan Park initiated the hemifacial clinic and personally performed over 5000 microvascular decompressions (MVD) for HFS [8]. We hereby present a concise overview with current updates of HFS.

#### *4.1. Overview of HFS*

HFS is defined as contractions on one side of the face. The clinical term, HFS, refers to involuntary facial contractions that are irregular, unilateral, and tonic or clonic. Those twitches usually start with the periorbital muscles and then they can spread to the perinasal, perioral, zygomaticus, and platysma muscles [9]. The diagnosis of HFS is primarily based on clinical history in accordance with the definition of HFS: involuntary facial contractions that are unilateral, irregular, and tonic or clonic. As an adjunctive maneuver, the "other Babinski sign", also known as the Babinski-2 sign, may be useful. It refers to a synchronized contraction of the frontalis muscle or orbicularis oculi muscle, induced by an attempt to lift up one's eyebrow while the eye is being closed [10]. This maneuver assists in the diagnosis of HFS with a sensitivity of 86% and a specificity of 100% [10]. Electromyography (EMG), computed tomography (CT), or magnetic resonance imaging (MRI) also can be adopted to confirm the diagnosis. Time of flight (TOF) of an MR angiography may delineate the proximity or contact of an offending vessel with the REZ. According to more recent studies where 3D MRI volumetric analysis was applied to evaluate the size of the CSF space in the posterior fossa, it appeared to be smaller in HFS patients compared to that of the control group [11]. The characteristic feature of EMG in HFS can be described as spontaneous and high-frequency synchronized firing, which may be helpful to differentiate HFS from other movement disorders, such as myokymia, blepharospasm, craniocervical dystonia (Meige syndrome), post-facial palsy synkinesia, tic disorders, myokymia, neuromyotonia, and tardive dyskinesias (TD) as well as phychogenic HFS [9]. According to an epidemiological study based on a Norwegian population, the prevalence of HFS was about 9.8 per 100,000 persons [12]. Another study from the USA reported the prevalence rate of HFS as 7.4 per 100,000 men and 14.5 per 100,000 women [13]. Data from our own institute revealed the male-to-female ratio to be 1:2.28 with an average age of 52.2 years [14].

The pathophysiology of HFS, as widely accepted, is explained by vascular compression on the root entry zone (REZ) of the facial nerve. When the compression of the REZ is the sole cause of HFS, it is defined as primary HFS, whereas any impairment of the facial nerve due to a pre-existing condition can constitute a secondary HFS. A modernday concept of vascular compression syndrome that included trigeminal neuralgia, HFS, and glossopharyngeal neuralgia was introduced by McKenzie in 1936 [9]. Based on its pathophysiological background, vascular decompression for HFS was first introduced by Gardner in 1962, following which, a more modern technique with a minimal approach, i.e., MVD via retrosigmoid craniotomy, was first performed by Bremond in 1974 [9,15,16]. The current concept of the pathophysiology and surgical treatment of HFS was established and popularized by Jannetta, and it started with his article in 1975, titled "*Neurovascular cross-compression in patients with hyperactive dysfunction symptoms of the eighth cranial nerve*" [9,17]. When a vascular curvature causes the compression on the REZ, the anterior inferior cerebellar artery (AICA) is most commonly involved, followed by posterior inferior cerebellar artery (PICA), and the vertebral artery (VA). A single artery could be the sole cause of the neurovascular compression, but it is rather infrequent (4.7%) according to our previous report [9,18]. In consideration of other additional factors, a total of six compressive patterns in HFS were proposed: loop, arachnoid, perforator, branch, sandwich, and tandem types [9,18]. Regarding a more detailed mechanism of HFS in addition to the microscopic disruption of myelin in the REZ, there are two major hypotheses: the central (hyperexcitability of the facial motor nucleus) vs the peripheral (ephaptic transmission between the facial nerve bundles) hypothesis [9,19]. An increasing amount of micro-anatomical and

neurophysiological research is dedicated to elucidating the precise pathway of HFS; but one hypothesis cannot explain all the phenomena without the other.

Pharmaceutical medicine in general has failed to provide long-term improvement for HFS. Anticonvulsants or GABAergic medicines may lessen symptoms partially and temporarily, but the effectiveness of those medicines cannot be comparable to botulinum neurotoxin (BTX) injection, not to mention microvascular decompression. BTX injection is the most preferred non-surgical treatment for HFS, yielding up to 85% of symptomatic relief, and among the seven serotypes of BTX, serotypes A and B are currently commercialized [9]. Following injections, symptomatic improvement occurs in 1–3 days, and it usually reaches its peak effect in 5 days [9,20]. The duration of clinical benefit varies center to center by 3–6 months [9,21,22]. Repeated injections of BTX are unavoidable, and tolerance can naturally develop in some subjects, although a 10-year multicenter study reported that the average duration of improvement did not change from the first year of injection to the 10th year of treatment with a similar dose of BTX [9,23]. Additionally, they insisted that the adverse responses derived from BTX injections decreased throughout the 10-year course. Local complications of BTX injection include ptosis, blurred vision, and diplopia, but they are rarely permanent [9,24]. Incidence of overall adverse effects was estimated as ranging from 20 to 53%, and the most frequent one was ptosis [21,22,25]. Despite its relatively high success rate of symptomatic improvement, one cannot ignore the fact that BTX injection fundamentally requires repeated sessions, which leads to emotionally and financially non-negligible burdens on the patients [9].

MVD is the only curative treatment option for HFS with a high success rate and low incidence of recurrence and complications. According to a systemic review on 22 studies with 5700 patients who underwent MVD, a complete resolution was achieved in 91.1% (95% CI: 90.3–91.8%) of patients [9,26]. Recurrence occurred in 2.4% (95% CI: 1.9–2.9%) of patients, and postoperative complications included transient complications including facial palsy (9.5% [95% CI:8.8–10.3%]), hearing deficit (3.2% [95% CI: 2.7–3.7%]), and cerebrospinal fluid leak (1.4% [95% CI: 1.1–1.7%]) [9,26]. Permanent complications included hearing deficit in 2.3% (95% CI: 1.9–2.7%) and facial palsy in 0.9% (95% CI: 0.7–1.2%) of patients, the risk of stroke was 1 in 1800, and risk of death was 1 in 5500 [9,26].

The basic concept of MVD is well described in the literature, but the detailed techniques vary depending on institutions and surgeons. Once a lateral retro-sigmoid suboccipital craniectomy or craniotomy is performed under a general anesthesia, the dura is incised to reveal the cerebellar cortex. With or without traction of the flocculus, the root entry zone (REZ) of the facial nerve is to be observed. Upon the identification of the compressing vessels, or the offending arteries, they are separated from the seventh nerve, which then can be perpetuated by insertion of Teflon pieces. A few more additional techniques, including transposition of the vessels, snare technique, vascular sling, etc., have been proposed [9,27–29]. Intraoperative EMG monitoring can be beneficial for improvement of surgical outcomes. Lateral spread response (LSR) is one of the most commonly employed neurophysiologic tests for HFS since Moller and Jannetta suggested that properly performed decompression would be accompanied with the disappearance of the LSR [9,30]. However, persistence of the LSR did not necessarily indicate a poor outcome, which precludes the LSR from being a reliable predictor for the long-term prognosis of HFS after MVD [9,31]. Furthermore, to properly monitor the integrity of the eighth nerve (CN VIII) during MVD, intraoperative brain stem auditory evoked potential (BAEP) can be employed, which has been accepted by numerous institutions in decreasing the risk of hearing impairment during MVD [9].

Clinical courses following MVD are not identical. According to our own report, 737 (92.8%) of 807 patients who had undergone MVD for HFS became absolutely or nearly spasm free by the 2-year postoperative follow-up [9,18]. However, not everyone became asymptomatic immediately after the surgery; 140 (19.0%) of 737 patients still experienced residual spasms for more than a month, and some of them lasted more than a year [9,18]. These inhomogeneous courses of MVD for HFS may indicate that microscopic

changes in the REZ, facial nerve, or facial nucleus in each patient can be diverse; some may have reversible compression without any structural changes, whereas others may have gone through microscopic changes in their facial nerve or nucleus. We believe that one cannot easily conclude the two aforementioned hypotheses of pathophysiology, i.e., hyperexcitability of the facial nucleus vs. ephaptic transmission of the facial nerve, are mutually exclusive.

#### *4.2. Compressive Patterns*

Since we introduced six compression patterns of HFS which included the loop, arachnoid, perforator, branch, sandwich, and tandem in 2008, we have received many questions concerning the significance of this categorization. As reported, one type was not necessarily associated with a better outcome compared to another, which indicated that a specific compression pattern could not determine indications for MVD. In 2020, we selected five types that can be technically challenging during MVD, but these also did not necessarily contribute to poorer results [7]. The only cases that significantly resulted in unfavorable outcomes are re-do surgeries, which underscores the importance of an accurate diagnosis, proper determination of surgical indication, thorough exploration around the REZ during MVD, and avoidance of iatrogenic compression on the REZ [7]. The significance of our categorization system can be rephrased as knowledge on these patterns may help in achieving a better surgical outcome. By being aware that there could be more than one artery causing the compression, as in the sandwich or tandem type, an incomplete decompression can be avoided. Additionally, when the vessel on the REZ is not easily movable, one must consider the possibility of the perforator type to prevent a devastating complication such as brain stem infarction or hemorrhage. Although each type did not directly impact the postoperative results, the understanding of these patterns seemed to improve the overall outcome in our institution.

A brief summary of the features and some useful tips for surgery are presented in Table 1. In the arachnoid type, the offending vessel may no longer be located on the REZ after the arachnoid dissection, which can coincide with the disappearance of the LSR. When the preoperative MRI delineates the VA near the REZ, the tandem type should be the first one to rule out. If the tandem type is confirmed during the surgery, one should be reminded of the fact that the simple insertion of Teflon pieces between the REZ and the smaller artery may not guarantee a satisfactory outcome, because the larger artery, most likely the VA, would probably continue to derive throbbing forces to the smaller one. The larger artery must be detached from the smaller one, and furthermore, it would be ideal if the larger one could be transposed so that the trajectory of the throbbing force can be re-directed. The loop type is the least challenging since the approach to the REZ is not hindered by any obstacles and there is enough working space for decompression. The involvement of PICA is most frequent in this type. The perforator type is one of the challenging ones. Owing to the tight perforators, manipulation of the compressing artery can be highly difficult and sometimes potentially dangerous. Any disruption to one or more of the perforators can result in permanent impairment to the brain stem. When the REZ is caught between the branches, the main trunk must be moved off the REZ, so that the inserted Teflon does not cause an iatrogenic compression. The sandwich type can be overlooked if the medial side of the REZ is not thoroughly inspected. After one compression on the dorsal side of the REZ is successfully decompressed, the medial side also should be carefully observed.

The arachnoid type was most frequently observed (27.9%), followed by the tandem (24.6%) and the perforator (22.0%) types based on our own research in 2008 [6]. The addendum classification was published in 2021, where 1527 (50.4%) of 3027 MVD for HFS cases were selected as such [7]. Among the challenging types, the tandem (40.2%) and the perforator (31.1%) accounted for the majority, and the remaining three types included cisternal, encircling, and penetrating ones in order of frequency; the penetrating type was the most extreme one (4 out of 3027, 0.13%). In the cisternal type, it is sometimes difficult to find the compression, because the usual compression site, i.e., the REZ, appears to be compression-free. Exploration towards the cisternal portion of the facial nerve often demands further retraction of the cerebellum, but special attention must be paid not to retract it excessively. BAEP monitoring is mandatory. When the encircling artery is found around the REZ, the decompression process should be carried out from the medial to the lateral side of the REZ, because the insertion of Teflon pieces on the lateral side may hinder the medial side from being properly maneuvered. The penetrating type is the most rare and probably the most challenging of all. These challenging types did not necessarily lead to poorer outcomes, whereas revision cases resulted in a significantly lower spasm-free rate (10 of 13, 76.9%) along with a substantially higher incidence of intraoperative BAEP change (7 of 13, 53.8%), postoperative facial palsy (6, 46.2%), and deafness (1, 7.7%) [7].


**Table 1.** Features and surgical tips for each type.

¶ : Posterior inferior cerebellar artery, <sup>β</sup>: vertebral artery, <sup>δ</sup> : root exit zone, \*: lateral spread response.

Veins were found to be responsible for HFS in 6 (1.1%) out of 528 HFS patients, according to our previous study; 2 of them were solely caused by venous compression, whereas the remaining 4 cases were due to a combination of arterial and venous compression [32]. We did not cauterize the vein as it may result in venous infarction or brain stem injury. After careful manipulation of the vein, the REZ was decompressed using a small piece of Teflon. Although we do not have statistically significant data, HFS due to venous compression appeared to be associated with rather unfavorable outcomes because it was not always possible to detach the vein from the REZ and manipulation of the facial nerve was sometimes inevitable.

The importance of a thorough 360◦ inspection around the REZ cannot be overemphasized. Even after one offending artery is successfully detached and insulated, the medial and cisternal sides of the REZ should also be free from any compression. The LSR tends to disappear when the REZ is no longer compressed, but it is questionable if the disappearance of the LSR can be a predictor of the long-term prognosis [18,24]. If the LSR still exists after a decompression process, it could indicate either incomplete decompression with a secondary cause of compression untreated, or complete decompression with lingering hyperexcitability of the facial nucleus or ephaptic transmission between nerve fibers [2,14,33]. Since it is not possible to discriminate the latter from the former

during a surgery, a hypothetical secondary cause should always be ruled out to achieve a long-term cure.

#### *4.3. Brief Summary of Current Updates*

A recent randomized clinical trial (RCT) on BTX injections for HFS revealed that bilateral injection of BTX decreased facial asymmetry more than ipsilateral injection did [34]. In a related study, MR tractography findings were evaluated in HFS patients following injection of BTX, where apparent diffusion coefficient (ADC) and fractional anisotropy (FA) values in the contralateral motor cortex were found to be close to those of the pathological side [35]. The authors suggested that this result might indicate an impact of peripherally injected BTX on the central nervous system [35]. Another double blinded, RCT demonstrated that pretarsal injection of BTX was more efficient than preseptal injection in terms of better symptom control and longer duration of efficacy [36].

Surgical techniques have evolved over time as well. Since the concept of MVD initiated by Jannetta et al. has been employed for HFS, trigeminal neuralgia, and glossopharyngeal neuralgia, newer techniques have emerged. The "transposition technique" differs from the simple insertion of Teflon pieces as it aims to alter the location and trajectory of the offending vessel so that the vessel can no longer convey the pulsating force to the REZ [37]. This technique can be roughly rephrased as "off the REZ", since the REZ is free not only from the offending vessel but also from any iatrogenic Teflon pieces [37]. MVD using this transposition technique, with or without the help of a fibrin coated sling, demonstrated a higher success and lower recurrence rate [37]. As we emphasized, the re-do surgeries resulted in significantly poorer outcomes [7]. During a revision MVD, previously inserted Teflon pieces were often found near the REZ, and they were thought to have accounted for the residual or recurrent symptoms. Iatrogenic compression must be avoided at all costs, and we believe that this "off the REZ" policy may be the key to prevent a recurrence, and accordingly, a re-do intervention.

Endoscope-assisted MVD and fully endoscopic surgery have gained increasing popularity. A meta-analysis comparing the traditional and endoscopic MVD, with a total of 12 studies and 1122 patients, reported that the endoscopic MVD yielded a higher success rate (97% vs. 89%), lower recurrence rate (5.7% vs. 0.3%), as well as a lower complication rate (12% vs. 27%) than the microscopic MVD did [38]. Another study in 2019 also insisted that a fully endoscopic MVD is both safe and feasible in the treatment of HFS since it can provide a better visualization of the neurovascular conflict, despite its original shortcomings, e.g., being prone to blood soiling, lacking 3D information, or having a longer learning curve [39].

The disappearance of the LSR is still a useful parameter during MVD, either traditional or endoscopic, as the majority of researchers concur. A meta-analysis on intraoperative monitoring of the LSR reported that an intraoperative disappearance of the LSR could predict a favorable clinical outcome with a high specificity of 90% at discharge and after 1 year, whereas the sensitivity was only 40% at discharge and after 1 year [40]. We believe this lower sensitivity of the LSR might be derived from the hyperexcitability of the facial motor nucleus in HFS, which could be sustained even after a successful decompression. A neurophysiological study using a novel parameter in the future may distinguish the hyperexcitability of the facial motor nucleus from the ephaptic transmission.

#### **5. Conclusions**

We believe the categorization of compressing patterns on the REZ of HFS patients as well as insightful technical tips in accordance with each individual pattern, may contribute to a safer and more efficient MVD for HFS. The golden rules of successful MVD for HFS are accurate diagnosis, proper indication for surgery, thorough and careful exploration around the REZ, and avoidance of iatrogenic compression on the REZ.

**Author Contributions:** Conceptualization, K.P.; Methodology, J.S.P.; Resources, J.S.P.; Data curation, J.S.P.; Writing—original draft, J.S.P.; Writing—review & editing, K.P.; Visualization, J.S.P.; Supervision, K.P. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Institutional review board of Samsung medical center: SMC (2020-04-008).

**Informed Consent Statement:** Patient consent was not necessary because of the retrospective nature of the study, and the validity of the findings would not be affected by the absence of patient consent. Moreover, no additional risk to patient safety was expected from this study without patient consent.

**Data Availability Statement:** Data available on request due to restrictions eg privacy or ethical.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

## *Review* **Lateral Spread Response: Unveiling the Smoking Gun for Cured Hemifacial Spasm**

**Kyung Rae Cho <sup>1</sup> , Sang Ku Park <sup>1</sup> and Kwan Park 1,2,\***


**\*** Correspondence: kwanpark@skku.edu

**Abstract:** Hemifacial spasm (HFS) is a rare disorder characterized by involuntary facial muscle contractions. The primary cause is mechanical compression of the facial nerve by nearby structures. Lateral spread response (LSR) is an abnormal muscle response observed during electromyogram (EMG) testing and is associated with HFS. Intraoperative monitoring of LSR is crucial during surgery to confirm successful decompression. Proper anesthesia and electrode positioning are important for accurate LSR monitoring. Stimulation parameters should be carefully adjusted to avoid artifacts. The disappearance of LSR during surgery is associated with short-term outcomes, but its persistence does not necessarily indicate poor long-term outcomes. LSR monitoring has both positive and negative prognostic value, and its predictive ability varies across studies. Early disappearance of LSR can occur before decompression and may indicate better clinical outcomes. Further research is needed to fully understand the implications of LSR monitoring in HFS surgery.

**Keywords:** lateral spread response; abnormal muscle response; hemifacial spasm

life13091825

**Citation:** Cho, K.R.; Park, S.K.; Park, K. Lateral Spread Response: Unveiling the Smoking Gun for Cured Hemifacial Spasm. *Life* **2023**, *13*, 1825. https://doi.org/10.3390/

Academic Editors: Zuzanna Nowak and Aleksandra Nitecka-Buchta

Received: 29 June 2023 Revised: 11 August 2023 Accepted: 26 August 2023 Published: 29 August 2023

**Copyright:** © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

#### **1. Introduction**

Hemifacial spasm (HFS) is a rare neuromuscular disorder characterized by involuntary contractions of the facial muscles, predominantly affecting one side of the face. The underlying pathophysiology of HFS involves mechanical compression of the facial nerve at the root exit zone (REZ) by adjacent structures, including arteries, veins, and tumors. Identification of the offending vessel can be achieved through magnetic resonance imaging and confirmed during surgical intervention. However, in some cases, there may be multiple compressing structures, or they may elude detection during the surgical procedure. Therefore, it is crucial for surgeons to accurately identify the primary cause when manipulating the surface of the brainstem. Intraoperative monitoring can be valuable in confirming successful decompression while minimizing the risk of unnecessary manipulation near delicate structures.

Lateral spread response (LSR), an abnormal muscle response (AMR), represents a distinctive neurophysiological characteristic of HFS that remains undetectable under normal conditions. When one branch of the facial nerve is stimulated, an atypical response is observed on the electromyogram (EMG) in other branches of the facial nerve. This anomalous response was initially reported by Janetta and Moller [1,2], and subsequent studies have been conducted to elucidate its significance and its connection to the pathophysiology of HFS. Although LSR holds substantial importance during microvascular decompression surgery (MVD) for HFS, its precise implications are not yet fully understood. The presence of LSR indicates the existence of aberrant cross-connections between facial nerve branches or fibers, although the exact nature of these abnormal connections remains to be identified [3].

The objectives of this article are to provide a comprehensive review of the pathophysiology of LSR, the appropriate techniques for its accurate monitoring, and its clinical

implications, and to address the controversies that have been previously discussed in the literature.

#### **2. Methods for Monitoring Lateral Spread Response (LSR)**

LSR monitoring is typically conducted throughout the entire surgical procedure, from the initiation of general anesthesia until its conclusion. The monitoring of LSR does not interfere with the surgical process, allowing for continuous observation while manipulating the facial nerve and adjacent vessels. Generally, monitoring of LSR begins after the insertion of electrodes, before dural opening, after dural opening, during REZ decompression, and after dural closure [4]. Even if LSR disappears immediately after dural opening or decompression, its recurrence may indicate unsuccessful decompression. Hence, continuous monitoring of LSR is recommended throughout the surgery, even if it disappears initially.

#### *2.1. Anesthesia*

Since LSR is an abnormal electromyogram (EMG), appropriate anesthetic agents must be used to avoid interfering with the accurate monitoring of this abnormal muscle response. During the induction of general anesthesia, anesthesiologists typically administer neuromuscular blockade (NMB) agents such as rocuronium or vecuronium to minimize patient stress during endotracheal intubation [5]. However, continuous administration of these agents can affect the precise detection of LSR. Therefore, careful titration of NMB is necessary, and many anesthesiologists prefer to delay its administration until the end of the surgery [6]. It is generally recommended that a train-of-four count of more than two be maintained, for accurate monitoring. However, maintaining partial NMB with a target T1/Tc ratio of 50% has proven to be clinically acceptable for LSR monitoring and surgical conditions during MVD [7]. Complete termination of NMB did not significantly enhance LSR monitoring when compared to maintaining a T1/Tc ratio of 50%; thus, it is not recommended in MVD surgery [8].

Inhalational anesthetics have the potential to inhibit or block LSR [9,10]. Studies have also demonstrated significant alterations in the chronaxie of human corticospinal axons when exposed to the inhalational anesthetic sevoflurane [10]. There are reports indicating that desflurane can suppress LSR amplitude by 43%, compared to total intravenous anesthesia alone [9].

#### *2.2. Electrode Position*

The accurate positioning of electrodes significantly affects the results of neuromonitoring. Misplaced electrodes may stimulate unintended structures, leading to monitoring artifacts instead of capturing meaningful waves. Therefore, ensuring correct electrode positioning is crucial for the precise monitoring of LSR. Typically, the stimulation electrode is placed between the ipsilateral tragus and the external canthus of the eye, where the zygomatic branches of the facial nerve are located. Since the zygomatic branch of the facial nerve innervates the orbicularis oculi muscle, the abnormal muscle response recorded in other facial muscles, such as the frontalis (temporal), orbicularis oris (buccal), and mentalis (marginal mandibular), is defined as LSR [11]. Some institutions have also explored stimulating the marginal mandibular branch, which is located at the border of the mandible and lateral to the mental tubercle, and recording from the orbicularis oculi muscle and the mentalis muscle [3,5].

The conventional stimulation method involves placing paired dermal electrodes in such a way that the cathode is positioned at the proximal branch and the anode is positioned at the distal branch, resulting in centripetal impulses toward the brainstem. However, Lee et al. [12] conducted a study in which they inverted this method by placing the cathode at the distal branch and the anode at the proximal facial nerve. They found that the innervated muscles responded more sensitively to this new stimulation method. In fact, the new method with the cathode at the distal branch demonstrated a higher detection

rate for the disappearance of LSR than that of the conventional method. The conventional method achieved a detection rate of 61.8%, while the new method achieved a detection rate of 98.2%. Additionally, after surgical decompression, the conventional method still showed a remaining LSR rate of 29.1%, while the new method had a remaining LSR rate of only 1.8% (Figure 1). tion rate for the disappearance of LSR than that of the conventional method. The conventional method achieved a detection rate of 61.8%, while the new method achieved a detection rate of 98.2%. Additionally, after surgical decompression, the conventional method still showed a remaining LSR rate of 29.1%, while the new method had a remaining LSR rate of only 1.8% (Figure 1).

cathode at the distal branch and the anode at the proximal facial nerve. They found that the innervated muscles responded more sensitively to this new stimulation method. In fact, the new method with the cathode at the distal branch demonstrated a higher detec-

*Life* **2023**, *13*, x FOR PEER REVIEW 3 of 12

**Figure 1.** A paired dermal stimulation electrode was positioned either at the zygomatic branch or at the marginal mandibular branch to induce stimulation. Typically, in conventional practice, the cathode is placed proximally while the anode is placed distally to elicit a centripetal impulse. In contrast, Lee et al. conducted a study in which they reversed the placement of the cathode and the anode. Interestingly, this alternative configuration resulted in more sensitive recording of lateral spread response. Their findings suggested that the directionality of the electrode placement can significantly impact the quality and precision of LSR recordings. **Figure 1.** A paired dermal stimulation electrode was positioned either at the zygomatic branch or at the marginal mandibular branch to induce stimulation. Typically, in conventional practice, the cathode is placed proximally while the anode is placed distally to elicit a centripetal impulse. In contrast, Lee et al. conducted a study in which they reversed the placement of the cathode and the anode. Interestingly, this alternative configuration resulted in more sensitive recording of lateral spread response. Their findings suggested that the directionality of the electrode placement can significantly impact the quality and precision of LSR recordings.

#### *2.3. Stimulation Parameters 2.3. Stimulation Parameters*

Currently, there is no standardized guideline for the stimulation and recording methods used in LSR monitoring. Due to variations in nerve excitability thresholds among individuals and the potential influences of anesthesia and other conditions, establishing specific parameters for stimulation presents challenges. However, in most studies, a pulse wave with a duration of 0.2–0.3 ms and an intensity ranging from 5 to 25 mA have been commonly employed. Within this intensity range, LSR can be consistently detected [11]. Currently, there is no standardized guideline for the stimulation and recording methods used in LSR monitoring. Due to variations in nerve excitability thresholds among individuals and the potential influences of anesthesia and other conditions, establishing specific parameters for stimulation presents challenges. However, in most studies, a pulse wave with a duration of 0.2–0.3 ms and an intensity ranging from 5 to 25 mA have been commonly employed. Within this intensity range, LSR can be consistently detected [11].

Nevertheless, although the results have not yet been published, the authors of this review paper discovered the presence of superficially spreading artifacts that can be Nevertheless, although the results have not yet been published, the authors of this review paper discovered the presence of superficially spreading artifacts that can be mistaken for LSR during high-intensity stimulation. In cases where LSR disappears early or goes undetected, examiners may be inclined to increase the stimulation intensity to reveal any hidden LSR. However, the authors' study revealed the presence of artifacts that

mimic LSR, particularly when the abnormal muscle response appears with a very short latency—specifically, less than 10 ms. Therefore, it is advisable that the stimulation intensity not be increased in such cases, where the abnormal muscle response seems to appear too quickly, as it may be due to these artifact responses (Figure 2). reveal any hidden LSR. However, the authors' study revealed the presence of artifacts that mimic LSR, particularly when the abnormal muscle response appears with a very short latency—specifically, less than 10 ms. Therefore, it is advisable that the stimulation intensity not be increased in such cases, where the abnormal muscle response seems to appear too quickly, as it may be due to these artifact responses (Figure 2).

mistaken for LSR during high-intensity stimulation. In cases where LSR disappears early or goes undetected, examiners may be inclined to increase the stimulation intensity to

*Life* **2023**, *13*, x FOR PEER REVIEW 4 of 12

**Figure 2.** The study conducted by the authors revealed that when the intensity of stimulation surpassed its upper threshold, a direct spreading artifact resembling the low-threshold sensory afferent response (LSR) was observed that did not disappear, even after decompression. This exceeding intensity was quantified by a reduction in the latency of the abnormal muscle response, which was shorter than 10 milliseconds according to their findings. These results suggest that care must be taken to differentiate true LSR from artifacts caused by excessive stimulation intensity, as these artifacts can lead to misleading interpretations and, potentially, affect the accuracy of experimental outcomes. **Figure 2.** The study conducted by the authors revealed that when the intensity of stimulation surpassed its upper threshold, a direct spreading artifact resembling the low-threshold sensory afferent response (LSR) was observed that did not disappear, even after decompression. This exceeding intensity was quantified by a reduction in the latency of the abnormal muscle response, which was shorter than 10 milliseconds according to their findings. These results suggest that care must be taken to differentiate true LSR from artifacts caused by excessive stimulation intensity, as these artifacts can lead to misleading interpretations and, potentially, affect the accuracy of experimental outcomes.

#### **3. Prognostic Value of LSR 3. Prognostic Value of LSR**

Zhang et al. conducted a meta-analysis of 14 papers that investigated the prognostic value of lateral spread response (LSR) during microvascular decompression (MVD) [13]. Their findings indicated that the disappearance of LSR is highly associated with shortterm outcomes. However, they did not find a significant predictive effect on long-term outcomes. Another systematic review, by Nugroho [4], also concluded that short-term outcomes are strongly correlated with the resolution of LSR. However, the resolution of Zhang et al. conducted a meta-analysis of 14 papers that investigated the prognostic value of lateral spread response (LSR) during microvascular decompression (MVD) [13]. Their findings indicated that the disappearance of LSR is highly associated with short-term outcomes. However, they did not find a significant predictive effect on long-term outcomes. Another systematic review, by Nugroho [4], also concluded that short-term outcomes are strongly correlated with the resolution of LSR. However, the resolution of LSR does not significantly impact long-term outcomes, as patient outcomes tend to improve over time with adequate decompression, even if LSR persists after surgery.

In contrast, a meta-analysis performed by Thirumala et al. [14] revealed that intraoperative LSR monitoring demonstrates high specificity but low sensitivity in predicting a postoperative hemifacial spasm (HFS)-free status at discharge, at 3 months after discharge, and at 1 year after discharge. According to their analysis, the sensitivity was calculated as 40%, 41%, and 40%, respectively, while the specificity was estimated as 89%, 90%, and 89%, respectively, at discharge, 3 months after discharge, and 1 year after discharge. They further calculated the negative predictive value, which indicates the probability of patients achieving LSR resolution, as 92.7%, 95.8%, and 96.0%, respectively, at discharge, 3 months after discharge, and 1 year after discharge. Additionally, the positive predictive value, representing cases where LSR persists, was determined as 47.8%, 40.8%, and 24.4%, respectively, at discharge, 3 months after discharge, and 1 year after discharge. These results suggest that both short-term and long-term outcomes can be predicted based on the resolution of LSR during surgery. However, it should be noted that even if LSR persists during surgery, long-term outcomes may still be positive.

#### *3.1. Positive Prognostic Value*

Lee et al. [12] reported that the AMR monitoring during MVD is beneficial for identifying the offending vessel and suggesting the most appropriate surgical endpoint. Kong et al. [15] reported that the monitoring of AMR is an effective tool when performing complete decompression, and it may help to predict the outcomes. Some patients still had residual spasm despite LSR disappearance. Various and complicated findings of the offending vessels, as stated in that report, may be the cause of spasm persistence. However, in follow-up visits at 1 year, the number of patients that were included in the category of HFS-free status increased remarkably, so that the correlation between LSR and outcome became significant. Nevertheless, divergent views on this issue have always existed.

Sekula et al. reported that the likelihood of achieving a cure is 4.2 times higher if LSR disappears during surgery than when it persists. However, it is important to note that that meta-analysis only evaluated the utility of LSR at the final follow-up visit and did not consider the postoperative measurements taken two days after surgery [16], Furthermore, there is a consensus among studies that there is a positive relationship between the resolution of LSR and the clinical outcome of HFS. Concerns regarding this relationship arise when there is no LSR observed or when there is early disappearance of LSR. Additionally, questions arise when LSR persists even after successful decompression. In cases where no LSR is observed, or where there is early disappearance, the predictive value becomes uncertain. It becomes challenging to determine the prognosis and the clinical outcome without the presence of LSR as a reliable indicator. Further research is needed to understand the implications and significance of these scenarios.

Similarly, when LSR persists despite successful decompression, the correlation between LSR and clinical outcomes becomes less straightforward. The persistence of LSR may indicate the presence of additional contributing factors or complexities that influence the overall outcome. These cases highlight the multifactorial nature of HFS and the need for a comprehensive assessment of various clinical factors to determine the prognosis and treatment outcomes accurately. However, the resolution of LSR followed by successful decompression of the vessel compressing the facial nerve REZ suggests a positive clinical outcome (Figure 3).

**Figure 3.** Captured image from intraoperative neuromonitoring program. The offending vessel was dissected from the facial nerve root exit zone, but teflon felt was not yet inserted. As a result, lateral spread response (LSR) is still seen (white oval). After placing teflon felt between the facial nerve and the dissected vessel, LSR disappeared (white rectangle). **Figure 3.** Captured image from intraoperative neuromonitoring program. The offending vessel was dissected from the facial nerve root exit zone, but teflon felt was not yet inserted. As a result, lateral spread response (LSR) is still seen (white oval). After placing teflon felt between the facial nerve and the dissected vessel, LSR disappeared (white rectangle).

#### *3.2. Negative Prognostic Value 3.2. Negative Prognostic Value*

While most studies concur that the disappearance of lateral spread response (LSR) is associated with favorable outcomes in hemifacial spasm (HFS), there are some studies that demonstrate a lack of correlation between LSR resolution and clinical outcomes. This discrepancy may be attributed to intraoperative findings of multiple vessel compressions, as well as to the presence of vessels that are not easily visible behind the facial nerve or vessels coursing around the root exit zone without compressing the nerve. These factors can contribute to residual spasms following microvascular decompression (MVD) [15]. In a study by Wei et al. [17], the efficacy of intraoperative auditory brainstem re-While most studies concur that the disappearance of lateral spread response (LSR) is associated with favorable outcomes in hemifacial spasm (HFS), there are some studies that demonstrate a lack of correlation between LSR resolution and clinical outcomes. This discrepancy may be attributed to intraoperative findings of multiple vessel compressions, as well as to the presence of vessels that are not easily visible behind the facial nerve or vessels coursing around the root exit zone without compressing the nerve. These factors can contribute to residual spasms following microvascular decompression (MVD) [15].

sponse (AMR) monitoring in improving the outcomes of MVD for HFS was evaluated. However, the findings indicated that intraoperative AMR monitoring did not significantly enhance the efficacy of MVD for HFS, when performed by skilled surgeons. Notably, studies by Kiya et al. [18] and Yamashita et al. [19] reported a high propor-In a study by Wei et al. [17], the efficacy of intraoperative auditory brainstem response (AMR) monitoring in improving the outcomes of MVD for HFS was evaluated. However, the findings indicated that intraoperative AMR monitoring did not significantly enhance the efficacy of MVD for HFS, when performed by skilled surgeons.

tion of patients who exhibited LSR persistence but were free from HFS symptoms. This finding resulted in an insignificant correlation between LSR resolution and HFS relief. It should be acknowledged that these studies may have been limited by small sample sizes. Kiya et al.'s study lacked remaining spasms in both the LSR-disappearance and the persistence groups, rendering their 3 month follow-up analysis inconclusive. Similarly, Yamashita et al. found no significant correlation in the 1 year evaluation. Indeed, there are studies that have reached different conclusions regarding the pre-Notably, studies by Kiya et al. [18] and Yamashita et al. [19] reported a high proportion of patients who exhibited LSR persistence but were free from HFS symptoms. This finding resulted in an insignificant correlation between LSR resolution and HFS relief. It should be acknowledged that these studies may have been limited by small sample sizes. Kiya et al.'s study lacked remaining spasms in both the LSR-disappearance and the persistence groups, rendering their 3 month follow-up analysis inconclusive. Similarly, Yamashita et al. found no significant correlation in the 1 year evaluation.

dictive value of intraoperative lateral spread response (LSR) monitoring in the outcomes Indeed, there are studies that have reached different conclusions regarding the predictive value of intraoperative lateral spread response (LSR) monitoring in the outcomes of microvascular decompression (MVD) [20]. One such study, by El Damaty et al. [21], prospectively analyzed 100 patients with hemifacial spasm (HFS) and found that while LSR could guide the appropriate decompression of the facial nerve during MVD, it did not serve as a reliable predictor of postoperative efficacy. Similarly, Hatem et al. [22] observed that all 10 patients in their study achieved clinical cures despite the persistence of LSR during MVD. That finding raised doubts about the practical usefulness of LSR in the context of MVD.

These studies indicated that there is conflicting evidence regarding the predictive value of intraoperative LSR monitoring in MVD outcomes, highlighting the need for further research and consideration of multiple factors in surgical decision-making [3].

In a review conducted by Neves, the abolition of lateral spread response (LSR) and its correlation with clinical outcomes were examined in a group of 32 patients. The study reported a sensitivity of 100% and specificity of 94% in predicting long-term outcomes based on LSR abolition. However, there was no observed relationship between intraoperative LSR changes and relief from hemifacial spasm (HFS) on the first day after surgery.

These findings suggested that LSR abolition may serve as a reliable predictor of long-term outcomes in HFS patients. Additionally, the presence of increased temporal dispersion in the direct response at the stimulated nerve branch could provide valuable insights regarding LSR status, especially in patients with a history of botulinum toxin treatment [23].

Additionally, the clinical course of hemifacial spasm (HFS) after microvascular decompression (MVD) is characterized by high variability. Although many patients experience immediate relief from spasms following MVD, there are instances in which facial spasms continue to persist for several months or even years after surgery, despite the disappearance of LSR. This variability underscores the complex nature of HFS and indicates that factors other than LSR status contribute to the persistence or recurrence of symptoms [24].

#### *3.3. Early Disappearance of LSR*

Furthermore, it has been observed that the disappearance of lateral spread response (LSR) can occur early in the surgical procedure, even before any vascular decompression of the facial nerve takes place. This early LSR disappearance can be attributed to various factors, such as changes in the dynamics of cerebrospinal fluid (CSF) upon dural opening, CSF drainage, or minimal cerebellar retraction [25]. In some cases, there may be a transient disappearance of LSR followed by its reappearance at a later stage [5,26].

Jiang et al. [5] conducted a retrospective review of 372 patients. Among them, 33 patients exhibited early disappearance of LSR. The study found that the injection of muscle relaxants could diminish LSR and, importantly, that early disappearance of LSR was associated with better clinical outcomes. Several studies have explored the mechanism behind early loss of LSR before decompression in HFS surgery [5,26,27]. This early loss of LSR suggests that the compression force exerted by the offending blood vessels is relatively mild and can be easily influenced by subtle environmental changes, such as CSF egress. Kim et al. [28] suggested that the disappearance of LSR during dural opening or after CSF drainage, prior to decompression, was correlated with poorer outcomes. They emphasized the importance of surgeons carefully identifying the exact offending vessels in order to optimize surgical outcomes. These findings highlight the dynamic nature of LSR changes during HFS surgery and the potential significance of early LSR disappearance as a predictor of surgical outcomes. Surgeons should be attentive to the timing and patterns of LSR changes to improve identification of the responsible vessels and to optimize treatment strategies.

Figure 4 presents a flow chart that outlines the surgical decision-making process with LSR monitoring. This flow chart provides a visual representation of the sequential steps involved in making informed surgical interventions based on the observed LSR patterns.

**Figure 4.** Flowchart for surgical decision-making based on lateral spread response (LSR) monitoring. If LSR disappears before decompression or persists after decompression, careful exploration is warranted. However, the delayed disappearance of LSR is common and, occasionally, early disappearance leads to better outcomes. Therefore, excessive exploration should be avoided to prevent unnecessary intervention. **Figure 4.** Flowchart for surgical decision-making based on lateral spread response (LSR) monitoring. If LSR disappears before decompression or persists after decompression, careful exploration is warranted. However, the delayed disappearance of LSR is common and, occasionally, early disappearance leads to better outcomes. Therefore, excessive exploration should be avoided to prevent unnecessary intervention.

#### *3.4. LSR Monitoring in Secondary HFS*

*3.4. LSR Monitoring in Secondary HFS*  LSR constitutes an electrophysiological manifestation that arises not only from neurovascular conflicts, but also from influences encompassing the REZ and, potentially, other unidentified mechanisms. Consequently, LSR can manifest within secondary HFS, LSR constitutes an electrophysiological manifestation that arises not only from neurovascular conflicts, but also from influences encompassing the REZ and, potentially, other unidentified mechanisms. Consequently, LSR can manifest within secondary HFS, where lesions exert an impact along the trajectory of the facial nucleus and the facial nerve REZ.

where lesions exert an impact along the trajectory of the facial nucleus and the facial nerve REZ. However, within the context of tumors originating from the facial nerve, such as facial nerve schwannomas, reports of facial spasm presentation are conspicuously absent However, within the context of tumors originating from the facial nerve, such as facial nerve schwannomas, reports of facial spasm presentation are conspicuously absent [29–31]. This discrepancy may be attributed to the propensity of facial nerve schwannomas to originate, predominantly, from the peripheral aspect of the facial nerve rather than from its

[29–31]. This discrepancy may be attributed to the propensity of facial nerve

intracranial segment. Consequently, the presence of abnormal electrophysiological signs such as LSR are not evident in such cases [32].

#### **4. Pathophysiology of LSR**

The exact pathophysiology of HFS remains unclear. While the most common cause is believed to be mechanical compression of the facial nerve at the REZ by an adjacent artery, there are cases where venous compression, or no specific vessel near the REZ, is identified. Thus, a more comprehensive explanation is needed beyond the direct compression of the REZ. Due to the vague nature of the disease's pathophysiology, the physiology of the AMR is also a subject of debate.

#### *4.1. Peripheral Theory*

The peripheral theory of HFS suggests an abnormal cross-transmission of facial nerve fibers at the site of vascular compression, known as the ephaptic transmission of neural impulses between different branches of the facial nerve [33]. Yamashita et al. [34] conducted a study using double stimulation of the AMR in 12 HFS patients to explain the pathophysiology of the condition. Their results showed a refractory period of 3.4 msec between stimulations, which remained constant within the same patients. This finding provided evidence of a peripheral mechanism underlying the pathophysiology of HFS, in contrast to the central nucleus theory. It indicated that the amplitude and latency intervals of the AMR stimulated at different time sequences remain consistent, suggesting no influence from the facial motor neuron. If the site of abnormal cross-transmission were in the facial nucleus, lateral spread responses would exhibit variable latency and amplitude, similar to the F wave [34]. Wilkinson used a strength–duration analysis to suggest that the AMR is likely mediated by antidromic afferent signaling along the facial nerve, as the chronaxies determined for the AMR and M wave were virtually identical [35].

#### *4.2. Central Theory*

The hyperexcitability of the facial nucleus has been proposed as a significant contributing factor in the pathogenesis of HFS, as suggested by Moller and Jannetta [36] and Poignonec et al. [37]. Yamakami et al. [38] suggested that the kindling-like hyperactivity of the facial nucleus induced by chronic electrical stimulation is the cause of the AMR. Several reports support the central theory, suggesting that the hyperexcitability of the facial nucleus is the origin of the lateral spread response. Ishikawa et al. [39–41] examined F waves pre- and postoperatively and during surgery, finding support for this central theory. By correlating F/M-wave amplitude ratios with lateral spread response F/M-wave amplitude ratios, they concluded that the origin of enhanced F waves is the same as that of the lateral spread response. The F wave exhibits variable latency and amplitude due to the hyperexcitability of the central nucleus [34,42].

#### **5. Further Research**

Thirumala and colleagues [14], in their meta-analysis, proposed that the definition of lateral spread response (LSR) resolution varies across different studies. They suggested that a prospective international multicenter study, with standardized and established definitions of LSR resolution, would yield more accurate and precise results regarding its prognostic value. This highlights the importance of having consistent criteria for determining LSR resolution in future research, to enhance the comparability and the reliability of findings.

Additionally, Kim's report [28] emphasized that the timing of LSR resolution may have prognostic significance. To further explore this aspect, more studies comparing the timing of LSR disappearance and its relationship to clinical outcomes should be conducted. Understanding the timing of LSR resolution in relation to patient outcomes can provide valuable insights into the prognostic implications of LSR changes during the course of hemifacial spasm (HFS) and subsequent microvascular decompression (MVD) surgeries.

#### **6. Limitations**

The primary aim of this paper was to provide a comprehensive overview of the monitoring methods, prognostic values, and pathophysiology of LSR. However, it is important to acknowledge certain limitations that were inherent in the scope and nature of this review. Unlike a traditional systematic review or meta-analysis, this study did not adhere to the conventional structure that includes detailed information about the methodology employed, such as the specific time frame of the literature search, the search terms used, and the inclusion and exclusion criteria. As this review focused on presenting a subjective overview of the topic rather than conducting an exhaustive analysis of the available literature, the absence of these methodological components is recognized. Therefore, it is acknowledged that the level of the evidence in the considered literature was limited, given that it did not fall within the domain of a systematic review or a meta-analysis. Nevertheless, significant efforts were made to maintain a neutral standpoint and to encompass a wide range of concerns related to LSR.

#### **7. Conclusions**

LSR monitoring has been used to identify the cause of HFS and to guide surgical interventions. This review article discussed the pathophysiology of LSR, the techniques for accurate monitoring, and the clinical implications. The presence and resolution of LSR were found to be associated with short-term outcomes, but their predictive value for long-term outcomes is less clear. Studies have shown conflicting results regarding the correlation between LSR resolution and HFS relief, highlighting the need for further research. Factors such as multiple vessel compressions and vessels that are not easily visible can contribute to residual spasms, even after successful decompression. The early disappearance of LSR before decompression can occur due to various factors. Overall, LSR monitoring is a valuable tool, but further research is needed to fully understand its implications and to optimize its use in HFS treatment.

**Author Contributions:** K.R.C. collected data, drafted the manuscript, and created the illustrations and figures. S.K.P. collected the electrophysiology data, interpreted the monitoring results, provided images from the monitoring device, and contributed to the conceptualization of the figures and the manuscript. K.P. collected data, revised the manuscript, and provided supervisory guidance on the literature. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

## *Article* **Penetrating Offenders in Hemifacial Spasm: Surgical Tactics and Prognosis**

**Hyun-Seok Lee <sup>1</sup> and Kwan Park 1,2,\***


**Abstract:** (1) Background: In cases of hemifacial spasm (HFS), there are various patterns related to the vascular compression of the facial nerve, including a very rare form that is seen when the offending vessel penetrates the facial nerve. However, there have been few reports in the literature regarding the associated surgical techniques and postoperative prognosis. (2) Methods: A retrospective review was conducted of 4755 patients who underwent microvascular decompression (MVD) surgery from April 1997 to June 2023. In total, 8 out of the 4755 patients (0.2%) exhibited a penetrating offending vessel; the medical and surgical records of these 8 patients were then analyzed. Surgery was then attempted to maximally decompress the penetrating offender. (3) Results: Seven out of the eight patients (87.5%) were spasm-free immediately after surgery, and one had only 10% residual spasm compared to their preoperative condition. That patient was also spasm-free one year later. Postoperative facial palsy occurred in one patient (12.5%) who was assessed as grade II in the House–Brackmann grading system. In another patient, the resection of a small facial nerve bundle did not result in facial palsy. There were no cases of hearing loss or other complications. (4) Conclusions: Decompressing the penetrating offender did not increase the incidence of facial palsy, and the prognosis for hemifacial spasms was good. Therefore, when a penetrating pattern was encountered during MVD surgery, decompression between the penetrating offender and the facial nerve may offer good results.

**Keywords:** hemifacial spasm; microvascular decompression; penetrating offender

#### **1. Introduction**

Hemifacial spasm (HFS) is a form of neurovascular syndrome that is usually due to neurovascular compression in the root exit zone (REZ) of the facial nerve. The disease presents as an intermittent, involuntary facial twitching movement that usually begins in the eyelids and progresses to involve the entire system of ipsilateral facial muscles. The result is an asymmetrical appearance of the face, due to the strengthening of the facial muscles on the side of the spasm. The pathogenesis of HFS is thought to be derived from the vascular compression of the facial nerve that emerges close to the brain stem, leading to demyelination and ephaptic transmissions [1,2].

There are various treatments for HFS, such as medications and *Botulinum toxin* injections, but, compared to these treatments, MVD surgery is the most effective treatment and one that completely resolves the symptoms [3–9]. The overall rate of being spasm-free after MVD surgery in patients with HFS is approximately 90%, with the other 10% of patients experiencing a recurrence of facial spasm or surgical failure [3,10,11]. In the 10% of cases where this facial spasm did not resolve itself, in rare instances, the surgeons may have encountered unusual patterns of compression that would make for a very difficult surgical challenge. In 2007, we categorized six different patterns of facial nerve compression, describing their clinical implications and prognosis [12]. As the number of cases increased, we began to see other rare and difficult cases [11], one of which was the perforating pattern, where the offending vessel penetrated the facial nerve.

**Citation:** Lee, H.-S.; Park, K. Penetrating Offenders in Hemifacial Spasm: Surgical Tactics and Prognosis. *Life* **2023**, *13*, 2021. https://doi.org/10.3390/ life13102021

Academic Editor: Alfredo Conti

Received: 3 August 2023 Revised: 22 September 2023 Accepted: 5 October 2023 Published: 7 October 2023

**Copyright:** © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

There are currently no reports as to how this pattern should be decompressed, whether complications such as postoperative facial palsy and hearing loss after decompression are likely, and with what factors they are associated; this is the focus of the current study.

#### **2. Materials and Methods**

*2.1. Patient Cohort*

We retrospectively analyzed the medical records of 4755 patients who underwent MVD with HFS from April 1997 to June 2023. All MVD surgeries were performed by a single surgeon (Kwan Park), with facial nerve motor-evoked potential (facial MEP), lateral spread response (LSR), and brain stem auditory evoked potential (BAEP) being monitored by an experienced neurophysiologist. The clinical information and details of the offending vessels of all 4755 patients are summarized in Table 1. Cases with 2 or 3 offending vessels present a sandwich pattern or a tandem pattern, in which the vessels are compressed together.

**Table 1.** Clinical characteristics of 4755 patients treated with MVD surgery for hemifacial spasm.


MVD: microvascular decompression, AICA: anterior inferior cerebellar artery, PICA: posterior inferior cerebellar artery, VA: vertebral artery.

Of these 4755 patients, 8 exhibited the penetrating type of pattern (0.2%). All penetrating offenders were confirmed via surgical microscopic findings. Of these 8 patients, 4 were male, 4 were female, 5 had right-sided lesions, and 3 had left-sided lesions. The median age of the patients was 41.5 years (an age range of 23–67). Of the 8 patients, the two oldest patients (67 and 55 years old) had underlying hypertension and were on anti-hypertensive medication, and the next-oldest patient (49 years old) was taking medication for diabetes mellitus. The other 5 patients had no underlying medical conditions.

The degree of preoperative facial spasm was analyzed using our previously published SMC grading system [13,14]. Grade I refers to situations when the spasm is localized to the periocular area, whereas grade II refers to situations when the involuntary movement spreads to other areas of the ipsilateral face and affects other muscles, such as the orbicularis oculi, frontalis, zygomaticus, mentalis, and platysma. Grade III represents the disruption of vision due to frequent spasms, whereas grade IV refers to a persistent spasm resulting in significant facial asymmetry. Among the 8 patients under study, the degree of preoperative spasm was classified as grade II in 3 patients (37.5%), as grade III in 3 patients (37.5%), and as grade IV in 1 patient (12.5%), whereas the grade of the final patient could not be determined.

The mean duration of postoperative spasm symptoms was 60.4 months (in a range of 18–120 months), with 4 patients being prescribed medication for preoperative treatment, 4 patients receiving botulinum toxin (Botox) injections, and 2 patients receiving both Botox injections and medication. One patient was prescribed medication and acupuncture at an oriental medicine clinic, whereas another received botulinum toxin treatment and acupuncture at an oriental medicine clinic. The final 2 patients received no other treatment prior to MVD surgery (see Table 2).


**Table 2.** Clinical characteristics of 8 patients with penetrating offenders.

No: number, M: male; F: female, R: right, L: left, BTX: botulinum toxin, HTN: hypertension, DM: diabetes mellitus, N/A: not available or not mentioned.

#### *2.2. Operative Technique and Intraoperative Monitoring*

The patients underwent MVD surgery via a retromastoid suboccipital craniotomy (RMSOC) while in the park bench lateral position. All patients underwent MVD surgery with intraoperative neuromonitoring, which consists of a real-time BAEP monitoring method [14] and the facial motor evoke potential (fMEP) and lateral spread response (LSR), i.e., abnormal muscle response (AMR). The evaluations of intraoperative LSR disappearance were categorized as follows: disappearance after durotomy and cerebrospinal fluid drain from the lateral medullary cistern; disappearance after immediate decompression, where the amplitude of LSR decreases but does not disappear completely; and no disappearance. In all patients, decompression of the neurovascular conflict zone was performed using Teflon felt.

#### *2.3. Assessment of Postoperative Outcomes*

All patients underwent preoperative intra-auditory canal (IAC) magnetic resonance imaging (MRI) to identify the most likely offender, the results of which were reviewed by an experienced neuroradiologist. Preoperative audiometry, including pure tone audiometry (PTA) and speech audiometry (SA), was performed, and the patients also underwent preoperative lateral spread response (LSR) testing conducted by an experienced neurophysiologist. Computed tomography (CT) of the brain was performed immediately after the operation in all patients, and a temporal bone CT scan was also taken postoperatively on day three in all patients. Postoperative pure tone audiometry (PTA) and SA (speech audiometry) was performed postoperatively on days 4–5, and the results were then compared with the preoperative results. The patient's spasms were assessed preoperatively, immediately after surgery (until 5 days after surgery), 1 month after surgery, 1 year after surgery, and up to 2 years after surgery. The lateral spread response (LSR) was examined just before the patient's outpatient department visit after their discharge.

#### **3. Results**

In the eight patients identified, the offending vessels were as follows: the anterior inferior cerebellar artery (AICA) in six patients (75%), the branch of the AICA in one patient (12.5%), and the posterior inferior cerebellar artery (PICA) in one patient (12.5%). The intraoperative LSR disappearance pattern was IIa (i.e., the vessel disappeared immediately after decompression) in seven patients (87.5%) and IIc (i.e., a 50% reduction with a residual presence) in one patient (12.5%).

An intraoperative change in BAEP was seen in three patients. Two had a 50% decrease in amplitude, prolonged by 1.6 ms and 2.0 ms, respectively, with full recovery by the end of surgery, and one exhibited the loss of all but wave I, with 80% recovery by the end of surgery. There was no long-term hearing loss after surgery. It is worth noting that three of the eight patients underwent surgery before our real-time BAEP monitoring method was established. All three exhibited no postoperative hearing loss.

Postoperative facial palsy was seen in one patient (12.5%). Immediately after surgery, the palsy was classified as House–Brackmann grade III–IV, and the patient showed gradual improvement to grade II but still demonstrated residual issues. The other seven patients (87.5%) exhibited no facial palsy.

The prognosis of postoperative spasm was evaluated immediately after surgery (around 5 days), 1 month after surgery, 1 year after surgery, and 2 years after surgery. In seven of the eight patients (87.5%), there was no spasm immediately after surgery. One patient had 10% residual spasms compared to their preoperative state. At 1 month after surgery, one of the seven patients who had no spasm immediately after surgery had 10% residual spasm compared to their preoperative state, and the one patient with 10% residual spasm had worsened slightly to 20%. At 1 year after surgery, one patient with a new 10% spasm at 1 month had similar symptoms, and one patient with symptoms shortly after surgery was then spasm-free. Of the other six patients who were asymptomatic immediately after surgery, three remained spasm-free, one patient was lost to follow-up, one patient with no spasm developed a spasm of approximately 10% compared to preoperative levels, and one had not yet reached the one-year postoperative mark. Upon follow-up in year 2, two patients with a residual 10% of spasms at 1 year after surgery were completely spasm-free after 2 years, and one patient who had spasms immediately after surgery and after 1 month remained spasm-free at 1 and 2 years after surgery. Of the three patients who were followed up, one was lost to follow-up, and two continued to be spasm-free (one of these two is a patient who developed facial palsy after MVD surgery) (see Table 3).


**Table 3.** Intraoperative findings and prognosis of 8 patients with penetrating offenders.


#### **Table 3.** *Cont.*

No: number, LSR: lateral spread response, BAEP: brain stem auditory evoked potential, D: day, M: month, Y: year, AICA: anterior inferior cerebellar artery, PICA: posterior inferior cerebellar artery, N/A: not available or not mentioned, H-B: House–Brackmann grade, CN: cranial nerve.

#### *Illustrative Case*

A 34-year-old man developed twitching of the right eye and mouth two and a half years prior to admission and was subsequently diagnosed with HFS. After medication and botulinum toxin treatments, he was referred for MVD surgery and visited our institution. His pre-operative spasm grade was SMC grade II; magnetic resonance imaging showed a complex REZ, which the radiologist interpreted as demonstrating that both the AICA and PICA were in contact in the REZ (Figure 1). LSR was observed in the right facial nerve during an examination with electrical stimulation (Figure 2a). *Life* **2023**, *13*, x FOR PEER REVIEW 6 of 13

**Figure 1.** A 34-year-old man with right HFS, shown in a preoperative proton density-weighted (PDweighted) magnetic resonance image (MRI). At the exit region of the right facial nerve root, the right anterior inferior cerebellar artery (AICA) and posterior inferior cerebellar artery (PICA) have a complex appearance (red circle). **Figure 1.** A 34-year-old man with right HFS, shown in a preoperative proton density-weighted (PD-weighted) magnetic resonance image (MRI). At the exit region of the right facial nerve root, the right anterior inferior cerebellar artery (AICA) and posterior inferior cerebellar artery (PICA) have a complex appearance (red circle).

Intraoperatively, the offending vessel was a branch of the AICA, which was penetrating the facial nerve (Figure 3a–c).

(**a**) (**b**)

tion, the LSR seen before surgery has disappeared.

trating the facial nerve (Figure 3a–c).

**Figure 2.** Right-side facial nerve conduction studies of a patient acting as an illustrative case: (**a**) the preoperative examination showed lateral spread response (LSR); (**b**) in the postoperative examina-

Intraoperatively, the offending vessel was a branch of the AICA, which was pene-

plex appearance (red circle).

**Figure 1.** A 34-year-old man with right HFS, shown in a preoperative proton density-weighted (PDweighted) magnetic resonance image (MRI). At the exit region of the right facial nerve root, the right anterior inferior cerebellar artery (AICA) and posterior inferior cerebellar artery (PICA) have a com-

**Commented [M1]:** Please provide the explanation

of the number in figure 2 in the caption.

**Commented [K2R1]:** I modified figure and

In both picture, each number is a reference point for measurement. '1' is onset latency, from '1' to '2' is onset to peak amplitude, and from '2' to '3' is

explained with caption.

peak to peak amplitude

*Life* **2023**, *13*, x FOR PEER REVIEW 6 of 13

**Figure 2.** Right-side facial nerve conduction studies of a patient acting as an illustrative case. In both picture, each number is a reference point for measurement. '1' is onset latency, from '1' to '2' is onset to peak amplitude, and from '2' to '3' is peak to peak amplitude: (**a**) the preoperative examination showed lateral spread response (LSR); (**b**) in the postoperative examination, the LSR seen before surgery has disappeared. Intraoperatively, the offending vessel was a branch of the AICA, which was pene-**Figure 2.** Right-side facial nerve conduction studies of a patient acting as an illustrative case. In both picture, each number is a reference point for measurement. '1' is onset latency, from '1' to '2' is onset to peak amplitude, and from '2' to '3' is peak to peak amplitude: (**a**) the preoperative examination showed lateral spread response (LSR); (**b**) in the postoperative examination, the LSR seen before surgery has disappeared. *Life* **2023**, *13*, x FOR PEER REVIEW 7 of 13

trating the facial nerve (Figure 3a–c).

(**c**)

**Figure 3.** Intraoperative microscopic findings for the illustrative case: (**a**) a branch of the anterior inferior cerebellar artery (AICA) passing through the facial nerve, with no other indentations visible in the root exit zone (yellow circle); (**b**) illustration of the surgical visualization before decompression. The illustration shows the offending artery passing through the facial nerve; (**c**) Microscopic findings with decompression in progress on both the left and right sides. The yellow arrow is the 8th cranial nerve, and the navy arrow shows a branch of the AICA penetrating the facial nerve, both images. **Figure 3.** Intraoperative microscopic findings for the illustrative case: (**a**) a branch of the anterior inferior cerebellar artery (AICA) passing through the facial nerve, with no other indentations visible in the root exit zone (yellow circle); (**b**) illustration of the surgical visualization before decompression. The illustration shows the offending artery passing through the facial nerve; (**c**) Microscopic findings with decompression in progress on both the left and right sides. The yellow arrow is the 8th cranial nerve, and the navy arrow shows a branch of the AICA penetrating the facial nerve, both images.

First, we checked whether there was a space between the facial nerve and the penetrating artery and carefully dissected the facial nerve and the penetrating branch of the

to decompress in all directions (Figure 4a–c).

(**a**)

First, we checked whether there was a space between the facial nerve and the penetrating artery and carefully dissected the facial nerve and the penetrating branch of the AICA

*Life* **2023**, *13*, x FOR PEER REVIEW 8 of 13

**Figure 4.** (**a**) First, decompression was started with small Teflon felt, from inferior side of the perforation site (yellow arrow); (**b**) Next, decompression was performed in supero-medial (supero-anterior) side (yellow arrow); (**c**) Lastly, the supero-lateral (supero-posterior) side decompression was performed (yellow arrow) to decompress in all directions of the perforation site, and the lateral spread response (LSR) subsequently disappeared. **Figure 4.** (**a**) First, decompression was started with small Teflon felt, from inferior side of the perforation site (yellow arrow); (**b**) Next, decompression was performed in supero-medial (superoanterior) side (yellow arrow); (**c**) Lastly, the supero-lateral (supero-posterior) side decompression was performed (yellow arrow) to decompress in all directions of the perforation site, and the lateral spread response (LSR) subsequently disappeared.

Immediately after decompression, the LSR disappeared, and the surgery was completed. The patient in question had no spasms and no facial palsy immediately after surgery and was discharged on day 5 with no further complications. Postoperative audiometry showed no hearing difficulty. One month after surgery, the patient still had no spasms, and the disappearance of LSR was confirmed via a postoperative neurophysiological study (Figure 2b).

#### **4. Discussion**

In 2008, we analyzed 236 cases of HFS, reporting several patterns of neurovascular compression of the facial nerve [12]. At the time, we categorized the patterns into six types: arachnoid (28.0%), loop (4.7%), and perforator (24.6%), which are generally compressed by a single causative vessel, and branch (7.6%), sandwich (11.9%), and tandem (22.0%), which are compressed by two or more vessels [12]. There were also three cases (1.3%) in which the pattern category was not clear at the time. As the number of surgeries grew, we added new neurovascular compression patterns. One is the encircling pattern, in which the vessel encircles the facial nerve in a 270-degree or even 360-degree loop. Another is the penetrating type, in which the culprit vessel passes completely through the facial nerve [11]. These patterns of neurovascular compression are particularly challenging to operate on, especially the penetrating type, as they are often not clearly identifiable from a magnetic resonance image, and cases are so rare that there are no established decompression methods.

There is a very limited corpus of literature on HFS caused by a penetrating offender; however, a report was published for a single case in 2015. Oh et al. reported a left HFS presenting in a 20-year-old male patient. The penetrating offender in this patient was the AICA; after decompression, the spasm disappeared, but definite facial palsy occurred [15]. From this case report and from our study, it is apparent that HFS caused by the penetrating type occurs at a relatively young age compared to the general population. Typically, HFS is reported to be more prevalent in those in their 40s and 50s [16,17]. Similarly, when analyzing all 4755 of our cases, the median patient age was 53 years old (range: 17–75), and the median age of the eight patients with the penetrating type was 41.5 years (range: 23–67). This has the clinical implication that symptoms occur at a younger age in the case of penetrating offenders (Table 4). It is conceivable that the symptoms occur at a younger age because the causative vessels (mostly the AICA) directly irritate the facial nerve. It is also possible that if the facial nerve is separated, myelination is less well developed than if it is not separated. Reports of duplication of the facial nerve in the mastoid segment or its distal part are rare in the ENT department [18–20], but there are few reports of duplication in the intracranial portion from the root exit zone of the facial nerve to the entry point of the IAC; therefore, further anatomic and pathologic studies are required.


**Table 4.** Factors correlating with penetrating offenders and hemifacial spasm.

In general, the same trend is seen in our 4755 cases; the AICA is the most common offending vessel for HFS, whereas the second most common is the PICA, followed by the AICA and other vessel compressions in conjunction (see Table 1) [12,21]. The PICA originates in the VA and emerges from the ventral to the dorsal side of the brain stem, usually at the level of the lower cranial nerves (at the 9th, 10th, and 11th cranial nerves), forming a hairpin structure (caudal loop) downward [21,22]. Therefore, while it is possible for a vessel to loop and compress the root exit zone of the facial nerve or encircle the facial nerve, it is structurally difficult for the vessel to pass through the facial nerve. In the cases

examined in this study, only one out of the eight (12.5%) showed penetration by the PICA, and almost all of them showed penetration by the AICA (87.5%) (see Table 3). The facioauditory primordium, the origin of the facial nerve, appears in the third week of gestation. It then splits into two sections at the end of the fourth week and is complete by the fifth or sixth week of gestation [15,23]. The AICA begins to develop later, in the gestational fourth or fifth month, when the facial nerve is already fairly well-developed [15,24]. Therefore, facial nerve penetration by the artery is thought to occur as the facial nerve attempts to pass through the divergence that appears after 5–6 weeks of gestation.

In all patients, we attempted complete decompression in all directions and a 360-degree inspection in all patients, as in the illustrative case. First, check for space in the perforating artery and facial nerves with a micro-dissector or micro-bayonet forceps. Then, carefully dissect, and, when space is made, put small Teflon felt in there to first decompress. Next, we decompressed in all possible directions (Figure 4a–c). Decompressing the location between the facial nerve and penetrating vessels was associated with a very low probability of postoperative facial palsy. In one case, we decompressed only the medial side in a patient presenting a difficult dissection. As the LSR improved when the medial side was decompressed in this patient, the lateral side, which was difficult to meticulous dissection and presented a high risk of facial nerve injury upon dissection, was left untreated, and the operation was terminated. Postoperatively, the facial spasm had been resolved. In another case, the fascicle of the facial nerves was divided into 90% and 10% divisions by a penetrating PICA. In this patient, decompression of the thin side of the divided nerve fascicle was not possible. We first stimulated the thin nerve fascicle with a direct nerve stimulator to check the facial muscle response, which showed a small response by the orbicularis oris muscle. The thin fascicle that could not be decompressed and dissected was then excised, and the neurovascular compression between the large fascicle and the offending vessel was decompressed with Teflon felt. Immediately after decompression, the LSR disappeared. After the surgery, facial spasms remained at 10–20% compared with preoperative conditions; notably, there was no postoperative facial palsy. The absence of facial palsy despite the resection of a portion of the facial nerve in this patient was thought to be related to the innervation of the distal branch of the facial nerve. The facial nerve has five branches, namely, the temporal, zygomatic, buccal, marginal, and cervical branches, each of which is known to have interconnections, and these connections vary considerably [25–28]. In addition, several studies have shown that facial nerve is anatomically connected to the trigeminal nerve [25,27–31], vestibulocochlear nerve [25,32], glossopharyngeal nerve [25,33], and vagus nerve [25], as well as the cervical plexus [25,34]. In particular, there are numerous interconnections and variations between the zygomatic and buccal branches of the facial nerve, which can confound LSR measurements; we have previously published a paper on how to measure LSR in this context [26]. As the orbicularis oris muscle is often innervated by both the zygomatic and buccal branches, we consider that facial palsy did not occur despite the partial resection of the nerve fascicle during the operation. Maximal decompression using Teflon felt placed between the penetrating vessel and the facial nerve was performed in all eight patients, as mentioned, and attempted decompression in all compression sites. One patient developed postoperative facial palsy that was classified as House–Brackmann grade II, but the other seven patients exhibited no facial palsy and had good prognoses regarding postoperative facial spasm (see Table 3). Therefore, even if the artery were going through the facial nerve, it is considered necessary to actively dissect and decompress it. Yet, the most important thing we and other very experienced surgeons of more than 4000 cases emphasize is to check the neurovascular conflict completely, it is important to check the full inspection of facial nerve and entire 360 degrees of REZ if possible [35]. Also, even if you encounter a perforating artery, you should look for offenders in the REZ. This is because MVD of HFS is not an emergency surgery, but a functional neurosurgery, and it is recommended to decompress all offenders safely.

There is also a nervus intermedius that can be confused with a split in the facial nerve. The nervus intermedius was first identified in 1563; it was first named "portio media intercommunicantem faciei et nervum auditorium" in 1777 by Heinrich August Wrisberg [36,37]. The nervus intermedius is thus named because it is located between the facial nerve and the superior portion of the vestibular cochlear nerve [37,38]. The nervus intermedius carries parasympathetic nerve fibers to the nasopalatine and lacrimal glands, conducting sensory information from facial areas such as the concha of the ear and the nose [37–41]. Irritation of the nervus intermedius can cause geniculate neuralgia, which is expressed as intermittent but severe sharp pains deep in the ear, accompanied by the disruption of salivation, the sense of taste, and lacrimation [42]. Geniculate neuralgia can also be caused by neurovascular compression, which is an indicator of the need for MVD surgery. Lovely and Jannetta reported on 14 cases of MVD surgery in patients with geniculate neuralgia [43]. In their paper, they reported that vascular compression of the nervus intermedius was present in the surgical findings and correlated with symptoms such as deep ear pain [43]. In surgical findings that appear to show a perforating offender, it is possible to differentiate between the perforating offender and the nervus intermedius by tracing the entire course of the facial nerve. It may be helpful to perform direct nerve stimulation to check the response of the facial muscles. As mentioned above, the nervus intermedius is mainly a sensory nerve; when stimulated, it manifests as ear pain [38], rather than as a facial muscle response. It is also thought to be helpful in the differentiation of preoperative symptoms.

A limitation of this study is the small cohort of eight cases. This refers to 8 out of 4755 cases, so a very small probability of 0.2%. This means that the chances of actually encountering a perforating offender during MVD surgery are very low. In addition, this study was based on data regarding the experiences of a single surgeon, and it was not possible to compare the findings with results from other institutions. This is because there are few reports regarding HFS with a perforating offender that are available from other institutions. However, this limitation also means that our surgical method can act as a reference for the operation. As more data are accumulated from continuing MVD surgeries, in the future, we expect to analyze the prognoses and casual factors in a large series.

#### **5. Conclusions**

In as little as 0.2% of the 4755 cases examined in this study, offenders of the penetrating type are very rare, but they do exist. Cases of penetrating offenders in the context of HFS have been reported very rarely, and there is no established policy for decompression. In our experience, when the penetrating offender and the facial nerve fascicle are delicately dissected and decompressed using a small piece of Teflon felt, the facial spasm is resolved, and the likelihood of postoperative facial palsy is very low. In the case where it was necessary to cut a minor portion of a facial nerve that was difficult to decompress, it was found that the patient recovered without exhibiting facial palsy.

The statistical analysis of the factors involved showed that such patients were treated at a relatively young age compared to the general age of HFS patients. For this reason, in HFS cases at a relatively young age, when the offending vessels in the root exit zone are unclear on an MRI scan, it is advisable to consider the possibility that the offending vessel is of the penetrating type. In addition, if encountering a penetrating offender during MVD surgery, it is recommended that the surgeon should not hesitate but instead conduct careful dissection and complete decompression.

**Author Contributions:** Conceptualization, K.P.; Writing—original draft, H.-S.L.; Writing—review & editing, H.-S.L.; Supervision, K.P. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** This study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board of Konkuk University Medical Center (2023-07-008).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** All data included in this study can be provided by contacting hs5937@hanmail.net.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

MDPI St. Alban-Anlage 66 4052 Basel Switzerland www.mdpi.com

*Life* Editorial Office E-mail: life@mdpi.com www.mdpi.com/journal/life

Disclaimer/Publisher's Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Academic Open Access Publishing

mdpi.com ISBN 978-3-0365-9294-7