Analysis of the mylohyoid nerve in elderly Japanese cadavers for dental implant surgery

Abstract Objectives Injury to the mandibular nerve (MN) branches may cause pain and irregular occlusal movement during mastication after mandibular dental treatments. Growing evidence indicates that the calcitonin gene‐related peptide (CGRP) plays a key role in the development of peripheral sensitization and the associated enhanced pain, suggesting it may be a sign to ensure a safe and reliable dental implant treatment. Our focus was on the distribution of the MN branches and their communication with the lingual nerve (LN), the localized expression of CGRP, and the identification of a pain area related to the mylohyoid muscle (MM) fascia in the mandibular floor. Material and Methods In this study, MM samples from 440 sides of 303 human cadavers aged 61–103 years were examined microscopically and immunohistochemically. These data were further evaluated by the use of principal component analysis. Results A complex but weak attachment site was identified for the fascia of the MM. CGRP expression was mainly located in small vessels and was scattered throughout the whole fascia of the MM. Communication between the MN and LN was found in 62.5% (275/440) of the samples. The results from the principal component analysis showed that the positive contributions were from the descending branch in the premolar region (correlation coefficient value R = 0.665), the ascending branch in the molar region (R = 0.709) and the intermediate branch of the digastric branch (R = 0.720) in component 1. In the fascia off the MM, strongly labeled CGRP‐positive cells were also found around the blood vessels and the nerve. Conclusions The findings reported in this study indicate that there is a risk of damage when pulling the fascia off the MM at the border of the molar and premolar regions during dental implant surgery.


| INTRODUCTION
The mandibular nerve is mainly composed of two roots that send branches to the mylohyoid muscle (MM) and the anterior belly of the digastric muscle. The posterior trunk of the MN gives rise to the main sensory nerves, that is, the auriculotemporal, lingual, and inferior alveolar nerves and motor branches to the MM and the anterior belly of the digastric muscle (Brennan, Webb, Kemidi, Spratt, & Standring, 2008).
Communication between the MN and LN was also found at some frequency (46. 3% Kameda & Uber den, 1952;Sato, Sunohara, Ueno, & Yoshida, 2004, 66%). In dental implant or oral surgery, the inner surface of the mandible has become an important area for lingual flap dental treatment (Urban et al., 2018).
A case report showed that the mylohyoid line somewhat deviates from that position, which makes precision peeling extremely difficult (a reference is needed here). Clinically, the submental triangle containing the MN and the submental artery is one of the important areas for dental implants or oral surgery using the lingual flap. Therefore, it is necessary to investigate the distribution of branches of the MN and the range of origin of the MM within the fascia.
The pain due to injury of the mandibular nerve branches may be related to local expression of the calcitonin gene-related peptide (CGRP). In general, the blood flow in submental blood vessels is controlled by the sympathetic and parasympathetic nerves. In particular, the dominant effect of the sympathetic nerve on blood flow decreases at this site. There is a large amount of evidence showing that CGRP mainly controls the smooth muscle in the small branches of blood vessels, such as the facial, superficial temporal and maxillary arteries. Throughout these branches, CGRP restricts the blood supply to the submandibular triangle area. CGRP is also a neurotransmitter that helps transmit signals related to pain from the muscle's vessels (Ambalavanar et al., 2006;Azuma, Miwa, & Sato, 2016;Onuoha & Alpar, 1999;Sakuma et al., 2016). There are several mechanisms by which CGRP induces vascular relaxation (Bell & McDermott, 1996;Brain & Cambridge, 1996;Marshall, 1992). The distribution of CGRP affects pain and hearing through its actions on the human tensor tympani muscle (Yamazaki & Sato, 2014). Therefore, it is necessary to examine the localization of CGRP as a neurotransmitter at the site of pain in the submandibular triangle.
Our focus was on the distribution of the MN branches and their communication with the lingual branches at the microscopic level. We also observed the expression of CGRP at the inner and outer surfaces of the fascia using immunohistochemical methods, which might be related to pain during the reflection of the MM fascia in the submandibular triangle. All data, including the sex and the left-right differences, were further evaluated using principal component analysis (PCA). These results may provide useful information for safer oral clinical treatments.

| Statistical methods
The data were assessed using a Pearson's chi-square test to determine the validity of the distribution. The level of significance (p < .05) was calculated using Pearson's chi-square tests. Principal component analysis (PCA) and cluster analysis were performed for the human MNs comprised of 13 elements, including age, sex, the difference between right and left, and the ratio of the MNs. These effects were assessed using one-way analysis of variance (ANOVA) followed by Bonferroni's post hoc test with one categorical independent variable and one continuous variable. These results are reported as the mean ± SD. The statistical analyses were performed using IBM SPSS Statistics Base, version 22.

| Attachment of the MM and mylohyoid line
In the dentulous samples, the MM origin was located along the mylohyoid line on the mandible, which curved widely on the inner MM surface (Figure 1). After peeling off the fascia (fa), the attachment of the MM was clearly found from the molar region to the anterior basal region of the inner surface of the mandible (Figure 2). The fa with numerous fine fibers was also observed between the periosteum and surface of the mandible (Figures 1 and 2). In the edentulous region, the MM origin was limited, and the mylohyoid line receded and became shorter from the infraorbital foramen to the posterior region of the inner surface of the mandible (Figure 2). The fa of the MM was thin, and the periosteum was also found in the attachment area of the mylohyoid line. The alveolar ridge was found in the anterior region of the mandible (Figure 2).

| Fascia structure
The origin of muscle fibers was found by peeling off the MM fascia from the mylohyoid line of the inner surface of the mandibular body ( Figure 1). The site between the muscle fiber and the thin serosa was weakly attached to the mylohyoid line ( Figure 1). The fascia covering the MM was also thin and formed a weak attachment border of the MM at the mylohyoid line on the inner surface of the human mandibular body. The muscle attachments at the site of the mylohyoid line were either long or short. A short attachment site was mainly observed at the inner surface of the anterior region of the edentulous human mandible in contrast to that of a dentulous human mandible ( Figure 1). In edentulous samples, the attachment site was mainly observed in the limited mylohyoid line, where the cortical bone was irregularly formed and showed a resorption site compared to the findings in dentulous samples (Figures 1 and 2). In particular, the fascia  (Figures 1 and 2).

| Branches of the MN division points
We identified the branches of the MN division points, such as the ascending (AB) and descending branch (GB) of the molar region, the ascending (BB) and descending branch (BH) of the premolar region, F I G U R E 2 The attachment of the fascia covering the mylohyoid muscle (MM) in dentulous (Dent) and edentulous (Edn) samples. On the lingual side at the base of the mandible, the fascia of the MM appears as a wide and very thin sheet that covers the MM. The border of the MM origin is clearly different from the origin of the muscle, since that looks like a border line angled obliquely downward from the posterior region to the anterior regions (a). After peeling off the fascia of the MM from its origin along the mylohyoid line on the inner surface in the mandibular body, very fine and thin fascia-composed fibers were found on the inner side of the mandible (b). The site between the muscle fiber and the thin serous membrane has a weak attachment at the mylohyoid line (arrows). Then, the periosteum (Ps) of the mandible body was removed from the mylohyoid line, and the small fine fibers were located between the periosteum and MM (c). On the lingual side at the base of the mandible, the border between the MM and the periosteum of the mandible body was almost indistinguishable at the site of the mylohyoid line (d). After peeling off the fascia of the MM from its origin along the mylohyoid line on the inner surface of the mandibular body, the periosteum of the mandible body was collected at this site, which is clearly not the fascia of the MM (e). Then, the periosteum of the mandible body was removed from the mylohyoid line, and the periosteum with thick connective tissue was collected between the periosteum and the MM (f).

| Communication between the MN and LN
We identified the site of communications between the MN and LN. The communications between the MN and LN were mainly found in the premolar region (P2) of the human mandible (Figures 5 and 7).
The ratio of the communication between the MN and LN (Com) was 62.5% (275/440). Negative correlations were found between the Com and BB in the MN (r > 0.283, p < .05).
3.6 | The correlation coefficient measurement data, cluster analysis and PCA were performed for human MM in different groups for 11 variables including aging, sex, appearances, and locations of the MN supply There were correlation coefficients between 13 measured elements from the human MN based on the macroscopic data (Table 1). We performed clustering analysis using a hierarchical classification model with these two components, demonstrating that optimal grouping was obtained with three clusters (Figure 8 between BB and Com. Component 1 was explained mainly by the CB and PB, IB, GB, and ANB, and the negative correlation between aging and these groups. Figure 9 shows a display of the MM variables and components 1 and 2 on a two-dimensional map (Table 1) Table 1). The following correlations among the appearance ratio of the MN branches of the MM are shown in Table 1: We observed a fine positive correlation between BH and PB (r > 0.627, p < .01), CB (r > 0.624, p < .01), AB (r > 0.581, p < .01), and INB (r > 0.437, p < .01); a positive correlation between GB and ANB (r > 0.662, p < .01) and a positive correlation with BH (r > 0.568, p < .01); a negative correlation between GB and age (r > −0.476, p < .01); a correlation between ANB and CB (r > 0.428, p < .01) and PB and CB (r > 0.532, p < .01); a correlation between IB and PB (r > 0.501, p < .01); a positive correlation between CB and AB (r > 0.421, p < .01); and a negative correlation between CB and aging (r > −0.325, p < .01) ( Table 1).

| Immunohistochemical observations of the localization of CGRP in whole mount samples and tissue sections with MM fascia
In the fascia of the mylohyoid muscle (MM), the c scatter was sometimes located around the blood vessels. In the lateral region of the tongue, the CGRP reaction was mainly located at small cells around the vessels beneath the submucosa and lamina propria. Moreover, a fine positive reaction was found around the muscle fiber of the MM at whole mount analysis levels ( Figure 10). In addition, we defined a

| DISCUSSION
In the present study, detailed observations were carried out on the innervation and arrangement of the MN traversing in various directions. The number of nerve branches entering the MM was evidently more than that of the anterior belly of the digastric muscle. Communications between the MN and LN were mainly found between the premolar and molar region of the mandible, which has been shown to occur at a frequency ranging from 45 to 66% (46. 3% Kameda & Uber den, 1952;66% Sato et al., 2004, this study, 63%). It is interesting to point out that the communication occurs between the LN and MN, both of which originate from the mandibular nerve. The tongue is supplied by the facial, glossopharyngeal, hypoglossal and vagus nerves.
This communication makes the innervation system of the tongue even more complicated, and such communication makes it more susceptible to injury during third molar extractions (Potu, D'Silva, Thejodhar, & Jattanna, 2010). It has been reported that during dental surgery, pain and dysesthesia resulting from injury to the LN would spread to the inner side of the tongue via the aforementioned system (Chossegros, Guyot, Cheynet, Belloni, & Blanc, 2002), which is consistent with our findings.
In general, the posterior region of the MM is in the vicinity of the molars, including the first and second molar teeth. It has been reported that the molar tooth location is the most important region for dental implant treatment (Huang et al., 2015;Lin et al., 2014). In the present study, the attachment site for the MM was clearly different between the dentulous and edentulous samples. In particular, F I G U R E 7 Microscopic features in the nonsublingual artery with a lateral view of the submental nerve, artery, and vein at the base of the mandible. In the lateral view of the basal region of the mandible, the submental nerve and artery and vein were located between the MM and anterior belly of the digastric muscle and the submandibular gland. The LN projected to the submandibular gland at the lateral surface of the MM. The main trunk of the submental vein and artery were clearly located in their basal regions near the hyoid bone (a,b). The MN was found in the upper region of the submandibular gland and extended many fine branches (arrowheads) to the MM. Various branches of the submental artery were mainly found in the posterior, medial, and anterior regions of the MM (a,b). On the medial side of the MM, the SMA branch (red arrowhead) extended to the lateral tongue through the MM. The LN (yellow arrowhead) was divided into branches in the lateral anterior region of the tongue (c,d). ABDM, anterior belly of the digastric muscle; APA, ascending pharyngeal artery; FA, facial artery; MM, mylohyoid muscle; MN, mylohyoid nerve; SD, sublingual duct; SG, submandibular lymphatic ganglion; SMA, submental artery; SMG, submandibular gland; SMV, submental vein. Bar = 1 cm structural differences in the MM fascia were found between the inner and outer side of the mandible. Despite the existence of many morphological risk factors during dental implant surgery (Lin et al., 2014;Moraschini & Porto Barboza, 2016;Nassar et al., 2014), our findings indicate that the difference in functional mastication should be a good guide to carefully observe the MM origin site on the mandible. Especially, when peeling off the fascia, the inner side of the MM fascia might be damaged easily due to its very fine and thin structure.
On the other hand, it is generally accepted that CGRP is a nerve pain marker in the smooth muscle of blood vessels (Ambalavanar et al., 2006;Azuma et al., 2016;Onuoha & Alpar, 1999;Sakuma et al., 2016) related to vascular relaxation (Bell & McDermott, 1996;Brain & Cambridge, 1996;Marshall, 1992). The location of CGRP indicates vasodilation or widening of the blood vessels related to the blood flow in human MM. CGRP is most frequently present among high-threshold mechanosensitive (presumably nociceptive) afferent neurons, suggesting that CGRP is also expressed in many other types of primary afferent neurons (Hoheisel, Mense, & Scherotzke, 1994). It has been reported that most CGRP-positive fibers are located in the outer layer of the fascia and the subcutaneous tissue around the MM (Tesarz, Hoheisel, Wiedenhöfer, & Mense, 2011). In the present study, the expression of CGRP was mainly detected around numerous vessels in the MM fascia and in the posterior region of the MM. The expression of CGRP at different locations may be a landmark that signifies a pain risk for peeling off the MM fascia and the periosteum of the mandible during oral surgery. On the lingual side of the scheduled implant site, a preoperative biopsy of the oral mucosa on the mylohyoid muscle could be performed to examine the positive activity level of CGRP, which might provide information about the prognosis and help guide the procedure. It has also been reported that CGRP originating from the trigeminal ganglion cells could amplify nociception and lead to migraines (Deen et al., 2017;Dussor et al., 2014). Therefore, rimegepant, which is a CGRP antagonist, and fremanezumab, which is a monoclonal antibody targeting CGRP, have been developed (Lipton et al., 2019;Scuteri et al., 2019). It would be interesting to determine if local administration of these reagents could suppress migraine vasodilatation and headaches.
In the cluster analysis, these elements were classified into three groups; cluster 1 comprised of PB, IB, ANB, GB, and Com; cluster  Note: Ascending (AB) and descending branch (GB) of molar region, ascending (BB) and descending branch (BH) of premolar region, ascending (CB) and descending branch ( Moreover, one of the purposes of this study was to accurately identify the MN pathway using some anatomical landmarks in its vicinity. In general, severe trigeminal neuralgic pain mainly occurs, compared to the pain that occurs via any other nerve in the body (Benoliel & Eliav, 2008;DuPont, 2008). The distribution of the fine small vessel supplies is also important for muscle pain conception (Wasner et al., 2002). We found that the attachment site of the MM