A Newly Discovered Tendon Between the Genioglossus Muscle and Epiglottic Cartilage Identified by Histological Observation of the Pre-Epiglottic Space

The epiglottis is a process that points superiodorsally from the root of the tongue at the boundary between the pharynx and the larynx. It shifts backward during swallowing to cover the laryngeal aperture in the movement known as the swallowing reflex [1], preventing the food bolus from being carried into the airway. The swallowing reflex is an involuntary movement, and if the reflex is mistimed, the food bolus may be carried into the airway, causing dysphagia (aspiration) [2, 3]. Conventionally, dysphagia was mostly regarded as a sequela of amyotrophic lateral sclerosis and other motor neuron diseases, but age-related structural changes around the epiglottis are also attracting attention as another cause [4‐6]. Sarcopenia, a condition characterized by a generalized decrease in muscle mass, is considered to be one of the major conditions facilitating aspiration, since the mass and function of the lingual muscles associated with swallowing are severely diminished [7‐10]. The functional and physiological relationships between the lingual muscles and the epiglottis have thus been well reported. However, there have been few morphological reports concerning the attachment of the lingual muscles to the epiglottis.

The skeleton of the epiglottis is made up of cartilage with high elastic fiber content, and this epiglottic cartilage is connected to the tissues around it by a number of ligaments. At the front of the epiglottis, these include (1) the hyoepiglottic ligament, which connects the epiglottis to the hyoid bone; (2) the thyrohyoid ligament, which connects the hyoid bone to the thyroid cartilage; and (3) the thyroepiglottic ligament, which connects the epiglottis to the thyroid cartilage. The space surrounding these ligaments is known as the pre-epiglottic space (PES) [11‐13] (Fig. 1a). The PES is known to be an extremely important structure that enables epiglottic retroversion, with the hyoepiglottic ligament in particular performing complex actions during the short reflex. The normal swallowing reflex mechanism comprises the first epiglottic movement at the start of retroversion, in which the epiglottis reaches the horizontal position (Fig. 1b), and the second epiglottic movement, in which the top third of the epiglottis folds downward to cover the laryngeal aperture completely (Fig. 1c). In the first epiglottic movement, the hyoepiglottic ligament extends to support the retroversion of the epiglottis, but in the second epiglottic movement, it draws the tip of the epiglottis forward together with the hyoid bone [14‐17]. To explain these actions of the hyoepiglottic ligament, Van Daele et al. showed that the hyoepiglottic ligament is attached not only to the hyoid bone but also to the fascia of the root of the tongue, suggesting that the hyoepiglottic ligament may consist of more highly active tissue. Reidenbach further named the fiber bundle running toward the tongue root the “hyoepiglottic membrane” and showed that the hyoepiglottic ligament and hyoepiglottic membrane are fiber bundles running anteriorly from the lingual surface of the epiglottic cartilage [18]. Recent studies have reported that the fibrous components of the hyoepiglottic ligament and hyoepiglottic membrane are decreased by smoking and aging, and the degradation of these fibers has been implicated in obstructive sleep apnea and dysphagia [19, 20]. However, the hyoepiglottic membrane is a thin-fiber bundle belonging to the hyoepiglottic ligament, and it is unclear whether it is a structure capable of controlling the movements of the epiglottis.

In this study, a comprehensive visualization of the fibrous structures present in the PES was conducted, with a focus on the hyoepiglottic ligament, with the objectives of identifying new fiber bundles and clarifying the functional and physiological mechanisms of epiglottic movements from the morphological perspective.

This study was conducted in compliance with the provisions of the 1995 Declaration of Helsinki (revised in Edinburgh in 2000). Twenty donated cadavers (10 men and 10 women) of mean age 79 years (range 72–88 years) were studied. These cadavers were donated to Tokyo Dental College for use in human dissection research and teaching, and their use in this study was approved by the university’s ethics committee (approval no. 891). The donated cadavers were fixed by arterial perfusion with 10% v/v formalin solution and preserved for more than 3 months with 70% v/v ethanol solution. Tissue blocks were prepared containing the tongue, epiglottis, thyroid cartilage, arytenoid cartilage, and upper cricoid cartilage. All samples were decalcified by incubation in 0.5 mol/l EDTA solution (pH 7.5; decalcifying solution B, Wako, Tokyo, Japan) for 30 days at room temperature. Paraffin embedding was carried out by the normal method, with cut planes comprising 10 samples in sagittal section, 5 in horizontal section, and 4 in coronal section (with all sections including both men and women). From each of the prepared paraffin blocks, four 5–10-μm-thick semi-continuous sections at approximately 0.3-mm intervals were prepared. Two of these prepared sections were stained with hematoxylin and eosin (H–E) stain and the other two with elastica-Masson stain (a variation of Masson–Goldner staining). The tissue slices were observed and photographed with a universal photomicroscope (UPM Axiophot 2, Carl Zeiss Microscopy, Jena, Germany), but ultra-low-magnification photographs (less than 1 × at the objective lens) were photographed with a high-grade flat scanner (Epson GTX970; Seiko Epson Corporation, Nagano, Japan) equipped with translucent lighting.

When the tongue and epiglottis were viewed from above, the glossoepiglottic tendon was composed of numerous collagen fiber bundles that radiated out from the center of the epiglottic cartilage (Fig. 8a). These collagen fibers ran along the surface of the epiglottic cartilage (Fig. 8b, red arrowhead), and at their terminals, they formed bridges in a reticular pattern with the fine fibers of the perichondrium of the epiglottic cartilage (Fig. 8c, red and blue arrowheads). In cross-sectional view, perichondrium was present in the superficial layer of the insertion site and cartilage matrix in its deep layer (Fig. 8d), and the collagen fibers of the glossoepiglottic tendon were only evident in the perichondrium (Fig. 8e, red arrowhead).

The findings seen in this investigation were observed in all 19 cadavers. However, because of differences in the size of the epiglottis between cadavers, the same structures were not always observed at specific distances apart. The epiglottis tended to be larger in men and smaller in women. The slit in the superior part of the hyoid bone tended to be shorter when the position of the larynx was higher and longer when it was lower. This may be related to age-associated ptosis of the larynx, but this was not clear from the results of the present study.

In the long history of anatomy to date, very few studies have discussed the relationship between the lingual muscles and the epiglottis in morphological terms [17, 18]. The structure of the hyoepiglottic ligament, which divides the tongue from the epiglottis, has therefore yet to be completely explained. Van Daele showed that the hyoepiglottic ligament starts from the lateral side of the lingual surface of the epiglottis and is attached to the greater cornu of the hyoid bone and the fascia of the lingual muscles of its upper portion. We observed similar results in the epiglottis in lateral sagittal section, and additionally observed the novel finding that the fascia is that of the superior longitudinal lingual muscle (Fig. 9a). The fact that the superior longitudinal lingual muscle attaches to the hyoid bone is also a new discovery, and this suggested that a simpler function of the hyoepiglottic ligament is exerted because the superior longitudinal lingual muscle prevents the genioglossus muscle from progressing backward. Reidenbach defined the hyoepiglottic ligament as a two-layer ligament and showed that the lower layer is attached to the periosteum of the superior edge of the hyoid bone, whereas the upper layer spreads out within the muscle tissue of the tongue root. Since the upper layer is also a semi-independent fiber bundle that is also found in the free part of the epiglottic cartilage, he named this the hyoepiglottic membrane. We also made similar observations in the mid-lateral section of the epiglottis and also found that the genioglossus muscle, which runs posteriorly while breaking through the superior longitudinal lingual muscle bundles, forms a muscle–tendon junction with the hyoepiglottic membrane (Fig. 9b). The major difference between Van Daele’s and Reidenbach’s results and ours is the structure of the median epiglottis. In the median section of the epiglottis, a tendon that forms a thick-muscle–tendon junction with the genioglossus muscle is broadly attached to the lingual side of the epiglottic cartilage. Because it formed a narrow slit with the thin-hyoepiglottic ligament directed toward the base of the epiglottic cartilage, we recognized this tendon as a completely independent fiber and have therefore named it the glossoepiglottic tendon (Fig. 9c). From the above results, the glossoepiglottic tendon forms a thick-muscle–tendon junction with the genioglossus muscle in the median epiglottis and is broadly attached to the middle of the lingual surface of the epiglottic cartilage. The hyoepiglottic ligament, on the other hand, runs between the hyoid bone and the epiglottis, and it was found to attach in a V shape directed from the lateral margin of the epiglottis to its median base (Fig. 9d). The glossoepiglottic tendon that was discovered has never been reported in anatomical studies of the larynx, such as those by Van Daele and Reidenbach. This new discovery required three unique research methods. The first was that the observation target was from the epiglottis to the entire tongue. This made it possible to more clearly observe the relationship between the epiglottic cartilage and the lingual muscle at the base of the tongue. The second was that sagittal, horizontal, and forehead cuts were examined; three cross-sections clarified three-dimensional fiber bundle orientation. The third was that continuous sections were prepared at equal intervals. For the larynx, where the structure changes significantly in a small area, continuous sections were effective.

The basic components required of a motor organ in the living body are the muscles that provide its motive power, the tendons that transmit this force, and the skeleton to exert its action. Among these components, the connections between tendons and bones are known as entheses [21]. Entheses are histologically classified into three different types: periosteal insertions, fibrous entheses, and fibrocartilage entheses. In periosteal insertions, the collagen fibers of the tendon are inserted into the periosteum; in fibrous entheses, they pass through the periosteum and are inserted into the bone matrix; and in fibrocartilage entheses, they pass through the layer of fibrocartilage on the surface of the bone and are inserted into the bone matrix [22]. In the masticatory muscles of the head and neck in particular, these three types of enthesis are all found together in a single muscle insertion [23, 24]. In the present experiments, however, the morphology of the insertion of the glossoepiglottic tendon into the epiglottic cartilage resembled that of a periosteal insertion, and no fibrous or fibrocartilage enthesis was observed. In addition, the skeleton that is the point of action is not hard bony tissue, but is composed of elastic cartilage, the softest form of cartilage, making this insertion unlike the previously reported concepts of entheses. These results suggested that in cartilage that undergoes significant distortion and periosteal insertions, which are attached across a broad area as if sewn into the perichondrium, may be better at preventing deformation of the cartilage when it functions are fibrous or fibrocartilage entheses, which are fitted deeply into a small spot.

Experiments concerning the computational analysis of swallowing mechanics use videofluoroscopy or barium X-rays in modified barium swallow studies to enable the amount of movement of anatomical structures to be measured and subjected to vector analysis or quantitative evaluation [25‐27]. On the basis of previous computational analysis of swallowing mechanics, the first epiglottic movement was formerly thought to occur in tandem with the movements of the hyoid bone and larynx [14‐16]. Recently, however, numerous studies have reported that the main component of the first epiglottic movement is in fact the backward movement of the tongue root [28‐30]. Orsbon et al. further used electromyography to show that the genioglossus muscle relaxes and the geniohyoid muscle contracts during the first epiglottic movement [31]. In the present study, the slit that was identified in median sagittal section was shown to be a structure that can separate the movements of the genioglossus muscle and the geniohyoid muscle. That is, the glossoepiglottic tendon in the upper layer of the slit transmits the relaxation of the genioglossus muscle to the upper part of the epiglottis, and the hyoepiglottic ligament in the lower layer of the slit transmits the contraction of the geniohyoid muscle to the lower part of the epiglottis. This suggests that these two opposing movements may generate a rotational force that causes the retroversion of the epiglottis. We also considered that the relaxation of the genioglossus muscle and the glossoepiglottic tendon may be involved in the previously reported extension of the hyoepiglottic ligament during the first epiglottic movement. In the second epiglottic movement, it is the forward movement of the hyoid bone that reportedly causes the hyoepiglottic ligament on the lateral side to draw the tip of the epiglottis forward and downward, so that it completely covers the laryngeal aperture [17]. In this study, the hyoepiglottic ligament was also identified as a broad, strong ligament in the lateral sagittal section, a result that supports this mechanism for the second epiglottic movement. However, the effects on the second epiglottic movement of pressure on the posterior wall of the pharynx and of the thyroepiglottic and aryepiglottic muscles must also be taken into account [15].

At rest, the swallowing reflex is inhibited by the filling of the valleculae with saliva close to the tongue root [32]. This function reportedly prevents the swallowing reflex from becoming excessive during sleep at night, maintaining the pharynx and larynx in a relaxed state, but the detailed mechanism of this movement has yet to be identified [33]. In this study, the glossoepiglottic tendon was found to be located in the submucosa of the glossoepiglottic fold, and the hyoepiglottic ligament was found to be located in the submucosa of the epiglottic valleculae. From these results, at rest, the glossoepiglottic tendon is tensed and the glossoepiglottic fold is clearly evident, enabling saliva to move smoothly into the epiglottic valleculae. This suggests that, in its resting position, the hyoid bone and the hyoepiglottic ligament form a deep epiglottic valleculae that collects saliva.

The genioglossus muscle is classified as an extrinsic lingual muscle, and it is found as a pair of muscles on either side of the lingual septum. The genioglossus muscle runs in the sagittal direction, tracing an arc from the mental spine toward the tip of the tongue, the dorsum of the tongue, and the tongue root [34, 35]. In this course, the posterior portion running toward the root of the tongue is thicker than the rest of the muscle, and it was previously regarded as a different muscle separated by connective tissue [36]. The posterior part of the genioglossus muscle also functions to move the hyoid bone and tongue root forward, dilating the upper airway, and numerous studies have described its involvement in respiration and swallowing actions [37‐40]. In addition, many studies of muscle attachment have been reported [41, 42], along with a paper describing a short-term experiment on the tongue muscle in mice [43]. In terms of the human oral cavity, there are reports on the soft palate and pharynx [44, 45]. However, no study has yet described the morphology of the terminus of the posterior part of the genioglossus muscle, and the mechanism of upper airway dilation remains unknown. In the present study, the existence of the glossoepiglottic tendon was identified, and the terminus of the genioglossus muscle was found not to be the hyoid bone. From these findings, we consider that the contraction of the posterior part of the genioglossus muscle directly draws the epiglottis forward, and this dilates the upper airway. These results suggest that this function works at rest and during sleep in particular and that the genioglossus muscle plays a role as an anti-gravity muscle to keep the upper airway consistently open.

Gross and microscopic observations of the submucosal structures from the tongue to the larynx of 20 elderly cadavers were performed. As a result, in the upper layer of the hyoepiglottic ligament, a new tendon that connects the genioglossus muscle and the epiglottic cartilage, named the glossoepiglottic tendon, was discovered. Interestingly, there was a slit between the hyoepiglottic ligament and the glossoepiglottic tendon that can separate the movements of the two fiber bundles. These results suggest that the genioglossus muscle may be involved in keeping the epiglottis upright and in its retroversion. Therefore, the function of the genioglossus muscle is a very important item to evaluate for rehabilitation of dysphagia.