Evaluation of swallowing function in patients with oropharyngeal secretions

Abstract Background Several studies have reported a strong association between the presence of oropharyngeal secretions in the laryngeal vestibule and the likelihood of aspiration of food or liquid. However, no previous studies have evaluated the accumulation of saliva and swallowing dynamics. Objective The objective of this study was to examine the factors related to decreased function that result in saliva accumulation based on images from videofluoroscopic examination of swallowing (VF) performed on the same day as videoendoscopic examination of swallowing (VE). Methods This retrospective study investigated 47 patients with dysphagia who underwent VF and VE on the same day. Saliva accumulation in the pharynx was assessed on VE and classified by the Murray secretion scale. Pharyngeal residue was assessed on VF. In addition, displacement of the hyoid bone and larynx on swallowing and the opening size of the esophageal orifice were measured, and contact between the base of the tongue and the posterior pharyngeal wall was examined on VF. Results Moderate correlations were found between saliva accumulation and perpendicular displacement of the larynx and upper esophageal sphincter opening. The percentage of patients showing contact between the base of the tongue and the posterior pharyngeal wall was significantly greater in those with a saliva accumulation score of 0 or 1. Conclusion Less laryngeal elevation and upper esophageal sphincter opening and absence of contact between the base of the tongue and the posterior pharyngeal wall when swallowing tended to result in accumulation of saliva in the pharynx.


| INTRODUCTION
Although saliva in the oral cavity is an important component in swallowing, saliva also plays a critical role as a medium for carrying oral bacteria to the lower respiratory tract. (Inglis, Sherratt, Sproat, Gibson, & Hawkey, 1993) Several studies have reported a strong association between the presence of oropharyngeal secretions in the laryngeal vestibule and the likelihood of aspiration of food or liquid. (Link, Willging, Miller, Cotton, & Rudolph, 2000;Mikushi et al., 2015;Murray, Langmore, Ginsberg, & Dostie, 1996;Takahashi, Kikutani, Tamura, Groher, & Kuboki, 2012), (Takahashi et al., 2012) in a study in nursing care facilities, showed that aspiration of saliva detected by videoendoscopy was a significant risk factor for pneumonia. Silent aspiration and laryngeal penetration of saliva have been postulated to occur in the elderly (Teramoto et al., 2008) and patients with Parkinson's disease, (Ebihara et al., 2003;Gross et al., 2008;Rajaei et al., 2015) leading to aspiration pneumonia. (Kohno et al., 2013) Decreased sensitivity of the oropharynx might be responsible for accumulation of saliva in the pharynx and subsequent silent aspiration and laryngeal penetration of saliva. However, little is known about the pathophysiological factors affecting saliva aspiration or laryngeal penetration of saliva. (Rodrigues, Nobrega, Sampaio, Argolo, & Melo, 2011) showed that the presence of hypoesthesia of the laryngeal structures and diminished protective reflexes might play major roles in the mechanisms underlying silent aspiration and silent laryngeal penetration. We have also hypothesized that reduced function of the oropharynx (i.e., incomplete pharyngeal constriction and impaired opening of the upper esophageal sphincter) may lead to accumulation or aspiration of saliva, similar to aspiration of food. No previous studies have evaluated the accumulation of saliva and swallowing dynamics. The present study examined the factors related to decreased function that result in saliva accumulation based on images from videofluoroscopic examination of swallowing (VF) performed on the same day as videoendoscopic examination of swallowing (VE).

| Subjects
This retrospective study investigated 47 patients with dysphagia who underwent VF and VE on the same day between October 2014 and January 2015 at Nagasaki University Hospital (Table 1). The patients were 35 men and 12 women, with a mean age of 72.8 ± 11.2 years (range, 27-97 years). Conditions underlying dysphagia were: head and neck tumor (n = 21), respiratory disease (n = 7), esophageal disease (n = 4), cerebrovascular disease (n = 3), Parkinson disease (n = 2), and others (n = 10).

| VF and VE
For VF examinations, an X-ray fluoroscopic table (C-vision Safire L, Shimadsu, Kyoto, Japan; 30 frames/sec) was used. The examiner served test foods into each patient's mouth. Patients were asked to swallow as usual in the sitting position. Test foods were made with a barium solution (Barium: barium/water 50/50% weight/volume, Barytgen HD, Fushimi Pharmaceutical Co., Ltd., Kagawa, Japan). Thickened barium solution was made using a thickening agent (Toromelin V, Nutri, Mie, Japan). A radiopaque ball (10 mm in diameter) was attached on the left side of the neck to calibrate the image for mechanical analyses.
For VE examinations, the following systems were used (Fiberscope: FNL-10RBS, Halogen light source: CLH-SC, CCD camera: PSV-4000, video system: OTV-SC, Olympus, Tokyo, Japan). Subjects were seated comfortably in the chair. The examiner applied a topical anesthetic to the nasal mucosa and inserted the endoscope to observe saliva and the behavior of the pharynx when swallowing and coughing. Observation was performed for about 5 min. VF and VE data were recorded with a DVD recorder (DIGA, DMR-EH50, Panasonic, Osaka, Japan) and copied onto a DVD.

| Data collection and analysis
Patient images were collected from the VF and VE image database and analyzed using a personal computer (iMac, Apple, Cupertino, CA, USA). They were analyzed using the slow motion and stop-frame function of Quicktime software (Version 7.76, Apple). Images were captured in JPEG format and measured by the software (Adobe Photoshop, Adobe Systems Inc., San Jose, CA, USA). Two dentists evaluated and measured the images. They had more than 8 years of clinical experience with eating and swallowing disorders, and they had training in the evaluation of dysphagia after graduation and had treated dysphagia patients. We evaluated the accumulation of saliva in the pharynx from VE, the pharyngeal residue of test food, the displacement of the hyoid bone and larynx on swallowing, the opening of the esophageal orifice, and the contact between the base of the tongue and the posterior pharyngeal wall from VF. Then, the correlations between saliva accumulation and the others were analyzed, respectively. We isolated head and neck tumor patients and analyzed them separately.

| Accumulation of saliva in the pharynx
Saliva accumulation in the pharynx was assessed based on VE images and classified into four levels according to the method described by (Murray et al., 1996) (Table 2). To measure the reliability of the scale, interrater and intrarater reliabilities were calculated with the kappa coefficient. The interrater reliability between the two raters was 0.75, and the mean intrarater reliability was 0.83. Substantial agreement was observed in the interrater and intrarater reliabilities of this scale. Differences in their ratings were discussed by the raters to reach a consensus on the final evaluation value.

| Pharyngeal residue of test food
Pharyngeal residue was assessed from the lateral view when 4 ml of honey-like thickness barium were consumed for VF. The participants swallowed in accordance with the examiner's instructions.
Each participant swallowed 4 ml of honey-like thick liquid once only, and the amount of residue was evaluated after this one swallow. Swallowing dynamics were evaluated as specified by the Japa-

| Opening of the esophageal orifice
The opening size of the esophageal orifice was measured in the lateral view at the narrowest point of the opening between C3 and C6 during maximal distention for bolus passage when swallowing 4 ml of honey-like thick liquid for VF. (Kendall & Leonard, 2002)

| Contact between the base of the tongue and the posterior pharyngeal wall
Contact between the base of the tongue and the posterior pharyngeal wall was examined on the lateral view when 4 ml of honeylike thick liquid were swallowed for VF. Following the method described by Fujiu and Logemann,(Fujiu & Logemann, 1996)  1 Any secretions evident upon entry or following a dry swallow in the channels surrounding the laryngeal vestibule that were bilaterally represented or deeply pooled. This rating would include cases where there is a transition in the accumulation of secretions during the observation segment. A subject could start with no visible secretions but accumulate secretions in an amount great enough to be bilaterally represented or deeply pooled. Likewise, a subject would be rated as a "1" if initially presenting with deeply pooled bilateral secretions and ending the observation segment with no visible secretions.
2 Any secretions that changed from a "1" rating to a "3" rating during the observation period. 3 Most severe rating. Any secretions seen in the area defined as the laryngeal vestibule. Pulmonary secretions were included if they were not cleared by swallowing or coughing at the close of the segment. presence or absence of contact was evaluated. Saliva accumulation was divided into those with a score of 0 or 1 and those with a score of 2 or 3, and the percentage of patients with contact was compared between these two groups.

| Statistical methods
Statistical analyses were performed using JMP (Version 13.0, IBM, Armonk, NY, USA). Associations between saliva accumulation and each assessed value were analyzed using Spearman's rank correlation coefficients. Values of p < .05 were considered significant. Contact between the base of the tongue and the posterior pharyngeal wall was compared using Fisher's exact test.

| Ethics
This study was conducted with the approval of the Ethics Committee of the Nagasaki University Hospital (approval number: 17041701). All participants gave written, informed consent before this study.

| Saliva accumulation and pharyngeal residue of test food
The mean score for residue was 2.6 ± 0.6 at the epiglottic vallecula and 2.7 ± 0.6 at the pyriform sinus. Spearman's rank correlation coefficients with saliva accumulation were −.308 and −.351, respectively.
There was a weak correlation between saliva accumulation and pharyngeal residue. A significant correlation with residue in the pyriform sinus was observed. In other words, pyriform sinus residue during VF increased with greater saliva accumulation and saliva penetration to the laryngeal vestibule on endoscopy.

| Saliva accumulation and displacement of the hyoid bone and larynx on swallowing
Mean hyoid bone displacement on swallowing was 7.4 ± 3.8 mm in the horizontal direction and 9.9 ± 5.5 mm in the perpendicular direction. Spearman's rank correlation coefficients with saliva accumulation were −.384 and −.250, respectively. There was a weak correlation between saliva accumulation and displacement of the hyoid bone. A significant correlation with horizontal displacement of the hyoid bone was observed (Figure 2). Mean displacement of the larynx on swallowing was 5.3 ± 3.9 mm in the horizontal direction and 18.8 ± 6.7 mm in the perpendicular direction. Spearman's

| Saliva accumulation and opening of the esophageal orifice
Mean upper esophageal sphincter opening on swallowing was 5.8 ±

| Head and neck tumor patients
In 21 patients with head and neck tumor, the score of Murray's secretion severity rating scale was 0 in two patients (9.5%), 1 in six patients (28.6%), 2 in four patients (19.0%), and 3 in nine patients (42.9%). The mean score for pharyngeal residue was 2.6 ± 0.6 at the epiglottic vallecula and 2.5 ± 0.7 at the pyriform sinus. Mean hyoid bone displacement on swallowing was 6.9 ± 3.2 mm in the horizontal direction and

| Accumulation and aspiration of saliva
The combination of saliva with food that has been masticated into small pieces forms a bolus that is easily swallowed. Saliva in particular plays a key role when low-moisture foods are consumed. In the oral cavity, countless bacteria are present and become incorporated with saliva, which is secreted not only during meals, but also at rest. Regular swallowing of saliva that has accumulated in the oral cavity and pharynx is therefore necessary. However, similar to all other fluids, there is a risk of saliva aspiration. Previously, saliva aspiration during bedtime has been reported as the phenomenon of "microaspiration." About half of healthy individuals (Gleeson, Eggli, & Maxwell, 1997;Huxley, Viroslav, Gray, & Pierce, 1978) and 71% of patients with community-acquired pneumonia (Kikuchi et al., 1994) show saliva aspiration while sleeping. Moreover, saliva aspiration can also occur when awake. A study that performed videoendoscopic swallowing tests in nursing home residents showed saliva aspiration in 12.1% of residents when they were awake, (Takahashi et al., 2012) and a different report showed saliva aspiration in 46% of outpatients.(Schröter-Morasch, Bartolome, Troppmann, & Ziegler, 1999) The videoendoscopic swallowing test that we performed for inpatients also showed saliva aspiration in 34.2% of dysphagic patients. (Mikushi et al., 2015) However, although the influx of saliva into the trachea with endoscopy can be viewed in some patients while awake, endoscopic evaluation of saliva aspiration when the volume involved is too small to view is diffi-

cult. Rodrigues et al. examined saliva aspiration in patients with
Parkinson's disease by applying pigment in the oral cavity. (Rodrigues et al., 2011) In particular, saliva aspiration that follows along the interarytenoid notch to the posterior wall of the trachea is difficult to evaluate due to restrictions in the angle of endoscopic viewing. For this reason, saliva accumulation, rather than aspiration, was used as the objective variable in this study. Clinically, patients with a large amount of saliva accumulation are predicted to also show saliva aspiration. proportion was similar to this study. The score of 3 on Murray's evaluation scale in the present study signifies constant saliva accumulation in the laryngeal vestibule, and saliva aspiration may thus be occurring in these patients. In the present study, 15 of 47 patients (31.9%) had a score of 3. In the future, we plan to clarify the association between the extent of saliva accumulation and aspiration of saliva.

| Saliva accumulation and food residue
On the basis of our experience, it appears that patients with accumulated saliva on endoscope insertion also show decreased pharyngeal function when swallowing food. The present results also showed that patients with accumulated saliva have a large amount of food residue in the pyriform sinus. Murray et al. showed that the percentage of food and liquid aspiration increases with greater severity of saliva accumulation, and that saliva accumulation is a predictor of aspiration. (Murray et al., 1996) Link et al. reported that saliva accumulation was associated with a history of pneumonia, laryngeal penetration of food, and aspiration in pediatric patients with neurologic disorders. (Link et al., 2000) In addition, we have previously reported that 46% of patients with saliva aspiration showed aspiration of honey-like thick liquid, and that aspiration after swallowing was observed in ≥70% of aspirations observed during VE. (Mikushi et al., 2015) Eisenhuber et al.
stated that 89% of patients with severe pharyngeal residue of food showed food aspiration after swallowing. (Eisenhuber et al., 2002) Food residue is also likely to be observed at the pyriform sinus in patients with accumulated saliva, indicating an elevated risk of aspiration. Patients with accumulated saliva are likely to show food aspiration after swallowing, signifying the importance of managing pyriform sinus residues and being aware of aspiration after swallowing during direct training and meals.

| Factors that cause saliva accumulation
The accumulation of saliva has been reported to be associated with decreased laryngeal sensation and a reduced number of swallows. (Gross et al., 2008;Rajaei et al., 2015;Rodrigues et al., 2011) (Olsson, Castell, Johnston, Ekberg, & Castell, 1997) The results of the present study showed that saliva accumulation was associated with perpendicular displacements of the larynx, contact between the base of the tongue and the posterior wall of the pharynx, and opening size of the upper esophageal sphincter. Perpendicular displacements of the larynx and opening size of the upper esophageal sphincter were also associated with saliva accumulation in head and neck tumor patients. (Logemann et al., 2000) reported that healthy older individuals showed hyoid bone displacement of 8.47 ± 1.05 mm in the horizontal direction and 14.58 ± 1.46 mm in the perpendicular direction and laryngeal displacement of 6.18 ± 0.93 mm in the horizontal direction and 24.25 ± 1.58 mm in the perpendicular direction. All of these values were larger than the present findings; we postulated that this was due to the inclusion of dysphagia patients in the present study. Regarding upper esophageal sphincter opening, (Kendall & Leonard, 2002) reported that in ≥65-year-old dysphagia patients, an opening of 7.6 ± 0.27 mm was observed with 1 ml of liquid and 8.1 ± 0.31 mm with 20 ml of liquid. Hattori et al. (Hattori, 2004) reported that in healthy older individuals, an opening of 8.18 ± 2.12 mm was observed with the consumption of 8 ml of paste. In the present study, 4 ml of paste were used, and the mean opening was 5.8 ± 2.3 mm, indicating a smaller opening compared with the above reports.
Upper esophageal sphincter opening requires relaxation of the sphincter and mechanical opening through pharyngeal pressure and laryngeal elevation on swallowing. The present results demonstrated that many patients with accumulated saliva did not show contact between the base of the tongue and the posterior wall of the pharynx on swallowing, indicating a deficiency in pharyngeal pressure. In the future, we plan to conduct manometry and measure tongue pressure to confirm this finding of insufficient pharyngeal pressure on swallowing. In addition, deficient laryngeal elevation are linked to deficient upper esophageal sphincter opening, causing saliva accumulation similarly to food accumulation. These findings indicated that saliva accumulation is associated not only with decreased sensation, but also with decreased pharyngeal function. To reduce saliva accumulation, improvement of laryngeal elevation, pharyngeal contraction, and upper esophageal sphincter opening is crucial, similar to the approach for reducing food residue in the pharynx.

| Limitations of the study
The limitations of this study are that it was a retrospective evaluation, used only one thick-liquid swallowing measurement that was performed during a clinical examination, and included a small sample size, resulting in poor reliability. With a larger sample size, comparisons by sex, age, and underlying disease can be carried out. We therefore plan to increase the number of study participants in the future. The evaluation of residue was conducted using a method that is used only within Japan. Furthermore, all study participants were Japanese; thus, there may also be differences in measurements based on ethnicity.

| CONCLUSION
Pharyngeal malfunction was associated with saliva accumulation in the pharynx. The tendencies of pharyngeal function that may lead to saliva accumulation included: small vertical laryngeal displacement during swallowing, small opening of the esophageal orifice, and absence of contact between the base of the tongue and the posterior pharyngeal wall.