The Role of the Epiglottis in the Swallow Process after a Partial or Total Glossectomy Due to a Neoplasm

The subjects were 95 patients, including 28 women aged 20-78 years (mean = 46.3) and 67 men aged 9-79 years (mean = 53.9). All had undergone surgical treatment for an oral cavity neoplasm at the Clinic of Maxillofacial Surgery at the Pomeranian Medical University which consisted of a partial or total glossectomy. None of our patients had ever had a condition that affected swallowing efficiency diagnosed or treated before. Hemiglossectomy was performed in 44 patients (46%), a nearly total glossectomy in 12 (12.6%), and a total glossectomy in 6 (6.3%). In all those cases, the quality of the videofluoroscopic records of the act of swallowing allowed for both the evaluation of the epiglottic movement pattern and determination of the time sequence of all actions involved in it. Neither the presurgical medical interview nor the clinical assessment revealed a condition, other than the oral cavity neoplasm, that might involve swallowing disorders. No videofluoroscopic study had been done before the surgery. The extent of both the oral tissue excision and the reconstruction performed as a one-step surgery are presented in Table 1. Tongue reconstruction was carried out in eight patients, and three patients had restoration of the tongue and floor of the mouth. Frontal, pectoral, and microvascular free flaps from the forearm were used for reconstruction purposes. The mandibular arch was maintained in 13 cases by means of a mandibular locking plate which stabilized in a nearly anatomical position the parts of the bone remaining after a partial mandibulectomy and, at the same time, served as a scaffold for suprahyoid muscles or parts thereof. All the other reconstructions were performed by means of tissues near the postoperative bone defect.

Videofluoroscopic evaluation of swallowing was carried out between 2 and 4 weeks postoperatively in patients who had complaints of difficulty swallowing saliva, in whom the extent of tissue excision implied a severe swallowing dysfunction, and in those who were afraid of the removal of the feeding tube and reinitiation of normal feeding.

The images obtained by means of the EDR 700 model (X-ray unit, Medicor Co., Medical Röntgen RT, Budapest, Hungary) were video-recorded. The recordings were made on magnetic video tape using VHS at the rate of 24 frames per second. This enabled us to watch multiple replays of the recording, including slow motion and frame-by-frame. The digital real-time recordings were saved and stored as JPEG files on a computer hard disc, which made it possible for us to evaluate the time sequences of the specific components of the recording. The images taken of the oral and pharyngeal phases of swallowing were both lateral and anterior-posterior views. The patients were seated comfortably with the head positioned to facilitate to feeding. The study started with the lateral view of the oral cavity (when possible), nasopharynx, midpharynx, laryngopharynx and the cervical part of the esophagus. When it was impossible to have all the above-mentioned structures visible in one image, the oral and pharyngeal phases of swallowing were studied on two different images. One image showed the whole oral cavity, nasopharynx, midpharynx, and part of the laryngopharynx. The other image was supposed to show the posterior part of the oral cavity, midpharynx, laryngopharynx and the cervical part of the esophagus.

For the radiological studies we used a barium suspension. For the first swallowing study, a thin suspension in an amount not exceeding 5 ml was used. If the patient undergoing assessment suffered from a massive aspiration, the study was immediately terminated. If there was no risk of serious aspiration, the patients underwent further swallowing evaluations carried out by means of an increased amount of suspension ranging from 20 to 25 ml. During the evaluations, the consistency of the suspension was pudding- or paste-like and the patients were free to decide how many times they wanted to swallow.

The swallowing act was analyzed based on findings included in two reports by Dodds et al. [7, 8] in which the methods of assessment of both normal and abnormal oral and pharyngeal phases of swallowing were described. The following criteria were taken into account: labial closure, palatopharyngeal closure, oral transit of the food bolus, mobility of the tongue, postswallow oral residue, additional swallows used for emptying the oral cavity, oral trajectory of the bolus, body posture and head movement during the swallows, preswallow leakage, mobility of the epiglottis, glossopharyngeal closure, pharyngeal trajectory of the bolus, additional swallows used for emptying the laryngopharynx, postswallow laryngopharyngeal residue, postswallow leakage into the lower airways, protection of the lower airways, intraswallow leakage, and aspiration.

The records were analyzed instantly on a real-time basis by a team comprising the patient, two radiologists, and a clinician (maxillofacial surgeon). The radiologists stratified the swallowing competence of patients using such descriptions as:

The above-mentioned evaluation phase was aimed at finding the answer to the following question: “Can we initiate oral feeding of the patient or shall we maintain tube-feeding while the patient’s re-education of swallowing is carried through?”

The video records were analyzed frame-by-frame without the participation of the patient. A qualitative evaluation of the swallowing act was carried out, including such parameters as mobility of the tongue and epiglottis; labial, palatopharyngeal, and glossopharyngeal closures, and postswallow pharyngeal residue. A value from 1 to 4 was assigned to each aspect, where 1 was “normal” and 4 was “absent.” The criteria of evaluation also included patients’ postures during the swallows, additional swallows affected by them while holding their breath, and the bolus trajectory. The said criteria were assigned values from 1 to 4 as well. Quantitative features of the act of swallowing, such as the moment of initiation, termination, and the duration of the formation of the bolus, oral and pharyngeal transfer, laryngeal closure, and opening of the upper esophageal sphincter, were described as moments in time or time spans. The point of reference for time measurements was referred to as 0.00 s and corresponded to initiation of the pharyngeal phase.

The evaluation phase mentioned above was aimed at identifying circumstances in which the penetration and/or aspiration occurred and for formulating indications for the re-education of swallowing.

The maxillofacial surgeon used her own 100-point scale [9] during the course of the observations of both the patient and the videofluoroscopic image of the swallowing act while keeping in mind the postoperative tissue damage, the compensatory mechanisms used by the patient, and the symptoms of impaired swallowing such as coughing, choking, and drooling. For the purpose of this evaluation, ten composed actions performed by the oral cavity, throat, and esophagus structures were taken into account. The activities on which swallowing efficiency depends include:

The maximum number of points to be scored for each of those activities was 10. When an activity was either not completed or was completed by means of supplementary mechanisms, 5 points were scored, and when a specific activity was absent, 0 points were given. All of the points scored were then summed up. The total score was the outcome of both a clinical and a radiological evaluation of the swallowing efficiency determined at the end of each examination.

A joint analysis of the descriptive evaluation results and the number of points scored revealed a repeatable regularity, and thus the established radiological description “incompetent” corresponded with a score of 0-20 points, “poorly functional” with 25-45 points, “moderately functional” with 50-65 points, “functional” with 70-85 points, and “correct” with 90-100 points.

Discontinuous variables were verified by means of a χ² test, continuous variables by means of the Mann-Whitney test, and correlations among variables by means of the Spearman’s rank correlation coefficient. The results were described using R correlation coefficients and P probability. In all of the tests, P < 0.05 was considered to be statistically significant while 0.05 < P < 0.10 was the limit of statistical significance.

Normal epiglottic movement (horizontal position of the epiglottis with a complete downfolding of its free border) was observed in 68 of 95 cases (72%), horizontal tilt in 9 cases, hardly detectable posterior translocation in 5 cases, and absence of movement in 10 cases. Movements difficult to assess were registered in three patients in whom the margin required for oncological safety reached pre-epiglottic valleculae. The three patients were not taken into account during evaluation of the results. No statistically significant relationship was found to exist between the sex or age of the patients, time since surgery, or type of neck surgery involving the lymphatic system and epiglottic movement patterns. The extent of excision of the oral cavity structures (Fig. 1) exerted a statistically significant influence on epiglottic movement (P = 0.00637).

In 68 patients (72%) who had undergone surgery for oral cavity cancer, irrespective of the extent of tissue excision a normal epiglottic movement was observed. In most cases, normal epiglottic movement occurred following the excision of the lateral part of the tongue, floor of the mouth, and adjacent parts of the mandibular body. Actions involved in the oral phase of swallowing such as tight lip closure, bolus formation, oral transfer, or glossopalatal closure were impaired but still the airway closure was tight.

Penetration occurred in 13% of patients. No aspiration cases were found. Postswallow residue was noted in 19% of subjects, mostly in the pre-epiglottic valleculae and pyriform sinuses. Nineteen percent of the patients changed the position of their head with respect to their body while swallowing and 15% used additional swallows (Table 2). Despite a delay in the initiation of the pharyngeal phase, the swallowing was “functional” (Table 3).

Abnormal epiglottic movement was observed in 24 (26%) patients. In 9 of them the epiglottis tilted horizontally, in 5 a hardly detectable posterior translocation was recorded, and in 10 cases it did not move at all. The horizontal tilt was found mostly in subjects who underwent hemiglossectomy and hemimandibulectomy. A total glossectomy was performed on 6 patients and a nearly total glossectomy was performed on 12 patients, although we failed to evaluate 3 of them. In 9 of the remaining 15 patients, the epiglottic movement was found to be normal and in 6 no movement was observed. Actions involved in the oral phase of swallowing such as tight lip closure, bolus formation, oral transfer, and glossopalatal closure were considerably impaired or absent. Airway closure was inefficient. Penetration was recorded in 67% of people with abnormal epiglottic movements and aspiration was recorded in 42%. Aspiration was found mainly in those whose epiglottis remained immobile in the course of swallowing. The presence of postswallow residue in the pre-epiglottic valleculae, pyriform sinuses, and pharynx was noted in 62.5% of subjects. In 42% of cases patients changed the position of their head with respect to their body while swallowing, and in 63% they used additional swallows (Table 2). A delay in the initiation of the pharyngeal phase of deglutition was 2.5 times longer and swallowing efficiency was almost 2.5 times worse than those in patients enjoying normal epiglottic movements (Table 3).

A joint analysis of epiglottic movement patterns and the moments of initiation and termination as well as the duration of the acts constituting the oral phase of swallowing enabled us to notice the following correlation: the more abnormal the epiglottic movement pattern was, the more delayed the laryngeal closure and the pharyngeal sphincter opening were, the longer the time needed for emptying the throat was, and the longer the pharyngeal phase and swallowing apnea (although statistically insignificant) were (Table 3).

The correlation between the swallowing competence evaluation carried out by means of a 100-point scale and the epiglottic movement pattern (R = -0.50, P = 0.0000) was highly statistically significant. An average rating of swallowing competence for patients with normal epiglottic movements was 2.5 times higher than its value for subjects with an absence of epiglottic movements (Table 3).

Following a partial or total glossectomy patients were tube-fed in the early postoperative period. At the time preceding the videofluoroscopic evaluation of swallowing, most had already started to apply their own methods of saliva swallowing without the risk of aspiration. Some of them had even made attempts to drink liquids disregarding the feeding tube. Validity of the compensatory mechanisms used by patients was assessed during the videofluoroscopic studies of the swallowing act. As a result, in several cases patients were advised against using ineffective compensatory mechanisms.

Swallowing with the head tilted backward was noted in 16 patients in whom the oral transfer proved to be incompetent. The patients used this maneuver by their own initiative. It enabled them to transfer the bolus from the oral cavity to the throat and to swallow it safely. In four patients with no epiglottic movements, a preswallow leakage was noted, which then turned into aspiration. In all those cases tube-feeding was temporarily maintained (Fig. 2).

Swallowing with the head tilted downward was noted in 7 patients in whom there was no glossopharyngeal closure. In five cases this maneuver enabled the patient to swallow safely, but in two cases in which the epiglottic movements were absent, intraswallow leakages occurred, which then turned into aspiration. In those cases the tube-feeding also was maintained (Fig. 2).

In all of the examined subjects, swallowing one dose of liquid was accompanied by an apnea lasting 0.15-8.08 s. A prolonged apnea (> 0.6 s) was noted in 74 patients. An apnea lasting 0.61-1.19 s was reported in the cases in which the highest swallowing competence (irrespective of the epiglottic movement pattern) was observed. The said apnea duration was more important for the swallowing competence in patients with absent or hardly detectable epiglottic movements. In subjects with absent epiglottic movements, a higher swallowing competence was found to be accompanied by an apnea lasting more than 1.20 s (Fig. 3).

Additional swallows were used by 25 patients who attempted to clear the pharynx during one apnea episode. The more profound the epiglottic movement impairment was, the more the subjects used additional swallows. In the case of an absence of epiglottic movements, additional swallows did not always ensure pharynx emptying before the opening of the larynx. In such circumstances, the compensatory mechanism did not protect the patients against aspiration and accompanied incompetent swallowing (Fig. 4).