Auris· Nasus· Larynx (Tokyo) 12 (Supp\. II) S 200-S 205,1985

PRIMARY TRACHEOESOPHAGEAL SHUNT OPERATION FOR POSTLARYNGECTOMY VOICE RESTORATION

-A SPHINCTER MECHANISM AGAINST ASPIRATION-

Mutsuo AMATSU, M.D., Kunihiko MAKINO, M.D., Minoru KrNISHI, M.D., Mitsutake T ANI, M.D., and Michiyo KOKUBU, M.D.

Department of Otorhinolaryngology, Kobe University School of Medicine, Kobe, 650 Japan

Out of a total of 113 consecutive tracheoeso­phageal (TE) shunt operations for postlaryngec­tomy voice restoration in the past 8 years performed at the Department of Otorhinolaryn­gology, 92 patients (81 %) succeeded in the post­operative TE speech. The essential part of this surgery consists of the construction of the TE shunt using the membraneous part of the trachea obtained at surgery. In the course of 8 years, important changes have been employed for the prevention of aspiration. We attempted to combine primary cancer surgery with the crea­tion of intelligible voice, but without aspiration. For the above purpose, we have employed the bilateral esophageal muscle flaps (BEMF) against aspiration in combination with the TE shunt construction for phonation. Sixteen of 18 pa­tients thus operated on complained of no aspira­tion even with a drop of saliva and dietary fluids. As far as the mechanism against aspiration is concerned, both dilatation and elevation of the cervical esophagus during deglutition, together with the BEMF, seem to approximate the sphinc­ter mechanism against tracheal reflux. A proper case selection may achieve high success rates for preserving normal deglutition and restoring speech after total laryngectomy.

The treatment of laryngeal cancer consists of both complete removal of the lesion and restoration of the voice. Since the progno­sis of laryngeal cancer can be highly guar­anteed by total removal of the larynx, resto­ration of the voice is indispensable for the better quality of the postoperative life.

Over the past 20 years, a number of surgi­cal voice restoration methods for this pur-

pose have been reported (CONLEY et ai., 1958; ASAI, 1965; CALCATERRA and JAFEK, 1971; KOMORN, 1974; STAFFIERI and SERA­FINI, 1976; AMATSU et ai., 1977; SINGER and BLOM, 1980). The principle behind these procedures is to divert the powerful pulmo­nary air into the pharynx or esophagus, to achieve an intelligible voice. The resultant speech compared satisfactorily with esopha­geal speech which had been the most fre­quently used method of vocal rehabilitation after total laryngectomy (ROBBINS et ai., 1982, 1984; AMATSU et ai., 1984a; KINISHI and AMATSU, 1985). However, the greatest limitation in these shunt procedures is aspi­ration inherent in any surgical methods. The only method reported till now incorporating a mechanism against aspiration is the use of prosthetic device with an one way valve (SINGER and BLOM, 1980; PANJE et ai., 1981).

During the past 8 years, we have been performing the primary tracheoesophageal (TE) shunt operation using the membraneous part of the trachea with satisfactory phona­tory function (AMATSU et al., 1977; AMATSU, 1978, 1980; AMATSU et al. 1984a, b), and important changes in the procedure have been directed at minimizing aspiration. This paper discusses a study of 113 consecutive cases which are divided into 3 categories according to the changes in the procedure employed for the prevention of aspiration.

Operative Techniques The membraneous part of the trachea ob­

tained at laryngectomy is used for the con-

M. AMATSU et al. S 201

struction of the TE shunt. To minimize aspiration, some changes in the procedure have been instituted during recent years.

1. Original method. Usually the larynx is removed just above the first tracheal ring. Anterior and lateral portion of the trachea above the inferior level of the tracheal open­ing are cut off. The remainder of the tra­cheal cartilage is carefully dissected from the tracheal mucosa on both sides. A tracheal flap attached to the esophagus, 3 cm in length and 2 cm in width, can be obtained.

With a sharp knife an 8 mm midline inci­sion is made vertically, starting just below the upper end of the flap until the esophageal lumen is entered. The incised margin of the esophageal mucosa is carefully approximated

Fig. 1. Creation of side-to-side anastomosis (in­ternal orifice of the shunt). a, side-to-side anastomosis; b, tracheal mucosa; c, esopha­geal mucosa; d, esophageal lumen.

Fig. 2. Construction of TE shunt.

to that of the tracheal flap to create a side­to-side anastomosis (internal orifice of the shunt) (Fig. 1). After a rubber catheter (12th Fr.) is inserted through this opening into the esophagus, both lateral margins of the tracheal flap are approximated to make a mucosal tube (TE shunt) with 4-0 chromic catgut (Fig. 2). Suction drains are placed effectively, and skin is closed. The rubber catheter is anchored to the chest wall for 2 weeks postoperatively before the voice pro­duction is initiated.

2. Modified method (1): a deviated and collapsed shunt. The fundamental goal of this technique is to construct a laterally de­viated and collapsed shunt that opens only by exhalation for voice production but that does not allow fluid leak into the trachea during deglutition. This modification con­sists of the following two points: (1) the

Fig. 3. The right stomal end of the tracheal flap is cut slightly upward until midline.

Fig. 4. The stomal end of the constructed shunt is sutured to the left stomal margin.

S202 M. AMATSU et al.

right stomal end of the tracheal flap is cut slightly upward until midline (Fig. 3); (2) the stomal end of the constructed shunt is sutured to the left stomal margin (Fig. 4). Phonation is initiated on the 14th postopera­tive day by occluding the tracheostoma with a finger to divert the exhaled air through the TE shunt into the esophagus, where the vi­bration occurs.

3. Modified method (2),' a sphincter mech­anism with bilateral esophageal muscle flaps ( BEM F ). On completing the larynx resec­tion, 5 cartilaginous tracheal rings are re­moved to obtain a 2.5 cm tracheal flap only with the tracheal posterior wall. After mak­ing a 10 mm vertical incision on the upper part of this flap entering the esophagus, a side-to-side anastomosis is created in the same fashion as described above.

Thyroid lobes are elevated with blunt dis­section to expose the esophageal wall. A cranially based esophageal muscle flap, which includes both outer and inner muscle layers (15 mm x 7 mm), is obtained from the post­erolateral wall (Fig. 5).

The base of this flap should not project beyond the level of the side-to-side anasto­mosis. This procedure is performed bilater­ally. After constructing the TE shunt with­out inserting a rubber stent catheter, BEMF are approximated with 3 stitches of 4-0 chro­mic catgut anchoring the TE shunt (Fig. 6). The dead space lateral to the TE shunt is carefully eliminated.

This surgical procedure is contraindicated in subglottic cancer and glottic cancer with

Fig. 5. The superiorly based esophageal muscle flap obtained.

Fig. 6. Approximation of BEMF.

subglottic extension in which 3 or 4 tracheal rings are included with the resected specimen. We also do not advise this surgery with reduced pulmonary reserves. Factors such as age, motivation, and need to speak are considered before this technique is applied. For irradiated patients, a minimum interval of 8 weeks is needed prior to surgery.

Results One-hundred-thirteen patients underwent

the primary TE shunt technique for voice restoration following total laryngectomy dur­ing the past 8 years: 59 patients for the original technique, 36 patients for the modi­fied technique (1), and 18 patients for the modified technique (2). Age ranged 33 years to 77 years. Fifty-eight patients underwent either unilateral or bilateral neck dissection at the time of laryngectomy: 34 were uni­lateral radical, 9 were unilateral elective, 8 were bilateral radical, 4 were unilateral elec­tive, and 3 were ipsilateral radical and con­tralateral elective. One-hundred-one were laryngeal, 1 was mesopharyngeal, and 11 were hypopharyngeal cancers, 38 patients were salvaged by laryngectomy for recurrent cancer following full-course radiation: 60 Gy for 35 patients, and 120 Gy (two times 60 Gy) for 3 patients. The preoperative radiation therapy group consists of 8 patients who underwent laryngectomy more than 4 weeks after planned radiation therapy (40 Gy). Three patients had radiation therapy (40 Gy) after laryngectomy (Table 1).

As far as postoperative shunt speech is

M. AMATSU et al. S 203

concerned, 92 of 113 patients (81 %) suc­ceeded in producing a voice suitable for con­versation: 46 of 59 patients (78 %) in the original method group, 30 of 36 patients (83%) in the modified method (1) group, and 16 of 18 patients (89%) in the modified method (2) group succeeded in the shunt speech (Table 2).

Twenty-one of 113 (19%) failed in shunt speech. Fourteen patients failed due to ste­nosis of the internal orifice of the shunt. In 2 patients operated on with the modified method (2), the shunt closed by circumfer­ential mucosal erosion which was caused by the excessive persistent pressure of the stent catheter placed in situ. Breakdown of the shunt due to infection of a hematoma was

Table 1. Patients undergoing TE shunt opera­tion.

Surgery TE shunt operations Neck dissections

Location of primary lesion Larynx Mesopharynx Hypopharynx

Irradiation Full course Preoperative Postoperative

Number of patients

113 58

101 1

11

38 8 3

seen in 2 patients of the original method group. Pharyngeal stricture was also seen in 2 patients in the same group. One pa­tient failed in the shunt speech due to a loss of motivation (Table 3).

The success rate of the irradiated group exceeded that of the nonirradiated group in the postoperative phonation (Table 4).

Concerning the aspiration, 19 of 46 shunt speakers in the original method group did not complain of aspiration. Although aspi­ration confined to dietary fluids was observed in the remaining 27 patients, it disappeared spontaneously in 9 patients by reducing the volume of each swallow. In 18 patients, aspiration was effectively controlled by ap­plying slight digital pressure on the shunt during deglutition. In the series of the modi­fied method (I) group, 15 of 30 experienced no aspiration, but it was observed in the remaining 15 patients. Eleven of these 15 patients overcame this difficulty with time. Sixteen of 18 patients in the modified method (2) group complained of no aspiration even with a drop of saliva and dietary fluids. In 2 patients of this group, a slight aspiration occurred with liquids. This was effectively controlled by reducing the volume of each swallow in I patient and by applying slight digital pressure on the shunt during degluti­tion in the another (Table 5).

Table 2. Success rate for shunt speech in 113 patients.

Methods Overall cases Success Failure Success rate (X )

Original method 59 46 13 78 Modified method (1) 36 30 6 83 Modified method (2) 18 16 2 89

Total 113 92 21 81

Table 3. Causes of failure in shunt speech.

Causes Original method Modified method (1) Modified method (2) Total

Stenosis Internal orifice 9 5 0 14 Whole tract 2 2

Breakdown of shunt 2 0 0 2 Pharyngeal stricture 2 0 0 2 Loss of motivation 0 0

Total 13 6 2 21

S 204 M. AMATSU et al.

Table 4. Relationship between radiation and success rate for shunt speech in 49 irradiated patients.

Modality of Success Failure Success irradiation rate (%)

Full course 33 5 87 Preoperative 8 0 100 Postoperative 2 1 67

Total 43 6 88

Table 5. Swallowing function in 92 shunt speakers.

Aspiration* Original Modified Modified method method (1) method (2)

19 15 16 +-->- 9 11 1

+ 18 4 1

Total 46 30 18

* , none; +-->-, aspiration was observed at an early postoperative period but disappeared with time; +, aspiration was controlled by digital pres­sure during deglutition.

Discussion The use of a primary tracheoplastic tech­

nique has been effective from the viewpoint of the voice restoration after total laryngec­tomy. This primary technique is simple, safe, minimally complicated, and compatible with the complete removal of the lesion with a sufficient safety margin. Thus it meets the criteria for the surgical voice restoration technique.

Since the stenosis of the internal orifice of the shunt is the most frequent cause of the failure in shunt speech, fine suturing with 8 stitches is advisable. Breakdown of the shunt due to the infection of a hematoma should be prevented by the effective use of the suc­tion drains and antibiotics. Two patients whose fistulas were patent complained of a fullness in the stomach after the effort of phonation by diverting air into the shunt. These patients seemed to have failed to achieve shunt speech because of the pharyn­geal stricture. To prevent a stricture of spasm, the inferior pharyngeal constrictor muscle should be left un sutured at the orig­inal surgery of the nonirradiated patients. One patient failed in speech in spite of the

patent shunt because of a loss of motivation. Careful preoperative screening is always nec­essary.

Based on the success rate of the irradiated patients group which surpassed the rate of the nonirradiated group, we can assume that radiation therapy does not interfere with the effective shunt operation if a proper post­operative interval is observed prior to sur­gery.

As far as the prevention of aspiration is concerned, the modified method (2) appar­ently surpassed the other 2 methods. Since there is no longer any doubt that voice can be produced by connecting the trachea to the pharynx or esophagus, aspiration should be prevented at all rates. To date, the use of a valve placed in a short side-to-side TE tunnel is the only safe and reliable way for both voice restoration and tracheal reflux prevention (SINGER and BLOM, 1980). In the modified method (2) described in this paper, we attempted to combine primary cancer surgery with the creation of intelligible voice, but without aspiration. In this method using the BEMF, aspiration is effectively avoided in most cases. Both dilatation and elevation of the cervical esophagus during deglutition, together with the BEMF, seem to approxi­mate the sphincter mechanism against tra­cheal reflux. The dilated cervical esophagus pushes the shunt from behind during deglu­tition, while the BEMF surrounding the shunt restrain it from moving forward (Fig. 7). This is reinforced by elevation of the cervical esophagus with the BEMF holding the shunt from behind.

at rest during deglutition

Fig. 7. The dilated cervical esophagus pushes the shunt from behind deglutition, while the BEMF surrounding the shunt restrain it from moving.

M. AMATSU el al. S 205

References AMATSU, M.: A new one-stage surgical technique

for postlaryngectomy speech. Arch. Olorhino­[aryngo[. 220: 149-152, 1978.

AMATSU, M.: A One Stage Surgical Technique for postlaryngectomy voice rehabilitation. Laryn­goscope 90: 1378-1386, 1980.

AMATSU, M., KINISH, M., and JAMIR, J.e.: Evalu­ation of speech of laryngectomees after the Ama­tsu tracheoesophageal shunt operation. Laryngo­scope 94: 696-701, 1984a.

AMATSU, M., MATsm, T., KANAGAWA, K., KINISHI, M., and MAKINO, K.: Six years' experience with Amatsu's primary voice restoration technique after total laryngectomy. In Plastic and Recon­structive Surgery of the Head and Neck, pp. 1091-1094, Mosby Publishing Co., St. Louis, 1984b.

AMATSU, M., MATsm, T., MAKI, T., and KANAGAWA, K: Vocal rehabilitation after total laryngec­tomy: A new one-stage surgical technique. J. Ololaryngol. Jpn. 80: 779-785, 1977.

ASAI, R.: Asai's new voice production method: A substitute for human speech. Proceedings of VIII International Congress of Otorhinolaryn­gology, Tokyo, p. 730, 1965.

CALCATERRA, J.C., and JAFEK, D.W.: Tracheoeso­phageal shunt for speech rehabilitation after total laryngectomy. Arch. Ololaryngol. 94: 124-128, 1971.

CONLEY, J.J., DEAMESTI, F., and PIERCE, M.K.: A new surgical technique for the vocal rehabilitation

of the laryngectomized patient. Ann. 0101. Rhinol. Laryngol. 67: 655-664, 1958.

KINISHI, M., and AMATSU, M.: Aerodynamic study of laryngectomees after the Amatsu tracheoeso­phageal shunt operation. Ann. 0101. Rhinol. Laryngol. 95(1): 1986 (in press).

KOMORN, R.M.: Vocal rehabilitation in the laryn­gectomized patient with a tracheoesophageal shunt. Ann. 0101. Rhinol. Laryngol. 83: 445-451, 1974.

PANJE, W.R., VANDEMARK, D., and MCCABE, B.F.: Voice button prosthesis rehabilitation of the laryngectomy. Ann. 0101. Rhinol. Laryngol. 90: 503-505, 1981.

ROBBINS, J., FISHER, H.B., BLOM, E.D., and SINGER, M.I.: A comparative acoustic study of normal esophageal, and tracheoesophageal, speech pro­duction. J. Speech Hear. Disord. 49: 202-210, 1984.

ROBBINS, J., FISHER, H.B., and LOOEMANN, J.A.: Acoustic characteristics of voice production after Staffieri's surgical reconstructive procedure. J. Speech Hear. Disord. 47: 77-84, 1982.

SINGER, M.I., and BLOM, E.D.: An endoscopic tech­nique for restoration of voice after laryngectomy. Ann. 0101. Rhinol. Laryngol. 89: 529-533, 1980.

STAFFIERI, M., and SERAFINI, Z.: La riabilitazione chirurgica deIIa voce er della respiratione dopo laryngectomia totale. 29th National Congress of the Associazione Otologi Ospedalieri Italiana, pp. 57-111, 1976.