- Email: [email protected]
Extended cervical esophagomyotomy for cricopharyngeal dysfunction
J THORAC
CARDIOVASC SURG
80:669-678, 1980
Extended cervical esophagomyotomy for cricopharyngeal dysfunction Forty adult patients have undergone a 7 to 10 em cervical esophagomyotomy (from the superior cornu of the thyroid cartilage to behind the cluvicleyfor cricopharyngeal dysfunction. A Zenker's diverticulum was present in 12 patients (30%) and in five was recurrent. Preoperative symptoms included cervical dysphagia (85%), expectoration of saliva (40%), and intermittent hoarseness (30%). Four patients were being fed through tubes because of total inability to swallow. "Heartburn" was experienced by one halt' of the patients, but only 12 had acid or food regurgitation. The duration of symptoms ranged from I month to /I years (average 3.9 years). Weight loss had occurred in 15 patients (38%) and ranged from 5.5 to 40.9 kg (average 16 kg). Barium swallows showed no abnormalities in 10 patients. Abnormal findings included a Zenker's diverticulum (12), prominent cricopharyngeal sphincter (II), nasopharyngeal reflux or incoordinated initiation of deglutition, or both (seven), a sliding hiatal hernia (II), and abnormal esophageal motility (seven). Esophageal manometry revealed abnormalities of upper esophageal sphincter (UES).fimction in only 16 patients. Of 36 patients undergoing standard acid reflux testing, one third had moderate-to-severe gastroesophageal reflux. Seven patients underwent staple resection of a Zenker's diverticulum at the time (~t' cervical esophagomyotomy, Postoperative complications included transient vocal cord paresis (four), vocal cord paralysis (one), and salivary fistula (one). There were no postoperative deaths. After 2 to 48 months (average 16 months) offollow-up, 34 patients (85%) have had a good to excellent result, and six (15%) have not been benefited by operation.
Mark B. Orringer, M.D., Ann Arbor, Mich.
Numerous terms have been used to define the symptom complex which results from abnormal function of the upper esophageal, or cricopharyngeal, sphincter. Such terms as cricopharyngeal chalasia, achalasia, or spasm have little more validity in these patients than the archaic diagnosis of globus hystericus, for in the vast majority, standard esophageal manometric techniques have failed to demonstrate true hypotonicity or hypertonicity of the upper esophageal sphincter (UES) or failure to relax with swallowing (achalasia). This failure may be a function of the limitations of existing instrumentaton in recording the rapid sequence of events that occurs with normal deglutition in a unique asymmetric sphincter that changes position with laryngeal excursions during swallowing. The term
From The Department of Surgery, Section of Thoracic Surgery, University of Michigan Medical Center, Ann Arbor, Mich. Read at the Sixtieth Annual Meeting of The American Association for Thoracic Surgery, San Francisco, Calif'., April 28 to 30, 1980. Address for reprints: Mark B. Orringer, M.D., Box 32, C7079, University Hospital, Ann Arbor, Mich. 48109.
"cricopharyngeal dysfunction" has been chosen to describe a spectrum of disorders of swallowing which involve the upper esophageal sphincter and occur in association with a variety of abnormalities of the central and peripheral nervous system, metabolic and inflammatory myopathy, and gastroesophageal reflux, as well as currently unidentified factors. This report describes experience at the University of Michigan Medical Center with the diagnosis and surgical treatment of cricopharyngeal dysfunction.
Patients Forty adult patients, including 18 men and 22 women ranging in age from 23 to 86 years (average 56.7 years), have undergone an extended cervical esophagomyotomy for symptoms of cricopharyngeal dysfunction. A Zenker's diverticulum was present in 12 patients (30%), and in five it was recurrent after previous resections. In two patients, cricopharyngeal dysfunction had followed midbrain cerebrovascular accidents. Additional associated conditions included symptomatic thoracic esophageal spasm (four), polymyositis (one), dermatomyositis (one), myotonic dystrophy (one), multiple
0022-5223/80/110669+10$01.00/0 © 1980 The C. V. Mosby Co.
669
The Journal of
670
Orringer
Thoracic and Cardiovascular Surgery
A
B
c
Fig. 1. Extended cervical esophagomyotomy. A, A 5 ern oblique incision anterior to the sternocleidomastoid muscle and centered over the cricoid cartilage. B, Esophagomyotomy is placed on left posterolateral aspect of esophagus to avoid injury to the recurrent laryngeal nerve, seen in the tracheoesophageal groove. C, Completed esophagomyotomy extends from the level of the superior cornu of the thyroid cartilage inferiorly to I to 2 em behind the clavicle.
sclerosis (one), multiple myeloma (one), and severe chronic cervical compression fracture (one). Four patients had undergone hiatal hernia operations 3 to 18 months earl ier, and although their reflux symptoms were controlled, cervical dysphagia had persisted; five patients had undergone prior partial or subtotal thyroidectomy for symptoms of cervical dysphagia, and three had persistent symptoms after incomplete cervical esophagomyotomies elsewhere. Four patients were being fed through feeding tubes because of total inability to swallow, and one had undergone a tracheostomy because of recurrent severe aspiration. Cervical dysphagia was the predominant symptom in 38 of these patients, being present in all but two who complained of a "pressure sensation" in the throat without associated difficulty swallowing. Among the 28 patients with intermittent symptoms, dysphagia was for solid foods in 27 and for both solids and liquids in eight. The 10 patients with constant dysphagia all had difficulty swallowing both solids and liquids. The duration of dysphagia ranged from I month to II years (average 3.9 years). Weight loss had occurred in 15 patients (37.5%) and ranged from 5.5 to 40.9 kg (average 16 kg). Nasopharyngeal regurgitation during swallowing was experienced by four patients. Respiratory symptoms were present in 33 patients (83%). Twenty-two (55%) described choking on saliva and 16 (40%) complained of "too much phlegm," with expectoration of excessive saliva. Eighteen (45%) had choked on food. A history of intermittent hoarseness
was present in 12 (30%). Three patients experienced marked shortness of breath as food slowly passed through the cervical esophagus, and in one, episodes of wheezing were precipitated by bouts of dysphagia. Although one half of these patients complained of intermittent "heartburn," in only 12 (30%) was there a history compatible with gastroesophageal reflux, that is, acid or food regurgitation, or both, aggravated by the recumbent position. Barium swallow examinations were performed in all patients and showed no abnormalities in 10. Among the 12 patients with a Zenker's diverticulum, the cricopharyngeal sphincter was reported to be prominent in II. Nasopharyngeal reflux or incoordinated initiation of deglutition, or both, was noted in seven, a sliding hiatal hernia in II, and abnormal esophageal motility in seven. Standard esophageal function tests;' performed with a combined, triple-lumen manometric-pH recording catheter, perfused with saline," were conducted in 37 patients. Twenty-four had normal thoracic esophageal peristalsis and 13 had abnormal motility patterns. Three had distal esophageal spasm, nine had frequent simultaneous contractions after swallowing, and one had changes of scleroderma. Relaxation of the distal esophageal sphincter was normal in all but the patient with scleroderma. Manometric evaluation of the UES was completely unremarkable in 21 patients. In our laboratory, peak UES pressures in asymptomatic "control" subjects
Volume 80 Number 5 November, 1980
range from 15 to 85 mm Hg and length from 2 to 5 cm (average 3.4 ern). The following manometric abnormalities of DES function were detected in 16 of the patients with cricopharyngeal dysfunction: marked delay in ability to initiate a swallow on command (five); weak pharyngeal contraction (four); incomplete relaxation of the DES to cervical esophageal baseline pressure (three); total lack of DES relaxation (achalasia) (two); hypotensive DES with maximum pressures of 5 to 7 mm Hg (two); hypertensive, spastic DES with resting pressure greater than 90 mm Hg , length between 6 and 7 em, and sustained (12 to 15 seconds) contractions after swallowing (two); and incoordination of the UES, with pharyngeal contraction occurring simultaneously with DES closure rather than relaxation (two). One third of the patients demonstrated moderate-tosevere gastroesophageal reflux with the standard acid reflux test. The acid clearance test was normal in only five of the patients, and the acid perfusion test was abnormal in 10. At the time of their cervical esophagomyotomy, seven patients underwent staple resection of a Zenker's diverticulum, one a Collis-Nissen hiatal hernia repair, and one a long thoracic esophagomyotomy for associated diffuse esophageal spasm. Technique
After induction of general anesthesia and endotracheal intubation, rigid esophagoscopy is usually performed to exclude neoplasm or inflammatory esophageal disease and to evacuate the contents of a Zenker's diverticulum. It is unwise to persist with attempts to advance the esophagoscope in a patient with an extremely tight cricopharyngeal sphincter or a large pharyngoesophageal diverticulum that prevents free passage of the instrument. The patient's neck is extended by placing a small rolled sheet beneath the scapulae. The head is turned toward the right side and the occiput supported on a head ring. The field extends from the mandible to the nipple line, and drapes are sutured to the skin to provide access to the entire left side of the neck. The cricoid cartilage is identified by palpation, and a 5 em oblique left cervical incision centered at the level of this cartilage is made parallel and anterior to the sternocleidomastoid muscle (Fig. 1, A). In young women, a transverse left cervical incision in a skin crease directly over the cricoid cartilage gives a slightly better cosmetic result. The platysma and omohyoid fascial layers are incised, and the sternocleidomastoid muscle and carotid sheath and its contents are retracted laterally and the trachea medially, The recurrent
Esophagomyotomy for cricopharyngeal dysfunction
67 I
laryngeal nerve coursing in the tracheoesophageal groove is identified and protected. The middle thyroid vein or inferior thyroid artery is ligated and divided as required. The dissection proceeds directly posterior through the cervical fascial layers to the prevertebral fascia. A No. 40 Fr. Hurst-Maloney bougie inserted into the oropharynx by the anesthetist is guided into the esophagus by the surgeon and aids in defining the posterolateral esophageal wall, which subsequently will be incised, as well as the diverticulum if it is present. When there is no diverticulum, the cervical esophagomyotomy is next performed. The incision extends from the level of the tip of the superior cornu of the thyroid cartilage inferiorly to I to 2 em below the clavicle and is 7 to 10 em long (Fig. I, B). The myotomy is deliberately placed on the left posterolateral aspect of the esophagus to avoid injury to the recurrent laryngeal nerve. Although the esophagomyotomy is longer than the skin incision, adequate exposure of the muscle being incised is achieved by exerting upward traction on the skin retractors when working superiorly and downward traction when approaching the clavicle. In addition to this shifting of skin retractors, if the first assistant applies direct upward or downward retraction on the esophagus with a small gauze square as the esophagomyotomy is performed, exposure onto the pharynx superiorly and behind the clavicle inferiorly is readily achieved (Fig. 1, C). In the presence of a Zenker's diverticulum, after the intraesophageal dilator is in position, the pouch is located and mobilized away from adjacent fibroareolar tissue. The inferior thyroid artery is an excellent landmark, as the diverticulum is inevitably found behind it when the vessel is divided. A Babcock clamp applied to the fundus of the diverticulum is used for traction that facilitates dissection of the pouch. The base of the mucosal pouch at its emergence through the esophageal muscle is identified, and the adjacent muscle is gently dissected away with a fine, right-angle clamp and divided with electrocautery and a needle-tip point (Fig. 2, A). As described earlier, the incision is extended inferiorly behind the clavicle and superiorly to the level of the superior cornu of the thyroid cartilage. When the diverticulum is 1.5 to 2 em or smaller, resection is not necessary. When larger, the neck of the pouch is crossed with a TA-30 surgical stapler using 4.8 mm staples, and the diverticulum is resected (Fig. 2, B). The staple suture line is not reinforced. A dilator present within the esophagus and avoidance of undue traction on the diverticulum prevent removal of excessive esophageal mucosa. After the neck wound has been irrigated and hemostasis along the divided muscle edge
The Journal of
672
Thoracic and Cardiovascular
Orringer
Surgery
B
A
Fig. 2. Cervical esophagomyotomy and concomitant resection of Zenker's diverticulum. A. After mobilization of the diverticulum, the esophagomyotomy is performed in either direction from the base of the pouch for the same distance as described in Fig. 1. B. After the esophagomyotomy is completed, the base of the diverticulum is crossed with a TA-30 stapler and amputated.
has been obtained, a nasogastric tube is inserted by the anesthetist and gently guided into the upper esophagus. Air is insufflated and the esophagomyotomy is tested under saline to ensure that no tear has occurred. If an air leak is identified, the tear is closed with interrupted 5-0 Prolene mucosal stitches. Silver clip markers for future roentgenographic evaluation are placed at either end of the esophagomyotomy so as to avoid contact with the mucosa that might cause erosion and perforation. The nasogastric tube is advanced into the stomach and utilized for gastric decompression until any postoperative nausea has subsided. The tube is usually removed the evening of operation, and clear liquids by mouth are begun the next day. A small Penrose drain is used to drain the wound for 24 hours. The patient is generally discharged on a soft diet by the second or third postoperative day.
Results There were no postoperati ve deaths. Transient left vocal cord paresis that resolved spontaneously in 3 to 6 weeks occurred in four patients. One patient, who had had a previous resection of a Zenker's diverticulum complicated by a salivary fistula, had a left vocal cord paralysis that was managed with a Teflon injection; this patient regained a near-normal voice. There was one salivary fistula in this series, and this occurred in a patient who had undergone three previous resections of a Zenker's diverticulum; the fistula closed after 5 days.
After 2 to 48 months (average 16 months) of follow-up, 22 patients have had an excellent result and have no difficulty swallowing a regular diet; 12 are markedly improved and experience only occasional mild dysphagia; and six (15%) have had a poor result, with no noticeable benefit from operation. Among these latter six patients are two elderly men, aged 70 and 86 years, with associated severe cerebrovascular disease and senility, and a 38-year-old woman whose chief complaint was hoarseness and throat pain, rather than dysphagia.
Discussion
In evaluating the patient with cervical dysphagia, one must exclude other local causes for this complaint: extrinsic compression of the esophagus by the thyroid gland, cervical lymphadenopathy, or hyperostosis of the cervical spine (bone spur) and intrinsic compromise of the cervical esophageal lumen by neoplasm, web, stricture, or an inflammatory process (abscess). A history of prior head and neck operations may be extremely important, as recurrent laryngeal nerve injury may result in cricopharyngeal dysfunction. This may occur not only after thyroid or parathyroid operations, but after pulmonary resections as well, since innervation of the cricopharyngeus is derived in part from the recurrent laryngeal nerve." Similarly, dysphagia is not infrequent after laryngectomy. 4 Finally, following tracheostomy, normal elevation and anterior rotation of the larynx that occurs with swallowing may be pre-
Volume 80 Number 5 November , 1980
Esophagomyotomy for cricopharyngeal dysfunction
67 3
Fig. 3. A, Prominence of the cricopharyngeal sphincter (a rrows) in a patient with cervical dysphagia and symptomatic gastroesophageal reflux. B, Detail of cervical esophagus. vented by local scar formation, resulting in cervical dysphagia." Both a barium swallow and esophagoscopy remain the most vital studies in patients with dysphagia at any level . The patient with cricopharyngeal dysfunction localizes his cervical dysphagia between the thyroid cartilage and the suprasternal notch . Expectoration of excessive saliva, a complaint of 40% of our patients, is common in those who are unable to swallow normally the I to 1.5 L of saliva produced in the mouth each day . Although the physician who elicits a history of cervical dysphagia and constant expectoration of saliva may be somewhat suspicious of the clinical significance of such complaints, the additional occurrence of hoarseness should suggest the probability of cricopharyngeal dysfunction. Hoarseness is not commonly associated with swallowing disorders, and the patient who describes both dysphagia and difficulty with his voice is seldom malingering . The inferior pharyngeal constrictor muscle, of which the cricopharyngeal sphincter is a part, may affect the vocal cords in two ways. First, contraction of the inferior pharyngeal constrictor adducts or approximates the alae of the thyroid cartilage, lengthening and tensing the vocal cords . Second, by pulling the cricoid cartilage posteriorly, the cricopharyngeal muscle acts with the lateral cricothyroid to stretch the vocal cords . Thus there is adequate physiological explanation
for the association between symptoms of cervical dysphagia and hoarseness in the patient with cricopharyngeal dysfunction . Concomitant weight loss completes the diagnostic symptom complex that in our experience justifies cervical esophagomyotomy, even in the absence of roentgenographic, manometric , or endoscopic abnormalities. Barium swallow findings of a prominent cricopharyngeal sphincter (Fig . 3), a prominent posterior cricopharyngeal bar on lateral cervical views (Fig. 4), or a Zenker diverticulum (Fig. 5) are evidence of cricopharyngeal dysfunction in the symptomatic patient; each of these findings , howe ver, has been reported in asymptomatic individuals. Despite our inability to demonstrate manometric abnormalities of the upper esophagus in the majority of patients with a Zenker diverticulum, it is our conviction that the underlying problem is neuromotor dysfunction of the sphincter. The basis for this belief is the inevitable occurrence of the mucosal pouch proximal to fibers of the cricopharyngeus muscle and the complete disappearance of 1.5 to 2 em diverticula after esophagomyotomy alone, with no resection of the diverticulum. It is seldom the size of the pouch which determines the degree of associated dysphagia; few are much larger than 3 to 4 em . The patient with a 2 mm Zenker diverticulum, however, may have as much or more cervical dyspha-
The Journal of
674
Orringer
Thoracic and Cardiovascu lar Surgery
Fig. 4. A. Typical appearance of hypertrophic cricopharyngeal sphincter (anteroposterior view). B. Lateral cervical view showing posterior cricopharyngeal bar.
Fig,S. A. Zenker 's diverticulum in a patient with an associated " patulous cardia" (arrow) and asymptomatic gastroesophageal reflux without esophagitis. B. Residual barium in 2.5 ern pouch (anteroposterior view). gia and aspiration as one with a much larger pouch . In treating the patient with a Zenker's diverticulum, therefore, the least important part of the operation is resecting the pouch; dysphagia is not the result of the pouch, but rather of the cricopharyngeal dysfunction that has produced the diverticulum , and unle ss the abnormal muscle is divided, the functional obstruction is not re-
lieved and symptoms may persist. We can only attri bute the excellent results report in patients undergoing resection of a Zenker diverticulum without concomitant esophagornyotomy" to the fact that, when a meticulous dis section of the neck of the pouch is performed, inadvertent division of the incoordinated UES muscle generally occurs . In five of our 12 patients with Zenker
Volume 80 Number 5 November, 1980
diverticul a, previous resections of the pouch had been performed elsewhere; one of these patients had undergone three prior resections . In each case , recurrence of the diverticulum had been evident within several months of the di verticulectomy . Cervical esophagomyotom y performed alone or concomitantly with diverticulecto my provided lasting control of both dysphagia and the pouch in all of the se patients. The admonition to resect the diverticulum completely in order to avoid recurrence of the pouch is nonsense . With a complete cervical esophagomyotomy and a 1.5 to 2 em residual pouch, there will be no demonstrable roentgenographic abnormality within 2 weeks of operation (Fig. 6). Surgical stapling devices greatly facilitate resection of large pouches with minimum tissue trauma , without the need to open the esophagus, and in a manner that is more efficient and superior to diverticuloplexy as described by Belsey . i Becau se the length of the upper esophageal sphincter may be quite variable on manometric evaluation, being 6 to 7 em in two of our patients , an extended 7 to 10 ern cervical esophagomyotomy is advocated as a means of ensuring complete relief of the functional obstruction resulting from the incoordinated sphincter muscle . Three patients in this series had undergone limited cervical esophagomyotomies elsewhere , but their symptoms had persisted . We stres s, therefore , (I) meticulous division of all cervical esophageal muscle fibers from the inferior pharynx to the cla vicle, (2) distending the muco sa with air insufflated through an intraesophageal tube to identify any remaining undivided muscle strands, and (3) marking either end of the myotomy with silver clips to permit accurate radiographic determination of the limits of the operation . Little damage can occur from an excessively long myotomy; the concern should be that dysphagia will persist because of too limited an esophageal incision . Standard esophageal manometric evaluation of the cricopharyngeus is made difficult by the unique characteristics of the UES. H. 9 Availability of intraesophageal microtransducer recording catheters.!" as well as the more recently described glycerin-filled spatially oriented manometry system , II may provide more precise documentation of the physiology of the cricopharyngeus than has been possible in the past. Thus, despite the well-described manometric patterns of UES dysfunction; " we have demonstrated objective motility abnormalities in only 16 of our patients . Recognizing the limitations of our current recording systems in evaluating UES function , we have nonetheless routinely evaluated our patients with standard esophageal function tests, not only to document demonstrable
Esophagomyotomy for cricopha ryngeal dysfun ction
675
Fig. 6. Anteroposterior ( A) and lateral ( B) cer vical views of esoph agus of patient shown in Fig. 5 two week s after extended cervical esophagomyotom y without resection of the diverticulum. Dysphagia was relieved and the pouch is no longer evident.
manometric abnormalities, but also to define the degree of any gastroesophageal reflux present. Cricopharyngeal dysfunction is common in patients with gastroesophageal reflux ," presumably the result of reflex muscular incoordination induced in the cervical esophagus by irritation of the lower esophagus by gastric acid . Henderson!" has reported that 50% of 200 consecutive patients with symptomatic gastroesophageal reflux had pharyngoesophageal dysphagia and that this symptom was relieved by correcting the reflux; only eight of 490 patients undergoing a hiatal hernia operation have required subsequent cricopharyngeal myotomy for persistent cervical dysphagia. Betsey? has admonished against performing a cervical esophagomyotomy in the presence of an incompetent lower esophageal sphincter, lest the patient be subject to the complications of aspiration from unchecked gastroesophageal reflux . Henderson and Marryatt;':' however , achieved successful relief of cervical dy sphagia in five patients with gastroesophageal reflu x who were either too ill for surgical correction of reflux or had dominant cricopharyngeal symptoms with minimal reflux symptoms. In our experience , the minority of patients with gastroesophageal reflux actually regurgitate acid to the level of the pharynx. Thus, when there is cricopharyngeal dysfunction in the elderly or debilitated patient or the patient with objectively demonstra-
The Journal of
676
Orringer
ble but asymptomatic reflux without esophagitis, it seems perfectly reasonable to perform a cervical esophagomyotomy , place the patient on a strict antireflux regimen, and not perform an antireflux procedure unless pulmonary symptoms from aspiration become evident. Among the 12 patients with abnormal gastroesophageal reflux in this series, four had persistent cervical dysphagia despite successful antireflux operations; only two of these had relief of dysphagia after cervical esophagomyotomy. One patient with scleroderma reflux esophagitis and marked cricopharyngeal dysfunction had an excellent response to concomitant Collis-Nissen hiatal hernia repair and cervical esophagomyotomy. The remaining seven patients were either too debilitated to tolerate an antireflux operation or had few if any typical reflux symptoms. All but one of these had a good or excellent response to cervical esophagomyotomy, and none has had difficulty with chronic aspiration. It is apparent that the diagnosis of cricopharyngeal dysfunction must frequently be made on clinical grounds in a patient complaining of cervical dysphagia, expectoration of saliva, and intermittent hoarseness, with or without weight loss. There may be no objective radiographic, manometric, or endoscopic abnormalities. A carefully performed, extended cervical esophagomyotomy relieves symptoms in a substantial number of these individuals (85%) by dividing all incoordinated pharyngoesophageal muscle, and it is a relatively low-risk procedure for which the gains may be great in appropriately selected patients. A general policy of operating upon every patient who complains of a "lump in the throat" is not being advocated. A careful history, corroborated at times by barium swallow and/or manometric findings, is essential to avoid unnecessary operations in patients in whom cervical symptoms in fact have a psychological basis.
2
3
4
5
REFERENCES Skinner DB, Hendrix TR: Diagnostic tests, Gastroesophageal Reflux and Hiatal Hernia, chap 7, DB Skinner, RH Belsey, TR Hendrix, GD Zuidema, eds., Boston, 1972, Little, Brown & Company, pp 87-106 Orringer MB, Lee R, Sloan H: Combined manometric-pH recording catheter for esophageal function tests. Ann Thorac Surg 26:581-585, 1978 Henderson RD, Boszko A, van Nostrand SWP: Pharyngoesophageal dysphagia and recurrent laryngeal nerve palsy. J THORAC CARDIOVASC SURG 68:507-512, 1974 Duranceau A, Jamieson G, Hurwitz AL, Jones RS, Postlethwait RW: Alteration in esophageal motility after laryngectomy. Am J Surg 131:30-35, 1976 Bonanno PC: Swallowing dysfunction after tracheostomy. Ann Surg 174:29-33, 1971
Thoracic and Cardiovascular Surgery
6 Welsh GF, Payne WS: The present status of the one-stage paryngoesophageal diverticulum. Rev Surg 30:370-371, 1973 7 Belsey R: Functional disease of the esophagus. J THORAC CARDIOVASC SURG 52:164-188, 1966 8 Dodds WJ, Steward ET, Hogan WJ, Stef 11, Arndorder RC: Effect of esophageal movement on intraluminal pressure recording. Gastroenterology 67:592-600, 1974 9 Dodds WJ, Stef 11, Hogan WJ: Factors determining pressure measurement accuracy by intraluminal esophageal manometry. Gastroenterology 70: 117-128, 1976 10 Pope CE II, Christensen J, Harris LD, Nelson TS, Motlet NK, Templeton F: Diseases of the esophagus (work group I). A survey of opportunities and needs in research on digestive diseases. Gastroentero logy 69: 1058-1070, 1975. II Welch RW, Luckmann J: The upper esophageal sphincter (UES) in man. Significance of radial asymmetry and precise measurement of closure strength. Gastroenterology 72:1168, 1977 12 Hurwitz AL, Duranceau A, Haddad JK: Oropharyngeal dysphagia, Disorders of Esophageal Motility, chap 7, Major Problems in Internal Medicine, vol 16, Philadelphia, 1979, W. B. Saunders Company, pp 67-84 13 Henderson RD: Disorders of the pharyngoesophageal junction, Motor Disorders of the Esophagus, chap 19, Baltimore, 1976, The Williams & Wilkins Company, pp 184-204 14 Henderson RD, Marryatt G: Cricopharyngeal myotomy as a method of treating cricopharyngeal dysphagia secondary to gastroesophageal reflux. J THORAC CARDIOVASC SURG 74:721-725, 1977
Discussion DR. DAVID B. SKINNER Chicago. l/l.
I would like to raise two questions, one of which is related to the indications for this operation. It appeared to me that Dr. Orringers series included 12 patients with a Zenker's diverticulum. Most of us would agree that myotomy is part of the treatment of that condition. There were six other patients with a neural or myopathic disorder which might indeed be associated with spasm of the sphincter. Dr. Orringer also listed six other reasons for operating, but he did not report the precise indications for operation in 16 patients. There were 21 patients in this group with normal esophageal function studies on manometry, 10 patients with a normal barium swallow, and II patients with a so-called prominent cricopharyngeal sphincter. Our radiologist tells us that this may be a normal finding. Thus I am left unconvinced about the indications for operation in these patients. I would agree that this operation can be highly valuable in patients with well-documented manometric spasm or obstruction related to neuromuscular disorders and Zenker's diverticulum. Second, I would like to question the use of the term "extended myotomy. " In the abstract Dr. Orringer says that he has used a 7 to 10 cm esophagomyotomy. The length of the esophagus from the cricopharyngeal sphincter to the aortic
Volume 80 Number 5 November, 1980
arch is 7 ern, Therefore, to do a myotomy or to advocate one of 7 to 10 em would extend the dissection below the aortic arch, which I doubt that he has done through a cervical incision. The only other way to extend a myotomy is upward, and I know of no reason for dividing the pharyngeal muscles in these conditions. It is often as much a lack of coordination and activity in the pharynx that causes the difficulty as it is the tightness of the cricopharyngeal sphincter. I would ask Dr. Orringer to clarify these two points related to indications and the meaning of the term "extended esophagomyotomy. " DR. CLEMENT HIEBERT Portland. Maine
Dr. Orringer 's paper illustrates the two principal problems-finding the patients and finding the cricopharyngeus muscle. With respect to the first problem, I am concerned over the array of esophageal conditions for which the operation is recommended. A more precise criterion would seem to be the radiologic finding of a posterior obstruction at the point where the pharynx joins the esophagus. The very thought of operating for a lump in the throat without confirmatory evidence gives me a lump in my throat. As for the second problem of identifying the cricopharyngeal muscle and sectioning it precisely, I respectfully disagree with both the concept of and the need for a long myotomy. To set about deliberately to weaken the pharynx even slightly runs counter to the whole purpose of the operation. I also disagree with the idea of removing a diverticulum. The diverticulum is but a mucosal balloon and needs only to be upended and plicated in combination with sectioning of the cricopharyngeal muscle. When the diverticulum is small it may be ignored. We have utilized these concepts in operating on 37 patients over the last 20 years and find the use of local anesthesia advantageous. These elderly patients remain alert and can swallow at once. Indeed, they can swallow (gelatin dessert) during the operation and thereby afford the surgeon a chance to view the disordered physiology. With each pharyngeal contraction the ordinarily ephemeral cricopharyngeal muscle declares itself as a posterior constricting band. Finally, I agree with Dr. Orringer that this upper sphincter of the esophagus (UES) is a legitimate concern of the thoracic surgeon; perhaps nowhere else in the entire gastrointestinal tract can so much be achieved with so short a cut. DR. VICTOR H. KAUNITZ Teton Village. Wyo.
I would like to address my remarks to the possible cause of this condition. Dr. Orringer's paper deals with swallowing difficulties referred to the neck area-95% of his patients had cervical dysphagia as their chief complaint. This is a relatively common symptom. I have had experience with some 50 patients with the same symptoms, referred often by ear, nose, and throat colleagues who could find no cervical disease. By taking a careful history, I elicited one or more symptoms typical of reflux, esophagitis in addition to the cervical dys-
Esophagomyotomy for cricopharyngeal dysfunction
67 7
phagia; namely, heartburn, lower substernal dysphagia, regurgitation of acid and/or food, and substernal pain. In Dr. Orringers series 50% had heartburn and 30% had acid or food regurgitation. I believe we are dealing with the same pathophysiological entity. What is its cause? In 75% of my group of patients, a sliding hiatal hernia was demonstrated by roentgenography. In the remainder, the presence of the hernia was proved by esophagoscopic visualization of the esophagogastric mucosal junction above the hiatal level. A sliding hiatal hernia was present at operation in all these cases, and it was corrected by means of the technique I described before this Association in 1974. All patients had complete relief of their symptoms of cervical dysphagia as well as those of reflux esophagitis. There have been no recurrence to date. These results would indicate that the cause of cricopharyngeal dysfunction, producing the symptoms of cervical dysphagia, is almost always a reflex spasm of the UES-the cricopharyngeal muscle. This spasm would seem to be induced by reflux esophagitis in the lower esophagus, caused by an incompetent lower esophageal sphincter resulting from a sliding hiatal hernia. DR. F. HENRY ELLIS, JR. Boston. Mass.
Our knowledge of esophageal physiology has increased greatly in the past 15 to 20 years but chiefly as it concerns the body of the esophagus and its lower sphincter. There is still confusion regarding normal and abnormal function of the upper esophagus, particularly the UES. Dr. Orringer has tried to bring some order to this chaos and for this we are indebted to him. I have been interested, as have others, in classifying the causes of upper esophageal dysphagia, and the slide illustrates some similarities in our respective experiences. [Slide] I have treated far fewer patients. This is because operations were restricted almost entirely to patients in whom abnormalities of esophageal function could be identified. Pharyngoesophageal diverticula, many of them with a hiatus hernia, predominate, and we are indebted to our distinguished guest, Mr. D'Arcy Sutherland, as well as to Mr. Belsey for having pointed out many years ago the importance of accompanying diverticulectomy with an esophagomyotomy. Nonetheless, we still see patients with postoperative recurrences due to failure to perform a myotomy at the time of diverticulectomy. Increasing numbers of patients with hypertension of the UES are being seen, some of them with incoordination of cricopharyngeal function. Some cases may represent early stages of a pharyngoesophageal diverticulum and they are relieved by myotomy. The poor results in my experience have occurred in patients with neuromuscular disease-three patients with amyotrophic lateral sclerosis. In these patients there was no demonstrable abnormality of motility, and cricopharyngeal myotomy failed to relieve their dysphagia. I would like to ask Dr. Orringer to elaborate on the patients with neuromuscular disease whom he described. Were his poor results in that group? I have always questioned the existence of cricopharyngeal
The Journal of
67 8
Orringer
Thoracic and Cardiovascular Surgery
"achalasia. " I note Dr. Orringer had two such patients in his series. Did their motility records really show lack of relaxation? I have never seen such a record. I have two other questions to ask Dr. Orringer. When he has been so insistent in other areas of esophageal surgery on the presentation of objective data, why has he in this instance departed from that policy? In only a minority of the patients that he operated upon could evidence of abnormality in function be presented. Second, what is the need for extending the myotomy to the extent that he has illustrated? In our patients, a short myotomy encompassing the sphincter and some of the fibers of the upper esophagus has given equally good results. DR. JOHN KENNEDY London, England
I would like to make the observation that one eighth of Dr. Orringer's patients were apparently hoarse postoperatively. This can be an annoying problem for both surgeon and patient. I wonder if he could be more specific about the patients in whom voice apparently returned. We all recognize that return of voice does not necessarily mean vocal cord function is normal again because of the remarkable ability of patients to compensate. DR. 0 R R IN G E R (Closing) I appreciate the discussers' comments, which I think help emphasize a number of key points about this topic of cricopharyngeal dysfunction. Dr. Skinner's concern about the variety of associated disorders in our patients with cricopharyngeal dysfunction allows me to emphasize again my belief that this is frequently a clinical diagnosis which may be made in a variety of illnesses. The spectrum of associated conditions in which cricopharyngeal dysfunction occurred in our patients was broad, and yet the clinical symptom complex of cervical dysphagia, respiratory symptoms, hoarseness, and weight loss was so common that it allowed us to approach all of these patients operatively in a uniform fashion. The need for an "extended" cervical esophagomyotomy , 7 to 10 cm in length, has been challenged, In our laboratory, the normal UES length in asymptomatic control patients is between 2 and 5 ern, Two of our patients with recurrent Zenker's diverticula, both of whom had had multiple operations, had a UES which measured 7 cm. I think the only way we can ensure that all abnormal cricopharyngeal muscle is divided is to be sure that a complete "extended" myotomy is performed. A Zenker's diverticulum herniates through fibers of cricopharyngeal sphincter, and if just the muscle inferior to the pouch is divided, there still may be undivided, incoordinated cricopharyngeal fibers superior to the origin or neck of the pouch. This area can then become a new site for the subsequent development of another pouch. Therefore, we carry the myotomy upward onto the pharynx to the level of the superior cornu of the thyroid cartilage and inferiorly be-
hind the clavicle. We have found no adverse effect from performing too long a cervical esophagomyotomy; in fact, four patients in our series represented failures of too conservative a cervical esophagomyotomy performed elsewhere, The average length between the silver clip markers that we place at either end of our cervical myotomy is 7 to 10 cm. As documented with postoperative radiographic studies, the inferior extent of the incision is not the aortic arch but rather just behind the clavicle. In identifying this sphincter which is supposed to be so elusive, I would like to emphasize that the incision should center on the cricoid cartilage. The cricopharyngeal sphincter arises from the cricoid cartilage, which is therefore the anatomic landmark for this operation. If the incision is centered over the cricoid cartilage and an adequate incision is made well above and well below this level, one will not miss the sphincter. Now, of course, comes the key question: Why do we advocate an operation without objective data when all of our other work has emphasized the need for objectivity in esophageal surgery? The problem is that the cricopharyngeus sphincter is unique in the gastrointestinal tract. It is not circumferential and symmetric like the pylorus or the anus. Rather, it is a muscular sling that arises from cartilage. In addition, with swallowing, this sphincter ascends and descends with laryngeal excursions. Thus, with our current techniques of manometric evaluation, it is extraordinarily difficult to obtain consistent, meaningful measurements of UES function, This is responsible for our inability to show objectively definable abnormalities in all of these patients, With some of the newer microtransducer techniques that are becoming available and the glycerin-filled, spatially oriented catheters which have recently been described, hopefully our ability to objectively document abnormalities of the UES will be improved. Finally, Dr. Kennedy has expressed concern about the reported incidence of postoperative hoarseness in our patients. We limit the skin incision to 5 cm in length and retract the larynx medially, I am certain that the four patients who had vocal cord paresis in our group had traction on the recurrent laryngeal nerve. In all cases the hoarseness resolved within 4 to 6 weeks after operation, and in each case indirect laryngoscopy confirmed that the initially paretic vocal cord had resumed its normal movement. We are not advocating an operation in every patient who complains of a lump in the throat. Both the barium swallow and esophagoscopy remain the most vital studies in patients with dysphagia at any level. A careful history, esophageal function tests, a barium swallow, and esophagoscopy are essential to exclude the various local causes for cervical dysphagia-aberrant thyroid tissue, lymphadenopathy, hyperostosis of the cervical spine, neoplasms, cervical weband to avoid unnecessary operations in patients who have a true psychological basis for their cervical complaints or a true globus hystericus.
CARDIOVASC SURG
80:669-678, 1980
Extended cervical esophagomyotomy for cricopharyngeal dysfunction Forty adult patients have undergone a 7 to 10 em cervical esophagomyotomy (from the superior cornu of the thyroid cartilage to behind the cluvicleyfor cricopharyngeal dysfunction. A Zenker's diverticulum was present in 12 patients (30%) and in five was recurrent. Preoperative symptoms included cervical dysphagia (85%), expectoration of saliva (40%), and intermittent hoarseness (30%). Four patients were being fed through tubes because of total inability to swallow. "Heartburn" was experienced by one halt' of the patients, but only 12 had acid or food regurgitation. The duration of symptoms ranged from I month to /I years (average 3.9 years). Weight loss had occurred in 15 patients (38%) and ranged from 5.5 to 40.9 kg (average 16 kg). Barium swallows showed no abnormalities in 10 patients. Abnormal findings included a Zenker's diverticulum (12), prominent cricopharyngeal sphincter (II), nasopharyngeal reflux or incoordinated initiation of deglutition, or both (seven), a sliding hiatal hernia (II), and abnormal esophageal motility (seven). Esophageal manometry revealed abnormalities of upper esophageal sphincter (UES).fimction in only 16 patients. Of 36 patients undergoing standard acid reflux testing, one third had moderate-to-severe gastroesophageal reflux. Seven patients underwent staple resection of a Zenker's diverticulum at the time (~t' cervical esophagomyotomy, Postoperative complications included transient vocal cord paresis (four), vocal cord paralysis (one), and salivary fistula (one). There were no postoperative deaths. After 2 to 48 months (average 16 months) offollow-up, 34 patients (85%) have had a good to excellent result, and six (15%) have not been benefited by operation.
Mark B. Orringer, M.D., Ann Arbor, Mich.
Numerous terms have been used to define the symptom complex which results from abnormal function of the upper esophageal, or cricopharyngeal, sphincter. Such terms as cricopharyngeal chalasia, achalasia, or spasm have little more validity in these patients than the archaic diagnosis of globus hystericus, for in the vast majority, standard esophageal manometric techniques have failed to demonstrate true hypotonicity or hypertonicity of the upper esophageal sphincter (UES) or failure to relax with swallowing (achalasia). This failure may be a function of the limitations of existing instrumentaton in recording the rapid sequence of events that occurs with normal deglutition in a unique asymmetric sphincter that changes position with laryngeal excursions during swallowing. The term
From The Department of Surgery, Section of Thoracic Surgery, University of Michigan Medical Center, Ann Arbor, Mich. Read at the Sixtieth Annual Meeting of The American Association for Thoracic Surgery, San Francisco, Calif'., April 28 to 30, 1980. Address for reprints: Mark B. Orringer, M.D., Box 32, C7079, University Hospital, Ann Arbor, Mich. 48109.
"cricopharyngeal dysfunction" has been chosen to describe a spectrum of disorders of swallowing which involve the upper esophageal sphincter and occur in association with a variety of abnormalities of the central and peripheral nervous system, metabolic and inflammatory myopathy, and gastroesophageal reflux, as well as currently unidentified factors. This report describes experience at the University of Michigan Medical Center with the diagnosis and surgical treatment of cricopharyngeal dysfunction.
Patients Forty adult patients, including 18 men and 22 women ranging in age from 23 to 86 years (average 56.7 years), have undergone an extended cervical esophagomyotomy for symptoms of cricopharyngeal dysfunction. A Zenker's diverticulum was present in 12 patients (30%), and in five it was recurrent after previous resections. In two patients, cricopharyngeal dysfunction had followed midbrain cerebrovascular accidents. Additional associated conditions included symptomatic thoracic esophageal spasm (four), polymyositis (one), dermatomyositis (one), myotonic dystrophy (one), multiple
0022-5223/80/110669+10$01.00/0 © 1980 The C. V. Mosby Co.
669
The Journal of
670
Orringer
Thoracic and Cardiovascular Surgery
A
B
c
Fig. 1. Extended cervical esophagomyotomy. A, A 5 ern oblique incision anterior to the sternocleidomastoid muscle and centered over the cricoid cartilage. B, Esophagomyotomy is placed on left posterolateral aspect of esophagus to avoid injury to the recurrent laryngeal nerve, seen in the tracheoesophageal groove. C, Completed esophagomyotomy extends from the level of the superior cornu of the thyroid cartilage inferiorly to I to 2 em behind the clavicle.
sclerosis (one), multiple myeloma (one), and severe chronic cervical compression fracture (one). Four patients had undergone hiatal hernia operations 3 to 18 months earl ier, and although their reflux symptoms were controlled, cervical dysphagia had persisted; five patients had undergone prior partial or subtotal thyroidectomy for symptoms of cervical dysphagia, and three had persistent symptoms after incomplete cervical esophagomyotomies elsewhere. Four patients were being fed through feeding tubes because of total inability to swallow, and one had undergone a tracheostomy because of recurrent severe aspiration. Cervical dysphagia was the predominant symptom in 38 of these patients, being present in all but two who complained of a "pressure sensation" in the throat without associated difficulty swallowing. Among the 28 patients with intermittent symptoms, dysphagia was for solid foods in 27 and for both solids and liquids in eight. The 10 patients with constant dysphagia all had difficulty swallowing both solids and liquids. The duration of dysphagia ranged from I month to II years (average 3.9 years). Weight loss had occurred in 15 patients (37.5%) and ranged from 5.5 to 40.9 kg (average 16 kg). Nasopharyngeal regurgitation during swallowing was experienced by four patients. Respiratory symptoms were present in 33 patients (83%). Twenty-two (55%) described choking on saliva and 16 (40%) complained of "too much phlegm," with expectoration of excessive saliva. Eighteen (45%) had choked on food. A history of intermittent hoarseness
was present in 12 (30%). Three patients experienced marked shortness of breath as food slowly passed through the cervical esophagus, and in one, episodes of wheezing were precipitated by bouts of dysphagia. Although one half of these patients complained of intermittent "heartburn," in only 12 (30%) was there a history compatible with gastroesophageal reflux, that is, acid or food regurgitation, or both, aggravated by the recumbent position. Barium swallow examinations were performed in all patients and showed no abnormalities in 10. Among the 12 patients with a Zenker's diverticulum, the cricopharyngeal sphincter was reported to be prominent in II. Nasopharyngeal reflux or incoordinated initiation of deglutition, or both, was noted in seven, a sliding hiatal hernia in II, and abnormal esophageal motility in seven. Standard esophageal function tests;' performed with a combined, triple-lumen manometric-pH recording catheter, perfused with saline," were conducted in 37 patients. Twenty-four had normal thoracic esophageal peristalsis and 13 had abnormal motility patterns. Three had distal esophageal spasm, nine had frequent simultaneous contractions after swallowing, and one had changes of scleroderma. Relaxation of the distal esophageal sphincter was normal in all but the patient with scleroderma. Manometric evaluation of the UES was completely unremarkable in 21 patients. In our laboratory, peak UES pressures in asymptomatic "control" subjects
Volume 80 Number 5 November, 1980
range from 15 to 85 mm Hg and length from 2 to 5 cm (average 3.4 ern). The following manometric abnormalities of DES function were detected in 16 of the patients with cricopharyngeal dysfunction: marked delay in ability to initiate a swallow on command (five); weak pharyngeal contraction (four); incomplete relaxation of the DES to cervical esophageal baseline pressure (three); total lack of DES relaxation (achalasia) (two); hypotensive DES with maximum pressures of 5 to 7 mm Hg (two); hypertensive, spastic DES with resting pressure greater than 90 mm Hg , length between 6 and 7 em, and sustained (12 to 15 seconds) contractions after swallowing (two); and incoordination of the UES, with pharyngeal contraction occurring simultaneously with DES closure rather than relaxation (two). One third of the patients demonstrated moderate-tosevere gastroesophageal reflux with the standard acid reflux test. The acid clearance test was normal in only five of the patients, and the acid perfusion test was abnormal in 10. At the time of their cervical esophagomyotomy, seven patients underwent staple resection of a Zenker's diverticulum, one a Collis-Nissen hiatal hernia repair, and one a long thoracic esophagomyotomy for associated diffuse esophageal spasm. Technique
After induction of general anesthesia and endotracheal intubation, rigid esophagoscopy is usually performed to exclude neoplasm or inflammatory esophageal disease and to evacuate the contents of a Zenker's diverticulum. It is unwise to persist with attempts to advance the esophagoscope in a patient with an extremely tight cricopharyngeal sphincter or a large pharyngoesophageal diverticulum that prevents free passage of the instrument. The patient's neck is extended by placing a small rolled sheet beneath the scapulae. The head is turned toward the right side and the occiput supported on a head ring. The field extends from the mandible to the nipple line, and drapes are sutured to the skin to provide access to the entire left side of the neck. The cricoid cartilage is identified by palpation, and a 5 em oblique left cervical incision centered at the level of this cartilage is made parallel and anterior to the sternocleidomastoid muscle (Fig. 1, A). In young women, a transverse left cervical incision in a skin crease directly over the cricoid cartilage gives a slightly better cosmetic result. The platysma and omohyoid fascial layers are incised, and the sternocleidomastoid muscle and carotid sheath and its contents are retracted laterally and the trachea medially, The recurrent
Esophagomyotomy for cricopharyngeal dysfunction
67 I
laryngeal nerve coursing in the tracheoesophageal groove is identified and protected. The middle thyroid vein or inferior thyroid artery is ligated and divided as required. The dissection proceeds directly posterior through the cervical fascial layers to the prevertebral fascia. A No. 40 Fr. Hurst-Maloney bougie inserted into the oropharynx by the anesthetist is guided into the esophagus by the surgeon and aids in defining the posterolateral esophageal wall, which subsequently will be incised, as well as the diverticulum if it is present. When there is no diverticulum, the cervical esophagomyotomy is next performed. The incision extends from the level of the tip of the superior cornu of the thyroid cartilage inferiorly to I to 2 em below the clavicle and is 7 to 10 em long (Fig. I, B). The myotomy is deliberately placed on the left posterolateral aspect of the esophagus to avoid injury to the recurrent laryngeal nerve. Although the esophagomyotomy is longer than the skin incision, adequate exposure of the muscle being incised is achieved by exerting upward traction on the skin retractors when working superiorly and downward traction when approaching the clavicle. In addition to this shifting of skin retractors, if the first assistant applies direct upward or downward retraction on the esophagus with a small gauze square as the esophagomyotomy is performed, exposure onto the pharynx superiorly and behind the clavicle inferiorly is readily achieved (Fig. 1, C). In the presence of a Zenker's diverticulum, after the intraesophageal dilator is in position, the pouch is located and mobilized away from adjacent fibroareolar tissue. The inferior thyroid artery is an excellent landmark, as the diverticulum is inevitably found behind it when the vessel is divided. A Babcock clamp applied to the fundus of the diverticulum is used for traction that facilitates dissection of the pouch. The base of the mucosal pouch at its emergence through the esophageal muscle is identified, and the adjacent muscle is gently dissected away with a fine, right-angle clamp and divided with electrocautery and a needle-tip point (Fig. 2, A). As described earlier, the incision is extended inferiorly behind the clavicle and superiorly to the level of the superior cornu of the thyroid cartilage. When the diverticulum is 1.5 to 2 em or smaller, resection is not necessary. When larger, the neck of the pouch is crossed with a TA-30 surgical stapler using 4.8 mm staples, and the diverticulum is resected (Fig. 2, B). The staple suture line is not reinforced. A dilator present within the esophagus and avoidance of undue traction on the diverticulum prevent removal of excessive esophageal mucosa. After the neck wound has been irrigated and hemostasis along the divided muscle edge
The Journal of
672
Thoracic and Cardiovascular
Orringer
Surgery
B
A
Fig. 2. Cervical esophagomyotomy and concomitant resection of Zenker's diverticulum. A. After mobilization of the diverticulum, the esophagomyotomy is performed in either direction from the base of the pouch for the same distance as described in Fig. 1. B. After the esophagomyotomy is completed, the base of the diverticulum is crossed with a TA-30 stapler and amputated.
has been obtained, a nasogastric tube is inserted by the anesthetist and gently guided into the upper esophagus. Air is insufflated and the esophagomyotomy is tested under saline to ensure that no tear has occurred. If an air leak is identified, the tear is closed with interrupted 5-0 Prolene mucosal stitches. Silver clip markers for future roentgenographic evaluation are placed at either end of the esophagomyotomy so as to avoid contact with the mucosa that might cause erosion and perforation. The nasogastric tube is advanced into the stomach and utilized for gastric decompression until any postoperative nausea has subsided. The tube is usually removed the evening of operation, and clear liquids by mouth are begun the next day. A small Penrose drain is used to drain the wound for 24 hours. The patient is generally discharged on a soft diet by the second or third postoperative day.
Results There were no postoperati ve deaths. Transient left vocal cord paresis that resolved spontaneously in 3 to 6 weeks occurred in four patients. One patient, who had had a previous resection of a Zenker's diverticulum complicated by a salivary fistula, had a left vocal cord paralysis that was managed with a Teflon injection; this patient regained a near-normal voice. There was one salivary fistula in this series, and this occurred in a patient who had undergone three previous resections of a Zenker's diverticulum; the fistula closed after 5 days.
After 2 to 48 months (average 16 months) of follow-up, 22 patients have had an excellent result and have no difficulty swallowing a regular diet; 12 are markedly improved and experience only occasional mild dysphagia; and six (15%) have had a poor result, with no noticeable benefit from operation. Among these latter six patients are two elderly men, aged 70 and 86 years, with associated severe cerebrovascular disease and senility, and a 38-year-old woman whose chief complaint was hoarseness and throat pain, rather than dysphagia.
Discussion
In evaluating the patient with cervical dysphagia, one must exclude other local causes for this complaint: extrinsic compression of the esophagus by the thyroid gland, cervical lymphadenopathy, or hyperostosis of the cervical spine (bone spur) and intrinsic compromise of the cervical esophageal lumen by neoplasm, web, stricture, or an inflammatory process (abscess). A history of prior head and neck operations may be extremely important, as recurrent laryngeal nerve injury may result in cricopharyngeal dysfunction. This may occur not only after thyroid or parathyroid operations, but after pulmonary resections as well, since innervation of the cricopharyngeus is derived in part from the recurrent laryngeal nerve." Similarly, dysphagia is not infrequent after laryngectomy. 4 Finally, following tracheostomy, normal elevation and anterior rotation of the larynx that occurs with swallowing may be pre-
Volume 80 Number 5 November , 1980
Esophagomyotomy for cricopharyngeal dysfunction
67 3
Fig. 3. A, Prominence of the cricopharyngeal sphincter (a rrows) in a patient with cervical dysphagia and symptomatic gastroesophageal reflux. B, Detail of cervical esophagus. vented by local scar formation, resulting in cervical dysphagia." Both a barium swallow and esophagoscopy remain the most vital studies in patients with dysphagia at any level . The patient with cricopharyngeal dysfunction localizes his cervical dysphagia between the thyroid cartilage and the suprasternal notch . Expectoration of excessive saliva, a complaint of 40% of our patients, is common in those who are unable to swallow normally the I to 1.5 L of saliva produced in the mouth each day . Although the physician who elicits a history of cervical dysphagia and constant expectoration of saliva may be somewhat suspicious of the clinical significance of such complaints, the additional occurrence of hoarseness should suggest the probability of cricopharyngeal dysfunction. Hoarseness is not commonly associated with swallowing disorders, and the patient who describes both dysphagia and difficulty with his voice is seldom malingering . The inferior pharyngeal constrictor muscle, of which the cricopharyngeal sphincter is a part, may affect the vocal cords in two ways. First, contraction of the inferior pharyngeal constrictor adducts or approximates the alae of the thyroid cartilage, lengthening and tensing the vocal cords . Second, by pulling the cricoid cartilage posteriorly, the cricopharyngeal muscle acts with the lateral cricothyroid to stretch the vocal cords . Thus there is adequate physiological explanation
for the association between symptoms of cervical dysphagia and hoarseness in the patient with cricopharyngeal dysfunction . Concomitant weight loss completes the diagnostic symptom complex that in our experience justifies cervical esophagomyotomy, even in the absence of roentgenographic, manometric , or endoscopic abnormalities. Barium swallow findings of a prominent cricopharyngeal sphincter (Fig . 3), a prominent posterior cricopharyngeal bar on lateral cervical views (Fig. 4), or a Zenker diverticulum (Fig. 5) are evidence of cricopharyngeal dysfunction in the symptomatic patient; each of these findings , howe ver, has been reported in asymptomatic individuals. Despite our inability to demonstrate manometric abnormalities of the upper esophagus in the majority of patients with a Zenker diverticulum, it is our conviction that the underlying problem is neuromotor dysfunction of the sphincter. The basis for this belief is the inevitable occurrence of the mucosal pouch proximal to fibers of the cricopharyngeus muscle and the complete disappearance of 1.5 to 2 em diverticula after esophagomyotomy alone, with no resection of the diverticulum. It is seldom the size of the pouch which determines the degree of associated dysphagia; few are much larger than 3 to 4 em . The patient with a 2 mm Zenker diverticulum, however, may have as much or more cervical dyspha-
The Journal of
674
Orringer
Thoracic and Cardiovascu lar Surgery
Fig. 4. A. Typical appearance of hypertrophic cricopharyngeal sphincter (anteroposterior view). B. Lateral cervical view showing posterior cricopharyngeal bar.
Fig,S. A. Zenker 's diverticulum in a patient with an associated " patulous cardia" (arrow) and asymptomatic gastroesophageal reflux without esophagitis. B. Residual barium in 2.5 ern pouch (anteroposterior view). gia and aspiration as one with a much larger pouch . In treating the patient with a Zenker's diverticulum, therefore, the least important part of the operation is resecting the pouch; dysphagia is not the result of the pouch, but rather of the cricopharyngeal dysfunction that has produced the diverticulum , and unle ss the abnormal muscle is divided, the functional obstruction is not re-
lieved and symptoms may persist. We can only attri bute the excellent results report in patients undergoing resection of a Zenker diverticulum without concomitant esophagornyotomy" to the fact that, when a meticulous dis section of the neck of the pouch is performed, inadvertent division of the incoordinated UES muscle generally occurs . In five of our 12 patients with Zenker
Volume 80 Number 5 November, 1980
diverticul a, previous resections of the pouch had been performed elsewhere; one of these patients had undergone three prior resections . In each case , recurrence of the diverticulum had been evident within several months of the di verticulectomy . Cervical esophagomyotom y performed alone or concomitantly with diverticulecto my provided lasting control of both dysphagia and the pouch in all of the se patients. The admonition to resect the diverticulum completely in order to avoid recurrence of the pouch is nonsense . With a complete cervical esophagomyotomy and a 1.5 to 2 em residual pouch, there will be no demonstrable roentgenographic abnormality within 2 weeks of operation (Fig. 6). Surgical stapling devices greatly facilitate resection of large pouches with minimum tissue trauma , without the need to open the esophagus, and in a manner that is more efficient and superior to diverticuloplexy as described by Belsey . i Becau se the length of the upper esophageal sphincter may be quite variable on manometric evaluation, being 6 to 7 em in two of our patients , an extended 7 to 10 ern cervical esophagomyotomy is advocated as a means of ensuring complete relief of the functional obstruction resulting from the incoordinated sphincter muscle . Three patients in this series had undergone limited cervical esophagomyotomies elsewhere , but their symptoms had persisted . We stres s, therefore , (I) meticulous division of all cervical esophageal muscle fibers from the inferior pharynx to the cla vicle, (2) distending the muco sa with air insufflated through an intraesophageal tube to identify any remaining undivided muscle strands, and (3) marking either end of the myotomy with silver clips to permit accurate radiographic determination of the limits of the operation . Little damage can occur from an excessively long myotomy; the concern should be that dysphagia will persist because of too limited an esophageal incision . Standard esophageal manometric evaluation of the cricopharyngeus is made difficult by the unique characteristics of the UES. H. 9 Availability of intraesophageal microtransducer recording catheters.!" as well as the more recently described glycerin-filled spatially oriented manometry system , II may provide more precise documentation of the physiology of the cricopharyngeus than has been possible in the past. Thus, despite the well-described manometric patterns of UES dysfunction; " we have demonstrated objective motility abnormalities in only 16 of our patients . Recognizing the limitations of our current recording systems in evaluating UES function , we have nonetheless routinely evaluated our patients with standard esophageal function tests, not only to document demonstrable
Esophagomyotomy for cricopha ryngeal dysfun ction
675
Fig. 6. Anteroposterior ( A) and lateral ( B) cer vical views of esoph agus of patient shown in Fig. 5 two week s after extended cervical esophagomyotom y without resection of the diverticulum. Dysphagia was relieved and the pouch is no longer evident.
manometric abnormalities, but also to define the degree of any gastroesophageal reflux present. Cricopharyngeal dysfunction is common in patients with gastroesophageal reflux ," presumably the result of reflex muscular incoordination induced in the cervical esophagus by irritation of the lower esophagus by gastric acid . Henderson!" has reported that 50% of 200 consecutive patients with symptomatic gastroesophageal reflux had pharyngoesophageal dysphagia and that this symptom was relieved by correcting the reflux; only eight of 490 patients undergoing a hiatal hernia operation have required subsequent cricopharyngeal myotomy for persistent cervical dysphagia. Betsey? has admonished against performing a cervical esophagomyotomy in the presence of an incompetent lower esophageal sphincter, lest the patient be subject to the complications of aspiration from unchecked gastroesophageal reflux . Henderson and Marryatt;':' however , achieved successful relief of cervical dy sphagia in five patients with gastroesophageal reflu x who were either too ill for surgical correction of reflux or had dominant cricopharyngeal symptoms with minimal reflux symptoms. In our experience , the minority of patients with gastroesophageal reflux actually regurgitate acid to the level of the pharynx. Thus, when there is cricopharyngeal dysfunction in the elderly or debilitated patient or the patient with objectively demonstra-
The Journal of
676
Orringer
ble but asymptomatic reflux without esophagitis, it seems perfectly reasonable to perform a cervical esophagomyotomy , place the patient on a strict antireflux regimen, and not perform an antireflux procedure unless pulmonary symptoms from aspiration become evident. Among the 12 patients with abnormal gastroesophageal reflux in this series, four had persistent cervical dysphagia despite successful antireflux operations; only two of these had relief of dysphagia after cervical esophagomyotomy. One patient with scleroderma reflux esophagitis and marked cricopharyngeal dysfunction had an excellent response to concomitant Collis-Nissen hiatal hernia repair and cervical esophagomyotomy. The remaining seven patients were either too debilitated to tolerate an antireflux operation or had few if any typical reflux symptoms. All but one of these had a good or excellent response to cervical esophagomyotomy, and none has had difficulty with chronic aspiration. It is apparent that the diagnosis of cricopharyngeal dysfunction must frequently be made on clinical grounds in a patient complaining of cervical dysphagia, expectoration of saliva, and intermittent hoarseness, with or without weight loss. There may be no objective radiographic, manometric, or endoscopic abnormalities. A carefully performed, extended cervical esophagomyotomy relieves symptoms in a substantial number of these individuals (85%) by dividing all incoordinated pharyngoesophageal muscle, and it is a relatively low-risk procedure for which the gains may be great in appropriately selected patients. A general policy of operating upon every patient who complains of a "lump in the throat" is not being advocated. A careful history, corroborated at times by barium swallow and/or manometric findings, is essential to avoid unnecessary operations in patients in whom cervical symptoms in fact have a psychological basis.
2
3
4
5
REFERENCES Skinner DB, Hendrix TR: Diagnostic tests, Gastroesophageal Reflux and Hiatal Hernia, chap 7, DB Skinner, RH Belsey, TR Hendrix, GD Zuidema, eds., Boston, 1972, Little, Brown & Company, pp 87-106 Orringer MB, Lee R, Sloan H: Combined manometric-pH recording catheter for esophageal function tests. Ann Thorac Surg 26:581-585, 1978 Henderson RD, Boszko A, van Nostrand SWP: Pharyngoesophageal dysphagia and recurrent laryngeal nerve palsy. J THORAC CARDIOVASC SURG 68:507-512, 1974 Duranceau A, Jamieson G, Hurwitz AL, Jones RS, Postlethwait RW: Alteration in esophageal motility after laryngectomy. Am J Surg 131:30-35, 1976 Bonanno PC: Swallowing dysfunction after tracheostomy. Ann Surg 174:29-33, 1971
Thoracic and Cardiovascular Surgery
6 Welsh GF, Payne WS: The present status of the one-stage paryngoesophageal diverticulum. Rev Surg 30:370-371, 1973 7 Belsey R: Functional disease of the esophagus. J THORAC CARDIOVASC SURG 52:164-188, 1966 8 Dodds WJ, Steward ET, Hogan WJ, Stef 11, Arndorder RC: Effect of esophageal movement on intraluminal pressure recording. Gastroenterology 67:592-600, 1974 9 Dodds WJ, Stef 11, Hogan WJ: Factors determining pressure measurement accuracy by intraluminal esophageal manometry. Gastroenterology 70: 117-128, 1976 10 Pope CE II, Christensen J, Harris LD, Nelson TS, Motlet NK, Templeton F: Diseases of the esophagus (work group I). A survey of opportunities and needs in research on digestive diseases. Gastroentero logy 69: 1058-1070, 1975. II Welch RW, Luckmann J: The upper esophageal sphincter (UES) in man. Significance of radial asymmetry and precise measurement of closure strength. Gastroenterology 72:1168, 1977 12 Hurwitz AL, Duranceau A, Haddad JK: Oropharyngeal dysphagia, Disorders of Esophageal Motility, chap 7, Major Problems in Internal Medicine, vol 16, Philadelphia, 1979, W. B. Saunders Company, pp 67-84 13 Henderson RD: Disorders of the pharyngoesophageal junction, Motor Disorders of the Esophagus, chap 19, Baltimore, 1976, The Williams & Wilkins Company, pp 184-204 14 Henderson RD, Marryatt G: Cricopharyngeal myotomy as a method of treating cricopharyngeal dysphagia secondary to gastroesophageal reflux. J THORAC CARDIOVASC SURG 74:721-725, 1977
Discussion DR. DAVID B. SKINNER Chicago. l/l.
I would like to raise two questions, one of which is related to the indications for this operation. It appeared to me that Dr. Orringers series included 12 patients with a Zenker's diverticulum. Most of us would agree that myotomy is part of the treatment of that condition. There were six other patients with a neural or myopathic disorder which might indeed be associated with spasm of the sphincter. Dr. Orringer also listed six other reasons for operating, but he did not report the precise indications for operation in 16 patients. There were 21 patients in this group with normal esophageal function studies on manometry, 10 patients with a normal barium swallow, and II patients with a so-called prominent cricopharyngeal sphincter. Our radiologist tells us that this may be a normal finding. Thus I am left unconvinced about the indications for operation in these patients. I would agree that this operation can be highly valuable in patients with well-documented manometric spasm or obstruction related to neuromuscular disorders and Zenker's diverticulum. Second, I would like to question the use of the term "extended myotomy. " In the abstract Dr. Orringer says that he has used a 7 to 10 cm esophagomyotomy. The length of the esophagus from the cricopharyngeal sphincter to the aortic
Volume 80 Number 5 November, 1980
arch is 7 ern, Therefore, to do a myotomy or to advocate one of 7 to 10 em would extend the dissection below the aortic arch, which I doubt that he has done through a cervical incision. The only other way to extend a myotomy is upward, and I know of no reason for dividing the pharyngeal muscles in these conditions. It is often as much a lack of coordination and activity in the pharynx that causes the difficulty as it is the tightness of the cricopharyngeal sphincter. I would ask Dr. Orringer to clarify these two points related to indications and the meaning of the term "extended esophagomyotomy. " DR. CLEMENT HIEBERT Portland. Maine
Dr. Orringer 's paper illustrates the two principal problems-finding the patients and finding the cricopharyngeus muscle. With respect to the first problem, I am concerned over the array of esophageal conditions for which the operation is recommended. A more precise criterion would seem to be the radiologic finding of a posterior obstruction at the point where the pharynx joins the esophagus. The very thought of operating for a lump in the throat without confirmatory evidence gives me a lump in my throat. As for the second problem of identifying the cricopharyngeal muscle and sectioning it precisely, I respectfully disagree with both the concept of and the need for a long myotomy. To set about deliberately to weaken the pharynx even slightly runs counter to the whole purpose of the operation. I also disagree with the idea of removing a diverticulum. The diverticulum is but a mucosal balloon and needs only to be upended and plicated in combination with sectioning of the cricopharyngeal muscle. When the diverticulum is small it may be ignored. We have utilized these concepts in operating on 37 patients over the last 20 years and find the use of local anesthesia advantageous. These elderly patients remain alert and can swallow at once. Indeed, they can swallow (gelatin dessert) during the operation and thereby afford the surgeon a chance to view the disordered physiology. With each pharyngeal contraction the ordinarily ephemeral cricopharyngeal muscle declares itself as a posterior constricting band. Finally, I agree with Dr. Orringer that this upper sphincter of the esophagus (UES) is a legitimate concern of the thoracic surgeon; perhaps nowhere else in the entire gastrointestinal tract can so much be achieved with so short a cut. DR. VICTOR H. KAUNITZ Teton Village. Wyo.
I would like to address my remarks to the possible cause of this condition. Dr. Orringer's paper deals with swallowing difficulties referred to the neck area-95% of his patients had cervical dysphagia as their chief complaint. This is a relatively common symptom. I have had experience with some 50 patients with the same symptoms, referred often by ear, nose, and throat colleagues who could find no cervical disease. By taking a careful history, I elicited one or more symptoms typical of reflux, esophagitis in addition to the cervical dys-
Esophagomyotomy for cricopharyngeal dysfunction
67 7
phagia; namely, heartburn, lower substernal dysphagia, regurgitation of acid and/or food, and substernal pain. In Dr. Orringers series 50% had heartburn and 30% had acid or food regurgitation. I believe we are dealing with the same pathophysiological entity. What is its cause? In 75% of my group of patients, a sliding hiatal hernia was demonstrated by roentgenography. In the remainder, the presence of the hernia was proved by esophagoscopic visualization of the esophagogastric mucosal junction above the hiatal level. A sliding hiatal hernia was present at operation in all these cases, and it was corrected by means of the technique I described before this Association in 1974. All patients had complete relief of their symptoms of cervical dysphagia as well as those of reflux esophagitis. There have been no recurrence to date. These results would indicate that the cause of cricopharyngeal dysfunction, producing the symptoms of cervical dysphagia, is almost always a reflex spasm of the UES-the cricopharyngeal muscle. This spasm would seem to be induced by reflux esophagitis in the lower esophagus, caused by an incompetent lower esophageal sphincter resulting from a sliding hiatal hernia. DR. F. HENRY ELLIS, JR. Boston. Mass.
Our knowledge of esophageal physiology has increased greatly in the past 15 to 20 years but chiefly as it concerns the body of the esophagus and its lower sphincter. There is still confusion regarding normal and abnormal function of the upper esophagus, particularly the UES. Dr. Orringer has tried to bring some order to this chaos and for this we are indebted to him. I have been interested, as have others, in classifying the causes of upper esophageal dysphagia, and the slide illustrates some similarities in our respective experiences. [Slide] I have treated far fewer patients. This is because operations were restricted almost entirely to patients in whom abnormalities of esophageal function could be identified. Pharyngoesophageal diverticula, many of them with a hiatus hernia, predominate, and we are indebted to our distinguished guest, Mr. D'Arcy Sutherland, as well as to Mr. Belsey for having pointed out many years ago the importance of accompanying diverticulectomy with an esophagomyotomy. Nonetheless, we still see patients with postoperative recurrences due to failure to perform a myotomy at the time of diverticulectomy. Increasing numbers of patients with hypertension of the UES are being seen, some of them with incoordination of cricopharyngeal function. Some cases may represent early stages of a pharyngoesophageal diverticulum and they are relieved by myotomy. The poor results in my experience have occurred in patients with neuromuscular disease-three patients with amyotrophic lateral sclerosis. In these patients there was no demonstrable abnormality of motility, and cricopharyngeal myotomy failed to relieve their dysphagia. I would like to ask Dr. Orringer to elaborate on the patients with neuromuscular disease whom he described. Were his poor results in that group? I have always questioned the existence of cricopharyngeal
The Journal of
67 8
Orringer
Thoracic and Cardiovascular Surgery
"achalasia. " I note Dr. Orringer had two such patients in his series. Did their motility records really show lack of relaxation? I have never seen such a record. I have two other questions to ask Dr. Orringer. When he has been so insistent in other areas of esophageal surgery on the presentation of objective data, why has he in this instance departed from that policy? In only a minority of the patients that he operated upon could evidence of abnormality in function be presented. Second, what is the need for extending the myotomy to the extent that he has illustrated? In our patients, a short myotomy encompassing the sphincter and some of the fibers of the upper esophagus has given equally good results. DR. JOHN KENNEDY London, England
I would like to make the observation that one eighth of Dr. Orringer's patients were apparently hoarse postoperatively. This can be an annoying problem for both surgeon and patient. I wonder if he could be more specific about the patients in whom voice apparently returned. We all recognize that return of voice does not necessarily mean vocal cord function is normal again because of the remarkable ability of patients to compensate. DR. 0 R R IN G E R (Closing) I appreciate the discussers' comments, which I think help emphasize a number of key points about this topic of cricopharyngeal dysfunction. Dr. Skinner's concern about the variety of associated disorders in our patients with cricopharyngeal dysfunction allows me to emphasize again my belief that this is frequently a clinical diagnosis which may be made in a variety of illnesses. The spectrum of associated conditions in which cricopharyngeal dysfunction occurred in our patients was broad, and yet the clinical symptom complex of cervical dysphagia, respiratory symptoms, hoarseness, and weight loss was so common that it allowed us to approach all of these patients operatively in a uniform fashion. The need for an "extended" cervical esophagomyotomy , 7 to 10 cm in length, has been challenged, In our laboratory, the normal UES length in asymptomatic control patients is between 2 and 5 ern, Two of our patients with recurrent Zenker's diverticula, both of whom had had multiple operations, had a UES which measured 7 cm. I think the only way we can ensure that all abnormal cricopharyngeal muscle is divided is to be sure that a complete "extended" myotomy is performed. A Zenker's diverticulum herniates through fibers of cricopharyngeal sphincter, and if just the muscle inferior to the pouch is divided, there still may be undivided, incoordinated cricopharyngeal fibers superior to the origin or neck of the pouch. This area can then become a new site for the subsequent development of another pouch. Therefore, we carry the myotomy upward onto the pharynx to the level of the superior cornu of the thyroid cartilage and inferiorly be-
hind the clavicle. We have found no adverse effect from performing too long a cervical esophagomyotomy; in fact, four patients in our series represented failures of too conservative a cervical esophagomyotomy performed elsewhere, The average length between the silver clip markers that we place at either end of our cervical myotomy is 7 to 10 cm. As documented with postoperative radiographic studies, the inferior extent of the incision is not the aortic arch but rather just behind the clavicle. In identifying this sphincter which is supposed to be so elusive, I would like to emphasize that the incision should center on the cricoid cartilage. The cricopharyngeal sphincter arises from the cricoid cartilage, which is therefore the anatomic landmark for this operation. If the incision is centered over the cricoid cartilage and an adequate incision is made well above and well below this level, one will not miss the sphincter. Now, of course, comes the key question: Why do we advocate an operation without objective data when all of our other work has emphasized the need for objectivity in esophageal surgery? The problem is that the cricopharyngeus sphincter is unique in the gastrointestinal tract. It is not circumferential and symmetric like the pylorus or the anus. Rather, it is a muscular sling that arises from cartilage. In addition, with swallowing, this sphincter ascends and descends with laryngeal excursions. Thus, with our current techniques of manometric evaluation, it is extraordinarily difficult to obtain consistent, meaningful measurements of UES function, This is responsible for our inability to show objectively definable abnormalities in all of these patients, With some of the newer microtransducer techniques that are becoming available and the glycerin-filled, spatially oriented catheters which have recently been described, hopefully our ability to objectively document abnormalities of the UES will be improved. Finally, Dr. Kennedy has expressed concern about the reported incidence of postoperative hoarseness in our patients. We limit the skin incision to 5 cm in length and retract the larynx medially, I am certain that the four patients who had vocal cord paresis in our group had traction on the recurrent laryngeal nerve. In all cases the hoarseness resolved within 4 to 6 weeks after operation, and in each case indirect laryngoscopy confirmed that the initially paretic vocal cord had resumed its normal movement. We are not advocating an operation in every patient who complains of a lump in the throat. Both the barium swallow and esophagoscopy remain the most vital studies in patients with dysphagia at any level. A careful history, esophageal function tests, a barium swallow, and esophagoscopy are essential to exclude the various local causes for cervical dysphagia-aberrant thyroid tissue, lymphadenopathy, hyperostosis of the cervical spine, neoplasms, cervical weband to avoid unnecessary operations in patients who have a true psychological basis for their cervical complaints or a true globus hystericus.