Esophageal Mucocele: A Complication of Blind Loop Esophagus

Transcription

1 Esophageal Mucocele: A Complication of Blind Loop Esophagus M. Vinayak Kamath, M.D., Robert G. Ellison, M.D., Joseph W. Rubin, M.D., H. Victor Moore, M.D., and Ganesh P. Pai, M.D. ABSTRACT Mucocele of the bypassed esophagus is an unusual complication of esophageal replacement and has been described only in isolated references. This report is based on our experience with 6 patients in whom a mucocele developed following esophageal replacement. Esophageal replacement was performed on 37 patients over a -year period at the Medical College of Georgia Hospital. A symptomatic mucocele requiring excision developed in 3 patients with achalasia, with congenital tracheoesophageal fistula, with esophageal atresia, and with inflammatory stricture. Conduits used included stomach (4), reversed gastric tube (l), and colon (). Our experience indicates that conversion of a closedloop esophagus into a symptomatic mucocele is more likely in the presence of functioning, chronically irritated mucosa. The clinical features were referable to the mucocele itself or respiratory embarrassment therefrom. Thoracic roentgenograms and computed tomographic scans were diagnostic in verifying the presence of the esophageal mucocele. All five mucoceles arose from squamous epithelium. One of 3 patients with achalasia in whom a mucocele developed following esophageal replacement had premalignant changes in the mucosa. Based on this experience, our treatment of choice is early, complete excision of the mucocele. Esophageal replacement provides excellent long-term results in patients with advanced benign esophageal disease [l-3 and offers some degree of palliation in patients with unresectable esophageal carcinoma [4]. The native esophagus does not require removal in most benign disease; the accepted indications for esophagectomy are esophagitis [l, 5) or development of malignancy [6]. Esophageal mucocele has not been reported in most large series of patients undergoing esophageal replacement [3, 7-9 and has been described only in isolated references [lo-. The occurrence of esophageal mucocele appeared to be more frequent among our patients undergoing esophageal replacement, and this finding prompted us to review our experience with esophageal replacement over the past decade. From the Section of Thoracic and Cardiac Surgery, Medical College of Georgia, Augusta, GA. Accepted for publication Apr, 986. Address reprint requests to Dr. Kamath, Section of Thoracic and Cardiac Surgery, Medical College of Georgia, Augusta, GA Material and Methods Thirty-seven patients underwent esophageal replacement at the Medical College of Georgia Hospital between July, 975, and June, 985. There were 9 patients with carcinoma of the esophagus, 7 with intractable corrosive strictures, 5 with complicated achalasia of the cardia, with a failed primary repair following ligation of a congenital tracheoesophageal fistula, with long-gap esophageal atresia, with extensive destruction of the thoracic esophagus from a neglected foreign body, and with acquired nonmalignant tracheoesophageal fistula. The 6 patients in whom an esophageal mucocele developed serve as the basis of this report (Table ). Of the 3 adult patients with achalasia, Patients and 3 had undergone numerous operative procedures with only transient relief of dysphagia. An atonic, dilated esophagus with long-standing achalasia necessitated esophageal replacement in Patient. Patient 5 was born with a tracheoesophageal fistula complicated by vascular anomalies. The fistula was ligated and a primary anastomosis performed. This broke down, and the patient eventually required esophageal replacement. Patient 6 was referred from another institution for management of an esophageal perforation caused by a neglected foreign body. The foreign body was removed along with the adjacent esophagus, and an end-to-end anastomosis was carried out. This broke down, and the patient required diversion and mediastinal drainage for three weeks before undergoing esophageal replacement. The esophageal replacement procedure was similar in all 4 patients undergoing a gastric interposition and has been described previously [ 3. The gastroesophageal junction was stapled before transection, and the staple line was reinforced with nonabsorbable suture. Thorough irrigation of the thoracic remnant was carried out through its cervical end before it was closed with interrupted sutures. Our technique of colon interposition and of reversed gastric tube interposition was similar to the techniques described by Gross and Firestone [lo] and Anderson and Randolph [], respectively. Two of the adult patients with esophageal mucocele were seen early after esophageal replacement with clinical features referable to the mucocele. The pediatric patients, on the other hand, were seen relatively late with respiratory symptoms (Table ). Computed tomographic (CT) scanning was performed in 3 patients whose standard chest roentgenograms were inconclusive (Table 3). 63 Ann Thorac Surg 43:63-69, Mar 987

2 ~ 64 The Annals of Thoracic Surgery Vol 43 No 3 March 987 Table I. Disease, Conduit, and Incidence of Mucoceles No. of Disease Conduit Patients Mucocele Carcinoma Corrosive stricture Achalasia Congenital tracheoesophageal fistula/ esophageal atresia Foreign-body perforation Acquired tracheoesophageal fistula Total Reversed gastric tube Table. Clinical Features of Patients with Esophageal Mucocele Patient Present Esophageal No. Age (yr) Diagnosis Initial Procedure Replacement Symptoms In tervala 6 Achalasia of cardia 5 Achalasia of cardia 3 4 Achalasia of cardia 4 7 Esophageal atresia 5 6 Tracheoesophageal fistula 6 7 Foreign-body perforation Modified Heller myotomy, 97 Modified Heller myotomy, 956 Esophagostomy, gastrostomy, 978 Esophagostomy, gastrostomy, 979 Esophagostomy, gastrostomy, 98 Gastric interposition, 98 Reversed gastric tube placement, 98 R colon interposition, 977 Gastric interposition, 98 Gastric interposihon, 98 Gastric interpqsition, 98 Pain, nausea Sepsis Pain, nausea Respiratory distress Respiratory distress, sepsis Respiratory distress 6 wk wk 7 yr 3 Y'.5 yr.5 yr "Interval is that between esophageal replacement and onset of symptoms of the mucocele Table 3. lnvestixations Patient Standard No. Leukocytosis Chest Roentgenogram Barium Swallow CT Scan Unremarkable Unremarkable Unremarkable Retrocardiac mass Yes Mediastinal air-fluid Unremarkable Not performed level 3 Unremarkable Redundant colon Slow transition Retrocardiac mass 4 Unremarkable Mediastinal mass, sta- Mild gastroesophageal Not performed ple separation reflex 5 Unremarkable Mediastinal mass, sta- Not performed Not performed 6 Yes ple separation Unremarkable, staple Unremarkable Mediastinal cyst separation CT = computed tomographic; staple separation = distal esophageal staple line separation.

3 65 Kamath, Ellison, Rubin, et al: Esophageal Mucocele Table 4. operative Findings, Procedure, and Follow-up Patient FOIIOW-UP No. Findings Procedure Histopathology (Yd 7-cm cystic mass 6-cm cyst 4-cm firm mass 6-cm cyst 6-cm ill-defined cyst 8-cm cyst Chronic inflammation, muscular hypertrophy Chronic inflammation, 3 muscular hypertrophy Severe inflammation, atypia, pseudoepitheliomas Esophageal and respiratory epithelium Esophageal remnant 3 Esophageal remnant 3 Results Operative findings were similar in all 6 patients. There was considerable reaction around the esophageal mucoceles, but all could be excised with minimum spillage (Table 4). Five of the 6 patients had stratified squamous epithelium lining the cysts, and varying degrees of associated inflahmation and muscular hypertrophy. The number and distribution of both mucosal and submucosal glands were within the usual limits in these 5 patients. Atypical epithelium was noted in Patient 3. In Patient 4, the mucosa was absent in large areas of the cyst and where present, consisted of foci of attenuated squamous and respiratory (ciliated columnar) epithelium with few glands visible. Symptoms reappeared in Patient 5 four months after excisioq of the mucocele. A CT scan, performed for the first time in this child, revealed a loculated cyst at the thoracic inlet. This was excised through a right supraclavicular incision. Microscopic examination confirmed the cyst to be of esophageal origin. Patients and have been lost to follow-up but were doing well two and three years, respectively, after the last operation. Patierit 3 has required hospitalization twice subsequently for partial small-bowel obstruction managed conservatively, but has otherwise done well. All 3 patients from the pediatric group have had symptoms due to reflux. Patient 4 required a modified intrathoracic Nissen fundoplication for relief. These 3 patients are asymptomatic now on an antireflux regimen. Comment Esophageal mucocele is considered to be a rare entity. There is experimental evidence to suggest that the excluded esophagus develops into a cyst within a matter of weeks and the growth of the cyst is regulated by the pressure within it [4]. The mucous glands atrophy with higher pressure, and hence the cyst fails to attain substantial size. The destruction of the esophageal mucosa by a combination of recurrent chronic inflammation and scarring probably accounts for the paucity of reports of esophageal mucocele in esophagi excluded for corrosive esophagitis, and perhaps too few patients with esophageal carcinoma survive long enough for a mucocele to develop. Indeed, in none of our patients with these diseases did an esophageal mucocele develop following esophageal replacement. All 6 of our patients with an esophageal mucocele had diseases that left the esophageal mucosa functionally intact. The 3 adult patients were drawn from our larger experience of patients with achalasia [5]. The necessity for subsequent resection of the esophagus was anticipated in them, but the time interval between the exclusion of the esophagus and the development of symptoms from the mucocele was short, perhaps because of the preexisting stimulus toward excess mucous production, that is, stagnation of food and the consequent irritation. With improvement in technique and experience gained over the years, primary repair of the tracheoesophageal fistula is now possible in most instances [6, 7. However, esophageal replacement is often required following the breakdown of the primary anastomosis [9, 8 and in patients with "long-gap" esophageal atresia [3, 9,, the usual procedure being a colon interposition or placement of a reversed gastric tube. Presumably the drainage of the partially excluded esophageal segment into the stomach prevents the formation of the "blind loop." Cervical esophagogastrostomy was one of the earliest treatments recommended to restore the continuity of the alimentary tract following ligation of the tracheoesophageal fistula []. Orringer and Sloan [3] and Mullen and co-workers [] recommended gastric interposition for the palliation of patients with advanced carcinoma and for benign disorders of the esophagus, respectively. The ease of mobilization and the consistent nature of the blood sllpply [3] make the stomach an attractive conduit. Atwell and Harrison [4] utilized the stomach as the conduit in 6 patients: 3 with tracheoesophageal fistula, with esophageal atresia, and with a high stricture. Two patients died of pulmonary complications and

4 66 The Annals of Thoracic Surgery Vol 43 No 3 March 987 Fig I. (Patient 6.) Standard (A) posteroanterior and ( 3) lateral chest roentgenograms demonstrating the separated distal esophageal staple line (black arrows) and the associated rriediastirid rtiass (white arrows). associated congenital defects. The remaining 4 patients did well and showed no evidence of reflux esophagitis or formation of an esophageal mucocele. In our series, gastric interposition was performed in 6 of 8 patients with benign esophageal disorders. The total exclusion of the esophagus resulting from the gastric interposition could have been an important etiological factor in the development of the mucocele, though it has not been observed in similar patients in other series [5, ). It has even been suggested that closure of the gastroesophageal junction be considered as prophylaxis against reflux esophagitis in patients undergoing colon interposition [l]. The adult patients were seen with signs and symptoms referable to the mucocele itself-pain, effusion, and leukocytosis. The pediatric patients, on the other hand, seemed to tolerate the slowly growing mass up to a point; then rapidly progressive respiratory distress requiring early surgical intervention developed. Preoperative evaluation had suggested the presence of sepsis in of our patients (Patients and 6), though none of the cultures were positive. Awareness of the possible existence of an esophageal mucocele is essential, since a properly elicited history could by itself suggest the diagnosis. Physical examination is usually noncontributory. Standard thoracic roentgenograms were most useful in following these patients for the development of an esophageal mucocele (Fig ). The separation of the double rows of staples at the caudal end of the excluded esophagus represents an early feature of a mucocele, the sign preceding the onset of symptoms in all 3 of our pediatric patients (Fig ). It is perhaps indicative of a clinically covert sinus formation [3]. Disruption of the distal closure has been attributed to the peristaltic activity in the excluded esophagus [5]. This occurred in of our patients (not included in this series) with unresectable carcinoma ten days postoperatively. The consequent inflammation precluded an internal drainage procedure. The esophageal remnant and the hiatal area were widely drained externally. The patient did well, and the sump drains were withdrawn over the next three weeks. There were no apparent ill effects. A CT scan is diagnostic in verifying the presence of a mucocele as well as defining the extent of it, especially in cases where the cyst may be loculated (Patient 5) (Fig 3). Our experience indicates that resection of the mucocele should be performed as soon as the patient becomes symptomatic (sooner in patients with achalasia). Computed tomography-guided needle aspiration has been described as an option available to rule out the presence of an infected mucocele and obviate the need for surgical intervention [, and may be indicated in patients who are seen with sepsis and a small mucocele. Olsen and associates and Mannell [6] found the internal drainage procedure to be useful in patients in whom the native esophagus could not be excised because of extensive scarring or tumor in the chest. We have only limited experience with this method. The mucocele was excised in all our patients with minimal or no spillage in spite of substantial reaction and fibrosis in the periesophageal tissue planes (Fig 4). Thus a more extensive internal drainage procedure was avoided, though it could perhaps have been considered at the time of the initial operation in the 3 young patients. Another alternative consists of drainage of the esopha-

5 67 Kamath, Ellison, Rubin, et al: Esophageal Mucocele Fig. (Patient 4.)Standard lateral chest roentgenograms denionstrating progressiveseparation ufthe distal esophqyeal staple line (arrows). geal remnant externally through a sump introduced through its cervical end. Haight (7 recommended a transhiatal excision of the distal esophageal segment in patients with long-gap esophageal atresia undergoing esophageal replacement, since patient in his series died of an intrathoracic perforation of the segment while awaiting elective esophageal replacement. We have used this approach in patients with carcinoma, but it could perhaps be considered for benign disease as well if the intrathoracic inflammation and fibrosis are expected to be minimal. The epithelium of the esophageal mucocele was stratified squamous in 5 patients and mixed (with ciliated columnar) in patient. This finding is in variance with the widely held belief that the columnar epithelium in the esophagus is more likely to produce mucocele than squamous epithelium, a belief based, perhaps, on the fact that esophageal glands are more profuse at both ends of the esophagus [8] and that the majority of the congenital esophageal cysts occur in the lower half of the esophagus and are lined by columnar epithelium [9]. Olsen and co-workers [ recommended that columnar epithelium be avoided in the excluded esophageal segment to prevent the formation of an infected esophageal mucocele. Our experience indicates that the squamous epithelium is also capable of producing an esophageal mucocele. However, it is possible that in Patient 4, had it not been for the location and the limited space in the mediastinum of the child, the cyst would not have produced symptoms, as the epithelium had undergone considerable degeneration at the time of excision.

6 68 The Annals of Thoracic Surgery Vol 43 No 3 March 987 In conclusion, we believe that the development of an esophageal mucocele is a distinct possibility in patients undergoing esophageal replacement for benign disease and is more likely in the presence of functioning, chronically inflamed mucosa. We suggest that patients undergoing esophageal replacement for achalasia should have early excision of the esophageal remnant in view of the additional risk of development of a clinically covert carcinoma in the excluded segment. Close followup of the asymptomatic patient with an excluded esophagus is essential if an expectant approach is to be taken, since symptoms may appear and progress with alarming rapidity. Complete excision of the mucocele is our treatment of choice. Fig 3. (Patient.) Cortiputd torrityraphir srati of the niediostiriurti dernoristrating the rsophagea/ ttiucocde (white arrows) arid gastric interposition conduit (black arrow). Of the 3 patients with achalasia who underwent excision of an esophageal mucocele following esophageal replacement, Patient 3 was the only patient in whom there was a long interval (seven years) between the two procedures. The premalignant changes discovered in this patient, though unexpected, are in accordance with the finding of others [3, 3) that patients with achalasia have an increased incidence of carcinoma of the esophagus. Indeed, the changes noted are similar to those found by Rake (3 in his review of 5 patients, 3 of whom had coexisting esophageal carcinoma. It may be speculated that premalignant changes would also have developed in the other patients with achalasia in our series had the esophagus been left in situ for an extended period. Reflux-related problems were encountered in all 3 of our pediatric patients with gastric interposition, and in our opinion, these problems are a major disadvantage of the procedure. This view is in concurrence with the opinion of other observers [33, 34. References. Neville WE, Najem A%: replacement of the esophagus for congenital and benign disease. Ann Thorac Surg 36:66, 983. Anderson KD, Randolph JG: Gastric tube interposition: a satisfactory alternative to the colon for esophageal replacement in children. Ann Thorac Surg 5:5, Kelly JP, Shackelford GD, Roper CL: Esophageal replacement with colon in children: functional results and longterm growth. Ann Thorac Surg 36:634, Mannell A: Presternal gastric bypass for unresectable carcinoma of the thoracic esophagus: a preliminary report. Br J Surg 675, Orringer MB, Kirsh M.M, Sloan H: New trends in esophageal replacement for benign disease. Ann Thorac Surg 3:49, Joske RA, Benedict EB: The role of benign esophageal obstruction in the development of carcinoma of the esophagus. Gastroenterology 36:749, Mansour KA, Hansen HA, Hersh T, et al: interposition for advanced nonmalignant esophageal stricture: experience with 4 patients. Ann Thorac Surg 3:584, Hankins JR, Cole FN, McLaughlin IS: interposition for benign esophageal disease: experience with 3 patients. Ann Thorac Surg 379, Lindahl H, Louhimo. Virkola K: interposition or gastric tube? Follow-up study of colon-esophagus and gastric tube-esophagus patients. J Pediatr Surg 8:58, 983. Gross RE, Firestone FN: ic reconstruction of the esophagus in infants and children. Surgery 6:955, 967. Johnson J, Schwegman CW, Kirby CK: Esophageal exclusion for persistent fistula following spontaneous rupture of the esophagus. J Thorac Surg 3:87, 956. Olsen CO, Hopkins RA, Postlethwait RW: Management of an infected mucocele occurring in a bypassed excluded esophageal segment. Ann Thorac Surg 4:73, Orringer MB, Sloan H: Substernal gastric bypass of the excluded thoracic esophagus for palliation of esophageal carcinoma. J Thorac Cardiovasc Surg 7:836, Deaton WR, Bradshaw HH: The fate of an isolated segment of the esophagus. J Thorac Surg 3:57, Pai GP, Ellison RG, Rubin JW, Moore HV: Two decades of experience with modified Heller s myotomy for achalasia. Ann Thorac Surg 38:, Holder TM: Current trends in the management of esophageal atresia and tracheoesophageal fistula. Am Surg 44:3, 978

7 69 Kamath, Ellison, Rubin, et al: Esophageal Mucocele 7. Strodel WE, Coran AG, Kirsh MM, et al: Esophageal atresia; a 4-year experience. Arch Surg 4:53, Richardson JV, Heintz SE, Rossi NP, et al: Esophageal atresia and tracheoesophageal fistula. Ann Thorac Surg 9:364, Freeman NV, Cass DT: interposition: a modification of the Waterston technique using the normal esophageal route. J Pediatr Surg 7:7, 98. May IA, Samson PC: Esophageal reconstruction and replacements. Ann Thorac Surg 7:49, 969. Sweet RH: A new method of restoring continuity of the alimentary canal in cases of congenital atresia of the esophagus with tracheoesophageal fistula not treated by immediate primary anastomosis. Ann Surg 7:757, 948. Mullen DC, Young WG Jr, Sealy WC: Results of twenty years experience with esophageal replacement for benign disorders. Ann Thorac Surg 5:48, Thomas DM, Langford RM, Russell RCG, LeQuesne LP: The anatomical basis for gastric mobilization in total oesophagectomy. Br J Surg 66:3, Atwell JD, Harrison GSM: Observation on the role of esophagogastrostomy in infancy and childhood with particular reference to the long-term results and operative mortality. J Pediatr Surg 5:33, Duranceau A, Jamieson GG: Malignant tracheoesophageal fistula (collective review). Ann Thorac Surg 37346, Mannell A: The Kirschner operation for cancer of the esophagus. Ann R Coll Surg Engl 64:56, Haight C: Some observations on esophageal atresias and tracheoesophageal fistulae of congenital origin. J Thorac Surg 34:4, Fawcett DW: The esophagus and stomach. In Bloom W, Fawcett DW (eds): A Textbook of Histology. Philadelphia, Saunders, 986, pp Totten RS, Stout AP, Humphreys GH, Moore RL: Benign tumors and cysts of the esophagus. J Thorac Surg 5:66, Hankins JR, McLaughlin JS: The association of carcinoma of the esophagus with achalasia. J Thorac Cardiovasc Surg 69:355, Just-Viera JO, Haight C: Achalasia and carcinoma of the esophagus. Surg Gynecol Obstet 8:8, Rake G: Epithelioma of the oesophagus in association with achalasia of the cardia. Lancet :68, Skinner DB: Esophageal reconstruction. Am J Surg 39:8, Hanna EA, Harrison AW, Derrick JR: Long-term results of visceral esophageal substitutes. Ann ThoracSurg3:,967