Ann Thorac Surg 2005;79:1890-1894
© 2005 The Society of Thoracic Surgeons
Original article: General thoracic
Pharyngocolonic Anastomosis for Esophageal Reconstruction in Corrosive Esophageal Stricture
Yao-Guang Jiang, MD*,*,
Yi-Dan Lin, MD*,
Ru-Wen Wang, MD,
Jing-Hai Zhou, MD,
Tai-Qian Gong, MD,
Zheng Ma, MD,
Yun-Ping Zhao, MD,
Qun-You Tan, MD
Thoracic Surgery Department, Daping Hospital, Chongqing, China
Accepted for publication December 28, 2004.
* Address reprint requests to Dr Jiang, Daping Hospital, Thoracocardiovascular Surgery Dept, Changjiang St 10#, Daping, Chongqing, 400042, China (E-mail: daniel{at}mail.dph-fsi.com).
 |
Abstract
|
|---|
BACKGROUND: The aim of our study is to observe the outcome of pharyngocolonic anastomosis in esophageal reconstruction for diffuse corrosive esophageal stricture involving hypopharynx.
METHODS: This is a retrospective report of the experience and results of 14 patients undergoing esophageal reconstruction with pharyngocolonic anastomosis without resection of the strictured intrathoracic esophagus. The left colonic segment was pulled up to the neck through the substernal space in all patients.
RESULTS: There was no operative or hospital death. Postoperative complications included cervical anastomotic leakage in 4 patients, rupture of abdominal incision in 1 patient, and aspiration pneumonia in 2 patients. The length of follow-up ranged from half a year to 10 years, with an average of 4 years. Anastomotic stenosis occurred in 2 patients. One patient was improved after dilatation and the other was relieved by plastic operation. One patient began to have vomiting after meals 7 months after surgery and was found to have redundant abdominal colon graft, which was corrected with a side-to-side anastomosis between the colon and the stomach.
CONCLUSIONS: A successful reconstruction for hypopharyngoesophageal stricture requires a sufficiently large hypopharyngocolonic anastomosis and a technique of good anastomosis. From our experience, this procedure is shown to be safe and effective.
|
Introduction
|
|---|
Caustic injuries to the esophagus remain more difficult to treat than other acquired esophageal disease because of the length and severity of cicatricial contraction of the strictured esophagus. Pharyngeal anastomosis is necessary in about 10% [1, 2] of these patients. In contrast to the use of the stomach as an esophageal substitute, the use of the colon for interposition is becoming less frequent. But long-segment colon interposition is ideally applied in patients with acquired esophageal disease in whom other visceral conduits are unavailable or their length cannot span the distance during transposition. In contrast to other visceral conduits, long-segment colon can be more easily transposed through different routes in reaching the pharyngeal site. Either the right or left colon may be used for interposition, and the middle colic artery is the sole route of blood supply. In comparison, the collateral circulation between the middle colic and left colic vessels is more abundant and persistent, whereas that between middle colic and right colic or ileocolic artery is not so constant, so sometimes it may be mandatory also to do a microvascular anastomosis to augment regional blood flow to the ascending colon for fear of graft ischemia. Therefore, long-segment left colon is more preferable to be chosen as an esophageal substitute, especially for those needing pharyngeal anastomosis; not only is it more preferable to the stomach but also to the right colon. Of course, the stomach is still commonly used as an esophageal substitute if it is found to be of sufficient length in reaching the hypopharynx. We therefore prefer to choose the left long-segment colon for esophageal substitution when pharyngeal anastomosis is necessary and, at the same time, the length of the stomach is inadequate.
|
Material and Methods
|
|---|
Medical records were reviewed for every patient admitted at Daping Hospital because of caustic injuries to the esophagus from August 1988 to June 2003. Operative mortality was defined as any death within 30 days after operation or during hospitalization. Follow-up, obtained by review of office records or telephone interview, was complete through July 2003.
One hundred forty-eight cases of burns to the esophagus were identified. Ninety-two underwent surgical intervention, including 25 using the stomach and 67 using a long-segment colon pulled up into the neck as an esophageal substitution. Fourteen of the 67 patients using long-segment colon interposition underwent pharyngocolonic anastomosis for esophageal reconstruction. Among these 14 cases there were 9 male and 5 female patients (mean age, 17±15 years; range, 4 to 35 years). The traumatic agents included alkali in 12 cases, unknown factor in 1, and melted boiling liquid iron in 1 (tracheotomy already performed at a community clinic because of concomitant severe burn to the larynx).
Preoperative evaluation included upper gastrointestinal endoscopy (Fig 1), direct laryngoscopy, bronchoscopy, barium meal or iodine contrast study (Fig 2), and barium enema. Attempts were made to explore the patency and integrity of the larynx, pharynx, esophagus, and colon. Mesenteric angiography was not performed routinely for economic reasons. Preparation of the colon consisted of clear liquid diet for 3 days and cleansing enema done in all patients. The time interval from burn to esophageal reconstruction was more than 6 months, with an average of 17.5 months. During the preoperative period, 13 patients underwent feeding gastrostomy or jejunostomy for enteral nutrition support; the remaining patient sustained oral intake with the help of repeated dilatations.
View larger version (123K):
[in this window]
[in a new window]
|
Fig 1. A picture taken during endoscopy of a female patient suffering corrosive esophageal stricture after a mistaken ingestion of sodium hydroxide. The hypopharynx was badly occluded; even saliva could not pass through it.
|
|
View larger version (161K):
[in this window]
[in a new window]
|
Fig 2. Photographs of iodine contrast study of the same patient before and after operation. (A) The diffuse corrosive stricture (left arrow) involved the upper esophagus and hypopharynx, and because of aspiration the left and right main bronchus were shown. (B) For esophageal reconstruction a left long-segment colon was used as a substitute, and the orifice of the pharyngocolonic anastomosis (right arrow) was sufficiently wide for a fluent passage of bolus.
|
|
Operation was performed under general anesthesia. The patient was placed in the supine position, the head was turned to the right, and the back was slightly elevated with a small pillow.
Exposure of the colon was performed through a midline laparotomy incision. Retroperitoneal attachments of the colon were incised to free the transverse colon, splenic flexure, and descending colon, and the middle and left colic arteries and their communication arcuates in the mesentery were identified through transillumination. The main trunk of the left colic artery was then temporarily occluded to assess the adequacy of collateral circulation. Thereafter it was ligated and divided after a careful confirmation of an adequate arterial supply to the selected conduit. The transverse colon was transected between intestinal clamps after the length of conduit required was determined. A colocolostomy was performed between the remaining segments of colon, and the mesocolic defect was closed. The conduit was then carefully transposed through the anterior mediastinum, avoiding entry into the pleural space. Mobilization of the colon segment through the mediastinum was facilitated with the use of liquid paraffin smeared on the graft. Care was taken during transposition to avoid any torsion of the vascular pedicle. The thoracic inlet was widened by blunt or sharp dissection of the regional ligaments to avoid any compression on the proximal part of the graft; partial excision of the osseous structures nearby was reserved only for the one case with a narrow thoracic inlet. An end-to-side cologastric anastomosis was performed using a two-layered suture method. A pyloromyotomy was usually performed for better gastric emptying.
A cervical incision was made along the anterior border of the left sternocleidomastoid muscle. The sternocleidomastoid muscle was retracted laterally together with the common carotid sheath. Blunt dissection of esophagus was extended downward along the prevertebral space and behind the trachea. The esophagus was encircled with a traction tape (Fig 3). Then the damaged cervical esophagus or hypopharynx was resected, and the upper end of the remaining intrathoracic esophagus was closed, leaving the strictured intrathoracic esophagus in situs.
View larger version (64K):
[in this window]
[in a new window]
|
Fig 3. A cervical incision was made along the anterior border of the left sternocleidomastoid muscle (2) to explore the damaged cervical esophagus (1) and the hypopharynx (4). The thyroid gland (5) and sternocleidomastoid muscle were mobilized and retracted laterally together with the common carotid sheath (3). The damaged esophagus was encircled with a traction tape after blunt dissection of the esophagus extended downward along the prevertebral space and behind the trachea.
|
|
Colopharyngeal anastomosis was performed in an end-to-end fashion (Fig 4). The proximal end of the colon graft was tailored, with a 1- to 1.5-cm vertical myotomy on its anterior wall, in accordance to the size of the remaining hypopharynx. An anastomosis was performed using a one-layered suture method, the margins of which were 0.5 to 1.0 cm wide, a skill that we called "wide margins, single-layered interrupted suture method" (Fig 3). This procedure helped to create a sufficiently large orifice of the colopharyngeal anastomosis (Fig 5). Also, an adjacent tissue buttress was applied to protect the anastomosis.
View larger version (64K):
[in this window]
[in a new window]
|
Fig 4. Colopharyngeal anastomosis was performed in an end-to-end fashion. A left long-segment colon (1) was used as the substitute. To make the orifice of the anastomosis sufficiently wide, a 1- to 1.5-cm vertical myotomy (2) was cut on the anterior wall of the proximal end of the colon graft to match the size of the defect of the hypopharynx (3).
|
|
View larger version (61K):
[in this window]
[in a new window]
|
Fig 5. Colopharyngeal anastomosis was completed using a skill that we called "wide margins, single-layered interrupted suture." This procedure helped to create a sufficiently wide orifice of the anastomosis.
|
|
A proximal nasogastric tube was used to provide postoperative decompression of the stomach. A feeding gastrostomy or jejunostomy was created or maintained to provide enteral nutritional support postoperatively. A Penrose drainage was placed at the anastomotic site, and the cervical incision was loosely closed.
In general, the abdominal and cervical procedures could be performed simultaneously by two groups of surgeons. However, we preferred to initiate our surgical procedure with cervical exploration if the anatomic structures in the hypopharynx and cervical esophagus had been presumed before the operation to be badly destroyed, or the residual space of the hypopharynx was occluded with scar tissues. In such conditions, a tracheal tube is introduced through the oropharynx downward to the hypopharynx as an indicator for anatomic marking. This tube guidance could facilitate the ease of safe exploration, cricopharyngeal myotomy, or even laryngectomy. Concomitant laryngectomy was performed only when the larynx was completely destroyed.
Usually, oral intake was begun after a barium meal radiographic study demonstrating an intact anastomosis on or after the 7th postoperative day. Diet was advanced from liquid to semiliquids and then to a regular diet.
At follow-up, all surviving patients were asked about body weight, any dysphagia, type of diet, and need for nutritional supplements.
|
Results
|
|---|
Preoperative endoscopic evaluation revealed the piriform fossa was occluded with scar tissue in 4 patients and the hypopharyngeal space was highly strictured in the other 10 patients. During operation, in all 14 patients, the left long-segment colon was selected and transposed as a substitute in the antiperistaltic position, pulled through the anterior mediastinum. The strictured cervical esophagus and hypopharynx were resected, and the strictured intrathoracic esophagus was left in situs. A proximal anastomosis was performed at the level of the hypopharynx, and a distal anastomosis was done to the stomach. A feeding gastric tube decompression or a feeding gastrostomy or jejunostomy was performed either before or concomitant with colon interposition. In 4 cases of occluded piriform fossa, a tracheal tube was applied as a guide to help the cervical procedure; in the case with an associated destroyed larynx, a concomitant laryngectomy was performed. Laryngeal function was well preserved in all patients except the one with concomitant laryngectomy. There was no operative death or graft ischemia, and the conserved esophagus was without any subsequent problem.
Early leakage at the colopharyngeal anastomosis site was seen in 4 patients on around the 10th postoperative day, and was successfully cured by conservative treatment. Disruption of the abdominal incision occurred in 1 patient on the 10th postoperative day, and was successfully cured by resuturing. Aspiration pneumonia, occurring in 2 patients, was ascribed mostly to the recurrent nerve damage, leading to glottic dysfunction in the early postoperative stage. Postoperative functional dysphagia observed in 4 patients was corrected after rehabilitative training for deglutition.
The length of follow-up ranged from a half to 10 years with an average of 4 years. All five cases of late complications occurred within 1 year. One patient each developed proximal anastomotic stenosis 6 and 7 months after operation. Both were initially managed by dilatation. This procedure was successful in 1 patient, whereas the other one required plastic surgery eventually. Distal anastomotic stenosis was seen in 1 patient 5 months after operation, also requiring anastomotic plastic surgery. In the last 2 patients, a progressive postprandial vomiting was seen 4 and 7 months after operation. By barium studies, the abdominal portion of the colon graft was found to be either excessively redundant in 1 patient or compressed in the other. The redundant graft appeared like a sac and was resolved by a side-to-side cologastrostomy. The compression proved during operation to be caused by a tight adhesion band was surgically relieved. At the end of follow-up, all 14 patients gained body weight having regular oral diet. Their larynx functioned well except the one having lost it eternally as a result of concomitant laryngectomy.
|
Comment
|
|---|
Alkali is the most common traumatic agent in this series, in which 12 of 14 (85.7%) patients were burned by sodium hydroxide. This phenomenon has a specific cultural background. In the city of Chongqing (with a population of 30 million), the fourth-largest municipality of China, the native people have a custom of using sodium hydroxide to pretreat foods, such as chitlings, to improve their taste. However, the storage of this strong corrosive agent is often merely in a drinking bottle carelessly placed without any warning sign. So sodium hydroxide could be mistaken for water because the two look alike. Under this condition, burns in the stomach rarely happen, but corrosive stricture may often be diffuse or also involve the hypopharynx. Thus, this constitutes a dilemma for medical intervention.
Colon interposition for esophageal substitution constituted 72.8% (67 of 92) of our patients undergoing esophageal replacement (Fig 6). In the context of the subtotal esophageal replacement when other visceral substitutes are available, colon conduits do not offer any inherent advantage outweighing in operative time, complexity, and technical demands of reconstruction. However, the primary advantage of a colon graft is its possible reach in length span. The unique anatomic attributes of long segments of the colon and their functional capabilities make them an attractive choice for total esophageal substitution. Replacement of the esophagus with a segment of isoperistaltic transverse colon was attempted by Kelling as early as 1911 [3]. In the same year, Vulliet [4] described the use of the transverse colon but in an antiperistaltic position. In 1950, Orsini and Toupet [5] transposed a left colon segment. In recent reports [6, 7], the ileocolon is favored by most surgeons. However, our experiences here have demonstrated that using the left long-segment colon for esophageal substitution in an antiperistaltic position is a relatively simple procedure, which can also assure successful anastomosis, few complications, and satisfactory postoperative deglutition function. The long-segment colon interposition is ideally applied in patients with acquired esophageal disease in whom other visceral conduits are not available or their length cannot span the distance during transposition. Also, a surgeon familiar with the technical demands of this procedure is a prerequisite component of this approach.
View larger version (31K):
[in this window]
[in a new window]
|
Fig 6. Decision tree for choosing different organs as an esophageal substitute according to the location of the strictured site on the esophagus.
|
|
The interval between the caustic burn and the diagnosis of scar carcinoma at strictured sites is reported to be as long as 30 to 46 years, and the highest incidence of scar carcinoma was reported to be 7.2% [8]. Considering the high incidence of cicatricial carcinoma, some surgeons are prone to performing simultaneous resection of the strictured esophagus [9, 10]. However, others still reserve it for fearing of neighboring organ damage when ablating the diffuse or multiple strictures [11]. In our practice, we prefer to reserve the damaged esophagus when it is diffuse or multiple, extending up above the aortic arch, and to pull up the long-segment colon into the neck as a substitute. We prefer to ablate the damaged segment of esophagus if the stricture is limited to the portion below the aortic arch, and use the stomach for substitution in the thorax (Fig 6). In this series, we left strictured intrathoracic esophagus in situs without noticing any relevant subsequent problem during follow-up. But it is of no doubt that these patients still need intensive follow-up in the future.
At the end of the oral phase of deglutition, the tongue propels the food bolus into the pharynx. This triggers the pharyngeal phase, during which the upper esophageal sphincter relaxes transiently, thus allowing the passage of the bolus. The movement of the bolus during the pharyngeal phase is very rapid (within 1 second) and in cooperation with epiglottis movement prevents aspiration. We therefore believe that a larger orifice of the anastomosis is of key importance in assuring a smooth swallowing, avoiding aspiration, and preventing subsequent anastomotic stenosis. In this series only 1 of 14 patients (7.1%) needed surgical revision because of subsequent stenosis at the site of pharyngocolonic anastomosis.
Four patients experienced subsequent dysphagia and aspiration but recovered rapidly after the rehabilitative course of training for deglutition. We believe that there should be a relearning course of how to swallow in such patients who had not been capable of swallowing for more than half a year. We are now planning to set up a rational program for judging and training for full functional recovery of deglutition after esophageal reconstruction with colon interposition and pharyngocolonic anastomosis.
|
Footnotes
|
|---|
* The authors wish it to be known publicly that the first two authors (Yao-Guang Jiang and Yi-Dan Lin) should be regarded as joint first authors. 
|
References
|
|---|
- Choi RS, Lillehei CW, Lund DP, et al. Esophageal replacement in children who have caustic pharyngoesophageal strictures J Pediatr Surg 1997;32:1083-1087.[Medline]
- Huang MD, Sung CY, Hsu HK, et al. Reconstruction of the esophagus with the left colon Ann Thorac Surg 1989;48:660-664.[Abstract]
- Kelling G. Ösophagoplastik mit Hilfe des Querkolon Zentralbl Chir 1911;38:1209-1212.
- Vulliet H. De l’oesophagoplastic et de ses diverse modifications Semain Med 1911;45:529-532.
- Orsini P, Toupe A. Utilization of descending colon and left portion of transverse colon in prethoracic esophagoplasty Presse Med 1950;58:804-808.
- Popovici Z. Results of the surgical treatment of severe caustic pharyngo-esophageal stenosis. The value of complete reconstruction of the pharynx by transposition of the ileum and colon Chirurgie 1998;123:552-559.[Medline]
- Wu MH, Tseng YL, Lin MY, et al. Esophageal reconstruction for hypopharyngoesophageal strictures after corrosive injury Eur J Cardiothorac Surg 2001;19:400-405.[Abstract/Free Full Text]
- Csikos M, Horvath O, Petri A, et al. Late malignant transformation of chronic corrosive esophageal strictures Langenbecks Arch Chir 1985;265:231-238.
- Kim YT, Sung SW, Kim JH, et al. Is it necessary to resect the diseased esophagus in performing reconstruction for corrosive esophageal stricture? Eur J Cardiothorac Surg 2001;20:1-6.[Abstract/Free Full Text]
- Gupta NM, Gupta R. Transhiatal esophageal resection for corrosive injury Ann Surg 2004;239:359-363.[Medline]
- Raffensperger JG, Luck SR, Reynolds M, et al. Intestinal bypass of the esophagus J Pediatr Surg 1996;31:38-46.[Medline]
This article has been cited by other articles:
|
|
|
|
 
P. Yannopoulos, D. Lytras, and K. I. Paraskevas
Esophageal reconstruction with intraoperative dilatation of the hypopharynx for the management of chronic corrosive esophageal strictures.: A technical tip
Eur. J. Cardiothorac. Surg.,
December 1, 2006;
30(6):
940 - 942.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
R.-W. Wang, J.-H. Zhou, Y.-G. Jiang, S.-Z. Fan, T.-Q. Gong, Y.-P. Zhao, Q.-Y. Tan, and Y.-D. Lin
Prevention of stricture with intraluminal stenting through laparotomy after corrosive esophageal burns.
Eur. J. Cardiothorac. Surg.,
August 1, 2006;
30(2):
207 - 211.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
