|
|
||||||||
Ann Thorac Surg 1999;67:313-317
© 1999 The Society of Thoracic Surgeons
a Division of Thoracic Surgery, Massachusetts General Hospital, Boston, Massachusetts, USA
Address reprint requests to Dr Wain, Division of Thoracic Surgery, Massachusetts General Hospital, Blake 1570, Fruit St, Boston MA 02114
e-mail: wain.john{at}mgh.harvard.edu
Presented at the Thirty-fourth Annual Meeting of The Society of Thoracic Surgeons, New Orleans, LA, Jan 2628, 1998.
| Abstract |
|---|
|
|
|---|
Methods. We reviewed the medical records, office visits, and operative reports of patients undergoing long colon interposition for acquired esophageal disease at our institution from 1956 to 1997.
Results. Long colon interposition was performed in 52 patients for caustic injury (n = 20), gastroesophageal disease (n = 16), previous irradiation (n = 8), primary motility disorders (n = 4), and acquired absence of the esophagus (n = 4). From 1976 to 1997, acquired diseases accounted for 62% of long colon interposition. The left colon was used in 46 patients and the right colon in 6. The in-hospital mortality rate was 4%. Early complications included graft ischemia in 5 patients, anastomotic leak in 3, and small bowel obstruction in 1. Late complications included anastomotic stenosis requiring dilation in 26 patients, with 2 requiring surgical revision, and bile reflux requiring surgical diversion in 1 patient. Swallowing function was excellent in 24% of patients, good in 66%, and poor in 10%.
Conclusions. Long colon interposition can be performed safely, with acceptable long-term functional results in patients with acquired esophageal disease.
| Introduction |
|---|
|
|
|---|
The unique anatomic attributes of long segments of colon and their functional capabilities make them an attractive choice for total esophageal substitution in patients with acquired esophageal disease. The underlying esophageal pathophysiology, the potential need for a longer interposition graft and the likely long-term survival of these patients may offset the technical complexity and inherent risks associated with colon interposition. Long-segment colon substitution for the esophagus has been performed at the Massachusetts General Hospital since 1956. Our purpose in the present study was to review our experience with this procedure in patients with acquired esophageal disease to assess our results in this population and to identify technical factors that may result in an improved outcome.
| Material and methods |
|---|
|
|
|---|
One hundred fifty-four cases of long-segment colon interposition were identified. Acquired esophageal disease was the indication for the procedure in 52 patients, including 20 of 102 procedures performed from 1956 to 1976 and 32 of 52 procedures performed from 1976 to 1997. Specific indications included intractable stricture in 44 patients (caustic injury in 20, gastroesophageal reflux disease in 16, previous irradiation in 8), primary motility disorders in 4 patients, and acquired absence of the esophagus in 4 patients (trauma in 2, previous neoplastic resection in 2). There were 32 male and 20 female patients with acquired esophageal disease (mean age 41 ± 7 years).
Preoperative evaluation included upper gastrointestinal endoscopy, direct laryngoscopy, and bronchoscopy. Selective mesenteric angiography of the celiac axis, superior mesenteric artery, and inferior mesenteric artery was performed routinely from 1977 to 1996. Attempts were made to identify the presence and continuity of a marginal colic artery; the patency of the left colic artery, particularly its ascending branch; and the anatomy of the bifurcation of the middle colic artery. Barium enema was usually performed unless specific symptoms, such as hematochezia or guaiac-positive stools, mandated colonoscopy. A mechanical preparation of the colon consisting of a clear liquid diet for 3 days and a antegrade colonic lavage was used in all patients.
Exposure of the colon was performed by laparotomy, usually a midline approach. The greater omentum was mobilized from the transverse mesocolon. Retroperitoneal attachments of the colon were incised, freeing the transverse colon, splenic flexure, and descending colon for left colon grafts and the terminal ileum, ascending colon, and transverse colon for right colon grafts. The marginal colic artery and middle colic arteries were identified in the mesentery by palpation and transillumination. For left colon grafts, the ascending branch of the left colic artery was isolated, and after division of the mesentery between this vessel and the middle colic artery, the main trunk of the middle colic artery was clamped to assess the adequacy of the circulation to the intervening colon (Fig 1). For right colon grafts, the main trunk of the middle colic artery was isolated, and after incision of the mesentery between this vessel and the ileocolic artery, the right and ileocolic arteries were clamped in an analogous maneuver. The collateral circulation is frequently incomplete between the middle colic vessels and the right and ileocolic vessels, which must be divided to complete conduit mobilization. Augmentation of the blood supply to the proximal (ileocolic) portion of the transposed segment may be performed by microvascular anastomosis of regional vessels to the divided ileocolic or right colic branches. After careful confirmation of an adequate arterial supply to the selected conduit, the clamped vessels were ligated close to their origins to preserve additional intramesenteric arterial and venous collateral flow.
|
Cologastric anastomosis was performed in an end-to-side fashion using two layers of suture material. The anastomosis was typically completed to the midportion of the posterior wall of the stomach for left colon conduits and on the anterior wall of the stomach for right colon conduits. A 10- to 12-cm length of intraabdominal colon was maintained to minimize the likelihood of significant cologastric reflux. In circumstances where the stomach was not available, the distal cologastric anastomosis was performed to a Roux-en-Y loop of jejunum. Gastric drainage procedures, either a pyloromyotomy or a Heineke-Mikulicz pyloromyotomy, were performed when vagal transection had occurred.
Esophagocolic or pharyngocolic anastomosis was performed in an end-to-end fashion using two layers of interrupted suture. When hypopharyngeal or pharyngeal anastomosis was required, the proximal end of the colon was tailored to the size of the mucosal defect. Selected sutures were used to fix the proximal colon to the prevertebral fascia and the distal portion of the graft to the diaphragm. A proximal nasogastric tube and a distal gastrostomy tube were used to provide diversion of native secretions. A feeding jejunostomy was frequently used to provide enteral nutrition postoperatively.
Oral intake was begun after Gastrografin (E.R. Squibb & Sons, Princeton, NJ) and barium radiographic studies demonstrated intact anastomoses, typically on or after the seventh postoperative day. Diet was advanced from liquids to pureed food and then solids. At follow-up, all surviving patients were queried with regard to weight, dysphagia, type of diet, and need for nutritional supplements.
| Results |
|---|
|
|
|---|
Colon conduits consisted of left colon segments in 46 cases and right colon segments in 6 cases. In 2 cases, both right colon conduits, augmentation of the distal arterial supply using a microvascular anastomosis to the internal mammary artery was required. In 1 patient this maneuver was performed electively because of limited conduit availability and in another patient after intraoperative evidence of ischemia of the proximal portion of the graft was noted (Fig 2).
|
Two (4%) of 52 patients died after the procedure during the initial hospital stay (1 of adult respiratory distress syndrome and 1 of multisystem organ failure after an aspiration pneumonia). Graft ischemia occurred in 3 of 46 left colon conduits and in 2 of 6 right colon conduits but did not result in death. In 3 patients, loss of the entire graft occurred. These ischemic events were ascribed to arterial anomalies in 2 patients operated on before the routine use of mesenteric angiography and to venous obstruction in the third patient from torsion of the vascular pedicle. In the remaining instances, ischemia was limited to the proximal end of the conduit. In 1 patient in whom graft ischemia was identified intraoperatively, microvascular anastomosis to a regional blood supply was used to salvage the graft. In the final patient, in whom the localized ischemic region was identified postoperatively, successful conservative management resulted in a proximal anastomotic stricture with salvage of the remainder of the graft.
Pneumonia was the most common acute complication, occurring in 24 patients, and was ascribed in part to abnormal glottic function in 16 patients. The median duration of postoperative intubation was 5 days (range, 2 to 33 days). Acute anastomotic leak was seen in 3 patients. Two leaks at the proximal anastomotic site were attributed to limited graft ischemia and were managed conservatively. A distal cologastric leak was ascribed to persistent carcinoma in the gastric wall at the site of the anastomosis and required reoperation. Other acute complications included graft ischemia in 5 patients (in 3 of 46 patients with a left colon conduit and in 2 of 6 patients with a right colon conduit); acute nonvascular perforation in 1, and sternal necrosis in 1.
Late complications occurring among the 50 operative survivors included proximal anastomotic stenosis in 24 patients, 12 of whom had caustic injuries. All 24 patients were initially managed by dilation. Two patients, both with caustic injuries, required surgical revision of the proximal anastomosis. The procedure was successful in 1 patient. Distal anastomotic stenosis was seen in 1 patient. Although routine late barium studies were not obtained, graft redundancy resulting in significant dysphagia was seen in 2 patients. Both patients were successfully palliated with segmental colonic resections and colocolostomy, preserving the marginal colic vascular arcade. Documented episodes of aspiration pneumonia occurring more than 6 months postoperatively were identified in 3 patients. All episodes were treated successfully, with no deaths. Small bowel obstruction was seen in 3 patients, all of whom required reoperation. In 1 patient, a presumed embolic event resulted in necrosis of a portion of the small intestine 2 weeks postoperatively, requiring a limited jejunal resection. In another patient, a limited volvulus about the site of a feeding jejunostomy required surgical reduction. In the third patient, who presented with acquired absence of the esophagus after a previous esophagectomy for carcinoma, peritoneal carcinomatosis developed 30 months postoperatively and was diagnosed by exploratory laparotomy. In 1 patient, intractable bile reflux developed postoperatively, requiring a distal biliary diversion, and in another patient a mucocele arising from the retained esophagus was treated by interval esophagectomy.
Late deaths occurred in 3 patients (suicide in 1 and recurrent metatstatic cancer in 2). Median survival was 11.5 years (range, 1 to 37 years). All patients were instructed in standard maneuvers to limit reflux and aspiration, including elevation of the head of the bed and avoidance of postprandial recumbency. Maintenance of a normal weight with no or rare episodes of dysphagia (less than once per year) was achieved in 11 of 50 patients. A diet modified for aspiration or moderate dysphagia, but allowing maintenance of weight without supplemental tube feeding, was necessary for 33 of 50 patients. However, in 5 patients continuous nutritional support by enteral tube feeding was required. Two of these patients had previous caustic injuries.
| Comment |
|---|
|
|
|---|
Long-segment colon interposition can be performed with a low risk and operative mortality, comparable to that for other types of esophageal substitution [8, 9]. Some surgeons do not require the performance of mesenteric angiography, but we continue to find it a useful adjunct in selecting the appropriate conduit for a particular patient [4, 12]. Although arterial anomalies are more common in the vascular supply to the ascending colon, we identified an abnormal vascular supply in 38% of our patients, despite a preponderance of left colon segments. Even with routine angiography, there remains a risk of graft ischemia, as noted in 9% of our patients. No patient, however, died as a result of the ischemic complication. Arterial ischemia can be corrected intraoperatively in some instances by microvascular anastomosis of arterial branches from the proximal conduit to a regional blood supply. In our series, this maneuver was more frequently required for right colon transpositions. Venous infarction may also occur and most likely accounts for the remaining instances of graft ischemia occurring in the later postoperative period [13]. Careful transposition of the conduit and avoidance of partial torsion of the vascular pedicle reduces the likelihood of this complication.
Isoperistaltic orientation and precision in estimating the required length of the graft are vitally important for a successful outcome. The primary advantage of the colon graft is its length, and an adequate segment to span the gap of gastrointestinal continuity is required. A sufficient length of intraabdominal colon is also required to minimize postoperative reflux symptoms. Excess redundancy of the colon graft, however, is to be avoided because it may compromise conduit perfusion acutely and may result in late postoperative dysphagia [14]. Reoperation and segmental resection of the colon, preserving the marginal vascular arcade, provide a satisfactory return of conduit function in such cases.
Caustic injuries to the esophagus remain more difficult to treat than other acquired esophageal disease because of the length of involvement and the severity of the injury. In our series, 60% of patients with caustic injury required pharyngeal anastomosis. Surgical revision of the proximal anastomosis was ultimately required in 10% of these patients. Poor swallowing function was seen in 20% of patients undergoing colon interposition for caustic injury. By comparison, for the other cases of acquired esophageal disease requiring colon interposition, pharyngeal anastomosis was necessary in only 11% of cases. No patient required surgical anastomotic revision, and only 10% of these patients had poor long-term functional outcomes. Similar results have been reported in other series of both pediatric and adult patients [15, 16]. Poor swallowing function after colon interposition appears to be due less to the conduit itself and is related more to the type of etiologic injury [15,16]. Although surgical revision of the proximal anastomosis for patients with caustic injuries has met with good results, an intriguing alternative approach is the initial use of a radial forearm free flap to create a staged composite conduit [17, 18]. Careful consideration should be given to this approach in a patient with extensive proximal pharyngoesophageal injury from caustic ingestion.
On the basis of our experience, the risks and long-term function of long-segment colon conduits are similar to those reported for other visceral substitutes for acquired esophageal disease [8, 9, 19, 20]. However, in the context of subtotal esophageal replacement when other visceral substitutes are available, colon conduits do not offer any inherent advantage to outweigh the operative time, complexity, and technical demands of their preparation. Colon grafts should not be used as a primary mode of reconstruction in such patients. Long-segment colon interposition is ideally applied in patients with acquired esophageal disease in whom other visceral conduits are unavailable or in whom a disproportionate length of conduit is required. A surgeon familiar with the technical demands of this procedure is a requisite component of this approach.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
P. A. Thomas, A. Gilardoni, D. Trousse, X. B. D'Journo, J.-P. Avaro, C. Doddoli, R. Giudicelli, and P. Fuentes Colon interposition for oesophageal replacement MMCTS, June 3, 2009; 2009(0603): 2956. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. E. de Delva, C. R. Morse, W. G. Austen Jr., H. A. Gaissert, M. Lanuti, J. C. Wain, C. D. Wright, and D. J. Mathisen Surgical management of failed colon interposition. Eur. J. Cardiothorac. Surg., August 1, 2008; 34(2): 432 - 437. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. X. Li, M. S. Levine, S. E. Rubesin, and I. Laufer Nonanastomotic Strictures After Colonic Interposition Am. J. Roentgenol., July 1, 2007; 189(1): 30 - 34. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Motoyama, M. Kitamura, R. Saito, K. Maruyama, Y. Sato, K. Hayashi, H. Saito, Y. Minamiya, and J.-i. Ogawa Surgical Outcome of Colon Interposition by the Posterior Mediastinal Route for Thoracic Esophageal Cancer Ann. Thorac. Surg., April 1, 2007; 83(4): 1273 - 1278. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. S. Kent, L. Gayle, L. Hoffman, and N. K. Altorki A New Technique of Subcutaneous Colon Interposition Ann. Thorac. Surg., December 1, 2005; 80(6): 2384 - 2386. [Abstract] [Full Text] [PDF] |
||||
![]() |
J.-H. Zhou, Y.-G. Jiang, R.-W. Wang, Y.-D. Lin, T.-Q. Gong, Y.-P. Zhao, Z. Ma, and Q.-Y. Tan Management of corrosive esophageal burns in 149 cases J. Thorac. Cardiovasc. Surg., August 1, 2005; 130(2): 449 - 455. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |