Colon Replacement of the Esophagus for Congenital and Benign Disease

Transcription

1 ORIGINAL ARTICLES Colon Replacement of the Esophagus for Congenital and Benign Disease William E. Neville, M.D., and Ahmad Z. Najem, M.D. ABSTRACT Over the past 28 years, one of us (W. E. N.) has reconstructed the esophagus with the right colon for congenital and benign disease in 84 patients. The first patient in the series, who was operated on in 1955, remains asymptomatic. Nine patients had congenital tracheoesophageal fistula with atresia; 4, esophageal varices; 30, advanced obliterative esophagitis; and 23, corrosive destruction. In 7, severe esophagitis followed esophagogastrectomy; 4 had unsuccessful operations for achalasia; and 7 had colon bypass following esophageal perforation. Eleven early nonfatal complications occurred. Late nonfatal complications were seen in 6 patients. There were 4 early deaths (4%): following dehiscence of an intrathoracic esophagocolic anastomosis and 1 due to peritonitis. Four individuals died over the years, and 5 patients were lost to follow-up. The late results in 7l patients show that 60 (84.5%) believe they have a satisfactory result. Nine (13%) individuals are symptomatic, and 2 (2.8%) must be classified as failures. Early complications have been minimized by using preoperative intestinal angiography, anastomotic stapling techniques, and the Doppler study intraoperatively to prognosticate colon blood flow. Several important observations have been made: anastomosis in the neck is preferable; the transplanted colon dilates from loss of motor activity but is functionally adequate; an isoperistaltic segment is preferable, but an antiperistaltic implant suffices; colonic mucosa is relatively resistant to acid-peptic digestion; and hyperalimentation is mandatory in very ill and debilitated patients. The past three decades have seen the emergence of the colon as an ideal substitute for the esophagus in a variety of complex congenital and benign conditions. During these years the From the UMDNJ-New Jersey Medical School, Newark, NJ. Presented at the Nineteenth Annual Meeting of The Society of Thoracic Surgeons, San Francisco, CA, Jan 17-19, Address reprint requests to Dr. Neville, UMDNJ-New Jersey Medical School, 100 Bergen St, Newark, NJ operation has been thoroughly evaluated by competent surgeons, who have attested to its efficacy Although the technique has not changed greatly, the indications are now more precise and information on the long-term results is becoming available. The purpose of this paper is to review the 28-year experience of one of us (W. E. N.) with 84 patients who had all or part of a nonneoplastic esophagus replaced with either the right or transverse colon. Complications were seen immediately and even late, but increasing experience, technical modifications, and advancing therapeutic techniques have made it possible to diminish the morbidity markedly and to achieve a reasonable mortality. Material and Method During the past 28 years, the esophagus was reconstructed in 84 patients with congenital and so-called benign disease. There were 62 male and 22 female patients ranging in age from 6 months to 67 years. Among the 30 children, 9 had tracheoesophageal fistula with atresia, 4 had esophageal varices, and 3 had iatrogenic perforation. Unyielding esophageal strictures developed in 14 following ingestion of caustics. Among the 54 adults, 30 individuals had obliterative esophagitis, primarily from reflux in 19 and from failure of operation for reflux in 11. Strictures developed in 9 patients after they swallowed caustics, in 7 following esophagogastrectomy for esophagitis, and in 4 after operation for achalasia. Four patients had a substitution after esophageal perforation. Follow-up information was obtained for 79 of the 84 patients, including the 4 who died late. The patient or the immediate family was contacted directly by telephone or by certified mail or was seen by us. Five patients with an initial good result for 2 to 8 years were lost to follow-up. Operative Technique Preoperative preparation consists of a barium enema, colonoscopy if necessary, and vascular 626

2 627 Neville and Najem: Colon Replacement of the Esophagus intestinal angiography in the elderly patients. Nutritional and chemical imbalances are restored with hyperalimentation, and the colon is prepared for two to three days with neomycin, Fleet Phospho-Soda (sodium biphosphate and sodium phosphate), and cleansing enemas. Our technique is essentially the same as that described in 1958 by Neville and Clowes [3]. The operation is performed by one or two surgical teams depending on the nature and extent of the disease. A prior cervical esophagostomy is dissected from the surrounding tissue; otherwise the esophagus is exposed in the neck through an incision along the anterior border of the sternocleidomastoid muscle. The entire right and transverse colon is mobilized from the cecum to the splenic flexure. The ileocolic, right, and middle colic vessels are identified and divided near their origin. Selectively, the terminal ileum is transected to cm from the cecum or at the ileocecal valve. Additional length of the segment is obtained by division of the mesentery to the origin of the superior mesenteric artery and by freeing the second, third, and fourth portions of the duodenum and pancreas [5]. The bowel is transplanted into the neck subcutaneously or through an anterior mediastinal tunnel. The thoracic inlet is enlarged when necessary by removing a portion of the manubrium, or by excising the head of the left clavicle, or both. The Doppler study is used to verify the adequacy of the circulation to the proximal and distal ends [61. In recent years the GIA and TA-55 staplers have been used for the cervical anastomosis whenever possible [7]. The distal end of the colon is stapled to the stomach anteriorly and a pyloromyotomy or preferably a pyloroplasty is performed. A tube gastrostomy is placed for decompression and alimentation, and the ileum is stapled to the transverse colon. The distal transverse colon and splenic flexure are used to replace the lower esophagus. Preferably the bowel is placed in an isoperistaltic position. The opening in the cardia is closed with the TA-55 stapler, and the distal end of the colon is stapled to the stomach anteriorly. A colocolostomy, pyloroplasty, and tube gastrostomy complete the procedure. Results Complications and Morfality There were 11 early major nonfatal complications. Nine involved the cephalad portion of the transplant. Two adults had partial necrosis of the colon at the thoracic inlet. In each, a median sternotomy was performed and the involved area was resected. The colon was lengthened, exteriorized in the neck, and sutured to the skin adjacent to the upper end of the esophagus. Later, the esophagus and colon were reunited. Intermittent dilations of the anastomosis were necessary for 2 to 3 years. Neither patient has required dilations for the past 10 years, and each is asymptomatic. Esophagocolic anastomotic leaks in the neck were observed in 7 patients. The fistulas closed within a few weeks, but a subsequent narrowing necessitated dilation and eventual revision in all of the patients. None has had severe anastomotic obstruction during follow-up of 10 to 28 years. Another patient had extensive destruction of the pharynx and larynx following an attempted suicide by drinking Drano. Seven months after operation, a stricture developed at the anastomosis between the mouth and ileocolic segment. Prior to the esophageal reconstruction the patient had a complicated hospital course that included a tracheostomy, gastrostomy, pharyngostomy, and open drainage of a right empyema from an iatrogenic perforation. A vagotomy, pyloroplasty, and plication of a bleeding duodenal ulcer were necessary at one point, following which a subdiaphragmatic abscess was drained. After a prolonged period of hyperalimentation, the ileocolic segment was transplanted subcutaneously to the mouth 24 months after the initial hospital admission. Five years after the anastomotic revision, he is swallowing normally, having undergone in the interim a total laryngectomy to prevent aspiration. One cologastric obstruction occurred in a 61- year-old woman in whom an antiperistaltic short segment of the transverse colon had been interposed following an esophagogastrectomy. She complained of intermittent dysphagia within a few months of operation, but the anastomotic narrowing was not substantiated until a year after operation. The stomach and adjacent

3 628 The Annals of Thoracic Surgery Vol 36 No 6 December 1983 colon appeared normal when the anastomosis was revised, but the microscopic diagnosis was gastritis and colitis. She has been asymptomatic for 22 years. Late nonfatal complications were observed in 6 individuals. Two patients had a bleeding gastric ulcer 3 and 5 years following operation. They both responded to medical treatment and have remained asymptomatic over the past 12 to 14 years. One adult and a child, 4 and 6 years, respectively, after esophageal reconstruction for corrosive stricture, underwent laparotomy for partial small bowel obstruction due to the adhesions. They have been asymptomatic for 12 and 15 years. In 1 asymptomatic child with a short-segment transplant, whose history has been reported previously 1111, acute dysphagia developed 5 years after operation while he was eating an apple. Esophagoscopy showed large pieces of apple impinged at the esophagocolic anastomosis. They were removed, and the patient had complete relief of symptoms. The anastomosis was widely patent, and there was no evidence of esophagitis or colitis. Just prior to this episode, cinefluorographic studies had shown unimpeded flow of barium through the esophagocolic segment into the stomach. The patient has been asymptomatic for the past 22 years. One man had a cologastric obstruction and anastomotic ulcer 28 years after esophageal replacement for a tracheoesophageal fistula with atresia. The intraabdominal segment of colon was huge and filled with a giant bezoar. This portion of the colon was resected, and the upper end was exteriorized in the right upper quadrant. The previous cologastrostomy was closed and a gastrostomy performed. Following a prolonged period of hyperalimentation, the mediastinal segment of the colon was reanastomosed to the stomach. The patient is asymptomatic 18 months later. There were 4 early deaths. Three were due to dehiscence of an intrathoracic supraaortic esophagocolic anastomosis. Another patient died of peritonitis as a result of a cologastric anastomotic leak. Four patients died months or years after operation. One elderly man died 18 months postoperatively of aspiration due to a cologastric obstruction. Another died of pulmonary emphysema 20 years after the operation. A third patient was reported to have had a carcinoma of the esophagus at postmortem examination 2 years after a colon bypass for a massively dilated obstructive esophagus. His primary pathological condition over a 20-year period had been esophageal achalasia for which he had had several pneumatic dilations and two modified Heller procedures. The fourth patient died of congestive heart failure 25 years after his lower esophagus and cardia had been replaced with an antiperistaltic colon segment. Following a few months of symptomatic reflux after operation, he had minimal digestive disturbances over the years. Postmortem examination showed no evidence of esophagitis, colitis, or gastritis. Five patients were lost to follow-up. Long-term Results At present, 60 patients consider themselves to have a satisfactory result regardless of the primary pathological condition, the length of the transplant, or whether it is isoperistaltic or antiperistaltic. They can eat a regular diet, according to their standards, without discomfort. They do not experience nocturnal regurgitation, indigestion, or heartburn and believe they are in good health, in general. However, they were not always asymptomatic. Previously, it was pointed out that babies with tracheoesophageal fistula, atresia, and long-segment replacement had specific symptoms relating to the abnormal position of the colon [8]. Nocturnal regurgitation, difficulty in swallowing, and diarrhea were observed. Growth patterns, however, were not altered. While the symptoms lasted, it was imperative to offer food in small pieces, to have the patients erect for at least an hour after meals, and not to permit them to drink before sleeping. These symptoms gradually abated within 12 to 24 months, and over the ensuing years the patients became relatively asymptomatic. In contrast, the older children and young adults who had obstruction tolerated the transplants exceeding well immediately after operation, regardless of the length, the route, or the underlying pathological involvement. Their early postoperative

4 629 Neville and Najem: Colon Replacement of the Esophagus complications were minor, and unlike the babies, they did not have annoying digestive disturbances. Four children with long-segment interpositions had angulation at the proximal end of the transplant. Although no overt obstruction was evident, they did complain of dysphagia for a time. These symptoms became less prominent as the children grew and the colon segment became longer. In 3 children, intermittent symptoms reminiscent of a dumping syndrome responded to moderate dietary restriction. Four asymptomatic children stated they had occasional loud noises in the chest that were embarrassing. None complained of belching, and foul breath was not noted by their parents. Eight adults with an antiperistaltic short segment complained for varying periods of difficulty in swallowing and of acid regurgitation in the prone position. Cinefluorography in the early postoperative period revealed evidence of reverse peristalsis, which gradually subsided as the colon lost its intrinsic motor activity. This coincided with the disappearance of symptoms in all but 2 patients, who still had some nonincapacitating digestive disturbances 10 and 14 years after operation. Less than satisfactory long-term results were evident in 9 patients, including the 2 individuals with antiperistaltic short segments. The remaining 7 patients had long-segment right colon interpositions. There is no common denominator separating these patients from the asymptomatic group except that they are older. All complained of intermittent nocturnal regurgitation, gaseous eructations and a vague, transient full feeling immediately after meals. There is no radiographic evidence of obstruction. The two failures were in patients with complicated problems antedating the colon interposition. One individual was first seen during an episode of severe hematemesis from an obstructing esophagogastric anastomotic ulcer. It was necessary to transect the esophagus above the ulcer and the fundus of the intrathoracic stomach distal to the ulcer. The upper end of the esophagus was exteriorized in the neck and a feeding jejunostomy performed. Subsequently, an ileocolic segment was interposed between the cervical esophagus and the second portion of the duodenum, since the pylorus was adjacent to the diaphragmatic hiatus. Although swallowing is not impaired, the patient complains of severe dumping syndrome and ifitermittent diarrhea, which is only temporarily relieved by diet regulation and medication. He has difficulty maintaining his weight, and several hospital admissions have been required over the years to correct recurring metabolic imbalances. The second patient initially had a high esophageal obstruction from swallowing alkali in an attempted suicide. He was transferred to our hospital from a psychiatric institution. He had aspiration pneumonitis and was malnourished. Intravenous hyperalimentation was initiated and a cervical esophagostomy done because of his inability to cope with excessive oral secretions. In due course, an ileocolonic bypass was performed to a markedly fibrotic stomach. The patient has been unable to eat solids in the 3 years since operation, but he subsists fairly well on liquids and pureed food. There is some narrowing at the upper anastomosis, which has required intermittent dilations. However, the main problem appears to be an atonic, contracted stomach that has a limited capacity and empties slowly over several hours. Comment Proximal cervical esophagocolic anastomotic leaks in long-segment replacements through the anterior mediastinum have universally been the paramount postoperative complication [ The resultant fistula may be vexing, but it is limited and does have an opportunity to close spontaneously, in contrast to one occurring in the free pleural space where it is usually fatal [12, 131. In addition, when a stricture occurs at the anastomotic site in the neck, dilation can be performed easily. When this is not permanently effective, revision of the stricture is not difficult and one can be reasonably sure that it will not recur. None of our patients required subsequent dilation or revision once the stricture was remedied. The etiology of cervical anastomotic leaks is conjectural. They appear to result from venous congestion of the bowel at the narrow thoracic inlet. This produces localized edema and necro-

5 630 The Annals of Thoracic Surgery Vol 36 No 6 December 1983 sis of the suture line. In our experience, the problem has been obviated by removing part of the manubrium and, if necessary, the head of the left clavicle and by using the ileum for the cervical anastomosis [14]. This method permits the bulky cecum to remain in the more spacious anterior mediastinum. Interval barium studies have shown redundancy and dilatation in about 20% of the anterior mediastinal transplants in growing children. Although these patients were well nourished and asymptomatic, prolonged emptying over two to three hours could be observed radiographically. Our experience suggests that progressive diminution of intrinsic motor activity of the colon and obstruction at the diaphragmatic level are mainly responsible. Any initial mediastinal redundancy of the colon should be corrected by replacing the segment in the abdomen and affixing it to the diaphragmatic hiatus. In addition, the retrosternal tunnel should be wide enough for the transplant to egress without the risk of being compressed or twisted. Despite this observation by us and others who have followed children over several years, only 1 of our patients had serious complications from enlargement of the colon and they occurred 28 years after replacement. Both the supradiaphragmatic and infradiaphragmatic colon segments were markedly dilated due to obstruction at the diaphragm and the stomach. The enlarged anterior mediastinal segment had shifted into the right chest and was atonic. Following decompression, this portion of the colon gradually assumed a normal configuration, which signifies that the translocated bowel is compliant regardless of the duration of its atonicity. A recent barium swallow showed that it was normal in caliber with a direct route in the anterior mediastinum to the stomach. Our experience with subcutaneous transplants is limited to 8 adults who have been observed for only 9 years, but there is noticeable enlargement of the transplant anteriorly. However, it is not angulated and is contained by the skin which prevents lateral displacement. In addition, the patients are afforded an opportunity to milk the bolus of food down to the stomach. Even though this type of reconstruction is not cosmetically appealing, it does offer some advantages. The propensity for dilatation and elongation of the bowel is negligible when a short segment of the transverse colon is substituted for the esophagus. This may be related to the recognizable contractions in the long proximal esophagus, the straight, nonangulated position of the transplant, and replacement of the interposed bowel into the bed of the resected esophagus. Although we favor using an isoperistaltic segment of bowel, it is apparent in lower esophageal replacements that antiperistaltic segments are eventually just as efficient. Cinefluorographic studies in the early postoperative period show some regurgitation and impedance to barium in an antiperistaltic segment. These patients complain of indigestion for varying periods, have occasional acid reflux, and have a sense of retardation of food passing into the stomach. In general, these symptoms gradually subside and usually can be correlated with the disappearance of active peristalsis. The spatial relationship between the stomach and the proximal esophagus by the interposed colon protects the susceptible esophageal mucosa from erosion. Of continuing concern has been reflux of gastric juice into the colon. Barium can be observed to reflux actively into the colon segment in the Trendelenburg position, but no inflammatory changes have been observed by esophagoscopy or at postmortem examination. However, the occurrence of a perforated anastomotic ulcer 28 years after colon substitution casts doubt on the perpetual protection of the colonic mucosa from erosion. Malcolm [15] found 11 cases of anastomotic ulcers in the literature. The common denominators in these patients were related to gastric outflow obstruction, incomplete vagotomy, and the presence of an anterior cologastrostomy. Considering the sizable number of patients who have had colon replacement of the esophagus over the years, this complication is rare but may be more prevalent now than was realized formerly. The absence of colitis in the majority of the transplants may be due to the recent observations of Clark and his associates [16]. They found prolonged periods of alkaline recordings during 24-hour ph monitoring of the infradiaphragmatic colon and postulated that the colon secretions of mucus, known to contain bicar-

6 631 Neville and Najem: Colon Replacement of the Esophagus bonate, could be a protection against gastric reflux. Our radiographic observations over the years as well as those of other groups conclusively demonstrate that the interposed colon is nonperistaltic within a few months and essentially becomes a conduit. Recently, however, this concept has been challenged by sophisticated manometric studies. It has been confirmed that there is a mechanical response of the transplanted colon to intraluminal administration of 0.1N HCl solution. Jones and colleagues [17] observed that type I1 colonic contractions in the transplant were similar to those observed when the colon was in normal position. However, they were unable to confirm whether these peristaltic waves were effective in propelling food from the proximal esophagus to the stomach. Hyperalimentation for malnourished and desperately ill patients has had a major impact. Since these patients are operated on for esophageal obstruction and are malnourished, they undergo operation in a better nutritional state. Except in babies with tracheoesophageal fistula and atresia and in patients with a perforated lower esophagus, the need for cervical esophagoscopy and feeding gastrostomy has virtually disappeared. The proximal esophagus can be decompressed in many instances with a polyethylene tube and nutrition maintained intravenously without violating the integrity of the neck and abdomen. The addition of the stapler to our surgical armamentarium has been a distinct asset. The surgical and anesthetic time has been shortened considerably, since the several visceral anastomoses can be more accurately and expeditiously performed with the stapler than with manual suture techniques. The wide, nonedematous esophagogastric and ileocolic anastomosis afforded by the GIA and TA-55 staplers has lessened the period of postoperative paralytic ileus, and suture line leakage or dehiscence has not been seen. The use of these instruments for the esophagocolic anastomosis in the neck is fast and immediately provides a smooth, nonconstricted lumen. There have been no anastomotic leaks or strictures, and subsequent narrowing and angulation of the cervical esophagoileocolic anastomotic area have not oc- curred in patients observed over several years. Despite these observations, it should be stressed that gastrointestinal stapling techniques cannot be applied safely without a period of learning in the surgical research laboratory or regular training with a knowledgeable surgeon. Preoperative intestinal angiography appears to be beneficial in the older patient. It reveals unsuspected arteriosclerotic areas in the abdominal aorta that may have a detrimental effect on the arterial blood supply to the colon transplant. With this knowledge, the surgeon can select an alternative section of the viscera for esophageal reconstruction. In addition, the variant arterial supply to the colon may make one segment preferable to another for replacement [4]. Prediction of intestinal viability by observing peristalsis or arterial pulsation to the colon is frequently unreliable. However, with the Doppler ultrasound device it is possible to relate viability to blood flow directly. The potential of this method was pointed out by Wright and Hobson [MI, who used it to assess colon blood flow during abdominal aortic reconstruction. Subsequently, Kurstin and associates [6] reported on its efficacy in esophageal bypass procedures. We have been impressed with the use of the Doppler device to assess venous as well as arterial flow in the small vessels of the colon transplants and thereby detect compression or occlusion of venous return, which is a major factor in cervical anastomotic leaks. It is a rapid, convenient, and accurate way to detect viability of the proximal anastomosis and has contributed substantially to the diminution of cervical leaks in our patients. The indications for colon replacement of the esophagus have gradually expanded over the years. What was initiated as a technique that would afford a reliable, well-vascularized visceral segment to supplant the esophagus after resection for cancer has become more applicable to patients with congenital and benign disease. In babies and children, only a segment of transverse colon is necessary for total esophageal reconstruction because of its length and thin, pliable mesentery. The bowel grows with the child and is durable, and the long-term functional results have been good in a sizable series of patients observed by different groups over many

7 632 The Annals of Thoracic Surgery Vol 36 No 6 December 1983 years [8, 19, 201. We have used the right colon exclusively, but it is evident that the left colon serves equally well [I, 10, 201. Although major advances have been made in the management of complications stemming from gastroesophageal reflux, intractable esophageal stenosis can be remedied only by resection and reestablishment of esophagogastric continuity [21,22]. In general, esophagogastrectomy has been abandoned because of the high incidence of recurrent stenosis. Skinner [23] reported excellent functional results in 25 of 27 patients with short-segment left colon replacements for benign stenosis in whom there were no anastomotic leaks or deaths. Orringer and co-workers [24] and Glasgow and associates [25] similarly found good results with colon interposition in patients following failure of esophagomyotomy. Our experience and functional results are similar to those of these two groups. All patients obtained relief of symptoms and have maintained their weight over the years. However, cervical anastomotic leaks developed in 4 individuals when the strictured esophagus was bypassed, and 2 patients with long-segment transplants died of dehiscence of a high intrathoracic anastomosis. The latter was utilized in the formative years of our experience and has since been abandoned. All high anastomoses now are done in the cervical area. Controversy still exists regarding the necessity to remove the bypassed stenotic esophagus. It was suggested by Joske and Benedict [26] that in a patient with a lye stricture, the likelihood of cancer developing is 22 times greater than in the average person, while Kiviranta [27] believed that the incidence of cancer was actually 1,000 times more. However, the interval between the ingestion of lye and the development of cancer was 25 to 40 years. On the other hand, Marchand [28] could not find a single case of cancer among 135 patients with caustic stricture of the esophagus. A more pertinent reason for removing the esophagus is the risk of bleeding from the distal end due to gastric reflux. Although this has not been seen in our patients, it might be prudent to occlude the intraabdominal esophagus at the time of the colon substitution. Even though we do not believe that removal of the esophagus is mandatory, 7 children in this series underwent a subsequent uncomplicated esophagectomy. There were no compelling reasons but since the risk of cancer is unproven, removal of the esophagus should be considered in this age group because of their life expectancy. Most of the early complications and the immediate postoperative deaths in these groups of patients occurred early during the development of a relatively new surgical operation. In the past decade, these incidences have been minimized through increasing experience, improved technical planning, aggressive perioperative hyperalimentation, careful patient selection, monitoring of intraoperative audible signs of vascular competence of the transplant, and lessening of anastomotic edema with stapling techniques. It is concluded from our long-term results that the colon can effectively replace the esophagus in a wide variety of pathological entities. This cannot be accomplished with other viable segments of the gastrointestinal tract. References 1. Belsey R: The long term clinical state after resection with colon replacement. In Smith RA, Smith RE (eds): Surgery of the Esophagus. New York, Appleton-Century-Crofts, pp Gross RE, Firestone FN: Colonic reconstruction of the esophagus in infants and children. Surgery 61:955, Neville WE, Clowes GHA Jr: Reconstruction of the esophagus with segments of the colon. J Thorac Surg 35:2, Wilkins SW: Long segment colon substitute for the esophagus. Ann Surg 192:722, Leininger B, La1 R, Neville WE: Technique for increasing the length of colon bypass segment. Surgery 62:88, Kurstin RD, Soltanzadeh H, Hobson RW 11, Wright CG: Ultrasonic blood flow assessment in colon esophageal bypass procedures. Arch Surg 112:270, Ravitch MM, Steichen RM: Technique of staple suturing in the gastrointestinal tract. Ann Surg 175:815, Neville WE, Clowes GHA Jr: Reconstruction of the esophagus in children for congenital and acquired disease. J Thorac Cardiovasc Surg 40:507, Battersby JS, Moore TC: Esophageal replacement and bypass with ascending and right half of transverse colon for treatment of congenital atre-

8 633 Neville and Najem: Colon Replacement of the Esophagus sia of the esophagus. Surg Gynecol Obstet 109:207, German JC, Waterston DJ: Colon interposition for the replacement of the esophagus in children. J Pediatr Surg 11:227, Neville WE, Smith AE, Storer J: Use of the transverse colon for reconstruction of the esophagus in tracheo-esophageal fistula. Ann Surg 144:1045, Buntain WL, Payne WS, Lynn HB: Esophageal reconstruction for benign disease: a long term appraisal. Am Surg 46:67, Dale AW, Sherman CD: Late reconstruction of congenital esophageal atresia by intrathoracic colon transplantation. J Thorac Surg 29:344, Neville WE: Esophageal carcinoma. NY State J Med 68:649, Malcolm JA: Occurrence of peptic ulcer in colon used for esophageal replacement. J Thorac Cardiovasc Surg 55:763, Clark J, Moraldi F, Moosa R, et al: Functional evaluation of the interposed colon as an esophageal substitute. Ann Surg 183:93, Jones EL, Skinner DB, DeMeester TR, et al: Response of the interposed human colonic segment to an acid challenge. Ann Surg 17775, Wright DA, Hobson RW 11: Prediction of intestinal viability using Doppler ultrasound technique. Am J Surg 129:642, Postlethwait RW: Surgery of the esophagus. New York, Appleton-Century-Crofts, Rodgers BM, Talbert JL, Moazam R, Felman AH: Functional and metabolic evaluation of colon replacement of the esophagus in children. J Pediatr Surg 13:35, Neville WE, Clowes GHA Jr: Surgical treatment of reflux esophagitis. Arch Surg 83:534, Neville WE, Clowes GHA Jr: The surgical treatment of complications resulting from cardioesophageal incompetence. Dis Chest 43:575, Skinner DB: Benign esophageal strictures. Adv Surg 10:177, Orringer MB, Kirsh MM, Sloan H: New trends in esophageal replacement for benign disease. Ann Thorac Surg 23:409, Glasgow JC, Cannon JP, Elkins RC: Colon interposition for benign esophageal disease. Am J Surg , Joske RA, Benedict EB: The role of benign esophageal obstruction in the development of carcinoma of the esophagus. Gastroenterology 36:749, Kiviranta UK: Corrosion carcinoma esophagus. Acta Otolaryngol (Stockh) 42:89, Marchand P: Caustic stricture of the esophagus. Thorax 10:171, 1956 [Discussion of this article appears on pages of this issue.]