Caustic Ingestion and Subsequent Damage to the Oropharyngeal and Digestive Passages

Transcription

1 CURRENT REVIEW Caustic Ingestion and Subsequent Damage to the Oropharyngeal and Digestive Passages Marvin M. Kirsh, M.D., and Frank Ritter, M.D. ABSTRACT The characteristics, diagnosis, and management of oropharyngeal and digestive passage lesions due to ingestion of caustic agents are presented. Previous experimental and clinical studies are reviewed with discussion of the peculiar qualities of the newer caustics. Differentiation in treatment of burns caused by solid and liquid agents is stressed. The management of corrosive ingestion has never been definitive or precise because of differences in the amount and concentration of the caustic ingested. Clinical experience and animal experimentation have provided a few specific guidelines for therapy, but even these must often be modified for various reasons, especially when changes occur in the type of corrosive ingested. For example, for the forty years preceding 1965 most of the caustics ingested in the United States were flakes or solid pellets of sodium hydroxide. In this form the caustic resulted in burns primarily of the oropharyngeal and upper digestive tract, with segmental stricture of the esophagus a common sequela. During these years medical efforts were expended toward more effective ways of neutralizing the caustic to prevent stricture formation. When strictures did develop, therapy consisted of repeated dilations, perhaps for the life of the patient. Toward the end of the era animal investigations revealed that if antibiotics and steroids are given early enough following ingestion, the likelihood of stricture is reduced. Thus, for some years therapy consisted of steroids and antibiotics given at the acute stage and chronic dilation for strictures that developed later [I, 6, 7, 9-11, 16, 17, 24, 25, 331. In 1967 lye in a liquid state was first marketed in the United States as a drain cleaner. This From the Department of Surgery, Section of Thoracic Surgery, and the Department of Otorhinolaryngology, The University of Michigan Medical Center, Ann Arbor, MI. Address reprint requests to Dr. Kirsh, Professor of Surgery, Section of Thoracic Surgery, The University of Michigan Medical Center, Ann Arbor, MI extremely concentrated substance produces severe tissue destruction so that the consequences are death or long-term morbidity, not isolated or segmental stricture formation. Animal studies were instituted in order to understand the pathophysiology of this ingested substance. These same studies also promoted an aggressive therapy of esophagogastrectomy. Since these initial efforts, the Federal Hazardous Substance Act has modified the concentration of these marketable corrosives so that the earlier recommended therapy has been modified. For this reason a more recent review of the present therapy of corrosive ingestion seems warranted. Ingestion of Solid Caustics Pa fhology Injuries of the esophagus due to caustic bums are classified in a manner similar to that used to describe thermal bums of the skin [lv. A firstdegree bum is characterized by superficial mucosal hyperemia, mucosal edema, and superficial sloughing. It represents minimal involvement. A second-degree bum implies deeper tissue damage, is transmucosal, and involves all layers of the esophagus, with exudate, ulceration, loss of mucosa, and erosion through the esophageal wall into the periesophageal tissues. A third-degree bum is one which, in addition to involving the esophageal wall, has eroded through the esophagus into the periesophageal tissues including the mediastinum and the pleural or peritoneal cavity. Alkaline caustic agents produce injury by liquefaction necrosis. Fats and proteins become saponified and blood vessels thrombose, leading to further cellular necrosis and degeneration. This characteristic of an alkali enhances its penetration into tissues and thwarts its neutralization by surface irrigation. Thus, fullthickness tissue bums commonly result. The extent and severity of tissue injury after caustic ingestion depends upon the corrosive nature of 74

2 75 Current Review: Kirsh and Ritter: Caustic Ingestion the ingested substance, its quantity, its concentration, and the duration of tissue contact [9]. Alkali swallowed in the solid state adheres on contact to the mucous membranes, especially those of the oropharynx and esophagus, causing ulceration, then deeper vascular thrombosis and liquefaction necrosis. Severe damage is more apt to occur in these areas, while the stomach is usually spared because the solids fail to reach it in sufficient quantity. If particles of alkali should enter the gastric lumen, they are diluted or neutralized by gastric acid. Should gastric damage occur, however, pyloric stricture with gastric outlet obstruction is the result [15, 26, 351. Physiology There is an initial physiological response to the ingestion of solid caustics. The burning produced in the tongue and lips by the caustic causes the patient to expectorate as much corrosive as possible. The same symptom often causes the victim to drink water, which tends to neutralize the burned tissues, dilute the remaining caustic, and wash it into the stomach, where neutralization usually takes place. Clinical Features The patient who ingests a solid caustic usually burns the lips, tip of the tongue, and fauces. If the corrosive is swallowed, the esophagus may also be burned. Initial drooling may cause the caustic to drain over the chin so that circumoral bums develop. These patients do not drool chronically and are usually able to manage their saliva. There is rarely deep burning, so perforation of the esophagus and resultant shock do not occur. Examination of the oropharynx within 24 hours of the burn usually demonstrates a white to yellow leukocytic membrane covering the ulcers produced by the caustic. The lesions bleed easily and are very painful [9]. Diagnosis While oral cavity bums can be assessed by visual examination, the question of whether the esophagus has been burned and to what degree can be answered only by esophagoscopy. This is done within the first 24 to 48 hours after the injury. Should a bum be noted, the endoscope must not be inserted past it in order to diminish the risk of esophageal perforation. A small nasogastric tube is passed to keep the lumen open and to permit alimentation. Ingestion of Liquid Caustics Pa tho Zogy In contrast to the solid form, ingested liquid corrosives expose all mucosal surfaces of the upper gastrointestinal tract to the caustic material. Using a solution of lye, Krey [20] demonstrated experimentally that the depth of penetration into tissue by the caustic agent is more severe. He found that a normal solution of sodium hydroxide (3.8%) in contact with the esophageal mucosa for 10 seconds produces necrosis of the mucosa, submucosa, and longitudinal muscle fibers. The area of necrosis penetrated into the outer longitudinal muscle when a 10.7O/0 (3 times normal) solution was used. When a solution of 22.5% sodium hydroxide (7 times normal) came in contact with the mucosa for the same length of time, necrosis of the entire esophageal wall with extension to the periesophageal tissues occurred. Similar experimental conclusions were reached by Leape and associates [21] using Liquid-plumr (30% sodium hydroxide) as the caustic a ent in cats. A 3-second exposure to 1 ml Ig of this concentrated caustic produced severe injury. Histological studies 30 minutes after the exposure confirmed marked swelling and necrosis of the epithelium, and after 24 hours the lesion was noted to ulcerate the esophageal mucosa, accompanied by an intense full-thickness inflammatory reaction. In some animals there was also extensive necrosis of the esophageal wall and widespread thrombosis of submucosal arterioles and venules. In a similar study in dogs, 15 ml of the same alkali caused fullthickness bums of the esophagus and stomach. The microscopical pattern is the same in human esophagi removed shortly after ingestion of Liquid-plumr [4]. In all studies, not only are second- and thirddegree burns apparent, but surrounding viscera such as the spleen, pancreas, aorta, and trachea are bumed and black in color. This indicates a high degree of penetration through the esophageal or stomach wall to adjacent organs. The

3 76 The Annals of Thoracic Surgery Vol 21 No 1 January 1976 pathological sequence of events (necrosis, ulceration-granulation, and scar formation) that occurs following liquid caustic ingestion is similar to that described following granular lye ingestion but with a greater frequency and severity of serious sequelae. Following granular lye ingestion only 10 to 25% of patients develop an esophageal stricture. In contrast, most patients with a history of liquid caustic ingestion are found to have a serious (near-circumferential) esophageal burn, and almost all progress to extensive stricture formation. Frequently the entire esophagus is damaged, with multiple stricture formations. Half or more of the length of the esophagus may be converted to a threadlike stricture [21]. Recently Ashcraft and Padula [3] demonstrated the hazardous nature of these newer caustics despite their diluted concentration. Using cats, they applied 8% potassium hydroxide (new Liquid-plumr) and 8.3% sodium hydroxide (Plunge diluted 1: 3) directly to the esophagus for 30 seconds. These agents were then washed off with copious quantities of water. After 2 hours the esophagus was removed and examined histologically. Extensive coagulation necrosis was noted under the microscope after this 30-second exposure. Microscopically the 8% potassium hydroxide was observed to produce complete liquefaction of the mucosa with edema formation and mild inflammation in the underlying muscularis mucosa and submucosal adventitial layers. Similarly, the 8.3% sodium hydroxide solution caused esophageal injury with destruction of superficial layers of squamous mucosa along with submucosal and even transmural thrombosis in blood vessels. These changes were not as severe as those that occurred with the application of 8% potassium hydroxide. Physiology The relationship of liquid corrosives to esophageal gastric injury has been amply demonstrated in animal studies [29, 301. Twenty seconds following the instillation of 15 ml of Liquid-plumr into an anesthetized dog s esophagus, violent regurgitation of the gastric contents occurred. As soon as the contents entered the pharynx, the cricopharyngeus immediately went into spasm. Esophageal peristalsis rapidly propelled the contents remaining in it back into the stomach. This was followed by reverse peristalsis, and a to-and-fro or seesaw action of gastric contents between esophagus and stomach developed that lasted between 3 and 5 minutes in each animal. After this the pylorus relaxed, antegrade peristalsis began, and the stomach contents entered the duodenum. At this point the gastric rugae were three times normal size, the esophagus was atonic and dilated, and the esophagogastric junction was reduced in size because of edema. The animals that were fed prior to the experiment had a toand-fro motion of shorter duration and the esophageal distention and atonicity were of lesser magnitude. At postmortem examination 24 hours later, all animals were noted to have fullthickness burns of the esophagus and stomach. The entire length of the esophagus and all of the stomach except the pylorus were involved. The mucosa of these tissues was gray or black in color and extremely edematous. Microscopical examination revealed that the esophageal mucosa contained ulcerated areas with coagulation necrosis, and the deeper tissues of the stomach showed a marked polymorphonuclear response. Vascular stasis and hemorrhage were seen in the gastric mucosa. Subserosal hemorrhage was noted opposite the necrotic central mucosal portion. In the animals fed prior to institution of the caustic, the microscopical changes from the burn were markedly less than in the fasting animals. Similar gross and microscopical findings in the stomach have been seen in people following ingestion of Liquid-plumr. The degree of damage to the stomach is dependent upon: (1) the amount and strength of caustic ingested, (2) whether or not food is present in the stomach, (3) duration of contact with the stomach mucosa, and (4) the presence or absence of pylorospasm. Clinical Features The clinical picture varies depending upon the concentration and amount of the liquid caustic agent ingested as well as the presence or absence of food in the stomach. Ingestion of the alkali in solid form usually causes severe burns of the mouth or pharynx or both. The solid form induces expectoration of the caustic so that little or

4 77 Current Review: Kirsh and Ritter: Caustic Ingestion none of it is swallowed. On the other hand the liquid alkali is often rapidly swallowed, causing less tissue injury within the mouth and pharynx but extensive damage to the esophagus, stomach, or both. Therefore caustic burns of the esophagus cannot necessarily be predicted from the presence and appearance or even the absence of external burns. If the patient vomits after ingesting the caustic, he will often drool continuously. This is probably secondary to atony and narrowing of the burned esophagus. Substernal pain, back and abdominal pain, and rigidity are usually indicative of mediastinitis, peritonitis, or both. Hoarseness, stridor, aphonia, and dyspnea suggest either associated laryngeal edema or actual epiglottic and laryngeal destruction. Other symptoms include dysphagia, odynophagia, aspiration pneumonia, recurrent emesis, and hematemesis. With burns of the esophagus, stomach, or both, the patient will be febrile and tachypneic and will have a rapid pulse. Hypotension secondary to hypovolemia may be an additional complicating factor. Examination of the lips, mouth, and oropharynx may reveal signs of tissue injury ranging from a few scattered areas of superficial mucosal erosions of the lips and tongue to deep and extensive destruction of the lingual, buccal, and pharyngeal mucosa. In the case of extensive injury there is usually a gray or gray black pseudomembrane covering the mucosal surfaces. Similar changes are found in the hypopharynx, with pooling of secretions and perhaps sufficient edema of the larynx to produce airway obstruction. Diagnosis Since it is not possible to correlate the findings in the mouth and pharynx with esophageal changes, the only means of establishing the extent and severity of esophageal injury is by esophagoscopy, preferably within the first 24 hours of injury. Although Hollinger [17] and Borja and co-workers [7] oppose such early esophagoscopy when caustic burns are suspected because of the risk of esophageal perforation, the majority of physicians treating these injuries favor early esophagoscopy. The risk of esophageal perforation is low if the procedure is performed under general anesthesia and the esophagoscope is passed to the first burned area but not beyond it. In three separate series with a combined total of 322 patients in whom esophagoscopy was performed soon after caustic ingestion was diagnosed, there were 3 instances of perforation [12]. One of the authors stated that none of these would have occurred if the esophagoscope had not been passed beyond the first burned area [ll]. In addition, early esophagoscopy permits those patients without esophageal bums to avoid prolonged hospitalization. Esophagoscopy is also beneficial in the continuing management of patients with caustic injury and may be of value in predicting which patients require esophageal substitution procedures. There are several limitations to esophagoscopy, however. (1) At times it may be difficult to evaluate the depth of any burn with absolute certainty by observing superficial epithelial necrosis. (2) When a severe burn is encountered in the upper third of the esophagus, the esophagoscope is not passed beyond this area and the involvement of the middle and lower thirds cannot be ascertained. (3) The area of burn may not be visualized, thus delaying the diagnosis. The areas of ulceration may be more easily visualized if a dilute solution of methylene blue is instilled into the esophagus at the time of esophagoscopy. The normal epithelium will be stained blue, whereas the areas of ulceration will not. Despite these limitations, the advantages of early esophagoscopy are great enough that it should be carried out in all patients with suspected caustic injury to the esophagus when the hypopharynx and larynx are free from burns. Hyperemia of the esophageal mucosa with superficial desquamation of epithelium occurs with a first-degree burn. Superficial blisters, ulcers, hyperemia, and patchy membranous exudate of the mucosa are characteristic of second-degree burns. Areas with deep loss of esophageal epithelium that are hyperemic and show evidence of granulation tissue are classified as third-degree burns. Usually the third-degree burn involves almost the entire circumference of the esophagus. Sometimes the extent and severity of involvement are difficult to ascertain.

5 78 The Annals of Thoracic Surgery Vol 21 No 1 January 1976 Radiological Features Contrast visualization of the esophagus and stomach is the best means of assessing the severity of the stomach injury [23]. The roentgenographic appearance of the burned esophagus and stomach following ingestion of a liquid caustic varies considerably and is dependent upon the severity of the injury and the time elapsed since it occurred. Plain roentgenograms of the chest and abdomen may disclose mediastinitis or peritonitis by demonstrating an extraalimentary soft tissue mass. As a result of the associated mediastinitis there is widening of the pleuroesophageal line and abnormal displacement of the pleural reflection. A radiocontrast esophagogram will provide additional evidence of liquid caustic injury. Water-soluble agents are preferable to barium for contrast examinations in view of the increased risk of perforation. The prominent radiocontrast features include: (1) diffusely blurred esophageal margins, reflecting mucosal ulceration, sloughing, and pseudomembrane formation; (2) linear streak and plaquelike collections of contrast material due to deep necrotic ulcers with intramural dissection; (3) scalloped or straightened esophagogastric margins, probably reflecting submucosal edema and hemorrhage; (4) intraluminal and intramural re- tention of contrast material for prolonged periods because of esophageal atony due to intramural dissection; and (5) persistent esophageal dilatation with intraluminal retention of gas secondary to diffuse muscular necrosis and possibly indicative of impending perforation. Stricture formation and lack of peristalsis are late manifestations due to fibrous replacement of the muscle. In the acute stage the ph of the gastric or emesis contents is also of value in determining the presence or absence of gastric damage. An alkaline ph of either one implies severe gastric damage, especially if it persists despite gastric lavage. When the stomach is entered at the time of esophagoscopy, its mucosa can be evaluated for the presence of burns. General Principles of Management The clinical picture varies widely, and each patient must be evaluated individually. The management of these patients must take into account both the desperate emergency phase in the be- ginning and, in some patients, the chronic course thereafter. All individuals with physical findings indicative of caustic ingestion should be admitted to the hospital; in fact, anyone in whom there is the least suspicion of caustic ingestion should be admitted and undergo evaluation. An attempt should be made to find the container that held the ingested caustic or some of the material itself for proper identification. In those patients who are hypotensive on admission, arterial and central venous catheters should be inserted for constant pressure monitoring. Hypovolemia should be treated with either whole blood or colloid-containing fluids. The patients should be carefully observed for airway obstruction. Laryngeal edema is frequently seen in the first 24 hours in association with severe caustic burns of the esophagus, and endotracheal intubation or tracheostomy may be needed. In view of the instantaneous nature of the injury produced by liquid caustics as well as analysis of our experimental studies, water or other juices to dilute the caustic are of little value. In fact, they may produce additional vomiting or retching that may aggravate the situation. This is especially true for patients who have ingested Liquid-plumr or other liquid caustics, because the stomach is already burned and adding neutralizing solutions only increases the chance of regurgitation. An additional risk of vomiting is, of course, aspiration of gastric contents. For the same reason, emetics and gastric lavage are contraindicated. The patient should be kept NPO and maintained on intravenous fluid. Clear fluids are begun by mouth only when the patient is able to swallow his saliva; he may then progress to a regular diet as tolerated. If the patient is unable to swallow, a gastrostomy should be performed. Because antibiotics have been shown to decrease both the risk of pulmonary infection and bacterial invasion through the injured esophagus into the mediastinum, they should be administered as soon as the diagnosis of esophageal injury is established [4, 10, 16, 31, 331. Our preference is for ampicillin (40 mg per kilogram of body weight per day in divided doses) because it is effective against oropharyngeal flora. It should be administered intravenously at

6 79 Current Review: Kirsh and Ritter: Caustic Ingestion first and then intramuscularly until oral feedings are begun. The antibiotics are discontinued at seven days if the patient is afebrile or continued until the fever disappears. If the patient remains febrile after a two-week course of antibiotic therapy, one should suspect an associated complication such as mediastinitis, peritonitis, or a localized abscess in either the mediastinum or peritoneum. Contrast visualization of the esophagus and stomach should be carried out shortly after admission. Esophagoscopy is done upon admission or not later than 24 hours after the ingestion. Those patients in whom there are no oropharyngeal or esophageal bums and in whom the barium swallow is normal can be discharged after 24 hours of observation. A repeat barium swallow should be obtained at one and two months so that an unrecognized esophageal stricture will not be overlooked. For many years treatment of caustic ingestion has included systemic corticosteroid therapy. The basis for steroid therapy was based on the observation by Spain and associates [32] that early administration of cortisone in mice inhibited fibroplasia and formation of granulation tissue. They also observed that, to be effective, glucocorticoids must be given within 48 hours of injury. Rosenberg and co-workers [31] demonstrated in dogs that while steroids inhibited stricture formation, the animals would die from infection if antibiotics were not used. Similar findings showing that steroids inhibit the formation of strictures after caustic ingestion in animals have been reported by Wieskuff [34], Haller and colleagues [16], and others [12]. These experimental studies have been supported by Haller s [16] report on 69 patients with esophageal bums after caustic ingestion. Stricture developed in 25% of the patients treated without steroids compared with only 9% of those treated with steroids plus antibiotics. Contrasting results have been reported by Middlekamp, Ashcraft, and Leape [l, 3, 6, 21, 251. In these series 18 of 19 patients with a severe (thirddegree) circumferential esophageal burn developed a stricture despite the early administration of steroids and antibiotics. In addition, an increased occurrence of perforation has been found when steroids were used following inges- tion of concentrated solutions of alkali. Steroids have also been found to mask symptoms of infection in other mediastinal structures and have delayed healing. As a result of these studies and our own personal experience, the efficacy of steroids in preventing stricture formation after caustic ingestion is highly questionable, especially with third-degree burns from liquid caustics. The previous studies, in addition to lacking uniform criteria for assessment of esophageal injury by esophagoscopy, showed some variability in treatment that might have affected the results. We agree with Ashcraft and Simm [4] that a well-controlled prospective clinical study is needed for the evaluation of steroids in human esophageal caustic injury. Until such a study is available, and because of the potentially serious complications that might result from their administration, it is our conviction that steroids should not be used in the treatment of esophageal injuries caused by liquid caustics. Solid-state Caustics As a rule, the ingestion of caustic crystals almost immediately produces severe pain that causes the patient to stop further ingestion. The crystals tend to adhere to the mucous membranes of the tissues with which they first come into contact, namely, the mouth, pharynx, and esophagus. Perforation of the esophagus can occur and should be constantly searched for, but is an infrequent occurrence following ingestion of caustic crystals. Emergency measures include early esophagoscopy and contrast roentgenographic visualization. The process of healing is observed by esophagoscopy and barium swallow repeated at 10- to 14-day intervals. The majority of patients who have ingested crystals do not suffer severe esophageal burns; however, patients who sustain severe (second- or third-degree) bums often develop an esophageal stricture. Those patients should swallow a string as soon as a liquid diet can be tolerated. Since early bougienage has been shown to be associated with an increased risk of perforation [24, 251, dilation is not started until healing is complete and an epithelialization mucosa is seen as far as the esophagoscope can be advanced, which may take four to five weeks [ll]. Dilation is then started and can be carried out with the

7 80 The Annals of Thoracic Surgery Vol 21 No 1 January 1976 mercury-filled, tapered, Hurst-Maloney bougie or the Tucker bougie, using the previously swallowed string as a guide. If a gastrostomy has been performed, a retrograde esophageal dilator with the Tucker bougie is an alternative method of dilation. Even if dilation is delayed until healing has occurred, there is still an increased risk of perforation, hemorrhage, or both when it is instituted. Therefore dilation should be carried out as carefully as possible. If bleeding occurs, dilation should be temporarily discontinued until it has stopped; dilation is then reinstituted cautiously. Although dilations restore the swallowing mechanism in patients with less severe strictures and in a few of those with severe lesions, it is not successful in restoring deglutition in the majority of those who have sustained a severe esophageal bum. Dilation fails not only because the strictures are deep and circumferential, but because they are multiple and long as well. When it becomes obvious that dilation is not going to be successful, it should be abandoned and esophageal substitution (colon interposition) carried out. At times, because of the proximal location of the burn, the anastomosis may have to be performed to the pharynx, and not infrequently extensive oropharyngeal reconstructive procedures are also necessary before an adequate and functioning swallowing mechanism is restored. Psychiatric support is sometimes needed during both the acute and the chronic phase. Liquid Caustics The newer liquid caustics, because of their high specific gravity, pass rapidly through the esophagus and reach the stomach soon after ingestion, destroying its full thickness and even adjacent tissues. Severe esophageal injury occurs frequently and with a higher incidence than is seen with ingestion of granular caustic substances. Emergency measures, especially treatment of hypovolemic shock, should be carried out. An outline of the management for ingestion of liquid corrosives appears below. I. Correction of hypovolemia 11. Contrast visualization 111. Aspiration and irrigation of stomach A. Alkaline aspirate-celiotomy 1. Necrotic stomach a. Esophagogastrectomy b. Colon interposition 2. Nonnecrotic stomachesoph agos copy a. Esophagus necrotic (1) Esophagectomy (2) Colon interposition b. Esophagus not necrotic (1) Antibiotics (2) Dilations B. Acid aspirate 1. Peritoneal signs-celiotomy; esophagogastrectomy if indicated 2. No peritoneal signs a. Esophagoscopy b. Antibiotics c. Dilations Clinical signs and contrast visualization of the esophagus and stomach are especially valuable in detecting impending damage or perforation of these viscera. Early in our experience these injuries were treated with the usual conservative regimen (esophagoscopy, steroids, antibiotics, and dilation) with poor results, namely, perforation, aortoesophageal fistula, and stricture. Consequently, a more aggressive plan of esophagectomy has been advocated in recent years that has resulted in a lessened morbidity and mortality and shortening of the hospital stay and convalescence [13]. With this therapy, a nasogastric tube is inserted cautiously and the stomach is aspirated. If the aspirate is basic and the alkalinity does not cease with gentle and limited irrigation, it is presumed that the stomach and esophagus are burned. Contrast visualization of the esophagus and stomach is also obtained. Esophagoscopy is not performed in this circumstance, as it does not give accurate information about the extent of necrosis. Instead a celiotomy is performed, and if the gastric wall is ecchymotic or black, indicating a necrotic stomach, a gastrectomy is performed. At times a burn extends through the stomach wall to involve the adjacent viscera such as colon, pancreas, spleen, or small bowel. If damaged, these structures are also removed. It has now been amply demonstrated that when

8 81 Current Review: Kirsh and Ritter: Caustic Ingestion the stomach is necrotic following caustic ingestion, the esophagus is also necrotic or is so severely burned that a nondilatable stricture will develop within several months [28]. Therefore a right thoracotomy is recommended, and if the esophagus is necrotic, total esophagostomy and left cervical esophagostomy should be carried out. Colon interposition is performed 4 to 6 weeks later. At present, with the less concentrated corrosives, the stomach may not be necrotic at celiotomy so that only a gastrostomy need be performed. If the distal intraabdominal segment of esophagus is necrotic or if contrast examination shows dilation of the esophagus with retention of either swallowed air or contrast material (which implies diffuse muscular necrosis and impending perforation), esophagectomy and left cervical esophagostomy are performed. If neither of these abnormalities is present, the extent and severity of the esophageal injuries are assessed by careful esophagoscopy and managed conservatively with antibiotics and later dilations. If the gastric aspirate is acidic and peritoneal signs are present, a celiotomy should be performed and gastrectomy, esophagostomy, or both carried out if indicated. When the gastric aspirate is acid (as is apt to occur when ingestion occurs on a full stomach) and there are no peritoneal signs, there is little likelihood that the stomach is seriously burned, and a celiotomy should not be performed. An esophagectomy is now recommended only if there are clinical signs of perforation or if a contrast examination shows evidence of impending perforation. When neither is present, an esophagoscopy is performed to assess the severity and extent of the injury. Strictures that develop following liquid caustic ingestion tend to begin just below the inferior constrictor and damage the entire esophagus, with multiple stricture formation in half or more of the esophagus that may convert it into a single, threadlike lumen. Because of their extensiveness, these strictures are exceedingly resistant to dilation. Therefore one should not persist with this form of management if it proves unsuccessful; instead, esophageal replacement with colon interposition should be performed before complications occur. Another controversial point concerning caustic injuries of the esophagus is whether the damaged esophagus should be removed at the time of or before colon interposition or left in place. The association of lye stricture and carcinoma of the esophagus has been known since 1896 [2, 5, 141. The exact incidence of its occurrence after lye ingestion is unknown. Kiviranta [19] believed that the incidence of esophageal cancer among victims of lye stricture is at least 1,000 times greater than that occurring in the normal population. A similar increased incidence of carcinoma in victims of lye ingestions has been reported by Joske and Benedict [18]. The interval between lye ingestion and the development of carcinoma ranges from 25 to 40 years. On the other hand, Marchand [22] failed to find a single case of carcinoma in 135 patients with caustic strictures of the esophagus, and Carver and coworkers [8] encountered only 2 patients with esophageal carcinoma among 233 patients with lye stricture. Although the development of carcinoma in a residual stenosed esophagus is always a possibility and a theoretical objection to leaving the organ in situ, we agree completely with Nardi [27] that the technical hazards certainly outweigh the theoretical desirability of removing a severely damaged esophagus. References 1. Aceto T, Terplan K, Fiore RR, et al: Chemical burns of the esophagus in children on a glucocorticoid therapy. J Med (Basel) lo:l, Alvarez AF, Colbert JC: Lye stricture of the esophagus complicated by cancer. Can J Surg 6:470, Ashcraft KW, Padula R: The effect of dilute corrosives on the esophagus. Pediatrics 53:226, Ashcraft KW, Simm JL: Accidental caustic ingestion in children: a review. Tex Med 68:86, Bigelow NA: Carcinoma of the esophagus developing in the site of lye stricture. Cancer 6: 1159, Bihkarzi H, Thompson ER, Shumrick DA: Caustic ingestion: current status. Arch Otolaryngol 89:112, Borja AR, Ransdell HT Jr, Thomas TV, et al: Lye injuries of the esophagus: analysis of ninety cases of lye ingestion. J Thorac Cardiovasc Surg 57:533, Carver CM, Sealy WC, Dillon MJ: Management of alkali burns of the esophagus. JAMA 160:447, 1956

9 82 The Annals of Thoracic Surgery Vol 21 No 1 January Citron B, Pincus JJ, Geokac M, et al: Chemical trauma of the esophagus and stomach. Surg Clin North Am 48:1303, Cotton R, Fearon B: Esophageal strictures in infants and children. Can J Otolaryngol1:225, Daly JF: Corrosive esophagitis. Otolaryngol Clin North Am June 1968, p Feldman M, Iben A, Hurley EJ: Corrosive injury to oro-pharynx and esophagus. Calif Med 118:6, Gago 0, Ritter FN, Martel W, et al: Aggressive surgical treatment for caustic injury of the esophagus and stomach. Ann Thorac Surg 13:243, Gerami S, Booth A, Pate JW: Carcinoma of the esophagus grafted on lye stricture. Chest 59:226, Gryboski W, Page R, Rush B: Management of total gastric necrosis following lye ingestion. Ann Surg 161:469, Haller JA, Andrew HG, White JJ, et al: Pathophysiology and management of acute corrosive burns of the esophagus: results of treatment in 285 children. J Pediatr Surg 6:578, Hollinger PH: Management of esophageal lesions caused by chemical burns. Ann Otol Rhinol Laryngol 77:819, Joske RA, Benedict EB: The role of benign esophageal obstruction in the development of carcinoma of the esophagus. Gastroenterology 36:749, Kiviranta NK: Corrosive carcinoma of the esophagus. Acta Otolaxyngol 42:82, Krey H: Treatment of corrosive lesions in the esophagus. Acta Otolaryngol [Suppl] 102: 1, Leape L, Ashcraft K, Scarpelli D, et al: Hazard to health-liquid lye. N Engl J Med 284:578, Marchand P: Caustic strictures of the esophagus. Thorax 10: Martel W: Radiologic features of esophagogastritis secondary to extremely caustic agents. Radiology 103:31, Middlekamp JN, Cone AJ, Ogura J, et al: Endoscopic diagnosis and steroid and antibiotic treatment of acute lye burns of the esophagus. Laryngoscope 71: 1354, Middlekamp JN, Ferguson TB, Roper CL, et al: The management and problems of caustic burns in children. J Thorac Cardiovasc Surg , Moody FG, Garrett JM: Esophageal achalasia following lye ingestion. Ann Surg 170:775, Nardi G: Surgical treatment of lye stricture of the esophagus by mediastinal colon transplant without resection. N Engl J Med 256:777, Ray J, Myers W, Saulter RD: Lye ingestion. JAMA 229: 765, Ritter F, Gago 0, Kirsh M, et al: The rationale of emergency esophagogastrectomy in the treatment of liquid caustic burns of the esophagus and stomach. Ann Otol Rhinol Laryngol 80:513, Ritter F, Newman MH, Newman DE: A clinical and experimental study of corrosive burns of the stomach. Ann Otol Rhinol Laryngol 77:830, Rosenberg W, Kunderman P, Vroman L: Prevention of experimental lye stricture by cortisone 11-control of suppurative complications by penicillin. Arch Surg 66: 593, Spain DM, Molomut N, Haber A: The effect of cortisone on the formation of granulation tissue in mice. Am J Pathol 26:710, Webb WR, Koutras P, Ecker RR, et al: An evaluation of steroids and antibiotics in caustic burns of the esophagus. Ann Thorac Surg 9:95, Weiskuff A: Effects of cortisone on experimental lye burns of the esophagus. Ann Otol61:681, Wilson WH, Winegarner FG, Jesseph JE: Gastric outlet obstruction following alkali ingestion. Ohio State Med J 65:1000, 1969