Esophageal Disorders

FIGURE 130.1 Barium esophagram shows food bolus impaction obstructing the distal esophagus.

Medical therapy has also been considered in the acute treatment of esophageal obstructions. The most extensively studied medical therapy is glucagon. Glucagon is a polypeptide produced in the pancreas that primarily functions as a regulator of euglycemia. It is theorized that due to structural similarities to the hormone secretin, glucagon is able to also relax smooth muscle of the gastrointestinal tract. Multiple retrospective and prospective trials have utilized glucagon at a dose between 0.5 and 2 mg alone or in combination with various effervescent contrast agents, anxiolytics, and contrast media with varying results (5). The only multicenter, placebo-controlled, double-blinded, randomized controlled trial utilized the combination of diazepam at varying doses with 1 mg of glucagon. In the active treatment arm, 38% of patients had their foreign bodies dislodge versus 32% in the placebo group. On critical analysis of the data, only approximately one-third of the patients in the active treatment group had their bolus dislodge within 1 hour. Given the short half-life of glucagon and studies showing the duration of effect of glucagon being less than 15 minutes, the authors concluded the bolus that were dislodged after 60 minutes were unlikely to be related to treatment effect (6). Systemic reviews have concluded that the six studies of the effect of glucagon have insufficient evidence to recommend is routine use and caution side effects of hypoglycemia and nausea and vomiting should be considered (5). The American Society for Gastrointestinal Endoscopy states in a recent practice guideline that glucagon is relatively safe and is an acceptable option. Its use, however, should not delay definitive endoscopic removal of a food impaction (1). Other medical therapies, including calcium channel blockers, have been suggested but experience is limited and use is not currently recommended in current gastrointestinal endoscopy guidelines.

Enzymatic therapy using papain or meat tenderizers has long been used in attempts to dissolve the food bolus (7). Given the risk for hypernatremia, local ulceration, and perforation proteolytic enzymes like papain are not recommended (1).

Nonendoscopic procedures for food bolus removal have involved tubes for suction or retrieval with radiographically guided graspers or balloons to pull out the bolus. Techniques have been described using a 34-F large-bore tube modified to aspirate the food bolus under fluoroscopic guidance (8). Special caution is advised because the bolus could potentially be dropped while passing through the hypopharynx, posing a risk of tracheal aspiration and obstruction. This suction technique should be attempted only by personnel experienced in gastrointestinal tube placement and airway management. With the widespread availability of endoscopy these nonendoscopic extraction techniques are seldom utilized and not discussed in current professional guideline statements.

Special Considerations for Subsequent Management and “Steakhouse Spasm”

Most authors have approached steakhouse syndrome and esophageal obstruction as disorders in which a food bolus impacts in or above a pre-existing esophageal lesion. Reported lesions include neoplastic, peptic, or caustic strictures, webs, distal rings, and vascular anomalies. Food impaction is also a common presentation of eosinophilic esophagitis (9). However, we reported data on several patients with food impaction and complete obstruction for 72 to 96 hours who had no underlying anatomic lesions (10). Subsequent endoscopy and barium radiographs were normal, but esophageal motility disorders were defined by esophageal manometry. A careful review of previous literature revealed that most reported cases of steakhouse syndrome had no anatomic explanation for obstruction, and we call this variant “steakhouse spasm” to emphasize the spastic nature of the obstruction.

FOREIGN BODIES

Presentation

It had been reported that over 1,500 people die yearly as a result of foreign-body ingestion (4,11). In the era of highly accessible flexible endoscopy, large case series have shown death in adults to be almost nonexistent (1). Commonly ingested items include coins, batteries, sharp and pointed objects, and cocaine packets. As previously discussed, food impaction is probably the most common upper GI foreign body that requires medical management in adults. In psychiatric patients or prisoners who ingest foreign bodies for secondary gain, nonfood objects are frequently ingested. Over 75% of foreign-body obstructions occur in pediatric patients (11). Children more often ingest coins and toys, whereas adults in the United States commonly ingest meat and animal bones as the cause for esophageal symptoms (12).

Treatment

Most objects pass spontaneously, but approximately 10% to 20% need to be removed endoscopically, and about 1% may require surgery (2). The preferred management for most foreign-body obstructions of the upper GI tract is removal with a flexible endoscope. Sharp objects in the esophagus and button batteries in the esophagus should always warrant emergent endoscopy. Although less than 1% of foreign bodies may perforate the gut, all sharp and pointed objects should be removed before they pass the stomach in an attempt to avoid distal intestinal perforation. Batteries, particularly the small button battery type, may cause caustic mucosal injury to the esophagus and removal is essential. Longstanding consequences of button battery damage to the esophagus include fistula formation and massive bleeding (13). Batteries that reach the stomach do not pose as serious a risk of mucosal damage because of the acid milieu. Batteries in the stomach may be observed with a radiograph every 3 or 4 days and endoscopic removal if symptoms develop or if the battery remains in the stomach for more than 48 hours (14). After an object is beyond the reach of the upper endoscope, it usually passes without difficulty. However, in cases where a high-risk object has passed the reach of a conventional endoscope, deep enteroscopy has been described as a potential therapeutic alternative to surgical exploration. In objects with low risk of perforation that have passed the reach of a conventional flexible endoscope, patients should be instructed to monitor their stool daily and radiography obtained every few days to assess progression. Patients should also be instructed on the symptoms of perforation and obstruction and need to seek medical intervention in a timely fashion. In cases where the object fails to progress or symptoms of obstruction or perforation develop surgical exploration is indicated (1). In recent years, drugs (most commonly cocaine) have been swallowed in packet form for transport or other reasons for concealment. Endoscopy is not recommended in these conditions because of the risk of packet rupture. Surgery is the safest way to remove these agents.

Foreign bodies lodged in the hypopharynx or proximal esophagus may require rigid esophagoscopy. Most other objects are amenable to removal with a flexible endoscope. It must be emphasized that in all cases the airway should be protected because of the risk of dropping and aspirating the object as it passes the hypopharynx. When there is any doubt or risk, tracheal intubation or rigid esophagoscopy with general anesthesia can be used.

CORROSIVE INJURY

Presentation

The clinical presentation of a patient with corrosive injury is dependent on the type (alkali or acid) and nature (solid or liquid) of the caustic substance (15). Liquid alkali is swallowed rapidly, causing less oropharyngeal injury but extensive damage to the esophagus and stomach. Solid alkali causes severe burns to the oropharynx and induces severe pain and expectoration such that little corrosive is actually swallowed. Acid ingestion injury is more localized to the gastric antrum, but systemic acidosis and toxicity have been reported. Thus, mouth pain, hoarseness, dysphagia, odynophagia, or abdominal pain can occur as determined by the agent ingested and location of the injury. Stridor, aphonia, dyspnea, and hoarseness suggest laryngeal edema. Substernal, abdominal, or back pain raises concern for mediastinitis or peritonitis (16).

Physical examination of the lips, mouth, and pharynx can reveal a spectrum of injuries from mild erythema to erosions, ulcers, and obvious severe burns. Some authors have graded the injury by the presence and severity of oropharyngeal findings at the time of admission. The presence or absence of oropharyngeal burns is not considered a reliable marker for esophageal damage (17).

Diagnosis

After the history and physical examination have been obtained, particular attention is devoted to systemic hemodynamics, oropharyngeal involvement, and airway status. The laboratory evaluation is directed at determining complications of the ingestion such as renal or hepatic insufficiency or anemia. Chest and abdominal radiographs should be performed to look for evidence of aspiration, visceral perforation, or mediastinal air. After the patient has been stabilized, the extent and severity of disease can be evaluated by fiberoptic endoscopy. Caustic injuries are graded using the commonly accepted Zargar classification and graded similar to cutaneous burns (Table 130.1) (18). The timing for endoscopy is influenced by the substance ingested, quantity of substance ingested, intention of ingestion, and symptoms at presentation. It is generally accepted that unless a third-degree burn is present in the posterior oropharynx, patients should undergo endoscopy in 12 to 24 hours to stage the involvement of the esophagus (19). Although some authors would suggest endoscopy could be performed safely up to 96 hours after ingestion, it is generally accepted that endoscopy be avoided between days 5 and 15 due to tissue softening (16). Despite this principle, successful endoscopic dilation has been described during this time frame (20).

When endoscopy was initially introduced as a diagnostic procedure for evaluating caustic ingestions, concern was raised about the risk of perforation. Authors opposed using early endoscopy or recommended not passing the endoscope beyond the first burned area. Recent work suggests that endoscopy can be safely performed early in the course and provides information about severity and extent of damage that may influence management (21,22). It remains controversial whether the endoscope can be passed through a circumferential third- or fourth-degree burn. Some authors suggest that aborting the procedure for a circumferential burn is not mandatory (16). When possible, a complete endoscopic examination evaluating the esophagus, stomach, and duodenum should be accomplished. Previous data suggest that endoscopic ultrasound may help predict risk of subsequent stricture formation; however, this does increase the risk of perforation and the time to complete the procedure. Because of these factors, endoscopic ultrasound is not considered standard of care after caustic ingestions (23).

TABLE 130.1 Zargar Endoscopic Classification of Caustic Injuries

Radiographic examination can be helpful, particularly when endoscopy is not available or is dangerous because of suspected perforation. In these situations, water-soluble contrast agents should be used. In the cases where perforation is not suspected, computed tomography has been investigated to determine the risk of mortality and surgery after a caustic ingestion. Lurie et al. showed CT could be used to predict mortality and need for surgery after a caustic ingestion, but had a substantially lower sensitivity in this small study (24). CT has also been investigated by Ryu et al. and was found to predict future stricture formation with higher sensitivity and specificity than endoscopy (25). At this time given the limited data, computed tomography should not be considered a suitable replacement of flexible endoscopy and should be used in cases where extraesophageal pathology is suspected (i.e., pulmonary aspiration).

Complications

Complications of corrosive ingestion depend on the nature of the agent, the quantity and concentration of the agent, and the contact time duration. Liquid alkali such as Liquid Plur was a 20% sodium hydroxide solution when introduced in the late 1960s and was subsequently reduced to a 5% solution after being implicated in 20% of reported caustic ingestions. Liquid alkalis have a high specific gravity and pass rapidly through the esophagus to the stomach. In dogs, violent regurgitation of gastric contents and pyloric and cricopharyngeal spasm cause a seesaw action that prolongs contact time. Solid alkali is usually in crystal form and causes severe pain in the oropharynx that limits further ingestion. Crystals adhere to mucous membranes of mouth, pharynx, and upper esophagus, causing predominantly proximal burns. Alkali produces injury by liquefaction necrosis. This type of injury enhances alkali penetration and prevents surface neutralization that results in full-thickness burns.

Concentrated acids produce a coagulative necrosis that forms eschar, which, with the coagulum, limits penetration to deeper muscular coats. Surface sloughing and perforation are therefore common problems. Late complications relate to location and extent of injury. Gastric injury may result in pyloric obstruction, antral stenosis, or hourglass deformity. Esophageal stricture may be proximal or distal, and despite careful management, develops in 10% to 20% of patients with caustic ingestion but varies depending on the severity of the insult. In patients with a grade 2B or 3 injury, stricture prevalence can be as high as 61% and 100%, respectively (16). In these patients, esophageal cancer has an estimated incidence of 2% to 30%, with an up to 3,000-fold increase over normal persons. Most cancers occur at the site of a stricture decades after the insult; however, cancers have been reported in as little as a year after corrosive ingestion (16,26).

In attempts to avoid cicatricial esophageal stenosis, several interventions have been tried with controversial results. Traditional approaches have used antibiotics, steroids, and early “prophylactic” dilations, but these cannot be supported by any well-controlled studies. Dilation remains the mainstay of therapy for patients with strictures after corrosive ingestions and the interval as well as method of dilation depends on the nature of the strictures (27). In the cases of recurrent benign strictures unresponsive to dilation after a corrosive ingestion, a number of authors have attempted intralesional corticosteroids to decrease the frequency of dilations. These studies have been limited by small sample size and differing techniques, so the utility of intralesional steroids is yet to be fully elucidated (28). Nasogastric tubes have been postulated to “stent” the esophagus open after caustic ingestion. These tubes have the benefit of giving a “roadmap” for future dilations as well as giving the ability to enterally feed a patient. Despite these advantages, given the risk of potentially serving as a nidus for long strictures, nasogastric tube placement is currently considered on an individualized basis (16). Prophylactic stenting with the use of plastic or self-expanding metal stents after caustic injury is not currently recommended. Stenting, however, remains an option for refractory benign strictures when other methods fail (27).

Treatment

Initial efforts are directed toward hemodynamic stabilization and airway support. The need for careful assessment of the airway cannot be overemphasized. Translaryngeal intubation or tracheostomy may be necessary. Evidence of esophageal perforation requires early surgical intervention. The use of neutralizing agents should be avoided because the exothermic chemical reaction can cause further damage by release of heat. Milk and water have both been used as diluting agents; however, due to the lack of data to suggest benefit and potential to obscure subsequent endoscopy both of these methods are not recommended. Gastric lavage and induced vomiting should be avoided due to potential for repeated damage to the esophageal mucosa. Nasogastric intubation should be avoided unless the tube is placed under direct visualization, and should not be routinely used unless needed to control the symptoms of emesis (16).

Early endoscopy is used when feasible. Complete examination of the esophagus, stomach, and duodenum should be attempted. If no significant injury is found and diet is tolerated, the patient can be discharged (29). In patients with significant injury, hospitalization and careful management are necessary (19). Intravenous acid suppression therapy is reasonable and commonly performed after caustic ingestion, although large well-designed trials have not been performed to date. A small prospective study showed that the initiation of high-dose intravenous proton pump inhibitors after caustic ingestion was associated with clinically significant mucosal healing; however, the study was limited by small sample size and lack of a control group (30). The prophylactic use of steroids or antibiotics is controversial and is not routinely recommended (31). Broad-spectrum antibiotic coverage is used for signs of aspiration, infection, or suspected perforation, or it may be used when deep ulcers are present and perforation seems imminent. Laryngeal edema is treated with short courses of high-dose steroids. Early bougienage using polyvinyl bougie dilators or pneumatic dilators can be used in an attempt to prevent strictures and is usually started 2 to 3 weeks after the ingestion. Patients are kept NPO until they can swallow their saliva. Generally patients with Zaragar classification less than 2a can be rapidly fed. Those with higher-degree Zargar classifications may require parenteral nutrition in the short term (19).

It must be remembered that corrosive injuries are often severe, causing full-thickness mucosal destruction and perforation. Even with direct visualization with flexible endoscopy the evaluation is imperfect and can overestimate the need for surgery in up to 15% of cases (32). For this reason, patients require both a clinical and endoscopic stratification as surgery is often the best option in severe cases (33,34).

ESOPHAGEAL PERFORATION

Presentation

Esophageal perforations may be iatrogenic or noniatrogenic. Iatrogenic injury from endoscopy is the most commonly reported cause of esophageal perforation (35). The development of small-bore fiberoptic endoscopy techniques has dramatically decreased the incidence of iatrogenic instrumental perforation from esophagoscopy (36). Previously used rigid esophagoscopy had a perforation rate of 0.2% to 1.9%. Perforation during esophagoscopy performed with modern flexible fiberscopes is approximately 0.033% (37). Other nonendoscopic iatrogenic causes include endotracheal intubation, obturator airway placement, and surgical complications. Noniatrogenic causes are usually barogenic ruptures. The most well-known “spontaneous” rupture occurred in the gluttonous Dutch Admiral Baron Van Wassanaer. The admiral gorged himself and induced forceful vomiting for relief. His autopsy by Hermann Boerhaave was published in 1724 and described the pathologic findings of barogenic esophageal rupture. Resultant signs and symptoms are similar for Boerhaave syndrome and iatrogenic perforations. Pain is a near-universal experience, and 70% of patients develop acute pain. Fever is the second most common symptom irrespective of location of esophageal perforation (35).

The site of esophageal perforation largely influences other associated symptoms in esophageal perforation. Patients with abdominal esophageal segment tears have had retroperitoneal air and vague epigastric pain. Patients with thoracic perforations often complain of abdominal and back pain. Cervical perforations are associated with subcutaneous emphysema and chest pain. Symptoms that occur during or shortly after an esophageal procedure should raise concern for iatrogenic perforations. Other clinical findings in patients with esophageal tears include pleural effusion, pneumothorax, dysphagia, nausea and vomiting, cervical crepitus, hematemesis, and shock. However, asymptomatic perforations have been demonstrated radiographically, which emphasizes the importance of an accurate history and a high index of suspicion.

Diagnosis

The diagnosis of esophageal perforation may be fairly obvious, particularly in the iatrogenic group. Plain chest radiographs may suggest perforation in over 90% of patients (Fig. 130.2) (38). Findings include mediastinal air, pneumothorax, pleural effusion, infiltrate, or subcutaneous emphysema. Hyperextended neck films can reveal widened spaces, air, or esophageal displacement. Fears of barium mediastinitis have traditionally led to the use of water-soluble contrast media (39). Reports indicate that iodinated water-soluble contrast radiographs may be normal in 20% to 25% of thoracic and 50% of cervical perforations (38). Therefore, negative or equivocal findings on water-soluble studies should be immediately re-examined with barium sulfate contrast radiographs. Other authors believe that barium does not potentiate mediastinal inflammation. Because it is more palatable than water-soluble agents and less dangerous if aspirated into the bronchial tree, many use dilute barium in the initial examination to take advantage of its better coating and definition. Regardless of the agent used, it is important to examine the entire esophagus in multiple positions.