Caustic esophageal injury in children

INTRODUCTION — Accidental ingestion of caustic agents continues to be a major concern for pediatric emergency department clinicians. Caustic ingestions are seen most often in young children between one and three years of age and can cause severe acute injury and long-term complications, especially the development of esophageal strictures [1-5].

The evaluation and management of a child with suspected ingestion of a caustic substance are described here. Caustic esophageal injury in adults is discussed separately. (See "Caustic esophageal injury in adults".)

EPIDEMIOLOGY — The National Poison Data System compiles data annually from a population of approximately 300 million people served by the American Association of Poison Control Centers (AAPCC). In 2018, 2.1 million toxic exposures in humans were reported in the United States, of which approximately one-half occurred in children five years or younger [6]. The most commonly ingested caustic substances were household cleaning products (particularly, household bleach), which accounted for 11 percent of the toxic ingestions in young children, and cosmetics or personal care products. In a retrospective series of more than 1000 cases in Germany, cleansers and alkaline liquids accounted for 13 percent of cases and the numbers of all ingested foreign bodies and substances increased between the study years of 2005 to 2017 [7]. During the coronavirus disease 2019 (COVID-19) lockdown, an Italian pediatric center reported decreased cases overall but increased hospitalization rates [7,8]. Caustic ingestion is most common in young children between one and three years of age [9], with boys accounting for 50 to 62 percent of cases [4,5].

Most ingestions by children are accidental, and the amounts ingested tend to be small. The opposite is the case in adolescents and adults, in whom ingestion often is deliberate and related to attempted suicide [10,11]. In such cases, the amount ingested may be large and the injury to the esophagus and stomach often severe [12]. Cases of alkali ingestion as a result of child abuse have been reported [13,14].

Common caustic ingestions are of household cleaning products, including household bleach, or oven and drain cleaners, which are strong lyes that contain sodium and potassium hydroxides [1,9,15,16]. Other common cleaning products include laundry and dishwasher detergents, and cleaning agents with sodium phosphate, sodium carbonate, and ammonia [2,3,17]. Other caustic ingestions include cosmetic products such as hair relaxer, swimming pool products, and toilet bowl cleaners [18]. Alkaline agents are more commonly ingested compared with acidic agents. (See 'Risks of specific substances' below.)

PATHOPHYSIOLOGY — A substance is caustic if it is capable of burning or corroding organic tissue by chemical action, typically due to strong acid or alkaline properties.

Acid versus alkali — Alkalis and acids produce tissue injury by different mechanisms:

●Alkali – Alkaline agents tend to cause esophageal injury if the pH is above 11.5 to 12.5 [19,20], via liquefaction necrosis [21]. This type of injury leads to early disintegration of the mucosa, allowing deep penetration and even perforation [22,23]. Penetration into the esophageal wall by alkaline agents varies with the concentration and length of time that the agent remains in contact with the mucosa [24]. Experimentally, a 10% solution of sodium hydroxide (lye) may require as long as one minute of contact to produce a deep burn, whereas a 30% sodium hydroxide can cause transmural esophageal necrosis within one second of contact [15]. In animal experiments, even weakly concentrated alkaline agents (eg, 1.8% sodium hydroxide) can produce superficial necrosis if left in contact with the mucosa for one hour [24].

●Acids – Acids or corrosive agents tend to cause esophageal injury if the pH is less than 2, via coagulation necrosis. As compared with alkali agents, esophageal injury from acids tend to be attenuated and perforation is less common because the coagulum that forms on the mucosal surface may limit deeper penetration of the caustic substance [22]. The alkaline pH and squamous epithelium of the esophagus also help limit the severity of esophageal injury from acids. Despite these mitigating factors, 6 to 20 percent of acid ingestions result in esophageal burns [4,23,25,26]. Upper airway injuries also are common with ingestion of acids [22]. (See 'Airway injury' below.)

Delayed injury — Following the initial necrosis (whether caused by alkali or acid), additional destruction takes place over the first week, caused by inflammation and vascular thrombosis (table 1) [27,28]. By 10 days, there is formation of granulation tissue and weakening of the esophageal wall. During this early period, the esophagus is increasingly vulnerable to perforation [9,29]. By three weeks, fibrogenesis and stricture formation are well under way and perforation is less likely.

RISKS OF SPECIFIC SUBSTANCES — The risk and severity of injury is affected by the corrosive properties of the ingested substance (eg, the pH) (see 'Acid versus alkali' above) but also by the amount, concentration, and physical form of the agent (eg, solid or liquid) and duration of contact with the mucosa. The approximate pH of various cleaning products are shown in the figure (figure 1).

Esophageal injury — Esophageal burns account for most of the serious injuries, lead to the chronic complications associated with toxic ingestions, and have been reported in 18 to 46 percent of caustic ingestions in children [9,25,30-34]. Injury to the lips, oropharynx, and upper airway also may occur.

The likelihood of esophageal injury depends in part on the substance ingested:

●Detergents – Standard liquid household detergents, phosphates, and bleaches have pHs ranging from 9 to 11 and rarely cause serious injuries, unless taken in large amounts [9,35].

Concentrated forms of laundry or dishwasher detergent, especially single-use packets ("capsules," "gel packs," or "pods") are increasingly available and occasionally cause esophageal injury and airway compromise, as documented in multiple case reports [36-41]. Other complications with these products are contact dermatitis and corneal injury [42]. In registries from the United Kingdom and United States, in which 30 to 40 percent of exposed individuals remained asymptomatic; the remainder were classified as minor (50 to 60 percent), moderate (2 to 3 percent), or severe (0.5 percent), and one patient died [43,44]. Gastrointestinal endoscopic evaluation should be considered for patients with significant symptoms, including airway compromise or dysphagia, specifically those being considered for airway evaluation. Endoscopic findings may not correlate with symptoms. Mild esophageal injury identified on endoscopy is probably not clinically important if the patient has few or no symptoms.

These products typically are small and colorful, looking somewhat like candy. Approximately 10,000 ingestions by children five years and younger are reported annually despite safety modifications to packaging [45]. In 2018, the AAPCC issued a "high alert" about a sharp increase in laundry pod ingestions among teenagers [46]. These were intentional exposures, apparently in response to a "challenge" on social media. In addition to esophageal injury, these exposures also have been associated with respiratory distress, pulmonary edema, and, occasionally, coma and death.

●Bleach – Household bleaches are relatively common ingestions but rarely cause esophageal burns. These are typically composed of approximately 5% sodium hypochlorite and have a pH of approximately 11, which is below the pH threshold of 11.5 to 12.5 that tends to cause esophageal injury [3,19,22]. When burns do occur with these agents, they tend to be mild and do not require treatment [19,23,47,48]. Industrial-strength bleaches or household bleaches from other countries may be much more corrosive because of a higher concentration of sodium hypochlorite [19].

●Batteries – Button batteries lodged in the esophagus can rapidly cause injury to the esophagus and critical surrounding structures, in part due to leakage of alkaline material. This issue is discussed in a separate topic review. (See "Button and cylindrical battery ingestion: Clinical features, diagnosis, and initial management".)

●Hair relaxer – Cosmetic products such as hair relaxers are ingested commonly, but they appear to be rare causes of esophageal injury [49,50]. In one review of 29 patients with hair relaxer ingestions who underwent upper endoscopy, six patients (20 percent) had esophageal mucosal injury and five patients (17 percent) had gastric mucosal injury, but these injuries were limited to grade 1 (superficial) injuries [51].

●Wart remover – Wart or callus removers containing salicylic acid have caused esophageal injury and scarring in case reports; the colloid suspension is viscous, which likely contributes to the injury by prolonging contact with the mucosa [52,53].

●Other household products – Common acid household products include toilet bowl cleaners, battery fluids, and muriatic (hydrochloric) acid used in swimming pools [2,54]. Children living on farms may be at particularly high risk for incurring serious burns because industrial-strength agents are more commonly used in this setting [55].

●Solids versus liquids – Solid caustic materials, which tend to adhere to the mucosa, can produce deep burns of the oral cavity and esophagus [23], but they are less likely to reach the stomach [56]. The immediate and severe pain produced by these products may limit the amount ingested and thus lessen the extent of injury [20]. Powdered or granular detergents tend to injure the upper airway; esophageal injury is less common but has been reported [20,38,57]. (See 'Airway injury' below.)

Gastric injury — The incidence and severity of gastric lesions in caustic ingestion have not been described extensively, in part because this type of injury is less common than esophageal injury. In addition, older reports may have underestimated the presence of gastric injury because of the use of rigid endoscopes that frequently were not advanced beyond the first visible esophageal burn [47]. In a review in which flexible endoscopes were used to evaluate 156 children with caustic ingestion, 11 percent had both esophageal and gastric burns and 9 percent had gastric burns only [34].

Among patients with alkaline ingestions, gastric injury is most common in those who ingest relatively large volumes (200 to 300 mL); in one series of such patients, gastric injury was seen in more than 80 percent [12]. Large ingestions have resulted in gastric perforation, hemorrhage, and death from erosion into a bronchial vein [23].

Gastric injury often is most severe with acids, such as sulfuric acid, which trigger antral spasm, leading to pooling of acid in the antrum. This may cause severe burns in the prepyloric area and result in pyloric obstruction [23]. In a review of 98 children with ingestion of acid substances, eight (8.2 percent) developed gastric outlet obstruction at a mean of 27 days after ingestion, necessitating gastrojejunostomy [58]. In a smaller series of 20 patients with gastric outlet obstruction, two also had esophageal strictures [59]. Duodenal erosions have been reported in occasional cases [12]. (See 'Pyloric stenosis' below.)

Airway injury — Upper airway injuries are more common with ingestion of acids, perhaps related to their bad taste, which tends to stimulate gagging, choking, and attempts to spit out the ingested material [22]. Powdered or granular detergents also tend to injure the upper airway, resulting in stridor and epiglottitis.

CLINICAL MANIFESTATIONS — The clinical features of caustic ingestions vary widely. Early signs and/or symptoms may not correlate with the severity and extent of tissue injury, particularly in young children.

The following symptoms are common after a caustic ingestion:

●Gastrointestinal tract injury

•The most common symptom is dysphagia, which can occur even with mild esophageal injury [4]. This symptom may be caused by loss of esophageal motility during the acute phase of injury [60], similar to what is seen in peptic esophagitis [61,62]. Esophageal dysmotility and prolonged esophageal transit time may persist for several weeks following severe esophageal burns. Persistent dysphagia also may be caused by fibrosis of deeper muscle layers, with or without stricture formation [60]. Fibrosis alone produces a decrease in the amplitude of esophageal contractions and elevated basal pressure. (See "Evaluation of acute dysphagia in children".)

•Patients may also present with drooling, retrosternal or abdominal pain, and hematemesis. Deep esophageal burns can be complicated by esophageal perforation, which can cause mediastinitis and the development of a tracheoesophageal fistula [22].

●Upper airway injury – Symptoms suggesting upper airway injury include stridor, hoarseness, nasal flaring, and retractions. Such symptoms suggest injury to the epiglottis, which can be severe (especially in children younger than two years of age) and may require emergency intubation or tracheotomy [20,63]. Although symptoms of upper airway injury usually begin rapidly after contact with caustic agents, there may be a one- to two-hour delay with ingestion of powdered agents [20]. Occasionally, the clinical presentation may mimic anaphylaxis, leading to misdiagnosis if the history of caustic ingestion is not revealed [64].

The presence or absence of any of the clinical features described above does not reliably predict ingestion or the presence or severity of esophageal or gastric burns. In a single series of 378 caustic ingestions in children, 12 percent had no symptoms but had severe esophageal burns, whereas 82 percent who were symptomatic at presentation had minimal or no esophageal involvement [4]. The presence or absence of oral lesions also is a poor predictor of esophageal injury. In a review of multiple case series totaling 489 children, 45 percent without oral burns had esophageal burns and 54 percent with oral burns had accompanying esophageal burns [47].

INITIAL EVALUATION — Diagnostic evaluation and management of a patient with a known or suspected caustic ingestion depends on whether the patient is symptomatic, tempered by the type of agent ingested.

History and examination — The history should seek to establish the timing of the exposure, including whether it was directly observed by a caretaker, and an estimation of the amount of the substance ingested. When possible, the exact type and brand of substance ingested should be identified and the pH measured or determined from the material safety data sheets or poison control center.

The physical examination includes rapid evaluation of mental status, vital signs, and pupils to evaluate for central nervous system excitation or depression. The patient should be closely evaluated for respiratory compromise, including the presence or absence of stridor, hoarseness, nasal flaring, retractions, and wheezing. Drooling, food refusal, or complaints of dysphagia suggest oropharyngeal or esophageal injury. The lips and oropharynx should be inspected for burns (picture 1). However, the presence or absence of oral burns is a poor predictor of esophageal injury, as noted above.

Signs that suggest the possibility of esophageal perforation include persistent severe retrosternal or back pain, especially if fever is present suggesting mediastinitis. Localized severe abdominal tenderness, rebound, and rigidity suggest the possibility of gastric perforation with peritonitis.

Other considerations that should be explored by the history and examination include:

●The possibility of toxic effects in addition to the potential caustic injury, which depends on the nature of the substance ingested. (See "Approach to the child with occult toxic exposure".)

●The possibility that the ingestion was intentional (eg, suicide attempt or child abuse) [13].

●The possibility of multiple ingestions in the same individual. This is more likely for those in whom the exposure was intentional (eg, adolescents or victims of child abuse).

Imaging — A chest radiograph is appropriate in any patient with respiratory symptoms. The goal is to evaluate for other possible causes of the respiratory symptoms (including foreign body or pneumonia), as well as for possible complications of the caustic ingestion, such as esophageal or gastric perforation. Signs of perforation include pneumomediastinum, mediastinal widening, subcutaneous emphysema in the neck, pleural effusions, hydropneumothorax, or subdiaphragmatic air. A chest radiograph generally is not necessary in asymptomatic patients.

Radiologic contrast studies (upper gastrointestinal series) generally are not valuable in the initial stages of the evaluation, because they are unreliable in detecting acute injury or in predicting stricture formation [10,65]. Contrast studies may be appropriate in the initial stages in selected patients in whom there is a clinical suspicion of esophageal perforation based on findings from plain radiography or endoscopy [66]; in this case, only water-soluble contrast should be used, rather than barium.

Computed tomography (CT) or magnetic resonance angiography are sometimes needed to evaluate patients for the possibility of esophageal perforation with erosion into vascular structures, including an aortoesophageal fistula. In adults, there are some data to suggest that CT findings may be more reliable than endoscopy in detecting transmural necrosis [67]. A few cases of life-threatening or fatal hemorrhage have been reported, most of which were caused by button battery ingestion and presented with a sentinel bleed (hematemesis or melena) up to two weeks after removal of the battery [68]. (See "Button and cylindrical battery ingestion: Clinical features, diagnosis, and initial management", section on 'Esophageal impaction'.)

INITIAL MANAGEMENT

Stabilization and supportive care — The initial step in management of caustic ingestion is supportive care and close observation, with an emphasis on preventing vomiting, choking, and aspiration (algorithm 1).

The following strategies should not be used:

●Induction of vomiting is contraindicated because vomiting may lead to additional esophageal injury if gastric contents come in contact with the esophageal mucosa [47].

●Neutralizing agents should not be used, because of concern about additional damage from heat injury during the neutralization process [69].

●Diluting agents (eg, milk or water) should not be given, because of safety considerations and lack of efficacy. The safety concern is that ingestion of a large volume of a diluting agent can induce vomiting, potentially leading to further complications. In the presence of acute airway swelling and obstruction, diluting agents clearly are contraindicated [47]. In addition, the amount of milk or other fluids that would be necessary to significantly dilute a caustic agent is too large to make such therapy practical.

●Activated charcoal generally should not be given to children who have ingested acidic or alkaline corrosives (eg, sodium or potassium hydroxide, or hydrochloric or sulfuric acid). This is because charcoal will obstruct the view of the endoscopist and because these small, highly ionized chemicals are poorly absorbed by charcoal. (See "Gastrointestinal decontamination of the poisoned patient", section on 'Contraindications'.)

Diagnostic evaluation and management depends on whether the patient is symptomatic, tempered by the type of agent ingested.

Asymptomatic patients — Patients who are asymptomatic (no evidence of oral lesions and no dysphagia, vomiting, or other symptoms) should be observed for several hours to monitor fluid intake and overall status (algorithm 1) [4,5,70]. The observation period is particularly important for children who may have ingested a powdered caustic agent, for which the effects may be delayed by several hours after the ingestion. If the patient remains asymptomatic with normal swallowing and if the suspected substance is known to be of low causticity (eg, household bleach), or if it is unclear from the history whether the patient ingested a caustic substance, then the patient is at low risk for complications and endoscopy usually is unnecessary [25,30,47,48,71,72]. By contrast, endoscopy is appropriate for patients who are suspected to have ingested a more caustic substance (eg, drain cleaner) or substance of questionable causticity, even if they are asymptomatic. This is because the presence or absence of symptoms in young children is not a reliable predictor of injury [32,34].

Symptomatic patients — Patients with symptoms should be admitted to the hospital and closely monitored. Those with significant airway compromise (stridor or retractions) may require intensive care, endotracheal intubation, and/or tracheostomy. All patients with symptoms also should be evaluated with upper endoscopy to evaluate the extent of the injury, unless endoscopy is contraindicated because of respiratory compromise (algorithm 1).

Endoscopic evaluation

Indications — Upper endoscopy should be performed in all patients who are symptomatic, who have oral burns, or who are known to have ingested a substance with a high risk of esophageal injury. For patients with alkali ingestion and minimal symptoms, such as vomiting or drooling alone, it is reasonable to observe overnight and proceed to endoscopy only if symptoms persist [73,74]. The endoscopy ideally should be performed within 24 hours of the ingestion to evaluate the extent of injury, predict prognosis, and guide management [32,34,75,76]. While there are few data to determine the ideal timing of endoscopy, very early endoscopy (eg, less than six hours after the ingestion) may not show the full extent of the injury and endoscopy after four days increases the risk of perforation [9,22,32,34,75]. Endoscopy is contraindicated in patients who are hemodynamically unstable, have severe respiratory distress or evidence of perforation on imaging (see 'Imaging' above), or exhibit severe oropharyngeal or glottic edema and necrosis.

Flexible video endoscopes are typically used for this evaluation. Unless the esophageal burns are very severe (raising concerns about perforation), it is important to pass the endoscope beyond the first burn and into the stomach to fully evaluate the extent of injury (picture 2A-B) [9,12,32,34]. The endoscopist should specifically observe the vocal cords to evaluate for airway injury. Flexible endoscopy may be performed under moderate or deep (intravenous) sedation in many patients, but general anesthesia should be used for those with severe burns or with upper airway signs or symptoms [17]. Rigid endoscopes are still used in some centers, but because they are associated with higher risks of perforation as compared with flexible endoscopes, they usually are not advanced beyond the first significant esophageal burn, which limits their diagnostic utility [34].

Staging — A grading system for esophageal injury helps to predict clinical outcomes and guide management (table 2). Grading should be documented in operative note to assist in follow-up assessments.

The most widely used grading system was developed based upon a study of 81 patients (primarily adults) with corrosive ingestion and is a modification of classification schemes used for cutaneous burns [75]. This grading scheme is widely used in adults, although few studies have examined its validity.

●Grade 0 – Normal mucosa

●Grade 1 (superficial) – Mucosal edema and hyperemia

●Grade 2 (transmucosal) – Friability, hemorrhages, erosions, blisters, whitish membranes, and superficial ulcerations

•Grade 2A – No deep focal or circumferential ulcers

•Grade 2B – With deep focal or circumferential ulcers

●Grade 3 – Areas of multiple ulceration and areas of brown-black or greyish discoloration suggesting necrosis

•Grade 3A – Small scattered areas of focal necrosis

•Grade 3B – Extensive necrosis

A novel scoring system in adults has been proposed that assigns a prognostic score based on intake of acidic (versus nonacidic) caustic substance, neutrophil count ≥75 percent, bicarbonate ≤22 mEq/L, and severe endoscopic injury (esophageal injury >grade 2b or stomach injury ≥grade 3) [77]. Similar scoring in children has not been well established; however, one study found that the risk of severe esophageal lesions increased with number of symptoms compared with asymptomatic patients [72].

Follow-up — If the endoscopy demonstrates significant esophageal burns (grade 2A or higher (table 2)), we suggest follow-up evaluation with a contrast study two to three weeks post-ingestion, even if the patient is asymptomatic (algorithm 1). Alternatively, it is reasonable to manage asymptomatic patients with close clinical follow-up and perform a contrast study only if they develop symptoms suggestive of a stricture (see 'Stricture formation' below). Collaboration with otolaryngology to evaluate the airway may be needed, depending on the symptoms.

Nasogastric tube — In patients in whom endoscopy reveals extensive circumferential burns (grade 2B or 3 using the system above), a nasogastric tube (NGT) should be placed under direct visualization during the endoscopic procedure. An NGT should not be inserted blindly, because perforation or additional injury can occur while passing the tube.

The NGT can provide a route for nutritional support during the healing phase and help maintain a lumen during stricture formation. It also can serve as a guide for esophageal dilation [78]. There is no need to put the NGT to suction, unless a patient has nausea, vomiting, or concern for obstruction or peritonitis. If the NGT is dislodged, it is reasonable to defer replacing the tube, depending on other clinical parameters. (See 'History and examination' above.)

Gastrostomy tube — In patients with severe extensive esophageal burns, consideration should be given to placement of a gastrostomy during or shortly after the initial endoscopic evaluation. A gastrostomy can be used for feeding and also for retrograde dilation of strictures. If retrograde dilation is anticipated, a string or suture can be placed from the nose to the gastrostomy, which is used to guide the dilators. Unlike an NGT, which can be pulled out inadvertently and may be impossible to replace, a string is less likely to be pulled out.

Medications

●Antibiotics – In our practice, we administer antibiotics for patients with suspicion of perforation on chest radiograph or endoscopy and for those with grade 2B or 3 burns (endoscopic evidence of necrosis). We use a third-generation cephalosporin and treat for approximately one week. Giving prophylactic antibiotics to patients with severe esophageal burns is common practice and has been used in key case series, but there are no clear standards or studies to guide patient selection [79].

●Acid suppression – For symptomatic patients, we treat with a proton pump inhibitor for approximately one week. This is common practice, as reported in case series, but has not been specifically been evaluated for efficacy [74].

●Corticosteroids – The use of corticosteroids for caustic esophageal injury is controversial, and practice varies among institutions. We suggest the following approach, based on the limited clinical evidence available:

•For patients with documented grade 2B lesions (deep focal or circumferential ulcers (table 2)) and no signs of esophageal perforation, we suggest a brief course of high-dose corticosteroids, eg, methylprednisolone 1 g/1.73 m² intravenously for three days. This approach is also suggested in a European guideline [80].

•For patients with any other stage lesion, we suggest against treatment with corticosteroids. Patients with grade 1 or 2A lesions are unlikely to benefit from corticosteroids since their risk for stricture formation is relatively low. For patients with grade 3 lesions, corticosteroids may soften the wound and increase the risk and/or mask the symptoms of perforation.

Based on limited clinical trial data, the role of corticosteroid therapy in children with caustic esophageal injury is uncertain and their efficacy may depend on burn severity. A meta-analysis of three randomized trials (244 patients) comparing corticosteroids versus no corticosteroids in children with caustic esophageal injury did not detect a significant reduction in esophageal stricture formation (19 versus 31 percent; risk ratio 0.63, 95% CI 0.29-1.37) [81]. However, there was important heterogeneity among the three trials. Most notably, one trial enrolled only patients with grade 2B lesions and used a brief course of corticosteroids [79], whereas the other two trials enrolled patients with grade 1, 2, and 3 lesions and used much longer courses of corticosteroids [32,82]. In the trial involving children with grade 2B esophageal injuries, subjects were treated with methylprednisolone (1 g/1.73 m² for three days) or placebo, in addition to ceftriaxone and ranitidine [79]. The corticosteroid-treated group had lower rates of stricture development compared with the placebo group (14 versus 45 percent, as assessed by radiography, and 11 versus 30 percent, as assessed by endoscopy) and shorter duration of parenteral nutrition (9 versus 17 days). These findings suggest a benefit of a brief course of high-dose corticosteroids in the subset of patients with grade 2B lesions. By contrast, in the other two trials, which included a mix of patients with esophageal burns from grades 1 to 3, rates of esophageal stricture formation were similar with and without corticosteroid treatment (23 versus 25 percent, respectively) [32,81,82]. The majority of esophageal strictures in these two trials occurred in patients with grade 3 burns, suggesting that stricture formation may be inevitable for these patients regardless of therapy.

LATE COMPLICATIONS — The most common complication of caustic ingestion is stricture formation, which can cause esophageal obstruction, presenting with dysphagia. Patients with severe gastric injury may develop pyloric stenosis. Caustic ingestion also confers an increased risk for developing esophageal carcinoma, as described below. Eosinophilic esophagitis developing after caustic ingestion has been described in a case report [83].

Stricture formation

Incidence and risk factors — Stricture formation is the primary complication of caustic injury, occurring in 3 to 57 percent of ingestions with documented esophageal burns and in virtually all patients with severe circumferential burns (grade 2B or 3 in the scheme outlined above) [4,5,9,32]. More superficial or non-circumferential burns (grade 1 or 2A) are less likely to result in strictures [65], although one report found that one-half of the children with non-circumferential transmucosal burns (grade 2A) went on to develop esophageal stricture [84]. A few older reports described patients with apparently normal endoscopic findings who went on to develop strictures [9], but it seems likely that a burn was present but was missed because the endoscopy was performed with a rigid endoscope.

Clinical manifestations and diagnosis — All patients with significant esophageal burns (grade 2A or higher) should be evaluated with a water-soluble contrast esophagogram to assess for stricture formation. The study should be performed two to three weeks after the ingestion, or sooner if the patient develops dysphagia (difficulty swallowing), which suggests esophageal obstruction. Although barium may still be used at some facilities, water-soluble contrast agents should always be used if a perforation is suspected.

Most patients who develop a stricture will have obstructive symptoms within two months. In patients with the most severe burns, there may be complete obstruction within three weeks [85]. At six weeks, the esophagogram shows a long constriction (image 1). Patients with severe strictures commonly have recurrent symptoms after two years, despite repeated dilation.

Dilation — Once a stricture is confirmed radiologically, esophageal dilation usually is required to maintain or reestablish normal swallowing. Balloon dilators under endoscopic control are our preferred approach; bougies also may be used (picture 3). The endoscopic appearance of a stricture prior to dilation demonstrates an abrupt tapering in the mucosa. This can be mild, with mucosal edema and erythema, or more pronounced, with friability, erosive changes, whitish plaques, and fibrinous material. In more severe cases, there may be blistering or frank necroinflammatory changes.

For patients who develop strictures, the optimal time for beginning dilation has not been established. Most clinicians wait three to six weeks after the initial injury before beginning dilation because there is some evidence that performing earlier dilation will increase the risk of perforation [86-88]. However, two retrospective series in children suggest that strictures were less likely to be recurrent if they were treated with early dilation (7 to 28 days after the injury) versus later dilation (beginning four weeks or more after the injury) [89,90].

A variety of dilators can be used, including balloon dilators, silicone or tungsten bougies, Maloney antegrade dilators, or Tucker dilators (used for retrograde dilation); some dilators can be passed over a string or guide wire. Because caustic strictures appear to perforate easily, retrograde dilation has been considered the safest method, although this method requires a gastrostomy and a string for guidance [78]. The risk of perforation probably depends on many factors, including the cause of the stricture (perforation appears to be more common in caustic strictures as compared with those due to esophageal atresia), interval of time since injury, length and diameter of the stricture, total number of previous dilations, and operative technique (operator experience and selection of maximal dilation) (see "Adverse events related to endoscopic dilation of esophageal strictures", section on 'Esophageal perforation'). The techniques used for children are similar to those for adults, except that the maximal dilation of the esophagus is smaller for a child. (See "Endoscopic interventions for nonmalignant esophageal strictures in adults".)

Balloon dilators under endoscopic control are commonly used in children and are our preferred approach [91,92]. In theory, perforation should be less likely with balloon dilation because only radially directed force is exerted, as compared with use of a bougie, in which there is also a longitudinal shearing force. The reported rates of esophageal perforation for balloon dilation range from 1 to 13 percent per procedure [93-97], compared with 5.6 percent for bougienage [96]. Balloons come in a variety of sizes and lengths. The size of the balloon dilator is selected using an estimate of the esophageal lumen diameter. One group selects a balloon catheter with a diameter equal to the interpeduncular distance of the child's thoracic spine, as measured on a preprocedure radiograph [96]. Dilation balloons are available as small as 2 mm (these are designed for biliary dilation), and these can be used for the tightest strictures. For strictures narrower than 2 to 3 mm, dilation balloons with a wire guide may prevent creating a "false track," in which the balloon may pass through damaged tissue along the mucosal surfaces rather than through the esophageal lumen.

The use of fluoroscopy varies widely among institutions. This technique may be of greatest benefit for patients with small-diameter lumen or near-complete occlusion, in patients where there is a high concern of perforation based on endoscopic findings, or in those with a prior history of perforation. Fluoroscopy can be used for any dilation technique and is used to identify the "waist" of the stricture (the area of narrowing relative to the non-dilated areas proximal and distal to the stricture). Fluoroscopy also can demonstrate leaks or other procedure-related adverse events. Due to the risk of ionizing radiation and recurrent use of fluoroscopy in these patients, tracking the fluoroscopy time and/or dose with each procedure is important.

Although esophageal dilation may be beneficial initially, repeated dilations usually are needed. Only 33 to 48 percent of patients with caustic strictures have long-term success with repeated dilations [5,32]. The remaining patients, who often have long strictures, have increasing difficulty in swallowing because of progressive obstruction.

Other interventions — Mitomycin (also known as mitomycin-C), an inhibitor of fibroblast proliferation, helps to prevent recurrent stricture formation in patients with severe caustic esophageal injury [98-102]. Mitomycin is applied topically to the strictured area of the esophagus at the end of a dilation session [103] or, in some cases, directly infiltrated at the stricture site [104]. Mitomycin typically is used as a second- or third-line agent for strictures deemed resistant to repeated dilation or steroid injections.

In a randomized study, patients treated with mitomycin and dilation therapy required fewer dilations compared with those treated with dilation therapy alone and were also more likely to be asymptomatic at six months' follow-up (81.6 versus 40 percent, respectively) [105]. Additionally, in a randomized trial in 40 patients with localized caustic esophageal strictures (<3 cm in length), treatment with mitomycin was associated with a reduced need for repeated dilation (3.85 versus 6.90 dilation sessions) and higher rates of complete resolution during the six-month follow-up period (80 versus 35 percent resolution) as compared with placebo [101]. Similar results were seen in two retrospective case series in which topical mitomycin was successful in eliminating or reducing the need for repeated dilations in 67 to 82 percent of the patients [100,106]. In another case series, direct infiltration of mitomycin significantly reduced the frequency of periodic dilations [104]. The relative risks and benefits of topical application versus infiltration have not been established.

Esophageal stenting has been described in limited series for use in management of refractory caustic esophageal strictures [107,108]. Stents have also been used in children after surgery for congenital strictures and esophageal injury from batteries, with mixed success [107,109]. Most data are from studies of fully covered nitinol self-expanding plastic in adults, but other nonmetal stents continue to be developed. The use of expandable uncovered metal stents for benign esophageal strictures is not recommended, because of a high incidence of complications and difficulty in removal. The use of biodegradable stents has been reported in adults with caustic ingestion but has not been studied in children [110]. (See "Endoscopic interventions for nonmalignant esophageal strictures in adults".)

Patients with progressive esophageal obstruction despite repeated dilation may ultimately require esophagectomy with colon interposition, sometimes within two years following the ingestion [5,32]. Although colon interposition can be successful, many patients have persistent difficulty swallowing. Complications of colon interposition surgery include anastomotic strictures that may require periodic dilation and gastrocolic ulceration caused by reflux of gastric contents into the neoesophagus [111,112].

Carcinoma — Esophageal squamous cell carcinoma develops in approximately 2 percent of individuals with severe caustic esophageal injury. Those who have had colonic interposition surgery face a risk of carcinoma in the bypassed esophagus, with a reported incidence of 2 to 8 percent [56]. As a result, some authors now recommend resection of the esophagus at the time of bypass with colon interposition.

In one study of 63 patients with lye ingestion, the mean latency period for development of esophageal cancer was 41 years with a range of 13 to 71 years [113]. For those who have had colonic interposition surgery, the time interval between ingestion and the detection of tumor has ranged from 16 to 42 years [56]. Based on these and other data, the American Society for Gastrointestinal Endoscopy recommends beginning endoscopic surveillance 15 to 20 years after the caustic ingestion.

Any patient with a history of caustic ingestion and late onset or worsening of dysphagia should be evaluated promptly, and carcinoma should be suspected until proved otherwise. Since carcinomas tend to occur in strictured areas, they are often resectable. (See "Caustic esophageal injury in adults", section on 'Complications'.)

Pyloric stenosis — Pyloric stenosis can occur with both acids and alkalis and often is associated with esophageal injury and strictures [26,114]. With severe injury to the stomach, gastric outlet obstruction may occur as early as 3 weeks or as late as 10 weeks [26,115]. Surgical bypass may be necessary, but endoscopic balloon dilation also has been used successfully in case reports [116,117].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Esophageal strictures, foreign bodies, and caustic injury".)

SUMMARY AND RECOMMENDATIONS

●Epidemiology and pathophysiology – Esophageal burns account for most of the serious injuries and chronic complications of caustic ingestion. Injury to the lips, oropharynx, upper airway, and stomach also may occur. (See 'Epidemiology' above and 'Esophageal injury' above.)

Following the initial necrosis, additional destruction takes place over the first week (table 1). The risk for esophageal perforation peaks at this stage. (See 'Delayed injury' above.)

●Risk by type of caustic substance

•The most common causes of caustic esophageal burns are alkaline household cleaning products, such as oven and drain cleaners, strong lyes, and concentrated laundry detergents and cleaning agents. Esophageal burns are rare with household bleaches.

•Some caustic ingestions are caused by acid household products including toilet bowl cleaners, battery fluids, and muriatic (hydrochloric) acid used in swimming pools. Esophageal injury from acids may be attenuated compared with alkalis.

(See 'Acid versus alkali' above and 'Risks of specific substances' above.)

●Clinical manifestations – The most common symptom of caustic ingestion is dysphagia, which can occur even without severe esophageal injury. Patients may also present with drooling, retrosternal or abdominal pain, hematemesis, and features suggesting upper airway injury. The presence or absence of any of the clinical features described above does not reliably predict ingestion or the presence or severity of esophageal or gastric burns. (See 'Clinical manifestations' above.)

●Stabilization and supportive care – Initial management involves supportive care and close observation, with an emphasis on preventing vomiting, choking, and aspiration. Induction of vomiting is contraindicated because vomiting may lead to additional esophageal injury if gastric contents come in contact with the esophageal mucosa. Similarly, attempts to dilute or neutralize the caustic agent, administration of activated charcoal, or gastric lavage should not be performed. (See 'Stabilization and supportive care' above.)

●Subsequent management – Subsequent management depends on the severity of the ingestion (algorithm 1):

•Asymptomatic patients with low-risk ingestion – Patients who are asymptomatic (no evidence of oral lesions, and no dysphagia, vomiting, or other symptoms), should be observed for several hours to monitor fluid intake and overall status.

If the patient remains asymptomatic with normal swallowing and if the suspected substance is known to be of low causticity (eg, household bleach), or if it is unclear from the history whether the patient ingested a caustic substance, then the patient is at low risk for complications and endoscopy generally is not necessary. (See 'Asymptomatic patients' above.)

•Symptomatic patients or high-risk ingestion – Upper endoscopy should be performed in all symptomatic patients, those who have oral burns, and those who are known to have ingested a substance with a high risk of esophageal injury. For patients with alkali ingestion and minimal symptoms (eg, vomiting or drooling alone), it is reasonable to observe overnight and proceed to endoscopy only if symptoms persist (algorithm 1). The endoscopy ideally should be performed within 24 hours of the ingestion.

The endoscopy evaluates and stages the esophageal injury (table 2), which guides subsequent management (see 'Endoscopic evaluation' above):

-All symptomatic patients – For all symptomatic patients, we suggest treatment with a proton pump inhibitor (PPI) for approximately one week (Grade 2C). (See 'Medications' above.)

-Grade 1 and 2A burns – Management of patients with grade 1 and 2A burns is generally supportive. We suggest not treating such patients with corticosteroids (Grade 2C). Patients with grade 1 and 2A burns have a low risk for strictures and thus are unlikely to benefit from corticosteroids. (See 'Medications' above.)

-Grade 2B and 3 burns – For patients with grade 2B and 3 burns, management includes placement of a nasogastric tube (NGT), which should be performed under direct visualization during the endoscopic procedure. In addition, for most patients with grade 2B or 3 burns, we suggest prophylactic antibiotic therapy (typically with a third-generation cephalosporin) for approximately one week (Grade 2C). For patients with 2B lesions without signs of esophageal perforation, we suggest a brief course of high-dose corticosteroids (Grade 2B). We avoid corticosteroids in patients with grade 3 lesions or those with suspected perforation because they may soften the wound and increase the risk and/or mask symptoms of perforation. (See 'Nasogastric tube' above and 'Medications' above.)

-Follow-up imaging – All patients with significant esophageal burns (eg, grade 2A and higher) noted on endoscopy, or those with persistent dysphagia, should be evaluated with barium contrast studies two to three weeks post-ingestion to assess for stricture formation. (See 'Stricture formation' above.)

●Late complications – Stricture formation is the primary complication of caustic injury. Most patients who develop strictures present with obstructive symptoms within two months. The presence of a stricture is confirmed with radiographic contrast studies. If strictures develop, esophageal dilation is performed to maintain or reestablish normal swallowing. (See 'Stricture formation' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges George D Ferry, MD, who contributed to earlier versions of this topic review.