In most teams, caustic ingestions are routinely given antibiotics, steroids, and H2 blockers (PPI) [2–4].

Antibiotics seem useless to prevent bacterial colonization of the esophageal lesions as it occurs in a devascularized tissue where microcirculation has been destroyed, so their routine use is debatable. They are indicated in case of perforation and respiratory involvement. Riffat suggests that there is evidence of a lower rate of stricture formation with the use of antibiotics: by decreasing bacterial counts in the necrotic tissue, superinfection is reduced which may lessen the stricture formation [2]. Occasionally, we have observed a peak of fever after dilatation in some children. As cerebral abscesses have been reported in such circumstances, those cases received a prophylactic dose of antibiotics before each dilatation without recurrence of fever [29].

The beneficial role of steroid on inflammation and scarring process is still debated. They could be used in first- and second-degree injuries but not in third degrees because of the potential increased risk of perforation. Some cases are reported who developed a gastric ulcer with associated hemorrhage after receiving systemic steroids [18]. A meta-analysis of 13 studies done by Fulton over 50 years has concluded that steroid use does not decrease the incidence of stricture formation following Grade 2 caustic ingestion, and therefore, the use of steroids was not advised [30]. Some reports, more specifically concerning infants and toddlers, have shown that the use of high dosage of corticosteroids, starting at the early phase of treatment, could be beneficial in decreasing the need for dilatations [31–33]. Our experience is that corticoids do not prevent from esophageal stenosis in serious caustic burns, but are helpful to achieve faster resolution of the edema, mainly on the airways.

The caustic burn induces a shortening of the esophagus and a motility disorder resulting in reflux with poor esophageal clearance, which adds a peptic stenosis to a caustic one as evidenced by histology (O. Reinberg, unpublished). For this reason, as many others, all our patients under conservative treatment with dilatations receive proton pump inhibitors (PPI) even if their efficiency has not been proven [2–4, 18].

Benign esophageal strictures usually produce dysphagia for solids, liquids, or both, with slow and insidious progression of weight loss and malnutrition. If the stenosis is important with subsequent dysphagia lasting for more than a month, a gastrostomy should be done to avoid long-lasting total parenteral nutrition with its potential complications. Most patients referred to us, even those with a previously done gastrostomy, were in poor nutritional conditions and must be placed under refeeding program before surgery.

Percutaneous endoscopic gastrostomy (PEG) is our favorite technique for feeding tube placement in children with inadequate nutritional intake. However, it is not feasible after caustic ingestion and a gastrostomy has to be performed. As most of our patients are in poor conditions, they require a gastrostomy including a gastropexy to avoid parietal disunion related to malnutrition. We have described a technique of a real Stamm gastrostomy performed by laparoscopy for these cases [34]. This laparoscopic technique combines the advantages of a minimal invasive procedure with the safety of an open operation and related gastric attachments to the abdominal wall.

The proper placement of the gastrostomy on the anterior stomach wall is a major concern. When intending to replace an esophagus, the surgeon never knows which transplant can be used: if the gastrostomy has been placed too close from the greater curvature, he may face an interruption of the gastroepiploic artery and a gastric tube cannot be achieved. When performing a gastrostomy for caustic stenosis, it is wise to place it far away from the great curvature, just in case a tube could be done.

Gastrostomies are our first choices of surgical access to the bowel, better than jejunostomies. However, if the stomach has been involved in the caustic injury, we must refrain from using it and then perform a jejunostomy.

In some cases, we used an interesting artifice suggested in 1974 by Papahagi and Popovici: when performing the gastrostomy, these authors ligated the middle colonic artery and sometimes the right one to stimulate the development of the left one, anticipating a transverse isoperistaltic colonic replacement [35].

About a month after caustic ingestion, once the edema has gone, the diagnosis of stenosis can be assessed by an esophagogram and an endoscopy. Then, according to the severity of the stenosis, a dilatation program can be started. The rate of stricture formation reported in literature varies from 2 to 63 % (!). Isolated short stenosis of the esophagus, i.e., 1–2 cm, can be treated by dilatations with good results. Long ones (more than 3 cm), multiple stenosis (more than two), or those with a tracheoesophageal fistula cannot be solved by dilatations and require an esophageal replacement [36, 37]. However, the decision should not be precipitated.

Of the various methods to dilate, we use three of them: the Tucker-Rehbein bougie on a never-ending loop, the Savary-Gillard bougie on an atraumatic guidewire (M. Savary was our Chief of ENT in Lausanne, Switzerland), or the balloon dilators similar to angioplasty. The Tucker-Rehbein bougie has the advantage of being done without endoscopy or chest X-rays control, with a very low risk of perforation. It requires placing a string from the nose, down into the esophageal lumen, and externalized through the gastrostomy. It can be used both ways, antegrade or retrograde. To dilate, the Tucker-Rehbein bougie is tied to either ends of the string and pushed or pulled using progressively larger dilators. When using balloon dilatators, a radial pressure on the stricture is performed that is thought to be better than a longitudinal direction of dilatations as done with the other methods. But balloon dilatation is not as safe as described, as it can be difficult to control the strength of expansion when the balloon inflates suddenly. For this reason the Savary or the Tucker-Rehbein’s techniques are softer and more progressive. Our belief is that all different techniques should be available in a team caring with caustic burns and be adapted to each case and dilatation.

The optimal frequency of dilatation is not well established in the literature, and our practice was to use a symptom-based approach, but an interval of 3 weeks seems appropriate in most cases. We encourage normal eating as soon as possible, as pieces are good self-dilatators, but with a high risk of entrapment. The scarring process of the esophagus is long, and the evolution of a stenosis must be confirmed by repeated esophagograms. An important apparent stenosis related to the inflammatory process can last for months before its disappearance. On the other hand, a dilatation program without significant improvement after a year can be considered as a failure. For these reasons we do not continue a dilatation program more than 1 year. However, some teams persist in dilating patients for years, up to 15 years [18]. Dilatation should be continued as long as a progressive increase in esophageal diameter is noted, along with the recovery of a normal feeding. Even after, dilatation has to be continued from time to time. Without improvement at 12 months, we consider doing an esophageal replacement. Indications for esophageal replacements and their timing vary widely. As a result, children are often subjected to prolonged courses of dilatations prior to esophageal replacement or, conversely, may be exposed to unnecessary surgery [36]. A strong predictor of poor outcome was the delay from ingestion to beginning of dilatations [36, 38].

The early insertion of stents was first proposed by Salzer in 1920 and later advocated by Fell [39, 40]. Early reports fell into disfavor because the strictures soon reformed after removal of the stent [18]. However, Coln as well as Estrera wrote that after stenting, the frequency of recurrences decreased and the strictures were easier to dilate [23, 41]. The use of a self-expanding esophageal stent for malignant strictures is well documented first on animals and now applied in human cases of malignant and recalcitrant benign strictures [42–45]. It has evolved toward early endoscopic esophageal stenting using removable plastic or metallic self-expandable stent [46], or better using biodegradable stents [47, 48]. These techniques are under evaluation and the new materials available are promising. They could play an interesting role to prevent stenosis; however, many migrations or displacements are described [3, 46]. Recently, Okata published the histology of a removed esophagus after self-expandable biodegradable stenting and was able to compare the histology of the esophageal wall under the stent and at distance from its ends. The resected specimens showed thickened scar formation at the level of the stricture, while the degree of esophageal wall damage, both at the proximal and distal ends of the stricture, was slight [48, 49].

The idea is that the stent prevents the adhesion of the facing surfaces of the esophagus, thus minimizing the stenosis, but they cannot restore the defect of the muscular layers. Even if the lumen remains open, a rigid, nonpropulsive segment of the esophagus will be left. Until now we refused to use them as we have been referred 11 children with major complications after esophageal stenting: migrations in the mediastinum and in the left bronchus and posterior erosion of the trachea; one of them 7 years old having had 42 previous procedures.

Many agents have been tried as adjuvant therapies in order to prevent excessive granulation tissue formation.

Mitomycin C is an antibiotic-cytostatic drug derived from Streptomyces caespitosus similar to antialkylating agents. It inhibits DNA and protein synthesis by inducing cross-linking, thus fibroblast proliferation. It is used in multidrug regimen in oncology for disseminated carcinoma as well as for transitional cell tumor of genitourinary tract, but it has a poor antimitotic effect. However, its properties have led to its use as an agent for reducing scar formation in ophthalmology for the treatment of pterygium surgery since 1963 and of refractory glaucoma since 1983, and it is commonly used today in those fields even in children. It has been used by ENT surgeons for recurrent laryngeal or tracheal stenosis both in adult and children. We first presented the use of mitomycin C in recurrent esophageal strictures in children in 2001 at the 33rd annual meeting of the Canadian Association of Paediatric Surgeons [50], followed by Afzal who published the first pediatric cases [51]. The use of mitomycin C is still limited in this indication with some cases or very small series reported [52–54]. In our team, first we dilate the stenosis with Tucker-Rehbein or Savary bougies, then the mitomycin C is applied by the ENT surgeons through a rigid esophagoscope. Two ml of mitomycin C Kyowa® solution 2 mg/ml are applied for 2 min using a peanut positioned on the area uncovered with mucosa under visual control. We have the experience of 25 pediatric cases treated with 1 to 4 topical application of mitomycin C after dilatation either for recurrent esophageal stenosis or for stenosis of the upper anastomosis after esophageal replacement with a success rate of 82 %. Other authors came to the same conclusions (El-Asmar 2015, 21 children, 86 % success [55]).

Hyaluronic acid is used in many clinical situations as diverse as neurosurgery and wound healing. A study showed that hyaluronic acid treatment could be effective in treating damage and preventing strictures after experimental caustic esophageal burn on rats [56]. Several other different chemical agents (heparin, vitamin E, caffeic acid phenethyl ester, tamoxifen, 5-fluorouracil) have also been used experimentally, but only a few of these have been added to clinical treatments. Most of these agents impair collagen metabolism and inhibit fibroblastic proliferation either by direct or indirect routes. Physical treatments have also been used such as argon plasma coagulation, but they remain anecdotal [57].

Surgical segmental resections followed by end-to-end anastomosis have a very high rate of failure even after adding enlargement procedures. Unlike the resection of a congenital esophageal stenosis where the anastomosis is performed in normal tissue on both sides of the malformation, the resection of a caustic stenosis is always done in an injured pathologic tissue and leads to recurrence of the stenosis as done under tension in a poorly vascularized tissue.

The ingestion of foreign bodies (FBs) is a common problem in infants, but fortunately the majority of them will pass through the digestive tract without any adverse effects. The peak incidence of FB ingestion is between 6 months and 3 years, and coins are the most common with an occurrence of >125,000 ingestions per year (2007) and 20 deaths reported in the United States during a 10-year period [58–60]. It has even been described in neonates (esophageal zipper in a 2 months old baby) [61]. Some kids continue to put unbelievable objects in their mouths after infancy. Coins, toys, crayons, and ballpoint pen caps are most often ingested during the childhood [62, 63]. Food impactions are not as frequent as in adults but not unusual [63].

The dangerous FBs are those who remain entrapped in the esophagus at the level of anatomic narrowings: esophageal omentum, aortic arch and left major bronchus, and above the esophagogastric junction. There are no guidelines available to determine which type of object will pass safely. The size depends on the age of the child. A study done by Tander on 62 ingestions in children tries to correlate the sizes of the FB with the ages: up to 5 years of age entrapments occurred for objects between 17 and 23 mm, and after 5, objects from 23 to 26 mm were involved [64].

A FB impacted in the esophagus leads to a pressure lesion and local necrosis resulting in stenosis or perforation. Once in the stomach, it may pass through the pylorus and be eliminated in the stools. But it can be retained anywhere along the bowel at places of anatomical narrowing or angulation such as duodenojejunal flexure or ileocecal valve causing mechanical obstruction. If the object has irregular or sharp edges, it may lodge anywhere in the GI tract. If the objects are elongated, they can become trapped in the appendix or ileocecal valve.

After ingestion, children can be asymptomatic at the time of presentation. If present, common symptoms include drooling, gagging, dysphagia, odynophagia, decreased appetite, food refusal, neck pain, chest pain, abdominal pain, cough, stridor, wheezing, and respiratory distress. Esophageal FBs often present with respiratory complaints. In most patients physical examination is normal [59, 63].

The priority is to assess the presence of a FB.

Chest radiographs including the neck, and a supine abdominal one, should be obtained to rule out ingestion. Two orthogonal projections are mandatory, because some FBs, especially those of discoid shape, could be shown only in one view [59, 62]. Limited chest radiograph not including the upper thoracic inlet may miss a higher-up foreign body. Radiological visualization depends on radiopacity. Radiograph detects as much as 80 % of all FBs [62]. Objects of metal, except aluminum, most animal bones, and glass are opaque on radiographs. Objects composed of plastic and most fish bones are radiolucent structures, and their diagnosis may be challenging. Careful attention should be placed on the edges of a presumed coin to exclude the double halo typical of a button battery, which may easily be mistaken for a coin. Regarding FBs, such as fish bones, chicken bones, and toothpicks, an X-ray has a sensitivity that ranges from 23 to 55 % for the first two and 9 % for the latter. In case of toothpicks, even other imaging studies have a low sensitivity, 15 % for MDCT and 29 % for ultrasound (US). But they are the methods of choice in the diagnosis of a FB that migrated from the GI tract and retained in the soft tissues [62, 65].

An expert panel from the North American Society for Pediatric Gastroenterology Hepatology and Nutrition (NASPGHAN) was convened and produced the following guidelines for practical clinical approaches to the pediatric patient with a variety of FB ingestions [63].

Symptomatic FBs impacted in the esophagus have to be removed urgently. Asymptomatic FBs in the esophagus should be removed within 24 h to reduce the risk of significant esophageal injury or erosion into neighboring structures.

Once in the stomach, FBs can generally be managed expectantly in asymptomatic patients. Parents should be instructed to monitor the stools for passage of the FBs. X-rays should be obtained every 1–2 weeks until clearance can be documented. If the FB is still retained in the stomach after 2–4 weeks of observation, elective endoscopic removal may be considered. Children with underlying anatomic or surgical changes, such as previous pyloromyotomy, have an increased risk for retained FBs [63, 66, 67]. Just before removal of a retained gastric FB, X-rays should be repeated to make sure that the FB has not passed just before.

The gold standard is endoscopic removal under general anesthesia. Most ingested FBs are best treated with flexible endoscopes, which allow retrograde exploration of the gastric fundus. However, rigid esophagoscopy may be helpful for proximal foreign bodies impacted at the level of the upper esophageal sphincter or hypopharyngeal region. Various retrieval devices are used, including rat-tooth and alligator forceps, polypectomy snares, polyp graspers, Dormier baskets, retrieval nets, magnetic probes, etc. Before endoscopy, practicing grasping test on an object similar to the ingested one may help to determine the most appropriate available retrieval device and in what fashion the object has to be seized.

However, nonendoscopic methods have been successfully used. The very high esophageal FBs can be retrieved with forceps under direct laryngoscopy. We have a good experience with the use of a Foley catheter under fluoroscopic guidance to “sweep” out coins lodged in the esophagus, while the patient is maintained in the prone Trendelenburg position [68–71]. It can be done without anesthesia in selected patients. The positioning of the Foley catheter can be helped with some contrast in the balloon. Esophageal bougienage is another technique that uses a blunt Hurst dilator to push down an esophageal FB into the patient’s stomach [72–74]. It is safe and cost-effective compared with endoscopic removal. The disadvantage is that direct inspection of the esophagus for underlying pathology is not done as well as inability to retrieve the FB, which “falls” in the stomach.