Michael Manfredi, Frederick Gottrand, Luigi Dall’Oglio, Mike Thomson, George Gershman, Antonio Quiros, and Thierry Lamireau Congenital esophageal strictures (CES) are caused by a malformation of the esophageal wall leading to narrowing of the esophageal lumen [1,2]. Four types are described: membranous; web or diaphragm (MD); fibromuscular stenosis (FMS); and tracheobronchial remnants (TBR). Congenital stenosis may also be revealed by progressive dysphagia with semisolid foods or even liquids and episodes of food impaction. Symptoms like regurgitation and aspiration are frequent, leading to cough, respiratory distress or choking during feeds. Feeding refusal or apnea can occur in small babies. Contrast esophagogram reveals abrupt stenosis of varying length, mostly located in the distal esophagus. Endoscopy confirms the stenosis, which can be tight, preventing the passage of the endoscope. Thoracic CT can also be helpful. Acquired strictures are mostly secondary to surgical repair of an esophageal atresia, occurring in up to 80% of cases in some series [3,4]. Factors implicated in the pathophysiology of anastomotic strictures include the creation of a long gap esophageal atresia leading to high‐tension esophageal anastomosis, ischemia, two suture layer anastomosis, use of silk suture material, anastomotic leak, and gastroesophageal reflux. Children with long gap esophageal atresia are especially at risk of stricture of the anastomosis and sometimes of the colonic segment used for esophagoplasty. Stricture is usually diagnosed on systematic contrast esophagogram performed in the follow‐up of these children. It is also suspected in case of feeding difficulties occurring sometimes early after surgical repair, with regurgitations and feeding refusal in small babies. It can be associated with aspiration, leading to cough, respiratory distress or choking during feeds. When feeding is impossible, nutrition via a gastrostomy is mandatory. Feeding difficulties are more often revealed at the time of transition to a more solid diet, with regurgitation, dysphagia with semisolid foods, and impaction of solid foods like meat in older children. Contrast esophagogram confirms the stricture at the location of the anastomosis. Caustic ingestion can be responsible for severe burning lesions Children with grade 2 or 3 caustic lesions of the esophagus often develop multiple or long strictures [5]. Ingestion of a button battery can lead to severe ulceration of the esophagus, sometimes leading to stenosis. These strictures are usually diagnosed on systematic contrast esophagogram performed during the follow‐up of these children. This is dealt with in more detail in Chapter 26. Peptic stenosis has become rare since proton pump inhibitors are now largely used in the treatment of gastroesophageal reflux, but it can occur in some children like those with cerebral palsy [6]. Eosinophilic esophagitis can be responsible for esophageal stenosis requiring dilation. Stenosis can also complicate esophageal involvement associated with skin diseases, such as dystrophic epidermolysis bullosa or Stevens–Johnson syndrome. For purposes of characterization and treatment planning, esophageal strictures may be differentiated into two structural types: simple and complex. Strictures that are short (<2 cm), focal, not angulated, and permitting the passage of an endoscope can be labeled as simple strictures. This type of stricture is amenable to through‐the‐scope (TTS) balloon dilation. Simple strictures tend to be due to peptic esophagitis, Schatzki’s ring and esophageal web. Complex strictures are those that are angulated, long (>2 cm), irregular and have a very narrow lumen (Figure 25.1). Complex strictures are more commonly seen in children with caustic ingestions, anastomotic strictures, post radiofrequency ablation and chest radiation therapy. ESPGHAN–ESGE guidelines on diagnostic and therapeutic endoscopy in pediatrics suggest the following definitions of refractory and recurrent esophageal strictures: “inability to successfully remediate the anatomic problem to obtain age‐appropriate feeding possibilities after a maximum of 5 dilation sessions with maximal 4‐week intervals” for refractory strictures and “inability to maintain a satisfactory luminal diameter for 4 weeks once the age‐appropriate feeding diameter has been achieved” for recurrent strictures [7–9]. The evaluation and classification of esophageal stenosis requires a multimodal approach. Contrast radiological studies provide information about the stricture’s shape, location, length, diameter of the residual lumen as well as associated anomalies, while endoscopy clarifies the etiology of the stricture: CES versus peptic, eosinophilic or corrosive esophagitis, and allows instant endoscopic treatment. However, even high‐resolution computed tomography is not sensitive enough for diagnosis of TBR in the esophageal wall, or to differentiate TBR from FMS prior to endoscopic or surgical intervention. With the advent of endoscopic ultrasonography (EUS), definitive diagnosis of FMS and TBR became feasible (Figure 25.2) [10–13]. The resolution of the high‐frequency EUS (15–30 MHz) miniprobe allows discrimination of cartilaginous tissue, fluid‐filled lesions, thickened muscular layer or external “compressing” vascular structures which is very helpful in guiding therapy. In addition, transendoscopic EUS is useful in assessing severity and degree of submucosal scarring in children with eosinophilic esophagitis (EE). It is also useful in selection of patients with EE and significant scarring who may carry an increased risk of perforation after dilation (see Chapter 21). Symptoms like dysphagia, regurgitation, feeding difficulties, and even food impaction may be caused by numerous diseases of the esophagus even in the absence of stenosis. Gastroesophageal reflux is frequently responsible for feeding difficulties. Complications such as peptic esophagitis have become rare and peptic stenosis exceptional since the widespread use of proton pump inhibitors. However, they may be encountered in children with repaired esophageal atresia and cerebral palsy. Eosinophilic esophagitis must be sought in atopic children presenting with persistent manifestations of gastroesophageal reflux, feeding difficulties or food impaction [14]. Endoscopy may show white deposits and a linear pseudotracheal appearance. Diagnosis will be confirmed by the presence of numerous eosinophils in the mucosae. Infectious esophagitis can occur in cases of primary or secondary immunodeficiency syndrome. Dysphagia and feeding difficulties are usually associated with painful deglutition. Endoscopy shows an inflammatory mucosa and biopsies reveal the microorganism. Achalasia is a rare condition which can be difficult to differentiate from stenosis of the distal esophagus. It can be isolated or associated with Sjögren’s syndrome, triple A syndrome (achalasia, alacrimia, and ACTH insensitivity), familial dysautonomia, and Ondine’s syndrome. Children present with progressive dysphagia with solid food or liquids, regurgitation, vomiting, and weight loss. Respiratory symptoms are frequent in small children, with nocturnal cough, chronic bronchopathy, and aspiration pneumonia responsible for delayed diagnosis [15]. Contrast esophagogram reveals a dilated esophagus, with a narrowed termination. Esophageal manometry demonstrates dyskinesia or akinesia of the corpus esophagus and increased pressure of the lesser esophageal sphincter associated with a fault in relaxation (see Chapter 27). Dysmotility of the esophagus is constant in children with repaired esophageal atresia, and can be responsible for dysphagia and food impaction in the absence of anastomotic stenosis. Other motor dysfunctions of the esophagus can occur in rare conditions, such as diabetic neuropathy, scleroderma, and infection with Trypanosoma cruzei. Tumors of the esophagus, such as leiomyoma, are exceptional in children. There are two types of bougie dilators: guidewire and nonguidewire types. The guidewire ones are tapered cylindrical flexible tubes made of polyvinyl chloride with a central channel to accommodate a guidewire. These dilators have variable length of tapered portion and also radiopaque markers for fluoroscopic guidance (e.g., Savary‐Gilliard® dilators, Cook Medical, Eder‐Puestow, or American Dilators and SafeGuide®). The nonguidewire tungsten dilators weighted for gravity assistance are inserted when patients are in an upright position. The two commonly used nonguidewire bougie dilators are Hurst and Maloney dilators. The Hurst dilators have a rounded blunt tip, while the Maloney dilators have a tapered tip. Both dilators were designed for self‐dilation at home. Another type of mechanical dilator is the Tucker dilator designed for patients with gastrostomy. Tucker dilators can remain inside the patient for periodic serial dilations [9,16–21]. The basic technique of mechanical dilation involves the passage of a bougie across the stricture. This results in both longitudinal shearing force as well as radial force on the stricture area. The goal of mechanical dilation is to pass serial bougie dilators of incremental size across the stricture site. Although fluoroscopy can be helpful for confirming the position and progression of the bougie dilator, it is not mandatory. It is generally recommended to use fluoroscopy in complex strictures. Mechanical dilation requires training the appreciate the feel of resistance while advancing the bougie across the stricture site. The goal is to feel and then overcome the resistance from the stricture with the minimal force. Once this goal is achieved, it is generally recommended to pass no more than three consecutive dilators in increments of 1 mm in a single session for a total of 3 mm. This approach, known as the “rule of 3,” is well established for mechanical dilation. Balloon dilators deliver equal radial force across the entire length of the stricture. They are designed to pass through the endoscope with or without a guidewire. Through‐the‐scope dilation allows the endoscopist to directly visualize the stricture during and immediately after dilation. However, through‐the‐scope balloon dilation requires endoscopes with a working channel of ≥2.8 mm. This can be problematic for infants under 10 kg. In this case, a 0.035 mm or smaller guidewire can be passed into the stomach either through the biopsy channel, using the “exchange” technique leaving the wire in place as the scope is removed, or under fluoroscopic guidance. The balloon is then passed over the wire. The first step of balloon dilation is estimation of the stricture size. The golden rule of the procedure is application of the balloons by 1 mm increments up to 3 mm above the diameter of the stricture. It is also interpreted as three times the diameter of the stricture, such that a 3 mm stricture should not be dilated more than 9 mm, etc. The middle portion of the balloon is centered across the stricture with either endoscopic and/or fluoroscopic guidance. Introducing into the balloon dilator water‐soluble contrast and the use of fluoroscopy allow one to observe the disappearance of the waist of the stricture which is a sign of successful dilation (Figure 25.3). In the setting of a complex stricture, fluoroscopy is useful in advancing the wire and balloon safely across the stricture. An immediate esophagogram under fluoroscopy post dilation can be useful for detecting and managing an early esophageal leak [9,16–21]. Despite more than half a century of endoscopic steroid injection as adjunct therapy to balloon dilation of esophageal strictures, the available data and recommendations remain controversial. Potential benefits have been linked to several biological properties of steroids: interference with collagen synthesis, fibrosis, and chronic scarring processes [22,23]. The preferred steroid for intralesional injection is triamcinolone acetate or acetonide (10–40 mg/mL solution). Betamethasone and dexamethasone preparations have also been used. The technique of intralesional injection is not standardized, especially in pediatrics. Currently, ESGE–ESPGHAN guidelines for pediatric gastrointestinal endoscopy do not support routine use of intralesional steroid injection for refractory esophageal stenosis in children [9]. Mitomycin C, an antifibrotic, has been proposed as an adjunct treatment to manage esophageal strictures [24–29]. Mitomycin C has mainly been used topically but there are also reports of injection of mitomycin C [27]. There are numerous methods of topical application of mitomycin C: soaking pledgets or cotton swab placement on the stricture area, dripping via an injection needle onto the affected area, and using a spray catheter. The dose of mitomycin C used in these studies is also variable, ranging from 0.004 mg/mL to 1 mg/mL. We use 0.5–1.0 mg/mL on 1–3 separate occasions and shield the normal tissue from the soaked cotton swab using a plastic sheath over the tip of the endoscope whilst holding the swab in place with biopsy forceps [24,25]. The effect of monotherapy with mitomycin C in comparison with dilation has been assessed in 30 children with caustic strictures [23]. Although this study was not blinded and does not appear to be randomized, there was a statistically significant improvement in dysphagia score (p = 0.005), and increased median interval between dilations in mitomycin C group [26]. In another retrospective study, the efficacy of topical application of mitomycin C as adjunct therapy to endoscopic dilation in preventing recurrence of anastomotic strictures after surgical repair was analyzed in 21 children with esophageal atresia (EA). Eleven children received mitomycin C concurrently with endoscopic dilations [29]. The authors documented no benefit of adding mitomycin C treatment in the resolution of the stricture compared with repeated esophageal dilations alone in historical controls. There is a hypothetical risk of secondary malignancy with mitomycin C, therefore long‐term follow‐up with esophageal biopsies at the site of mitomycin C application should be recommended. Endoscopic electrocautery incisional therapy (EIT) has been reported as an alternative treatment for refractory strictures in a small number of adult series. There are variations in the EIT technique reported in the literature. The technique involves the use of a needle knife to make incisions into a stricture at its densest points. Typically, multiple radial incisions are made around the stricture site (Figure 25.4). An electrosurgical generator applies a cut current to make the incision. One approach is to make several incisions in the stricture site followed up with balloon dilation. The balloon allows for preferential tearing along the incision site. A large pediatric experience looking at EIT in the treatment of refractory anastomotic strictures reported a 61% success rate. In this study of 36 refractory strictures, the median number of dilations prior to EIT was eight. This group had a median of two dilations in the two years post EIT. In this same study, EIT was performed on 22 nonrefractory strictures with 100% success. The median number of dilations prior to EIT was three and the median number of dilations in the two years after EIT was one. In this study the severe adverse event rate was 2.3% with three perforations [30]. Electrocautery incisional therapy shows promise as a treatment option for pediatric refractory strictures and may be considered prior to surgical resection even in severe cases. The complication rate, albeit low, is significant and EIT should only be considered in experienced hands with surgical consultation. Further prospective longitudinal studies are needed to validate this treatment. Endoscopic dilations are the mainstay of the conservative approach for esophageal strictures of any etiology. However, some patients may experience recurrent or refractory esophageal stricture despite multiple dilation sessions and therefore further treatments are needed [9,31–38]. Even though primarily used as a palliative procedure to relieve dysphagia associated with malignant esophageal strictures, temporary stent placement has increasingly been used as a conservative treatment for refractory and recurrent benign esophageal strictures in adults – and recently in children (Figure 25.5). To date, no evidence‐based indications exist about the timing of stent placement, but most experts agree that stent placement should be considered when other treatment options have failed [9]. The rationale for esophageal stenting for recalcitrant strictures is that continuous radially oriented pressure over a long period allows the esophagus to maintain luminal patency while simultaneously stretching the stricture. Remodeling of scar tissue may occur while the stent is in place, which can result in persistent luminal patency and reduced risk of recurrent stricture formation. Multiple types of stents are commercially available from various manufacturers. These devices differ in material (metal, plastic or biodegradable polymer), luminal diameter, and flexibility and have the same conceptual design: exert centrifugal force on the esophageal wall while allowing the passage of food into the stent lumen. Unfortunately, appropriately sized esophageal stents designed for small pediatric patients are lacking, therefore airway or biliary stents use has been reported in children [31]. Currently, three main categories of esophageal stents are commercially available: self‐expandable metal stents (SEMSs), self‐expandable plastic stents (SEPSs), and biodegradable stents (BDSs). Self‐expandable metal stents consist of woven, knitted, or laser‐cut metal mesh cylinders that exert self‐expansive forces until their maximum fixed diameter is reached. All SEMSs are made of nitinol, a nickel and titanium alloy that has unique properties of shape memory and superelasticity. To prevent tissue ingrowth through the stent mesh, SEMSs can be fully or partially covered by a plastic membrane or silicone. In fully covered SEMs (FCSEMSs), the entire length of the stent is covered while in partially covered SEMs (PCSEMSs), the proximal and distal ends of the stent are devoid of covering. Partial or fully covered SEMSs are currently recommended for palliation of malignant dysphagia; only fully covered stent designs can safely be removed after a prolonged time of stenting [38]. Self‐expandable metal stents can be placed under endoscopic and/or fluoroscopic guidance. After endoscopic and/or fluoroscopic assessment of length and degree of stricture, a guidewire is passed endoscopically, the SEMS delivery system is advanced over the wire and then deployed across the stricture under continuous fluoroscopic monitoring. To guide accurate stent placement, the proximal and distal ends of the stenotic tract should be marked appropriately with radiopaque markers (metal or plastic radiopaque bands placed on the patient’s skin, hemoclips deployed endoscopically, or submucosal injection of a radiopaque agent). To decrease the risk of migration, stent length should be 4 cm longer than the stricture. The ESGE suggests that FCSEMSs should be preferred over PCSEMSs for the treatment of refractory benign esophageal stricture in adults, because of their lack of embedment and ease of removability. Indeed, hyperplastic tissue reaction of the esophageal mucosa to the bare metal mesh may preclude safe stent removal. Self‐expandable plastic stents are constituted by a woven polyester skeleton completely covered with a silicone membrane. Radiopaque markers are positioned at the middle and ends of the stent to assist fluoroscopy‐guided delivery. To reduce the risk of migration, SEPSs are designed to have a wider proximal flange. Compared to SEMSs, complete silicone coating prevents granulation tissue growing through SEPSs, allowing for easier removal even after they have been in place for several months [9,38]. Biodegradable stents are made of biomaterial composed of synthetic polymers (generally polydioxanone) that are free of toxic effects (including carcinogenicity, immunogenicity, teratogenicity), are completely degraded or absorbed by the body and hence do not have to be removed. The polydioxanone BDS lasts for about three weeks and begins to degrade in 11–12 weeks. As for SEPS, BDSs are radiolucent but have radiopaque markers at both ends and in the middle and need to be assembled and loaded onto the delivery system prior to insertion. Placement technique is similar to that previously described for SEMSs and SEPSs. The theoretical advantage over SEMS and SEPS is that BDSs do not require removal. The Dynamic Stent® (DS) consists of a silicone tube of varying sizes coaxially built over a nasogastric tube. The main difference with previously described stents is that instead of passing within the lumen of the stent, foods pass between the stent and the esophageal wall. This new concept in stent design derives from the thirty years’ experience of the inventors. Since 1988, the Digestive Surgery and Endoscopy Unit of Bambino Gesù Children’s Hospital in Rome has produced the DS manually, using silicone tubes of increasing size coaxially overlapped on each other to fashion the stent until the desired size is reached (7, 9 or 12.7 mm outer diameter) (Figure 25.6). To ensure correct positioning of the stent over the stricture, the device is mounted on a nasogastric tube (Figure 25.7). The DS is currently undergoing the European CE medical device approval process; a commercial version will be available in the market in the next few years [36]. The placement technique is comparable to that of previous stents. The stent, customized according to stricture size, is inserted through the mouth and advanced over a guidewire under endoscopic and fluoroscopic guidance through the stricture (skin markers and radiopaque markers at both ends of the stent are used as reference point). The nasogastric tube of the stent is passed with a backward movement through the nasopharynx and nose, and then fixed externally. The DS should remain in place for at least eight weeks, but according to patient tolerability, longer durations (up to six months) are advocated to achieve better results. Indeed, passing the food between the esophageal wall and the stent, there is no risk of tissue reaction and stent embedment [9,31–38]. The ESGE clinical guideline for esophageal stenting in adults suggests consideration of temporary stent placement for refractory benign esophageal strictures. Although only SEPSs have received formal approval for this indication in adults, the guideline does not recommend a specific type of expandable stent (SEMS, SEPS, BDS) because none has been shown to be superior to any other. However, the ESGE suggests that FCSEMSs be preferred over PCSEMSs because of their lack of embedment and ease of removability. Moreover, although no studies have compared different strategies in terms of stenting duration, the ESGE guideline suggests that stents should remain in place for at least 6–8 weeks and no more than 12 weeks, to maximize success and minimize the risk of hyperplastic tissue reaction and stent embedment. Overall, pediatric data on stricture resolution are scarce and heterogeneous, reported success rates ranging from 26% to 86%. In a study by Manfredi et al., 24 children with esophageal atresia underwent a total of 41 stent placement (SEPSs 14, FCSEMSs 27), with a success rate of 39% and 26% at ≥30 and ≥90 days after stent removal, respectively. The mean duration of stent placement was 9.7 days (range 2–30 days). In this study, the authors reported that esophageal stenting is a safe and effective approach in closing esophageal perforations, especially post dilation [30]. In a series of predominantly small children (median age 1 year old), Best et al. reported that esophageal stenting using an airway FCSEMS was successful in all seven patients (five esophageal atresia, one battery ingestion and one major congenital cardiac anomaly) [37]. Lange et al. published their experience with FCSEMS (biliary, bronchial and colonic stents) use in 11 children with esophageal strictures of different etiologies. Median duration of stent placement was 29 days (range 17–91 days). Six children (55%) were successfully treated without any further intervention, two needed one single dilation after stent removal, while two did not improve and required surgery [38]. Recent ESPGHAN‐ESGE guidelines for pediatric gastrointestinal endoscopy suggest temporary stent placement (or application of topical mitomycin C) for refractory esophageal stenosis in children [9]. In relation to the Dynamic Stent, the group at the Bambino Gesù Children’s Hospital reported an overall success rate of 89% in a series of 79 children, mostly with caustic strictures. High‐dose systemic steroid therapy (dexamethasone 2 mg/kg/day for three days) was administered in all children after stent placement [36].
25
Endoscopic management of esophageal strictures
Stricture presentation
Classification
Diagnosis
Differential diagnosis
Treatment
Bougie dilation
Balloon dilation
Adjunct therapies
Intralesional steroid injection
Mitomycin C
Incisional therapy
Esophageal stenting
Self‐expandable metal stents
Self‐expandable plastic stents
Biodegradable stents
Dynamic Stent
Outcome