Dysphagia attributable to benign esophageal strictures is an important cause of morbidity and death in veterinary medicine.[1-6] These strictures most commonly occur as a result of esophagitis secondary to peri-anesthetic gastroesophageal reflux. The impact of benign esophageal strictures on pets and their owners is considerable as treatment can be costly and is often associated with a poor prognosis. The goal of treating esophageal strictures is to restore normal esophageal function; however, in most animals, this is not a clinically realistic outcome. A more practical expectation with treatment is to reduce the frequency of regurgitation such that oral feeding can be used to maintain adequate nutrition.
Conventional treatment for benign esophageal strictures includes medical management, bougienage, or balloon dilatation. Medical management is aimed at controlling esophageal reflux and subsequent mucosal damage; however, resolution of strictures with medical management alone is unlikely. A subjectively good outcome has been reported after balloon dilatation or bougienage (70–88%) for benign esophageal strictures in dogs and cats.[2-6] Some degree of stricture often remains with only 14–25% of animals regaining the ability to eat dry food without occasional regurgitation.[2-6] A subset of strictures recur despite multiple attempts at dilatation. Strictures are considered refractory benign esophageal strictures (RBES) if they fail to maintain a satisfactory luminal diameter despite greater than 3 dilatation sessions.[7, 8]
Indwelling esophageal stents have been used in humans for the palliation of dysphagia resulting from RBES.[9, 10] Gentle dilatation through gradual expansion of the stent causes less trauma to the stricture and the continual presence reduces the risk of restenosis.[9, 10] Furthermore, placement of stents might avoid the serious complications of acute dilatation therapy such as iatrogenic mucosal trauma or esophageal perforation.[2-6] Although long-term clinical resolution of strictures only occurs in 50% of human patients, stenting can offer patients with RBES improved palliation.[9, 10]
The use of esophageal stents has been recently described in separate reports encompassing 3 dogs, 2 cats, and a ferret with benign esophageal strictures.1,[11-13] The purpose of this study was to describe a minimally invasive, potentially long-term treatment option for dogs with RBES. Esophageal stenting was offered as a salvage option for dogs that had already undergone multiple dilatation procedures without a satisfactory outcome. The authors hypothesized that esophageal stent placement would be a safe and effective technique for the management of RBES and would be associated with low morbidity.
- Top of page
- Materials and Methods
Nine dogs met the criteria for inclusion and no dogs were excluded during the study period. The study cohort included 6 castrated male dogs, 2 spayed female dogs, and 1 intact female dog. Breeds represented included 3 mixed breed dogs, and one of each of the following: Boxer, Golden Retriever, Labrador Retriever, Pug, Rottweiler, and Vizsla. Median age at the time of evaluation was 4 years (mean 5, range 1–9 years). Median body weight of dogs was 30 kg (mean 26.8, range 9.3–45 kg).
All dogs developed clinical signs associated with esophageal stricture within 1–4 days of receiving general anesthesia. Two dogs had possible trauma to the esophagus that might or might not have been related to the stricture including cervical neck laceration repair and ventral slot procedure. Clinical signs included regurgitation (9/9), vomiting (4/9), anorexia (3/9), ptyalism (3/9), diarrhea (2/9), weight loss (2/9), hematemesis (1/9), pyrexia (1/9), and gastric dilatation (1/9).
Initial medical management of suspected esophageal dysfunction included H-2 receptor antagonists or proton pump inhibitor drugs (9/9), promotility agents (4/9), sucralfate (3/9), and corticosteroids (2/9). before esophageal stent placement, dogs underwent esophageal dilatation procedures a median of 10 times (mean 10.2, range 4–24 times) over a median of 2 months (mean 2.97, range 0.75–8 months). Three dogs had intralesional triamcinolone8 injections performed during dilatations, and 2 dogs had topical mitomycin-C therapy in the region of the stricture. Immediately before the stent placement, 3 dogs (3/9) had a dysphagia score of 4, 4 dogs (4/9) had a score between 3 and 4, 1 dog (1/9) had a score of 3, and 1 dog (1/9) had a score of 2, although dogs were routinely receiving dilatation procedures during this time.
All dogs received diagnostic esophagoscopy at the time of stent placement to confirm the presence and location of the stricture. Biopsies of the esophagus were not obtained in any of the dogs. Either single (6/9 dogs) or multiple (3/9 dogs) esophageal strictures were observed with a total of 14 separate strictures documented. Single stricture locations included caudal intrathoracic (3/14), cranial intrathoracic (2/14), or cervical (1/14) areas. Two dogs had multiple cranial intrathoracic strictures, and 1 dog had 1 cranial and 1 caudal intrathoracic stricture.
Esophageal stents were chosen based on availability and size needed for each dog. Substantial dilatation before stent placement (more than simply allowing the endoscope to pass) using a balloon catheter4 was performed in 3 dogs. Nine dogs initially had 10 esophageal stents placed. One dog required 2 stents simultaneously because of multiple strictures in different locations (1 cranial and 1 caudal intrathoracic stricture). Stents used included uncovered biodegradable polydioxanone stents9 (6/10), polyurethane covered mesh nitinol stents10 (3/10), and a silicone covered plastic stent11 (1/10). Seven dogs (7/9) had 8 stents sutured or anchored in place (3/4 covered stents and 5/6 uncovered stents) and 2 dogs did not (2/9). Two dogs had gastrostomy tubes placed and 1 dog had an esophagostomy tube placed. All dogs survived the procedures without intraoperative or anesthetic complications.
Postoperatively, all dogs received medications including, but not limited to, omeprazole12(1 mg/kg PO Q24h), metoclopramide (0.5 mg/kg PO Q8h), and tramadol (2–4 mg/kg PO Q8h). Antiemetics (ondansetron13 2–4 mg/kg PO q24h or maropitant citrate14 2 mg/kg PO Q24h) were prescribed in 3/9 dogs and antibiotics (ampicillin/clavulanate15 13.75 mg/kg PO Q12h) were prescribed in 2/9 dogs. After stent placement, all dogs were fed a high-calorie gruel or wet food,1617 until discharge, after which owners were instructed to gradually increase the consistency of the diet. Four dogs were supplemented via feeding tube.
Complications occurred <7 days after stent placement in 7 dogs (7/9), 2 (2/7) of which required repeat procedures. Complications included severe ptyalism, nausea, gagging, and regurgitation (4/7), regurgitation only (2/7), ptyalism only (1/7), covered SEMS migration (nonsutured) into stomach requiring removal (1/7), covered SEMS (sutured) caudal stent shortening requiring repositioning (1/7), and incisional infection (1/7). In the dog that experienced stent migration into the stomach 2 hours after placement, endoscopy was performed to remove the stent. In the dog that experienced stent shortening, the distal portion of the stricture was no longer engaged by the stent. The stent was removed, placed back onto a delivery system, repositioned more caudally, and resutured.
All nine dogs were available for short-term (7–30 days poststent placement) follow-up evaluation. Short-term clinical dysphagia scores were improved in all 9 dogs (9/9) as compared with preoperative scores (median improvement 1, mean 1.3). Short-term complications occurred in 7/9 dogs, 2/7 of which required repeat procedures. Complications included severe ptyalism, regurgitation, gagging, and nausea (4/7), regurgitation alone (2/7), vomiting (1/7), megaesophagus requiring SEMS removal (1/7), and progressive incision infection/drainage (1/7). One dog that experienced severe ptyalism, regurgitation, gagging, and nausea remained hospitalized and balloon dilatations within the stent were performed 6 times over 30 days with no improvement in clinical dysphagia score. The dog that developed megaesophagus aboral to the covered SEMS required endoscopic stent removal. This dog had excessive discomfort from the time of stent placement, which resolved immediately after the stent was removed.
Long-term (>30 days poststent) follow-up information was available for all 9 dogs. Compared with pre-operative score, 6 (6/9) dogs had some degree of improvement in dysphagia score (median improvement of 2.25, mean 2) during the long-term follow-up period. Long-term complications occurred in 8 (8/9) dogs, 7 (7/9) of which required additional procedures. Complications included BDS stent dissolution and restricture (3/8 dogs; 3/6 BDS stents), stricture recurrence after previous stent removal because of migration in 1 dog and intolerance in another (2/8), covered (sutured) SEMS migration with restricture leading to euthanasia (1/8), covered SEPS (sutured) migration requiring removal (1/8), vomiting (1/8), regurgitation which resolved 6 weeks after stent placement (1/8), hyperplastic tissue ingrowth into uncovered (unsutured) BDS with euthanasia (1/8), and tracheoesophageal fistula in a dog with covered (sutured) SEPS leading to euthanasia (1/8). All 3 dogs with a BDS that had dissolution of the stent and restricture within 2–4 months had another stent placed using covered (sutured) SEMS. Two dogs with stent removal before the long-term follow-up period required intermittent balloon dilatation therapy to alleviate dysphagia.
Median final follow-up time for all 9 dogs was 8 months (range 2–48 months). One dog lost to follow-up at 12 months after stent placement required frequent balloon dilatations. Although dysphagia scores improved in 6/9 dogs for at least 30 days, by 6 months, only 3/9 dogs maintained a score of less than 2 without severe stent adverse effects or additional dilatation procedures. In one of these 3 dogs, RBES resolved after the stent migrated, approximately 3 months after stent placement. In the 2nd dog, resolution of clinical signs was achieved with an indwelling stent for 4 years; however; this dog eventually experienced severe complications (stent dissolution, restricture, restenting and subsequent tracheal-esophageal fistula) and was euthanized. The 3rd dog was able to eat semisolid food for 8 months until the stent migrated and esophageal stricture recurred. Median survival times for the remaining 6 dogs euthanized after esophageal stent placement was 383 days (range 60–1440 days). One dog (1/6) was euthanized for an unknown reason 10 months postoperatively and 1 dog (1/6) was euthanized because of pre-existing neoplasia (1/6). The remaining 4 dogs (4/6) were euthanized because of recurrence of clinical signs from strictures or complications associated with esophageal stents. Reasons cited for euthanasia included recurrence of strictures after stent migration (1/6), stent discomfort and intolerance (1/6), hyperplastic tissue ingrowth (1/6), and tracheal–esophageal fistula (1/6).
- Top of page
- Materials and Methods
The results of this study suggest that esophageal stenting is a technically feasible procedure for the treatment of RBES, but is associated with a high complication rate and poor long-term success for the majority of dogs treated. Although all dogs showed some degree of short-term improvement, placement of the stents alone was successful in providing palliation longer than 6 months in only 3 dogs with RBES. Seventy-eight percent (7/9) of dogs in this study experienced 1 or more major complications necessitating intervention, and 50% of dogs in this study (4/8) were ultimately euthanized because of stricture recurrence or stent-related complications. Based on these results, esophageal stenting is not a safe and efficacious procedure for treatment of recurrent benign esophageal strictures.
The use of stents for esophageal disease has been reported in 2 cats,[11, 12] a ferret, and 3 dogs using a variety of stent types.1, A BDS was successfully used in 1 cat with an RBES that resulted in complete resolution of the stricture. In another report, an SEMS was used in a cat with an RBES which improved clinical signs for 12 months until hyperplastic tissue ingrowth and foreign body obstruction occurred. An SEMS was successfully used to palliate esophageal neoplastic obstruction in a dog with relief of clinical signs for at least 12 weeks after placement.
The ideal esophageal stent would be easy to place, resistant to migration, cause minimal tissue reaction, prevent tissue ingrowth, be easily retrievable, and be well tolerated by patients without discomfort or nausea. There are primarily 3 major types of stents available for esophageal use (SEMS, SEPS, BDS), which can be covered or uncovered. Stent designs often include modifications to prevent more common complications of migration or tissue ingrowth. Modifications can include flared or uncovered ends, or stents can be either partially or fully covered with silicone or polyurethane. Uncovered stents are less likely to migrate than covered stents but can be associated with a higher rate of hyperplastic tissue or stricture ingrowth.[9, 10] Perceived advantages of SEPS compared with SEMS include easier retrieval, lower local tissue reaction, softer material so possibly better tolerated, and lower costs; however, they can be associated with higher migration rates.[15, 16] BDS are a relatively new design that is bioabsorbable and can eliminate the need for additional procedures when only temporary placement is desired.[17, 18]
Ancillary procedures performed during esophageal stenting were associated with a low complication rate. Substantial dilatation using a balloon was performed before stent placement in 3 dogs, and all three stents were sutured in place. In humans, dilatation before stent placement has been associated with a higher risk of peri-operative stent migration; however, this was not observed in these dogs, potentially because of tacking suture placement.[9, 10] Antimigrational sutures were placed in 6 dogs, which resulted in one incisional infection as a peri-operative complication. This dog had a recent surgical procedure using a similar approach (ventral slot surgery) and required resuturing after stent shortening, which might have increased the risk for infection. Anchoring of an esophageal stent to an esophagostomy tube was performed in 1 dog that had a caudal intrathoracic stent placed to avoid surgical morbidity associated with accessing the intrathoracic region of the esophagus. This anchoring method was effective in preventing stent migration in this dog; however, position was dependent on the integrity of the esophagostomy tube.
Stent-associated discomfort was a frequently reported complication in the present study. In 7 dogs, some degree of ptyalism, regurgitation, gagging, nausea, or vomiting was noted in the peri-operative period, and this continued into the short-term, and even long-term, periods in some dogs. These signs were often more severe than the clinical signs dogs had before stent placement. In humans, stent discomfort is often characterized by nausea and chest pain, which are thought to be secondary to both esophageal dilatation and the radial force exerted by stent expansion.[9, 10] Globus, or a feeling of a lump in the throat, is also commonly reported in humans with esophageal stents, which might explain some of the discomfort seen in these dogs.[9, 10] Stents with lower expansile strength are thought to cause less esophageal discomfort; however, these stents might not fully expand when the compressive force of the stricture is greater.[9, 10]
Stent migration occurred as a peri-operative and long-term complication in this study. Stent migration is believed to be attributable to a combination of natural peristalsis and lack of tissue interface and traction between the stent and the esophageal wall. Of the 7 dogs that had sutures placed to prevent migration of the stent, two (2/7) had stent migration as a long-term complication at approximately 3 months (covered SEPS) and 8 months (covered SEMS). Of the 2 dogs that did not have sutures placed, 1 dog had migration of a covered SEMS during the peri-operative period (1/3). No dogs with uncovered stents had migration as a complication (although all but one were sutured), and no stents that were sutured had migration as a peri-operative or short-term complication. In this group of dogs, stents were intended to be placed permanently; however, placement of a tacking suture did not appear to prevent migration in the long-term follow-up period. Reasons for this might include pull-through of the suture through the thin-walled esophagus or stent, failure of suture knots, or less likely dissolution of the suture because of exposure to an acidic environment. In human studies, stent migration occurred in approximately 12–67% of patients who received esophageal stents for treatment of RBES.[9, 10] Although migration was not a complication in the 3 dogs that had dilatation before stent placement performed, it has been associated with increased risk of migration in humans and should perhaps only performed if the delivery system cannot be passed through the stricture. Suturing the stent in place might be effective in permitting predilatation in dogs without risk of early stent migration.
Stent shortening was a short-term complication in 1 dog in this study. Stent shortening might occur because of inappropriate stent diameter selection, inappropriate stent placement, postoperative esophageal relaxation, or progressive dilatation of the esophageal stricture postoperatively. Predilatation with a balloon might minimize stent shortening, but if performed, the stent should be sutured in place. Fluoroscopy and endoscopy are recommended by the authors to visualize the extent of the stricture(s), estimate the proper stent size and length, and confirm adequate placement after stent deployment. When shortening occurs, stent replacement or adjustment is often necessary.
Recurrence of RBES was a common complication, occurring in 3 dogs with BDS (3/5 BDS dogs, 3/6 stents) after dissolution and 2 dogs with SEMS (2/3 SEMS dogs) after migration or removal. Ultimately, only 1 BDS stent (the proximal stent in the dog with 2 stents) was successful in providing stricture resolution after stent dissolution. Half the BDS mechanical strength is lost after approximately 3 weeks and stent structure begins to disintegrate within 8 weeks. In the present study, strictures most likely recurred as BDS degraded as clinical signs returned at approximately 2–3 months after stenting was performed. Strictures only recurred in dogs with covered SEMS after stents migrated. It is important to note that typically, the use of BDS in humans is aimed at temporary relief with the expectation that the stricture will resolve, or that the stent will be replaced if the stricture recurs or persists after dissolution. In 1 study, BDS provided comparable success to SEPS for relief of dysphagia; however, sequential stenting was required for long-term palliation and could help avoid serial dilatations.
Hyperplastic ingrowth of tissue into the esophageal stent and recurrent dysphagia was a long-term complication in 1 dog with a BDS in this study. This complication occurred in a cat that received a BDS for a RBES. In humans, this complication is more common with uncovered SEMS, occurring in approximately 30% of patients.[9, 10] The type of metal used in the stents, excessive radial force, or the prolonged presence of the foreign body could predispose to the inflammation and the formation of tissue ingrowth or outgrowth. Covered or partially covered stents are designed to avoid the high incidence of hyperplastic tissue growth seen in uncovered stents.[9, 10] Furthermore, placement of stents across high-motion regions such as the thoracic inlet in the patient in this study could increase the risk of hyperplastic tissue formation because of the constant mobility and friction between the stent and esophageal wall.
Formation of tracheal–esophageal fistulae is a rare complication of esophageal stents in humans, and occurred in one of the dogs in this study. In that dog, a tracheal–esophageal fistula was diagnosed several years after stent placement. Rapidly progressive dysphagia and severe halitosis were noted and the presence of the fistula was diagnosed via combined bronchoscopy and esophagoscopy. In humans, covered tracheal stents can be used to occlude the fistula and lead to rapid clinical resolution. However, even when clinical signs are controlled, these fistulas often continue to persist and outcome is generally poor.
Megaesophagus was a short-term complication in one dog. In that dog, clinical dysphagia score increased and dilatation of the esophagus was noted aboral to the stent on thoracic radiographs. Removal of the stent resolved the clinical signs. Megaesophagus secondary to esophageal stenting for RBES has not been reported as a complication in humans. A possible cause might include esophageal dysmotility secondary to temporary achalasia. The outward radial force of the stent could cause focal neuropraxia within the esophagus. With achalasia, a loss of ganglion cells in the myenteric plexus leads to a lack of contraction within peristaltic muscles and constant contraction of the lower esophageal sphincter. Food retention and subsequent esophageal dilatation can result, leading to the radiographic appearance of megaesophagus. Removal of the stent and resolution of segmental esophageal neuropraxia might explain the resolution of clinical signs.
A major limitation of the present study is its retrospective nature and small population size. This led to considerable variability in pre- and postoperative management. In addition, many different types of stents were used, at many different locations and secured in many different ways. This makes comparisons of treatment very difficult. Selection bias should be considered as a reason for poor outcomes as dogs presented for esophageal stenting as a salvage procedure when the typical alternative was euthanasia. The outcome of these dogs should not be compared with those animals with the more typical (nonrecurrent) strictures receiving standard balloon dilatation therapy. To the authors' knowledge, the rates of recurrence and success of conventional dilatation therapies for treatment of RBES (>3 dilatations) in dogs have not been examined. Survival times for dogs with RBES are highly dependent on many factors, including owner's financial means to continue therapy as well as their assessment of quality of life with the ongoing clinical signs. It is necessary to consider not only patient factors when considering stent choice but also the characteristics of strictures including, but not limited to, their length, circumference, and degree of mucosal inflammation. It is impossible to make definitive conclusions about the ideal candidates or the contraindications for esophageal stenting for RBES on the basis of the low number of dogs in this study; however; these results can provide guidelines for placement when no other options are available. Further understanding of the biologic factors that cause stricture recurrence and the development of medical interventions that target these mechanisms is imperative to the ultimate success of dilatation therapy.
For treatment of RBES requiring stent placement, the authors currently recommend a covered stent with tacking sutures. Placement of a covered stents might help avoid the major complications of hyperplastic tissue ingrowth and would be easier to remove than uncovered stents if not well tolerated by the dog. SEPS might have less discomfort than SEMS, and when BDS are used, endoscopic evaluation 2–3 months after placement is recommended because of the high rate of stricture recurrence after stent dissolution. Advances in stent technology will improve the future outcome when used in dogs with RBES. With further investigation and development, esophageal stenting could become an effective modality in improving dysphagia score and prolonging survival time for some dogs with RBES when owners are unable to continue frequent dilatation therapy or long-term enteral feeding.