Esophageal foreign body entrapment is fairly common in dogs and occurs occasionally in cats. This difference in frequency has been attributed to the more indiscriminate eating habits of dogs. Studies1–5 have shown that bones are the most common EFB (47% to 100%) in dogs, and other entrapped objects include toys, wooden sticks, fishhooks, and food material.
Clinical signs vary with size, location, and duration of the entrapment. The most common clinical manifestations are hypersalivation, retching, gagging, vomiting, regurgitating, signs of pain, respiratory distress, and restlessness.6,7 Chronically affected dogs may remain bright and alert but lose weight and have periodic bouts of regurgitation and inappetence. Sharp or chronic EFB entrapment can result in esophageal perforation, pneumomediastinum, pneumothorax, mediastinitis, pleuritis, pyothorax, mediastinal abscess, esophageal diverticulum formation, or bronchoesophageal or tracheoesophageal fistulas, with resultant pyrexia, depression, and respiratory distress.4,7,8–11 Respiratory distress may also be associated with aspiration pneumonia or impingement of the EFB on the upper airway.
Small-breed dogs are predisposed to EFB entrapment,4 with terrier breeds overrepresented.4,5,12,13 Many (64%) affected dogs are < 3 years of age.7 No obvious sex predilection has been identified.4,7
Esophageal foreign bodies most commonly lodge at the thoracic inlet, heart base, and caudal esophageal region, where extraesophageal structures restrict esophageal dilation. Retrospective case series4,14 have revealed that EFB entrapments in canine referral patient populations are most commonly located between the heart and diaphragm (58% to 88%) and over the base of the heart (12% to 29%), and less commonly located in the cervical region (13% to 14%). On the other hand, a case series7 revealed that 41% of dogs in a non-referral population had cervical EFB entrapment. This difference in findings between populations suggests that patients with more complicated and distally located EFBs are likely referred to secondary or tertiary centers for treatment. Type of EFB may also influence site of the entrapment, with the pharyngeal portion of the esophagus (34%) and heart base (30%) being the most common sites for fishhooks, and with the thoracic inlet and caudal esophageal region being less common sites (11% and 5%, respectively) in a case series.3
Endoscopic retrieval of the EFB is the recommended initial noninvasive treatment option for affected dogs. Reported success rates for retrieval or pushing of the digestible material into the stomach range from 65% to 92%.2,4,12,15 Balloon catheter–assisted removal, by which the balloon tip is passed distal to the EFB and inflated to allow traction of both the catheter and EFB, has also been described.16 Surgical intervention is recommended in situations in which endoscopic retrieval is unsuccessful or substantial perforation of the esophagus is identified.
The veterinary literature1,2,15,17 indicates complication rates ranging from 15% to 22% for EFB removal from dogs. Reported short-term complications associated with EFB include esophagitis, perforation, aspiration pneumonia, pyothorax, pneumothorax, pneumoperitoneum, pneumomediastinum, mediastinitis, pleuritis, abscess formation, and damage to the adjacent organs.2,8,17–19 Esophageal stricture, diverticulum formation, esophageal fistula, and mediastinal abscess formation are reported longer-term complications.8,13,20 These complications typically require medical or surgical intervention, increase morbidity and mortality rates, and may affect the final outcome. Esophageal stricture is the most commonly encountered longer-term complication.
Despite the use of dilatory procedures, such as ballooning or bougienage and medical management, 10% to 30% of animals with esophageal stricture reportedly die or are euthanatized and approximately 15% to 30% must be maintained on slurry diets.21–25 A negative association between duration of clinical signs and complications has been shown in previous studies.2,15 The objective of the study reported here was to identify factors that could predict the need for surgical intervention, complications, and outcome in dogs with an EFB.
Materials and Methods
Animals
Medical records at the Ontario Veterinary College Health Sciences Center were searched electronically to identify dogs treated for EFB from January 1, 1995, to December 31, 2014. Dogs that passed the EFB without any intervention were excluded.
Data collection
Data were obtained from the records of each dog regarding age, sex, breed, body weight, concurrent or historical medical problems or previously removed EFB, the approximate interval from object ingestion to intervention, type of EFB, type and duration of reported clinical signs, and diagnostic and interventional procedures performed prior to referral. Physical examination findings were also obtained, including rectal temperature and clinical signs at hospital admission as well as results of diagnostic imaging performed at the center and results obtained prior to referral but reviewed by a board-certified radiologist at the center.
Performed therapeutic interventions were classified as endoscopy (esophagoscopy), surgery, or both. Endoscopy, when performed, had involved a flexible gastrointestinal endoscopea appropriate for the size of the dog. In each situation, durations of the anesthetic session and procedure were noted as well as whether antimicrobials had been administered. Intraoperative findings, including visual assessment of the degree of esophageal damage, were recorded when available. When endoscopy had been performed, degree of esophageal wall integrity and evidence of perforation were noted. If radiography had been performed after EFB removal, the findings were recorded. Finally, feeding-tube placement, postoperative medications, feeding method, in-hospital complications, and duration of hospitalization as well as status at the time of discharge from the hospital were recorded.
Follow-up data were obtained retrospectively from medical records and when possible via telephone interviews with owners, the referring veterinarians, or both in 2 stages; for dogs treated from 1995 through 2009, follow-up telephone calls were made in early 2010 and for dogs treated from 2010 through 2014, follow-up calls were made in early 2015. When available, information regarding dysphagia, vomiting, regurgitation, nasal discharge, and surgical wound complications up to the time of follow-up was recorded. Outcome was classified as excellent when no clinical signs attributable to stricture formation were reported, such as the need for a special diet or regurgitation after ingestion of the dog's usual diet.
Statistical analysis
Statistical softwareb was used for all analyses. Pearson χ2 goodness-of-fit tests with exact binomial probabilities computed for each cell were used to compare breed distributions of dogs with EFB with those of the canine patient population at Ontario Veterinary College Health Sciences Center during the study period. Exact confidence intervals were computed for all reported breed proportions for dogs with EFB.
Among dogs with EFB, the dependent variables examined included treatment outcome (death related to EFB entrapment, excellent outcome, or stricture) and whether surgery had been required to remove the EFB. The standard Pearson χ2 test was used to compare distributions of categorical variables between groups. However, because expected frequencies were often quite small, exact P values were computed when feasible; otherwise, P values were estimated through Monte Carlo simulation based on 100,000 samples.
For continuous variables (eg, age [for outcome comparisons], body weight, and duration of hospitalization), standard ANOVA for outcome comparisons or the Student t test for treatment (surgery or endoscopy) comparisons was used; some data required transformation to achieve a normal distribution. The Kruskal-Wallis test or Wilcoxon-Mann-Whitney test was also performed for comparisons involving age (for treatment comparisons), rectal temperature, duration of EFB entrapment, and duration of procedure. To assess the ANOVA assumptions, the residuals were analyzed through plotting the residuals versus the predicted values and assessment for normality of distribution via the Shapiro-Wilk, Kolmogorov-Smirnov, Cramér-von Mises, and Anderson-Darling tests. Such residual analysis may reveal outliers, unequal variances, or the need for data transformation. Values of P < 0.05 were considered significant.
On the basis of the results of univariable analysis, multivariable logistic regression was used to predict the probability of requiring surgery for a few explanatory variables (ie, the continuous variables age, rectal temperature, EFB duration, and BW; and the categorical variables foreign body type, anorexia, lethargy, sex, and EFB location). All 2-way interactions were included in the model as well as quadratic terms for the continuous variables.
For evaluations of factors associated with treatment outcomes, dogs considered to have undergone successful endoscopic EFB removal included those for which the EFB was either removed successfully or pushed into the stomach and did not require surgical intervention. Dogs considered to have required surgery included those that underwent surgery without prior endoscopy, dogs that had their EFB pushed into the stomach and required surgical removal by gastrotomy, and dogs for which endoscopic removal of the EFB failed, resulting in surgery or euthanasia because of the need for surgical intervention.
Results
Dogs
Over the 20-year study period, 224 incidents of EFB entrapment in 223 dogs met the inclusion criteria and were therefore included in the study. One dog was evaluated twice for this condition within a 1-year period. Dogs were classified as mixed breed (n = 34), West Highland White Terrier (29), Shih Tzu (25), Yorkshire Terrier (18), Pomeranian (13), Cairn Terrier (8), Labrador Retriever (8), Boxer (6), Jack Russell Terrier (5), Lhasa Apso (5), Bichon Frise (4), Chihuahua (4), German Shepherd Dog (4), Maltese (4), Miniature Poodle (3), Miniature Schnauzer (3), Shetland Sheepdog (3), Airedale Terrier (2), Dachshund (2), Keeshond (2), Miniature Pinscher (2), Papillon (2), pit bull–type (2), Portuguese Water Dog (2), Pug (2), Scottish Terrier (2), Chinese Shar-Pei (2), Standard Poodle (2), Toy Poodle (2), and various other breeds (1 each). Terrier breeds (n = 71 [31.8% of all dogs]) were overrepresented, compared with other breeds. West Highland White Terriers, Shih Tzus, and Yorkshire Terriers were significantly (P < 0.001 for all comparisons) overrepresented, compared with their distributions in the entire canine patient population at the time of the study, with West Highland White Terriers 18.5 (95% CI, 12.3 to 27.6) times as likely to be admitted for EFB entrapment as any other breed.
Eighty-eight (39.5%) dogs were neutered males, 82 (36.8%) were spayed females, 32 (14.3%) were sexually intact males, and 21 (9.4%) were sexually intact females. Median age at the time of initial evaluation was 4.0 years (range, 0.2 to 14.5 years). Median body weight at the time of initial evaluation was 7.7 kg (16.9 lb; range, 0.8 to 61.5 kg [1.8 to 135.3 lb]).
Historical and clinical variables
Median duration of EFB entrapment (based on clinical signs) at initial evaluation was 1.0 day (range, 0.1 to 14.0 days). This variable was significantly (P = 0.03) associated with the need for surgical intervention (Table 1). Duration of EFB entrapment was also significantly (P = 0.04) associated with long-term outcome, whereby dogs that failed to survive had a significantly (P = 0.02) longer entrapment period than dogs with an excellent outcome (Table 2). In 5.8% (13/224) of EFB entrapment incidents, owners were not aware that a foreign body had been ingested and had consulted their veterinarian for associated clinical signs.
Descriptive data for and results of significant univariable comparisons of dogs with EFB entrapment that underwent successful endoscopy (EFB removed or pushed into the stomach with no gastrotomy required; n = 183) or required surgery (surgery alone, EFB pushed into the stomach but gastrotomy required, or failed endoscopic attempt requiring surgery or leading to euthanasia; 41).
| Variable | Endoscopy only | Surgery required | P value* | Â | |
|---|---|---|---|---|---|
| Mean (SD or 95% CI) | Median (range) | Mean (SD or 95% CI) | Median (range) | Â | |
| Age (y) | 4.58 (3.62) | Â | Â | Â | Â |
| Body weight (kg)†| 7.80 (6.92–8.80) |  |  |  |  |
| Rectal temperature (°C) | 38.64 (0.71) |  |  |  |  |
| Duration of EFB entrapment (d) | 1.65 (2.03) | Â | Â | Â | Â |
| Duration of hospitalization (d) | 1.59 (1.57) | Â | Â | Â | Â |
P values for body weight and duration of hospitalization represent comparisons of means; those for the remaining variables represent comparisons of medians.
Logarithmically transformed data are presented as back-transformed mean (geometric mean) and 95% CI.
— = Not applicable.
Descriptive data for and results of significant univariable comparisons of dogs with EFB entrapment that had an excellent long-term outcome (n = 90), developed a stricture (16), or failed to survive (12) following veterinary intervention for EFB.
| Variable | Excellent outcome | Stricture | Nonsurvival | Â | |||
|---|---|---|---|---|---|---|---|
| Mean (SD or 95% CI) | Â | Â | Â | Â | Â | ||
| Age (y) | 4.46 (3.38) | Â | Â | Â | Â | Â | Â |
| Body weight (kg)†| 8.90 (7.61–10.48) |  |  |  |  |  |  |
| Rectal temperature (°C) | 38.55 (0.67) |  |  |  |  |  |  |
| Duration of EFB entrapment (d) | 1.66 (2.06) | Â | Â | Â | Â | Â | Â |
| Duration of procedure (min)†| 31.6 (3.4–36.8) |  |  |  |  |  |  |
P values for body weight and duration of procedure represent comparisons of means among all 3 outcomes; those for the remaining variables represent comparisons of medians among all 3 outcomes.
See Table 1 for remainder of key.
The most common clinical signs at the time of initial evaluation were vomiting, retching, and lethargy (Table 3). The odds of dogs requiring surgery were 2.6 (95% CI, 1.2 to 5.7) times as great for dogs with versus without signs of anorexia (P = 0.02) and 2.5 (95% CI, 1.2 to 5.2) times as great for dogs with versus without signs of lethargy (P = 0.02).
Number (%) of EFB events with various clinical signs associated with EFB entrapment that were treated by endoscopy alone (n = 174) or required surgical intervention (36).
| Clinical sign | Endoscopy alone | Â | Â |
|---|---|---|---|
| Anorexia | 45 (25.9) | Â | Â |
| Gagging or retching | 69 (39.7) | Â | Â |
| Regurgitation | 46 (26.4) | Â | Â |
| Vomiting | 102 (58.6) | Â | Â |
| Hypersalivation | 35 (20.1) | Â | Â |
| Pain | 26 (14.9) | Â | Â |
| Dyspnea | 15 (8.6) | Â | Â |
| Coughing | 31 (17.8) | Â | Â |
| Sneezing | 1 (0.1) | Â | Â |
| Lethargy | 54 (31.0) | Â | Â |
| Esophageal perforation | 6 (3.3)* | Â | Â |
Data available for 183 dogs.
Data available for 40 dogs.
Mean rectal temperature at initial evaluation for 190 dogs for which it was recorded was 38.7°C (101.7°F; SD, 0.67°C [1.2°F]). Dogs in which endoscopy was successful had significantly more variance in rectal temperatures than dogs that required surgery, which also had a significantly (P = 0.047) higher mean rectal temperature (Table 1). Body weight was recorded for all 233 dogs; those that required surgery were significantly (P = 0.02) heavier than dogs for which endoscopic removal was successful. None of the other evaluated variables were associated with the requirement for surgical treatment.
Diagnostic tests
For 99.6% (223/224) of EFB entrapments, plain radiography, contrast radiography, or both were used to make the diagnosis. One dog that had ingested a bamboo skewer was treated successfully with endoscopic retrieval without any prior imaging. Repeated radiography at the time of initial evaluation at the referral center revealed that 99.2% (128/129) of EFB entrapments had an unchanged position, compared with the position in the radiographs from the referring veterinarian. In the 1 instance of foreign body migration, the piece of bone had moved aborally from the heart base to the caudal esophageal region.
Foreign body types
The most common types of EFB were bone (n = 134 [59.8%]), rawhide (23 [10.3%]), fishhook (20 [8.9%]), meat (11 [4.9%]), and apple (8 [3.6%]). Other types of EFBs included toys (n = 6), vegetables (4), dog biscuits or other treats (4), needles or safety pins (3), plant material (2), pig ears (2), endotracheal or feeding tubes (2), skewers (2), mixed materials (2), and basket material (1). The most common locations for EFB entrapment were the caudal esophageal region (n = 130 [94 bone, 12 rawhide, 2 fishhook, 5 apple, 6 meat, and 11 other]), heart base (51 [27 bone, 8 rawhide, 3 fishhook, 2 apple, 2 meat, and 9 other]), cervical esophageal region (25 [8 bone, 1 rawhide, 13 fishhook, 1 meat, and 2 other]), and thoracic inlet (9 [4 bone, 2 fishhook, 1 apple, 1 meat, and 1 other]) or in multiple locations (1 piece of bone in each of the cervical esophageal and heart base regions). For 3 bone, 2 rawhide, 1 meat, and 2 other foreign bodies, no location was recorded.
Location of EFB entrapment was associated with intervention method (pull by endoscopy vs push by endoscopy into stomach with or without gastrotomy vs failed endoscopy), with cervical EFBs 30.6 (95% CI, 4.8 to 637.9; P < 0.001) times as likely to be retrieved by pulling than those located in the caudal esophageal region, which were pulled out orally or pushed into the stomach more equally.
Fishhooks were 60.8 (95% CI, 11.9 to 430.2) times as likely to be lodged in the cervical esophageal region as were bones, and bones were more likely to be lodged in the caudal esophageal region. Dogs with fishhook entrapment were significantly (P < 0.001) heavier (mean body weight, 15.99 kg [35.18 lb]; 95% CI, 11.21 to 22.81 kg [24.66 to 50.18 lb]) than dogs with bone (7.35 kg [16.17 lb]; 95% CI, 6.48 to 8.44 kg [14.26 to 18.57 lb]), rawhide (7.59 kg [16.70 lb]; 95% CI, 5.45 to 56.00 kg [11.99 to 123.20 lb]), or other types of EFB (8.85 kg [19.47 lb]; 95% CI, 7.04 to 11.13 kg [15.49 to 24.49 lb]).
Treatment
Two hundred nineteen (97.8%) EFB entrapments were initially treated by endoscopy and 5 (2.2%) by surgical retrieval. For EFBs treated by endoscopy, 183 (83.6%) had a successful outcome with endoscopy alone and 36 (16.4%) did not (30 bone, 2 fishhook, 1 needle, 1 safety pin, 1 toy, 1 pig ear, 0 rawhide, 0 apple, and 0 meat). One hundred twenty-eight of 219 (58.4%) EFBs initially treated by endoscopy were successfully retrieved through the oral cavity, with the most common locations for entrapment in these instances being the caudal esophageal (n = 54) and heart base (37) regions. For 68 of 219 (31.1%) EFBs initially treated by endoscopy, the object was advanced into the stomach because oral retrieval was unsuccessful; for 55 (80.9%) of these EFBs, dogs were left to pass or digest the foreign body (considered successful attempts at initial endoscopic removal), whereas the other 13 (19.1%) EFBs (11 bone material, 1 toy, and 1 pig ear) were removed by gastrotomy because they were deemed too large or sharp to safely pass through the gastrointestinal tract (considered unsuccessful attempts at initial endoscopic removal). The remaining 23 (10.5%) EFBs for which endoscopic removal was initially attempted required surgical intervention other than gastrotomy. For 18 of these 23 EFBs, surgical retrieval was performed, and for the remaining 5 EFBs, each in a different dog (2.2% of all dogs with EFB), owners elected euthanasia instead.
In total, gastrotomy was successfully performed for 20 (8.9%) EFBs (all of which had an initial retrieval attempt via endoscopy). For 7 of these EFBs, prior attempts at endoscopic retrieval had been unsuccessful. One additional dog had a cervical EFB (bone) removed endoscopically; however, a second FB obstructing the pylorus could not be retrieved endoscopically and required gastrotomy for removal.
Esophagotomy was performed for 16 (7.1%) EFBs (11 for which endoscopic removal was attempted first and 5 for which surgery was attempted first) and was successful for 15 of these. A cervical approach to removal via esophagotomy was used for 6 EFBs (2 in the cervical esophageal region, 1 at the thoracic inlet, 2 at the heart base, and 1 in the caudal esophageal region). The caudally located EFB for which esophagotomy was performed could not be retrieved through this approach and was therefore retrieved from the esophagus via the cardia through a gastrotomy approach. Thoracotomy was successfully performed to remove 8 (3.6%) EFBs in the caudal esophageal region and 2 (0.1%) at the heart base (7 for which the initial removal attempt was made via endoscopy, and 3 via surgery).
Exact univariate logistic regression revealed that type of EFB was significantly (P = 0.005) associated with failure or success of endoscopic attempts at removal. Specifically, bones were more likely to require surgery than rawhides (OR, 9.57; 95% CI, 2.00 to infinity; P = 0.008), which never required surgery, or foreign bodies classified as other (OR, 3.20; 95% CI, 1.04 to 13.29; P = 0.04). Furthermore, the odds of successful endoscopic removal of bones were 16.13 (95% CI, 3.17 to infinity; P < 0.001) times as high when the bones were located in the cervical versus caudal esophageal region and 8.71 (95% CI, 2.51 to 37.03; P < 0.001) times as high when located at the heart base versus the caudal esophageal region.
Exact conditional (multivariable) logistic regression modeling revealed that the odds of requiring surgical intervention were associated with dog body weight (P = 0.003), anorexia (P = 0.01), and EFB type (P = 0.01). These odds increased by 35% (OR, 1.35; 95% CI, 1.11 to 1.63) for each 5-kg (11-lb) increase in body weight. For dogs with signs of anorexia, the odds of requiring surgical intervention were 2.88 (95% CI, 1.27 to 6.54) times as great as those for dogs without these signs. In the multivariate model, bone EFBs continued to have higher odds of requiring surgical intervention (OR, 5.02; 95% CI, 1.32 to 19.09; P = 0.02) than EFBs of other types (eg, needles, safety pins, toys, or pig ears) but also had greater odds of requiring surgical intervention than fishhooks (OR, 8.34; 95% CI, 1.01 to 69.18; P = 0.049). No significant interactions or quadratic terms were identified for these variables.
Univariate analysis revealed that esophageal perforation was associated with the need for surgical intervention (OR, 12.41; 95% CI, 4.27 to 36.50; P < 0.001) and with nonsurvival (OR, 19.70; 95% CI, 4.96 to 80.58; P < 0.001). Existing esophageal perforation was identified endoscopically for 17 EFBs and occurred iatrogenically in 1 dog during foreign body removal, resulting in a total of 18 dogs with perforation (12 bone, 4 fishhook, 1 rawhide, and 1 needle). Type of EFB was not significantly (P = 0.07) associated with perforation. Perforations had occurred in the caudal esophageal region (n = 10), at the heart base (5), in the cervical esophageal region (2), and at the thoracic inlet (1). Three dogs with esophageal perforation were euthanized after owners declined surgical intervention. Five dogs with esophageal perforation received no surgical treatment after endoscopic EFB removal and survived to discharge from the hospital; 3 were alive at the time of last follow-up, and 2 were lost to follow-up. The remaining 10 dogs underwent surgical repair of the esophageal perforation after endoscopic EFB removal; 4 of these dogs had no complications at the time of follow-up (12 to 52 months after surgery). Six dogs developed postprocedural complications related to esophageal perforation (aspiration pneumonia or pneumothorax); 4 dogs were subsequently euthanized, 1 was alive at the time of last follow-up, and the remaining dog was discharged after 6 days of hospitalization but was subsequently lost to follow-up. Overall, 7 of 18 (38.9%) dogs with esophageal perforation (6 treated initially with endoscopy and 1 treated initially with surgery) were euthanized.
Postprocedural treatments, when recorded, included various analgesics (ie, opioids and NSAIDs), gastroprotectants, and antimicrobials. Protocols differed according to clinician preference, severity of esophagitis, and concurrent medical problems.
Outcome
Median duration of hospitalization for all dogs was 1 day (range, 0.2 to 14.0 days). Dogs requiring surgery to treat the EFB had a significantly (P < 0.001) longer hospitalization period than dogs requiring no surgery (Table 1). Twelve dogs died or were euthanized in the hospital (11 related to EFB). Eleven of the in-hospital deaths or euthanasias were directly associated with the EFB: 7 related to esophageal perforation (3 for which endoscopic removal failed), 1 owing to cardiorespiratory arrest, 1 owing to aspiration pneumonia and respiratory failure, and 2 owing to the owner declining surgery to remove an EFB (without perforation) that could not be removed endoscopically. One dog was euthanized 1 day after discharge because of poor recovery related to the EFB event, resulting in a 5.4% (12/223) overall mortality rate related to EFB entrapment.
The most common complication recorded after discharge from the hospital was esophageal stricture, which was noted in 16 of 143 (11.2%) dogs for which follow-up information was available. Dogs that developed esophageal stricture were significantly (P = 0.01) older than those with an excellent outcome (Table 2).
Follow-up information was not consistently available for the first 10 years of the 20-year study period. Of the 134 dogs evaluated during the last 10 years, 7 died in hospital and 25 were lost to follow-up or had a follow-up period < 1 month. Ninety of 102 (88.2%) dogs with a median follow-up period of 27 months (range, 1 to 64 months) after EFB treatment had an excellent outcome, and 12 (11.8%) dogs developed esophageal stricture in the latter 10 years of the study period.
Mean procedure duration was significantly (P = 0.03) shorter for dogs with a good outcome (Table 2). A significant (P = 0.02) association was identified between EFB type and stricture formation, with food, toys, and household items resulting in most esophageal strictures. None of the dogs with fishhook EFBs developed a stricture, and very few dogs with bone (n = 6) and rawhide (2) EFBs developed a stricture.
Discussion
As previously reported,4,5,12,13 terrier breeds were common among dogs with EFB entrapment in the present study. West Highland White Terriers, Shih Tzus, and Yorkshire Terriers were significantly overrepresented, compared with their distribution in the entire canine patient population of the referral center. Older age was associated with a poor outcome, which could have reflected a greater likelihood of older dogs being euthanized in such situations or developing more complications, which was the situation for stricture.
Clinical signs of dogs in the present study were consistent with those in previous reports,12,13 and anorexia, lethargy, and high rectal temperature were associated with the need for surgical intervention. Of these variables, only anorexia remained significant in the multivariable model, indicating that anorexia was unrelated to the other examined factors. Although mean rectal temperature differed significantly between dogs requiring surgery rather than endoscopy only, the actual difference was clinically unimportant.
Mean duration of EFB entrapment as indicated by clinical signs was brief for both dogs treated by endoscopy alone and those requiring surgery; however, a longer entrapment period was associated with a greater likelihood of surgical intervention and a poorer outcome, which is also consistent with previous reports.2,15 This finding could be explained by the presumption that prolonged (vs briefer) EFB entrapment would be more likely to cause esophageal ulceration or perforation or lead to postentrapment stricture.
Body weight was associated with surgical intervention in both univariable and multivariable models in the present study, whereby a higher proportion of large dogs required surgical EFB removal. This was likely attributable to the fact that access to the EFB with endoscopic equipment may have been more difficult in larger dogs owing to the length of the esophagus and the presumed relatively larger-sized EFB, even when large nonendoscopic forceps or balloon catheters were used outside of the flexible endoscope working channel.
Similar to findings of previous studies,4,14,15 bone material was the most commonly encountered EFB (134/224 [59.8%]) in the present study. Most of the complications and failures of endoscopic removal pertained to bone material. This may have been a result of properties of the bone that made endoscopic grasping and manipulation difficult or sharp edges that lodged the EFB firmly within the esophageal wall. Dogs with bone EFBs more frequently required surgical removal, although in many situations this was for retrieval by routine gastrotomy of a large bone or joint fragment that had been pushed into the stomach. In contrast, rawhide EFBs required no surgical intervention at all, possibly because of their soft and easily grasped nature, which allowed for retrieval or pushing into the stomach for digestion. More than half of the EFBs retained in the cervical esophageal region (13/25 [52.0%]) were fishhooks, which may have contributed to the ease of removal with infrequent need for surgical intervention and the lack of stricture development in affected dogs. Foreign body location was associated with successful endoscopic removal, with EFBs located in the cervical esophageal region 30.6 times as likely to be removed by endoscopy as those in the caudal region. However, several bone EFBs that could not be retrieved orally via endoscopy were instead pushed into the stomach and did not require surgical intervention.
In the present study, endoscopic removal was successful for 83.6% (183/219) of EFBs, which is consistent with previous reports14,15,26,27 of success rates ranging from 89% to 91%. Combined endoscopic and surgical intervention increased the success rate to 97.8% (219/224), with many of the surgical interventions being a simple gastrotomy via laparotomy. Endoscopy therefore remains the most valuable initial treatment option for EFB removal in dogs.
Reported rates for esophageal stricture formation following EFB removal range from 24% to 29.4%.2,15,26 In the present study, 16 of 143 (11.2%) dogs with follow-up information available developed esophageal stricture after EFB removal. This low stricture formation rate could have been attributable to improvements in endoscopic retrieval techniques, more standardized postprocedural management directed at feeding protocols and the use of gastroprotectants, or the overall brief duration of EFB entrapment prior to intervention. The small number of EFBs resulting in stricture formation also precluded identification of risk factors for stricture formation in the present study.
Procedure duration was negatively associated with long-term outcome in the study reported here. Although procedure duration and iatrogenic trauma can be affected by the experience level of the attending endoscopist, we could not quantify this variable retrospectively. The observed association between procedure duration and outcome may have been attributable to the possibility that EFBs that were more difficult to remove caused greater tissue trauma and required a greater amount of manipulation for removal. Longer-standing foreign bodies may also have been lodged more tightly and therefore more difficult to dislodge or retrieve. The inflammation associated with the trauma from EFB entrapment and removal likely resulted in esophageal stricture and a poorer long-term outcome. Esophageal perforation was also negatively associated with outcome. Six of 18 dogs with perforation developed complications related to the perforation, and 7 of 18 dogs with perforation were ultimately euthanized.
The overall mortality rate related to EFB in the present study (5.4%) was lower than previously reported (21% to 26%).7,12,15,26 In the situations in which owners elected euthanasia, both surgical and medical management were offered but declined. Therefore, the true survival rate could not be determined. A potential explanation for the lower mortality rate in the present versus other studies could have been related to the overall briefer duration of EFB entrapment prior to therapeutic intervention.
One last finding worthy of discussion was that all but 1 EFB were found in the same radiographic location as those indicated in prereferral radiographs. This may have suggested that repeated imaging of dogs with EFBs is not necessary, but this possibility should be considered with caution. Many dogs with EFBs in the present study had been referred quickly after diagnosis and underwent endoscopy shortly after hospital admission, no matter the time of day or night. A prolonged delay between imaging and endoscopy could certainly result in movement of the EFB. Furthermore, failure to recognize that a bone EFB has passed naturally into the stomach could result in unnecessary general anesthesia and endoscopy. In contrast, failure to identify pneumothorax related to perforation could result in endoscopy with the associated risks when proceeding directly to thoracotomy was warranted.
The main limitation of the present study was its retrospective nature. Standardized grading for esophageal lesions at the time of EFB removal was not performed, and endoscopic or imaging follow-up information was not available in all situations to evaluate for the presence of esophageal stricture. Such information could have provided valuable data regarding the true rate of esophageal stricture formation after EFB removal, particularly when subclinical. In addition, most of the consistent follow-up data were available for only the last 10 years of the study period and therefore the rate of stricture formation was not reflective of all dogs in our study.
In the study reported here, longer duration of entrapment, higher body weight, anorexia, lethargy, higher rectal temperature, and esophageal perforation were associated with the need for surgical intervention (vs endoscopy only) in dogs with EFB entrapment. Older dogs, longer duration of EFB entrapment, and longer procedure duration were also associated with a poorer prognosis. No other factors were associated with the need for surgical intervention or outcome. On the basis of these results, endoscopic retrieval remains the initial treatment option of choice, and recommendations regarding when to forego endoscopy in favor of proceeding directly to surgery cannot be made except in situations in which perforation is evident.
Acknowledgments
The authors thank Gabrielle Monteith for assistance with statistical analyses.
ABBREVIATIONS
| CI | Confidence interval |
| EFB | Esophageal foreign body |
Footnotes
GIF 130, GIF P140, or GIF XP160, Olympus America Co, Melville, N Y.
SAS, version 9.3, SAS Institute, Cary, NC.
References
1. Sale CS, Williams JM. Results of transthoracic esophagotomy retrieval of esophageal foreign body obstructions in dogs: 14 cases (2000–2004). J Am Anim Hosp Assoc 2006;42:450–456.
2. Rousseau A, Prittie J, Broussard JD, et al. Incidence and characterization of esophagitis following esophageal foreign body removal in dogs: 60 cases (1999–2003). J Vet Emerg Crit Care 2007;17:159–163.
3. Michels GM, Jones BD, Huss BT, et al. Endoscopic and surgical retrieval of fishhooks from the stomach and esophagus in dogs and cats: 75 cases (1977–1993). J Am Vet Med Assoc 1995; 207:1194–1197.
4. Houlton JEF, Herrtage ME, Taylor PM, et al. Thoracic oesophageal foreign bodies in the dog: a review of ninety cases. J Small Anim Pract 1985;26:521–536.
5. Moore AH. Removal of oesophageal foreign bodies in dogs: use of the fluoroscopic method and outcome. J Small Anim Pract 2001;42:227–230.
6. Tams TR. Diseases of the esophagus. In: Handbook of small animal gastroenterology. Philadelphia: Saunders, 2003;118–158.
7. Ryan WW, Greene RW. The conservative management of esophageal foreign bodies and their complications: a review of 66 cases in dogs and cats. J Am Anim Hosp Assoc 1975;11:243–249.
8. Parker NR, Walter PA, Gay J. Diagnosis and surgical management of esophageal perforation. J Am Anim Hosp Assoc 1989;25:587–594.
9. Pearson H, Gibbs C, Kelly DF. Oesophageal diverticulum formation in the dog. J Small Anim Pract 1978;19:341–355.
10. Dodman NH, Baker GJ. Tracheo-oesophageal fistula as a complication of an oesophageal foreign body in the dog—a case report. J Small Anim Pract 1978;19:291–296.
11. Faulkner RTHD, Sammons ML, Hilmas DE, et al. A case of esophageal foreign body with mediastinal abscess formation in a dog. J Am Anim Hosp Assoc 1976;12:70–76.
12. Pearson H. Symposium on conditions of the canine oesophagus. I. Foreign bodies in the oesophagus. J Small Anim Pract 1966;7:107–116.
13. Spielman BL, Shaker EH, Garvey MS. Esophageal foreign body in dogs: a retrospective study of 23 cases. J Am Anim Hosp Assoc 1992;28:570–574.
14. Thompson HC, Cortes Y, Gannon K, et al. Esophageal foreign bodies in dogs: 34 cases (2004–2009). J Vet Emerg Crit Care (San Antonio) 2012;22:253–261.
15. Juvet F, Pinilla M, Shiel RE, et al. Oesophageal foreign bodies in dogs: factors affecting success of endoscopic retrieval. Ir Vet J 2010;63:163–168.
16. Hawe RS. Catheter technique for removing a foreign body from the canine esophagus. Vet Med Small Anim Clin 1979;74:1615–1618.
17. Gianella P, Pfammatter NS, Burgener IA. Oesophageal and gastric endoscopic foreign body removal: complications and follow-up of 102 dogs. J Small Anim Pract 2009;50:649–654.
18. Keir I, Woolford L, Hirst C, et al. Fatal aortic oesophageal fistula following oesophageal foreign body removal in a dog. J Small Anim Pract 2010;51:657–660.
19. Zimmer JF. Canine esophageal foreign bodies: endoscopic, surgical, and medical management. J Am Anim Hosp Assoc 1984;20:669–677.
20. Kyles AE, Schneider TA, Clare A. Foreign body intestinal perforation and intra-abdominal abscess formation as a complication of enteroplication in a dog. Vet Rec 1998;143:112–113.
21. Leib MS, Dinnel H, Ward DL, et al. Endoscopic balloon dilation of benign esophageal strictures in dogs and cats. J Vet Intern Med 2001;15:547–552.
22. Bissett SA, Davis J, Subler K, et al. Risk factors and outcome of bougienage for treatment of benign esophageal strictures in dogs and cats: 28 cases (1995–2004). J Am Vet Med Assoc 2009;235:844–850.
23. Adamama-Moraitou KK, Rallis TS, Prassinos NN, et al. Benign esophageal stricture in the dog and cat: a retrospective study of 20 cases. Can J Vet Res 2002;66:55–59.
24. Wilson DV, Walshaw R. Postanesthetic esophageal dysfunction in 13 dogs. J Am Anim Hosp Assoc 2004;40:455–460.
25. Melendez LD, Twedt DC, Weyrauch EA, et al. Conservative therapy using balloon dilation for intramural, inflammatory esophageal strictures in dogs and cats: a retrospective study of 23 cases (1987–1997). Eur J Comp Gastroenterol 1998;3:31–36.
26. Leib MS, Sartor LL. Esophageal foreign body obstruction caused by a dental chew treat in 31 dogs (2000–2006). J Am Vet Med Assoc 2008;232:1021–1025.
27. Jankowski M, Spuzak J, Kubiak K, et al. Oesophageal foreign bodies in dogs. Pol J Vet Sci 2013;16:571–572.
