APT is a surgically created permanent stoma between the trachea and overlying skin to bypass an upper airway obstruction, and PTs are used to treat various upper airway disorders, including laryngeal paralysis, collapse, neoplasia, trauma; persistent inflammation or edema of the upper airway; and other causes of permanent laryngeal dysfunction.1–6 Substantial postoperative care is required for patients that receive a PT, particularly in the immediate postoperative period when the risk of tracheal obstruction and death is high because of increased mucus production.5,7 Postoperative complications include mucus secretion, coughing, aspiration pneumonia, and death (from asphyxiation caused by obstruction of the stoma).5,6 Revision surgery may be required if stenosis develops at the stoma site or if excessive skinfolds obstruct the stoma.2,5 Despite these risks, a PT can be an effective option for treating obstructive upper airway diseases.2,5,6
Few reports have documented the long-term outcomes in dogs with PTs. A recent study5 involving 21 dogs with PTs shows a median survival time of 328 days, noting that the initial diagnosis was not associated with survival time and that neither body weight nor sex were associated with cause of death. Major complications occurred in 10 of 20 (50%) dogs and included most commonly aspiration pneumonia (n = 5 [25%]) or obstruction of the stoma requiring revision surgery (4 [20%]).5 The most common cause of stoma obstruction in those 4 affected dogs was skinfold occlusion (n = 3), and overall, the causes of death were respiratory related in 9 of 19 dogs.5 In a study2 involving 23 dogs and 11 cats with PTs, skinfold occlusion was the most common complication (6/22 [27%]); however, the median survival time was not reported.
Given the limited data available on long-term outcomes in dogs with PTs, the purposes of the study presented here were to evaluate long-term outcomes and identify factors associated with death or the need for revision surgery in dogs with PTs. The hypothesis was that dogs with PTs would have longer survival times than previously reported and that brachycephaly would be a predictive factor for shorter survival time.
Materials and Methods
Medical records from 4 veterinary teaching hospitals (University of Georgia, Texas A&M University, North Carolina State University, and Auburn University) were searched to identify dogs that received a PT between January 2002 and June 2016. Any dog that died or was lost to follow-up < 2 weeks after receiving a PT was excluded. Data obtained from medical records included signalment, date of surgery for a PT, preexisting respiratory disease, reason for a PT, previous medical or surgical treatment, laryngeal examination findings, preoperative radiographic findings, preoperative presence of esophageal abnormalities (diagnosed with contrast esophagography or thoracic radiography), major and minor postoperative complications, duration of hospitalization, cause of death, and survival time. The standard surgical procedure for creating a PT has been previously reported3; however, confirmation that the procedure was performed to the exact specification as that report was not possible. Major postoperative complications were defined as those requiring revision surgery or affecting quality of life (eg, mucus plug obstruction, skinfold occlusion, stenosis, collapse, or dehiscence of the stoma; syncope; aspiration pneumonia; and dyspnea). Aspiration pneumonia was diagnosed on the basis of thoracic radiographic findings. Minor postoperative complications were defined as those that did not require revision surgery to resolve or that did not result in risk of death (eg, infection at the stoma, vomiting, and regurgitation). Data on complications were obtained from the medical records or through follow-up with owners.
Causes of death were recorded as respiratory-related or nonrespiratory-related causes, similar to how causes were grouped in a previous report.5 Respiratory-related causes of death included death from the underlying respiratory disease for which treatment included a PT or from complications associated with the PT itself. All other causes of death were categorized as nonrespiratory-related causes. Dogs for which the cause of death could not be determined were excluded from statistical analyses pertaining to cause of death. All owners with available contact information and whose pets were discharged from the hospital alive were contacted by telephone or email and given a questionnaire to gather additional information on postoperative complications, cause of death, and approximate date of death if not available in the medical record.
Statistical analysis
Continuous data were evaluated for normality with the Kolmogorov-Smirnov test. Data with normal distribution were reported as mean ± SEM, and data not normally distributed were reported as median (range). Categorical and ordinal data were reported as percentages of the total. The Fisher exact test (n ≤ 5) and χ2 test (n > 5) were used to identify potential associations between preoperative factors and postoperative major complications and cause of death, with results reported as OR (95% CI of OR). Minor postoperative complications were not evaluated statistically because of the inability to accurately and repeatably define severity and frequency on retrospective evaluation. Only factors that included ≥ 6 dogs were considered for statistical analysis. Between categories (eg, dogs that did vs dogs that did not have extraluminal tracheal support ring prostheses placed), the Mann-Whitney U test was used to compare the number of tracheal rings removed. The Kaplan-Meier method and log-rank tests were used to evaluate correlations between survival time and pre- and postoperative factors, signalment, and cause of death. Dogs that were alive or lost to follow-up at the time of contact for the questionnaire and those that died of nonrespiratory-related causes were censored. The Cox proportional hazard model was used to evaluate continuous variables relative to survival time, and results were reported as hazard ratio with 95% CI. Values of P < 0.05 were considered significant.
Results
Animals
A search of the medical records identified 78 client-owned dogs that received a PT during the study period; however, 9 dogs were excluded from the study because they died (n = 7) or were lost to follow-up (2) < 2 weeks after receiving a PT. Of the remaining 69 dogs that met the inclusion criteria, 28 (41%) were spayed females, 25 (36%) were castrated males, 11 (16%) were sexually intact males, and 5 (7%) were sexually intact females. Breeds were recorded as Labrador Retriever (16/69 [23%]); Pekingese and Yorkshire Terrier (6 [9%] each); Pug (5 [7%]); Pomeranian (4 [6%]); Boston Terrier, Jack Russell Terrier, and Miniature Pinscher (3 [4%] each); Chihuahua, English Bulldog, French Bulldog, Golden Retriever, and Norwich Terrier, (2 [3%] each); and 13 other breeds (1 [1%] each). Twenty-two of the 69 (32%) dogs were brachycephalic breeds. Overall, mean ± SEM age at the time of initial evaluation was 7.6 ± 0.45 years.
Clinical signs
Respiratory-related clinical signs on initial examination included dyspnea (40/69 [58%]), stridor (7 [10%]), coughing (6 [9%]), collapse (6 [9%]), stertor (6 [9%]), wheezing (2 [3%]), and nasal discharge (1 [1%]), with > 1 clinical sign in some dogs. Nonrespiratory clinical signs included vomiting or regurgitation (3 [4%]), lethargy (2 [3%]), seizures (2 [3%]), anorexia (1 [1%]), abdominal distension (1 [1%]), neck pain (1 [1%]), hematochezia or hematemesis (1 [1%]), and polyuria and polydipsia (1 [1%]). Median duration of clinical signs before surgery for a PT was 102 days (range, 1 to 2,555 days).
Diagnosis, prior treatment, and diagnostic procedures
There were various underlying respiratory conditions diagnosed (Table 1), with laryngeal paralysis (40/69 [58%]) or laryngeal collapse (19 [28%]) most common. Of the 40 dogs with laryngeal paralysis, 18 (45%) were small-breed dogs, 16 (40%) were large-breed dogs (including 14 Labrador Retrievers), 4 (10%) were medium-breed dogs, and 2 (5%) were giant-breed dogs. Five of 40 (13%) dogs with laryngeal paralysis also had concurrent laryngeal collapse, which in the medical record was noted as stage I8 (n = 2) or not characterized (3). Overall, only 5 of the 69 (7%) dogs did not have a laryngeal examination recorded in the medical record before receiving a PT, and 4 of these 5 dogs had presumed laryngeal paralysis on the basis of findings from examinations by referring veterinarians (n = 2) or because of trauma to the neck (2). The remaining dog did not have results of a laryngeal examination noted in the medical record; however, tracheal collapse was diagnosed in this dog.
Diagnoses of underlying respiratory conditions in 69 client-owned dogs that survived > 2 weeks after surgery for a PT between January 2002 and June 2016.
| Diagnoses | No. (%) of dogs* |
|---|---|
| Laryngeal paralysis | 40 (58) |
| Laryngeal collapse | 19 (28) |
| Tracheal collapse | 12 (17) |
| Mass or neoplasia | 8 (12) |
| Edema or inflammation | 7 (10) |
| Trauma | 2 (3) |
Some dogs had > 1 condition diagnosed.
Before receiving a PT, 42 of the 69 (61%) dogs had received medical treatment and 31 (45%) had received surgical treatment (Table 2). For instance, 3 of the 4 dogs that underwent previous surgical treatment with placement of extraluminal rings for tracheal collapse developed laryngeal paralysis and required PTs. Overall, 13 of 69 (19%) dogs had received a temporary tracheostomy before a PT, and creation of a temporary tracheostomy was the most common previous surgery (13/31 [42%]) performed in dogs receiving a PT.
Number (%) of the 69 dogs in Table 1 that received various medical (n = 42) or surgical (31) treatments before receiving a PT.
| Previous treatments | No. (%) of dogs* |
|---|---|
| Medical | 42 (61) |
| Corticosteroids | 32 (46) |
| Antimicrobials | 19 (28) |
| Bronchodilators | 14 (20) |
| Sedatives | 14 (20) |
| Allergy medications† | 10 (14) |
| Antitussives | 5 (7) |
| Gastrointestinal medications‡ | 5 (7) |
| Surgical | 31 (45) |
| Temporary tracheostomy | 13 (19) |
| Unilateral arytenoid lateralization | 9 (13) |
| Staphylectomy | 9 (13) |
| Rhinoplasty | 5 (7) |
| Laryngeal sacculectomy | 4 (6) |
| Ventriculocordectomy | 4 (6) |
| Tracheal ring placement | 4 (6) |
Some dogs received > 1 treatment.
Antihistamines and antipruritics
Antacids, antiemetics, and prokinetics
Fifty-two of the 69 (75%) dogs underwent preoperative thoracic radiography, and results indicated that 10 of the 52 (19%) dogs had evidence of aspiration pneumonia and that 7 (13%) had megaesophagus. In addition, contrast esophagography was performed on 4 dogs, and results indicated that all 4 dogs had evidence of dysmotility, but no evidence of megaesophagus on static radiography. Further, laryngeal paralysis was also diagnosed in all 4 of these dogs. Thus, in total, esophageal abnormalities were diagnosed on the basis of radiographic evidence in 11 of the 69 (16%) dogs, and laryngeal paralysis was also diagnosed in these same 11 dogs.
Surgery
Surgical reports were available for 65 of the 69 (94%) dogs, and the number of tracheal rings removed (median, 4 rings; range, 1 to 7 rings) was recorded in 53 of the 65 (82%) dogs. Concurrent with the surgery for a PT, neck skin resection was performed in 38 of 65 (58%) dogs (23 nonbrachycephalic dogs and 15 brachycephalic dogs). Extraluminal tracheal support ring prostheses were placed in 7 of 65 (11%) dogs, and only 1 of the 7 dogs had tracheal collapse. The remaining 11 dogs with tracheal collapse did not have prostheses placed at the time of surgery for a PT; however, 4 of the remaining 11 dogs with tracheal collapse had prostheses placed in earlier surgeries. In dogs that received extraluminal support rings at the time of surgery for a PT, placement was noted in the medical records as follows: at both ends of the PT site (n = 2), location not specified (2), throughout the PT site with 3 rings (1), throughout the PT site with 1 large ring spanning the site (1), and 1 ring proximal to the PT site (1). There was no meaningful difference in the median number of tracheal rings removed from dogs in which extraluminal tracheal support ring prostheses were (5 rings; range, 3 to 7 rings) or were not (4 rings; range, 1 to 5.5 rings) placed. In addition, the glottis was sutured closed in 3 of 65 (5%) dogs, and all 3 of these dogs had laryngeal paralysis. One of these 3 dogs also had megaesophagus, and another also had esophageal dysfunction detected with contrast esophagography. Duration of hospitalization was recorded in 67 of the 69 (97%) dogs, and median duration of hospitalization was 6 days (range, 1 to 26 days).
Complications
Minor complications were reported in 12 of 69 (17%) dogs, including 9 (13%) that vomited or regurgitated and 3 (4%) that had a PT site infection. Major complications were reported in 42 of 69 (61%) dogs, with aspiration pneumonia (13/69 [19%]), skinfold occlusion of the stoma (13 [19%]), and stenosis of the stoma (12 [17%]) most common (Table 3). No meaningful differences in complications were detected between dogs with versus without a temporary tracheostomy, between dogs that had versus had not undergone unilateral arytenoid lateralization, or between dogs that had versus had not undergone staphylectomy before receiving a PT. Further, no dogs with a prior temporary tracheostomy developed stenosis after receiving a PT.
Number (%) of the 69 dogs in Tables 1 and 2 that developed various major postoperative complications (n = 42 [61%]) after receiving a PT.
| Major postoperative complications | No. (%) of dogs* |
|---|---|
| Aspiration pneumonia | 13 (19) |
| Skinfold occlusion of the stoma | 13 (19) |
| Stenosis of the stoma | 12 (17) |
| Dyspnea | 9 (13) |
| Collapse of the stoma | 6 (9) |
| Syncope | 4 (6) |
| Dehiscence | 2 (3) |
Some dogs developed > 1 complication.
Twenty-four of 69 (35%) dogs underwent revision surgery. Twelve of the 22 (55%) brachycephalic dogs underwent revision surgery, whereas 12 of the 47 (26%) nonbrachycephalic dogs underwent revision surgery. Reasons for revision surgery included stenosis (9/24 [38%]; 6 brachycephalic dogs and 3 nonbrachycephalic dogs), skinfold occlusion (9 [38%]; 4 brachycephalic dogs and 5 nonbrachycephalic dogs), collapse (5 [21%]; 1 brachycephalic dog and 4 nonbrachycephalic dogs), or stomal dehiscence (2 [8%], 2 brachycephalic dogs) or excessive skin tension after skin resection (1 [4%]; 1 nonbrachycephalic dog), with some dogs having had > 1 reason for revision surgery. Brachycephalic dogs were significantly (OR, 3.5; 95% CI, 1.2 to 10.2; P = 0.02) more likely to require revision surgery than were nonbrachycephalic dogs.
All 9 dogs that underwent revision surgery for skinfold occlusion of the stoma had also undergone skinfold resection when they received a PT. The duration between receiving a PT and the first revision surgery was noted in the medical records for 23 of the 24 affected dogs, and the median duration was 103 days (range, 1 to 3,334 days).
In addition, 4 dogs (Chinese Shar-Pei, Labrador Retriever, Pekingese, and Wire Fox Terrier) underwent a second revision surgery for stenosis alone (n = 2) or stenosis combined with skinfold occlusion (2). Median duration between the first and second revision surgeries for these 4 dogs was 162 days (range, 41 to 624 days). One dog (Chinese Shar-Pei) had a third revision surgery for collapse of the stoma 37 days after the second revision surgery.
Aspiration pneumonia—Aspiration pneumonia was diagnosed in 13 of the 69 (19%) dogs after receiving a PT, and 8 of these 13 dogs had laryngeal paralysis. Thus, in total, 8 of 40 (20%) dogs with laryngeal paralysis developed aspiration pneumonia after surgery. In addition, 4 of the 8 dogs with laryngeal paralysis and aspiration pneumonia also had megaesophagus or esophageal dysmotility. The remaining 5 of 13 dogs that developed aspiration pneumonia after surgery did not have confirmed megaesophagus, esophageal dysmotility, or laryngeal paralysis, but had laryngeal collapse (n = 2, including 1 with concurrent elongated soft palate), tracheal collapse (1), laryngeal mass (1), or laryngeal trauma (1). Of the 13 dogs with aspiration pneumonia, 9 dogs had only a single episode of aspiration pneumonia, whereas 4 dogs had 2 episodes. Median duration between receiving a PT and the first episode of aspiration pneumonia was 59 days (range, 1 to 522 days). In 3 dogs, aspiration pneumonia was diagnosed on the basis of radiographic evidence 1 day (n = 2) or 3 days (1) after surgery for a PT, and it was unclear whether the episodes were truly postoperative episodes or could have occurred before surgery. A second episode of aspiration pneumonia occurred in 4 dogs at a median duration of 329 days (range, 100 to 948 days) after receiving a PT. In addition, the medical records of 52 dogs indicated whether dogs had (n = 6) or did not have (42) megaesophagus, and 4 of the 6 dogs with megaesophagus developed aspiration pneumonia after surgery, compared with 6 of the 46 (13%) dogs without megaesophagus. Dogs with megaesophagus were significantly (OR, 13.3; 95% CI, 2.0 to 89.3; P = 0.01) more likely to develop aspiration pneumonia after surgery than were dogs without megaesophagus. None of the 3 dogs that had their glottis sutured closed when they received a PT developed aspiration pneumonia after surgery.
Skinfold occlusion—Skinfold occlusion of the stoma occurred in 13 of the 69 (19%) dogs, and 11 of these 13 dogs had undergone skinfold resection at the time of surgery for a PT. Ten of these 13 dogs underwent a revision surgery to correct the skinfold occlusion of the stoma, and all 10 had also undergone skinfold resection at the time of the initial surgery for a PT. Three dogs with skinfold occlusion of the stoma did not undergo revision surgery to correct it because the owners indicated that they believed that the occlusion was not bad enough to require additional surgery (n = 2) or because of unknown reasons (1). Median duration from receiving a PT to undergoing the first revision surgery for dogs with skinfold occlusion was 99 days (range, 1 to 1,434 days). Three dogs underwent revision surgery for skinfold occlusion 1 day (n = 1) and 3 days (2) after receiving a PT. All 3 of these dogs had skin resected when they received a PT, but skinfold occlusion of the stoma still occurred after surgery. In addition, skinfold occlusion was the reason for a second revision surgery in 2 dogs, 1 that had and 1 that had not undergone previous revision surgery because of skinfold occlusion. Skinfold resection during the surgery for a PT was performed in 38 of the 69 (55%) dogs, not performed in 27 (39%), and unknown because it was not mentioned in the medical records of 4 (6%). Eleven of the 38 (29%) dogs that had skinfolds resected at the time of surgery for a PT developed skinfold occlusion of the stoma, whereas only 2 of 27 (7%) dogs that did not have skinfolds resected at the time of surgery for a PT developed skinfold occlusion of the stoma. Dogs that had skinfolds resected at the time of surgery for a PT were significantly (OR, 5.1; 95% CI, 1.0 to 25.3; P = 0.03) more likely to develop skinfold occlusion of the stoma than were dogs that did not have skinfolds resected at the time of surgery for a PT.
Stenosis of the stoma—Stenosis of the stoma occurred in 12 of 69 (17%) dogs, and 11 of these 12 dogs underwent revision surgery. The remaining dog did not undergo revision surgery and was euthanized because of the combination of stoma stenosis and overall poor health. Median duration between receiving a PT and revision surgery for stenosis of the stoma was 147 days (range, 30 to 3,334 days). Mean ± SEM age was significantly (P = 0.03) younger for dogs with stenosis of the stoma (5.5 ± 1.2 years), compared with dogs without stenosis of the stoma (8.1 ± 0.5 years). However, there was no meaningful difference in the median number of tracheal rings removed from dogs that did (3 rings; range, 1 to 5 rings) versus did not (4 rings; range, 1 to 7 rings) develop stenosis of the stoma.
Survival times
At study end, 16 of the 69 (23%) dogs were lost to follow-up at a median of 143 days (range, 22 to 1,690 days) after surgery for a PT, 8 (1%) dogs were still alive with a median follow-up time of 745.5 days (range, 259 to 1,608 days), and 45 (65%) dogs were deceased with a median follow-up time of 354 days (range, 14 to 2,663 days). After censoring, the overall median survival time was 1,825 days (range, 14 to 2,663 days). Median survival time was significantly (P = 0.02) shorter for dogs with tracheal collapse (median survival time, 1,336 days; n = 12) than for dogs without tracheal collapse (median survival time not reached; 51; Figure 1). In addition, median survival time was significantly (P = 0.02) shorter for dogs that received corticosteroids before receiving a PT (1,475 days; n = 32) than for dogs that had not received corticosteroids before receiving a PT (1,825 days; 29; Figure 2). However, there was no meaningful difference in the number of dogs that received corticosteroids when considered on the basis of underlying diagnosis. Older dogs had a significantly (P = 0.02) higher hazard of nonsurvival than younger dogs, in that for every 1-year increase in age, the risk of nonsurvival over the study period increased by 20% (hazard ratio, 1.20; 95% CI, 1.03 to 1.41).
Kaplan-Meier survival curves for 63 client-owned dogs with (n = 12, dashed line) or without (51, solid line) tracheal collapse that survived ≥ 2 weeks after surgery for a PT between January 2002 and June 2016 and for which cause of death was known. Steps on each curve represent the death of ≥ 1 dog.
Citation: Journal of the American Veterinary Medical Association 254, 9; 10.2460/javma.254.9.1086
Kaplan-Meier survival curves for 63 client-owned dogs with (n = 12, dashed line) or without (51, solid line) tracheal collapse that survived ≥ 2 weeks after surgery for a PT between January 2002 and June 2016 and for which cause of death was known. Steps on each curve represent the death of ≥ 1 dog.
Citation: Journal of the American Veterinary Medical Association 254, 9; 10.2460/javma.254.9.1086
Kaplan-Meier survival curves for 63 client-owned dogs with (n = 12, dashed line) or without (51, solid line) tracheal collapse that survived ≥ 2 weeks after surgery for a PT between January 2002 and June 2016 and for which cause of death was known. Steps on each curve represent the death of ≥ 1 dog.
Citation: Journal of the American Veterinary Medical Association 254, 9; 10.2460/javma.254.9.1086
Kaplan-Meier survival curves for 61 client-owned dogs that did (n = 32; solid line) or did not (29; dashed line) receive treatment with corticosteroids before surgery for a PT between January 2002 and June 2016 and that survived ≥ 2 weeks after surgery. Steps on each curve represent the death of ≥ 1 dog.
Citation: Journal of the American Veterinary Medical Association 254, 9; 10.2460/javma.254.9.1086
Kaplan-Meier survival curves for 61 client-owned dogs that did (n = 32; solid line) or did not (29; dashed line) receive treatment with corticosteroids before surgery for a PT between January 2002 and June 2016 and that survived ≥ 2 weeks after surgery. Steps on each curve represent the death of ≥ 1 dog.
Citation: Journal of the American Veterinary Medical Association 254, 9; 10.2460/javma.254.9.1086
Kaplan-Meier survival curves for 61 client-owned dogs that did (n = 32; solid line) or did not (29; dashed line) receive treatment with corticosteroids before surgery for a PT between January 2002 and June 2016 and that survived ≥ 2 weeks after surgery. Steps on each curve represent the death of ≥ 1 dog.
Citation: Journal of the American Veterinary Medical Association 254, 9; 10.2460/javma.254.9.1086
Cause of death
Cause of death was recorded in the medical records for 40 of the 69 (58%) dogs. Of these 40 dogs, 16 (40%) died or were euthanized for respiratory-related causes (worsening respiratory disease [n = 6], acute obstruction of the stoma from a mucus plug [3], stoma collapse [2], stoma stenosis [2], drowning [2], and infection of the stoma site [1]), whereas 24 (60%) died or were euthanized for conditions unrelated to their respiratory disease (general health decline [7], neoplasia [5], neurologic disease [4], cardiopulmonary arrest while undergoing a general anesthetic procedure for unrelated conditions [2], trauma [2], infection [2], and gastrointestinal disease [2]). Of the 40 dogs with known causes of death, 6 of 7 dogs with tracheal collapse died or were euthanized because of respiratory-related causes, compared with 10 of 32 (31%) dogs without tracheal collapse. In addition, 2 dogs with tracheal collapse that later died from respiratory-related causes also had extraluminal tracheal support ring prostheses placed before receiving a PT. Dogs with tracheal collapse were significantly (OR, 14.7; 95% CI, 1.5 to 139.8; P = 0.01) more likely to die of respiratory-related causes than were dogs without tracheal collapse.
Discussion
Results of the present study indicated that for dogs surviving ≥ 2 weeks after receiving a PT, long-term survival was possible. Despite major complications (most commonly aspiration pneumonia, skinfold occlusion of the stoma, and stenosis of the stoma) that occurred in 42 of 69 (61%) dogs in the present study, overall median survival time was 1,825 days, which further suggested that dogs could do well for a long time after receiving a PT.
Revision surgery was performed in 24 of 69 (35%) dogs, with few dogs needing a second or third revision surgery. The most common reasons for revision surgery were skinfold occlusion or stenosis of the stoma. The findings in the present study that brachycephalic dogs were more likely to have required revision surgery, but were not more likely to have had skinfold occlusion or stenosis of the stoma, compared with nonbrachycephalic dogs, could have been a type II error owing to low numbers of dogs in each group. Brachycephalic dogs tend to have short necks with excessive skin and often have hypoplastic tracheas that may predispose them to skinfold occlusion and stenosis, respectively, of the stoma.
In addition, the finding in the present study that skinfold occlusion of the stoma was associated with dogs having had skin resected at the time of surgery for a PT likely reflected that surgeons suspected these same dogs of having had increased risk of developing skinfold occlusion and therefore preemptively excised skin to mitigate that risk. Although skin resection can be a temporary solution, as the skin heals, elasticity returns and may lead to recurrence of occlusion over time.9 Further, dogs in the present study with excessive skinfolds could have had inadequate amounts of skin resected. For instance, 3 dogs that had skin resected when they received a PT developed skinfold occlusion and then underwent revision surgery within 3 days after the initial surgery. Surgeons should be aware of this potential complication when preparing patients for a PT. Before surgery and to help guide resection, affected dogs could be observed in standing position so that the edges of the skinfolds may be marked with a permanent marker. In addition, cervical region skin should be pulled ventrally, liberally shaved, aseptically prepared, and draped into the surgical field to allow for adequate resection, if indicated.
Stenosis of the stoma was more common in younger dogs than in older dogs of the present study, and this finding could have been because the healing capacity of younger animals is greater than that of older animals.10 In addition, stenosis could have also been related to the number of tracheal rings removed; however, results of the present study suggested that this association was not meaningful. Two dogs had only 1 tracheal ring removed for their PT, and although one dog developed stenosis, the other did not. The recommended number of tracheal rings removed for a PT ranges from 2 to 4 rings,2,11 consistent with the median number of tracheal rings removed for all dogs in the present study (4 rings; range, 1 to 7 rings). The reason for removing only 1 tracheal ring could not be determined retrospectively in the present study, and the authors believe that removal of only 1 ring may not provide an adequate stoma size. In dogs with > 4 tracheal rings removed, the reason for removing > 4 rings was likely an intraoperative decision when removal of fewer did not appear adequate to provide a large enough stoma.
Collapse of the stoma was an uncommon reason for revision surgery (5/24 [21%]) in the present study. Placement of extraluminal tracheal support ring prostheses cranial and caudal to the stoma site may help to relieve signs of tracheal collapse, but collapse at the stoma itself may still occur.4 In the present study, tracheal collapse had been diagnosed in 12 dogs, 1 of which received extraluminal tracheal support ring prostheses at the time of surgery for a PT and 4 of which received the prostheses during earlier surgeries. It was unknown why extraluminal support was not placed along the entire length of the cervical aspect of the trachea in all dogs with tracheal collapse. It was possible that the surgeons suspected that collapse would not have occurred at the PT site, thus reducing the need for extraluminal support. In addition, the number of tracheal rings removed could have influenced the decision. Although the number of tracheal rings removed to create the PT did not differ substantially between dogs that did versus did not receive extraluminal tracheal support ring prostheses in the present study, clinically, the authors appreciated that longer stomas may have been more likely to have collapsed, and a similar clinical appreciation by the surgeons who treated dogs in the present study could have influenced their decisions. Further, the decision to use extraluminal tracheal support ring prostheses to prevent stoma collapse could have been influenced by the circumferential amount of tracheal rings removed, which could not be ascertained from the medical records.
Aspiration pneumonia was one of the most common major postoperative complications, occurring in 13 of 69 (19%) dogs at a median of 59 days after receiving a PT. Because it was unknown whether all dogs in the present study underwent thoracic radiography to evaluate for aspiration pneumonia after receiving a PT, the number of dogs with aspiration pneumonia could have been underestimated. Additionally, the 3 dogs with aspiration pneumonia diagnosed within 3 days after receiving a PT could have had aspiration pneumonia before surgery, resulting in a misclassification. Despite reports that indicate aspirated contents are expelled through the stoma in dogs with PTs,2 findings that dogs with megaesophagus were more likely to develop aspiration pneumonia in the present study supported the theory that not all aspirated contents may be expelled through the stoma. In addition, the finding that 8 of the 40 (20%) dogs with laryngeal paralysis developed aspiration pneumonia after receiving a PT was similar to recent studies7,12,13 that report between 19% and 32% of dogs with laryngeal paralysis develop aspiration pneumonia after undergoing unilateral arytenoid lateralization. Some surgeons suture the glottis closed when performing surgery for a PT, which would eliminate aspiration of gastrointestinal contents as a complication; however, only 3 dogs had their glottis sutured closed in the present study. Further investigation is warranted to evaluate whether suturing the glottis closed as a component of surgery for a PT reduces the risk of aspiration pneumonia. Because air is no longer warmed and filtered by the upper airway in dogs with PTs, aspiration pneumonia may also occur subsequent to direct inhalation of foreign material through the stoma. Therefore, care should be taken to protect the stoma from irritants and foreign material.
In the present study, factors affecting the survival time of dogs after receiving a PT included the presence of tracheal collapse and previous medical treatment with corticosteroids. Dogs with tracheal collapse in the present study had a shorter median survival time than did dogs without tracheal collapse and were more likely to die of respiratory-related causes than nonrespiratory-related causes. Dogs with tracheal collapse have weakened tracheal cartilages attributed to loss of proteoglycans and water, and their tracheal cartilage may not be able to support the tracheostoma.1 Additionally, tracheal collapse can still occur with a PT, particularly if extraluminal tracheal support ring prostheses or tracheal stents are not used to prevent collapse. This underscores the fact that creation of a PT is a palliative measure and does not correct the underlying respiratory disease. The shorter survival time associated with previous treatment with corticosteroids for dogs in the present study could have been related to a frequent use of corticosteroids in medical management of dogs with tracheal collapse. Although on the basis of underlying diagnosis there was no difference in the number of dogs that did or did not receive corticosteroids before receiving a PT, this may have been a type II error. In addition, because corticosteroids are frequently administered to treat various respiratory conditions, the results that suggested an association between corticosteroid treatment and survival time could have been spurious findings. No other previous medical treatments were meaningfully associated with survival time of dogs in the present study, which indicated that medical management of the underlying condition could be attempted until a PT is necessary.
The most common reasons for PTs in dogs of the present study were laryngeal paralysis and laryngeal collapse, consistent with previous studies.2,5 However, a higher proportion of dogs received a PT for treatment of laryngeal paralysis in the present study than has been previously reported.5 Surgeon preference could have accounted for the difference. In addition, results of the present study were inconsistent with previous studies2,7 that show that unilateral arytenoid lateralization may increase the risk of aspiration pneumonia and that the risk of aspiration with a PT is low. Future studies should be done to compare the risk of aspiration pneumonia in dogs with laryngeal paralysis undergoing unilateral arytenoid lateralization with that of dogs receiving a PT and to determine whether closure of the glottis reduces the risk of aspiration pneumonia in dogs with PTs.
Before surgery for a PT, 31 of the 69 (45%) dogs in the present study had undergone other surgical treatments (most commonly temporary tracheostomy, unilateral arytenoid lateralization, and multimodal surgery for brachycephalic obstructive airway disease) for their respiratory disease. However, results indicated that complications after receiving a PT did not meaningfully differ between dogs with versus without a temporary tracheostomy, between dogs that had versus had not undergone unilateral arytenoid lateralization, or between dogs that had versus had not undergone staphylectomy.
Older dogs in the present study had significantly higher hazard of nonsurvival after receiving a PT than did younger dogs, and this finding could have been associated with a typically shorter remaining lifespan of older animals, compared with younger animals. In addition, 40 of the 69 (58%) dogs in the present study had laryngeal paralysis, classically associated with geriatric onset laryngeal paralysis polyneuropathy, which may lead to a shorter lifespan because of the associated progressive polyneuropathy.7 Although treatment with creation of a PT is palliative, it should be considered a viable option in younger dogs because, in the present study, younger dogs had a lower hazard of nonsurvival after receiving a PT than did older dogs.
Limitations of the present study were related to its multi-institutional retrospective nature. Various intraoperative techniques and postoperative care could not be specifically evaluated. For instance, information regarding peristomal skin resection and peritracheal suturing was not specifically defined in all surgery reports. Although the authors of the present study assumed these procedures were performed as part of the standard technique of surgery for a PT, this could not be confirmed with complete accuracy. Additionally, for dogs that had peristomal skin resection, as opposed to skinfold resection, the amount of skin removed from each dog could not be quantified. Similarly, the number of sutures placed from the subcutaneous tissues to the peritracheal tissues in each dog could not be confirmed. Additionally, owners and referring veterinarians were contacted to obtain follow-up information, and information obtained from owners may have been less reliable. Although this represents a large collection of data on dogs with PTs, when divided according to variables of interest (eg, dogs with tracheal collapse vs dogs without tracheal collapse) overall numbers of dogs in resulting specific groups were small, which could have affected statistical analyses. A prospective study with consistent surgical technique and postoperative care is necessary to further define risk factors associated with survival and complications in dogs with PTs.
Creation of a PT is a valid treatment for severe obstructive upper airway disease in dogs. Results of the present study indicated that older dogs and dogs with tracheal collapse had shorter survival times after receiving a PT than did younger dogs or dogs without tracheal collapse, respectively. Results also suggested that when compared with dogs without megaesophagus, dogs with megaesophagus were more likely to develop aspiration pneumonia after receiving a PT. The proportion of dogs in the present study that developed aspiration pneumonia after receiving a PT was comparable with the proportions of dogs reported7,11,12 to develop aspiration pneumonia after undergoing unilateral arytenoid lateralization for laryngeal paralysis, which indicated that PT may not reduce the risk of aspiration pneumonia as previously reported. Further, results suggested that brachycephalic dogs, compared with nonbrachycephalic dogs, were more likely to require revision surgery. When performing skin resection for a PT, aggressive skin resection should be performed to minimize the risk of needing revision surgery later. Despite these factors, results of the present study indicated that long-term outcomes were good, with overall median survival time of 5 years for dogs surviving ≥ 2 weeks after receiving a PT.
Acknowledgments
The authors thank Tara Denley for technical assistance.
The authors declare that there were no conflicts of interest.
ABBREVIATIONS
| CI | Confidence interval |
| PT | Permanent tracheostomy |
References
1. Sura PA, Durant AM. Trachea and bronchi. In: Tobias KM, Johnston SA, eds. Veterinary surgery small animal. 2nd ed. St Louis: Elsevier Saunders, 2012;1739–1751.
2. Hedlund CS, Tangner CH, Waldron DR, et al. Permanent tracheostomy: perioperative and long-term data from 34 cases. J Am Anim Hosp Assoc 1988;24:585–591.
3. Hedlund CS, Tangner CH, Montgomery DL. A procedure for permanent tracheostomy and its effects on tracheal mucosa. Vet Surg 1982;11:13–17.
4. Hedlund CS. Tracheostomies in the management of canine and feline upper respiratory disease. Vet Clin North Am Small Anim Pract 1994;24:873–886.
5. Occhipinti LL, Hauptman JG. Long-term outcome of permanent tracheostomies in dogs: 21 cases (2000–2012). Can Vet J 2014;55:357–360.
6. Stepnik MW, Mehl ML, Harie EM. Outcome of permanent tracheostomy for treatment of upper airway obstruction in cats: 21 cases (1990–2007). J Am Vet Med Assoc 2009;234:638–643.
7. Stanley BJ, Hauptman JG, Fritz MC, et al. Esophageal dysfunction in dogs with idiopathic laryngeal paralysis: a controlled cohort study. Vet Surg 2010;39:139–149.
8. Leonard HC. Collapse of the larynx and adjacent structures in the dog. J Am Vet Med Assoc 1960;137:360–363.
9. Held M, Rothenberger J, Tolzmann D, et al. Alteration of biomechanical properties of skin during the course of healing of partial-thickness wounds. Wounds 2015;27:123–128.
10. Shyh-Chang N, Zhu H, Yvanka de Soysa T, et al. Lin28 enhances tissues repair by reprogramming cellular metabolism. Cell 2013;155:778–792.
11. Colley P, Huber M, Henderson R. Tracheostomy techniques and management. Compend Contin Educ Pract Vet 1999;21:44–53.
12. Wilson D, Monnet E. Risk factors for the development of aspiration pneumonia after unilateral arytenoid lateralization in dogs with laryngeal paralysis: 232 cases (1987–2012). J Am Vet Med Assoc 2016;248:188–194.
13. Bahr KL, Howe L, Jessen C, et al. Outcome of 45 dogs with laryngeal paralysis treated by unilateral arytenoid lateralization or bilateral ventriculocordectomy. J Am Anim Hosp Assoc 2014;50:264–272.
