Introduction
In dogs, TC occurs most commonly in small and toy breeds and can result in substantial respiratory difficulty.1 Tracheal collapse can include various disorders, such as tracheal chondromalacia, weakened or redundant dorsal tracheal membrane, and tracheal ring malformations, which result in labored breathing, coughing, and airway obstruction. Tracheal collapse is classified into 2 types: traditional and malformation.2 Traditional-type TC is caused by a combination of progressive chondromalacia and laxity of the trachealis dorsalis muscle, which leads to dynamic airway collapse. Malformation-type TC is a static form of focal obstruction of the tracheal lumen at the thoracic inlet caused by the tracheal rings at that location becoming invaginated such that they resemble the shape of a W rather than a C as is typical.3 Dogs with TC that fail to respond to aggressive multimodal medical management, such as administration of antitussives, anti-inflammatories, and antimicrobials, may benefit from the placement of tracheal ring prostheses or endoluminal stents, but both of those surgical options can lead to serious complications.4 For some dogs with TC, scheduled administration of medications can alleviate clinical signs of the disease for over a year.5 Currently, it is unclear how to distinguish between patients with TC that will and will not respond to multimodal medical management at the time of diagnosis. Therefore, it is challenging to provide owners of dogs with TC prognostic information to help inform treatment decisions.
The goal of the retrospective study reported here was to compare the short-, intermediate-, and long-term results between dogs with TC that received multimodal medical management only and those that underwent tracheal endoluminal stent placement (stent placement). We hypothesized that dogs with TC that were prescribed a multimodal medical management regimen would have improved (lower) clinical scores in the short and intermediate terms but would eventually become refractory to the medical regimen and develop progressive disease. We compared the outcomes for those dogs with the outcomes for a group of dogs with TC that underwent stent placement as described in another study.4 All dogs were examined over the same period by the IR service at a referral veterinary hospital. We hypothesized that dogs with TC that received multimodal medical management only would have higher clinical scores (ie, more severe clinical signs) and a shorter MST from TC diagnosis, compared with dogs with TC that underwent stent placement, and that dogs with a high clinical score at the time of TC diagnosis would have an improved outcome if they underwent stent placement as soon as possible.
Materials and Methods
Animals
All dogs with TC that were evaluated by the IR service of The Animal Medical Center in New York City between September 1, 2009, and August 1, 2018, were considered for study enrollment. Multimodal medical management of TC was recommended for each patient before stent placement was considered. In general, multimodal medical management consisted of administration of a combination of antitussives, anti-inflammatories, antimicrobials, and anxiolytics or sedatives; was adjusted as necessary dependent on the severity of the clinical signs; and was similar between patients that did and did not undergo stent placement. Stent placement was not recommended for patients with mild or moderate signs of TC and for which multimodal medical management had not been attempted or those with inadequately controlled comorbidities.
Dogs with TC that were managed by multimodal medical treatment only were included in the MM group unless they had undergone a previous tracheal surgery (eg, placement of tracheal ring prostheses or endoluminal stent), had < 2 sets of clinical scores available for evaluation, or were not reevaluated by the IR service > 3 weeks after the initial examination. None of the dogs enrolled in the MM group required stent placement or other surgical intervention for management of TC during the observation period.
Seventy-five dogs that underwent stent placement to alleviate the signs of TC after failing to respond or becoming refractory to multimodal medical management were included in the SM group for comparison with the dogs of the MM group. All dogs in the SM group were examined and treated by the same IR service as the dogs in the MM group and were described in another study.4 The dogs of the SM group frequently received antitussives, corticosteroids (generally a tapering dosage that was discontinued when possible), and antimicrobials (generally for 10 to 14 days unless the dog developed pneumonia, in which case antimicrobial administration was extended to 4 to 6 weeks) in addition to stent placement.
Medical records review
For each dog eligible for study inclusion, information obtained from the medical record was aggregated into 5 periods for analysis (initial evaluation by the IR service; short-term [first 45 days after TC diagnosis], intermediate-term [46 to 180 days after TC diagnosis], and long-term [> 180 days after TC diagnosis] follow-up periods; and final evaluation). Information collected for the initial evaluation included the hospital service of admission and patient history, including duration (unknown or unclear, < 1 year, 1 to 3 years, or > 3 years) and severity of clinical signs and previous medical treatment; physical examination and diagnostic imaging (eg, thoracic radiography, fluoroscopy, tracheoscopy, and echocardiography) findings; comorbidities; presence of bronchial collapse or compression (yes or no); medical regimen prescribed; changes in medical treatments; and type of TC (traditional or malformation). For dogs in the SM group, date of stent placement was also extracted from the record. The type of TC was determined from evaluation of thoracic radiographs as described.6,7,8 Common comorbidities recorded included heart murmurs, cardiomegaly, pulmonary hypertension, pneumonia, and cardiac valvular disease (ie, chronic degenerative valvular disease, mitral valve insufficiency, and tricuspid valve insufficiency). Multimodal medical management was modified for each patient on the basis of clinical signs and included combinations of antitussives, corticosteroids, anxiolytics or sedatives, antimicrobials, and bronchodilators.
For each patient during each examination performed by the IR service, the owner was asked to complete a questionnaire4 regarding the severity of the patient's clinical signs. Briefly, goose-honking or raspy breathing, coughing, and dyspnea were each assessed on a scale from 0 to 10. Owners were advised that a score of 0 indicated that the clinical sign under consideration was never observed, whereas a score of 10 indicated that the clinical sign was persistent or intractable.
Following diagnosis of TC, medical records regarding any visit related to TC were requested and obtained from the owner or primary care veterinarian. Information extracted from those records included data regarding severity of clinical signs, medications administered, examination frequency, complications (incidences of pneumonia or emergent veterinary visits associated with TC), survival time, and cause of death when available. The final evaluation was defined as the last known examination by the IR service or the last point of contact with the patient owner.
Data analysis
Descriptive data were generated for the MM and SM groups. For analysis purposes, cause of death was categorized into 4 categories (TC [confirmed by necropsy], suspected TC [not confirmed by necropsy], TC unlikely [not confirmed by necropsy], and not TC [confirmed by necropsy]). Also, the dogs of the MM group were grouped into 4 categories (0, 1, 2, or 3) on the basis of how many severe scores they were assigned at the initial evaluation as well as the final evaluation. A severe score was defined as a score ≥ 8 for any 1 of the 3 clinical signs assessed (goose-honking or raspy breathing, coughing, and dyspnea) and was considered consistent with severe and potentially life-threatening disease. Normally distributed continuous variables were summarized as the mean ± SD, and nonnormally distributed variables were summarized as the median (range). The frequency distribution was reported for each categorical variable.
Normally distributed dependent variables were compared between dogs of the MM and SM groups by use of ANOVA. The distribution of the error residuals was assessed for normality with the Kolmogorov-Smirnoff test, and all variables assessed by ANOVA were deemed to be normally distributed. The Wilcoxon rank sum test was used to compare nonnormally distributed continuous variables between dogs of the MM and SM groups. For categorical variables, proportions were compared between dogs of the MM and SM groups by means of the χ2 test or the Fisher exact test when individual cell counts were ≤ 5.
The MSTs and associated 95% CIs were calculated for dogs with various characteristics within the MM group. The mean age for the dogs of the MM group was used as the cutoff to define dichotomous categories for calculation of the MSTs for age, males, and females. Other variables were dichotomized as present or absent for calculation of the MST.
Additionally, for dogs of the MM group, the Kaplan-Meier method was used to compare the survival time from initial evaluation for TC by the IR service with that from initial TC diagnosis by the referring or primary care veterinarian and to compare survival times among dogs with severe clinical scores of 0, 1, 2, and 3 at the initial evaluation and final evaluation by the IR service. The Kaplan-Meier method was also used to compare survival times between dogs of the MM and SM groups. For all Kaplan-Meier analyses, dogs lost to follow-up were censored on the most recent date that they were known to be alive. All analyses were performed with commercially available statistical software,a and values of P < 0.05 were considered significant.
Results
Dogs
One hundred eighty-five dogs were evaluated for TC by the IR service at the referral hospital between September 1, 2009, and August 1, 2018, of which 110 (59%) received multimodal medical management only and 75 (41%) underwent stent placement in addition to multimodal medical management. Twenty-six of the 110 (24%) dogs that received multimodal medical management only were excluded from inclusion in the MM group because of loss to follow-up owing to incomplete medical records from the referring veterinarian or inability to contact the owner because of an invalid telephone number in the patient record. Thus, only 84 dogs were included in the MM group and are described in detail here. The 75 dogs that underwent stent placement were evaluated in the SM group and are described in detail elsewhere.4
The 84 dogs of the MM group included 45 castrated males (54%), 2 (2%) sexually intact males, and 37 (44%) spayed females. The group had a mean ± SD age of 8.8 ± 3.7 years (range, 1.1 to 17.7 years) and body weight of 4.4 ± 2.6 kg (9.7 ± 5.7 lb; range, 1.3 to 17.0 kg [2.9 to 37.4 lb]; Table 1) and a median body condition score of 5/9 (range, 3/9 to 9/9) at the time of initial evaluation by the IR service (baseline). Fifteen breeds were represented in the group, with Yorkshire Terrier (n = 33 [39%]) and Pomeranian (20 [24%]) being the most common.
Descriptive statistics for 159 dogs with TC that were treated by multimodal medical management alone (MM group; n = 84) or underwent tracheal endoluminal stent placement after failing to respond or becoming refractory to multimodal medical management (SM group; 75).
| Variable | Category | MM group | SM group |
|---|---|---|---|
| Age (y) | — | 8.8 ± 3.7 | 8.3 ± 3.4 |
| Weight (kg) | — | 4.4 ± 2.6* | 3.5 ± 1.3 |
| Sex | Sexually intact male | 2 (2) | 3 (4) |
| Castrated male | 45 (54) | 42 (56) | |
| Sexually intact female | 0 (0) | 3 (4) | |
| Spayed female | 37 (44) | 27 (36) | |
| Type of TC | TTC | 70 (83)* | 37 (49) |
| MTC | 5 (6)* | 38 (51) | |
| Undetermined | 9 (11)* | 0 (0) | |
| Clinical score for goose-honking or raspy breathing at baseline | — | 5 (0–10)* | 6 (0–10) |
| Clinical score for cough at baseline | — | 4.8 (0–10) | 4 (0–10) |
| Clinical score for dyspnea at baseline | — | 1.8 (0–10)* | 6 (0–10) |
| Clinical score for goose-honking or raspy breathing at final evaluation | — | 3 (0–9) | 2 (0–10) |
| Clinical score for cough at final evaluation | — | 3 (0–9) | 3 (0–10) |
| Clinical score for dyspnea at final evaluation | — | 2 (0–9) | 2 (0–10) |
Values represent the mean ± SD, number (percentage), or median (range). All dogs were treated by the IR service of a veterinary referral hospital between September 1, 2009, and August 1, 2018. The initial evaluation by the IR service of the referral hospital was considered baseline. At each evaluation, the owner of each dog was asked to score each of 3 clinical signs (goose-honking or raspy breathing, coughing, and dyspnea) on a scale of 0 to 10, where 0 indicated that the clinical sign was never observed and 10 indicated that the clinical sign was persistent or intractable.
Within a row, value differs significantly (P < 0.05) from that for the SM group.
Tracheal collapse was diagnosed by the primary care or referring veterinarian for 34 of the 84 (40%) dogs of the MM group. Tracheal collapse was diagnosed by a veterinarian of the IR service at the referral hospital for the remaining 50 (60%) dogs. The median duration between initial TC diagnosis by the primary care or referring veterinarian and initial evaluation by the IR service was 87 days (mean, 228 days; range, 0 to 1,825 days). The type of TC was categorized as TTC for 70 dogs and MTC for 5 dogs and was undetermined for the remaining 9 dogs. Among the dogs that were > 8.8 years old (the mean age for the group), the proportion with TTC (39/40) was significantly (P < 0.01) greater than the proportion with MTC (1/40). Among the 35 dogs that were ≤ 8.8 years old, 31 (89%) had TTC and 4 (11%) had MTC, but this difference was not significant (P = 0.10). During evaluation of thoracic radiographs, 32 of the 84 (38%) dogs had evidence of mainstem bronchial collapse or compression.
Initial evaluation by the IR service
Of the 84 dogs of the MM group, 66 (79%) were initially evaluated by the IR service, 12 (14%) were initially evaluated by the emergency service, and 6 (7%) were initially evaluated by another service at the referral hospital. Sixty-seven (80%) and 17 (20%) of the 84 dogs did and did not receive medical treatment, respectively, before the initial evaluation by the IR service (baseline). The number of medications received prior to baseline by the dogs of both the MM and SM groups was summarized elsewhere (Supplementary Table S1, available at: avmajournals.avma.org/doi/suppl/10.2460/javma.258.3.279). The median duration of medication administration before baseline did not differ significantly (P = 0.80) between dogs with TTC and MTC and was 86.5 days (mean, 228 days; range, 0 to 1,825 days) for all dogs of the MM group.
Among the 84 dogs of the MM group, the duration of clinical signs associated with TC before baseline was unknown or unclear for 3 (4%), < 1 year for 50 (60%), between 1 and 3 years for 25 (30%), and > 3 years for 6 (7%). Thirty-six (43%), 35 (42%), 27 (32%), 10 (12%), and 6 (7%) dogs had a history of cardiomegaly, heart murmur, cardiac valvular disease, pneumonia, and pulmonary hypertension, respectively. The median clinical score at baseline was 5 for goose-honking or raspy breathing, 4.8 for coughing, and 1.8 for dyspnea. Multimodal medical management was prescribed for all 84 dogs of the MM group after the baseline examination. Eighty-one (96%) dogs received antitussives, 43 (51%) received corticosteroids, 27 (32%) received anxiolytics or sedatives, and 28 (33%) received antimicrobials; most dogs received multiple medication types.
Short-term follow-up
Eight of the 84 (10%) dogs of the MM group were reevaluated by the IR service during the first 45 days after baseline (ie, during the short-term follow-up period). Medical records were requested from the referring veterinarians for the remaining 76 dogs and were received for 75. Thus, short-term follow-up data were available for 83 of the 84 dogs.
Commonly administered medications during the short-term follow-up period included antitussives (78/83 [94%]), corticosteroids (51/83 [61%]), anxiolytics or sedatives (28/83 [34%]), and antimicrobials (35/83 [42%]). Clinical score data for the short-term follow-up period were available for 8 dogs. The median clinical scores for those 8 dogs were 1.5 for goose-honking or raspy breathing, 0 for coughing, and 0 for dyspnea. Compared with baseline, goose-honking or raspy breathing and coughing had lessened in severity (improved) for 7 of the 8 dogs and worsened for 1, whereas dyspnea had improved for 5 dogs, worsened for 1 dog, and remained unchanged for 2 dogs.
Data regarding the incidence of pneumonia during the short-term follow-up period were available for 79 of the 84 dogs of the MM group. Pneumonia was diagnosed during the short-term follow-up period for 3 of the 79 (4%) dogs. Data regarding the number of emergent visits were available for 70 of 84 (83%) dogs. Three of 70 (4%) dogs underwent an emergency examination during the short-term follow-up period. None of the dogs in the MM group died during the short-term follow-up period.
Intermediate-term follow-up
Twenty-one of the 84 (25%) dogs of the MM group were reevaluated by the IR service between 46 and 180 days after baseline (ie, during the intermediate-term follow-up period). Medical records were requested from the referring veterinarians for the remaining 63 dogs and were received for 61. Thus, intermediate-term follow-up data were available for 82 of the 84 dogs.
Commonly administered medications during the intermediate-term follow-up period included antitussives (71/82 [87%]), corticosteroids (39/82 [48%]), anxiolytics or sedatives (27/82 [33%]), and antimicrobials (14/82 [17%]). Clinical score data for the intermediate-term follow-up period were available for 21 dogs. The median clinical scores for those 21 dogs were 3 for goose-honking or raspy breathing, 3.5 for coughing, and 0.5 for dyspnea. Compared with baseline, goose-honking or raspy breathing had improved for 12 of the 21 dogs, worsened for 6, and remained static for 3; coughing had improved for 16, worsened for 4, and remained static for 1; and dyspnea had improved for 9, worsened for 3, and remained static for 9.
Data regarding the incidence of pneumonia during the intermediate-term follow-up period were available for 78 of the 84 (93%) dogs of the MM group. Pneumonia was diagnosed during the intermediate-term follow-up period for 1 of the 78 (1%) dogs. Data regarding the number of emergent visits were available for 70 of 84 (83%) dogs. Three of 70 (4%) dogs underwent an emergency examination during the intermediate-term follow-up period. One of the dogs in the MM group died during the intermediate-term follow-up period.
Long-term follow-up
Seventy-one of the 83 (86%) surviving dogs of the MM group were reevaluated by the IR service > 180 days after baseline (ie, during the long-term follow-up period). Medical records were requested from the referring veterinarians for the remaining 12 dogs and were received for 11. Thus, long-term follow-up data were available for 82 of the 83 surviving dogs. Dogs were examined during the long-term follow-up period at a median of 838 days (mean, 1,028 days; range, 106 to 2,616 days) after baseline.
Commonly administered medications during the long-term follow-up period included antitussives (74/82 [90%]), corticosteroids (56/82 [68%]), anxiolytics or sedatives (34/82 [41%]), and antimicrobials (29/82 [35%]). Clinical score data for the long-term follow-up period were available for 72 dogs. The median clinical scores for those 72 dogs were 4 for goose-honking or raspy breathing, 3 for coughing, and 2 for dyspnea. Compared with baseline, goose-honking or raspy breathing had improved for 41 of the 72 dogs, worsened for 26, and remained static for 5; coughing had improved for 36, worsened for 29, and remained static for 7; and dyspnea had improved for 23, worsened for 31, and remained static for 18.
Data regarding the incidence of pneumonia during the long-term follow-up period were available for all 83 surviving dogs of the MM group. Pneumonia was diagnosed during the long-term follow-up period for 4 of the 83 (5%) dogs. Data regarding the number of emergent visits were available for 69 of 83 (83%) dogs. Fifteen of 69 (22%) dogs underwent an emergency examination during the long-term follow-up period. Forty-eight dogs of the MM group died during the long-term follow-up period.
Final evaluation
For the 84 dogs of the MM group, the final evaluation for TC occurred at a median of 743 days (mean, 936 days; range, 28 to 2,616 days) after baseline. Information regarding the medications administered and clinical scores at the time of the final evaluation were available for all 84 dogs.
Medications commonly administered at the time of the final evaluation included antitussives (67/84 [80%]), corticosteroids (42/84 [50%]), anxiolytics or sedatives (21/84 [25%]), and antimicrobials (17/84 [20%]). The median clinical scores at the final evaluation were 3 for goose-honking or raspy breathing, 3 for coughing, and 2 for dyspnea. Compared with baseline, goose-honking or raspy breathing had improved for 49 of the 84 dogs, worsened for 31, and remained static for 4; coughing had improved for 43, worsened for 29, and remained static for 12; and dyspnea had improved for 29, worsened for 40, and remained static for 15. Aggregation of the final evaluation data for all 84 dogs of the MM group revealed that pneumonia was diagnosed in 8 (10%) dogs and that 21 (25%) dogs underwent an emergency examination.
Outcome and MSTs
Forty-nine of the 84 (58%) dogs of the MM group died during the study period. For those 49 dogs, the cause of death was confirmed to be TC during necropsy for 2 (4%), suspected to be TC for 26 (53%; ie, necropsy was not performed, or the results were inconclusive), and unlikely to be TC for 21 (43%). Within the MM group, the proportion of dogs with MTC that died (4/5) did not differ significantly (P = 0.27) from the proportion of dogs with TTC that died (39/70).
For the 84 dogs of the MM group, the MST from initial TC diagnosis (1,363 days; 95% CI, 1,156 to 1,728 days) did not differ significantly (P = 0.13) from the MST from baseline (1,124 days; 95% CI, 702 to 1,528 days; Figure 1). All remaining MSTs for the MM group were calculated from baseline (Table 2). Factors associated with a significantly shorter MST included age > 8.8 years (regardless of sex; P < 0.01), cardiac disease (P < 0.01), an emergency examination during the intermediate-term follow-up period (P < 0.01), and diagnosis of pneumonia during the intermediate-term follow-up period (P < 0.02). The MST for dogs ≤ 8.8 years old with MTC (2,474 days) did not differ significantly (P = 0.10) from the MST for dogs ≤ 8.8 years old with TTC (1,528 days). Likewise, the survival time for the dog > 8.8 years old with MTC (3,010 days) did not differ significantly (P = 0.25) from the MST for dogs > 8.8 years old with TTC (893 days).
Kaplan-Meier survival curves for 84 dogs with TC that received multimodal medical management only (MM group) from initial TC diagnosis (red line) and from initial evaluation by an IR service at a referral hospital (baseline; blue line). Dogs lost to follow-up were censored on the most recent date that they were known to be alive. The 2 survival curves did not differ significantly (P = 0.13).
Citation: Journal of the American Veterinary Medical Association 258, 3; 10.2460/javma.258.3.279
Median survival times (95% CIs) from baseline for the dogs of the MM group of Table 1 on the basis of sex, age, and various comorbidities and complications.
| Variable | Category | MST (95% CI; d) |
|---|---|---|
| Males | > 8.8 y | 897 (729–1,301)* |
| ≤ 8.8 y | 2,141 (1,359–2,474) | |
| Females | > 8.8 y | 719 (514–1,299)* |
| ≤ 8.8 y | 2,151 | |
| Age | > 8.8 y | 893 (719–1,204)* |
| ≤ 8.8 y | 2,151 (1,528–2,527) | |
| History of pneumonia | Yes | 1,363 (107–2,474) |
| No | 1,359 (1,124–1,728) | |
| Cardiac disease | Yes | 752 (518–1,299)* |
| No | 1,883 (1,359–3,010) | |
| Cardiomegaly | Yes | 1,204 (752–1,728) |
| No | 1,528 (1,156–2,151) | |
| Yorkshire Terrier | Yes | 1,301 (1,107–1,494) |
| No | 1,359 (1,017–1,700) | |
| Pulmonary hypertension | Yes | 824 (276–1,371) |
| No | 1,363 (1,273–1,452) | |
| Bronchial collapse | Yes | 1,299 (778–1,414) |
| No | 1,528 (1,156–2,151) | |
| Emergency examination during the short-term follow-up period | Yes | 339 (0–710) |
| No | 1,330 (1,151–1,508) | |
| Emergency examination during the intermediate-term follow-up period | Yes | 339 (0–710)* |
| No | 1,363 (1,267–1,458) | |
| Emergency examination during the long-term follow-up period | Yes | 897 (372–1,421) |
| No | 1,363 (1,128–1,597) | |
| Pneumonia during the short-term follow-up period | Yes | 517 |
| No | 1,371 (1,156–1,883) | |
| Pneumonia during the intermediate-term follow-up period | Yes | 517* |
| No | 1,371 (1,124–1,883) | |
| Pneumonia during the long-term follow-up period | Yes | 940 (295–1,363) |
| No | 1,371 (1,156–1,883) |
The mean age for the dogs in the MM cohort was used to define the categories for MST calculations for the male, female, and age variables. The short-term, intermediate-term, and long-term follow-up periods were defined as 0 to 45, 46 to 180, and > 180 days after baseline, respectively. The lack of a 95% CI indicates that only 1 dog was in that category.
Within a variable, the MST differs significantly (P < 0.05) from that for the other category.
Factors not associated with the MST of dogs in the MM group included history of pneumonia, cardiomegaly, pulmonary hypertension, and mainstem bronchial collapse. Also, the MST for Yorkshire Terriers with TC (1,301 days; 95% CI, 1,107 to 1,494 days) did not differ significantly from that for other breeds of dogs (1,359 days; 95% CI, 1,017 to 1,700 days).
Clinical severity of disease was significantly associated with the MST for dogs of the MM group. Compared with the MST from baseline for all dogs of the MM group (1,124 days), the MST for dogs with 0 (1,363 days) severe scores at baseline was significantly (P < 0.01) longer, whereas the MSTs for dogs with 1 (726 days), 2 (446 days), and 3 (55 days) severe scores at baseline were significantly (P < 0.01 for all comparisons) shorter (Figure 2). The MSTs for dogs with 0 (1,114 days), 1 (731 days), and 3 (6 days) severe scores at the final evaluation were significantly (P < 0.01 for all comparisons) shorter than the MST from baseline for all dogs in the MM group.
Kaplan-Meier survival curves for the dogs of Figure 1 that had 0 (blue line), 1 (red line), 2 (green line), or 3 (black line) severe scores at baseline (A) and at the final evaluation (B). At each evaluation, the owner of each dog was asked to score each of 3 clinical signs (goose-honking or raspy breathing, coughing, and dyspnea) on a scale of 0 to 10, where 0 indicated that the clinical sign was never observed and 10 indicated that the clinical sign was persistent or intractable. A severe score was defined as a score ≥ 8 for any 1 of the 3 clinical signs assessed and was considered consistent with severe and potentially life-threatening disease. Fifty-six, 20, 5, and 3 dogs had 0, 1, 2, and 3 severe clinical scores, respectively, at baseline. Sixty, 8, 3, and 13 dogs had 0, 1, 2, and 3 severe clinical scores, respectively, at the final evaluation. Notice that the MST was not reached for dogs with 2 severe scores at the final evaluation. See Figure 1 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 258, 3; 10.2460/javma.258.3.279
Comparisons between the MM and SM groups
Descriptive statistics for both the MM and SM groups were summarized (Table 1). Compared with the dogs of the SM group, the dogs of the MM group had a significantly greater mean body weight (P < 0.01) and lower median clinical score for goose-honking or raspy breathing (P = 0.02) and dyspnea (P < 0.01) at baseline but a greater median clinical score for goose-honking or raspy breathing (P < 0.01) and dyspnea (P < 0.01) at the final evaluation.
The TC type was identified for a total of 150 dogs in the MM and SM groups. The proportion of dogs with TTC (70/84) in the MM group was significantly (P < 0.01) greater than the proportion of dogs with TTC in the SM group (37/75), whereas the proportion of dogs with MTC in the MM group (5/84) was significantly (P < 0.01) lower than the proportion of dogs with MTC in the SM group (38/75). Overall, the proportion of dogs with MTC that received a tracheal stent (38/43) was significantly (P < 0.01) greater than the proportion of dogs with TTC that received a tracheal stent (37/107). The odds for predicting the likelihood of requiring a tracheal stent were significantly (P < 0.01) less for TTC dogs relative to MTC dogs (OR, 0.068).
The MST from initial TC diagnosis did not differ significantly (P = 0.27) between dogs of the MM group (1,363 days [3.7 years]) and dogs of the SM group (1,913 days [5.2 years]). Additionally, the MST from initial TC evaluation by the IR service did not differ significantly (P = 0.28) between dogs of the MM group (1,124 days [3.1 years]) and dogs of the SM group (1,005 days [2.8 years]; Figure 3). However, the MST from the initial TC diagnosis by the referring veterinarian for dogs of the MM group (1,363 days) was significantly (P < 0.01) longer than the MST from stent placement for the dogs of the SM group (1,005 days).
Kaplan-Meier survival curves for 84 dogs with TC that received multimodal medical management only (MM group; blue line) and 75 dogs with TC that underwent stent placement after failing to respond or becoming refractory to multimodal medical management (SM group; red line) from initial TC diagnosis (A) and from initiation of treatment or stent placement (B). Notice that the MST from initial TC diagnosis was numerically but not significantly (P = 0.27) longer for the dogs of the SM group (1,913 days [5.2 years]) than for the dogs of the MM group (1,363 [3.7 years]); however, the MST from initiation of treatment for the dogs of the MM group (1,363 days) was significantly (P < 0.01) longer than the MST from stent placement for the dogs of the SM group (1,005 days [2.8 years]).
Citation: Journal of the American Veterinary Medical Association 258, 3; 10.2460/javma.258.3.279
The mean clinical scores for the MM and SM groups for each study interval were summarized (Table 3). The mean clinical score for coughing did not differ significantly between the 2 groups during any interval. However, the mean clinical score for goose-honking or raspy breathing for the SM group was significantly greater than that for the MM group at baseline but significantly lower during the intermediate-term and long-term follow-up periods and at the final evaluation. Similarly, the mean clinical score for dyspnea for the SM group was significantly greater than that for the MM group at baseline but significantly lower during the long-term follow-up period and at the final evaluation.
Comparison of the mean ± SD clinical scores at various study intervals between the dogs of the MM and SM groups described in Table 1.
| Goose-honking or raspy breathing | Coughing | Dyspnea | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Study interval | MM group | SM group | P value | MM group | SM group | P value | MM group | SM group | P value |
| Baseline | 5 ± 2.9 | 6 ± 3.0 | 0.02 | 4 ± 2.9 | 4 ± 3.4 | 0.27 | 3 ± 3.0 | 6 ± 3.0 | < 0.01 |
| Short-term follow-up period | 2 ± 1.8 | 1 ±1.5 | 0.35 | 1 ± 2.5 | 3 ± 2.1 | 0.10 | 1 ± 1.8 | 1 ± 1.6 | 0.72 |
| Intermediate-term follow-up period | 4 ± 2.3 | 1 ± 1.4 | < 0.01 | 4 ± 2.5 | 3 ± 1.9 | 0.13 | 1 ± 1.5 | 1 ± 1.4 | 0.97 |
| Long-term follow-up period | 4 ± 2.9 | 2 ± 1.4 | < 0.01 | 4 ± 3.0 | 3 ± 1.7 | 0.23 | 3 ± 3.2 | 1 ± 1.6 | < 0.01 |
| Final evaluation | 4 ± 3.0 | 2 ± 2.2 | < 0.01 | 4 ± 2.9 | 3 ± 2.1 | 0.07 | 3 ± 3.2 | 2 ± 2.5 | < 0.01 |
The MST after stent placement was 776, 1,005, 1,087, and 1,694 days for dogs of the SM group with 0, 1, 2, and 3 severe scores at baseline, respectively, and did not differ significantly (P = 0.26) from the MST from stent placement for all dogs in the SM group (1,005 days). The MST did not differ significantly (P = 0.32) between dogs of the MM (306 days) and SM (1,338 days) groups that had ≥ 2 severe scores at baseline. However, the MST did differ significantly (P < 0.01) between dogs of the MM (12 days) and SM (1,338 days) groups that had ≥ 2 severe scores at the final evaluation (Figure 4).
Kaplan-Meier survival curves for 10 dogs of the MM group (blue line) and 15 dogs of the SM group (red line) that were assigned ≥ 2 severe clinical scores at the final evaluation. The MST was significantly (P < 0.01) longer for dogs of the SM group (1,338 days [3.7 years]) than for dogs of the MM group (12 days). See Figure 3 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 258, 3; 10.2460/javma.258.3.279
Discussion
In the present study, clinical characteristics and short-, intermediate-, and long-term outcomes were compared between dogs with TC that received multimodal medical management only (MM group) and those that underwent tracheal endoluminal stent placement after failing to respond or becoming refractory to multimodal medical management (SM group). All dogs were evaluated because of mild to life-threatening signs associated with TC and treated by the same IR service at a veterinary referral hospital over the same 9-year period. The age and breed distributions for the dogs of the MM group were similar to those of the SM group.4 The age at the time of TC diagnosis for the dogs of the present report ranged from 1 to 17 years, which was consistent with findings of other reports1,9,10,11 and suggested that the condition can manifest at any age. Yorkshire Terriers and Pomeranians were the breeds most commonly represented in both the MM (53/84 [63%]) and SM (63/75 [84%])4 groups. The median follow-up time was 743 days (range, 28 to 2,616 days) for the 84 dogs of the MM group and 728 days (range, 0 to 2,119 days) for the 75 dogs of the SM group.4 Therefore, we believe that a sufficient number of dogs with TC at various stages of clinical severity was followed for a sufficiently long enough time to provide insight into the anticipated response to multimodal medical management, potential complications, and estimated longevity for similarly affected dogs.
In previous studies,8,12,13 71% to 93% of dogs with TC responded well to medical management for at least 1 year, with up to 50% of patients eventually weaned off all medications. Other reports9,14,15 suggest that up to a third of dogs with TC may not respond favorably to medical management and require tracheal surgery to alleviate clinical signs. For dogs with TC, medical management is currently deemed palliative,16 and surgery is generally reserved for dogs that fail to respond to medical management.10 Among the 84 dogs of the MM group in the present study, 64 (76%) had mild or non–life-threatening clinical signs during the initial evaluation at the referral hospital (baseline), and initially, multimodal medical management, which included administration of various combinations of antitussives, corticosteroids, anxiolytics or sedatives, and antimicrobials, lessened the severity of the clinical signs associated with TC. However, the clinical severity of the disease worsened (as evidenced by an increase or return to the baseline clinical scores for goose-honking or raspy breathing, coughing, and dyspnea) during the long-term follow-up period and at the final evaluation. The ongoing need for medication is an important consideration for the management of dogs with TC, regardless of whether tracheal surgery is performed. In fact, patients that undergo stent placement often require higher doses of medications more frequently than do medically managed dogs, likely owing to the combined effects of the presence of the stent and the severe manifestation of TC that led to its placement.
In the present study, 67 (80%) and 42 (50%) of the 84 dogs of the MM group were still receiving antitussives and corticosteroids, respectively, at the time of the final evaluation. Additionally, 21 (25%) dogs underwent emergency examination, and 8 (10%) dogs developed pneumonia during multimodal medical management. These data suggested that, although medical management may help alleviate clinical signs of TC for a period, the condition generally progresses, and the worsening of clinical signs can often be acute and severe and occasionally life-threatening.
For the dogs of the MM group of the present study, development of pneumonia or an emergency examination during the intermediate-term follow-up period (46 to 180 days after baseline) and diagnosis of cardiac disease were significantly associated with a shortened MST. That finding, along with the fact that 36 (43%), 35 (42%), and 27 (32%) of the 84 dogs in the MM group had a history of cardiomegaly, heart murmur, or cardiac valvular disease at baseline, suggested that a thorough cardiac evaluation should be performed for dogs with TC. Mainstem bronchial collapse was not significantly associated with a shortened MST for the dogs of the MM group; however, diagnosis of bronchial collapse was made solely on the basis of evaluation of thoracic radiographs, so that finding should be interpreted with caution.
We were disappointed that the MST from initial evaluation at our hospital (baseline) was only 1,124 days (3.1 years) for the dogs of the MM group in the present study. Given that 42 of the 84 (50%) dogs of the MM group were receiving ≥ 2 medications and only 17 (20%) were unmedicated at baseline, we decided to also calculate the MST from the initial TC diagnosis on the basis of review of medical records from the referring or primary care veterinarians. The MST from initial TC diagnosis was 1,363 days (3.7 years) for the dogs of the MM group, but it did not differ significantly (P = 0.13) from the MST from baseline. Because the primary purpose of the present study was to evaluate the prognosis for dogs with TC following initiation of multimodal medical management, we believed that the MST from initial evaluation at our hospital provided a more conservative and accurate assessment of survival time for medically managed patients. It could be argued that dogs with TC referred to our hospital for evaluation had more severe disease than those managed by their primary care veterinarians. However, the median duration of clinical signs prior to baseline for the dogs of the MM group was 87 days (mean, 228 days; range, 0 to 1,825 days), which suggested that many dogs were evaluated at our hospital fairly soon after TC was diagnosed and were presumably in the early stages of the disease process.
An important finding of the present study was that, for the dogs of the MM group, the MST decreased as the number of severe (≥ 8) clinical scores assigned at baseline increased from 0 (MST, 1,363 days) to 3 (MST, 55 days). Although the clinical scoring system used in the present study was subjective and has yet to be validated, it appeared to be useful for identification of patients most at risk for disease progression. It was also used in the previous study4 in which the dogs of the SM group were described. In general, the dogs of the MM group had less severe clinical disease than did the dogs of the SM group at baseline. For each dog, the severity of goose-honking or raspy breathing, coughing, and dyspnea was assessed by the owner on a scale of 0 (sign never observed) to 10 (sign persistent or intractable). We defined a severe score as any clinical score ≥ 8 on the basis of our clinical impression that dogs with scores ≥ 8 required immediate alterations to the multimodal medical regimen (ie, an increase in dose or frequency of a particular medication or the addition of a medication) more frequently than did dogs with clinical scores < 8. From a clinical standpoint, we believe it is crucial to distinguish dogs with severe clinical signs of TC from those with only mild to moderate clinical signs of the disease to facilitate identification of patients that will benefit from surgical intervention. We have found serial use of the clinical scoring system described in this study helpful for identifying patients with disease progression so that more timely interventions or frequent examinations can be performed, thereby reducing the risk for emergency examinations and respiratory crises. To our knowledge, a standardized questionnaire for accurately assessing dogs with TC has yet to be developed.
Owners of dogs with TC want to know the risks, prognoses, and costs associated with various interventions so they can make informed decisions regarding treatment. The present study was the first to compare survival times between dogs with TC that received multimodal medical management only and those that underwent stent placement. Results suggested that the MST from initial TC diagnosis for the dogs of the MM group (1,363 days [3.7 years]) was significantly (P < 0.01) longer than the MST from stent placement for the dogs of the SM group (1,005 days [2.8 years]). However, the MST from initial TC diagnosis for the dogs of the SM group (1,913 days [5.2 years]) was numerically, albeit not significantly (P = 0.27), longer than that for the dogs of the MM group. It should be noted that there were several significant differences between the dogs of the MM and SM groups. Specifically, dogs in the SM group generally had more severe clinical disease than did the dogs in the MM group and underwent stent placement only after multimodal medical management failed. Additionally, the mean ± SD body weight for the dogs of the MM group (4.4 ± 2.6 kg) was significantly greater than that for the dogs of the SM group (3.5 ± 1.3 kg [7.7 ± 2.9 lb]). The proportions of dogs with TTC and MTC also differed significantly between the MM and SM groups. Finally, some dogs in the MM group had severe clinical disease, but the owners declined stent placement because of concerns about complications associated with the procedure or stents or they were dissuaded from the procedure by their primary care veterinarian. When severely compromised dogs in the MM group (dogs with ≥ 2 severe scores at the final evaluation; n = 10 dogs) were compared with similarly compromised dogs in the SM group (15 dogs), the MST for the dogs in the MM group (12 days) was significantly shorter than that for the dogs of the SM group (1,338 days). That finding supported our hypothesis that dogs with severe clinical disease at TC diagnosis would have a better outcome if they underwent stent placement as soon as possible.
In the present study, 38 of the 75 (51%) dogs of the SM group but only 5 of the 84 (6%) dogs in the MM group had MTC. That finding supported our theory that dogs with MTC were more likely to require stent placement or some other surgical intervention because the static airway obstruction characteristic of MTC was less likely to respond to medical management than the more dynamic airway obstruction characteristic of TTC. Conversely, 70 of 84 (83%) dogs in the MM group and 37 of 75 (49%) dogs in the SM group had TTC. For the overall study population, the proportion of dogs with MTC that underwent stent placement (38/43 [88%]) was significantly (P < 0.01) greater than the proportion of dogs with TTC that underwent stent placement (37/107 [35%]). The odds for predicting the likelihood of requiring a tracheal stent were significantly (P < 0.01) less for TTC dogs relative to MTC dogs (OR, 0.068). Thus, determining the type of TC at the time of initial diagnosis provides important prognostic information.
The present study had several limitations. Because of the retrospective nature of the study, the data available for analysis were not standardized for each patient; therefore, potentially relevant comorbidities that may have affected survival were not analyzed. Cox proportional hazard models, which may have elucidated potential confounders, were not performed because the identification of confounders was beyond the scope of the present study. The type of TC was determined by evaluation of thoracic radiographs instead of by fluoroscopy or endoscopy. Research indicates that radiography underestimates the extent of TC at the thoracic inlet and thoracic and carinal regions17; however, assessing the extent of TC was not an important aspect of the present study or the previous study4 in which the dogs of the SM group were described. Owner compliance with the multimodal medical regimen prescribed by the IR service varied. Some owners occasionally discontinued a medication without consulting the IR service or intentionally chose to not follow the regimen. Referring veterinarians also occasionally altered the medical regimen without consulting the IR service, and the IR service had to make subsequent adjustments to the regimen accordingly. Although compliance issues complicated interpretation of the data, it reflected the reality of evaluating a group of patients with a complex disease and multiple comorbidities. Finally, the small number of dogs with MTC evaluated in this study may not have accurately represented the general population of dogs with that condition. Additional studies that assess a larger number of dogs with TC are necessary to better elucidate signalment variables, comorbidities, clinical signs, and treatments that significantly affect patient survival. Ideally, those studies should be prospective in nature with all dogs examined at the same hospital so that treatment and follow-up protocols can be standardized to the extent possible.
Results of the present study indicated that the clinical signs of dogs with TC that received multimodal medical management (MM group) generally improved during the short term (within the first 45 days after treatment initiation) but regressed and worsened over time. In the present study, clinical scores for the MM group during the intermediate-term (46 to 180 days after treatment initiation) and long-term (> 180 days after treatment initiation) follow-up periods increased (ie, clinical signs became worse) relative to those at baseline (prior to treatment initiation) for approximately 15% to 30% and 35% to 43% of patients, respectively. This supported our clinical impression that multimodal medical management was ineffective for long-term treatment of TC in many dogs. The MST from TC diagnosis was 3.7 years for dogs of the MM group and 5.2 years for dogs with TC that underwent stent placement. For dogs with severe clinical disease (≥ 2 severe scores), the MST was only 12 days for dogs that received multimodal medical management only, compared with 1,338 days for dogs that underwent stent placement. That finding supported our hypothesis that dogs with a high clinical score at TC diagnosis would have an improved outcome if they underwent stent placement as soon as possible. Other factors associated with a significantly shortened MST for dogs of the MM group included age > 8.8 years and a history of cardiac disease. Additionally, dogs with MTC appeared to be more likely to require stent placement than were dogs with TTC. In dogs, the etiology of TC is multifactorial,18 and findings of the present study indicated that multimodal medical management can alleviate clinical signs for months to years in dogs with mild to moderate disease, but surgical intervention, such as stent placement, should be considered for dogs with clinically severe TC.
Acknowledgments
Supported by the Egg Roll Fund.
Dr. Weisse is a consultant for, and minority partner of, Infiniti Medical LLC.
Preliminary findings were presented as research reports at the Veterinary Interventional Radiology and Interventional Endoscopy Society Summit, Squaw Valley, Calif, May 2019, and at the American College of Veterinary Internal Medicine Summit, Phoenix, June 2019.
The authors thank Ken Lamb for suggestions and assistance with data acquisition.
Footnotes
SAS, version 9.4, SAS Institute Inc, Cary, NC.
Abbreviations
| IR | Interventional radiology |
| MM | Medical management only |
| MST | Median survival time |
| MTC | Malformation-type tracheal collapse |
| SM | Surgical management with stent placement |
| TC | Tracheal collapse |
| TTC | Traditional-type tracheal collapse |
References
- 1. ↑
Payne JD, Mehler SJ, Weisse C. Tracheal collapse. Compend Contin Educ Pract Vet 2006;28:373–382.
- 2. ↑
Weisse C. Insights in tracheobronchial stenting and a theory of bronchial compression. J Small Anim Pract 2014;55:181–184.
- 3. ↑
Weisse C. Intraluminal tracheal stenting. In: Weisse C, Berent A, eds. Veterinary image-guided interventions. Ames, Iowa: Wiley Blackwell, 2015;73–82.
- 4. ↑
Weisse C, Berent A, Violette N, et al. Short-, intermediate-, and long-term results for endoluminal stent placement in dogs with tracheal collapse. J Am Vet Med Assoc 2019;254:380–392.
- 5. ↑
Raske M, Weisse C, Berent A, et al. Immediate, short-, and long-term changes in tracheal stent diameter, length, and positioning after placement in dogs with tracheal collapse syndrome. J Vet Intern Med 2018;32:782–791.
- 7. ↑
Slatter DH, Pettit GD. A surgical method of correction of collapsed trachea in the dog. Aust Vet J 1974;50:41–44.
- 8. ↑
White RAS, Williams J. Tracheal collapse in the dog—is there really a role for surgery? A survey of 100 cases. J Small Anim Pract 1994;35:191–196.
- 9. ↑
Buback JL, Boothe HW, Hobson HP. Surgical treatment of tracheal collapse in dogs: 90 cases (1983–1993). J Am Vet Med Assoc 1996;208:380–384.
- 10. ↑
Moritz A, Schneider M, Bauer N. Management of advanced tracheal collapse in dogs using intraluminal self-expanding biliary wallstents. J Vet Intern Med 2004;18:31–42.
- 11. ↑
Sura PA, Durant AM. Trachea and bronchi. In: Tobias KM, Johnston SA, eds. Veterinary surgery: small animal. St Louis: Elsevier Saunders, 2012;1746–1750.
- 13. ↑
Ayres SA, Holmberg DL. Surgical treatment of tracheal collapse using pliable total ring prostheses: results in one experimental and 4 clinical cases. Can Vet J 1999;40:787–791.
- 14. ↑
Johnson L. Tracheal collapse. Diagnosis and medical and surgical treatment. Vet Clin North Am Small Anim Pract 2000;30:1253–1266.
- 15. ↑
Tangner CH, Hobson HP. A retrospective study of 20 surgically managed cases of collapsed trachea. Vet Surg 1982;11:146–149.
- 16. ↑
Weisse CW, Berent AC, Todd KL, et al. Potential applications of interventional radiology in veterinary medicine. J Am Vet Med Assoc 2008;233:1564–1574.
- 17. ↑
Macready DM, Johnson LR, Pollard RE. Fluoroscopic and radiographic evaluation of tracheal collapse in dogs: 62 cases (2001–2006). J Am Vet Med Assoc 2007;230:1870–1876.
- 18. ↑
Marolf A, Blaik M, Specht A. A retrospective study of the relationship between tracheal collapse and bronchiectasis in dogs. Vet Radiol Ultrasound 2007;48:199–203.
