• Open Access

Clinical Evaluation and Endoscopic Classification of Bronchomalacia in Dogs


Corresponding author: Paola Gianella, Department of Veterinary Sciences, University of Turin, Via L. Da Vinci, Grugliasco, TO 44 10095, Italy; e-mail: paola.gianella@unito.it.



Little information is available about the association between bronchomalacia and historical or clinicopathologic data. Also, studies applying an endoscopic classification scheme that differentiates between static and dynamic bronchial collapse and based on a scoring system are lacking.


To describe the clinical presentation of bronchomalacia in dogs, to classify endoscopic findings, and to evaluate associations among historical, clinicopathologic data, and endoscopic findings.


Fifty-nine client-owned dogs with an endoscopic diagnosis of bronchomalacia.


In this retrospective study, medical records were analyzed and video documentation was reviewed to assign a score to endoscopic findings. Univariate analysis was performed on categorical variables organized in contingency tables, and a stepwise logistic regression model was used for multivariate analysis.


Of the 59 dogs included in the study, 2 were affected by static bronchial collapse (SBC), 35 by dynamic bronchial collapse (DBC), and 22 by both SBC and DBC. The association between SBC and DBC was more frequently seen in the dogs with higher body weight, pulmonary hypertension, a bronchial type of radiographic pattern, and nodularity at endoscopic examination. Thirty-one dogs were presented with tracheomalacia and bronchomalacia; an association emerged between these concurrent disorders in dogs living indoors. Multivariate analysis of the endoscopic scores showed a correlation between DBC severity and cough duration.

Conclusion and Clinical Importance

Results of this study provide evidence for 2 different types of bronchial collapse. Endoscopic scoring scheme has proved to be promising in the bronchomalacia classification, although further evaluation of its applicability in larger canine populations is needed.


bronchoalveolar lavage


body condition score




dynamic bronchial collapse


International Small Animal Cardiac Health Council


left principal bronchus


left cranial lobe


left caudal lobe


right principal bronchus


right cranial lobe


right middle lobe


accessory lung lobe


right caudal lobe


static bronchial collapse






vertebral heart size

Bronchomalacia (BM) refers to weakness of the bronchial walls because of softening of the supportive cartilage and hypotonia of myoelastic elements.[1, 2] This results in a reduction in airway lumen diameter, leading to a respiratory syndrome encompassing chronic cough, wheezing, intermittent or continuous dyspnoea, difficulty in clearing secretions, and recurrent bronchitis and pneumonia.[3-5]

Tracheomalacia (TM) refers to weakness of the tracheal wall, such that the airway is softer and more susceptible to collapse. The weakness may be localized to 1 portion or involve the entire trachea. If both bronchi and trachea are involved, the term tracheobronchomalacia (TBM) is used.[6, 7] As an isolated clinical entity, BM is much less common than TM and TBM in human medicine,[7] in contrast to its occurrence reported in veterinary medicine.[8]

The precise cause of malacia in humans is not fully understood, although several etiopathogenetic factors have been proposed, including congenital conditions,[9] endotracheal intubation,[10] long-term ventilation,[11] closed-chest trauma,[12] chronic airway irritation and inflammation,[13, 14] malignancy,[15] asthma,[3] mechanical anatomic factors,[4] and thyroid diseases.[16] In veterinary medicine, similar causative factors have also been hypothesized,[5, 8, 17] although the role of airway inflammation or cardiomegaly has yet to be confirmed.[8, 18]

Bronchoscopy is the gold standard for diagnosis in humans and animals,[5, 8, 17, 19, 20] with the left cranial lung lobe affected more frequently in brachycephalic and nonbrachycephalic dogs.[5, 17, 18] In human medicine, 2 different forms of dynamic airway obstruction are distinguished: BM and excessive dynamic airway collapse.[6] These entities might or might not coexist; however, they are rarely described in terms of extent, severity, location, and associated anomalies because their understanding has been limited over the years by uncertainties about their definition, diagnosis, and management. Furthermore, comparison between studies and different therapeutic protocols has been hampered for want of a standard method to quantitate the severity of airway collapse.[6]

For these reasons, quantitative or morphometric bronchoscopy has been proposed,[21] but owing to a lack of resources, time, information, or technical limitations, it has seldom, if ever, been used. In veterinary medicine, despite the recent increasing recognition of BM in dogs,[5, 8, 18] studies addressing functional classes of airway narrowing, as well as an endoscopic systematic classification, are lacking.

The aims of this study were (1) to retrospectively evaluate the history and clinical presentation of a series of dogs with BM; (2) to attempt an endoscopic classification of BM by differentiating between static and dynamic bronchial collapse and with the use of a scoring system; and (3) to identify possible associations between historical and clinicopathologic data and endoscopic findings.

Materials and Methods

The medical records of dogs with an endoscopic diagnosis of BM between June 2010 and September 2011 were retrospectively reviewed. Information gleaned from the medical records included breed, sex, age, body weight, body condition score (BCS), lifestyle, the presence of smoking owners, cough duration, type of thoracic radiographic pattern, radiographic evidence of tracheal and bronchial collapse, vertebral heart size (VHS), ISACHC heart failure classification, the presence of pulmonary hypertension, and cytology of the bronchoalveolar lavage (BAL) fluid.

Bronchoscopic examination was performed under general anesthesia in all dogs. A 5-minute preoxygenation period was used. Dogs weighing ≤15 kg were not intubated; a stable plane of anesthesia was maintained with a constant rate infusion of propofol at 0.1–0.4 mg/kg/min or intermittent boluses. Oxygen was delivered through the working channel of the bronchoscope or by use of jet ventilation. Dogs weighing >15 kg were intubated after endoscopic tracheal examination and maintained on gas anesthesia. Balanced isotonic crystalloid fluids were administered IV to all dogs; ECG, blood pressure, and pulse oximetry were constantly monitored. During recovery, supplemental oxygen was provided as needed.

All bronchoscopic procedures were performed in standardized fashion by 2 of the authors (E.B., P.R.).[22] Flexible video endoscopes (5.9 mm × 110 cm1 and 4.9 mm × 60 cm2) were used. Videotaping was done with a digital video converter3 connected to a FireWire 800-equipped Mac computer.4 Video documentation for each dog was independently reviewed by 2 authors (E.B., P.G.) in a blind, separate fashion to grade and map airway collapse and assign an endoscopic score. A discussion to reach a consensus opinion was done only in cases of divergent opinion.

The study population was composed of dogs with BM, with or without endoscopically detectable TM, and was divided into 3 subgroups according to the type of bronchial airway collapse (dynamic [subgroup 1], static + dynamic [subgroup 2], static [subgroup 3]). Tracheal collapse (cervical, intrathoracic, or both) was graded as proposed by Tangner and Hobson.[23] The canine endobronchial map was used for lower airways examination and identification.[22]

Normal bronchial openings were described as smooth, round, or slightly oval in appearance, with a minimal reduction in luminal diameter during phases of respiration.[24] Bronchial collapse was defined as static if a reduction in luminal diameter persisted during all phases of respiration, and as dynamic if changes in luminal diameter were observed on respiration. The reduction in luminal diameter was identified as flattening of the bronchial openings, circumferential narrowing, or distortion of their normal round appearance. Collapse was reported at the left principal bronchus, left cranial lobe, or left caudal lobe and the right principal bronchus, right cranial lobe, right middle lobe, accessory lung lobe, or right caudal lobe. Grade 1 static and dynamic collapse was visualized and defined as a diameter reduction ≤50%, grade 2 collapse as a diameter reduction >50% and ≤75%, and grade 3 collapse as a diameter reduction >75%, with contact between the dorsal and ventral mucosa of the collapsed bronchus. Furthermore, type A extension of static and dynamic collapse was defined as collapse affecting ≤3 bronchi (principal left and right, lobar, segmental),[22] type B as collapse affecting 4–5 bronchi (principal left and right, lobar, segmental),[22] and type C as collapse affecting >5 bronchi (principal left and right, lobar, segmental).[22] Each dog was classified based on the worst grade of collapse and total number of airways affected. Final endoscopic scores were then obtained by evaluating separately the severity and extension (grade 1, 2, 3 and type A, B, C, respectively) of static and dynamic bronchial collapse. Additional bronchoscopic findings including the presence of airway secretions, hyperemia, or nodularity were also recorded.

Bronchoalveolar lavage was performed by wedging the endoscope into 1 or 2 of the smallest bronchial segments, followed by instillation and continuous vacuum aspiration of 2 mL/kg of warmed, sterile saline solution through the biopsy channel of the endoscope. Bronchoalveolar lavage fluid was processed immediately after sampling for nucleated cell count by means of a standard hemocytometer and cytologic differentiation on slides stained with Wright-Giemsa stain. Cytology results were classified according to the predominant cell type in the sample.[25]

Statistical Analysis

Statistical analysis was performed by a freeware statistical software package.5

Nominal data were expressed as frequency, percentage, or both. TM, dynamic bronchial collapse (DBC), and dynamic + static bronchial collapse (DBC + SBC) were classified as being present or absent. For the endoscopic score, each grade of severity (1, 2, 3) and type of extension (A, B, C) were classified as present or absent in the DBC and SBC groups, respectively. The historical and clinicopathologic data were transformed into class variables (Table 1). Univariate analysis of nominal data was performed with contingency table analysis by Fisher's exact test. In addition, for each 2 × 2 contingency table, the odds ratio (OR) and the 95% confidence intervals of the odds ratio (95% ORCI) were calculated. Variables that meet a cut-off of P < .20 at the univariate analysis were entered into a logistic regression model for the multivariate analysis. Three separate logistic regression models, one for TM, one for DBC, and one for DBC + SBC, were constructed, having TM, DBC, and DBC + SBC as the dependent variables and historical and clinicopathologic data as the independent variables. For the endoscopic score, separate logistic regression models for each grade of severity (1, 2, 3) and type of extension (A, B, C) were constructed in the SBC and DBC groups, respectively, having the grade of severity and type of extension as the dependent variables and historical and clinicopathologic data as the independent variables. The most parsimonial final model was selected, via backward elimination, with a Wald P value of .05 as the removal threshold, given an acceptable log-likelihood ratio test value. Model fit was evaluated by Pearson's and Hosmer-Lemeshow's goodness-of-fit test. A value of P > .05 indicates that the data adequately fit the model used.

Table 1. Historical and clinicopathologic variables recorded and tested for their association with TM, in addition to bronchomalacia, dynamic bronchial collapse, and dynamic + static bronchial collapse
Historical and Clinicopathologic VariablesVariables
  1. a

    Reference values for normal BAL cell and differential count were 200–450 cells/μL comprising up to 5–7 ± 5% neutrophils, lymphocytes, and eosinophils, and 65–85% macrophages.[27]

  2. b

    Type of mixed inflammation at cytologic analysis of BAL fluid: neutrophilic-eosinophilic; neutrophilic-lymphocytic.

  3. c

    TM endoscopically detected.

Age (years)<5≥5 and ≤10>10 
Body weight (kg)≤10>10 and ≤20>20–≤30>30
BCS (1–9)1–56–97–9
Smoking ownersYesNo
Duration of cough (months)<6≥6
Radiographic patternNormalBronchialBronchointerstitialOther
Cardiologic examinationNormalAbnormal
Pulmonary hypertensionYesNo
Nodularity (endoscopic examination)YesNo


Between June 1, 2010 and September 30, 2011, 59 dogs had a bronchoscopic diagnosis of BM with or without TM. The study population (n = 59) consisted of dogs from 22 different breeds and 21 mixed-breed dogs. Breeds accounting for at least 3 cases were Miniature Poodle (n = 5), Yorkshire Terrier (n = 5), Pugs (n = 4), and Epagneul Breton (n = 3). Thirty dogs were male (4 neutered) and 29 were female (17 spayed). The median age was 11 years (range 2–16). Three dogs were aged between 2 and 4 years at presentation, 25 between 5 and 10 years, and 31 were more than 10 years old. The median body weight was 10 kg (range 2.5–60 kg). Twenty-five dogs weighed 10 kg or less, 23 between 11 and 20 kg, 8 between 21 and 30 kg, and 3 more than 30 kg. The median BCS was 7 (range 3–9). Of the 48 dogs with a BCS >5, 31 had a BCS ≥7. Forty dogs lived indoors, 8 outdoors, and 11 indoors/outdoors. Thirty-three dogs were from households where smoking had regularly occurred indoors in the past year. Chronic cough was the chief complaint in all dogs; the duration was <6 months in 33 dogs and ≥6 months in 26. Radiographic diagnosis of an abnormal pulmonary pattern was made in 40/59 dogs, with bronchial and bronchointerstitial patterns found in 23/40 and 13/40 dogs, respectively. In the remaining 4 dogs, an interstitial pattern (2), an alveolar pattern (1), and bronchiectasis (1) were identified. Static radiography was considered supportive of tracheal collapse in 4/59 dogs. There was no radiographic evidence of bronchial collapse.

The VHS was ≤9.7 (normal, 8.5–10.7)[26] in all 52/59 dogs in which it was measured.

Of the 32 dogs with concurrent heart disease, 22 were classified as ISACHC Class I, of which 8/22 Class Ia and 14/22 Class Ib, 9 as ISACHC Class II, and 1 as ISACHC Class III. Doppler echocardiography demonstrated pulmonary hypertension in 6/59 dogs.

BM without tracheal involvement was identified on bronchoscopy in 28/59 dogs (48%), and BM + TM in 31/59 (52%), of which 22 (71%) with a dynamic type of bronchial collapse. TM was primarily associated with movement of the dorsal tracheal membrane, and BM with dorsoventral flattening of the bronchial wall. Among the dogs with TM, grade I tracheal collapse was present in 9/31 (29%), grade II in 14/31 (45%), grade III in 7/31 (23%), and grade IV in 1/31 (3%). Overall, 2 dogs (3%) were affected by SBC alone, 35 (59%) by DBC alone, and 22 (37%) by both static and dynamic bronchial collapse (DBC + SBC). Concurrent static collapse of 1–3 lobar bronchi was documented in 9/59 dogs (15%); concurrent dynamic collapse of 1–6 lobar bronchi was documented in 22/59 (37%). In 1 of these dogs, both static and dynamic lobar bronchial collapse were identified. Segmental static and dynamic airway collapse were documented in 15/59 (25%) and in 35/59 (59%) dogs, respectively.

Overall, collapse of the left cranial lobe bronchus was most commonly identified (22/59 dogs, 37%), followed in descending order by collapse of the left caudal (16/59, 27%), right cranial (14/59, 24%), right caudal (12/59, 20%), right middle (10/59, 17%), and accessory lung lobe bronchus (8/59, 14%). Table 2 reports the final endoscopic score; the number of dogs in the different categories for static and dynamic collapse has been listed. All categories of severity and extension of SBC and DBC were represented, except for grade 3 type B SBC. With regard to DBC, 20 dogs were scored 3C, the most severe and extensive type of collapse.

Table 2. Severity and extension of static and dynamic bronchial collapse (endoscopic score) in 59 dogs with bronchomalacia
 Static Collapse
  1. The numbers (1, 2, 3) indicate the grade of severity of bronchial collapse; the letters (A, B, C) refer to the extension of bronchial collapse.

Endoscopic score1A1B1C2A2B2C3A3B3C
Number of dogs71232432
 Dynamic Collapse
Endoscopic score1A1B1C2A2B2C3A3B3C
Number of dogs10511592420

On bronchoscopy, all dogs presented gross evidence of airway inflammation with hyperemia and mucus accumulation; 20 dogs presented bronchial nodules. Bronchoalveolar lavage was performed in 45 dogs: cytology was normal in 7; in the remaining 38, analysis revealed neutrophilic (n = 23), lymphocytic (n = 1), eosinophilic (n = 1), mixed neutrophilic-eosinophilic (n = 12), and mixed neutrophilic-lymphocytic (n = 1) inflammation. Intracellular bacteria were identified in 25 dogs. No parasitic larvae or ova were found in any of the samples examined. Culture of BAL fluid samples was not performed because of financial and logistic concerns. Fourteen dogs did not undergo BAL because of dramatic airway collapse and hemoglobin desaturation.

There were no significant associations detected for DBC by multivariate analysis (Table 3 and 4).

Table 3. Results of univariate analysis of historical and clinicopathologic variables tested for the association with tracheomalacia (TM), dynamic (DBC), and dynamic + static bronchial collapse (DBC + SBC) in 59 dogs with bronchomalacia
Historical and Clinicopathologic VariablesTM (n = 31) (95% ORCI; P)DBC (n = 35) (95% ORCI; P)DBC + SBC (n = 22) (95% ORCI; P)
  1. BAL, bronchoalveolar lavage; BCS, body condition score; ORCI, odds ratio 95% confidence intervals.

  2. a

    Significant association.

Sex0.34–3.34; 10.24–2.55; .80.38–4.17; .7
Age <5 years0.00–2.13; .1a0.06–85.72; 10.013–16.98; 1
Age ≥5 years and ≤10 years0.29–3.29; 10.13–1.60; .30.57–6.74; .3
Age >10 years0.48–5.09; .40.58–6.31; .30.16–1.81; .3
Body weight ≤10 kg2.31–32.98; <.01a0.72–7.84; .1a0.17–1.87; .4
Body weight >10 and ≤20 kg0.10–1.61; .20.64–12.70; .20.04–1.30; .1a
Body weight >20 and ≤30 kg0.002–0.90; .02a0.001–0.65; .05a1.02–71.02; .04a
BCS ≤1–5 versus ≥6–90.31–6.71; .70.43–9.49; .30.08–1.94; .3
BCS 7–90.43–4.37; .60.72–7.84; .1a0.17–1.87; .4
Indoor4.01–216.83; <.01a1.10–14.96; .02a0.07–1.00; .04a
Outdoor0.00–0.43; <.01a0.01–1.18; .05a1.01–71.02; .04a
Indoor/outdoor0.01–0.83; .01a0.10–2.31; .30.31–6.96; .7
Smoking owners0.26–2.63; .80.64–7.24; .20.17–2.00; .4
Duration of cough <6 months versus ≥6 months0.34–3.48; 10.15–1.62; .30.79–9.10; .1a
Abnormal radiographic pattern0.4–4.83; .60.94–17.8; .04a0.03–0.91; .2
Radiographic bronchial pattern0.16–1.79; .30.07–0.85; .01a1.50–20.14; .01a
Radiographic bronchointerstitial pattern0.38–7.15; .50.27–5.08; 10.23–4.42; 1
Abnormal cardiologic examination0.33–21.98; .40.45–2.97; .30.81–6.22; .3
Pulmonary hypertension0.51–254.15; .20.002–1.14; .03a1.02–512.75; .02a
Neutrophilic inflammation (BAL)0.13–1.56; .1a0.29–3.33; 10.26–3.17; .8
Neutrophilic-eosinophilic inflammation (BAL)0.12–2.49; .50.33–7.64; .70.15–3.57; 1
Nodularity0.18–2.13; .40.11–1.41; .1a0.64–7.84; .1a
Endoscopically detectable TM0.85–9.50; .30.13–1.56; .5
Table 4. Results of multivariate analysis of historical and clinicopathologic variables tested for the association with tracheomalacia (TM), dynamic (DBC), and dynamic + static bronchial collapse (DBC + SBC) in 59 dogs with bronchomalacia
Study PopulationHistorical and Clinicopathologic VariablesOR; (95% CI)
  1. CI, 95% confidence intervals; OR, odds ratio.

  2. a

    Variables removed in backward selection for DBC + SBC group: Body weight >10 and ≤20 kg; Indoor lifestyle; Outdoor lifestyle; Duration of cough <6 months versus ≥6 months.

  3. b

    Variables removed in backward selection for TM group: Age <5 years; Body weight ≤10 kg; Body weight >20 and ≤30 kg; Outdoor lifestyle; Indoor/outdoor lifestyle; Neutrophilic inflammation (BAL).


P for Hosmer-Lemeshow

Goodness-of-fit statistic = .44

Body weight >20–≤30 kg8.4; (1.4–74.3)
Radiographic bronchial pattern7.0; (1.9–30.6)
Pulmonary hypertension25.7; (3.0–579.9)
Nodularity3.4; (1.5–18.4)


P for Hosmer-Lemeshow

Goodness-of-fit statistic = .53

Indoor lifestyle22.7; (4.8–176.4)

Results of the univariate analysis of the endoscopic scores revealed that the historical and clinicopathologic data that met the criteria for the inclusion in the multivariate analysis were as follows: age >10 years (CI 1.084–35.72; P = .04) for type C extension of DBC; BCS 7–9 (CI 0.27–21.13; P = .16) for type B extension of SBC; body weight between 10 and 20 kg (CI 1.36-infinite; P = .03) for SBC grade 1 severity; cough ≥6 months duration (CI 1.32–16.48; P = .02), smoking owners (CI 0.59–45.07; P = .18) and radiographic bronchial pattern (CI 1.12–14.01; P = .04) for DBC grade 3 severity. Multivariate analysis showed a correlation between DBC grade 3 severity and cough ≥6 months duration (CI 4.5–37.7; OR = 11.35). The goodness-of-fit statistic (Hosmer–Lemeshow) suggested that the data adequately fit the final model (P = .24).


This study provides a systematic description of the concurrent presence of different types of BM, with or without TM, in a population of dogs with chronic cough, and explores the applicability of an endoscopic scoring scheme. The occurrence of BM alone or in association with TM has been described elsewhere.[5, 8, 17, 18] As an isolated clinical entity, BM was reported in 87.5% of dogs with brachycephalic airway syndrome,[17] in 100% and in 50% of dogs examined elsewhere,[5, 8] and in 48% of the dogs in this study, most of which were of small to medium size. Sixty-seven percent of the dogs examined in a previous study were small-breed dogs,[5] whereas medium- and large-breed dogs were common in another population.[8] Taken together, these data suggest that BM is a common, increasingly recognized clinical condition in canine breeds.

A substantial number of dogs in this study (59%) were affected by DBC alone or in association with SBC (37%), with most dogs presenting with concurrent DBC and TM (71%). Only 2 were diagnosed with SBC alone (3%). Factors contributing to the association between DBC and TM are not known, although similar histopathologic alterations, such as atrophy of myoelastic elements and cartilage, causing dorsoventral flattening of the tracheal and bronchial wall along with dorsal tracheal membrane movements, can be hypothesized. In contrast, the lack of SBC in dogs with TM could suggest a different pathogenesis. In this regard, the static collapse could result from external compression or increased chronic expiratory efforts, as hypothesized elsewhere.[17]

In contrast to what observed in this study, and although clear data about the occurrence of DBC or SBC are largely unavailable, SBC has been more frequently reported than DBC in the canine BM population,[8, 18] as well as in brachycephalic dogs.[17] This could simply reflect either differences among dog populations or support the hypothesis for the presence of 2 different clinicopathologic entities that might or might not coexist. The presence of continuous forced exhalation in young brachycephalic dogs, in which immature airways are more easily compressed and deformed in comparison with those of adults, was hypothesized as the causative mechanism for developing the static type of bronchial collapse.[17] However, the presence of SBC also in older, nonbrachycephalic dogs, as found in this study population, may reflect the existence of other pathogenetic mechanisms. Furthermore, it cannot be excluded that, over time, a dynamic type of bronchial collapse could evolve into a static type, thus representing an initially less severe form of airway narrowing.

In this study, no predilection of BM for sex or age was observed. As described in previous studies,[5, 8] the majority of the dogs were adult, elderly, overweight, or obese, suggesting that a congenital component of BM may not exist or may clinically manifest over time in association with other comorbidities or clinical signs such as severe chronic cough and that weight gain may be a contributing factor or a consequence of this condition, as hypothesized elsewhere.[5] In this regard, DBC in association with SBC was more often diagnosed in dogs weighing between 20 and 30 kg; however, the BCS did not differ between the weight groups.

Although tracheal collapse is relatively common in small-breed dogs,[27] no information is available about the association between BM and medium- and large-breed dogs. It is possible that, as compared with smaller size dogs, severe chronic cough eventually associated with underlying comorbidities or more severe alterations of the bronchial cartilage and myoelastic elements could have contributed to the development of BM in the larger dogs.

TM was more often found in the dogs living indoors. This association might have been influenced by the fact that small-breed dogs, like the majority of the dogs with TM in this study population, and in which TM is relatively common,[27] are suited for indoor living. Unexpectedly, however, no significant association with chronic smoke exposure emerged. While in line with available data,[28] whether these results might have been related to less smoking among the dog owners over time is unknown and warrants further investigation.

Different abnormal radiographic patterns were identified in 68% of the dogs and a significant association emerged between the bronchial pattern and the presence of DBC + SBC. This could be explained by greater airway narrowing compared with that caused by the presence of DBC alone, thus leading to a more severe airway inflammation. In turn, the more severe airway inflammation might more likely lead to radiographic infiltrates. Concurrent parenchymal pathologies cannot be excluded, however. In agreement with other reports,[5] different radiographic patterns were observed in this study and the absence of radiographic changes did not exclude the presence of BM. Tracheal collapse was diagnosed in only 7% of the dogs and lower airway collapse in none, contrasting with other studies in which the radiographic identification of airway collapse was more accurate.[5, 8]

A substantial percentage of dogs (54%) in this study had concurrent heart disease, but no significant associations with any type of airway narrowing were found. A recent study failed to document a role of cardiomegaly in airway collapse.[18] In the dogs examined here, the available VHS values were not greater than normal, thus supporting the hypothesis that factors other than cardiomegaly might have contributed to the airway collapse.

Interestingly, an association was found between pulmonary hypertension and DBC + SBC, suggesting that the presence of the 2 different types of bronchial collapse might represent a severe clinicopathologic entity. It should be noted, however, that no attempt was made to identify the contribution of chronic obstructive pulmonary disease or hypoxia to the development of pulmonary hypertension, in addition to that caused by an underlying heart disease.

In this population of dogs, BM was found to more commonly involve the left-side bronchi, as observed in other studies.[5, 17, 18] For this particular distribution, besides thoracic conformation hypothesized for brachycephalic dogs,[17] other factors such as concurrent pulmonary disorders may have played a role. In addition, compared with observations of a previous study,[8] a larger number of dogs in this study were presented with smaller airway collapse in the absence of lobar bronchial collapse (49.1%). Whether a reduction in thoracic volume because of obesity or hepatomegaly might have contributed to this condition, as elsewhere suggested,[8] merits further investigation.

It is known that long-standing inflammation may lead to nodular proliferation of fibrous tissue protruding into the airway.[29] In this regard, an association was found here between nodularity at endoscopic examination and DBC + SBC, suggesting that the concurrent presence of the 2 different types of bronchial collapse might cause severe inflammation.

All dogs, regardless of the type and severity of airway narrowing, displayed bronchoscopic evidence of airway inflammation, and the majority showed neutrophilic inflammation, thus confirming the previously reported high percentage of inflammatory airway disease associated with BM.[5, 18] No statistical differences among different types of airway narrowing and results of airway cytology were found. Intracellular bacteria were identified in 25 dogs, for which a lower respiratory tract infection was considered possible.[30] Similar to recent observations,[5] but in contrast with other studies,[8, 18] our results suggest a potential association between infection and BM. However, because of the lack of BAL fluid culture, this link could not be clearly demonstrated and remains speculative at best.

As evidenced by the endoscopic score, most dogs appeared to be affected by DBC of much greater severity and extension than SBC. A possible explanation of this difference is that DBC may represent an initial form of BM, thus far more common than SBC. Furthermore, if we consider that many of the dogs here were of medium and large size, therefore with a larger bronchial tree potentially requiring higher compression forces to cause the same grade of narrowing compared with the smaller dogs, the highly severe DBC found remains surprising. It is possible that serious, chronic cough, the chief complaint in all the dogs, played a significant causative role, perhaps in association with marked alterations of the bronchial tree, especially in the medium and large size dogs.

Significant associations were found between the endoscopic score and several clinicopathologic variables, thus supporting further investigation of the scheme's applicability in larger canine populations.

This study has some limitations, primarily related to its retrospective design and the relatively small sample size, considering the different types of airway narrowing addressed. The increasing recognition of this syndrome[5, 8, 17, 18] means that a very large study population would be needed to identify significant differences between dogs with different types of airway narrowing. Another limitation is the lack of information about histologic diagnosis, which could have shed light on the etiology of the different forms of airway obstruction compared here.

Finally, further studies applying the endoscopic scoring scheme are needed to compare its use in populations of dogs of different size and age. This has practical implications because, in large-breed dogs, in which deeper segments of the bronchial tree can be examined, BM might be erroneously classified as being more severe than in small-breed dogs.

The question remains as to whether the entity of cough and the presence of severe alterations of the bronchial tree could have influenced the severity of bronchial collapse and its progression from a dynamic to a static type, especially in the medium-large-breed dogs. In light of these considerations, the use of an endoscopic scoring scheme could facilitate a comprehensive approach to BM patients and eventually guide further therapeutic interventions.


The authors thank Dr Elisa Manassero for her assistance in the initial data collection.

Conflict of Interest: Authors disclose no conflict of interest.


  1. 1

    EG-270N5 Fujinon®, AMEDIC, Lyon, France

  2. 2

    EG-270N5 Fujinon®, AMEDIC, Lyon, France

  3. 3

    Canopus ADVC110, Canopus®, Rekeo, Modena, Italy

  4. 4

    Apple MacBook Pro-Core i7 processor 2.2 GHz 15.4-inch, Apple, Turin, Italy

  5. 5

    R, version 1.2.0, R Foundation for Statistical Computing, Vienna, Austria. Available at: www.r-project.org. Accessed February 6, 2012.