The cough reflex is complex, involving multiple and interacting afferent receptor types in the larynx, trachea, and bronchi.1 Primary respiratory diseases that cause cough in dogs include dynamic airway collapse (tracheobronchomalacia), bronchitis, pneumonia, heartworm pneumonitis, and pulmonary neoplasia, among others.2 Cough also is commonly cited as a clinical sign of CPE resulting from left-sided congestive heart failure in dogs.3,4 However, given the distribution of cough receptors in the airways, stimulation of terminal airways by interstitial or alveolar edema would not be expected to induce cough.5,6
The cause of cough in elderly small-breed dogs with advanced DMVD is particularly challenging to identify and could be related to concomitant airway disease (tracheobronchomalacia or chronic bronchitis), CPE, or mainstem bronchial compression from severe left atrial enlargement.7–9 Nevertheless, it is clinically important to determine whether such dogs have CPE so that diuretic treatment can be either initiated or withheld.
Thoracic radiography is generally recommended as a commonly available, first-line diagnostic test for identifying causes of cough2,3 and is considered particularly useful for identifying pneumonia10 and CPE.11 However, thoracic radiography is less useful for the diagnosis of airway diseases such as bronchitis and dynamic airway collapse12,13 and has limited sensitivity for the detection of metastatic neoplasia.14 Furthermore, the presence of radiographic cardiomegaly (particularly left atrial enlargement) in a coughing dog could lead to a misdiagnosis of CPE, particularly when evaluation of the pulmonary parenchyma is complicated by the expiratory phase of respiration.7–9 Advanced radiographic techniques such as fluoroscopy and CT may provide additional diagnostic information for coughing patients, but availability is limited to larger referral institutions; furthermore, performance of CT requires that animals be sedated or anesthetized, which can affect cardiac and respiratory assessment. Adjunctive, noninvasive thoracic imaging modalities could therefore provide useful information in the diagnostic workup for coughing dogs.
Point-of-care LUS is a noninvasive thoracic imaging modality with potential in the identification of pulmonary disease. Findings can suggest the presence of alveolar or interstitial abnormalities via ultrasound artifacts known as B lines.15,16 B lines (also called ring-down artifacts or lung rockets) are created by the high impedance gradient between small fluid-filled alveoli and surrounding air.17 Stacks of horizontal ring-down artifacts coalesce to create the ultrasonographic appearance of narrow hyperechoic vertical lines that extend without fading from the pleural-pulmonary interface to the far aspect of the ultrasound screen. B lines obliterate normal pulmonary A lines and move synchronously with respiration.15–17 The presence of multiple bilateral B lines has been shown to have high diagnostic accuracy in the differentiation of CPE from other causes of respiratory distress in humans16,18 and dogs and cats.19–21 In addition to B lines, LUS can also allow identification subpleural abnormalities that suggest peripheral lung consolidation. In humans, specific subpleural ultrasonic patterns (shred, tissue, and nodule signs) have been associated with pneumonia, pulmonary thromboembolism, and pulmonary neoplasia, respectively.16,22 In the human emergency setting, particularly in patients with pleuritic chest pain, LUS subpleural signs have diagnostic accuracy for pneumonia and pulmonary thromboembolism equal to or better than that of thoracic radiography.23,24 Subpleural signs in small animals have also been described25; however, the diagnostic usefulness of these findings for dogs and cats with respiratory signs has not yet been assessed.
Although LUS aids the diagnosis of pulmonary abnormalities in small animals with respiratory distress, its usefulness for evaluating dogs with a primary clinical complaint of cough has not been evaluated to the authors' knowledge. The aims of the study reported here were to describe LUS findings in a population of coughing dogs and to determine whether these LUS results differed on the basis of the underlying cause of cough. We hypothesized that B lines would be most common in dogs with CPE, the shred sign most common in dogs with pneumonia, and the nodule sign most common in dogs with pulmonary neoplasia.
Materials and Methods
Animals
Dogs were prospectively enrolled from among dogs evaluated at the Hixson-Lied Small Animal Hospital of Iowa State University from January 2016 through April 2017. Dogs were considered for inclusion in the study if the primary complaint reported by the owner was cough, one of the authors trained and experienced in performing LUS (JLW or WAW) was available to perform an LUS examination, and thoracic radiography was planned and could be performed within 2.5 hours before or after the LUS examination. Dogs that received medical treatment between completion of LUS and thoracic radiography were excluded, as were dogs with moderate to severe tachypnea or respiratory distress identified on physical examination (to ensure inclusion of coughing dogs vs those with overt respiratory distress). Dogs with other clinical signs of upper or lower respiratory tract disease (stertor or stridor, mildly increased respiratory rate or effort, wheezing, sneezing, or nasal discharge) were not excluded. Prior treatment for cough was not an exclusion criterion provided owners still reported a clinical complaint of cough at the time of initial evaluation (ie, cough persisted despite treatment).
The study protocol was approved by the Institutional Animal Care and Use Committee of Iowa State University. Informed consent was obtained from the owner of each participating dog.
Clinical examination
Routine signalment, history, and physical examination data for each dog were collected, including patient age, body weight, heart rate, respiratory rate, rectal temperature, and presence or absence of heart murmur on cardiac auscultation. Owners were also asked to list any medications that had been prescribed previously for cough. Two- or 3-view thoracic radiographs (right lateral and dorsoventral views, with or without a left lateral view) were obtained within 2.5 hours before or after an LUS examination. No treatments were administered from the time of the initial imaging examination (LUS or thoracic radiography) until after the second imaging examination. Additional diagnostic tests and therapeutic interventions were performed at the discretion of the attending clinician.
LUS
The standardized Vet BLUE protocol for LUS examination was performed for each dog as described elsewhere,23–25 with dogs standing or in sternal recumbency, by a board-certified cardiologist trained and experienced in performing LUS (JLW or WAW). A single portable ultrasonography machinea programmed with standardized settings (ultrasound frequency, 5 to 8 MHz; depth, 4 to 6 cm) and a curvilinear probeb were used. Images were acquired at 4 sites on each hemithorax at standardized locations (caudal, perihilar, middle, and cranial) for a total of 8 sites/dog. Hair over each site was parted (not shaved), and alcohol was used to facilitate probe contact. The ultrasound probe was held horizontally at each site at an approximate 90° angle of insonation and moved slightly (cranially or caudally 1 to 2 intercostal spaces) to optimize visibility within the intercostal spaces at that site. The ultrasound probe was then held stationary at each site while a 3-second cine loop was acquired. In addition, a right parasternal 2-D short-axis image of the heart base, optimized for the left atrium and aorta, was obtained to assess left atrial size by calculating the LA:Ao as described elsewhere.26 The operator noted whether the procedure appeared to have been well tolerated by the dog (ie, the examination did not need to be terminated for medical or behavioral reasons) as well as the procedure duration (ie, ≤ 5 minutes or > 5 minutes).
Analysis of LUS images was performed at a later date by a single investigator (JLW) who was unaware of (blinded to) the patient's other diagnostic findings and final diagnosis. For each Vet BLUE site, the investigator recorded the presence and number of B lines and other subpleural abnormalities (shred, tissue, and nodule signs).25 In sites where B lines were visible, the maximum number of B lines within an intercostal space was recorded as 0, 1, 2, 3, > 3, or infinite (Figure 1).21 Infinite was defined as B lines so numerous that they were no longer discernable as individual B lines. An individual Vet BLUE site where > 3 or infinite B lines were recorded was scored as a strong-positive site. For each dog, the total number of B lines from all 8 individual Vet BLUE sites was summed to create a total B-line score for the overall LUS examination. For that calculation, > 3 B lines within an intercostal space were counted as 4 B lines for that site and infinite B lines were counted as 10 B lines for that site, consistent with protocols in human medicine.27 For example, a dog that had 3 Vet BLUE sites each containing 2 B lines, 3 Vet BLUE sites devoid of B lines, and 2 Vet BLUE sites with infinite B lines would have a total B-line score of 26 (2 + 2 + 2 + 0 + 0 + 0 + 10 + 10).
Representative still B-mode ultrasonographic images of healthy (A) and abnormal (B through D) lungs in dogs. In panel A, proper orientation of rib shadows (hypoechoic vertical columns) and pulmonary-pleural interface (hyperechoic curved line in the near field) is visible; no B lines are seen. In subsequent panels, B lines are visible as linear hyperechoic artifacts extending from the pulmonary-pleural interface to the far edge of the screen. These images show a single B line (B), > 3 B lines that can still be recognized individually (C), and infinite B lines where artifacts coalesce and individual B lines can no longer be discerned (D).
Citation: Journal of the American Veterinary Medical Association 255, 5; 10.2460/javma.255.5.574
Subpleural abnormalities were defined as described in the human LUS literature. The shred sign was defined as an irregular discontinuity of the expected linear pulmonary-pleural interface with hyperechoic foci within the consolidation, representing lung consolidation with aeration analogous to a radiographic air bronchogram (Figure 2).22,28 The tissue sign was defined as a linear to triangular discontinuity or deviation from the expected pulmonary-pleural interface, devoid of hyperechoic foci within the consolidation, representing consolidation without aeration described as hepatization of lung.22,28 The nodule sign was defined as a small, well-marginated hypoechoic circular or ellipsoid structure, with or without a hyperechoic distal border, and with or without acoustic enhancement appearing as a B line extending from its distal border through the far field of the ultrasonograph screen.22,28 These subpleural abnormalities have also been described in small animals.25 Combinations of B lines and subpleural abnormalities could be noted within a single Vet BLUE site (eg, a site could contain both a shred sign and B lines).
Representative still B-mode ultrasonographic images of dogs showing the various subpleural abnormalities (arrows) associated with pulmonary disease. A—The shred sign appears as an irregular discontinuity of the expected linear pulmonary-pleural interface, representing consolidation with aeration (analogous to a radiographic air bronchogram). B—The tissue sign appears as a linear to triangular discontinuity of the pulmonary-pleural interface, representing consolidation without aeration (also described as hepatization of lung). C—The nodule sign appears as a well-marginated hypoechoic circular or ellipsoid structure with hyperechoic outline and acoustic enhancement and a B line extending into the far-field, representing organized lung infiltration (analogous to a radiographic nodule or mass).
Citation: Journal of the American Veterinary Medical Association 255, 5; 10.2460/javma.255.5.574
Determination of cause of cough
The final clinical diagnosis for cause of cough was considered the reference (gold) standard for comparison of LUS results and other clinical variables. This diagnosis was determined by examination of the electronic medical record for each dog and consideration of all diagnostic test results from the study visit (excluding LUS results), including combinations of results of thoracic radiography, echocardiography, CT, fluoroscopy, and bronchoscopy; results of cytologic evaluation of tracheal wash or bronchiolar lavage samples; response to treatment; and findings on postmortem examination, if applicable.
Statistical analysis
Statistical analyses were performed by use of commercial software.c An a priori sample size calculation based on results from previous study21 of LUS in dyspneic dogs and cats revealed that 98 subjects would be required to estimate diagnostic accuracy of LUS for the diagnosis of CPE within 10% and with 95% confidence.
Quantitative clinical and LUS variables used in statistical analyses included dog age, body weight, heart rate, respiratory rate, and rectal temperature on the day of the study visit; LA:Ao; total B-line score; and number of strong-positive B-line sites. Qualitative (categorical) variables included presence (yes or no) of a heart murmur, prior treatment for CPE, B lines, strong-positive B-line sites, and the shred, tissue, and nodule signs. Quantitative data were visually inspected for normality of distribution by histogram creation. Categorical data are reported as frequencies and proportions. Quantitative data are reported as mean ± SD (normally distributed data) or median and range (nonnormally distributed data).
The Student t test (quantitative data) and Fisher exact or χ2 test (categorical data) were used to compare clinical data and LUS findings between groups of dogs. Sensitivity and specificity of LUS for the diagnosis of CPE were calculated, with the final clinical diagnosis used as the reference standard. For all statistical tests, a value of P < 0.05 was considered significant.
Results
Animals
A total of 100 dogs (51 castrated males, 41 spayed females, 6 sexually intact males, and 2 sexually intact females) were enrolled in and completed the study. Mixed-breed dogs were most common (n = 26), followed by Cavalier King Charles Spaniels (11), Pomeranians (6), Beagles (5), and Pugs (4). Thirty-three additional dog breeds were represented by < 4 dogs each.
Other signalment and clinical data were summarized (Table 1). Thirty-two (32%) dogs panted throughout the examination, precluding accurate determination of respiratory rate, so respiratory rate data were available for only 68 (68%) dogs. A systolic heart murmur was detected via auscultation in 58 (58%) dogs. Murmur intensity was graded as II/VI (n = 10), III/VI (9), IV/VI (29), V/VI (9), or VI/VI (1).
Clinical and LUS data for 100 dogs for which the primary clinical complaint was cough.
| Characteristic | All dogs (n = 100) | DAC without LAE (n = 22) | DAC with (n = 15) | CPE (n = 12) | Bronchitis (n = 10) | Bacterial pneumonia (n = 7) | EBP (n = 5) | Pulmonary neoplasia (n = 4) | Fungal pneumonia (n = 2) | Other diagnosis (n = 4) | Unknown diagnosis (n = 19) |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Age (y) | 9.5 ± 3.1 | 10.1 ± 2.0 | 10.6 ± 1.6 | 9.0 ± 1.7 | 9.5 ± 3.4 | 10.2 ± 3.2 | 4.5 ± 4.9* | 10.4 ± 2.5 | 1.5 ± 0.7* | 6.9 ± 4.3 | 9.8 ± 3.0 |
| Body weight (kg) | 15.6 ± 14.0 | 8.1 ± 4.0* | 9.4 ± 5.1 | 15.1 ± 13.6 | 26.1 ± 16.7† | 32.6 ± 28.1† | 20.4 ± 13.1 | 29.1 ± 12.4† | 21.3 ± 18.0 | 23.2 ± 13.0 | 11.6 ± 7.0 |
| Rectal temperature (°C) | 38.6 ± 0.5 | 38.7 ± 0.6 | 38.6 ± 0.3 | 38.3 ± 0.3 | 38.6 ± 0.6 | 38.4 ± 0.4 | 38.6 ± 0.7 | 39.2 ± 0.8† | 39.9 ± 0.6† | 38.3 ± 0.4 | 38.6 ± 0.4 |
| Heart rate (beats/min) | 123 ± 32 | 131 ± 34 | 123 ± 26 | 157 ± 18† | 107 ± 24 | 113 ± 35 | 90 ± 19* | 126 ± 34 | 100 ± 28 | 101 ± 22 | 118 ± 30 |
| Respiratory rate (breaths/min) | 41 ± 16 | 37 ± 13 | 35 ± 11 | 53 ± 16† | 38 ± 16 | 39 ± 12 | 60 ± 24† | 27 ± 4 | 57 ± 13 | 36 ± 10 | 38 ± 14 |
| Heart murmur | 58 (58) | 8 (36) | 15 (100)† | 12 (100)† | 5 (50) | 2 (29) | 0 (0)* | 1 (25) | 0 (0) | 1 (25) | 14 (74) |
| LA:Ao | 1.5 ± 0.5 | 1.3 ± 0.1* | 2.2 ± 0.3† | 2.2 ± 0.4† | 1.4 ± 0.3 | 1.1 ± 0.1* | 1.2 ± 0.2 | 1.2 ± 0.2 | 1.2 ± 0.1 | 1.3 ± 0.5 | 1.3 ± 0.3 |
| LA:Ao > 1.6 | 30 (30) | 0 (0)* | 15 (100)† | 12 (100)† | 2 (20) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 1 (25) | 3 (16) |
| Definitive diagnosis | 72 (72) | 22 (100) | 11 (73) | 12 (100) | 6 (60) | 7 (100) | 5 (100) | 4 (100) | 2 (100) | 3 (75) | 0 (0)* |
| Previous treatment for CPE | 16 (16) | 9 (41)† | 0 (0) | NA | 2 (20) | 2 (29) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 3 (16) |
| B lines present | 66 (66) | 12 (55) | 12 (80) | 12 (100)† | 6 (60) | 6 (86) | 3 (60) | 3 (75) | 2 (100) | 2 (50) | 8 (42) |
| Total B-line score | 1 (0–74) | 1 (0–12) | 1 (0–3) | 22.5 (6–74)† | 1 (0–5) | 6 (0–31) | 2 (0–9) | 2.5 (0–6) | 2.5 (1–4) | 0.5 (0–1) | 0 (0–4) |
| Strong-positive B-line sites present | 22 (22) | 4 (18) | 0 (0)* | 11 (92)† | 1 (10) | 4 (57)† | 1 (20) | 1 (25) | 0 (0) | 0 (0) | 0 (0)* |
| No. of strong-positive B-line sites | 0 (0–8) | 0 (0–2) | 0 (0–0) | 3 (0–8)† | 0 (0–1) | 1 (0–3) | 0 (0–2) | 0 (0–1) | 0 (0–0) | 0 (0–0) | 0 (0–1) |
| Shred sign | 11 (11) | 5 (23) | 1 (7) | 0 (0) | 0 (0) | 4 (57)† | 1 (20) | 0 (0) | 0 (0) | 0 (0) | 0 (0) |
| Tissue sign | 1 (1) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 1 (20) | 0 (0) | 0 (0) | 0 (0) | 0 (0 |
| Nodule sign | 9 (9) | 0 (0) | 1 (7) | 0 (0) | 1 (10) | 1 (14) | 0 (0) | 4 (100)† | 1 (50) | 1 (25) | 0 (0) |
Values for qualitative (categorical) data represent number (%) of dogs with the indicated characteristic. Values for normally distributed quantitative data (age, body weight, rectal temperature, heart rate, respiratory rate, and LA:Ao) represent mean ± SD, and those for nonnormally distributed quantitative data (total B-line score and number of strong-positive B-line sites) represent median (range).
Within a row, results for the indicated diagnosis group are significantly (P < 0.05) lower than results for all other dogs combined.
Within a row, results for the indicated diagnosis group are significantly (P < 0.05) higher than results for all other dogs combined.
BC = Bronchial collapse. DAC = Dynamic airway collapse. LAE = Moderate to severe left atrial enlargement. NA = Not applicable.
Diagnosed causes of cough included dynamic airway collapse (tracheal or bronchial collapse without [n = 22] or with [15] moderate to severe left atrial enlargement; total, 37 [37%]), CPE (12 [12%]), bronchitis (10 [10%]), bacterial pneumonia (7 [7%]), EBP (5 [5%]), pulmonary neoplasia (primary pulmonary carcinoma [3] or lymphoma [1]; total, 4 [4%]), fungal pneumonia (2 [2%]), gastroesophageal reflux disease (2 [2%]), peritoneopericardial diaphragmatic hernia (1 [1%]), and congenital pulmonary anomaly (1 [1%]). For statistical comparisons, the 4 dogs with gastroesophageal reflux disease, peritoneopericardial diaphragmatic hernia, or congenital pulmonary anomaly were grouped together in an “other diagnosis” category. For 19 (19%) dogs, the cause of cough could not be determined. Of the 37 dogs with dynamic airway collapse, 15 (41%) had concurrent moderate to severe left atrial enlargement (ie, LA:Ao > 1.6 on LUS), introducing a possible component of physical bronchial compression from cardiac enlargement.7,8 The CPE in the 12 dogs with that diagnosis was attributed to DMVD (n = 11) or dilated cardiomyopathy (1).
These diagnoses were considered definitive as the underlying cause of cough for 72 of the 100 (72%) dogs on the basis of results of thoracic imaging, cytologic evaluation, or both. In particular, the diagnosis of CPE was considered definitive for all 12 dogs given results of echocardiography and thoracic radiography as well as the clinical response of cough to treatment for congestive heart failure. Primary airway disease (dynamic airway collapse or bronchitis) was suspected in 9 of the 37 (24%) dogs within that diagnostic category by exclusion of other diseases and response to empirical treatment. All 19 dogs with an unknown cause of cough had unremarkable thoracic radiographic findings, allowing exclusion of pulmonary parenchymal diseases. No targeted treatments were pursued for these dogs because the owners perceived that the cough had no clinically important effect on the dogs' quality of life.
Sixteen of the 88 (18%) dogs with noncardiac causes of cough had been previously prescribed diuretics for suspected left-sided congestive heart failure by their referring veterinarians. Diuretic regimens for those 16 dogs included furosemide monotherapy (n = 9), furosemide and enalapril (2), furosemide and pimobendan (2), or all 3 medications (4). Median time between the presumptive diagnosis of congestive heart failure by the veterinarian and the visit of study inclusion was 90 days (range, 6 to 930 days). This diagnosis had been made on the basis of thoracic radiography interpretation (n = 12) or clinical findings alone (4). Owners reported no change in their dog's cough (n = 11) or mild improvement followed by worsening (5) in response to diuretic treatment. Final diagnoses for the underlying cause of cough in these 16 dogs included dynamic airway collapse (n = 9), bronchitis (2), bacterial pneumonia (2), and unknown cause (3). Ten of the 16 dogs had a heart murmur noted on physical examination. Fourteen dogs had mild structural heart disease identified on echocardiography (DMVD in 13 dogs and pulmonary hypertension in 1 dog), whereas 2 dogs had unremarkable echocardiograms. For the 13 dogs with DMVD, left atrial size was normal (n = 7) or mildly enlarged (6); mean ± SD LA:Ao in this group was 1.2 ± 0.2, and none of these dogs had an LA:Ao > 1.6. Because clinically important structural heart changes (including left atrial enlargement) were lacking, and little to no clinical response to treatment for congestive heart failure was observed, it was considered unlikely that these dogs truly had CPE in the past. Furosemide administration was discontinued for all dogs except 1, the owner of which disregarded the recommendation; no adverse consequences were noted in association with cessation of diuretic treatment.
LUS
The LUS examination was technically feasible (all images obtained at all described sites) and rapid (< 5 minutes) for all 100 dogs. No examinations were terminated early because of patient instability or behavioral noncompliance.
Overall, 66 (66%) dogs had at least 1 B line identified on LUS and 22 (22%) had at least 1 strong-positive Vet BLUE site (ie, site with > 3 or infinite B lines). Subpleural LUS abnormalities were noted in 20 (20%) dogs. Eleven dogs had the shred sign, 1 dog had the tissue sign, and 9 dogs had the nodule sign (1 dog had both the shred and nodule signs).
For the 12 dogs with CPE, strong-positive sites were most common in the right perihilar Vet BLUE site (n = 10) and least common in the left cranial and left perihilar sites (4 each). For the 7 dogs with bacterial pneumonia, strong-positive sites were most common in the right cranial Vet BLUE site (n = 3) but were also detected in the right middle and left middle sites (2 each); the shred sign was noted most commonly in the left middle (n = 3), right middle (2), or right cranial (2) Vet BLUE sites. The nodule sign was noted in all 4 dogs with pulmonary neoplasia (diagnosis confirmed in all dogs by cytologic evaluation of a fine-needle aspirate sample). Subpleural abnormalities were identified less commonly in dogs with fungal pneumonia (1/2), EBP (2/5), and dynamic airway collapse (5/37), and no shred, tissue, or nodule signs were noted in dogs with CPE (0/12).
Comparisons by diagnosis category
Compared with all other dogs combined, dogs with EBP (P < 0.001) or fungal pneumonia (P < 0.001) were younger. Dogs with dynamic airway collapse in the absence of left atrial enlargement (P = 0.003) weighed less, whereas dogs with bronchitis (P = 0.01), bacterial pneumonia (P < 0.001), or pulmonary neoplasia (P = 0.048) weighed more. Mean rectal temperature was higher in dogs with pulmonary neoplasia (P = 0.02) or fungal pneumonia (P < 0.001), mean heart rate was higher in dogs with CPE (P < 0.001) and lower in dogs with EBP (P = 0.02), and mean respiratory rate was higher in dogs with CPE (P = 0.01) or EBP (P = 0.005) than in all other dogs combined.
Heart murmurs were more prevalent in dogs with CPE (P = 0.001) or dynamic airway collapse with moderate to severe left atrial enlargement (P < 0.001) than in all other dogs combined. Mean LA:Ao and the prevalence of an LA:Ao > 1.6 were also higher in these 2 groups (P < 0.001 for both analyses) than among all other dogs combined.
Compared with all other dogs combined, dogs with CPE were more likely to have B lines (P = 0.007) and had a higher median total B-line score (22.5 vs 1; P < 0.001). Strong-positive sites were more common in dogs with CPE (P < 0.001) or bacterial pneumonia (P = 0.04). Dogs with CPE also had a higher median number of strong-positive sites (3) than all other dogs combined (0; P < 0.001).
Dogs with bacterial pneumonia were more likely than all other dogs combined to have the shred sign on LUS (P = 0.002). Dogs with pulmonary neoplasia were more likely than all other dogs combined to have the nodule sign (P < 0.001). Dogs with dynamic airway collapse were more likely to have had a prior presumptive diagnosis of congestive heart failure than other dogs (P = 0.002).
Diagnostic performance of LUS
Sensitivity and specificity of a total B-line score ≥ 10 for the diagnosis of CPE were 92% (11/12; 95% CI, 62% to 100%) and 94% (83/88; 95% CI, 87% to 98%), respectively. Sensitivity and specificity of ≥ 2 strong-positive sites for the diagnosis of CPE were also 92% (11/12; 95% CI, 62% to 100%) and 94% (83/88; 95% CI, 87% to 98%).
Sensitivity and specificity of the shred sign (present or not) for the diagnosis of bacterial pneumonia were 57% (4/7; 95% CI, 18% to 90%) and 92% (86/93; 95% CI, 85% to 97%), respectively. Sensitivity and specificity of the nodule sign (present or not) for the diagnosis of pulmonary neoplasia were 100% (4/4; 95% CI, 40% to 100%) and 95% (91/96; 95% CI, 88% to 98%). When B-line and subpleural findings were combined, sensitivity and specificity of a total B-line score ≥ 10 with no shred sign for the diagnosis of CPE were 92% (11/12; 95% CI, 62% to 100%) and 99% (87/88; 95% CI, 94% to 100%) and of ≥ 2 strongly positive Vet BLUE sites with no shred sign were 92% (11/12; 95% CI, 62% to 100%) and 98% (86/88; 95% CI, 92% to 100%).
Discussion
To the authors' knowledge, the present study provided the first description of LUS imaging findings in a large population of dogs with a primary clinical complaint of cough. Results indicated that LUS findings differed on the basis of the underlying cause of cough. Specifically, data supported our hypotheses that B lines were more common in dogs with CPE, subpleural shred signs were more common in dogs with bacterial pneumonia, and subpleural nodule signs were more common in dogs with pulmonary neoplasia than in other dogs.
The finding that B lines were more prevalent and numerous in dogs with (vs without) CPE echoed results of other LUS-based studies20,21,29,30 involving dogs. Previous studies have shown that the presence of multiple bilateral B lines can differentiate dogs with CPE from dogs with other causes of respiratory distress21 or from dogs with DVMD without active CPE.30 Accuracy of LUS for the diagnosis of CPE was higher in the 100 coughing dogs of the present study (sensitivity, 92%; specificity, 94%) than in a previous study21 that included 76 dogs with acute dyspnea (sensitivity, 83%; specificity, 70%). The higher sensitivity in the present study likely reflected the fact that the cutoff criterion for the number of B lines indicating CPE was lower for coughing dogs than for dyspneic dogs. In coughing dogs, the cutoff criterion that maximized diagnostic accuracy for CPE was ≥ 2 strong-positive (> 3 B lines/site) sites; in dyspneic dogs, the cutoff criterion was ≥ 2 strong-positive sites on each hemithorax (total of at least 4 strong-positive sites). The lower cutoff value for B lines in the present study could have suggested that dogs with CPE and cough as the predominant clinical sign are likely to have less pulmonary edema than dogs with a predominant clinical sign of acute dyspnea. However, we made no attempt to estimate radiographic lung water31 or directly compare results with those for dyspneic dogs with CPE.
The higher specificity of LUS for the diagnosis of CPE in coughing versus dyspneic dogs, despite a more sensitive cutoff criterion, could have been related to differences in the common non-CPE causes of coughing and dyspnea in those dogs. Non-CPE causes of dyspnea in the previous study21 included acute respiratory distress syndrome, pulmonary thromboembolism, and severe pneumonia; all of these diseases cause diffuse interstitial or alveolar disease and thus numerous B lines on LUS images.21 In contrast, the most common diagnosis for the coughing dogs of the present study was dynamic airway collapse, which would not be expected to cause B lines pathophysiologically. It follows that primary respiratory conditions associated with cough in dogs would be less likely to produce false-positive LUS results than would respiratory diseases that cause dyspnea.
Results of the study reported here suggested that identification of subpleural abnormalities on LUS could be useful in determining the cause of cough in dogs. The shred sign was noted in 4 of 7 dogs with bacterial pneumonia and was statistically overrepresented in that diagnosis group; presence of the shred sign was 57% sensitive and 92% specific for the diagnosis of bacterial pneumonia. In humans, the finding of subpleural lung consolidation is > 90% sensitive and specific for the diagnosis of pneumonia in children and adults with consistent clinical signs when thoracic CT is used as the reference standard.24 The lower diagnostic sensitivity for the dogs of the present study could have reflected differences in the severity, distribution, or underlying cause of pulmonary infection. Interestingly, the shred sign was also identified in 6 of 37 (16%) dogs with dynamic airway collapse. The explanation for this finding is not entirely clear; 1 possibility could involve mucous plugging with focal atelectasis, as reported for some cats with primary airway disease.32 Because shred signs were typically associated with B lines, dogs with bacterial pneumonia were also more likely than other dogs to have LUS sites with strong-positive scoring for B lines. Among all included coughing dogs, the only 2 diseases commonly associated with strong-positive LUS sites were CPE and bacterial pneumonia. These conditions could be differentiated on LUS on the basis of the distribution of strong-positive sites (with bacterial pneumonia preferentially affecting cranial and middle sites) and presence of the shred sign (occurring exclusively in dogs with bacterial pneumonia).
The nodule sign was detected in all 4 dogs with pulmonary neoplasia, suggesting that LUS was quite sensitive for the diagnosis of neoplasia in this small group of dogs. This sign was also detected in dogs with fungal pneumonia, bacterial pneumonia, bronchitis, dynamic airway collapse, and congenital pulmonary anomaly (1 dog each). In the dogs with pneumonia and congenital pulmonary anomaly, we suspect that these subpleural lesions were focal circular to ellipsoid areas of lung consolidation or inflammation related to the underlying pulmonary disease. In the dogs with bronchitis and bronchial collapse, no pulmonary abnormalities were noted on thoracic radiography; it remains unknown whether the nodule signs on LUS represented radio-occult pulmonary lesions or false-positive LUS results. Patterns of nodule signs in our study were similar to results from a case series of 27 dogs and cats, in which LUS was also 100% sensitive (but only 68% specific) for the diagnosis of pulmonary metastasis when thoracic radiography was used as the reference standard.d
A single dog in the present study had a subpleural tissue sign noted on LUS. This sign represents a more severe form of lung consolidation than the shred sign and may be associated with various types of pulmonary abnormalities.22,33 The single dog with the tissue sign had EBP, and the triangular area of lung consolidation seen on LUS was likely a focal area of severe consolidating inflammation or atelectasis. Overall, the patterns of subpleural abnormalities in different disease groups suggested that subpleural findings could have diagnostic value for dogs. However, given the small patient numbers within individual disease categories, additional research is needed to assess the diagnostic usefulness of LUS shred, nodule, and tissue signs for dogs with various pulmonary diseases.
Not surprisingly, several clinical values differed between disease groups in the study reported here. Compared with the larger cohort of all other coughing dogs, patients with fungal pneumonia or EBP were younger, dogs with dynamic airway collapse had a lower body weight, dogs with neoplasia and fungal disease had a higher rectal temperature, and dogs with CPE had a higher heart rate. Mild tachypnea was noted in some dogs across disease groups, but respiratory rates were highest in dogs with CPE, EBP, and fungal pneumonia (Table 1). These findings were all consistent with signalment or pathophysiologic factors associated with various causes of cough. Heart murmurs were common, observed in 58 (58%) dogs. The most common cause of heart murmurs was DMVD, accounting for 55 of these 58 (95%) dogs. Presence of heart disease with moderate to severe left atrial enlargement (LA:Ao > 1.6) was a necessary criterion for dogs to be considered to have CPE (n = 12) or dynamic airway collapse associated with left atrial enlargement (15). Milder heart disease was noted commonly (25% to 50% of dogs) across all other diagnosis groups, except that no concurrent heart disease was identified in dogs with fungal disease or EBP, likely owing to their younger age. The prevalence of concurrent heart disease was not surprising; in the authors' experience, dogs with DMVD are generally older small-breed dogs, a patient population also commonly affected by concurrent airway disease (dynamic airway collapse, bronchitis, or both). Indeed, the clinical scenario of a coughing geriatric small-breed dog with a heart murmur is common and can represent a diagnostic and therapeutic challenge. The cause of cough in such patients could be multifactorial, possibly involving concurrent primary respiratory disease (intrinsic tracheobronchomalacia or chronic bronchitis), CPE, or bronchial compression from cardiomegaly.7–9
Controversy exists surrounding the question of whether CPE itself actually causes cough. No cough receptors are present in the alveoli or interstitium of the lung, suggesting that edema would have to be severe enough to fill airways up to the level of bronchioles to cause cough.5,6 Indeed, CPE was not a significant predictor of cough in a study7 of dogs with DMVD; in that study, presence of cough was significantly associated only with left atrial enlargement and an abnormal radiographic airway pattern.7 In the present study, only 12 (12%) coughing dogs eventually received a diagnosis of CPE. During the period of study enrollment, a total of 53 dogs received a diagnosis of CPE at the same institution. The finding that a minority (23%) of dogs with left-sided congestive heart failure had cough as a notable clinical feature supported the notion that cough is a possible, but not common, clinical sign of CPE in dogs.7,9 However, veterinary textbooks3,4 continue to list cough as a common primary clinical complaint for dogs with CPE.
An important finding of the present study was that a relevant proportion (18%) of dogs had previously received furosemide treatment for presumptive congestive heart failure prior to referral to our hospital. These diagnoses were considered misdiagnoses given results of subsequent thoracic imaging (echocardiography and thoracic radiography) that showed minimal to no cardiac disease, lack of a positive response to diuretic treatment, and lack of adverse outcome after furosemide treatment was discontinued. Referral records indicated that congestive heart failure was suspected in these dogs on the basis of clinical findings (heart murmur and cough) and, for 12 of 16 dogs, radiographic interpretation (cardiomegaly and suspicion of pulmonary edema). Review of previously obtained thoracic radiographs often revealed that the perceived abnormalities actually reflected normal breed variation in heart size combined with suboptimal radiographic exposure. These observations suggested that general practice veterinarians might have an overly high index of suspicion for CPE in coughing dogs, particularly dogs with heart murmurs or cardiomegaly, which could result in radiographic overdiagnosis of congestive heart failure. Lung ultrasonography could be particularly useful in this specific diagnostic situation, given that the absence of B lines on LUS in the coughing dogs of the present study made CPE highly unlikely. Use of LUS could thus increase diagnostic confidence in situations when thoracic radiographs are suboptimal or related findings are equivocal, allowing practitioners to avoid unnecessary use of diuretics and to focus on other relevant differential diagnoses for cough.
The present study has several limitations. First, although the overall sample size was fairly large (n = 100), the number of dogs with each of the specific diagnoses was low (< 10 in many cases). This low number likely limited statistical power, making it difficult to draw conclusions about different patterns of LUS findings for certain diagnoses. Additional research is needed to investigate the diagnostic usefulness of LUS in larger cohorts of dogs with specific respiratory conditions.
Second, the reference standard used in determinations of the sensitivity and specificity of LUS for the diagnosis of various diseases was the composite clinical diagnosis based on review of all available medical record (except LUS) data. Although an ideal study design would have involved requiring a standardized suite of diagnostic tests for all dogs, not all owners opted for a thorough workup to elucidate the specific underlying respiratory disease. Nineteen dogs never received a final diagnosis for their cough, although all had unremarkable findings on thoracic radiography, allowing exclusion of many differential diagnoses. These patients were still included in the overall sample of coughing dogs to avoid overestimation of prevalence of other underlying causes of cough. Dogs lacking definitive diagnosis were also included in statistical comparisons of LUS and clinical findings across diagnosis categories because such dogs could at least be categorized as not having certain diseases (eg, CPE, bacterial pneumonia, or other pulmonary parenchymal diseases). However, further testing to confirm an underlying diagnosis would have augmented our understanding of LUS findings in individual respiratory diseases.
Third, the cause of cough was determined on the basis of clinical evidence of an underlying respiratory disease and improvement in cough following targeted treatment for that disease. However, such a diagnosis may not have reflected the true underlying cause of cough. For example, a dog with pulmonary neoplasia might have been coughing because of the tumor itself or from concurrent chronic bronchitis, either of which might have improved with glucocorticoid treatment. Similarly, a dog with a diagnosis of congestive heart failure might have had coughing related to CPE itself or to bronchial compression from cardiomegaly, either of which might have responded to preload reduction with furosemide. For study purposes, the law of parsimony was followed in presuming that the primary pulmonary diagnosis was also the cause of cough. However, we acknowledge that the cough reflex is complex and causes of cough may be multifactorial.
Fourth, in humans, the distribution of LUS B lines can be affected by gravity and positioning. For example, in humans imaged in a semirecumbent position, lower (more gravity-dependent) lung regions reportedly have higher B-line scores than higher (nondependent) areas.34 To minimize patient discomfort and the need for restraint, all dogs in the present study were imaged when standing or in sternal recumbency. It is therefore possible that distribution of B lines might have been different if dogs had been imaged when in lateral or dorsal recumbency. An additional study limitation was that LUS images were each reviewed by only 1 blinded investigator with experience in LUS; operator experience may thus have played a role in improving accuracy of LUS in this study. Previous findings suggest that interobserver agreement for identifying and counting B lines is excellent, even when comparing novice and experienced observers.21 However, interobserver agreement for identification of other subpleural abnormalities has not been established, and results of the present study may have been different with other observers.
A general limitation of LUS is the requirement for a point-of-care ultrasonography machine, which may not be widely available in a primary care setting. However, portable ultrasonography units are becoming increasingly common in veterinary hospitals. In a 2016 email survey conducted by one of the authors (TCD), > 50% of responding North Carolina general veterinary practitioners reported having an ultrasonography machine in their practice. Given the frequency with which CPE appears to be misdiagnosed in coughing dogs as indicated by the findings reported here, LUS offers practitioners another tool to help increase or decrease index of suspicion for CPE and prevent unnecessary use of diuretics. Certainly, complete echocardiography remains the reference standard for establishing the presence of severe structural heart disease that may be an underlying cause of CPE. Furthermore, advanced echocardiographic measurements (eg, ratio of early diastolic mitral inflow velocity to isovolumic relaxation time and ratio of early diastolic mitral inflow velocity to velocity of early diastolic mitral annular motion) are useful noninvasive surrogate measurements of increased left atrial pressure that can suggest the presence of CPE.35,36 However, performance of an accurate and comprehensive echocardiogram, including spectral and tissue Doppler measurements, requires expensive equipment (eg, phased-array transducers and advanced cardiac ultrasound software) and operator expertise (eg, board-certified cardiologist). The major benefit of LUS in this setting is allowing rapid cageside detection of CPE by practitioners in general practice settings by use of a typical portable ultrasonography machine.
In the present study of 100 coughing dogs, LUS results differed on the basis of the underlying cause of cough. B lines were more numerous in dogs with CPE, subpleural shred signs were more common in dogs with bacterial pneumonia, and subpleural nodule signs were noted in all dogs with pulmonary neoplasia. Cardiogenic pulmonary edema was a relatively uncommon cause of cough (12% of dogs), although a relevant proportion of study dogs had previously been treated with a diuretic because of prior CPE misdiagnosis. Lung ultrasonography therefore appeared to be a valuable diagnostic tool in the evaluation of coughing dogs, particularly to identify or exclude CPE.
Acknowledgments
This study was performed at the Lloyd Veterinary Medical Center, College of Veterinary Medicine, Iowa State University.
This study was not supported by any grant or funding source other than the corresponding author's start-up funds from Iowa State University.
Dr. Lisciandro is the owner of FASTVet.com, a private corporation that provides veterinary ultrasound training to practicing veterinarians. He teaches ultrasonography courses for Sound, SonoSite, EI Medical, and scil animal care and has licensed education materials to EI Medical. He has also received ultrasound equipment on loan from SonoSite, Sound, EI Medical, and scil animal care.
Presented in abstract form at the American College of Veterinary Internal Medicine Forum 2017, National Harbor, Md, June 2017.
The authors thank Lori Moran for technical assistance.
ABBREVIATIONS
| CI | Confidence interval |
| CPE | Cardiogenic pulmonary edema |
| DMVD | Degenerative mitral valve disease |
| EBP | Eosinophilic bronchopneumopathy |
| LA:Ao | Left atrial diameter-to-aortic root diameter ratio |
| LUS | Lung ultrasonography |
| Vet BLUE | Veterinary bedside lung ultrasound examination |
Footnotes
CX50 CompactXtreme ultrasound (model No. 101855), Philips Healthcare, Andover, Mass.
C8-5 transducer (part No. FUS5124), Philips Healthcare, Andover, Mass.
MedCalc, version 17.6, MedCalc Software, Ostend, Belgium.
Kulhavy DA, Lisciandro G. Use of a lung ultrasound examination (Vet BLUE) to screen for metastatic lung nodules in the emergency room (abstr). J Vet Emerg Crit Care 2015;25:S14.
References
1. Canning BJ. Anatomy and neurophysiology of the cough reflex. Chest 2006;129:33S–47S.
2. Ferasin L. Coughing. In: Ettinger SJ, Feldman EC, Cote E, eds. Textbook of veterinary internal medicine. 8th ed. St Louis: Elsevier, 2017;107–111.
3. Ware WA. Cough. In: Ware WA, ed. Cardiovascular diseases in small animal medicine. London: Manson Publishing, 2011;101–105.
4. Fox PR. The history. In: Fox P, Sisson D, Moise N, eds. Textbook of canine and feline cardiology. 2nd ed. London: Saunders, 1999;41–45.
5. Mazzone SB. An overview of the sensory receptors regulating cough. Cough 2005;1:2.
6. Widdicombe J. Airway receptors. Respir Physiol 2001;125:3–15.
7. Ferasin L, Crews L, Biller DS, et al. Risk factors for coughing in dogs with naturally acquired myxomatous mitral valve disease. J Vet Intern Med 2013;27:286–292.
8. Guglielmini C, Diana A, Pietra M, et al. Use of the vertebral heart score in coughing dogs with chronic degenerative mitral valve disease. J Vet Med Sci 2009;71:9–13.
9. Singh MK, Johnson LR, Kittleson MD, et al. Bronchomalacia in dogs with myxomatous mitral valve degeneration. J Vet Intern Med 2012;26:312–319.
10. Tart KM, Babski DM, Lee JA. Potential risks, prognostic indicators, and diagnostic and treatment modalities affecting survival in dogs with presumptive aspiration pneumonia: 125 cases (2005–2008). J Vet Emerg Crit Care (San Antonio) 2010;20:319–329.
11. Atkins C, Bonagura J, Ettinger S, et al. ACVIM consensus statement guidelines for the diagnosis and treatment of canine chronic valvular heart disease. J Vet Intern Med 2009;23:1142–1150.
12. Mantis P, Lamb CR, Boswood A. Assessment of the accuracy of thoracic radiography in the diagnosis of canine chronic bronchitis. J Small Anim Pract 1998;39:518–520.
13. Johnson LR, Singh MK, Pollard RE. Agreement among radiographs, fluoroscopy and bronchoscopy in documentation of airway collapse in dogs. J Vet Intern Med 2015;29:1619–1626.
14. Armbrust LJ, Biller DS, Bamford A, et al. Comparison of three-view thoracic radiography and computed tomography for detection of pulmonary nodules in dogs with neoplasia. J Am Vet Med Assoc 2012;240:1088–1094.
15. Lichtenstein D, Mézière GA, Biderman P, et al. The comet-tail artifact: an ultrasound sign of alveolar-interstitial syndrome. Am J Respir Crit Care Med 1997;156:1640–1646.
16. Volpicelli G, Elbarbary M, Blaivas M, et al. International evidence-based recommendations for point-of-care lung ultrasound. Intensive Care Med 2012;38:577–591.
17. Baad M, Lu ZF, Reiser I, et al. Clinical significance of US artifacts. Radiographics 2017;37:1408–1423.
18. Al Deeb M, Barbic S, Featherstone R, et al. Point-of-care ultrasonography for the diagnosis of acute cardiogenic pulmonary edema in patients presenting with acute dyspnea: a systematic review and meta-analysis. Acad Emerg Med 2014;21:843–852.
19. Hassdenteufel E, Henrich E, Hildebrandt N, et al. Assessment of circulating N-terminal pro B-type natriuretic peptide concentration to differentiate between cardiac from noncardiac causes of pleural effusion in cats. J Vet Emerg Crit Care (San Antonio) 2013;23:416–422.
20. Lisciandro GR, Fosgate GT, Fulton RM. Frequency and number of ultrasound lung rockets (B-lines) using a regionally based lung ultrasound examination named Vet BLUE (veterinary bedside lung ultrasound exam) in dogs with radiographically normal lung findings. Vet Radiol Ultrasound 2014;55:315–322.
21. Ward JL, Lisciandro GR, Keene BW, et al. Accuracy of point-of-care lung ultrasonography for the diagnosis of cardiogenic pulmonary edema in dogs and cats with acute dyspnea. J Am Vet Med Assoc 2017;250:666–675.
22. Lee PM, Tofts RP, Kory P. Lung ultrasound interpretation. In: Soni NJ, Arntfield R, Kory P, eds. Point-of-care ultrasound. Philadelphia: Elsevier, 2015;59–69.
23. Pereda MA, Chavez MA, Hooper-Miele CC, et al. Lung ultrasound for the diagnosis of pneumonia in children: a meta-analysis. Pediatrics 2015;135:714–722.
24. Chavez MA, Shams N, Ellington LE, et al. Lung ultrasound for the diagnosis of pneumonia in adults: a systematic review and meta-analysis. Respir Res 2014;15:50–59.
25. Lisciandro GR. The Vet BLUE lung scan. In: Lisciandro GR, ed. Focused ultrasound techniques for the small animal practitioner. Ames, Iowa: Wiley Blackwell, 2014;166–187.
26. Hansson K, Häggström J, Kvart C, et al. Left atrial to aortic root indices using two-dimensional and M-mode echocardiography in Cavalier King Charles Spaniels with and without left atrial enlargement. Vet Radiol Ultrasound 2002;43:568–575.
27. Gargani L. Lung ultrasound: a new tool for the cardiologist. Cardiovasc Ultrasound 2011;9: 6.
28. Lichtenstein D. Should lung ultrasonography be more widely used in the assessment of acute respiratory disease? Expert Rev Respir Med 2010;4:533–538.
29. Rademacher N, Pariaut R, Pate J, et al. Transthoracic lung ultrasound in normal dogs and dogs with cardiogenic pulmonary edema: a pilot study. Vet Radiol Ultrasound 2014;55:447–452.
30. Vezzosi T, Mannucci T, Pistoresi A, et al. Assessment of lung ultrasound B-lines in dogs with different stages of chronic valvular heart disease. J Vet Intern Med 2017;31:700–704.
31. Agricola E, Bove T, Oppizzi M, et al. Ultrasound comet-tail images: a marker of pulmonary edema. Chest 2005;127:1690–1695.
32. Gadbois J, Anjou M, Dunn M. Radiographic abnormalities in cats with feline bronchial disease and intra- and interobserver variability in radiographic interpretation: 40 cases (1999–2006). J Am Vet Med Assoc 2009;234:367–375.
33. Mathis G, Blank W, Reissig A, et al. Thoracic ultrasound for diagnosing pulmonary embolism: a prospective multicenter study of 352 patients. Chest 2005;128:1531–1538.
34. Cortellaro F, Ceriani E, Spinelli M, et al. Lung ultrasound for monitoring cardiogenic pulmonary edema. Intern Emerg Med 2017;12:1011–1017.
35. Schober KE, Hart T, Stern J. Detection of congestive heart failure in dogs by Doppler echocardiography. J Vet Intern Med 2010;24:1358–1368.
36. Oyama MA, Sisson DD, Bulmer BJ, et al. Echocardiographic estimation of mean left atrial pressure in a canine model of acute mitral valve insufficiency. J Vet Intern Med 2004;18:667–672.
