The coughing dog commonly presents both a diagnostic and therapeutic dilemma to the practicing veterinarian, the former often begetting the latter. This is particularly true in middle-aged to aged, small breed dogs which are often afflicted with chronic respiratory disease and mitral insufficiency, each of which may result in coughing. The problem, then becomes one of distinguishing the degree to which respiratory and/or cardiovascular disease are contributing to the cough. This is important because the therapies are quite different and, in some instances, appropriate therapy for one worsens the other.

Causes of coughing are many, but in the population of dogs under discussion, cough most commonly results from upper airway disease, tracheobronchial collapse, chronic bronchitis of varying etiologies, pulmonary fibrosis, bronchiectasis, pulmonary neoplasia, pneumonia, and mitral regurgitation (the latter, usually but not always with heart failure). In mitral regurgitation, the cause of cough is not entirely clear, but left atrial impingement on the left main bronchus and/or recurrent laryngeal nerve, edema of the respiratory mucosa, interstitial fluid pressure on airways, and excessive bronchial mucous production have all been suggested as being contributory.

The diagnosis of chronic, unresponsive cough is challenging, particularly when concurrent cardiac disease is identified. Often, referral of such cases is warranted, providing such diagnostic opportunities as fluoroscopy, electrocardiography, echocardiography, oximetry, bronchoscopy, bronchoalveolar lavage, culture and sensitivity, cytological evaluation, lung biopsy or fine needle aspiration, serology (e.g., fungal, toxoplasma, heartworm, etc), specialized fecal examinations (sedimentation and Baerman examination for respiratory parasites), and arterial and venous blood gas analysis. In many instances however, referral centers are inconvenient, owners decline referral, or referral is unnecessary. Virtually all small animal practices can, through careful historical, physical, and special procedural examination, provide an excellent diagnostic workup for dogs with cough and mitral regurgitation.

Special procedures typically available to private practitioners include inspiratory and expiratory thoracic and cervical radiographs to evaluate for airway collapse or infiltrate, cardiac size, evidence of heart failure, and the pulmonary parenchyma; electrocardiography to determine heart rate and the presence of left and/or right atrial and/or ventricular enlargement; transtracheal wash as a substitute for bronchoscopy, with culture and cytological evaluation; and trial diuresis.

Historical and physical findings which may indicate a respiratory cause for chronic cough include obesity, harsh cough, mucopurulent nasal discharge, harsh, "honking" and/or productive cough, and respiratory wheezes. Contrarily, loss of condition, soft, non-productive cough, often worse at night, accompanying dyspnea (may occur with severe respiratory disease as well), and tachycardia with weak pulses are more indicative of a cardiac cause for the cough.

The complete blood count may be useful in indicating eosinophilia and possibly basophilia which are most compatible with allergic or parasitic disorders, such as parasitic or eosinophilic pneumonias or heartworm disease. With bacterial pneumonia, there may be indicators of sepsis, such as neutrophilia, toxic changes in neutrophils, and monocytosis. It is emphasized that hematological changes are inconsistent, at best. In heartworm-endemic areas, appropriate testing (antigen Elisa for dogs on monthly preventative; antigen Elisa and/or Knott test if on no or daily preventative).

Overall, the most useful tool is the thoracic radiograph. It is beyond the scope of this manuscript to review thoracic radiography. Specific changes useful in differentiating cardiac from respiratory causes of cough include distinction of left vs right heart enlargement, indicative respectively of cardiac and respiratory causes. The status of the lung, airways, and pulmonary vasculature is, likewise, often elucidating. Bronchial thickening, pulmonary infiltrate (other than pulmonary edema), pulmonary arterial enlargement (apical pulmonary artery larger than accompanying vein or proximal onethird of the fourth rib where they intersect) suggest respiratory disease with the latter finding most consistent with pulmonary hypertension (typically heartworm disease). Contrarily, enlargement of the pulmonary vein (apical pulmonary vein greater than accompanying artery or greater than the proximal one-third of the fourth rib where they intersect) suggests pulmonary venous congestion (left heart failure). Evaluation of the airways for dynamic function is best performed with fluoroscopy but can be approximated by maximizing the information gleaned from conventional radiography. This can be accomplished by supplementing routine full inspiratory lateral and ventrodorsal radiographs with a lateral exposure at full expiration, and ideally, a full inspiratory exposure focused over the cervical trachea. The former will demonstrate intrathoracic and the latter, extrathoracic airway collapse. It should be emphasized that a negative dynamic study, such as this, does not definitively rule out airway collapse.

Electrocardiographic findings compatible with respiratory disease (and heartworm disease which in reality is respiratory--or pulmonary arterial--disease) include slow rate, sinus arrhythmia, P-pulmonale (P waves >0.4mv, indicating right atrial enlargement), and occasionally a right ventricular enlargement pattern (S waves in leads 1, 2 and 3, right axis deviation, and deep S waves in V₃). Alternatively, dogs with mitral valvular insufficiency typically have normal mean electrical axis, tachycardia if in heart failure, P mitrale (widened and sometimes notched P waves), and possibly, left ventricular enlargement pattern (tall R waves and/or prolonged QRS complexes). Specific information about the nature of respiratory disease can be obtained routinely with transtracheal aspiration (wash). This is accomplished under mild sedation (or none in debilitated animals) so that the cough reflex is not blunted. Standard surgical preparation is performed over the larynx. The cricothyroid ligament is palpated by feeling the indentation; this is accentuated by flexing and extending the neck. Lidocaine is infiltrated subcutaneously over this site. The dog is positioned in sternal recumbency or in a sitting position. It is important to keep the dog symmetrically positioned to allow accurate assessment of anatomic landmarks. The neck is extended and a small stab incision made over the cricothyroid membrane. A 12 to 18 inch 14 to 16 gauge intravenous catheter is used to puncture the membrane, with the trachea digitally stabilized. The catheter is directed down into the airway and the catheter advanced. Coughing indicates successful entry into the airway. Once the catheter has been fully advanced, the needle is backed from the trachea and the needle guard applied. The metal stylet is removed and nonbacterostatic saline (approximately .4 ml/kg) is rapidly infused and re-aspirated. Only a small fraction of the infusate is retrieved. This washing procedure may be repeated once or twice. The evidence of mucous or pus in the retrieved material suggests a successful wash. The material is placed in appropriate transport media and submitted for bacterial (and possibly fungal) culture and for cytological evaluation. Pressure should be applied to the entry site for a full 5 minutes after the catheter is remove. I prefer not to perform this procedure on dogs less than approximately 10-15 kg; in smaller dogs, the procedure is performed through a sterile endotracheal tube, using a red rubber feeding tube and brief general anesthesia. Complications to this procedure are uncommon but hospitalization or close observation at home are advised. Sedation for the first 12 hours is useful in excitable dogs or those with intractable, violent coughing.

Bronchoscopy provides the advantage of allowing direct visualization of the airways and selective sampling of specific sites. Disadvantages include cost, the need for special expertise and equipment, and the necessity of anesthesia. Superior cytological samples can be obtained using bronchoalveolar lavage, a technique which also requires general anesthesia and is best performed using bronchoscopy. This technique, which may compromise animals with severe respiratory disease, samples specific lung lobes by filling the alveoli in the region with sterile saline and aspirating the contents. Large numbers of cells are recovered and, because samples are diluted, clumping of cells with mucous is minimized. Bronchoalveolar lavage is generally reserved for cases in which less invasive methods are unsuccessful. Like the techniques mentioned above, bronchoalveolar lavage is most advantageously employed in cases with other than pure interstitial disease, as the samples represent primarily the airways and alveoli.

Lastly, a therapeutic trial of an off-loading drug, such as furosemide, can be useful in ruling in or out cardiac disease as the cause for coughing. Such therapy will clear edema and shrink the left atrium, relieving the cause of cardiac cough. If the cough is purely respiratory in origin, this maneuver is unlikely to be of benefit. Several concepts are important if this approach is used. First, the dosage of furosemide must be adequate and the duration of therapy long enough (.5-1 mg/kg tid for 2-4 days) to allow firm conclusion as to its efficacy in controlling the cough. Second, this approach should not involve any other drugs, so that the exact cause of a favorable response is evident. Lastly, the owner needs to be educated to the fact that this is a diagnostic test, that his or her evaluation of the response is important, and that if successful, the dosage will be lowered and other drugs, such as angiotensin converting enzyme inhibitors, will be added to the regimen.

It should be emphasized that, in some cases, respiratory and cardiac diseases may coexist and work in concert to produce the cough. In addition, neither respiratory or chronic mitral valvular diseases are static, so conclusions drawn regarding the cause of a cough correctly drawn today may no longer be valid over time.

Differentiating Cough Due to Cardiac and Respiratory

Note that findings vary and not all abnormalities are seen in a given case and that overlap of diseases and signs may occur. NSR = normal sinus rhythm, NSA = normal sinus arrhythmia, MR = mitral regurgitation, RA = right atrium, RV = right ventricle, LA = left atrium, LV = left ventricle, TR = tricuspid regurgitation, PI = pulmonary insufficiency, PHT = pulmonary hypertension.