Clinical Signs, Clinicopathological Findings, Etiology, and Outcome Associated With Hemoptysis in Dogs: 36 Cases (1990–1999)

Introduction

Hemoptysis, derived from the Greek words haima (meaning blood) and ptysis (meaning spitting), is defined as the expectoration of blood or bloody mucus from the respiratory tract at or below the larynx.¹ Hemoptysis can arise from primary disorders of the lung or from diseases of other organ systems that are manifest through the respiratory tract. Hemoptysis must be differentiated from pseudohemoptysis (i.e., coughing up hemorrhagic material from bleeding sites cranial to the larynx, including oral or nasal regions) and hematemesis (i.e., vomiting blood). History and physical examination findings alone may not be sufficient to distinguish these conditions. Observation of the hemoptysis is ideal, but not always possible. Further diagnostic tests, including thoracic radiography, endoscopic examination of the respiratory tract, cytopathology of material obtained from fine-needle aspiration (FNA), transtracheal wash (TTW) or bronchoalveolar lavage (BAL), histopathology of the lung, or a combination of the above, are instrumental in differentiating hemoptysis from pseudohemoptysis and hematemesis.

Hemoptysis in dogs has been reported as a clinical sign of primary or metastatic pulmonary neoplasia; bacterial, fungal, parasitic, and foreign-body pneumonia; thoracic trauma; lung-lobe torsion; heartworm disease; pulmonary thromboembolism; coagulopathies; left-sided congestive heart failure; and from iatrogenic FNA biopsy procedures.^2–6 In humans, retrospective studies of hemoptysis have been beneficial in evaluating underlying etiologies, providing prognostic information, and establishing a foundation for further diagnostic and therapeutic investigation.⁷ To the authors’ knowledge, no retrospective study has been performed to characterize features of hemoptysis in dogs. The purpose of this study was to determine the signalment, clinical manifestations, laboratory and radiographic data, underlying diseases, and outcomes for dogs presenting with a primary complaint of hemoptysis.

Materials and Methods

Medical records of dogs seen at the Veterinary Medical Teaching Hospital at the University of California at Davis between January 1990 and December 1999 were searched for cases in which hemoptysis was a primary complaint. For inclusion in the study, the underlying cause of bleeding into the respiratory tract had to have been determined. Diagnostic tests useful in the confirmation of the underlying disease included complete blood count (CBC); bone-marrow aspiration cytopathology; activated clotting time (ACT), prothrombin time (PT), activated partial thromboplastin time (aPTT), and proteins induced by vitamin K antagonism (PIVKA); heartworm antigen test; microfilarial test; thoracic radiography; TTW with cytopathology, culture, or both; bronchoscopic examination; BAL fluid cytopathology, culture, or both; bronchial mucosal biopsy; FNA cytopathology; echocardiography; pulmonary arterial catheterization; and lung biopsy (antemortem or postmortem).

Specific criteria to confirm the underlying disease associated with hemoptysis were used in this study. Trauma was diagnosed based on history and compatible clinical findings. Immune-mediated thrombocytopenia (ITP) was determined based on the finding of decreased platelet numbers and megakaryocytic hyperplasia on bone-marrow aspiration, with or without a positive antimegakaryocytic antibody test. A diagnosis of rodenticide poisoning was based on history; prolonged ACT, PT, aPTT, PIVKA, or a combination of the above; and response to therapy with vitamin K. Bacterial pneumonia was diagnosed if a dog had positive culture of lavage fluid from the lower respiratory tract, supportive septic cytopathology (as determined by TTW or BAL) or histopathology, and response to therapy with antibiotics. Foreign-body pneumonia was diagnosed if a dog had foreign material at the time of bronchoscopic evaluation. Lung-lobe torsion was confirmed by thoracic radiography and thoracotomy. Heartworm disease was confirmed with a positive antigen or microfilarial test, visualization of worms on echocardiography, or a combination of the above. Left-sided congestive heart failure was diagnosed with thoracic radiography and compatible changes seen on echocardiography. Pulmonary hypertension was diagnosed via echocardiography and pulmonary artery catheterization with pulmonary arterial pressure measurement. Neoplasia was confirmed on histopathological examination. Chronic bronchitis was diagnosed based on clinical signs, bronchoscopy with BAL, histopathological examination, or a combination of the above. Disseminated intravascular coagulation (DIC) was diagnosed based on clinical findings, elevated clotting times, thrombocytopenia, and decreased fibrinogen concentrations.

In all dogs, signalment, duration of hemoptysis, clinical signs, laboratory results, thoracic radiographic findings, underlying diseases, and outcome were obtained from each record.

Results

Thirty-six dogs met the criteria for inclusion in the study. The median age was 5 years (range, 6 months to 13 years). There were 24 males (eight castrated) and 12 females (eight spayed). Purebred dogs that were represented included Labrador retriever (n=10), chow chow (n=4), Australian cattle dog (n=2), German shorthaired pointer (n=2), Brittany spaniel (n=2), border collie (n=2), and one each of the following breeds: rottweiler, Shetland sheepdog, miniature poodle, Belgian sheepdog, Rhodesian ridgeback, American cocker spaniel, Belgian Malinois, and basset hound. Additionally, there were six mixed-breed dogs.

Clinical signs related to the respiratory tract that were reported by owners included episodes of coughing without hemoptysis (n=18), dyspnea (n=4), tachypnea (n=3), exercise intolerance (n=3), and collapse (n=2). Other clinical signs reported included lethargy (n=3), weight loss (n=3), hematochezia (n=3), melena (n=3), hematuria (n=2), and pallor (n=1). The duration of hemoptysis began <1 week prior to presentation in all cases, except in one case that had a 3-month history of hemoptysis. For dogs with a cough noted prior to the onset of hemoptysis, the duration of cough ranged from 1 day to 1 year [see Table].

Abnormalities detected on physical examination that were associated with the respiratory system included tachypnea (n=7), dyspnea (n=7), spontaneous coughing (n=7), hemoptysis (n=7), crackles (n=5), wheezes (n=3), cyanosis (n=2), and diminished lung sounds (n=1). Other abnormalities included hyperthermia (n=7), pallor (n=6), cardiac murmurs (n=5), dehydration (n=3), physical signs of trauma (n=3), petechia (n=3), ecchymoses (n=1), melena (n=1), hypothermia (n=1), palpable abdominal fluid wave (n=1), ventral and limb edema (n=1), and cardiac arrhythmia (n=1). Eight patients had signs of systemic bleeding historically or at the time of physical examination, which included melena, hematochezia, ecchymoses, petechiation, and hematuria.

Diagnostic tests performed during the investigation of the cause of hemoptysis varied among dogs and included thoracic radiography (n=35), hematocrit (HCT)/total protein concentration (TP) (n=35), platelet count (n=25), PT/aPTT (n=11), heartworm antigen test (n=10), microfilarial test (n=9), culture of TTW or BAL fluid (n=9), echocardiography (n=8), open thoracotomy with biopsy (n=6), PIVKA (n=6), bronchoscopy (n=6), TTW or BAL cytopathology (n=5 each), necropsy (n=5), ACT (n=5), bone-marrow aspirate (n=4), antimegakaryocyte antibody test (n=3), bronchoscopic mucosal biopsy (n=3), pulmonary FNA cytopathology (n=2), transthoracic lung biopsy (n=1), and pulmonary artery catheterization (n=1).

Many dogs had abnormal clinical pathology results. The median HCT was 42.5% (reference range, 37% to 55%),and the range was 9.4% to 57%. Although 11 dogs had a HCT value below the low end of the reference range, only one dog had a HCT <20%. The median TP was 6.3 mg/dL (reference range, 6.0 to 8.0 mg/dL), and the range was 3.7 to 9.3 mg/dL. The median platelet count was 241 × 10³/μL (reference range, 200 to 500 × 10³/μL), and the range was 1 to 497 × 10³/μL. Twelve dogs had platelet counts below the reference range, but only four dogs had values <30 × 10³/μL. Bone-marrow aspirates from these four dogs revealed megakaryocytic hyperplasia and positive antimegakaryocytic antibody results in three. The ACT was elevated in three dogs at 217, 360, and >999 seconds (reference range, <120 seconds). The PIVKA was within the reference range (15 to 18 seconds) in four dogs and was >300 seconds in two dogs tested. Only one dog had elevated PT and aPTT (PT, 97.4 seconds; aPTT, 45.5 seconds; reference range, 7.5 to 10.5 seconds and 9 to 12 seconds, respectively). This one dog had thrombocytopenia, elevated PT and aPTT, and PIVKA as well as a decreased fibrinogen of 100 mg/dL, and was subsequently diagnosed with disseminated intravascular coagulation. A heartworm antigen test was positive in four dogs, one of which also had a positive microfilarial test.

Lateral and dorsoventral radiographs were obtained in 35 dogs; they revealed abnormalities in all patients. The predominant radiographic patterns were alveolar (n=15), interstitial (n=6), mixed interstitial-alveolar (n=5), and bronchial (n=1); in addition, a pulmonary mass (or masses) was seen in four dogs, and lobar consolidation was seen in six dogs. The distribution of lesions on thoracic radiography was focal (n=18), diffuse (n=13), and multifocal (n=4). Other radiographic findings included cardiomegaly (n=8), pleural effusion (n=5), enlarged pulmonary arteries (n=4), pneumothorax (n=3), subcutaneous emphysema (n=1), fluid-filled bullae (n=1), microcardia (n=1), mediastinal mass (n=1), missile foreign body (n=1), perihilar lymphadenopathy (n=1), and chest wall mass (n=1).

Echocardiography was useful in obtaining a definitive diagnosis in three dogs. It documented a ruptured cordae tendinae in one dog with severe mitral valve regurgitation and radiographic evidence of left heart failure. In another dog with a positive heartworm antigen test, echocardiography was used to visualize worms in the right pulmonary artery. In the remaining dog, echocardiography revealed severe pulmonary hypertension causing cor pulmonale. Cardiac ultrasound did not identify an intracardiac cause of pulmonary hypertension, and a heartworm antigen test was negative. Cardiac catheterization was performed and documented a mean pulmonary artery pressure of 105 mm Hg.

Cytopathology of TTW and BAL in four dogs each demonstrated septic purulent inflammation. The samples from two TTW and one BAL also had evidence of erythrophagia by macrophages, indicating previous hemorrhage. No definitive diagnosis of a specific underlying disease could be drawn from the other cytopathological samples. Bronchoscopy showed diffuse erythema (n=5), increased friability (n=4), increased mucus (n=3), purulent debris (n=2), a cobblestone appearance (n=1), and a grass awn foreign body (n=1). One dog had diffuse and severe airway edema with extremely friable tissue during bronchial mucosal biopsy; the procedure was aborted because of severe bleeding. This dog died following bronchoscopy, and although the exact cause of death was undetermined, the dog had chronic bronchitis and a resistant Acinetobacter spp. pneumonia on necropsy examination. None of the other dogs having bronchoscopic examinations showed visible bleeding. Cultures of airway lavage fluid were positive in eight of nine dogs; in the remaining dog, there was evidence of gram-negative rods and septic inflammation on a direct smear. Three dogs with positive cultures and the dog with a negative culture were receiving antibiotics at the time of culture submission. All positive cultures grew single bacterial isolates including Pasteurella canis (n=3), Bacillus spp. (n=2), Escherichia coli (n=1), Pseudomonas spp. (n=1), and Acinetobacter calcoaceticus var. anitratus (n=1). Two additional cases of pneumonia were diagnosed based on histopathological findings without culture submission.

Fine-needle aspiration cytopathology of the pulmonary parenchyma was diagnostic for epithelial neoplasia in one dog and was suggestive, but not conclusive, for neoplasia in another dog. Histopathological examination of the lung specimens obtained antemortem revealed primary or metastatic neoplasia (n=5), chronic lobar pneumonia (n=2), chronic bronchitis (n=2), and lung-lobe torsion and traumatic lung injury (n=1 each). Histopathology of lung tissue obtained from necropsy revealed metastatic neoplasia (n=2); severe, diffuse, pulmonary hemorrhage without apparent structural damage (n=2); and bacterial bronchopneumonia (n=2).

A single underlying disease predisposing to hemoptysis was identified in 32 dogs and included bacterial bronchopneumonia (n=7), primary or metastatic neoplasia (n=5), trauma (n=5), ITP (n=4), heartworm disease (n=4), rodenticide poisoning (n=3), lung-lobe torsion (n=1), left-sided congestive heart failure (n=1), pulmonary hypertension (n=1), and foreign-body pneumonia (n=1). Four dogs had more than one underlying disease process that could have caused hemoptysis; this included chronic bronchitis with bacterial bronchopneumonia (n=2), neoplasia and bacterial bronchopneumonia (n=1), and chronic bronchopneumonia and DIC (n=1). Of the dogs diagnosed with neoplasia, three had primary lung tumors (all carcinomas), and three had metastatic pulmonary neoplasia (i.e., malignant melanoma, telangiectatic osteosarcoma, and histiocytic sarcoma [n=1 each]).

Of the 36 dogs in the study, seven dogs were euthanized during the course of their initial hospital visit, and one dog died in the hospital following bronchoscopy. The remaining 28 dogs were considered to have either recovered from their episode of hemoptysis or were felt to be stable and were discharged from the hospital. Of these dogs, five either died or were euthanized because of their underlying disease between 1 and 6 months following discharge; seven dogs recovered completely from their respiratory disease according to their owners, and they died or were euthanized as a result of unrelated causes; eight dogs were alive 3 months to 7 years following discharge (at the time of this study); and the remaining eight dogs were lost to follow-up.

Discussion

Hemoptysis can be caused by pulmonary, cardiovascular, or systemic disease. Possible pathophysiological explanations for the occurrence of hemoptysis include laceration of pulmonary structures; ulceration of the walls of the tracheobronchial tree; erosion of blood vessels due to inflammatory or infiltrative disease; formation and subsequent rupture of aneurysms; increased intravascular pressure causing leakage from capillary beds; absence of normal vascular wall integrity due to loss of hemostatic components, necrosis of infected or malignant tissue, or embolic obstruction to blood flow with resulting extravasation of blood.^8–10 Ultimately, plasma and red blood cells are leaked into alveoli and distal airways where they can be cleared by the mucocilliary apparatus or be coughed up and expectorated. Because of the varied causes of hemoptysis, a thorough clinical workup is often required to determine the underlying cause. Knowledge of the underlying etiology is important in directing the appropriate treatment and in providing prognostic information.

More than 75 causes of hemoptysis have been identified in humans.¹¹ The underlying etiology of hemoptysis and its frequency varies depending on the study, and includes bronchitis (18% to 37%), bronchogenic carcinoma (19% to 29%), bronchiectasis (1% to 20%), bacterial pneumonia (4.5% to 16%), metastatic neoplasia (1% to 9%), tuberculosis (1.4% to 7%), heart failure (1% to 4%), pulmonary contusions/trauma (1% to 4%), hemorrhagic diathesis (1% to 4%), fungal infection (1% to 3%), pulmonary abscess (1% to 2%), pulmonary thromboembolism (<1% to 2%), systemic hypertension (≤1%), pulmonary hypertension (≤1%), and idiopathic (3% to 24.9%).⁷¹²¹³ All of the diseases found in the current study have been reported in the human literature as causes for hemoptysis, including heartworm disease, although the pathogenesis of this infection differs in humans.¹⁴

Bacterial pneumonia was the most common cause of hemoptysis in this study, accounting for 19.4% of the cases. Hemoptysis has been associated with bacterial pneumonia in dogs, but one study reported its occurrence in only 2% of cases.³ Although the frequency of hemoptysis in dogs with bacterial pneumonia is uncommon, it was a primary cause or contributing factor in >30% of dogs in this study; therefore, steps to obtain respiratory tract material for culture, or the use of empirical antibiotics, or both, should be strongly considered. The bacterial organisms identified in this study were consistent with those previously reported in canine pneumonias.³¹⁵ The pathogenesis of hemoptysis in dogs and humans may be similar, in that some bacterial organisms can cause inflammation, necrosis, and vascular leakage into surrounding air spaces. The clinical significance of the particular species of bacteria seen and the lack of others, such as Staphylococcus spp. or Streptococcus spp., in the population of this study is unknown but may reflect the type of damage that different bacteria can cause.

Primary or metastatic neoplasia was the second most common single cause of hemoptysis, comprising 13.9% of cases. There were equal numbers of dogs with primary and metastatic neoplasia in this study. Dogs with neoplasia in this study tended to be older animals, consistent with the mean age of animals reported with primary pulmonary neoplasia.² Primary lung tumors can cause hemoptysis as they arise from bronchi and either become ulcerated, erode major vessels, or become necrotic and cavitary. Primary pulmonary neoplasia is a well-documented cause of hemoptysis in dogs and has been reported in up to 9% of cases.² In comparison, little specific information exists in the veterinary literature regarding the incidence of hemoptysis in dogs that results from metastatic neoplasia in the thoracic cavity. Metastatic nodules are often located more peripherally if as a result of hematogenous spread, or they are near the hilus if spread is via the lymphatic system. Endobronchial invasion is one mechanism by which metastatic neoplasia can lead to hemoptysis.¹⁶ In humans, hemoptysis has been associated with tracheal invasion by neoplasia of other structures within the thoracic cavity, such as the esophagus.¹⁷

Hemoptysis has been previously reported with intrathoracic trauma in the dog.⁴ In this study, 13.9% of dogs were diagnosed with trauma as the cause of hemoptysis, although the exact nature of the pulmonary bleeding was not determined (i.e., contusions versus lung-lobe laceration). Pulmonary contusions have been found to be the most common injury in dogs with thoracic trauma, although the incidence of hemoptysis was not specifically reported.¹⁸ Hemoptysis following trauma is not a localizing sign. Chest trauma can result in fractures of trachea or bronchi, laceration of a lung lobe due to thoracic wall puncture, or rupture of alveoli secondary to massive pressure changes. The authors’ hospital is generally not considered a major trauma center, which may have falsely lowered the number of dogs with hemoptysis secondary to trauma represented in this study.

Immune-mediated thrombocytopenia caused hemoptysis in 11.1% of cases. Each of the dogs with ITP had platelet counts below 25 to 30 × 10³/μL, values that are considered the cutoff for observation of spontaneous hemorrhage.¹⁹ All of these dogs had evidence of bleeding at other sites, including skin or mucous membranes, bladder, and gastrointestinal tract. Previous studies of thrombocytopenia in dogs have not specifically reported hemoptysis as an associated clinical sign, although thrombocytopenia and thrombocytopathies have been recognized causes of both pulmonary hemorrhage and hemoptysis in humans.²⁰²¹ The pulmonary parenchyma has a diffuse capillary network with a large surface area, which could allow for leakage of blood into alveoli if defects in primary hemostasis exist. Radiographically, hemorrhage secondary to thrombocytopenia in dogs appears commonly as a diffuse alveolar pattern.²² Although often not radiographically apparent, hemorrhage can also come from the mucosal surfaces of the airways. In humans, DIC can cause an acute onset of dyspnea due to hemoptysis during the hypocoagulable stage.²³ In this study, spontaneous hemorrhage into the pulmonary parenchyma in a dog with DIC was documented on necropsy.

Hemoptysis resulted from heartworm disease in 11.1% of dogs. Heartworm disease is well documented as a cause of hemoptysis, especially during treatment when dead worms lead to pulmonary thromboemboli (PTE).²⁴ It is not clear whether hemoptysis seen in dogs with heartworm disease results from embolism formation, secondary pulmonary hypertension, or inflammatory changes within the parenchyma.²⁵²⁶ While PTE is known to occur in heartworm disease, other diseases associated with PTE (such as neoplasia, sepsis, immune-mediated hemolytic anemia, hyperadrenocorticism, protein-losing nephropathies, and cardiac disease including dilative cardiomyopathy and chronic mitral valve insufficiency) have not reported hemoptysis as a clinical finding in dogs.²⁷ Pulmonary thromboembolism was not diagnosed in any dog in this study, perhaps reflecting a low frequency of occurrence in dogs, a low percentage of canine patients that have PTE causing hemoptysis, or the poor ability to make a diagnosis of PTE. Further study is required to differentiate hemoptysis due to heartworm disease with and without PTE.

A total of 8.3% of dogs had hemoptysis due to anticoagulant rodenticide toxicity. In a recent study, hemoptysis was observed in 30% of dogs with anticoagulant rodenticide toxicity.²⁸ Other causes of hemoptysis in the dogs of this study included lung-lobe torsion, congestive heart failure, foreign bodies, and pulmonary hypertension. Hemoptysis as a sequela of lung-lobe torsion in the dog has been reported previously.⁵ Evidence of gross hemorrhage in a dog with lung-lobe torsion has been observed during bronchoscopic examination.²⁹ Lung-lobe torsion may compromise venous return in the low-pressure pulmonary veins while maintaining a high-pressure arterial supply. This can cause vascular congestion of the torsed lung lobe with subsequent rupture of capillaries and loss of blood into alveoli and airways. Left-sided congestive heart failure can result in hemoptysis because of increased left ventricular pressure, congestion of pulmonary veins and capillaries, and subsequent rupture of distended capillaries due to intravascular pressure.¹⁰ Alveolar hemorrhage was found in 63% of humans with severe cardiac disease.³⁰ Foreign bodies can cause hemoptysis by creating physical damage to the pulmonary parenchyma or by acting as a nidus of infection. In this study, the dog with a foreign body had concurrent purulent debris in its airways, making clear determination of the cause of hemoptysis difficult. Humans with foreign bodies causing hemoptysis may present with an acute onset or with prolonged, recurrent infections that fail to resolve with conservative antibiotic therapy.³¹ Pulmonary hypertension, defined as mean pulmonary arterial pressure >25 mm Hg, has been associated with hemoptysis in humans.⁷ Pulmonary hypertension in dogs has been associated with intra-alveolar hemorrhage (based on histopathological examination), although hemoptysis was not specifically reported as an observed clinical sign.³²

In humans, chronic bronchitis is the first or second leading cause of hemoptysis.⁷¹³ One explanation for hemoptysis in humans is rupture of bronchial pulmonary anastomoses, which develop in patients with chronic respiratory disease, including chronic bronchitis.³³ These vessels lead to high-pressure blood flow within the pulmonary circulation that may predispose to rupture and hemorrhage. This vascular change is not recognized in dogs with chronic bronchitis, nor has hemoptysis been a reported clinical sign associated with this disease.³⁴ The significance of chronic bronchitis as a concurrent disease is unknown at this time, but it may serve as a predisposing factor to the development of hemoptysis.

History and physical examination may provide important clues as to the etiology involved. Dogs in this study that had a cough without blood for a month or longer prior to the onset of hemoptysis were found to have diseases commonly thought of as chronic in nature. Some of these include neoplasia, heartworm disease, left-sided congestive heart failure, and chronic bronchitis. Additionally, the eight dogs with systemic bleeding disorders had evidence either historically or during physical examination that was consistent with the ultimate diagnosis. This reiterates the importance of complete physical examinations and testing for disease processes outside of the thoracic cavity in dogs with hemoptysis.

The severity of hemoptysis (as reflected by a decrease in both HCT and TP) was not useful in determining the underlying disease process or in providing prognostic information in the dogs of this study. Hemoptysis may cause death by one of two ways: exsanguination due to severe blood volume loss, or hypoxia due to alveolar flooding. Only one dog, which was diagnosed with ITP, required blood transfusions; in the remainder of the dogs, hemoptysis did not cause life-threatening anemia. One dog had severe hemorrhage during bronchial mucosal biopsies that necessitated termination of that diagnostic procedure. The dog subsequently died, although on necropsy examination, the cause of death was undetermined but not believed to be due to alveolar flooding by hemorrhage.

All dogs presenting with hemoptysis had abnormal thoracic radiographic findings with a wide spectrum of radiographic patterns. Based on thoracic radiography alone, pulmonary hemorrhage could not be definitively differentiated from primary pulmonary disease. Pulmonary hemorrhage can have a variable appearance but commonly causes an interstitial pattern as a result of increased vascular permeability or an alveolar pattern from the pooling of blood within alveoli. Hemorrhage may be localized when caused by trauma to one lung lobe, or it may be more diffuse, as is typically seen with disorders of hemostasis. The latter may have radiographic evidence of concurrent, hemorrhagic, pleural and abdominal effusions. Also to be considered is the fact that when it is localized, hemorrhage may obscure discrete neoplastic masses. When it is diffuse, hemorrhage is indistinguishable from pulmonary edema and interstitial pneumonia.²²

Bronchoscopy has become a standard diagnostic test in humans with hemoptysis.³⁵ Bronchoscopy can be used in the bleeding patient and in a patient between bleeding episodes. Bronchoscopy allows visual examination of the upper respiratory tract, trachea, and bronchi. It is useful in the localization and retrieval of foreign bodies and in the identification of infection by cytopathology and culture.³⁶ Cytopathology of BAL fluid may reveal cells with features of malignancy; chronic hemorrhage can be detected by the presence of erythrophagocytosis, hemosiderin-laden macrophages, or both.³⁷

Additional diagnostic tests, such as computed tomography (CT), have been used in humans, particularly in patients with nonlocalizing thoracic radiographs, to help determine an etiology of hemoptysis.³⁸ Computed tomography may identify lesions that are not apparent on survey thoracic radiography. Further studies are indicated to define the role of CT in the workup of hemoptysis in dogs.

Given the number of dogs in this study and the multiple underlying causes of hemoptysis, sample sizes were considered too small to provide statistical evidence regarding outcomes. However, patients with ITP appeared to have the poorest outcomes, as 100% were euthanized during their visit. This may be due to difficulties in management of these patients and rapid progression of their disease. In addition, dogs with neoplasia or with multiple etiologies contributing to their hemoptysis had poor outcomes. Further study is warranted to identify possible prognostic indicators.

This study has several limitations. First, only cases with a definitive diagnosis of an underlying disease causing hemoptysis were included in this study. Second, the spectrum of cases (e.g., infectious, traumatic, etc.) seen at the authors’ referral hospital may differ from those seen at other hospitals. Finally, because dogs may swallow their sputum instead of coughing it out, the true incidence of hemoptysis may be higher than that reported due to the lack of observable sputum in dogs.

Conclusion

Because of the diversity of causes of hemoptysis, a broad approach to diagnostics should be considered. History and physical examination findings will continue to be paramount in reaching a diagnosis in most cases. Efforts to characterize hemoptysis may be of value in prioritizing some tests, especially when empirical treatment is found necessary because of economic restrictions, and in providing prognostic information.