Chapter 41. Massive Hemoptysis

• Assure adequate oxygenation.
• Look for and correct coagulation abnormalities.
• For localized bleeding, bronchoscopy and/or computed tomography scanning can generally establish the region of the lung from which the bleeding is coming.
• Kidney or lung biopsy may be needed to establish the cause of diffuse bleeding.
• Treatment depends on cause, but bronchial arterial embolization should generally be considered prior to surgery.

Hemoptysis is defined as coughing of blood, and can be caused by numerous diseases and conditions (Table 41-1). Most series indicate that bronchiectasis is the most common cause, but the prevalence of any specific causal entity is a function of the time frame studied (with tuberculosis being more common in older series), and the nature of the population encountered (e.g., bronchiectasis in patients with cystic fibrosis, cancer in older smokers).

Most patients with hemoptysis do not require intensive care. Those who do generally have either such rapid rates of bleeding that they are hemodynamically unstable, or they have life-threatening hypoxemia as a result of diffuse parenchymal hemorrhage or extensive aspiration of blood that originates from a localized source.

“Massive” hemoptysis has been variably defined as production of more than 300 to 600 mL of blood in 12 to 24 hours, depending on the study. Although this connotation is perhaps important from a descriptive standpoint, the distinction has little clinical utility, as it is difficult for patients to accurately quantify the volume of blood they are producing, and the volume of hemoptysis may vary considerably from hour to hour (or even from minute to minute). In addition, some patients with diffuse pulmonary hemorrhage can present with life-threatening hypoxemia and diffuse parenchymal infiltrates, meeting all the criteria for the acute respiratory distress syndrome (ARDS), yet have no hemoptysis. Although any of the conditions listed in Table 41-1 may cause hemoptysis, those more commonly associated with massive hemoptysis are bronchiectasis, mycetoma, tuberculosis, bronchogenic carcinoma, lung abscess, and vascular-bronchial fistulas. A recent study of 29 patients from Singapore found that 62% had bronchiectasis, 13% had mycetomas, 10% had tuberculosis, 6% had tumors, and 3% had a pulmonary hemorrhage syndrome.1

Older reviews of hemoptysis comment that bleeding from the nose, pharynx, or even the gastrointestinal system may present with what was thought to be hemoptysis. The point has been made that blood coming from the lungs is usually bright red and has an alkaline pH, whereas that from the stomach will be dark and acidic. Clinically, the distinction is only rarely difficult.

Attention should initially be directed at assuring adequate ventilation and perfusion as with any patient requiring intensive care. The most common cause of death in patients with hemoptysis is asphyxia from aspirated blood. Clearance of blood from the airway may be facilitated by utilizing the Trendelenburg position. If the site of bleeding is known, right or left lateral decubitus positioning may serve to protect the opposite lung.

Some patients (particularly those with diffuse parenchymal hemorrhage) may require intubation and mechanical ventilation. While intubation generally facilitates removal of blood from the lower airway, the tube may become obstructed by coagulated blood with a resultant limitation in ventilation. In extreme circumstances the mainstem bronchus of the nonbleeding lung can be intubated in an attempt to protect the other lung. Alternatively, many reports describe placing occluding catheters in the mainstem bronchus of the bleeding lung. In the setting of extensive bleeding, however, intubation and/or precise placement of endotracheal tubes may be extremely difficult because of the inability to visualize the airway.

Small doses of codeine or morphine may be used to blunt the cough reflex and perhaps slow the rate of bleeding, but agents that depress the sensorium must be used cautiously as they may also increase the risk of aspiration.

Bleeding disorders should be sought; in the appropriate setting the patient should be screened for conditions associated with pulmonary hemorrhage syndromes, and attempts made to determine the site of bleeding. The nose and mouth should be carefully inspected to exclude an upper airway source of bleeding. Rhinoscopy and/or laryngoscopy may at times be useful. A history consistent with rheumatic fever might lead to the suspicion of mitral stenosis, which is readily diagnosed by echocardiography.

Coagulation screening should include a platelet count, a prothrombin time, and a partial thromboplastin time. If patients have had substantial bleeding for a prolonged period of time, a fibrinogen level should also be determined.

Blood should be sent for urea nitrogen and creatinine, a urinalysis should be obtained, and if clinical indications point to diffuse parenchymal hemorrhage, serology should be sent to screen for connective tissue disease and vasculitis (e.g., antinuclear antibodies, rheumatoid factor, complement levels, cryoglobulins, the anti–glomerular basement antibody causing Goodpasture syndrome, anti-phospholipid antibodies, and the antinuclear cytoplasmic antibody that is seen in Wegener granulomatosis and pauci-immune vasculitis) (see Table 41-1).

Determining the Site of Bleeding

The chest x-ray will identify the region of bleeding in approximately 60% of patients. Fiberoptic or rigid bronchoscopy can be used to determine the site of bleeding, as can CT imaging. In most instances the site may be investigated electively after the patient stabilizes and the rate of bleeding slows. However in selected patients (e.g., trauma), immediate surgical intervention may be needed. In others, especially those with rapid rates of ongoing bleeding, the choice of diagnostic evaluation may be more difficult. A recent study of 29 consecutive patients admitted to intensive care with more than 300 mL/d of hemoptysis found that fiberoptic bronchoscopy localized the bleeding in 90% of cases.1

Fiberoptic bronchoscopy may demonstrate a specific endobronchial lesion, or may at least indicate a specific lung lobe or segment from which the bleeding is coming. However, with rapid rates of bleeding, visualization through a fiberoptic scope is likely to be suboptimal because of an inability to clear blood from the airway. Rigid bronchoscopy allows for better suctioning, but is more invasive and only the central airways can be inspected. Nonetheless, when the rate of bleeding precludes accurate assessment of the airway anatomy by fiberoptic bronchoscopy, there should be no delay in moving to rigid bronchoscopy.

A retrospective study of 80 patients with “large or massive” hemoptysis compared the utility of chest x-rays, fiberoptic bronchoscopy, and chest computed tomography (CT).2 The results indicated that chest CT was superior in determining the cause of bleeding, and suggested that it should become the first-line procedure (Table 41-2). A recent study of 40 consecutive patients assessed the accuracy of CT angiography for determining whether patients with massive hemoptysis (i.e., 300 to 600 mL/d) had a systemic artery as the source of the bleeding.3 Using conventional angiography as the gold standard, CT angiography had an 80% sensitivity, an 84% specificity, and an 84% positive predictive value. More than 3 mm of pleural thickening in a region located adjacent to a parenchymal abnormality, and contrast-enhancing vascular structures in the extrapleural fat correlated with the presence of a nonbronchial systemic arterial supply.3

Patients with diffuse alveolar hemorrhage more commonly have diffuse rather than localized infiltrates on their chest x-rays or CT scans, but diffuse infiltrates may also result from localized bleeding if the blood is extensively aspirated. The diagnosis of diffuse alveolar hemorrhage is suggested by finding blood in bronchoalveolar lavage fluid.

General Measures

The platelet count should be maintained above 50,000 mm3 in an actively bleeding patient. Immune platelet destruction may be treated with intravenous gamma globulin prior to platelet transfusion. The platelet dysfunction associated with uremia may be treated with dialysis or by administration of cryoprecipitate and desmopressin (DDAVP). DDAVP has also recently been shown to be remarkably effective in stopping the massive hemoptysis associated with leptospirosis,4 and may, along with cryoprecipitate, be useful in treating patients with von Willebrand disease and hemophilia A.

The prothrombin time and partial thromboplastin time should be corrected to near normal. Deficiencies are treated with vitamin K, or for a more rapid effect, with large volumes of fresh frozen plasma.

Specific Interventions

A number of treatment modalities may be used to treat life-threatening hemoptysis. These include endobronchial ablation of specific lesions, bronchial arterial embolization, external beam irradiation, and surgical resection.

Endobronchial ablation is generally reserved for tumors affecting the larger airways as a form of palliative therapy for patients with unresectable disease. External beam irradiation has been used successfully in a few patients with massive hemoptysis resulting from mycetomas5 and can also be used for patients with unresectable neoplasms if the rate of bleeding is sufficiently slow that the course of treatment can be completed.

Bronchial arterial embolization is the treatment of choice for most patients with life-threatening hemoptysis resulting from a localized parenchymal lesion. Recent studies indicate that the hemoptysis is controlled in as many as 85% of patients up to 1 month following the procedure, although recurrence is seen in approximately 15% of patients within the first 30 days.6

Surgical versus Medical Management

When hemoptysis occurs in the setting of diffuse disease (e.g., cystic fibrosis, extensive pulmonary tuberculosis, or other forms of bronchiectasis), surgery is precluded. For patients with localized disease the decision about when to operate continues to be based on clinical judgement and individual patient assessment, as there are no randomized trials comparing the outcomes achieved with surgery versus repeated embolization. The major indications for surgery are: 1. recurrence of bleeding after bronchial arterial embolization, 2. inability to perform the embolization because of anatomic problems, and 3. multiple bleeding vessels seen on angiography.7 Accordingly, with the possible exception of endobronchial carcinomas that require ablation, bronchial arterial embolization should generally be the initial treatment of choice for patients with localized disease who can be sufficiently stabilized to attempt the procedure.

Diffuse Alveolar Hemorrhage

Diffuse alveolar hemorrhage (DAH) occurs in conjunction with, and may be the initial manifestation of, numerous diseases, conditions, and effects of medications (see Table 41-1). Up to one-third of patients with DAH will not have hemoptysis, but many will have other systemic manifestations of a systemic disease (e.g., rash, myalgias, arthralgias, or conjunctivitis). Chest x-rays generally show diffuse infiltrates, but localized disease may also be seen. DAH occurs in up to 10% of patients with Wegener granulomatosis, and up to 33% of patients with microscopic polyangiitis (all of whom will have evidence of glomerulonephritis). DAH is also seen in approximately 5% of patients with systemic lupus erythematosus, and rarely in conjunction with polymyositis, rheumatoid arthritis, and mixed connective tissue disease. Most patients with Goodpasture syndrome (which is caused by an antibody to the type 4 collagen found in the basement membranes of alveolar walls and glomeruli) present with DAH and glomerulonephritis, but up to 10% may have only DAH.

Diagnosis

Establishing a diagnosis of DAH is not difficult when patients present with hemoptysis, diffuse pulmonary infiltrates, a falling hematocrit, and other manifestations of systemic disease. However, in those without hemoptysis or diffuse infiltrates the diagnosis becomes much more difficult. Bronchoalveolar lavage showing progressively more blood with serial aspirations may be considered diagnostic, but DAH must be suspected before this procedure is considered. When DAH is diagnosed, a kidney biopsy or an open or thoracoscopic lung biopsy is frequently recommended to allow differentiation among the various types of vasculitis, and to definitively either diagnose or exclude Goodpasture syndrome, as patients with vasculitis are treated with high doses of systemic corticosteroids and with cyclophosphamide, whereas those with Goodpasture should receive plasmapheresis.

Unfortunately, despite treatment, over half of patients with DAH resulting from a systemic vasculitis or a collagen vascular disease require mechanical ventilation, and the mortality ranges from 25% in patients with Wegener granulomatosis to 50% in patients with lupus. The 5-year survival rate for these conditions is also reduced.