Show Summary Details
Page of

Pathophysiology and causes of haemoptysis 

Pathophysiology and causes of haemoptysis
Pathophysiology and causes of haemoptysis

Francesco Blasi

and Paolo Tarsia

Page of

PRINTED FROM OXFORD MEDICINE ONLINE ( © Oxford University Press, 2020. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Medicine Online for personal use (for details see Privacy Policy and Legal Notice).

date: 19 October 2020

Key points

  • Haemoptysis varies in intensity from a minor event to a life-threatening condition.

  • Criteria for the definition of massive/life-threatening haemoptysis should be better defined.

  • The epidemiology of haemoptysis has changed over time, particularly in industrialized countries, where bronchiectasis and lung cancer have surpassed tuberculosis as the most frequent causes.

  • The lungs are furnished with a dual blood supply, the bronchial arteries and the pulmonary arteries. Although the former account for only about 1% of arterial supply to the lung, bronchial arteries are involved in approximately 90% of haemoptysis cases.

  • The mechanisms leading to haemoptysis are being better elucidated, particularly in chronic inflammatory conditions, where release of angiogenic growth factors leads to neovascularization with thin-walled fragile new vessels that are prone to rupture into the airways.


Haemoptysis is defined as the expectoration of blood or blood-streaked sputum from the lower respiratory tract. The term derives from the ancient Greek words haima, meaning blood, and ptysis, meaning spitting. The presence of haemoptysis, even in the case of minor events, is a frightening symptom for the patient. The clinical spectrum may vary from minor blood-stained sputum to major bleeding causing respiratory failure and haemodynamic instability. Underlying causes may vary from benign, self-limiting conditions to severe, potentially lethal diseases.

There is a lack of universal consensus on the quantification and severity of haemoptysis events. Haemoptysis is considered scant when involving <5 mL, mild when <20 mL, and moderate when >20 mL. Massive haemoptysis has been varyingly defined as 100 mL/24 hours to more than 1000 mL/24 hours. Most authors apply the term massive haemoptysis to bleeding >600 mL/24 hours or >25 mL/hour. The term exsanguinating haemoptysis refers to blood loss >1000 mL/24 hours (>150 mL/hour) or >300 mL for a single expectoration event [1]‌. Given the unreliability in both the patient’s and physician’s estimates of expectorated volume, and lack of consensus cut-off volume definition, other authors define haemoptysis as massive in the presence of clinical consequences, such as respiratory failure due to airway obstruction or haemodynamic instability [2]. It has been estimated that volumes of blood in the alveoli above 400 mL are sufficient to significantly alter gas exchange. However, it must be considered than, in many situations, haemoptysis arises in patients with underlying cardiorespiratory disease. These subjects may suffer considerable worsening of gaseous exchange even for smaller quantities of blood. It has been proposed that the term ‘massive’ haemoptysis be substituted with ‘life-threatening.’ Life-threatening haemoptysis may be defined as any haemoptysis that:

  • Is >100 mL in 24 hours.

  • Causes abnormal gas exchange/airway obstruction.

  • Causes haemodynamic instability.

Independent of its definition, massive/life-threatening haemoptysis involves only a minority of clinical events (generally 5–10%), but related mortality may exceed 50% [1,2]. Mortality is generally more the result of airway compromise with asphyxiation, rather than exsanguination.


Haemoptysis may derive from a variety of very different conditions, such as infections, pulmonary diseases, neoplastic conditions, cardiovascular alterations, vasculitis, traumatic events, haematological derangements, and iatrogenic or drug-induced events (see Box 126.1). The relative importance of different causes of haemoptysis has changed over time. For centuries, haemoptysis was considered virtually pathognomonic for pulmonary tuberculosis. The following Hippocratic aphorism: ‘The spitting of pus follows that spitting of blood, consumption follows the spitting of this and death follows consumption’ shows how far rooted in history is the association between haemoptysis and tuberculosis. During the course of the last century, however, effective antimycobacterial treatment and the rise in prevalence of cigarette smoking have changed the epidemiology of haemoptysis. TB continues to be a leading cause of haemoptysis in the developing world, but in industrialized countries bronchial carcinoma and bronchiectasis are more commonly reported [3]‌.

It has been estimated that in bronchogenic carcinoma, haemoptysis presents at some point in the natural history of the disease in up to 20% of patients. Conversely, haemoptysis may be the presenting symptom in only 7% of patients with lung malignancy [4]‌. Haemoptysis is more likely to occur, and may manifest earlier in the disease course, when carcinoma originates in a major bronchus compared with more peripheral sites.

Among mycobacteria, haemoptysis is mainly related to Mycobacterium tuberculosis, with few reports on the involvement of non-tuberculous mycobacteria. Haemoptysis may be the result of active tuberculosis, generally with small and chronic bouts of blood, although massive haemoptysis has also been described in this context. Inactive mycobacterial disease may be associated with bleeding arising from post-tuberculous thick-walled cavities or bronchiectasis. Rarely, peri-bronchial calcified lymph node may erode into or distort an adjacent bronchus. This condition is known as broncholithiasis and may be associated with symptoms such as cough, recurrent episodes of fever and purulent sputum, and sometimes massive haemoptysis [5]‌.

Aspergilloma is a mycotic colonization of a pre-existing lung cavity or cyst. Post-tuberculous cavities or idiopathic pulmonary fibrosis cavities are examples of pre-existing lung disease conditions that may be prone to fungal colonization. Aspergillus fumigatus and Aspergillus niger are the most commonly encountered species. The reported incidence of haemoptysis in patients with aspergilloma ranges from 54 to 87.5%. Most patients will experience mild and recurrent episodes of haemoptysis, although roughly 10% may develop massive events [6]‌. Invasive pulmonary aspergillosis in immunocompromised subjects is also associated with haemoptysis events, although rarely fatal.

Cystic fibrosis, usually in the context of extensive bronchiectasis, has become an increasingly common aetiology of haemoptysis, probably due to the longer survival of affected patients into adulthood. Approximately 40% of patients with cystic fibrosis will develop an episode of haemoptysis during the course of their life, with an average annual incidence of 0.87% [7]‌.

Necrotizing pneumonia, lung abscess, and lung gangrene caused by bacteria, such as Klebsiella pneumoniae, Pseudomonas aeruginosa, Staphylococcus aureus, Streptococcus pneumoniae, other Streptococcus spp. and Actinomyces spp. may be associated with haemoptysis in up to 15% of cases [8]‌.

Bevacizumab is a monoclonal antibody that targets angiogenic factors and has been shown to be effective in the treatment of various forms of cancer. The use of this drug in patients with lung malignancy may be complicated by a high incidence of pulmonary haemorrhage, mainly in patients with squamous cell histology [9]‌.

The event of pulmonary artery rupture during pulmonary artery catheterization (PAC) is relatively rare with an incidence of 0.01–0.47% [10]. However, when it does happen, associated mortality is considerable, being over 70%. Rupture may be due either to the catheter tip being directed into the vessel wall, or to catheter migration to a smaller calibre branch and subsequent rupturing during balloon inflation.

Pulmonary hypertension is reported as the underlying cause of haemoptysis in 0.2–4% of cases [11]. Conversely, in patients with Eisenmenger syndrome, haemoptysis is a more common finding.

In female patients with thoracic endometriosis haemoptysis represents an early clinical manifestation, occurring at an earlier age, whereas pneumothorax tends to manifest in more advanced disease. In over 80% of cases the right hemithorax is involved [12].

Diffuse alveolar haemorrhage is a rare cause of haemoptysis associated with disruption of the alveolar–capillary barrier. It should be suspected in case of haemoptysis, anaemia and bilateral infiltrates on the chest radiograph. The infiltrates are the result of distal inhalation of blood. Diffuse alveolar haemorrhage may be associated with Goodpasture’s syndrome, where antibodies to type IV collagen are deposited along the alveolar and glomerular basement membranes, giving rise to haemoptysis. Diffuse alveolar haemorrhage is also associated with systemic vasculitis, such as Wegener’s granulomatosis or microscopic polyangiitis. Less commonly, diffuse alveolar haemorrhage may be associated with other immunological conditions, such as systemic lupus erythematosus, Henoch–Schönlein purpura, IgA nephropathy, rheumatoid arthritis, Behçet’s syndrome, and cryoglobulinaemia. Infectious processes, such as leptospirosis, malaria, and cytomegalovirus infection may present alveolar haemorrhage. A number of drugs are also associated with diffuse alveolar haemorrhage, such as propylthiouracil, carbimazole, and crack cocaine [13].

In a sizable number of cases, even after extensive diagnostic work-up, a definitive cause for haemoptosyis is not found. These cases are termed as cryptogentic haemoptysis. The incidence of cryptogenetic haemoptysis varies between different reports, with most being between 15 and 30% [14]. Most of these studies have, however, not systematically used CT evaluation on all patients. Wider availability and technical developments in CT imaging will likely result in reduced prevalence of unknown origin cases of haemoptysis in the future.


The lungs are furnished with a dual blood supply, the bronchial arteries and the pulmonary arteries. The former account for only about 1% of arterial supply to the lung and bring nutrients to the lung parenchyma, the bronchi, and vasa vasorum of the pulmonary arteries and veins. The bronchial arteries are a high-pressure circulation system. They have a variable anatomy in terms of origin and branching distribution. They generally originate from the descending thoracic aorta, at the level of the 3rd–8th thoracic vertebral body, more commonly between T5 and T6 (70–83.3% of cases) [15]. The right bronchial artery often arises together with the first aortic intercostal artery to form the intercostobronchial (ICBT) trunk, which usually branches off from the right lateral surface of the descending thoracic aorta. The left bronchial arteries conversely tend to arise from the more anterior surface of the descending thoracic aorta. The four most common bronchial artery branching patterns include type I with one right bronchial artery (rising from the ICBT) and two left bronchial arteries (present in 40.6% of cases), type 2 with one right (rising from the ICBT), and one left bronchial artery (21.3% of cases), type 3 with two right (only one rising from the ICBT), and two left bronchial arteries (20.6%), and type 4 with two right (only one rising from the ICBT), and one left bronchial artery (9.7% of cases) [15]. In approximately 20–30% of cases aberrant bronchial arteries branch off from other systemic arteries, such as the aortic arch, brachiocephalic artery, subclavian artery, internal mammary artery, inferior phrenic artery, or abdominal aorta. In addition, during chronic inflammatory processes collateral blood supply may be recruited from non-bronchial systemic arteries through transpleural vessels. These collateral non-bronchial vessels may derive from ramifications of subclavian, axillary, internal mammary arteries, as well as from subdiaphragmatic arteries. True bronchial arteries (both normal variants and aberrant) can be distinguished from these non-bronchial systemic arteries in that their trajectory into the pulmonary parenchyma parallels the bronchovascular axes. In contrast, non-bronchial systemic collateral vessels do not run parallel to the airways and have a more unpredictable origin from infradiaphragmatic arteries or from the supra-aortic great vessels or their branches, and enter the parenchyma through the inferior pulmonary ligament or through the adherent pleura; their course is not parallel to that of the bronchi.

The pulmonary arteries are a low-pressure circulation system that account for 99% of the arterial supply to the lungs and are responsible for gas exchange. The bronchial and pulmonary systems are in close proximity at the level of the vasa vasorum where they are interconnected by numerous anastomoses. These communications cause a physiological right-to-left shunt that involves approximately 5% of the total cardiac output.

Bleeding in the lungs may originate from bronchial arteries, pulmonary arteries, bronchial capillaries, and alveolar capillaries. Bronchial arteries are the most common site of bleeding causing haemoptysis, being involved in approximately 90% of cases [16]. In only 5% of cases does haemoptysis origin from the pulmonary vessels. In the remaining 5% of cases bleeding arises from ectopic bronchial arteries or other non-bronchial systemic arteries (including the aorta).

Bleeding from the pulmonary artery circulation may be found in disease processes, such as tuberculosis, mycetoma, cavitating lung carcinoma, lymphoma, Behçet disease, pulmonary arteriovenous malformations, and trauma during right heart catheterization. The most common cause of bleeding from the pulmonary circulation is Rasmussens’s aneurysm. In the walls of cavitary lesions, bronchial or pulmonary arteries may be subject to deformation causing pear-shaped dilatations, which are, in truth, pseudo-aneurysmatic lesions, which may be eroded due to chronic inflammatory derangement. These aneurysms are traditionally considered responsible for haemoptysis in TB patients. However, the development of hypervascularized, dilated, tortuous bronchial vessels, often anastomotic with the pulmonary circulation, in the absence of true aneurysms, may be equally involved in these patients.

In certain situations, the thin-walled capillary communications between the high-pressure systemic bronchial arterial system, and the lower pressure pulmonary arterial system can vasodilate and enlarge. Conditions causing reduced pulmonary arterial perfusion, such as chronic thromboembolic disease and vasculitic disorders, in which there is a reduction in pulmonary arterial supply distal to the emboli, can lead to a gradual increase in the bronchial arterial contribution [3]‌, thereby increasing the importance of bronchial-to-pulmonary artery anastomoses in regions of the lung that are deprived of their pulmonary arterial blood flow. Experimental studies have suggested that the increased bronchial arterial blood flow is due to neovascularization [3]. The anastomotic vessels, which are subjected to increased systemic arterial pressure, are often thin-walled, and prone to rupture into the alveoli or bronchial airways, giving rise to haemoptysis. During acute or chronic tracheobronchitic events, inflammation of the mucosa with vascular engorgement, desquamation, atrophy, and erosion, may lead to bleeding. Infective agents, such as Aspergillus spp. may release fungal endotoxins that exert haemolytic activity. Inflammatory processes release angiogenic growth factors, thus promoting neo-angiogenesis and recruitment of collateral supplies from adjacent vessels and increased anastomoses between bronchial and pulmonary arterial systems [17]. Similar mechanisms are in play in haemoptysis caused by lung cancer. Cancer growth requires new vessel formation derived from pro-angiogenic tumour-secreted cytokines becoming dominant over naturally-occurring angiogenesis inhibitors. Most common factors involved are vascular endothelial growth factor (VEGF), angiopoietin 1, fibroblast growth factor, hepatocyte growth factor, transforming growth factors α‎ and β‎, platelet-derived growth factor (PDGF), tumour necrosis factor-α‎, and interleukin-8 [18]. Furthermore, there is evidence that VEGF elevation correlates significantly with the presence of haemoptysis in patients with pulmonary aspergilloma [19]. The hypoxic micro-environment associated with tumour growth causes cellular activation of hypoxia-inducible factor (HIF). HIF up-regulation promotes transcription of pro-angiogenic cytokines [20]. Hypoxaemia-induced angiopoietin 2 release plays an important role in initiating vessel sprouting in concert with VEGF. In contrast with angiopoietin-1, which stabilizes blood vessels, angiopoietin 2 destabilizes blood vessels, thus favouring bleeding. The new vessels associated with chronic inflammatory or tumoural pro-angiogenesis are usually thin-walled and fragile, and thus prone to rupture into the airways, therefore causing haemoptysis.

Other specific mechanisms may be involved in generating haemoptysis, in particular disease conditions. In patients with depressed left ventricular ejection fraction or mitral stenosis elevated pulmonary intravascular pressure may cause rupture of pulmonary veins or capillaries, resulting in blood-stained sputum. In patients with Goodpasture’s syndrome, deposition of antibodies directed against the alveolar–capillary basement membrane disrupts capillary integrity leading to bleeding.


1. Garzon AA, Cerruti M-M, and Golding ME. (1982). Exsanguinating hemoptysis. Journal of Thoracic and Cardiovascular Surgery, 84, 829–33.Find this resource:

2. Dweik RA and Stoller JK. (1999). Role of bronchoscopy in massive hemoptysis. Clinical Chest Medicine, 20, 89–105.Find this resource:

3. Mal H, Rullon I, Mellot F, et al. (1999). Immediate and long-term results of bronchial artery embolization for life-threatening hemoptysis. Chest, 115, 996–1001.Find this resource:

4. Hirshberg, B., Biran, I., Glazer, M., et al. (1997). Hemoptysis: etiology, evaluation, and outcome in a tertiary referral hospital. Chest, 112, 440–4.Find this resource:

5. Reza B and Ziaollah H. (2007). Surgical management of tuberculous broncholithiasis with hemoptysis: experience with 5 operated cases. Annals of Thoracic and Cardiovascular Surgery, 13, 185–90.Find this resource:

6. Soubani AO and Chandrasekar PH. (2002). The clinical spectrum of pulmonary aspergillosis. Chest, 121, 1988–99.Find this resource:

7. Flume PA, Yankaskas JR, Ebeling M, Hulsey T, and Clark LL. (2005). Massive hemoptysis in cystic fibrosis. Chest, 128, 729–38.Find this resource:

8. Reimel BA, Krishnadasen B, Cuschieri J, et al. (2006). Surgical management of acute necrotizing lung infections. Canadian Respiratory Journal, 13, 369–73.Find this resource:

9. Johnson DH, Fehrenbacher L, Novotny WF, et al. (2004). Randomized phase II trial comparing bevacizumab plus carboplatin and paclitaxel with carboplatin and paclitaxel alone in previously untreated locally advanced or metastatic non-small-cell lung cancer. Journal of Clinical Oncology, 22, 2184–91.Find this resource:

10. Bussieres JS. (2007). Iatrogenic pulmonary artery rupture. Current Opinion in Anaesthesiology, 20, 48–52.Find this resource:

11. Reesink HJ, van Delden OM, Kloek JJ, Jansen HM, Reekers JA, and Bresser P. (2007). Embolization for hemoptysis in chronic thromboembolic pulmonary hypertension: report of two cases and a review of the literature. Cardiovascular Interventional Radiology, 301, 136–9.Find this resource:

12. Channabasavaiah AD and Joseph JV. (2010). Thoracic endometriosis: revisiting the association between clinical presentation and thoracic pathology based on thoracoscopic findings in 110 patients. Medicine (Baltimore), 89, 183–8.Find this resource:

13. Wiik A. (2008). Drug-induced vasculitis. Current Opinion in Rheumatology, 20, 35–9.Find this resource:

14. Set PA, Flower CD, Smith IE, Chan AP, Twentyman OP, and Shneerson JM. (1993). Hemoptysis: comparative study of the role of CT and fiberoptic bronchoscopy. Radiology, 189, 677–80.Find this resource:

15. Cauldwell EW, Siekert RG, Lininger RE, et al. (1948). The bronchial arteries: an anatomic study of 150 human cadavers. Surgery, Gynecology and Obstetrics, 86, 395–412.Find this resource:

16. Yoon W, Kim JK, Kim YH, et al. (2002). Bronchial and nonbronchial systemic artery embolization for life-threatening hemoptysis: a comprehensive review. Radiographics, 22, 1395–409.Find this resource:

17. McDonald DM. (2001). Angiogenesis and remodeling of airway vasculature in chronic inflammation. American Journal of Respiratory and Critical Care Medicine, 164, S39–45.Find this resource:

18. Sun S and Schiller JH. (2007). Angiogenesis inhibitors in the treatment of lung cancer. Critical Reviews in Oncology Hematology, 62, 93–104.Find this resource:

19. Inoue K, Matsuyama W, Hashiguchi T, et al. (2001). Expression of vascular endothelial growth factor in pulmonary aspergilloma. Internal Medicine, 40, 1195–9.Find this resource:

20. Pugh CW and Ratcliffe PJ. (2003). Regulation of angiogenesis by hypoxia: role of the HIF system. Nature Medicine, 9, 677–84.Find this resource: