Bronchial artery embolization in patients with hemoptysis including follow-up

Hemoptysis may present as an acute medical emergency as even small amounts of blood (e.g. 150 mL) can lead to asphyxia. Hemoptysis refers to the expectoration of blood from the bronchial tree. Despite the lack of a uniform definition, most authors agree that a blood loss of approx. 200–600 mL during a 24 h period is massive. In recurrent hemoptysis, a blood loss that exceeds 100 mL/d over 3 days in a week can be regarded as massive hemoptysis, as well (1). Massive hemoptysis occurs in approx. 1.5% of patients presenting with hemoptysis. Regardless of these definitions, any hemoptysis that threatens the airways or compromises ventilation is significant. If imminent or potential impact on airways or oxygenation is present or suspected, active intervention is indicated.

Bronchial artery embolization has become a relevant and common treatment option in patients with massive or recurrent hemoptysis (2–9). Hemoptysis may be caused by various underlying conditions including airway diseases such as pulmonary malignancies or bronchiectasis, parenchymal disease such as cystic fibrosis, tuberculosis and vasculitides, and cardiovascular disorders such as congenital heart disease and pulmonary artery hypertension (primary and secondary). Tuberculosis remains the leading cause in developing countries, whereas cystic fibrosis, chronic inflammatory lung diseases and bronchogenic carcinoma are usually the main causes in the Western world (10, 11). Chest trauma, mostly including formation of pulmonary pseudocysts, is a rare cause of hemoptysis (12), and iatrogenic causes, e.g. pulmonary artery rupture due to Swan-Ganz catheters, may occur but are low in incidence as well (13).

Massive hemoptysis may be fatal, caused by asphyxiation rather than blood loss. In these patients, mortality risk in only conservatively treated patients was reported to be between 50% and 85%, death occurring within the first hour most of the time (14). Previously, resection of the affected lung has been considered definitive treatment of massive hemoptysis. However, emergency surgery has a mortality rate of 17–35% and the majority of patients are high-risk surgical candidates due to a combination of acute hypoxemia and limited lung capacity caused by severe diffuse chronic pulmonary disease (15, 16). As mortality rates are high in patients with massive hemoptysis, rapid multidisciplinary assessment and therapeutic intervention is mandatory. Bronchoscopy prior to embolization may be helpful and might allow locoregional therapy, however, overall accuracy of bronchoscopy in evaluating patients with hemoptysis was as low as 0–31% (17).

The purpose of this study was to report our experience with selective bronchial artery embolization in patients with acute or recurrent hemoptysis and to evaluate the safety and effectiveness of this technique including follow-up for a maximum of eight years post initial embolization.

Material and Methods

Patients

Twenty-eight patients underwent embolization therapy for management of hemoptysis at a large tertiary care center and were included in this study. Nine women and 19 men with a mean age of 41.8 years (age range 20–82 years) were treated for hemoptysis between January 1998 and October 2008. This retrospective study was approved by our local ethics committee and informed consent was waived. Patients were assessed by a pulmonary care specialist, a thoracic surgeon, and an interventional radiologist. Patients were included in the study if hemoptysis was (a) recurrent despite conservative management or (b) acute or imminent airway or ventilation compromise was present or clinically suspected. Prior medical management included conservative strategies such as bronchoscopic assessment and locoregional intervention, e.g. using vasoactive drugs or argon plasma coagulation. Vital signs and oxygen saturation were monitored in all patients. Additionally, hemoglobin, coagulation parameters, electrolytes, blood typing and cross-match were determined. Renal and liver function tests were also conducted. Etiologies of hemoptysis were cystic fibrosis (n = 9), lung cancer (n = 6), aspergillosis (n = 2), tuberculosis (n = 2), bronchiectasis (n = 3), chronic obstructive pulmonary disease (n = 2), iatrogenic post bronchoscopy (n = 1), coagulation disorder (n = 2), and idiopathic (n = 1). For follow-up, medical charts were reviewed and primary care physicians were interviewed by telephone.

Interventional technique

Angiography was performed via femoral access in all patients. An initial angiogram was performed using a 5-F diagnostic pigtail catheter and power injector to identify bronchial supplying arteries. A cobra-shape catheter (C-1 or C-2, Cook Medical Inc., Bloomington, IN, USA) or reverse-curve catheter (VS-2 or Simmons, Cook Medical Inc., Bloomington, IN, USA) was used for bronchial artery catheterization. Digital subtraction angiography was performed using Axiom Artis (Siemens Medical Solutions, Erlangen, Germany) or Integris V-5000 (Philips Medical Systems, Hamburg, Germany) standard angiography suites. For selective peripheral bronchial artery angiography and embolization, a coaxial microcatheter system (Renegade; Boston Scientific, Natick, MA, USA) was used. For angiography, non-ionic contrast material (Iopamiro 300; Bracco Diagnostic Inc., Princeton, NJ, USA) was used. A potential source of bronchial bleeding was determined if bronchial artery hypervascularization was present. Mapping of branches potentially supplying the spinal cord was performed and if present, a more selective catheterization was performed. After safe positioning of the catheter tip, embolization with particles including polyvinyl-alcohol particles (Contour; Boston Scientific, Natick, MA, USA), Embospheres (Biosphere Medical, Rockland, MA, USA), or Bead Block (Terumo Inc., Somerset, NJ, USA) sized between 350–900 μm suspended in contrast medium was performed through a microcatheter. A postembolization angiogram was performed to document procedural technical success. The endpoint of embolization was stagnation or stasis of antegrade flow within embolized branches. Clinical success of bronchial artery embolization was determined by cessation of bleeding within 24 hours.

In the case of suspected collaterals from the internal thoracic artery, formerly known as internal mammary artery or recurrent bleeding, selective catheterization of the internal thoracic artery was performed using a 5-F Berenstein catheter (Cook Medical Inc., Bloomington, IN, USA) and selective angiography was performed. In the case of collateral arterial supply, a microcatheter (Renegade; Boston Scientific, Natick, MA, USA) was used to catheterize this vessel close to the lesion and embolize with particles (Contour; Boston Scientific, Natick, MA, USA), Embospheres (Biosphere Medical, Rockland, MA, USA), or Bead Block (Terumo Inc., Somerset, NJ, USA) sized between 350–900 μm suspended in contrast medium.

Statistical analysis

Occurrence of post-interventional bleeding and cumulative survival were evaluated using the Kaplan-Meier life-table method, and survival curves were plotted (16). Statistical analysis was performed using GraphPad statistical software (Prism v. 4.01 and InStat v 3.0; GraphPad Software, San Diego, CA, USA).

Results

Angiography and transcatheter embolization was performed early during assessment of hemoptysis. Embolization was performed to prevent confinements in oxygenation and clinical deterioration. Twenty-six of 28 patients (93%) had recurrent hemoptysis prior to embolization. In two patients with bronchial carcinoma, hemoptysis leading to transcatheter embolization was the first event. Four of 28 patients (14%) required blood transfusions and one patient had to be intubated and mechanically ventilated due to airway obliteration by blood. Clinical success of bronchial artery embolization, determined by cessation of bleeding within 24 hours, was achieved in all patients. Stagnation or stasis of blood flow in feeding vessels was achieved using particles in all patients (Figs. 1 and 2). Hypervascularization was localized as follows: right upper lobe (n = 15), left upper lobe (n = 12), right middle lobe (n = 8), lingula (n = 1), right inferior lobe (n = 4) and left inferior lobe (n = 7). Bronchial arteries included 10 intercostobronchial trunks, 10 isolated bronchial arteries, and six common trunks from the right and left bronchial arteries. Additional collateral arterial supply from the internal mammary artery was found in nine patients and embolization was performed. Selective embolization was performed in a total of 35 interventions.

OPEN IN VIEWER

Fig. 1

43-year old patient with hepatocellular carcinoma metastasis to the right lung and recurrent hemoptysis. (a) Pigtail catheter aortic angiogram demonstrates the right bronchial artery originating from the descending aorta at the level of the carina (arrow); (b) An angiogram using a 2.7-F microcatheter positioned peripherally in the right bronchial artery demonstrates pathologic vessels in the middle lobe due to metastasis (arrows); (c) Angiographic image post embolization with particles demonstrates vessel pruning (arrow)

OPEN IN VIEWER

Fig. 2

A 30-year-old woman with severe pulmonary manifestation of cystic fibrosis. (a) Angiographic image demonstrates arterial supply (arrow) originating from the left internal thoracic artery. A 5-F single-curved catheter was used to select the left internal thoracic artery; (b) A 2.7-F microcatheter was used to catheterize the left internal thoracic artery. Angiogram demonstrates a blush of contrast located in the left upper lobe (arrow). Embolization was performed using particles sized 300–500 μm and 500–700 μm.

The follow-up period ranged from one month to eight years (mean 23 months). Twenty-four patients had no evidence of recurrent bleeding. In the remaining four patients, recurrent bleeding was found in two men and two women, all with cystic fibrosis. One patient received embolization with Embospheres (Biosphere Medical, Rockland, MA, USA), three patients received Bead Block (Terumo Inc., Somerset, NJ, USA) particles. Two patients had recurrent bleeding from the same vascular bed and two patients from a different vessel. The first episode of re-bleeding occurred at 1, 16, 24 or 48 months post initial embolization. Multiple episodes (×4) of recurrent bleeding occurred in one of these patients. This was a 20-year old woman with cystic fibrosis who had her first embolization of the middle lobar artery in 2004. Subsequently, she had re-interventions in 2006 for the right upper lobar artery (×2) and middle lobar artery (×1).

Rates of patients without re-bleeding were 96% (standard error of mean 3.5) at one month, 91% (standard error of mean 6.0) at 16 months, 86% (standard error of mean 7.9) at 19 months, 71% (14.6 standard error of mean) at eight years (Fig. 3).

OPEN IN VIEWER

Fig. 3

Kaplan-Meier plot demonstrating cumulative recurrence-free rates post bronchial artery embolization. Cumulative rate of recurrent hemoptysis was 28.6% at 8 years

Seven patients died between 1 and 17 months (mean 6.6 months) post initial intervention. One patient with bronchial carcinoma died on post-interventional day 1 due to his underlying disease; there was no hemoptysis present at the time of death. All other patients also died due to underlying disease. Patient survival rates were 89% (standard error of mean 5.8) at one month, 86% (standard error of mean 6.6) at two months, 78% (standard error of mean 8) at five months, and 74% (standard error of mean 8.7) at eight years (Fig. 4).

OPEN IN VIEWER

Fig. 4

Kaplan-Meier plot demonstrating cumulative survival following bronchial artery embolization. Cumulative survival rate was 74% at 8 years

Discussion

The present study was performed to evaluate effectiveness and safety of bronchial artery embolization in patients with hemoptysis including follow-up post initial intervention. Bronchial artery embolization was effective in most of our patients with hemoptysis. The post-interventional survival rate was 74% from 19 months until eight years of follow-up. In our study, bronchial artery embolization was demonstrated to be an effective non-invasive treatment of hemoptysis.

In patients with acute and chronic lung disease, smaller vessels originating from the pulmonary arteries may already have a constricted lumen or occlusion due to hypoxic vasoconstriction, thrombosis, or inflammatory changes. This induces compensatory hypertrophy of the small vessels originating from the bronchial arteries which can rupture following erosion of the vessel wall caused by inflammation, tumor growth, or regional increase of blood pressure (8, 9, 18–20).

Recently, Poyanli et al. found that in up to 90% of patients with hemoptysis, bronchial arteries are the source of bleeding (21). However, other authors report that 41–88% of hemoptysis cases stem from non-bronchial systemic arteries (9, 20, 22), underlining the high variability of bleeding sources.

Bronchial artery anatomy is highly variable in terms of origin, branching pattern, and course. In most patients, bronchial arteries originate from the descending thoracic aorta at the level T5/6 vertebrae. Bronchial arteries originating outside this area are considered ectopic; this ranges from 8.3–35% (17). In the majority of cases, bronchi are supplied by two separate bronchial arteries, one on the left and another one on the right side as intercostobronchial trunk. The second most common variant is two bronchial arteries on both sides including an intercostobronchial trunk on the right (22). Possible anatomic variants include origin of the mammary artery, thyrocervical trunk, subclavian artery, costocervical trunk, brachiocephalic artery, pericardiophrenic artery, inferior phrenic artery, or abdominal aorta (17). The most common abnormal origins are off the aortic arch, internal mammary artery, or subclavian artery. The most common anatomical variant regarding the number of bronchial arteries is: one right as intercostobronchial trunk and two left.

There are different causes of recurrent hemoptysis. Immediate recurrent hemoptysis may occur due to multiple feeding arteries that went untreated, whereas later recurrence may be a result of re-canalization or collateralization of the feeding artery or other source of bleeding. Recurrent bleeding in this study was found in four patients, all with cystic fibrosis. Two patients had recurrent bleeding from the same vascular bed and two patients from a different vessel. The first episode of re-bleeding occurred at 1, 16, 24 and 48 months post initial embolization, respectively. Multiple episodes (×4) of recurrent bleeding occurred in one of these patients. The recurrence rate of 18% in our study is in the lower range of recurrence rates reported in other studies (14–42%) (3, 4, 6, 8). Cystic fibrosis patients tend to have hemoptysis during the course of the disease. In a retrospective series, almost 10% of patients had hemoptysis, 20% of those more than once a month (22). Flume et al. estimate that 4.1% of CF patients will experience massive hemoptysis at least once in their life (22). Despite bronchial artery embolization, there is a high recurrence rate of ≥50% within 4 months (23).

The complication rate reported for bronchial artery embolization is generally low. In our series, there was no major complication including non-target embolization or death. Coaxial microcatheters enable superselective catheterization of small-diameter vessels without damaging the proximal portion of the bronchial artery (8, 19).

Embolic materials used for bronchial artery embolization are particles, as used in the present study. The possibility of retrograde embolization due to reflux or non-target antegrade embolization may cause lung infarction, esophageal necrosis, aortic wall necrosis, or spinal cord ischemia. Therefore, recommendation for particle size is 350 µm or larger mainly to avoid neurological complications (9).

There are several limitations of this study. First, the study is retrospective in nature and therefore not controlled for selection bias, detection of events, and data collection. Second, although a total of 35 vessels were embolized, complications that occur at a low rate could have been missed. Third, although 28 patients with hemoptysis were included in this study, indications and underlying diseases for bronchial artery embolization varied in our patient cohort. Finally, a control group is lacking.

In conclusion, bronchial artery embolization therapy was effective in patients with hemoptysis. It is a useful therapy to control hemoptysis and may help to avoid surgery in patients who are poor surgical candidates. Should hemoptysis recur in these patients, repeated embolization can be performed.

Conflict of interest: None.