- Email: [email protected]
Value of fiberoptic bronchoscopy and angiography for diagnosis of the bleeding site in hemoptysis
Value of Fiberoptic Bronchoscopy and Angiography for Diagnosis of the Bleeding Site in Hemoptysis J. Saumench, MD, J. Escarrabill, MD, L. Padro, MD, J. MontaiiA, MD, A. Clariana, MD, a n d A. Cant6, MD Departments of Thoracic Surgery, Respiratory Endoscopy, and Angioradiology, Hospital Principes de EspaAa, Hospitalet de Llobregat, Barcelona, Spain
We evaluated 36 patients during active hemoptysis and compared the diagnostic yield of different procedures. Twenty-seven patients (75%)had one or more previous episodes of hemoptysis, most of which were not massive. Fiberoptic bronchoscopy was performed in 25 patients, and the bleeding site was identified in 17 (68%). The likelihood of localizing the bleeding site was significantly higher with early versus delayed fiberoptic bronchoscopy (91%versus 50%). Bronchial arteriography was
0
ne of the most demanding diagnostic problems is identification of the bleeding site for surgical control of massive hemoptysis. Occasionally, only the hemorrhaging lung can be localized, but this is of little value when one performs pulmonary resection. Early diagnostic procedures may increase the likelihood of identifying the bleeding site. We reviewed the management of 36 patients during active hemoptysis, with particular attention to evaluating the efficacy of various diagnostic modalities.
Material and Methods Thirty-six hospitalized patients with hemoptysis of nonneoplastic origin were studied between January 1979 and December 1984. There were 29 men and 7 women with a mean age of 51 years. Fourteen patients had bronchiectasis, 12 exhibited sequelae to tuberculosis, and 4 had active tuberculosis (positive sputum culture). Two other patients had arteriovenous fistula and silicosis, respectively. In the remaining 4 patients, the underlying disease was unknown. Three patients bled more than 600 mL/24 h; 9 patients had losses between 300 and 600 mL/24 h, and 10 had losses between 100 and 300 mL/24 h. The daily quantity of hemoptysis could not be estimated in 14 patients. Distribution of patients according to cause of hemoptysis and quantity of bleeding is presented in Table 1. Twenty-seven patients (75%)had one or more previous episodes of hemoptysis, most of which were not massive. In all cases, a selective bronchial arteriography was performed with 6F to 7F catheters and nonosmotic hydrosoluble iodine contrast agent. ”Positive” arteriographies were those in which the pathological findings were lim-
performed in all 36 patients; positive arteriographies were observed in 20 cases (55.5%). Arteriography was positive in only 2 of the 8 cases in which fiberoptic bronchoscopy did not localize the site of bleeding. Bronchial artery embolization was successfully performed in 10 patients. However, the bleeding recurred within the first week after embolization in 2 patients, and 1 of them died. (Ann Thorac Surg 1989;48:2724)
ited to only one pulmonary lobe. In 10 patients with severe hemoptysis at the time of arteriography in whom the bleeding site was accurately identified because of extravasation of contrast medium and/or the presence of bronchopulmonary shunts (7 patients), and/or hypertrophy and hypervascularization of the bronchial artery (3 patients), embolization was performed using a fibrin sponge. Anterior spinal arteries were never visualized. In 25 patients, fiberoptic bronchoscopy was performed under local anesthesia with the Olympus model BFB3. Bronchoscopic examination was considered positive only if it identified the lobe or segment from which the bleeding originated; it was considered negative when only the side of bleeding was localized or when bilateral or unilateral blood remnants without persistent hemorrhage in either lobe were observed. Early fiberoptic bronchoscopy, ie, within 24 hours after the onset of hemoptysis, was performed in 11 patients. In the remaining 11 patients, fiberoptic bronchoscopy was not performed because of technical difficulties, excessive delay after bleeding had subsided, or patient’s refusal. A complete bronchographic examination was performed in 8 patients. In addition, chest roentgenograms were evaluated in all cases.
Results Fiberoptic Bronchoscopy In 17 patients (68%), the bleeding site was identified by fiberoptic bronchoscopy. In 13 of 17 patients with positive bronchoscopic examination, the bleeding site coincided with the suspicious image on the chest roentgenogram. The likelihood of localizing the bleeding site was significantly higher with early versus delayed fiberoptic bronchoscopy (91% versus 50%) (Table 2).
Accepted for publication March 31, 1989.
Bronchial Arteriography
Address reprint requests to Dr Saumench, Porvenir 43, 08021, Barcelona, Spain.
Bronchial arteriography was considered abnormal when hypervascularization, vascular hypertrophy, aneurys-
0 1989 by The Society of Thoracic Surgeons
0003-4975/89/$3.50
Ann Thorac Surg 1989:48272-4
SAUMENCH ET AL BRONCHOSCOPY AND ANCIOGRAPHY FOR HEMOPTYSIS
Table 1. Distribution of Patients According to Cause and Quantity of Hemoptysis
Table 2. Results of Fiberoptic Bronchoscopy in 25 Patients With Hemoptysis Time of Bronchnscopy
Blood Loss (mL/h)
Cause Active tuberculosis Sequelae to tuberculosis Bronchiectasis Arteriovenous fistula Silicosis Unknown Total
Nonmeasur>600 30WOO <300 able Total 2
...
1
5 2
...
... 1 . . . . . . ... 1 3
9
2 2 4
273
...
4
5
12 14
7
. . . . . . 1 ... 1
2
1 1 4
10
14
36
matic and angiomatous formations, bronchopulmonary fistulas, and bronchial extravasation of the contrast agent were observed. Abnormal arteriographies were observed in 31 of 36 patients (86%),but arteriography was considered “positive” in only 20 patients (64.5%).Seven of the nine early arteriographic examinations performed within 24 hours after the onset of hemoptysis were positive.
Correlation Between Bronchoscopy and Arteriography When positive arteriographies were compared with chest roentgenograms, the affected area was strictly coincidental in 16 cases; thus, in these cases, the bronchial arteriography did not identify a bleeding site different from that suspected by chest roentgenogram. Arteriographic and bronchoscopic findings coincided in 11 of the 17 patients in whom the bleeding site was identified by fiberoptic bronchoscopy. Of the 6 noncoincidental cases, a normal arteriography was demonstrated in 3, bilateral abnormal arteriographic findings were demonstrated in 2, and unilateral abnormal arteriographic findings affecting the whole lung were demonstrated in 1. Arteriography was positive in only 2 of the 8 patients in whom fiberoptic bronchoscopy did not localize the site of bleeding. In the remaining 6 patients, arteriographic findings were pathological (bilateral in 5 and unilateral in 1). Bronchial arteriography performed in the 25 patients who also underwent fiberoptic bronchoscopy proved to be a useful diagnostic tool before operation in only 2 patients (Table 3).
Result
fn = 11)
Early
Delayed (n = 14)
Active bleeding observed Side localized
10 (91%) 1
7 (50%) 7
Comment Clinical signs useful for determination of the severity of hemoptysis are not easy to establish. Quantitative criteria for defining arbitrarily the timing of operation in patients with active hemoptysis vary between 240 and 600 mLl24 h to as much as 1,000 mL/8 days (1-81. In most of the present patients it was difficult to assess bleeding rates before admission to the hospital. Only 3 patients in this series bled more than 600 mL in 24 hours [9]. The patients who underwent operation were those who bled uninterruptedly during the next 24 hours after admission, who had rapid flooding of the tracheobronchial tree after the initial bronchoscopy and washings, or those with both conditions. Bronchial arteriography was particularly valuable in the 10 cases in which embolization procedure was feasible. In these the purposes of bronchial artery embolization were to achieve immediate control of bleeding and to improve clinical conditions for a prospective operation. According to other investigators [2, 3, 61, the safe period of time before revascularization after bronchial artery embolization is approximately 4 weeks, with a 22.2%to 71% rate of recurrent bleeding. In our series, however, the bleeding recurred within the first week after embolization in 2 patients, and 1 of them died. Bronchial arteriography identified the bleeding site in 55.5% of patients, whereas fiberoptic bronchoscopy achieved the same result in 68% of patients. Arteriography was useful in only 2 of the 8 patients with negative bronchoscopic findings. In the remaining 23 patients, arteriography did not lead to a change in the prefiberoptic bronchoscopy diagnoses. Therefore, we consider the main advantage of angiography to be in the context of bronchial artery embolization. Patients should be referred for embolization after clinical and endoscopic evaluation. Bronchographic examination cannot be performed during the acute bleeding episode. In the present study,
Bronchography Bronchographic examination was also performed in 4 of the 17 patients in whom fiberoptic bronchoscopy identified the site of bleeding. In 2 of these 4 patients, bronchoscopic, bronchographic, and arteriographic findings coincided; in the other 2 patients with noncoincidental data, bronchial arteriography was negative. Bronchography was useful in localizing the site of bleeding in 4 of the 8 patients with negative fiberoptic bronchoscopy. In these 4 patients, arteriographic findings were abnormal, although “negative.”
Table 3. Correlation between Fiberoptic Bronchoscopy and Bronchial Arteriography in 25 Patients Results of
Arteriography
Results of Bronchoscopy Positive
Negative
Positive Negative
11 6
2
Total
17
8
6
274
SAUMENCH ET AL BRONCHOSCOPY AND ANGIOGRAPHY FOR HEMOPTYSIS
Ann Thorac Surg
bronchography yielded more diagnostic information than a delayed bronchial arteriography. Therefore, the indication of arteriography in the remission period does not appear to add any additional information to that obtained by bronchographic examination. Endoscopic and arteriographic findings frequently coincided with the "suspicious" areas already detected on chest roentgenograms. Our results are comparable with those reported by Fellows and associates [l]in 6 patients with cystic fibrosis and massive hemoptysis; in most cases, the combination of chest roentgenograms and fiberoptic bronchoscopy is considered potentially diagnostic for bleeding site and treatable surgical approach.
3. Uflacker R, Kaemmerer A, Neves C, Picon I'D. Management of massive hemoptysis by bronchial artery embolization. Radiology 1983;146:627-37. 4. Gong H, Salvatierra C. Clinical efficacy of early and delayed fiberoptic bronchoscopy in patients with hemoptysis. Am Rev Respir Dis 1981;124:221-5. 5. Ivanick MJ, Thonvarth W, Donohue J, Mandell V, Delany D, Jaques PF. Infarction of the left main-stem bronchus: a complication of bronchial artery embolization. AJR 1983; 141:5357. 6. Grenier PH, Cornud F, Lacombe P, Viau F, Nahum H. Bronchial artery occlusion for severe hemoptysis: use of isobutyl-2-cyanoacrylate. AJR 1983;140:467-71. 7. Remy J, Arnaud A, Fardou H, et al. Treatment of hemoptysis by embolization of bronchial arteries. Radiology 1977;122: 33-8. 8. Portes DK, Van MJ, Mack JW. Emergency lobectomy for massive hemoptysis in cystic fibrosis. J Thorac Cardiovasc Surg 1983;86:409-11. 9. Crocco JA, Rooney JJ, Fankushen DS, DiBenedetto RJ, Lyons HA. Massive hemoptysis. Arch Intern Med 1968;121:495-8. 10. Thoms NW, Puro HE, Arbulu A. The significance of hemoptysis in lung abscess. J Thorac Cardiovasc Surg 1970;59: 617-29.
We thank Dr Marta Pulido for copyediting.
References 1. Fellows KF, Sthigol L, Shuster S, et al. Selective bronchial arteriography in patients with cystic fibrosis and massive hemoptysis. Radiology 1975;114:551-6. 2. Milleron M, Carette MF, Marsac J, Helenon Ch. Traitement en urgence des hemoptysis graves pour embolisation des arteres systemiques. Presse Med 1983;12:625-8.
1989;48:2724
We evaluated 36 patients during active hemoptysis and compared the diagnostic yield of different procedures. Twenty-seven patients (75%)had one or more previous episodes of hemoptysis, most of which were not massive. Fiberoptic bronchoscopy was performed in 25 patients, and the bleeding site was identified in 17 (68%). The likelihood of localizing the bleeding site was significantly higher with early versus delayed fiberoptic bronchoscopy (91%versus 50%). Bronchial arteriography was
0
ne of the most demanding diagnostic problems is identification of the bleeding site for surgical control of massive hemoptysis. Occasionally, only the hemorrhaging lung can be localized, but this is of little value when one performs pulmonary resection. Early diagnostic procedures may increase the likelihood of identifying the bleeding site. We reviewed the management of 36 patients during active hemoptysis, with particular attention to evaluating the efficacy of various diagnostic modalities.
Material and Methods Thirty-six hospitalized patients with hemoptysis of nonneoplastic origin were studied between January 1979 and December 1984. There were 29 men and 7 women with a mean age of 51 years. Fourteen patients had bronchiectasis, 12 exhibited sequelae to tuberculosis, and 4 had active tuberculosis (positive sputum culture). Two other patients had arteriovenous fistula and silicosis, respectively. In the remaining 4 patients, the underlying disease was unknown. Three patients bled more than 600 mL/24 h; 9 patients had losses between 300 and 600 mL/24 h, and 10 had losses between 100 and 300 mL/24 h. The daily quantity of hemoptysis could not be estimated in 14 patients. Distribution of patients according to cause of hemoptysis and quantity of bleeding is presented in Table 1. Twenty-seven patients (75%)had one or more previous episodes of hemoptysis, most of which were not massive. In all cases, a selective bronchial arteriography was performed with 6F to 7F catheters and nonosmotic hydrosoluble iodine contrast agent. ”Positive” arteriographies were those in which the pathological findings were lim-
performed in all 36 patients; positive arteriographies were observed in 20 cases (55.5%). Arteriography was positive in only 2 of the 8 cases in which fiberoptic bronchoscopy did not localize the site of bleeding. Bronchial artery embolization was successfully performed in 10 patients. However, the bleeding recurred within the first week after embolization in 2 patients, and 1 of them died. (Ann Thorac Surg 1989;48:2724)
ited to only one pulmonary lobe. In 10 patients with severe hemoptysis at the time of arteriography in whom the bleeding site was accurately identified because of extravasation of contrast medium and/or the presence of bronchopulmonary shunts (7 patients), and/or hypertrophy and hypervascularization of the bronchial artery (3 patients), embolization was performed using a fibrin sponge. Anterior spinal arteries were never visualized. In 25 patients, fiberoptic bronchoscopy was performed under local anesthesia with the Olympus model BFB3. Bronchoscopic examination was considered positive only if it identified the lobe or segment from which the bleeding originated; it was considered negative when only the side of bleeding was localized or when bilateral or unilateral blood remnants without persistent hemorrhage in either lobe were observed. Early fiberoptic bronchoscopy, ie, within 24 hours after the onset of hemoptysis, was performed in 11 patients. In the remaining 11 patients, fiberoptic bronchoscopy was not performed because of technical difficulties, excessive delay after bleeding had subsided, or patient’s refusal. A complete bronchographic examination was performed in 8 patients. In addition, chest roentgenograms were evaluated in all cases.
Results Fiberoptic Bronchoscopy In 17 patients (68%), the bleeding site was identified by fiberoptic bronchoscopy. In 13 of 17 patients with positive bronchoscopic examination, the bleeding site coincided with the suspicious image on the chest roentgenogram. The likelihood of localizing the bleeding site was significantly higher with early versus delayed fiberoptic bronchoscopy (91% versus 50%) (Table 2).
Accepted for publication March 31, 1989.
Bronchial Arteriography
Address reprint requests to Dr Saumench, Porvenir 43, 08021, Barcelona, Spain.
Bronchial arteriography was considered abnormal when hypervascularization, vascular hypertrophy, aneurys-
0 1989 by The Society of Thoracic Surgeons
0003-4975/89/$3.50
Ann Thorac Surg 1989:48272-4
SAUMENCH ET AL BRONCHOSCOPY AND ANCIOGRAPHY FOR HEMOPTYSIS
Table 1. Distribution of Patients According to Cause and Quantity of Hemoptysis
Table 2. Results of Fiberoptic Bronchoscopy in 25 Patients With Hemoptysis Time of Bronchnscopy
Blood Loss (mL/h)
Cause Active tuberculosis Sequelae to tuberculosis Bronchiectasis Arteriovenous fistula Silicosis Unknown Total
Nonmeasur>600 30WOO <300 able Total 2
...
1
5 2
...
... 1 . . . . . . ... 1 3
9
2 2 4
273
...
4
5
12 14
7
. . . . . . 1 ... 1
2
1 1 4
10
14
36
matic and angiomatous formations, bronchopulmonary fistulas, and bronchial extravasation of the contrast agent were observed. Abnormal arteriographies were observed in 31 of 36 patients (86%),but arteriography was considered “positive” in only 20 patients (64.5%).Seven of the nine early arteriographic examinations performed within 24 hours after the onset of hemoptysis were positive.
Correlation Between Bronchoscopy and Arteriography When positive arteriographies were compared with chest roentgenograms, the affected area was strictly coincidental in 16 cases; thus, in these cases, the bronchial arteriography did not identify a bleeding site different from that suspected by chest roentgenogram. Arteriographic and bronchoscopic findings coincided in 11 of the 17 patients in whom the bleeding site was identified by fiberoptic bronchoscopy. Of the 6 noncoincidental cases, a normal arteriography was demonstrated in 3, bilateral abnormal arteriographic findings were demonstrated in 2, and unilateral abnormal arteriographic findings affecting the whole lung were demonstrated in 1. Arteriography was positive in only 2 of the 8 patients in whom fiberoptic bronchoscopy did not localize the site of bleeding. In the remaining 6 patients, arteriographic findings were pathological (bilateral in 5 and unilateral in 1). Bronchial arteriography performed in the 25 patients who also underwent fiberoptic bronchoscopy proved to be a useful diagnostic tool before operation in only 2 patients (Table 3).
Result
fn = 11)
Early
Delayed (n = 14)
Active bleeding observed Side localized
10 (91%) 1
7 (50%) 7
Comment Clinical signs useful for determination of the severity of hemoptysis are not easy to establish. Quantitative criteria for defining arbitrarily the timing of operation in patients with active hemoptysis vary between 240 and 600 mLl24 h to as much as 1,000 mL/8 days (1-81. In most of the present patients it was difficult to assess bleeding rates before admission to the hospital. Only 3 patients in this series bled more than 600 mL in 24 hours [9]. The patients who underwent operation were those who bled uninterruptedly during the next 24 hours after admission, who had rapid flooding of the tracheobronchial tree after the initial bronchoscopy and washings, or those with both conditions. Bronchial arteriography was particularly valuable in the 10 cases in which embolization procedure was feasible. In these the purposes of bronchial artery embolization were to achieve immediate control of bleeding and to improve clinical conditions for a prospective operation. According to other investigators [2, 3, 61, the safe period of time before revascularization after bronchial artery embolization is approximately 4 weeks, with a 22.2%to 71% rate of recurrent bleeding. In our series, however, the bleeding recurred within the first week after embolization in 2 patients, and 1 of them died. Bronchial arteriography identified the bleeding site in 55.5% of patients, whereas fiberoptic bronchoscopy achieved the same result in 68% of patients. Arteriography was useful in only 2 of the 8 patients with negative bronchoscopic findings. In the remaining 23 patients, arteriography did not lead to a change in the prefiberoptic bronchoscopy diagnoses. Therefore, we consider the main advantage of angiography to be in the context of bronchial artery embolization. Patients should be referred for embolization after clinical and endoscopic evaluation. Bronchographic examination cannot be performed during the acute bleeding episode. In the present study,
Bronchography Bronchographic examination was also performed in 4 of the 17 patients in whom fiberoptic bronchoscopy identified the site of bleeding. In 2 of these 4 patients, bronchoscopic, bronchographic, and arteriographic findings coincided; in the other 2 patients with noncoincidental data, bronchial arteriography was negative. Bronchography was useful in localizing the site of bleeding in 4 of the 8 patients with negative fiberoptic bronchoscopy. In these 4 patients, arteriographic findings were abnormal, although “negative.”
Table 3. Correlation between Fiberoptic Bronchoscopy and Bronchial Arteriography in 25 Patients Results of
Arteriography
Results of Bronchoscopy Positive
Negative
Positive Negative
11 6
2
Total
17
8
6
274
SAUMENCH ET AL BRONCHOSCOPY AND ANGIOGRAPHY FOR HEMOPTYSIS
Ann Thorac Surg
bronchography yielded more diagnostic information than a delayed bronchial arteriography. Therefore, the indication of arteriography in the remission period does not appear to add any additional information to that obtained by bronchographic examination. Endoscopic and arteriographic findings frequently coincided with the "suspicious" areas already detected on chest roentgenograms. Our results are comparable with those reported by Fellows and associates [l]in 6 patients with cystic fibrosis and massive hemoptysis; in most cases, the combination of chest roentgenograms and fiberoptic bronchoscopy is considered potentially diagnostic for bleeding site and treatable surgical approach.
3. Uflacker R, Kaemmerer A, Neves C, Picon I'D. Management of massive hemoptysis by bronchial artery embolization. Radiology 1983;146:627-37. 4. Gong H, Salvatierra C. Clinical efficacy of early and delayed fiberoptic bronchoscopy in patients with hemoptysis. Am Rev Respir Dis 1981;124:221-5. 5. Ivanick MJ, Thonvarth W, Donohue J, Mandell V, Delany D, Jaques PF. Infarction of the left main-stem bronchus: a complication of bronchial artery embolization. AJR 1983; 141:5357. 6. Grenier PH, Cornud F, Lacombe P, Viau F, Nahum H. Bronchial artery occlusion for severe hemoptysis: use of isobutyl-2-cyanoacrylate. AJR 1983;140:467-71. 7. Remy J, Arnaud A, Fardou H, et al. Treatment of hemoptysis by embolization of bronchial arteries. Radiology 1977;122: 33-8. 8. Portes DK, Van MJ, Mack JW. Emergency lobectomy for massive hemoptysis in cystic fibrosis. J Thorac Cardiovasc Surg 1983;86:409-11. 9. Crocco JA, Rooney JJ, Fankushen DS, DiBenedetto RJ, Lyons HA. Massive hemoptysis. Arch Intern Med 1968;121:495-8. 10. Thoms NW, Puro HE, Arbulu A. The significance of hemoptysis in lung abscess. J Thorac Cardiovasc Surg 1970;59: 617-29.
We thank Dr Marta Pulido for copyediting.
References 1. Fellows KF, Sthigol L, Shuster S, et al. Selective bronchial arteriography in patients with cystic fibrosis and massive hemoptysis. Radiology 1975;114:551-6. 2. Milleron M, Carette MF, Marsac J, Helenon Ch. Traitement en urgence des hemoptysis graves pour embolisation des arteres systemiques. Presse Med 1983;12:625-8.
1989;48:2724