- Email: [email protected]
The Incidence, Etiology, and Prevention of Postresectional Bronchopleural Fistula
The Incidence, Etiology, and Prevention of Postresectional Bronchopleural Fistula Robert James Cerfolio Bronchopleural fistula (BPF) is a life-threatening complication after pulmonary resection. The incidence varies from 4.5% to 20% after pneumonectomy and is only 0.5% after lobectomy. Certain patient characteristics increase this incidence. These include preoperative radiation to the chest, destroyed or infected lung from inflammatory disease, immunocompromised host, and insulin-dependent diabetes. Certain surgical techniques also increase the incidence. These include pneumonectomy, right-sided pneumonectomy, a long bronchial stump, residual cancer at the bronchial margin, devascularization of the bronchial stump, prolonged ventilation, or reintubation after resection and surgical inexperience. The best treatment of a BPF is prevention. Prevention centers around meticulous surgical technique and the liberal use of prophylactic, pedicled muscle flaps for the patient at increased risk. Survival of BPF depends on a high index of suspicion, early diagnosis, and aggressive surgical intervention. Copyright © 2001 by W.B. Saunders Company Key words: Bronchopleural fistula, incidence, etiology, prevention.
O
ne of the most dreaded complications after pneumonectomy is a bronchopleural fistula (BPF). Although there is no established "first reported case" of a BPF, surgeons have dealt with this devastating complication since the advent of elective pulmonary resection. BPF compromises the airway, infects the pleural space, and spills contaminated fluid into the only remaining lung, causing life-threatening pneumonia. Despite the advent of new heavy wire stapling devises, longer staples, and muscle flaps used to buttress bronchial closures, BPFs continue to occur. It continues to be a major cause of morbidity and mortality. This is especially true if early diagnosis and careful management are not preformed. Like any postoperative complication, prevention is the key to management. In this article we focus on the incidence, etiology, and prevention of BPF that occur after pulmonary resection.
From the Department of Surgery, Division ofCardiothoraci[ Surgery, University of Alabama at Birmingham, Birmingham, AL. Address reprint requests to Robert James Cerfolio, MD, Associate Professor of Surgery, Department of Surge~y, Division ofCardiothoracic Surgery, Universi~y of Alabama at Birmingham. 1900 University Blvd., THT Room 712, Birmingham. AL 35291. COjJyright © 2001 by WB. Saunders Compan)· 1043-0679/01/1301-0001$35.00/0 doi: 1 0.1 053/stcs.2001.22493
Definition BPF is defined as a communication between a lobar or segmental pulmonary bronchi and the pleural space. Despite this definition, the term BPF continues to be twisted and misused by many physicians. Persistent air leaks that originate from the pulmonary parenchyma are NOT bronchopleural fistulas. We believe these are best termed alveolar-pleural fistulas as described by us in 1998. 1 This distinction is not just one of semantics. The treatment of these two types of fistula is quite different. A BPF almost always requires some sort of reoperation whereas alveolar-pleural fistulas almost never do. The latter are best treated with patience and a Heimlich valve.
Incidence of Bronchopleural Fistula The incidence of a BPF after lobectomy is quite low. In 1992, Asamura et aF reported an incidence of only 1.2% for 1,586 patients who underwent lobectomy for lung cancer. However, the incidence is significantly higher after pneumonectomy. The exact reason for this is unknown. There are several explanations. The most likely explanation is that the remaining lung after lobectomy helps cover and indirectly buttress the closed bronchi. The incidence of BPF after pneumonectomy depends on whether it is a right or left and on the
Seminars in Thoracic and Cardiovascular Surgery, Vol 13, No 1 (january), 2001: pp 3-7
3
4
Robert James Cerfolio
indication for surgery. Asamurua et aF reported an incidence for BPF after pneumonectomy for lung cancer of 4.5%. However, if one stratifies his data, the incidence of BPF was 8.6% for the 162 patients that had a right pneumonectomy and 2.3% for the 302 patients that had a left pneumonectomy. The incidence increases even further when pneumonectomy is performed for benign disease (ie, inflammatory disease and/or destroyed lung). In 1991, Pomerantz et aP reported on patients who had pneumonectomy for resistant Mycobacterium infections. In that study, there were 40 pneumonectomies performed and 8 patients had a BPF (incidence 20%) despite the use of muscle flaps in many. Four of these 8 patients died. Other large reports have found similar results. In 1991, Vester et a1 4 reported an incidence for BPF (4.5%) on 506 patients who underwent pneumonectomy for cancer. He found that the incidence was 4.1 % when the bronchus was stapled and 17.6% when sutured. Careful review, however, shows that he prefers stapling and only sutured when needed. Only 17 patients had sutured closures and 489 were stapled. Of the 23 patients in that report who had a BPF, 17 patients had a right pneumonectomy and 6 had a left. In that same article, Vester 4 also reported a 1.0% BPF rate for 965 patients that had a lobectomy or bilobectomy. His incidence for a BPF after 650 segmentectomies of 0.3% was even less. This lends credence to the theory that the surrounding lung helps cover the stump and avoids a postresectional BPF. Finally, in 1994 AI-Kattan et aP reported an incidence of 1.3% for BPF after 530 consecutive pneumonectomies. The bronchus was closed with suture in that report. In the year 2000 the rate of BPF after elective lobectomy for lung cancer should be no more than 1%, and for pneumonectomy it should be around 3% to 5%.
Etiology of Bronchopleural Fistula Several factors have been shown to increase the risk of developing a BPF as described in Table l. Some are controversial but others have been proven over and over again. Vester et aI,4 Yasher et al,6 Asamura et aI,2 al-Kattan et al,5 and alKattan et aF in 1995, have all shown that radiation increases the risk of BPF. Any bronchial stump that has been exposed to radiation should be covered with a muscle flap as described in the
Table 1. Factors That Contribute to BPF Preoperative risk factors Radiation and/or chemotherapy Immunocompromised host Systemic steroids Diabetes Inflammatory diseased lung/destroyed lung Intraoperative factors Surgical inexperience Pneumonectomy Long stump Right-sided resections Right pneumonectomy Devascularization of stump Residual cancer at stump Failure of recognition of BPF before closure Tension on bronchial closure Postoperative factors Prolonged ventilation Systemic steroids Reintubation
prevention section of this article. Yamamoto et a1 8 in 2000 has shown that preoperative chemoradiation decreases the bronchial mucosal blood supply and adversely affects healing. The pathophysiology of a BPF in this setting probably is ischemia. Right pneumonectomy versus left pneumonectomy has also been shown to increase the risk for BPF. In fact, Asamura et aF showed in 1992 that any type of right-sided pulmonary resection had a higher incidence ofBPF. This included lobectomy or bronchoplastic procedures when compared with similar surgeries on the left. Patel et a1 9 in 1992 and Wright et apo in 1996 have also found that right-sided pneumonectomy is a risk factor for BPF. We have had a similar experience. We have had 23 patients referred to us for treatment of a BPF. Twenty have had a previous pneumonectomy and 19 of these have been right sided. The pathophysiology of this increased incidence may be attributable to the aortic arch that is usually on the left. The bronchial stump after left pneumonectomy retracts under the arch and it, as well as other surrounding tissue, helps to cover it. This is in direct contrast to the right main stem bronchial closure after right pneumonectomy, which is left open in the empty pleural space. However, this theory does not explain why the incidence of BPF is higher for a right lower lobectomy than a left lower lobectomy. Some theorize that the lymphatic drainage may be a factor but the pathophysiology remains unclear.
Prevention of Brollchopleural FiJtula
Surgical inexperience is probably the most common cause of a BPF. Vester et al{ and alKattan et al s,7 have found that the incidence of BPF falls dramatically as the experience of the surgeon increases. Of course, the term experienced is never defined. Bjork 11 in 1956 and Kaplan et ajI2 in 1987, as well as others mentioned earlier, have shown a long bronchial stump to be another risk factor for a BPF. This is one surgical fact that most all surgeons agree on. The stump should be kept short to prevent contaminated oral-pharyngeal secretions from pooling on the closure. These pooled secretions lead to chronic infection and eventual can lead to separation and a BPF. Prolonged mechanical ventilation and/or reintubation are also risk factors as shown by Mitsudomi et al 13 in 1996 and by Wright et ajIo in 1996. This reason for this is probably twofold. The first is that the ventilator delivers high airway pressure and this can traumatize bronchial tissue. The second is the iatrogenic introduction of bacteria. It is of little surprise that any factor that negatively impacts on healing may also increase the risk of BPF. Insulin diabetes mellitus has been cited by Asamura et ai2 to be a risk factor as well as residual cancer at the stump. Similarly, the use of systemic steroids and other patients that are immunocompromised are at increased risk.
Prevention of Postresectional Bronchopleural Fistula Like most postoperative complications, the key to management is prevention. Prevention is best achieved in the operating room. Several of these factors are listed in Table 2. Early recognition is crucial. The stump must be checked for a leak before chest closure. If the patient has any of the Table 2. Intraoperative Techniques To Help Prevent BPF Intraoperative techniques Plan to harvest, or harvest muscle before pulmonary resection in any patient with risk factors for BPF Short stump Prevention of devascularization of stump Cover all right pneumonectomy Check frozen section of bronchial margin Avoid tension on bronchial closure
5
risk factors described earlier, one should plan to harvest a muscle preoperatively. Potential muscles available include: serratus anterior, latissimus dorsi, intercostal muscles, and pectoralis major. We do not believe that a pericardial fat pad or a pleural flap provides enough protection to the bronchus that is at increased risk. The former has unreliable length and blood supply and the latter is too thin. A patient at increased risk for a BPF should have a pedicled muscle flap. In true clinical practice in the year 2000, one rarely plans to do a pneumonectomy. The plan more often is for a sleeve lobectomy with or without a concomitant sleeve of the pulmonary artery. In this situation, a muscle is required to either wrap the bronchial anastomosis or to separate it from the pulmonary artery anastomosis. This helps avoid a BPF as well as a bronchialpulmonary artery fistula. Therefore, anytime we plan for a sleeve lobectomy (which always has the potential to turn into a pneumonectomy) an intercostal muscle flap is harvested. Some argue that this muscle also has an unreliable blood supply and they may choose a larger muscle, such as a serratus anterior or a latissimus dorsi. These muscles are larger, but take longer to harvest, require a partial rib resection to create a window to introduce it into the thorax, and have some cosmetic deformity. We prefer an intercostal muscle that is taken off the rib we enter over. It is taken with the cautery set on low. Special care is needed when it is harvested off of the superior rib. It must be devoid of all periosteum so it does not calcify later. It must also be harvested before the chest retractor is placed so it is not crushed. Therefore, the surgeon must plan on taking it before exploring the chest or another intercostal or extrathoracic muscle is necessary. When harvested carefully, it is a reliable, long, healthy pedicled muscle flap that takes less than 10 minutes to prepare. There is little to no morbidity or cosmetic deformity and the same surgeon that is doing the pulmonary resection can easily harvest it himself. Before taking the bronchus, all the lymph nodes around it should be resected. The bronchus should be cleaned of all surrounding tissue as well. Many surgeons worry about overdissection of the bronchus and fear that this may render it ischemic. Some therefore choose to leave some lymph nodes and/or surrounding tissue on it. We disagree with this practice. The bronchus should
6
Robert James Ceifolio
be devoid of all surrounding tissue. If a pneumonectomy is to be performed the arterial blood supply to the subcarinal (number 7 lymph node) should be taken at its base in the angle of the carina. This does not lead to an ischemic stump and it allows one to visualize the left main stem bronchus from the right chest to ensure a short stump. This is more difficult to do on the left but probably is not as important on that side. Once prepared, the bronchus is divided. If the cancer is close to the bronchus it should be cut with a knife and a stapling device should be avoided. Jack 14 reported a technique in 1965 of cutting a posterior flap of the membranous main stem bronchus and using this to close the main stem onto the trachea. The advantage of this technique is the lack of tension on the frail membranous portion of the bronchus and no residual stump. Sarsam and Moussali 15 in 1989 reported on their success with this technique. He did 332 pneumonectomies and closed the main stem bronchus with chromic catgut by using Jack'sl4 method. He had no BPFs. This concept makes sense because right-sided BPFs almost always occur in the middle of the bronchial closure. The membranous trachea tears or pulls away from the staple line probably because of tension. Jack's technique avoids tension. We currently use this technique for right-sided pneumonectomies only and have had no BPF with it. We use interrupted 2-0 or 3-0 Prolene, double-armed suture and position the knots away from the pulmonary artery. We use the stapling device (when otherwise appropriate) for left pneumonectomies and all other lobectomies. There still is controversy over whether a hand-sutured or stapled closure of the bronchus impacts on the incidence of BPF. We believe the concepts and technical points described earlier are more important then the material chosen. Once the bronchus is closed, the muscle (which is kept wrapped in a warm gauze) is brought into the field. First, one ensures there is no bleeding from the area it is to cover because it cannot be evaluated later. Second, one must ensure that the window used to introduce it into the chest (for extrathoracic muscles like the serratus and latissimus) is dry and of sufficient size. If it is too small it can compromise the arterial or venous blood supply. Finally, one must ensure the muscle itself is not bleeding. Once ready, the muscle is sewn over the stump as a buttress. We
used interrupted 3-0 Prolene double-armed sutures. The first stitch is carefully placed with a thin tacking bite of some surrounding tissue below the bronchial closure. The other two ends are sewn through the edge of the muscle. This is performed at the 3, 6, 9, and 12 o'clock positions around the bronchus and the muscle is then parachuted down on top of the bronchus. If properly harvested and tacked down it will not only grow onto the bronchus but it can be used to grow into the bronchus and plug very large fistulas in an infected field. In conclusion, a BPF is a life-threatening complication after pulmonary resection. The incidence is dramatically higher after pneumonectomy, especially right pneumonectomy. The key to treatment of a BPF is prevention. Prevention depends on meticulous intraoperative technique and the recognition of a small leak before chest closure. Any patient at increased risk should have a muscle placed over the stump, prophylactic at the time of initial pulmonary resection. This includes all patients that have a right pneumonectomy.
References I. Cerfolio RJ, Tummala RP, Holman WL, et al: A prospective algorithm for the management of air leaks after pulmonary resection. Ann Thorac Surg 66: 1726-1731, 1998 2. Asamura H, Naruke T, et al: Bronchopleural fistulas associated with lung cancer operations.J Thorac Cardiovasc Surg 104:1456-1464, 1992 3. Pomerantz M, Madsen L, et al: Surgical management of resistant mycobacterial tuberculosis and other mycobacterial pulmonary infections. Ann Thorac Surg 52:11081112, 1991 4. Vester, SR, Faber LP, et al: Bronchopleural fistula after stapled closure of bronchus. Ann Thorac Surg 52:12531258, 1991 5. AI-Kattan K, Cattalani L, Goldstraw P: Bronchopleural fistula after pneumonectomy with a hand suture technique. Ann Thorac Surg 58:1433-1436, 1994 6. Yashar J, Weitberg AB, et al: Preoperative chemotherapy and radiation therapy for stage IlIa carcinoma of the lung. Ann Thorac Surg 53:445-448, 1992 7. AI-Kattan K, Cattelani L, Goldstraw P: Bronchopleural fistula after pneumonectomy for lung cancer. Eur J Cardiothorac Surg 9:479-482, 1995 8. Yamamoto R, Tada H, et al: Effects of preoperative chemotherapy and radiation therapy on human bronchial blood flow. J Thorac Cardiovasc Surg 119:939-945, 2000 9. Patel RL, Townsend ER, Fountain SW: Elective pneumonectomy: Factors associated with morbidity and operative mortality. Ann Thorac Surg 54:84-88, 1992 10. Wright CD, Wain JC, Mathisen DF, et al: Postpneumonectomy bronchopleural fistula after sutured closure: In-
Prevention qf Bronchopleural Fistula
cidence, risk factors , and management.] Thorac Cardiovasc Surg 112:1367-1371 , 1996 II. Bjork YO: Suture mat erial and technique for bronchial closure and bronchial anastomosis.] Thorac Surg 32:2227, 1956 12. Kaplan DK, Whyte RI, Donne lly RD: Pulmonary resection using automatic stapling devices. Eur] Cardiothorac Surg 1:152, 1987
7
13. Mitsudomi T, Mizoue T, Yoshima tsu T, et al: Postoperative complications after pneumonectomy for treatment of lung cancer: Multivariate analysis. ] Surg Oncol 61:218222, 1996 14. Jack CD: Bronchial closure. Thorax 20:8, 1965 15 . Sarsam MAl, Moussali H: Technique of bronchial closure after pneumonectomy.] Thorac Cardiovasc Surg 98:220223, 1989
O
ne of the most dreaded complications after pneumonectomy is a bronchopleural fistula (BPF). Although there is no established "first reported case" of a BPF, surgeons have dealt with this devastating complication since the advent of elective pulmonary resection. BPF compromises the airway, infects the pleural space, and spills contaminated fluid into the only remaining lung, causing life-threatening pneumonia. Despite the advent of new heavy wire stapling devises, longer staples, and muscle flaps used to buttress bronchial closures, BPFs continue to occur. It continues to be a major cause of morbidity and mortality. This is especially true if early diagnosis and careful management are not preformed. Like any postoperative complication, prevention is the key to management. In this article we focus on the incidence, etiology, and prevention of BPF that occur after pulmonary resection.
From the Department of Surgery, Division ofCardiothoraci[ Surgery, University of Alabama at Birmingham, Birmingham, AL. Address reprint requests to Robert James Cerfolio, MD, Associate Professor of Surgery, Department of Surge~y, Division ofCardiothoracic Surgery, Universi~y of Alabama at Birmingham. 1900 University Blvd., THT Room 712, Birmingham. AL 35291. COjJyright © 2001 by WB. Saunders Compan)· 1043-0679/01/1301-0001$35.00/0 doi: 1 0.1 053/stcs.2001.22493
Definition BPF is defined as a communication between a lobar or segmental pulmonary bronchi and the pleural space. Despite this definition, the term BPF continues to be twisted and misused by many physicians. Persistent air leaks that originate from the pulmonary parenchyma are NOT bronchopleural fistulas. We believe these are best termed alveolar-pleural fistulas as described by us in 1998. 1 This distinction is not just one of semantics. The treatment of these two types of fistula is quite different. A BPF almost always requires some sort of reoperation whereas alveolar-pleural fistulas almost never do. The latter are best treated with patience and a Heimlich valve.
Incidence of Bronchopleural Fistula The incidence of a BPF after lobectomy is quite low. In 1992, Asamura et aF reported an incidence of only 1.2% for 1,586 patients who underwent lobectomy for lung cancer. However, the incidence is significantly higher after pneumonectomy. The exact reason for this is unknown. There are several explanations. The most likely explanation is that the remaining lung after lobectomy helps cover and indirectly buttress the closed bronchi. The incidence of BPF after pneumonectomy depends on whether it is a right or left and on the
Seminars in Thoracic and Cardiovascular Surgery, Vol 13, No 1 (january), 2001: pp 3-7
3
4
Robert James Cerfolio
indication for surgery. Asamurua et aF reported an incidence for BPF after pneumonectomy for lung cancer of 4.5%. However, if one stratifies his data, the incidence of BPF was 8.6% for the 162 patients that had a right pneumonectomy and 2.3% for the 302 patients that had a left pneumonectomy. The incidence increases even further when pneumonectomy is performed for benign disease (ie, inflammatory disease and/or destroyed lung). In 1991, Pomerantz et aP reported on patients who had pneumonectomy for resistant Mycobacterium infections. In that study, there were 40 pneumonectomies performed and 8 patients had a BPF (incidence 20%) despite the use of muscle flaps in many. Four of these 8 patients died. Other large reports have found similar results. In 1991, Vester et a1 4 reported an incidence for BPF (4.5%) on 506 patients who underwent pneumonectomy for cancer. He found that the incidence was 4.1 % when the bronchus was stapled and 17.6% when sutured. Careful review, however, shows that he prefers stapling and only sutured when needed. Only 17 patients had sutured closures and 489 were stapled. Of the 23 patients in that report who had a BPF, 17 patients had a right pneumonectomy and 6 had a left. In that same article, Vester 4 also reported a 1.0% BPF rate for 965 patients that had a lobectomy or bilobectomy. His incidence for a BPF after 650 segmentectomies of 0.3% was even less. This lends credence to the theory that the surrounding lung helps cover the stump and avoids a postresectional BPF. Finally, in 1994 AI-Kattan et aP reported an incidence of 1.3% for BPF after 530 consecutive pneumonectomies. The bronchus was closed with suture in that report. In the year 2000 the rate of BPF after elective lobectomy for lung cancer should be no more than 1%, and for pneumonectomy it should be around 3% to 5%.
Etiology of Bronchopleural Fistula Several factors have been shown to increase the risk of developing a BPF as described in Table l. Some are controversial but others have been proven over and over again. Vester et aI,4 Yasher et al,6 Asamura et aI,2 al-Kattan et al,5 and alKattan et aF in 1995, have all shown that radiation increases the risk of BPF. Any bronchial stump that has been exposed to radiation should be covered with a muscle flap as described in the
Table 1. Factors That Contribute to BPF Preoperative risk factors Radiation and/or chemotherapy Immunocompromised host Systemic steroids Diabetes Inflammatory diseased lung/destroyed lung Intraoperative factors Surgical inexperience Pneumonectomy Long stump Right-sided resections Right pneumonectomy Devascularization of stump Residual cancer at stump Failure of recognition of BPF before closure Tension on bronchial closure Postoperative factors Prolonged ventilation Systemic steroids Reintubation
prevention section of this article. Yamamoto et a1 8 in 2000 has shown that preoperative chemoradiation decreases the bronchial mucosal blood supply and adversely affects healing. The pathophysiology of a BPF in this setting probably is ischemia. Right pneumonectomy versus left pneumonectomy has also been shown to increase the risk for BPF. In fact, Asamura et aF showed in 1992 that any type of right-sided pulmonary resection had a higher incidence ofBPF. This included lobectomy or bronchoplastic procedures when compared with similar surgeries on the left. Patel et a1 9 in 1992 and Wright et apo in 1996 have also found that right-sided pneumonectomy is a risk factor for BPF. We have had a similar experience. We have had 23 patients referred to us for treatment of a BPF. Twenty have had a previous pneumonectomy and 19 of these have been right sided. The pathophysiology of this increased incidence may be attributable to the aortic arch that is usually on the left. The bronchial stump after left pneumonectomy retracts under the arch and it, as well as other surrounding tissue, helps to cover it. This is in direct contrast to the right main stem bronchial closure after right pneumonectomy, which is left open in the empty pleural space. However, this theory does not explain why the incidence of BPF is higher for a right lower lobectomy than a left lower lobectomy. Some theorize that the lymphatic drainage may be a factor but the pathophysiology remains unclear.
Prevention of Brollchopleural FiJtula
Surgical inexperience is probably the most common cause of a BPF. Vester et al{ and alKattan et al s,7 have found that the incidence of BPF falls dramatically as the experience of the surgeon increases. Of course, the term experienced is never defined. Bjork 11 in 1956 and Kaplan et ajI2 in 1987, as well as others mentioned earlier, have shown a long bronchial stump to be another risk factor for a BPF. This is one surgical fact that most all surgeons agree on. The stump should be kept short to prevent contaminated oral-pharyngeal secretions from pooling on the closure. These pooled secretions lead to chronic infection and eventual can lead to separation and a BPF. Prolonged mechanical ventilation and/or reintubation are also risk factors as shown by Mitsudomi et al 13 in 1996 and by Wright et ajIo in 1996. This reason for this is probably twofold. The first is that the ventilator delivers high airway pressure and this can traumatize bronchial tissue. The second is the iatrogenic introduction of bacteria. It is of little surprise that any factor that negatively impacts on healing may also increase the risk of BPF. Insulin diabetes mellitus has been cited by Asamura et ai2 to be a risk factor as well as residual cancer at the stump. Similarly, the use of systemic steroids and other patients that are immunocompromised are at increased risk.
Prevention of Postresectional Bronchopleural Fistula Like most postoperative complications, the key to management is prevention. Prevention is best achieved in the operating room. Several of these factors are listed in Table 2. Early recognition is crucial. The stump must be checked for a leak before chest closure. If the patient has any of the Table 2. Intraoperative Techniques To Help Prevent BPF Intraoperative techniques Plan to harvest, or harvest muscle before pulmonary resection in any patient with risk factors for BPF Short stump Prevention of devascularization of stump Cover all right pneumonectomy Check frozen section of bronchial margin Avoid tension on bronchial closure
5
risk factors described earlier, one should plan to harvest a muscle preoperatively. Potential muscles available include: serratus anterior, latissimus dorsi, intercostal muscles, and pectoralis major. We do not believe that a pericardial fat pad or a pleural flap provides enough protection to the bronchus that is at increased risk. The former has unreliable length and blood supply and the latter is too thin. A patient at increased risk for a BPF should have a pedicled muscle flap. In true clinical practice in the year 2000, one rarely plans to do a pneumonectomy. The plan more often is for a sleeve lobectomy with or without a concomitant sleeve of the pulmonary artery. In this situation, a muscle is required to either wrap the bronchial anastomosis or to separate it from the pulmonary artery anastomosis. This helps avoid a BPF as well as a bronchialpulmonary artery fistula. Therefore, anytime we plan for a sleeve lobectomy (which always has the potential to turn into a pneumonectomy) an intercostal muscle flap is harvested. Some argue that this muscle also has an unreliable blood supply and they may choose a larger muscle, such as a serratus anterior or a latissimus dorsi. These muscles are larger, but take longer to harvest, require a partial rib resection to create a window to introduce it into the thorax, and have some cosmetic deformity. We prefer an intercostal muscle that is taken off the rib we enter over. It is taken with the cautery set on low. Special care is needed when it is harvested off of the superior rib. It must be devoid of all periosteum so it does not calcify later. It must also be harvested before the chest retractor is placed so it is not crushed. Therefore, the surgeon must plan on taking it before exploring the chest or another intercostal or extrathoracic muscle is necessary. When harvested carefully, it is a reliable, long, healthy pedicled muscle flap that takes less than 10 minutes to prepare. There is little to no morbidity or cosmetic deformity and the same surgeon that is doing the pulmonary resection can easily harvest it himself. Before taking the bronchus, all the lymph nodes around it should be resected. The bronchus should be cleaned of all surrounding tissue as well. Many surgeons worry about overdissection of the bronchus and fear that this may render it ischemic. Some therefore choose to leave some lymph nodes and/or surrounding tissue on it. We disagree with this practice. The bronchus should
6
Robert James Ceifolio
be devoid of all surrounding tissue. If a pneumonectomy is to be performed the arterial blood supply to the subcarinal (number 7 lymph node) should be taken at its base in the angle of the carina. This does not lead to an ischemic stump and it allows one to visualize the left main stem bronchus from the right chest to ensure a short stump. This is more difficult to do on the left but probably is not as important on that side. Once prepared, the bronchus is divided. If the cancer is close to the bronchus it should be cut with a knife and a stapling device should be avoided. Jack 14 reported a technique in 1965 of cutting a posterior flap of the membranous main stem bronchus and using this to close the main stem onto the trachea. The advantage of this technique is the lack of tension on the frail membranous portion of the bronchus and no residual stump. Sarsam and Moussali 15 in 1989 reported on their success with this technique. He did 332 pneumonectomies and closed the main stem bronchus with chromic catgut by using Jack'sl4 method. He had no BPFs. This concept makes sense because right-sided BPFs almost always occur in the middle of the bronchial closure. The membranous trachea tears or pulls away from the staple line probably because of tension. Jack's technique avoids tension. We currently use this technique for right-sided pneumonectomies only and have had no BPF with it. We use interrupted 2-0 or 3-0 Prolene, double-armed suture and position the knots away from the pulmonary artery. We use the stapling device (when otherwise appropriate) for left pneumonectomies and all other lobectomies. There still is controversy over whether a hand-sutured or stapled closure of the bronchus impacts on the incidence of BPF. We believe the concepts and technical points described earlier are more important then the material chosen. Once the bronchus is closed, the muscle (which is kept wrapped in a warm gauze) is brought into the field. First, one ensures there is no bleeding from the area it is to cover because it cannot be evaluated later. Second, one must ensure that the window used to introduce it into the chest (for extrathoracic muscles like the serratus and latissimus) is dry and of sufficient size. If it is too small it can compromise the arterial or venous blood supply. Finally, one must ensure the muscle itself is not bleeding. Once ready, the muscle is sewn over the stump as a buttress. We
used interrupted 3-0 Prolene double-armed sutures. The first stitch is carefully placed with a thin tacking bite of some surrounding tissue below the bronchial closure. The other two ends are sewn through the edge of the muscle. This is performed at the 3, 6, 9, and 12 o'clock positions around the bronchus and the muscle is then parachuted down on top of the bronchus. If properly harvested and tacked down it will not only grow onto the bronchus but it can be used to grow into the bronchus and plug very large fistulas in an infected field. In conclusion, a BPF is a life-threatening complication after pulmonary resection. The incidence is dramatically higher after pneumonectomy, especially right pneumonectomy. The key to treatment of a BPF is prevention. Prevention depends on meticulous intraoperative technique and the recognition of a small leak before chest closure. Any patient at increased risk should have a muscle placed over the stump, prophylactic at the time of initial pulmonary resection. This includes all patients that have a right pneumonectomy.
References I. Cerfolio RJ, Tummala RP, Holman WL, et al: A prospective algorithm for the management of air leaks after pulmonary resection. Ann Thorac Surg 66: 1726-1731, 1998 2. Asamura H, Naruke T, et al: Bronchopleural fistulas associated with lung cancer operations.J Thorac Cardiovasc Surg 104:1456-1464, 1992 3. Pomerantz M, Madsen L, et al: Surgical management of resistant mycobacterial tuberculosis and other mycobacterial pulmonary infections. Ann Thorac Surg 52:11081112, 1991 4. Vester, SR, Faber LP, et al: Bronchopleural fistula after stapled closure of bronchus. Ann Thorac Surg 52:12531258, 1991 5. AI-Kattan K, Cattalani L, Goldstraw P: Bronchopleural fistula after pneumonectomy with a hand suture technique. Ann Thorac Surg 58:1433-1436, 1994 6. Yashar J, Weitberg AB, et al: Preoperative chemotherapy and radiation therapy for stage IlIa carcinoma of the lung. Ann Thorac Surg 53:445-448, 1992 7. AI-Kattan K, Cattelani L, Goldstraw P: Bronchopleural fistula after pneumonectomy for lung cancer. Eur J Cardiothorac Surg 9:479-482, 1995 8. Yamamoto R, Tada H, et al: Effects of preoperative chemotherapy and radiation therapy on human bronchial blood flow. J Thorac Cardiovasc Surg 119:939-945, 2000 9. Patel RL, Townsend ER, Fountain SW: Elective pneumonectomy: Factors associated with morbidity and operative mortality. Ann Thorac Surg 54:84-88, 1992 10. Wright CD, Wain JC, Mathisen DF, et al: Postpneumonectomy bronchopleural fistula after sutured closure: In-
Prevention qf Bronchopleural Fistula
cidence, risk factors , and management.] Thorac Cardiovasc Surg 112:1367-1371 , 1996 II. Bjork YO: Suture mat erial and technique for bronchial closure and bronchial anastomosis.] Thorac Surg 32:2227, 1956 12. Kaplan DK, Whyte RI, Donne lly RD: Pulmonary resection using automatic stapling devices. Eur] Cardiothorac Surg 1:152, 1987
7
13. Mitsudomi T, Mizoue T, Yoshima tsu T, et al: Postoperative complications after pneumonectomy for treatment of lung cancer: Multivariate analysis. ] Surg Oncol 61:218222, 1996 14. Jack CD: Bronchial closure. Thorax 20:8, 1965 15 . Sarsam MAl, Moussali H: Technique of bronchial closure after pneumonectomy.] Thorac Cardiovasc Surg 98:220223, 1989