- Email: [email protected]
Treatment of bronchopleural fistula after pneumonectomy
J THORAC
CARDIOV ASC SURG
90:813-817, 1985
Treatment of bronchopleural fistula after pneumonectomy Breakdown of the closure of the main-stem bronchus after pneumonectomy is a dreaded complication,
and empyema and bronchopleural fistula frequently developin patients who survive. Management of these fistulas remaim a formidable therapeutic challenge, which has been approached with a variety of surgical tecbniques. We report our experience with anterior transpericardial closure, emphasizing the ability to expose either main-stem bronchus by this approach. The case histories of three patients who had bronchopleural fIStula after pneumonectomy are presented. The first patient had left penumonectomy for complicated tuberculosis; the second had right pneumonectomy for neoplasm; and the third had right pneumonectomy for trauma. All fistulas were treated surgically via a median sternotomy and ~pericardial approach to the distal trachea. The posterior pericardium was divided between the superior vena cava and aorta. In-continuity staple closure (with two lines of staples) of the proximal main-stem bronchus was employed in all cases. Two patients remain clinically weD 21 and 17 montlIS after the operation. The third patients did weD initially but developed a recurrent bronchopleural fIStula 21h montlIS after the operation and has required repeat closure with pedicled muscle flaps. In postpneumonectomy bronchopleural fIStula, the anterior, transpericardial approach to bronchial closure bas several advantages: the relatively weD-tolerated median sternotomy, the avoidance of dealing directly with areas of postoperative scarring and the devascularized bronchial stump, the avoidance of areas of chronic sepsis, and the avoidance of thoracoplastic surgical deformity of the chest wall, with possible associated compromise in pulmonary function. Our experience also indicates that either main-stem bronchus is accessible through an approach between the superior vena cava and aorta, without division of either pulmonary artery. .
John C. Baldwin, M.D. (by invitation), and James B. D. Mark, M.D., Stanford. Calif.
BronchoPleural fistula after pulmonary resection is associated with very substantial morbidity and mortality, and it is a considerable technical challenge to the surgeon to select the appropriate timing and technique for operative intervention. 1-] The occurrence of this complication after pneumonectomy is of particular concern because of the magnitude of the leak, with its associated respiratory insufficiency and risk of flooding and contamination of the contralateral lung. When bronchopleural fistula does occur, it is essential to establish adequate, dependent drainage.' In cases in which very early disruption occurs, it is sometimes From the Department of Cardiovascular Surgery, and the Department of Surgery, Division of Thoracic Surgery. Stanford University Medical Center Stanford, Calif- 94305. Read at the Sixty-fifth Annual Meeting of The American Association for Thoracic Surgery, New Orleans, La., April 29-May 1, 1985.
possible to resuture the stump, although such efforts may carry considerable morbidity and mortality.' However, chronic bronchopleural fistula with associated empyema and scarring is a technically more formidable and clinically disabling problem. Some argue for the value of drainage alone and long-term conservative management, but definitive operative closure is usually preferable.l' Extensive experience with pedicled flaps of muscle and omentum has been reported; these methods are demonstrably effective, especially when definitive excision of the sinus tract back to normal bronchial tissue is feasible." In 1983, Anderson and Li 9 reported a case of postpneumonectomy bronchopleural fistula, in which they used the anterior transpericardial approach to the right main-stem bronchus. Noting difficulties with devascularization, scarring, and chronic sepsis, they? described the successful use of this approach in a young patient with bronchopleural fistula occurring after right 813
The Journal of
814 Baldwin and Mark
Thoracic and Cardiovascular Surgery
laryngeal nerve, and esophagus. The TA-30 stapler (with 4.8 rom staples) was then applied, flush with the take-off ofthe bronchus from the trachea-on the left in one case and on the right in the other two (Fig. 2). Two applications of the stapling device were employed, and the bronchus was not divided in these cases.
Patients and results
Fig. 1. Exposure via the standard median sternotomy. Retraction of the superior vena cava (SVC) to the right and the aorta (Ao) to the left provides the approach to the tracheal bifurcation. RPA, Right pulmonary artery.
pneumonectomy for trauma. This approach to the tracheal bifurcation was reported by Padhi and Lynn'? in their 1960 series of bronchopleural fistulas. They suggested ligation of the pulmonary artery and veins inside the pericardium and dissection of the bronchial stump up to the level of the carina." The technique has received considerable attention in Europe and particularly in the Soviet Union.":"
Operative technique All patients in this group underwent preoperative bronchoscopic examinations, principally to assess the length of the bronchial stump. We believe that a stump of at least 10 rom is necessary to permit secure transpericardial bronchial closure. Standard median sternotomy incisions were employed and the pericardium was opened in a stellate fashion (Fig. 1). Tapes were placed around the aorta and superior vena cava, and the vessels were retracted apart (Fig. 1, inset). A vertical incision was made in the posterior pericardium above the right pulmonary artery; dissection of lymphoid and areolar tissue anterior to the carina permitted exposure of both main-stem bronchi. Care must be taken to avoid injury to the right pulmonary artery, left recurrent
CASE 1. A 64-year-old man had a history of treated pulmonary tuberculosis. In 1982, he underwent left upper lobectomy at another hospital for an undiagnosed pulmonary nodule. He had a prolonged air leak postoperatively and underwent completion pneumonectomy; a bronchopleural fistula developed. He was then treated with drainage procedures and thoracoplasty, but the fistula and chronic septic drainage persisted. He was transferred to our care, and after optimizing his nutritional status and external drainage, we operated on July 18, 1983. At bronchoscopy, we noted a 5 em residual left main bronchial stump. In-continuity staple closure of the left main bronchus via the transpericardial approach was performed. The patient remains well, without evidence of recurrent fistula, 21 months after the operation. CASE 2. A 41-year-old man underwent right pneumonectomy in October of 1976 at another hospital for carcinoid adenoma of the bronchus intermedius. During his postoperative course he had fever and cough, with development of a bronchopleural fistula and empyema. He was treated by drainage followed by rib resection and primary closure. without success. In December of 1978, the patient was transferred to our hospital, and an open window thoracostomy was created. He did well subsequently with intermittent dilatations to ensure adequate drainage. The bronchopleural fistula remained open. On Dec. 5, 1983, the right main-stem bronchus was closed by the technique described herein. The patient remains well 17 months after the operation. CASE 3. A 33-year-old man had a gunshot wound to the right side of the chest, which necessitated right pneumonectomy in April of 1984. He had a complicated postoperative course, with reexploration for bleeding, cardiac arrest, adult respiratory distress syndrome, disseminated intravascular coagulation, sepsis, bronchopleural fistula, and pericardial effusion treated by pericardiectomy. Three months after the original injury he was transferred to our hospital, where open window thoracostomy drainage was performed. Sepsis subsided, and the patient's general condition improved. Two months later, he was readmitted for bronchial closure. At bronchoscopy, a I to 2 ern right main-stem bronchial stump was noted. Dense intrapericardial adhesions were noted from the patient's previous operation, and the esophagus was densely adherent to the posterior aspect of the distal trachea and proximal right main-stem bronchus. In-continuity staple closure of the right main-stem bronchus was performed as described herein. The patient did well postoperatively; bronchoscopy and thoracoscopy performed 2 months postoperatively revealed no evidence of fistula recurrence. However, approximately 2 weeks later, the patient noted a "Whistling" sound in the region of the thoracostomy and was found to have a recurrent fistula. He has subsequently undergone reclosure of the bronchus via a lateral thoracotomy, with placement of pectoralis major and latissimus dorsi myocutaneous pedicled flaps. He is currently clinically well 2 months after the operation.
Volume 90 Number 6 December, 1985
Postpneumonectomy bronchopleural fistula
8 15
Discussion The critical importance of prompt, adequate, and dependent drainage for postpneumonectomy bronchopleural fistula and empyema cannot be overstressed.' 15, J6 However, persistent patency of the fistula usually necessitates subsequent additional operative intervention. Some have placed emphasis on the potential role of thoracoplasty in reducing the size of the empyema space, but cosmetic and functional deformity, as well as potential for compromised pulmonary function, are drawbacks to this approach.' When the lateral, transthoracic approach to definitive closure is selected, it is important to excise completely the sinus tract, providing for closure in the region of normal, well-vascularized bronchial tissue. Widespread experience has been accumulated with respect to use of pedicled flaps to enhance security of closure in this setting, with many groups reporting high rates of success with single-stage closure.i'? Pairolero and associates7• 18 have reported impressive results with a variety of extrathoracic skeletal muscles transposed into the thorax for fistula closure. They have described the use of pedicles of latissimus dorsi, pectoralis major, serratus anterior, pectoralis minor, rectus abdominis, and combinations of these. These sometimes discouraging results and the formidable technical challenges associated with operating in areas of scarring, chronic infection, and devascularized bronchial tissue prompt consideration of alternative surgical strategies for treatment of postpneumonectomy bronchopleural fistula. The approach described here, as pointed out by Anderson and Li,9 has received attention in the European literature but has not been widely emphasized in this country. The advantages of the less morbid median sternotomy and the avoidance of areas of chronic sepsis and scarring in previous surgical fields are evident. Of course, pneumonectomy should be performed at the carinallevel, and this technique is applicable only in cases in which a residual bronchial stump is present. Preoperative radiographs may indicate approximate length, but we suggest that 10 mm of the residual bronchus should be documented by preoperative bronchoscopy. Division of the bronchial stump with placement of viable tissue between the closure lines is the established and preferable technique; moreover, division of the bronchus is necessary for healing to occur. However, short segments of available bronchus within the operative field may severely limit the possibility for secure closure of both ends of the denuded bronchus. In cases in which exposure is particularly limited, division may increase the risk of soilage of the operative field.
Fig. 2. Application of the stapling device to the left main bronchus, SVC, Superior vena cava. Ao, Aorta. RPA, Right pulmonary artery.
Strictly speaking, unless combined with excision of the distal segment, division does leave a persistent blind stump, possibly containing mucus-producing bronchial mucosa. Cognizance of the proximity of the esophagus posteriorly is essential during dissection, and esophageal injury has been reported with this technique." In cases in which previous intrapericardial dissection has been performed, the dissection may be more difficult, negating some advantages of the operation, as in the case in which a recurrent fistula developed. Ilakim and Milstein" have reported on the incidence of bronchopleural fistula after pneumonectomy, comparing results for patients in whom bronchial closure was accomplished with the parallel jaw stapler (TA55) with those in whom the newer hinged stapler (Premium TA-55) was u.-d. They found a higher incidence (15.2% versus 4.2%, p <, '"1.05) of bronchopleural fistula when the hinged device was used, with incomplete staple closure near the hinge end of the closure shown in radiographic studies in cadavers. This finding is noteworthy; the Premium, hinged device was used in this group, and the one patient in whom the fistula recurred did have leakage in the area of the "heel" or hinge on bronchoscopic examination. We suggest that either main-stem bronchus can be approached between the superior vena cava and aorta. The approach to the left main-stem bronchus to the left of the aorta and above the left atrial appendage requires
The Journal of Thoracic and Cardiovascular Surgery
816 Baldwin and Mark
displacement of the heart to the right and may not be well-tolerated hemodynamically. We have not found division of either pulmonary artery to be necessary or advantageous. REFERENCES Hankins JR, Miller JE, Attar S, Satterfield JR, McLaughlin JS: Bronchopleural fistula. Thirteen-year experience with 77 cases. J THORAC CARDIOVASC SURG 76:755-762,1978 2 Barker WL, Faber P, Ostermiller WE, Langston HT: Management of resistant bronchopleural fistulas. J THORAC CARDIOVASC SURG 62:393-401, 1971 3 Maksimov lA, Osmakov NA: Results of treatment of postoperative bronchopleural complications in lung cancer. Vestn Khir 114:63-68, 1975 4 Conlan AA, Boyd AD, Spencer FC: The operative management of acute postpneumonectomy bronchopleural fistula after flush bronchial amputation. S Afr Med J 61:792-794, 1982 5 Weissberg D: Empyema and bronchopleural fistula. Experience with open window thoracostomy. Chest 82:447-450, 1982 6 Van der Heyde MN, Verwers HR, Van Leusen R: Closure of a bronchial fistula after pneumonectomy in the care of carcinoma in a patient in chronic dialysis. Acta Chir Belg 77:271-274, 1978 7 Pairolero PC, Arnold PG: Bronchopleural fistula. Treatment by transposition of pectoralis major muscle. J THORAC CARDIOVASC SUR? 79:142-145, 1980 8 Virkkula L, Ferola S: Use of omental pedicle for the treatment of bronchial fistula after lower lobectomy. Report of two cases. Scand J Thorac Cardiovasc Surg 9:287-290, 1975 9 Anderson RP, Li W: Anterior transpericardial closure of a main bronchus fistula after pneumonectomy. Am J Surg 145:630-632, 1983 10 Padhi RK, Lynn FB: The management of bronchopleural fistulas. J THORAC CARDIOVASC SURG 39:385-393, 1960 11 Bogush LK, Semenenkow JL, Polianskii VA: Surgical treatment of bronchopleural complications after pneumonectomy. Khirurgiia (Mosk) 11:115-120, 1976 12 Slepukha 1M: Treatment of postoperative bronchopleural fistulas by transpericardical occlusion of the main bronchus using the L.K. Bogush transthoracic approach. Grudn Khir 6:85-88, 1976 13 Maasen W: The transstemal and transpericardial approach for surgical treatment of fistulas of the main bronchus after pneumonectomy. Thoraxchirurgie 23:257261, 1975 14 Bogush LK, Travin AA, Semenenkow JL: Transperikardiale Operationen an den Hauptbroachien and Langengefassen, Stuttgart, 1971, Hoppokrates-Verlag 15 Shamji FM, Ginsberg RJ, Cooper JD, Spratt EH, GoWberg M, Waters PF, lives R, Todd TR, Pearson FG: Open window thoracoscopy in the management of postpneumo-
16
17
18
19
nectomy empyema with or without bronchopleural fistula. J THORAC CARDIOVASC SURG 86:818-822, 1983 Goldstraw P: Treatment of postpneumonectomy empyema. The case for fenestration. Thorax 34:740-745, 1979 Miller JI, Mansour KA, Nahai F, Jurkiewicz MJ, Hatcher CR: Single stage complete muscle flap closure of the postpneumonectomy empyema space. A new method and possible solution to a disturbing complication. Ann Thorac Surg 38:227-231, 1984 Pairolero PC, Arnold PG, Piehler JM: Intrathoracic transposition of extra thoracic skeletal muscle. J THoRAe CARDIOVASC SURG 86:809-817, 1983 Hakim M, Milstein BB: Role of automatic staplers in the aetiology of bronchopleural fistula. Thorax 40:27-31, 1985
Discussion DR. HERMES C. GRILLO Boston, Mass.
This technique has been used more in Europe. Perelman and Ambatjello ' described its use transstemally and transpericardially for the left main bronchus in 1970 and Bogusch, Travin, and Semenenkow' for the right in 1971. While Padhi and Lynn) had suggested the approach in 1960, I believe Abruzzini' first described it in detail in 1961. Though he did not dwell on the division of the pericardium, he illustrated the basic technique extensively. We have done a small number of closures of the right main bronchus by this approach, including one in a patient who had had 5,000 rads, which made it more difficult but still feasible. I feel emphatically that in addition to stapling, one should divide the bronchus to avoid recurrence of the fistula. We use staples distally and suture the proximal end. We also make a point of interposing tissue between the bronchial ends. For this I found it convenient to use a collar incision, pedicle the sternohyoid muscle, and turn it down. It easily reaches a point between the two bronchial ends. Then there is no question of refistulization. For fistulas of the left main bronchus-and the authors have properly emphasized that on both sides one needs a long stump to use this technique-we have preferred a right thoracotomy in a fresh field. We divide the left main bronchus at its origin and interpose an intercostal muscle flap between the bronchial ends. The authors did not describe what they did after bronchial closure. In our cases, we have followed fistula repair in a few weeks with a Clagett procedure. This has been a successful way of handling the residual cavity. I shall close by mentioning that Professor Perelman,' in a recent communication, points out that he now believes it is better to approach all main bronchial fistulas by thoracotomy and reamputate the long stumps because he has had difficulty with recurrent empyema from the residual stump and from foreign material in the stump. We have not experienced this in our small series.
Volume 90 Number 6
Postpneumonectomy bronchopleural fistula
8 17
December, 1985
2
3 4
5
REFERENCES Perelman MI, Ambatjello GP: Transpleuraler, transtemaler und Kontralateraler Zugang bei Operationen Wegen Bronchialfistel nach Pneumonektomie. Thorac Chir Chir Vaskul 18:45, 1970 Bogush LK, Travin AA, Semenenkow JL: Transperikardiale Operationen an den Hauptbroachien and Langengefassen, Stuttgart, 1971, Hoppokrates-Verlag Padhi RK, Lynn FB: The management of bronchopleural fistulas. J THORAC CARDIOVASC SURG 39:385-393, 1960 Abruzzini P: Trattamento chirurgico delle fistule del bronco principale consecutive a pneumonectomia per tubercolosi. Chir Torac 14:165, 1961 Perelman MI: Late treatment of chronic bronchopleural fistula with long stump after pneumonectomy, International Trends in General Thoracic Surgery, Vol 2, H Eschapasse, H Grillo, eds., Philadelphia, W. B. Saunders Company (in press)
DR. MARK M. RA VITCH Pittsburgh. Pa.
I address myself to the question of in-continuity stapling of the bronchus. There is ample experimental evidence and clinical experience to show that in the gastrointestinal tract in-continuity stapling is not dependable. The cut bronchus heals not where the sutures or the staples are placed but at its cut end, and in these patients that cut end is a long way off, is avascular, and is surrounded by infection. Mucosa does not heal to mucosa. If a person kept his mouth closed for a year, the lips would not heal together. Finally, in about 1908, William Stewart Halsted reported that in dogs a ligature of the bronchus, could be removed again many months later and the bronchus restored to its normal patency. The approach, of course, is beautiful. Two thirds of the results are outstanding. DR. KAMAL A. MANSOUR Atlanta. Ga.
I believethis technique might have its place in early or small bronchopleural fistulas. However, I wonder if mere secondary closure of a stump that failed to heal is the proper approach and whether the factor or factors known or unknown that caused disruption in the first place are still active.
With few exceptions bronchopleural fistulas are usually associated with postpneumonectomy empyemas, and our aim therefore should not only be directed toward closure of the fistula but also to definite management of a large infected empyema space. With the advent of muscle flaps, the management of the notorious postpneumonectomy bronchopleural fistulas has been made much easier. I would like to ask the authors whether effort was made to assess the size of the fistula and also the time interval between onset of the fistula and reoperation. Most of us, I believe, have seen tiny fistulas that closed spontaneously in a few weeks. The cases presented today are few indeed and only time will tell. DR. RICHARD P. ANDERSON Seattle, Wash.
Main bronchial fistula after pneumonectomy is, fortunately, an unusual situation that is extremely difficult to manage successfully. I particularly enjoyed the authors' description of the transpericardial approach to the left main bronchus. I must agree with the other discussants, however, about the possible relationship of in-continuity stapling of the bronchus and the development of recurrent fistula. In chronic fistula the area where the bronchus penetrates the mediastinal wall is fixed and rigid scar and must place considerable tension on the staple line, especially on the right side. If at all possible, division of the bronchus is probably best. This also permits interposition of a viable muscle flap, as suggested by Dr. Grillo, to further promote healing in a scarred and devascularized area. DR. BALDWIN (Closing) I would like to thank Dr. Grillo for bringing his wisdom and experience to bear on this topic, and I appreciate the thoughtful comments made by the other discussants. We are aware that the most controversial aspect of these three cases is the technique of in-continuity closure of the bronchus. Certainly, the established and conservative technique involving division of the bronchus would be preferable when it is feasible technically and when it is possible to accomplish, without soilage of the mediastinum. Particularly in those cases in which the retained bronchial stump is especially long, division may be technically feasible.
CARDIOV ASC SURG
90:813-817, 1985
Treatment of bronchopleural fistula after pneumonectomy Breakdown of the closure of the main-stem bronchus after pneumonectomy is a dreaded complication,
and empyema and bronchopleural fistula frequently developin patients who survive. Management of these fistulas remaim a formidable therapeutic challenge, which has been approached with a variety of surgical tecbniques. We report our experience with anterior transpericardial closure, emphasizing the ability to expose either main-stem bronchus by this approach. The case histories of three patients who had bronchopleural fIStula after pneumonectomy are presented. The first patient had left penumonectomy for complicated tuberculosis; the second had right pneumonectomy for neoplasm; and the third had right pneumonectomy for trauma. All fistulas were treated surgically via a median sternotomy and ~pericardial approach to the distal trachea. The posterior pericardium was divided between the superior vena cava and aorta. In-continuity staple closure (with two lines of staples) of the proximal main-stem bronchus was employed in all cases. Two patients remain clinically weD 21 and 17 montlIS after the operation. The third patients did weD initially but developed a recurrent bronchopleural fIStula 21h montlIS after the operation and has required repeat closure with pedicled muscle flaps. In postpneumonectomy bronchopleural fIStula, the anterior, transpericardial approach to bronchial closure bas several advantages: the relatively weD-tolerated median sternotomy, the avoidance of dealing directly with areas of postoperative scarring and the devascularized bronchial stump, the avoidance of areas of chronic sepsis, and the avoidance of thoracoplastic surgical deformity of the chest wall, with possible associated compromise in pulmonary function. Our experience also indicates that either main-stem bronchus is accessible through an approach between the superior vena cava and aorta, without division of either pulmonary artery. .
John C. Baldwin, M.D. (by invitation), and James B. D. Mark, M.D., Stanford. Calif.
BronchoPleural fistula after pulmonary resection is associated with very substantial morbidity and mortality, and it is a considerable technical challenge to the surgeon to select the appropriate timing and technique for operative intervention. 1-] The occurrence of this complication after pneumonectomy is of particular concern because of the magnitude of the leak, with its associated respiratory insufficiency and risk of flooding and contamination of the contralateral lung. When bronchopleural fistula does occur, it is essential to establish adequate, dependent drainage.' In cases in which very early disruption occurs, it is sometimes From the Department of Cardiovascular Surgery, and the Department of Surgery, Division of Thoracic Surgery. Stanford University Medical Center Stanford, Calif- 94305. Read at the Sixty-fifth Annual Meeting of The American Association for Thoracic Surgery, New Orleans, La., April 29-May 1, 1985.
possible to resuture the stump, although such efforts may carry considerable morbidity and mortality.' However, chronic bronchopleural fistula with associated empyema and scarring is a technically more formidable and clinically disabling problem. Some argue for the value of drainage alone and long-term conservative management, but definitive operative closure is usually preferable.l' Extensive experience with pedicled flaps of muscle and omentum has been reported; these methods are demonstrably effective, especially when definitive excision of the sinus tract back to normal bronchial tissue is feasible." In 1983, Anderson and Li 9 reported a case of postpneumonectomy bronchopleural fistula, in which they used the anterior transpericardial approach to the right main-stem bronchus. Noting difficulties with devascularization, scarring, and chronic sepsis, they? described the successful use of this approach in a young patient with bronchopleural fistula occurring after right 813
The Journal of
814 Baldwin and Mark
Thoracic and Cardiovascular Surgery
laryngeal nerve, and esophagus. The TA-30 stapler (with 4.8 rom staples) was then applied, flush with the take-off ofthe bronchus from the trachea-on the left in one case and on the right in the other two (Fig. 2). Two applications of the stapling device were employed, and the bronchus was not divided in these cases.
Patients and results
Fig. 1. Exposure via the standard median sternotomy. Retraction of the superior vena cava (SVC) to the right and the aorta (Ao) to the left provides the approach to the tracheal bifurcation. RPA, Right pulmonary artery.
pneumonectomy for trauma. This approach to the tracheal bifurcation was reported by Padhi and Lynn'? in their 1960 series of bronchopleural fistulas. They suggested ligation of the pulmonary artery and veins inside the pericardium and dissection of the bronchial stump up to the level of the carina." The technique has received considerable attention in Europe and particularly in the Soviet Union.":"
Operative technique All patients in this group underwent preoperative bronchoscopic examinations, principally to assess the length of the bronchial stump. We believe that a stump of at least 10 rom is necessary to permit secure transpericardial bronchial closure. Standard median sternotomy incisions were employed and the pericardium was opened in a stellate fashion (Fig. 1). Tapes were placed around the aorta and superior vena cava, and the vessels were retracted apart (Fig. 1, inset). A vertical incision was made in the posterior pericardium above the right pulmonary artery; dissection of lymphoid and areolar tissue anterior to the carina permitted exposure of both main-stem bronchi. Care must be taken to avoid injury to the right pulmonary artery, left recurrent
CASE 1. A 64-year-old man had a history of treated pulmonary tuberculosis. In 1982, he underwent left upper lobectomy at another hospital for an undiagnosed pulmonary nodule. He had a prolonged air leak postoperatively and underwent completion pneumonectomy; a bronchopleural fistula developed. He was then treated with drainage procedures and thoracoplasty, but the fistula and chronic septic drainage persisted. He was transferred to our care, and after optimizing his nutritional status and external drainage, we operated on July 18, 1983. At bronchoscopy, we noted a 5 em residual left main bronchial stump. In-continuity staple closure of the left main bronchus via the transpericardial approach was performed. The patient remains well, without evidence of recurrent fistula, 21 months after the operation. CASE 2. A 41-year-old man underwent right pneumonectomy in October of 1976 at another hospital for carcinoid adenoma of the bronchus intermedius. During his postoperative course he had fever and cough, with development of a bronchopleural fistula and empyema. He was treated by drainage followed by rib resection and primary closure. without success. In December of 1978, the patient was transferred to our hospital, and an open window thoracostomy was created. He did well subsequently with intermittent dilatations to ensure adequate drainage. The bronchopleural fistula remained open. On Dec. 5, 1983, the right main-stem bronchus was closed by the technique described herein. The patient remains well 17 months after the operation. CASE 3. A 33-year-old man had a gunshot wound to the right side of the chest, which necessitated right pneumonectomy in April of 1984. He had a complicated postoperative course, with reexploration for bleeding, cardiac arrest, adult respiratory distress syndrome, disseminated intravascular coagulation, sepsis, bronchopleural fistula, and pericardial effusion treated by pericardiectomy. Three months after the original injury he was transferred to our hospital, where open window thoracostomy drainage was performed. Sepsis subsided, and the patient's general condition improved. Two months later, he was readmitted for bronchial closure. At bronchoscopy, a I to 2 ern right main-stem bronchial stump was noted. Dense intrapericardial adhesions were noted from the patient's previous operation, and the esophagus was densely adherent to the posterior aspect of the distal trachea and proximal right main-stem bronchus. In-continuity staple closure of the right main-stem bronchus was performed as described herein. The patient did well postoperatively; bronchoscopy and thoracoscopy performed 2 months postoperatively revealed no evidence of fistula recurrence. However, approximately 2 weeks later, the patient noted a "Whistling" sound in the region of the thoracostomy and was found to have a recurrent fistula. He has subsequently undergone reclosure of the bronchus via a lateral thoracotomy, with placement of pectoralis major and latissimus dorsi myocutaneous pedicled flaps. He is currently clinically well 2 months after the operation.
Volume 90 Number 6 December, 1985
Postpneumonectomy bronchopleural fistula
8 15
Discussion The critical importance of prompt, adequate, and dependent drainage for postpneumonectomy bronchopleural fistula and empyema cannot be overstressed.' 15, J6 However, persistent patency of the fistula usually necessitates subsequent additional operative intervention. Some have placed emphasis on the potential role of thoracoplasty in reducing the size of the empyema space, but cosmetic and functional deformity, as well as potential for compromised pulmonary function, are drawbacks to this approach.' When the lateral, transthoracic approach to definitive closure is selected, it is important to excise completely the sinus tract, providing for closure in the region of normal, well-vascularized bronchial tissue. Widespread experience has been accumulated with respect to use of pedicled flaps to enhance security of closure in this setting, with many groups reporting high rates of success with single-stage closure.i'? Pairolero and associates7• 18 have reported impressive results with a variety of extrathoracic skeletal muscles transposed into the thorax for fistula closure. They have described the use of pedicles of latissimus dorsi, pectoralis major, serratus anterior, pectoralis minor, rectus abdominis, and combinations of these. These sometimes discouraging results and the formidable technical challenges associated with operating in areas of scarring, chronic infection, and devascularized bronchial tissue prompt consideration of alternative surgical strategies for treatment of postpneumonectomy bronchopleural fistula. The approach described here, as pointed out by Anderson and Li,9 has received attention in the European literature but has not been widely emphasized in this country. The advantages of the less morbid median sternotomy and the avoidance of areas of chronic sepsis and scarring in previous surgical fields are evident. Of course, pneumonectomy should be performed at the carinallevel, and this technique is applicable only in cases in which a residual bronchial stump is present. Preoperative radiographs may indicate approximate length, but we suggest that 10 mm of the residual bronchus should be documented by preoperative bronchoscopy. Division of the bronchial stump with placement of viable tissue between the closure lines is the established and preferable technique; moreover, division of the bronchus is necessary for healing to occur. However, short segments of available bronchus within the operative field may severely limit the possibility for secure closure of both ends of the denuded bronchus. In cases in which exposure is particularly limited, division may increase the risk of soilage of the operative field.
Fig. 2. Application of the stapling device to the left main bronchus, SVC, Superior vena cava. Ao, Aorta. RPA, Right pulmonary artery.
Strictly speaking, unless combined with excision of the distal segment, division does leave a persistent blind stump, possibly containing mucus-producing bronchial mucosa. Cognizance of the proximity of the esophagus posteriorly is essential during dissection, and esophageal injury has been reported with this technique." In cases in which previous intrapericardial dissection has been performed, the dissection may be more difficult, negating some advantages of the operation, as in the case in which a recurrent fistula developed. Ilakim and Milstein" have reported on the incidence of bronchopleural fistula after pneumonectomy, comparing results for patients in whom bronchial closure was accomplished with the parallel jaw stapler (TA55) with those in whom the newer hinged stapler (Premium TA-55) was u.-d. They found a higher incidence (15.2% versus 4.2%, p <, '"1.05) of bronchopleural fistula when the hinged device was used, with incomplete staple closure near the hinge end of the closure shown in radiographic studies in cadavers. This finding is noteworthy; the Premium, hinged device was used in this group, and the one patient in whom the fistula recurred did have leakage in the area of the "heel" or hinge on bronchoscopic examination. We suggest that either main-stem bronchus can be approached between the superior vena cava and aorta. The approach to the left main-stem bronchus to the left of the aorta and above the left atrial appendage requires
The Journal of Thoracic and Cardiovascular Surgery
816 Baldwin and Mark
displacement of the heart to the right and may not be well-tolerated hemodynamically. We have not found division of either pulmonary artery to be necessary or advantageous. REFERENCES Hankins JR, Miller JE, Attar S, Satterfield JR, McLaughlin JS: Bronchopleural fistula. Thirteen-year experience with 77 cases. J THORAC CARDIOVASC SURG 76:755-762,1978 2 Barker WL, Faber P, Ostermiller WE, Langston HT: Management of resistant bronchopleural fistulas. J THORAC CARDIOVASC SURG 62:393-401, 1971 3 Maksimov lA, Osmakov NA: Results of treatment of postoperative bronchopleural complications in lung cancer. Vestn Khir 114:63-68, 1975 4 Conlan AA, Boyd AD, Spencer FC: The operative management of acute postpneumonectomy bronchopleural fistula after flush bronchial amputation. S Afr Med J 61:792-794, 1982 5 Weissberg D: Empyema and bronchopleural fistula. Experience with open window thoracostomy. Chest 82:447-450, 1982 6 Van der Heyde MN, Verwers HR, Van Leusen R: Closure of a bronchial fistula after pneumonectomy in the care of carcinoma in a patient in chronic dialysis. Acta Chir Belg 77:271-274, 1978 7 Pairolero PC, Arnold PG: Bronchopleural fistula. Treatment by transposition of pectoralis major muscle. J THORAC CARDIOVASC SUR? 79:142-145, 1980 8 Virkkula L, Ferola S: Use of omental pedicle for the treatment of bronchial fistula after lower lobectomy. Report of two cases. Scand J Thorac Cardiovasc Surg 9:287-290, 1975 9 Anderson RP, Li W: Anterior transpericardial closure of a main bronchus fistula after pneumonectomy. Am J Surg 145:630-632, 1983 10 Padhi RK, Lynn FB: The management of bronchopleural fistulas. J THORAC CARDIOVASC SURG 39:385-393, 1960 11 Bogush LK, Semenenkow JL, Polianskii VA: Surgical treatment of bronchopleural complications after pneumonectomy. Khirurgiia (Mosk) 11:115-120, 1976 12 Slepukha 1M: Treatment of postoperative bronchopleural fistulas by transpericardical occlusion of the main bronchus using the L.K. Bogush transthoracic approach. Grudn Khir 6:85-88, 1976 13 Maasen W: The transstemal and transpericardial approach for surgical treatment of fistulas of the main bronchus after pneumonectomy. Thoraxchirurgie 23:257261, 1975 14 Bogush LK, Travin AA, Semenenkow JL: Transperikardiale Operationen an den Hauptbroachien and Langengefassen, Stuttgart, 1971, Hoppokrates-Verlag 15 Shamji FM, Ginsberg RJ, Cooper JD, Spratt EH, GoWberg M, Waters PF, lives R, Todd TR, Pearson FG: Open window thoracoscopy in the management of postpneumo-
16
17
18
19
nectomy empyema with or without bronchopleural fistula. J THORAC CARDIOVASC SURG 86:818-822, 1983 Goldstraw P: Treatment of postpneumonectomy empyema. The case for fenestration. Thorax 34:740-745, 1979 Miller JI, Mansour KA, Nahai F, Jurkiewicz MJ, Hatcher CR: Single stage complete muscle flap closure of the postpneumonectomy empyema space. A new method and possible solution to a disturbing complication. Ann Thorac Surg 38:227-231, 1984 Pairolero PC, Arnold PG, Piehler JM: Intrathoracic transposition of extra thoracic skeletal muscle. J THoRAe CARDIOVASC SURG 86:809-817, 1983 Hakim M, Milstein BB: Role of automatic staplers in the aetiology of bronchopleural fistula. Thorax 40:27-31, 1985
Discussion DR. HERMES C. GRILLO Boston, Mass.
This technique has been used more in Europe. Perelman and Ambatjello ' described its use transstemally and transpericardially for the left main bronchus in 1970 and Bogusch, Travin, and Semenenkow' for the right in 1971. While Padhi and Lynn) had suggested the approach in 1960, I believe Abruzzini' first described it in detail in 1961. Though he did not dwell on the division of the pericardium, he illustrated the basic technique extensively. We have done a small number of closures of the right main bronchus by this approach, including one in a patient who had had 5,000 rads, which made it more difficult but still feasible. I feel emphatically that in addition to stapling, one should divide the bronchus to avoid recurrence of the fistula. We use staples distally and suture the proximal end. We also make a point of interposing tissue between the bronchial ends. For this I found it convenient to use a collar incision, pedicle the sternohyoid muscle, and turn it down. It easily reaches a point between the two bronchial ends. Then there is no question of refistulization. For fistulas of the left main bronchus-and the authors have properly emphasized that on both sides one needs a long stump to use this technique-we have preferred a right thoracotomy in a fresh field. We divide the left main bronchus at its origin and interpose an intercostal muscle flap between the bronchial ends. The authors did not describe what they did after bronchial closure. In our cases, we have followed fistula repair in a few weeks with a Clagett procedure. This has been a successful way of handling the residual cavity. I shall close by mentioning that Professor Perelman,' in a recent communication, points out that he now believes it is better to approach all main bronchial fistulas by thoracotomy and reamputate the long stumps because he has had difficulty with recurrent empyema from the residual stump and from foreign material in the stump. We have not experienced this in our small series.
Volume 90 Number 6
Postpneumonectomy bronchopleural fistula
8 17
December, 1985
2
3 4
5
REFERENCES Perelman MI, Ambatjello GP: Transpleuraler, transtemaler und Kontralateraler Zugang bei Operationen Wegen Bronchialfistel nach Pneumonektomie. Thorac Chir Chir Vaskul 18:45, 1970 Bogush LK, Travin AA, Semenenkow JL: Transperikardiale Operationen an den Hauptbroachien and Langengefassen, Stuttgart, 1971, Hoppokrates-Verlag Padhi RK, Lynn FB: The management of bronchopleural fistulas. J THORAC CARDIOVASC SURG 39:385-393, 1960 Abruzzini P: Trattamento chirurgico delle fistule del bronco principale consecutive a pneumonectomia per tubercolosi. Chir Torac 14:165, 1961 Perelman MI: Late treatment of chronic bronchopleural fistula with long stump after pneumonectomy, International Trends in General Thoracic Surgery, Vol 2, H Eschapasse, H Grillo, eds., Philadelphia, W. B. Saunders Company (in press)
DR. MARK M. RA VITCH Pittsburgh. Pa.
I address myself to the question of in-continuity stapling of the bronchus. There is ample experimental evidence and clinical experience to show that in the gastrointestinal tract in-continuity stapling is not dependable. The cut bronchus heals not where the sutures or the staples are placed but at its cut end, and in these patients that cut end is a long way off, is avascular, and is surrounded by infection. Mucosa does not heal to mucosa. If a person kept his mouth closed for a year, the lips would not heal together. Finally, in about 1908, William Stewart Halsted reported that in dogs a ligature of the bronchus, could be removed again many months later and the bronchus restored to its normal patency. The approach, of course, is beautiful. Two thirds of the results are outstanding. DR. KAMAL A. MANSOUR Atlanta. Ga.
I believethis technique might have its place in early or small bronchopleural fistulas. However, I wonder if mere secondary closure of a stump that failed to heal is the proper approach and whether the factor or factors known or unknown that caused disruption in the first place are still active.
With few exceptions bronchopleural fistulas are usually associated with postpneumonectomy empyemas, and our aim therefore should not only be directed toward closure of the fistula but also to definite management of a large infected empyema space. With the advent of muscle flaps, the management of the notorious postpneumonectomy bronchopleural fistulas has been made much easier. I would like to ask the authors whether effort was made to assess the size of the fistula and also the time interval between onset of the fistula and reoperation. Most of us, I believe, have seen tiny fistulas that closed spontaneously in a few weeks. The cases presented today are few indeed and only time will tell. DR. RICHARD P. ANDERSON Seattle, Wash.
Main bronchial fistula after pneumonectomy is, fortunately, an unusual situation that is extremely difficult to manage successfully. I particularly enjoyed the authors' description of the transpericardial approach to the left main bronchus. I must agree with the other discussants, however, about the possible relationship of in-continuity stapling of the bronchus and the development of recurrent fistula. In chronic fistula the area where the bronchus penetrates the mediastinal wall is fixed and rigid scar and must place considerable tension on the staple line, especially on the right side. If at all possible, division of the bronchus is probably best. This also permits interposition of a viable muscle flap, as suggested by Dr. Grillo, to further promote healing in a scarred and devascularized area. DR. BALDWIN (Closing) I would like to thank Dr. Grillo for bringing his wisdom and experience to bear on this topic, and I appreciate the thoughtful comments made by the other discussants. We are aware that the most controversial aspect of these three cases is the technique of in-continuity closure of the bronchus. Certainly, the established and conservative technique involving division of the bronchus would be preferable when it is feasible technically and when it is possible to accomplish, without soilage of the mediastinum. Particularly in those cases in which the retained bronchial stump is especially long, division may be technically feasible.