| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1997;63:1391-1396
© 1997 The Society of Thoracic Surgeons
Department of Thoracic Surgery, Pulmologisches Zentrum, Vienna, Austria
Accepted for publication November 12, 1996.
 |
Abstract |
|---|
 TopFootnotes Abstract IntroductionPatients and MethodsResultsCommentConclusionsReferences |
|---|
Methods. The cases of 96 patients with bronchopleural fistula after pneumonectomy seen over a 13-year period (1982 to 1995) were retrospectively analyzed. Cancer, TNM stage and histology, age, sex, side and size of the fistula at primary bronchoscopic diagnosis, time of occurrence after operation (days), cause of death, and survival after fistula formation (days) were analyzed. Management consisted of bronchoscopic closure with fibrin sealant or decalcified spongy calf bone or both, repeat thoracotomy with resection of the bronchial stump, thoracoplasty, or open window thoracostomy.
Results. Except for one instance, all total stump dehiscences occurred within 90 days after operation. Sixty-four patients (67%) died during the observation period; in 25, the cause of death was aspiration pneumonia. Only 2 patients who died of aspiration pneumonia had development of a fistula after 90 postoperative days. The aspiration rate dropped with increasing interval between operation and fistula occurrence (p = 0.000). Patient survival after fistula formation was positively correlated to this interval (p = 0.002). Successful fistula closure was achieved by surgical intervention in 21 patients and endoscopically in 11 patients. The overall postoperative mortality rate irrespective of treatment method was 31%.
Conclusions. The incidence of aspiration pneumonia declines sharply if bronchopleural fistula occurs more than 3 months after operation. Formation of fibrothorax apparently represents a natural protection against fistula formation and subsequent fatal aspiration pneumonia. Close follow-up during the first 3 postoperative months should detect bronchopleural fistula before aspiration occurs.
|
Introduction |
|---|
|
TopFootnotesAbstractIntroductionPatients and MethodsResultsCommentConclusionsReferences |
|---|
Bronchopleural fistula (BPF) after pneumonectomy represents the most dreaded complication in thoracic surgery. Nevertheless, diagnosis and therapeutic management remain major problems. Mortality rates range from 20% to 70% in the literature, the most common cause of death being aspiration pneumonia with subsequent adult respiratory distress syndrome. Most authors therapeutic approaches, and to date, no data have been reported about the incidence and the clinical course of aspiration pneumonia.
In a retrospective analysis of 96 patients, the natural history of BPF after pneumonectomy is examined. The objective of the study was to obtain data characterizing the natural history of this type of fistula, thus identifying risk factors and the best therapeutic management to prevent aspiration pneumonia.
|
Patients and Methods |
|---|
|
TopFootnotesAbstractIntroductionPatients and MethodsResultsCommentConclusionsReferences |
|---|
). Of these 797 patients, 96 (90 men and 6 women) (12%) had development of BPF. These 96 patients are the subject of this retrospective study.
|
For identification of risk factors, a group of 177 patients who underwent pneumonectomy from 1992 to 1995 and did not have development of BPF was compared with the study group. In both groups, tangential resection with automatic stapling devices was performed. No routine mechanical ventilation was employed postoperatively.
Fistulas larger than 5 mm were treated with total thoracoplasty including resection of the first rib or with open window thoracostomy in the case of multiple-organ failure resulting from sepsis. Repeat thoracotomy with resection and resuturing of the bronchial stump followed by coverage with a muscular flap was considered only for patients seen with early empyema. Small fistulas (<5 mm) underwent endoscopic treatment with fibrin sealant or decalcified spongy calf bone. If bronchoscopic treatment failed, total thoracoplasty or open window thoracostomy was performed, depending on the patient's general condition.
All data were entered in a personal computer (Compaq PC, IBM compatible, Windows 95, Pentium) and evaluated by a statistical program (SPSS for Windows, version 6.1). Evaluation of data was performed by crosstabulation and correlation. The applied significance level was 0.05.
|
Results |
|---|
|
TopFootnotesAbstractIntroductionPatients and MethodsResultsCommentConclusionsReferences |
|---|
, see Table 1
). Male sex was predominant in the fistula group (male to female ratio, 15:1). Thus, right side and male sex were significant risk factors (p = 0.000). Age showed no significant relevance for fistula formation. In the fistula group, pneumonectomy had been performed because of neoplastic disease in 93 patients (97%), destroyed lung in 2 (2%), and pulmonary atresia in 1 patient (1%).
|
). Therefore, TNM stage was not identified as a risk factor for postoperative fistula development. Nor did histology show any correlation (Table 4). Fistula size at the time of diagnosis was documented in 87 patients. Group 1 included 28 patients (32%); group 2, 26 patients (30%); and group 3, 33 patients (38%). Fistula size was not documented for 9 patients.
|
|
The interval between operation and fistula formation ranged from 2 to 1,199 days (mean ± standard deviation, 88 ± 178 days) and survival after fistula formation, from 1 day to 2,208 days (258 ± 465 days) (Fig 1). Seventy-six fistulas (79%) developed within 90 days after operation, 28 of them during the first 2 postoperative weeks before discharge from the hospital. Twenty fistulas (21%) occurred between postoperative days 90 and 1,199. Of these 20 patients, only 1 patient died of aspiration pneumonia. Except for 2 patients, all patients who died of aspiration pneumonia had development of the BPF within 90 days after operation (interval, 2 to 82 days; mean time, 28.2 days). Patient survival ranged from 1 day to 78 days (mean survival, 21.5 days) even though aggressive treatment had been given (Fig 2). The mortality rate was highest during the first 2 postoperative weeks (29%). Forty-eight fistulas (50%) developed between postoperative days 15 and 90. The mortality rate was still high in this subset (35%).
|
|
). Sixty-four patients (67%) died during the observation period. Forty-two died within 90 days after operation. Death was caused by aspiration pneumonia in 25 patients, cardiac failure in 7, empyema with sepsis in 4, and local or distant tumor recurrence in 19. One patient died of pulmonary embolism, 1 of cerebral hemorrhage, and 2 of cachexia. No postmortem examination was done on 5 outpatients. Only 1 patient (1%) in whom the BPF developed after 90 postoperative days died of aspiration pneumonia (Table 5, Fig 4).
|
|
|
Eleven patients (11%) were treated with thoracic drainage only. In 1 patient, the BPF closed without reoperation. Seven patients died before any further surgical treatment could be started, and 3 patients had local tumor recurrence that led to fistula blockage.
Forty-nine fistulas (51%) were treated surgically. Twenty-one of the patients underwent total thoracoplasty. The fistula was larger than 5 mm in 18, and the remaining 3 patients had small fistulas and advanced empyema at admission. Nine patients died in the early postoperative course. Causes of death were aspiration pneumonia in 5, cardiac failure in 2, and pulmonary embolism and progressive sepsis originating from the empyema in 1 each. The remaining 12 patients having thoracoplasty were cured.
Nine patients underwent repeat thoracotomy. Three died of aspiration pneumonia, and 1 died of cardiac failure. Successful fistula closure was achieved in 3 patients. One patient left the hospital with chronic empyema and permanent drainage (ie, a 50-mL drainage bottle without a water seal; the bottle is changed daily and can be worn under clothing, and the chest tube is changed every 3 weeks). Fistula closure failed in 2 patients, who had successful treatment by total thoracoplasty.
In 19 patients, open window thoracostomy was performed as an emergency procedure. Five died of pneumonia and 2 of cardiac failure. The condition of 3 patients improved after operation, and they subsequently underwent successful thoracoplasty. Nine patients were discharged with an open window thoracostomy. During the observation period, three thoracic windows closed spontaneously, and 3 patients had substantial reduction in window size.
Thirty-six patients (38%) were treated bronchoscopically. Fibrin sealant and decalcified spongy calf bone [13] were used to achieve fistula closure. Five patients died in the hospital of aspiration pneumonia. Fistula closure was successful in 11 patients, but only 4 had healing of the concomitant empyema. Thus, only 4 patients left the hospital without drainage; the remaining 7 patients were discharged with chronic empyema and permanent drainage. In 8 patients, permanent fistula closure failed, and they were also discharged with chronic empyema and permanent drainage. In 4 patients, open window thoracostomy became necessary because of progressive sepsis originating from the empyema; they survived. Eight patients required thoracoplasty; 2 died postoperatively of sepsis and empyema, and 6 left the hospital without drainage.
The overall postoperative mortality rate regardless of treatment method was 31%. After successful fistula closure and definitive clearance of empyema, no recurrence was observed.
|
Comment |
|---|
|
TopFootnotesAbstractIntroductionPatients and MethodsResultsCommentConclusionsReferences |
|---|
Various groups [3–7] have covered the subject of adequate therapy for bronchial stump dehiscence and fistula formation. All agree on the importance of early chest tube drainage to prevent aspiration. If a patient is seen with aspiration pneumonia, even immediate operation cannot save his or her life. This emphasizes the importance of early fistula diagnosis and treatment. Sepsis originating from the empyema was the cause of early postoperative death in only 3 patients. Thus, even advanced empyema can be successfully controlled by early drainage and operation. The literature dealing with fibrin sealing of BPF is sparse. Except for reports about fistulas after smaller lung resections [8, 9], only a small number of patients with fistulas after pneumonectomy have been studied [9–11]. All the fistulas were small, and most were successfully closed. Becker [12] reported successful closure in 35 of 45 patients. However, no distinction was made between patients who had pneumonectomy and those who had lesser resections. All groups considered endoscopic sealing a worthwhile alternative to surgical treatment, as there is a good chance to avoid thoracotomy.
We achieved endoscopic fistula closure in 11 of 36 patients. However, fistula closure does not necessarily mean eradication of empyema. The 4 patients we reported as free from fistula and empyema were discharged and are still alive. Although when first seen, they had positive bacteriologic cultures in the chest tube drainage, intrathoracic infection was still at an early phase and responded well to pleural antibiotic irrigation after successful fistula closure. Therefore, endoscopic treatment with fibrin sealant or spongy bone may result in complete cure only in early empyema.
Persistent empyema after fistula closure should not be regarded as a treatment failure, as the patient is saved from life-threatening aspiration. Even if the fistula cannot be closed permanently, temporary fistula closure allows irrigation of the pleural space with antibiotics without danger of aspiration. Reduction of drainage volume as a sign of empyema control allows discharge with permanent drainage. This method seems preferable to open window thoracostomy or thoracoplasty, which are the only alternatives in many instances of advanced empyema. Progression of empyema during treatment makes surgical intervention (ie, thoracoplasty or open window thoracostomy) inevitable.
Our overall perioperative mortality rate of 31% is in accordance with the literature, which reports mortality rates of 19.5% [4], 23% [5], and up to 71% [13], which includes fistulas after smaller lung resections such as lobectomy, bilobectomy, and segmentectomy. Hankins and associates [7] reported a mortality rate of 46% in a subset of patients with postpneumonectomy fistula. One of the most important factors influencing mortality is the interval to postoperative fistula formation. Incidence reaches its peak in the second and third postoperative weeks and declines rapidly later. Many patients in whom BPF develops represent a serious problem because they simply do not get a diagnosis. They may have to travel long distances for treatment. If a fistula develops after discharge, it is very likely the patient will not see a surgeon early enough and will die of aspiration pneumonia at home. Another problem is the difficulty of diagnosing BPF. Many physicians are not familiar with its clinical symptomatology, and the danger of aspiration pneumonia is underestimated; aspiration is often misdiagnosed as pulmonary edema, and diuretics are applied. Thus, early drainage as prophylaxis for aspiration is missed.
In some patients fistulas develop 5 to 10 years after operation. Because postoperative follow-up has concluded, the patient often does not get adequate treatment. To date, only very few risk factors have been reported. As mentioned by Asamura and associates [13], the right side is a significant risk factor for fistula formation, which is in agreement with our data. Fistulas develop significantly more often in men than in women. The pathogenetic mechanism remains unclear.
The influence of TNM stage on fistula formation could not be verified from our data unlike data of others [13]. Only malignant invasion of the bronchial stump itself promotes fistula formation [13].
Aspiration pneumonia was the main cause of death early in the disease. Only 2 of 25 patients with aspiration pneumonia early survived. There is a strong correlation between the interval from operation to fistula formation and survival (p = 0.002) during the first 3 months after operation. The high mortality rate in the early postoperative period may be explained by postoperative ventilation/perfusion mismatch and mucous congestion caused by the procedure and anesthesia; this exposes the contralateral lung to a high risk of aspiration damage.
In many patients, fistulas are at an initial stage when diagnostic bronchoscopy is performed and often enlarge before definitive treatment is started. Patients with late fistula formation often show a poor clinical symptomatology and are not initially referred to a thoracic surgeon; therefore, early fistulas are usually diagnosed more easily than late ones. However, size at primary bronchoscopy influences the therapeutic management of many thoracic surgeons and pulmonologists, as it is the only evaluable information about the fistula itself. The earlier the fistula forms, the bigger the leak is. The finding that only one total bronchial stump dehiscence occurred more than 3 months after operation suggests that formation of fibrothorax, which should be sufficiently established after this period, helps to stabilize the bronchial stump. Even if BPF occurs, the danger of fatal aspiration is virtually excluded by the many small compartments that have formed in the empty hemithorax. These are filled with smaller amounts of fluid that can be coughed up easily, even if they are aspirated.
The hypothesis that total stump dehiscence bears a higher risk of massive aspiration than minifistulas could not be verified statistically. Fistula size and aspiration rate did not correlate. In the early postoperative period, even minifistulas can cause silent aspiration with subsequent pneumonia and are as dangerous as total stump dehiscence, although air leakage is much smaller. The risk of aspiration virtually disappears 3 months after operation. Before that, aspiration occurs independently of fistula size and is responsible for a high mortality, no matter which treatment is chosen. Even aggressive intensive care therapy does not improve this discouraging outcome.
Aspiration more than 3 months after pneumonectomy is better tolerated because patients have almost regained their normal coughing and muscular capability, and this allows them to remove small amounts of aspirated fluid more easily. We observed 1 patient with a documented fistula of 2 mm who refused any therapeutic intervention over 5 years and is still alive.
Thus, the interval between operation and onset of BPF appears to be the most important factor influencing clinical course and outcome. Patients who undergo pneumonectomy and are free from fistula formation for 3 months have escaped death caused by aspiration.
Although BPF occurs more frequently on the right side, we could not find any difference in its natural history that could be attributed to the affected side. Right- and left-sided fistulas showed exactly the same patterns in size, survival, interval between operation and their formation, and aspiration. Formation of fibrothorax seemingly develops on both sides with comparable progress, although the volume of the left hemithorax is smaller. Whether the bigger size of the right hemithorax represents one of the natural disadvantages promoting fistula formation remains unresolved.
|
Conclusions |
|---|
|
TopFootnotesAbstractIntroductionPatients and MethodsResultsCommentConclusionsReferences |
|---|
The major goal of therapeutic intervention must be the prevention of aspiration pneumonia, as it is the main cause of death in patients with postpneumonectomy BPF. This goal can easily be accomplished by chest tube drainage. Repeat thoracotomy or endoscopic intervention should be performed as soon as possible in early empyema. If advanced empyema is present at admission, such as in late fistulas, disinfection of the pleural space remains the primary approach. Fistula closure can then be achieved simultaneously at bronchoscopy or later by surgical intervention. Progressive sepsis originating from the empyema during conservative or endoscopic treatment can be successfully controlled with open window thoracostomy.
With special regard to prognosis and postoperative time, we propose the following classification of BPF:
|
Footnotes |
|---|
|
TopFootnotesAbstractIntroductionPatients and MethodsResultsCommentConclusionsReferences |
|---|
|
References |
|---|
|
TopFootnotesAbstractIntroductionPatients and MethodsResultsCommentConclusionsReferences |
|---|
Related Article
This article has been cited by other articles:
|
|
 |
|
  D. Schneiter, T. Grodzki, D. Lardinois, P. B. Kestenholz, J. Wojcik, B. Kubisa, J. Pierog, and W. Weder Accelerated treatment of postpneumonectomy empyema: A binational long-term study J. Thorac. Cardiovasc. Surg., July 1, 2008; 136(1): 179 - 185. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. A. Kesler, Z. T. Hammoud, K. M. Rieger, L. E. Kruter, M. Yu, and J. W. Brown Carinaplasty Airway Closure: A Technique for Right Pneumonectomy Ann. Thorac. Surg., April 1, 2008; 85(4): 1178 - 1186. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. West, A. Togo, and A. J.B. Kirk Are bronchoscopic approaches to post-pneumonectomy bronchopleural fistula an effective alternative to repeat thoracotomy? Interactive CardioVascular and Thoracic Surgery, August 1, 2007; 6(4): 547 - 550. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Chichevatov, A. Gorshenev, and E. Sinev Preventive diaphragm plasty after pneumonectomy on account of lung cancer. Asian Cardiovasc Thorac Ann, August 1, 2006; 14(4): 265 - 272. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Tao, M. Araki, T. Sato, S. Morino, R. Kawanami, M. Yoshitani, and T. Nakamura Bronchoscopic treatment of postpneumonectomy bronchopleural fistula with a collagen screw plug J. Thorac. Cardiovasc. Surg., July 1, 2006; 132(1): 99 - 104. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Feller-Kopman, R. Bechara, R. Garland, A. Ernst, and S. Ashiku Use of a removable endobronchial valve for the treatment of bronchopleural fistula. Chest, July 1, 2006; 130(1): 273 - 275. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Chichevatov and A. Gorshenev Omentoplasty in Treatment of Early Bronchopleural Fistulas after Pneumonectomy Asian Cardiovasc Thorac Ann, September 1, 2005; 13(3): 211 - 216. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. P. Madden, A. Sheth, T. B.L. Ho, G. R. McAnulty, and R. E. Sayer A Novel Approach to the Management of Persistent Postpneumonectomy Bronchopleural Fistula Ann. Thorac. Surg., June 1, 2005; 79(6): 2128 - 2130. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Takaoka, S. Inoue, and S. Ohira Central Bronchopleural Fistulas Closed by Bronchoscopic Injection of Absolute Ethanol* Chest, July 1, 2002; 122(1): 374 - 378. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Kiriyama, Y. Fujii, Y. Yamakawa, I. Fukai, M. Yano, M. Kaji, and H. Sasaki Endobronchial neodymium:yttrium-aluminum garnet laser for noninvasive closure of small proximal bronchopleural fistula after lung resection Ann. Thorac. Surg., March 1, 2002; 73(3): 945 - 948. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Schneiter, P. Cassina, S. Korom, I. Inci, M. Al-Abdullatief, A. Dutly, P. Kestenholz, and W. Weder Accelerated treatment for early and late postpneumonectomy empyema Ann. Thorac. Surg., November 1, 2001; 72(5): 1668 - 1672. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. H Khan, S. B Rahman, D. B McElhinney, A. L Harmon, J. P Anthony, T. S Hall, D. M Jablons, J. H Khan, S. B Rahman, D. B McElhinney, et al. Management Strategies for Complex Bronchopleural Fistula Asian Cardiovasc Thorac Ann, March 1, 2000; 8(1): 78 - 84. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. S. Topcuoglu, C. Kayhan, and T. Ulus Transsternal transpericardial approach for the repair of bronchopleural fistula with empyema Ann. Thorac. Surg., February 1, 2000; 69(2): 394 - 397. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. de Perrot and A. Spiliopoulos Postpneumonectomy bronchopleural fistula Ann. Thorac. Surg., November 1, 1999; 68(5): 1886 - 1887. [Full Text] [PDF]
|
|
|
|
 |
|
 P. H. Hollaus, F. Lax, P. N. Wurnig, D. Janakiev, and N. S. Pridun Videothoracoscopic debridement of the postpneumonectomy space in empyema Eur. J. Cardiothorac. Surg., September 1, 1999; 16(3): 283 - 286. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Klepetko, S. Taghavi, A. Pereszlenyi, T. Birsan, J. Groetzner, N. Kupilik, O. Artemiou, and E. Wolner Impact of different coverage techniques on incidence of postpneumonectomy stump fistula Eur. J. Cardiothorac. Surg., June 1, 1999; 15(6): 758 - 763. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
U. Acikel, E. Silistreli, N. Ozelsancak, O. Karabay, E. S. Ucan, E. Hazan, and O. Oto Postpneumonectomy Bronchopleural Fistula Formation and Surgical Management Asian Cardiovasc Thorac Ann, March 1, 1999; 7(1): 49 - 51. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. H. Hollaus, F. Lax, P. N. Wurnig, and N. S. Pridun VIDEOTHORACOSCOPIC TREATMENT OF POSTPNEUMONECTOMY EMPYEMA J. Thorac. Cardiovasc. Surg., February 1, 1999; 117(2): 397 - 398. [Full Text] [PDF]
|
|
|
|
|
|
A. El-Gamel, G.M.K. Tsang, and D.C.T. Watson The threshold for air leak: stapled versus sutured human bronchi, an experimental study Eur. J. Cardiothorac. Surg., January 1, 1999; 15(1): 7 - 10. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. H. Hollaus, F. Lax, D. Janakiev, P. Lucciarini, E. Katz, A. Kreuzer, and N. S. Pridun Endoscopic treatment of postoperative bronchopleural fistula: experience with 45 cases Ann. Thorac. Surg., September 1, 1998; 66(3): 923 - 927. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
