|
|
||||||||
Ann Thorac Surg 2005;79:433-437
© 2005 The Society of Thoracic Surgeons
a Division of Thoracic Surgery, Toronto General Hospital, Toronto, Ontario, Canada
b Department of Biostatistics, Princess Margaret Hospital, Toronto, Ontario, Canada
Accepted for publication July 6, 2004.
* Address reprint requests to Dr Darling, Division of Thoracic Surgery, Toronto General Hospital, 200 Elizabeth St, Toronto, Ontario M5G 2C4, Canada (E-mail: gail.darling{at}uhn.on.ca).
| Abstract |
|---|
|
|
|---|
METHODS: We used a retrospective review of pneumonectomies performed during the period 1990 to 2000 and included a meta-analysis of relevant literature.
RESULTS: There were 187 pneumonectomies: 68 right, 119 left. The primary study end point was in-hospital death. There were 11 deaths: 7 (10.3%) right, 4 (3.3%) left (p = 0.10). Six deaths were attributable to bronchopleural fistula and its subsequent complications. The risk of bronchopleural fistula was higher on the right (9 [13.2%]) versus left (6 [5.0%]; p = 0.0472), as was the mortality associated with bronchopleural fistula (4 of 9 [44%] right versus 2 of 6 [33%] left). Right pneumonectomies were more likely to require an intrapericardial or extended dissection (p = 0.003), hand-sewn bronchial closure (p < 0.0001), or the closure buttressed (p < 0.0001). By univariate analysis, factors associated with an increased mortality were bronchopleural fistula (p < 0.0001), hand-sewn closure (p = 0.001), and a history of smoking (p = 0.01). By multivariate analysis, the most important factor was bronchopleural fistula (odds ratio, 43.3; 95% confidence limits, 4.2 to 441.9; p = 0.002). A meta-analysis combining our results with those from the literature found increased mortality of right pneumonectomy with a relative risk of 3.39 (95% confidence limits, 2.10 to 5.48; p < 0.00001).
CONCLUSIONS: Right pneumonectomy is associated with a higher mortality even in the absence of induction therapy. This is primarily related to the increased risk of bronchopleural fistula on the right. The increased number of bronchopleural fistulas on the right may be attributable to more extensive resection. Addressing technical factors that contribute to early bronchopleural fistula may reduce the mortality of right pneumonectomy.
| Introduction |
|---|
|
|
|---|
Advances in perioperative management have improved surgical outcome for pulmonary resections, but pneumonectomy continues to be associated with increased mortality compared with lobectomy [2]. It has been suggested that the mortality of a right pneumonectomy is higher than that of a left. Of the reports examining perioperative mortality after pneumonectomy, few have specifically analyzed or compared right and left pneumonectomy. The purpose of this study was to determine the operative morbidity and mortality associated with right-sided versus left-sided pneumonectomy.
| Material and Methods |
|---|
|
|
|---|
The data retrieved included age, sex, medical comorbidities, smoking history, preoperative pulmonary function tests, echocardiograms, ventilationperfusion scans, and resting and exercise pulse oximetry when available. If used, preoperative chemotherapy or radiotherapy was noted. The final pathologic stage was classified according to the 1997 international system for staging lung cancer [3]. Operative reports were reviewed in detail in an effort to determine the rationale for pneumonectomy versus lobectomy, intrapericardial or other extended dissection, the extent of lymph node dissection, hand-sewn versus stapled closure, and what, if any, tissue was used to reinforce the bronchial closure.
Intraoperative and postoperative complications occurring during hospitalization were noted, including empyema, pneumonia, arrhythmias, congestive heart failure, myocardial infarction, adult respiratory distress syndrome (ARDS), pulmonary embolus, sepsis, bronchopleural fistula (BPF), respiratory failure requiring intubation longer than 48 hours, and hemorrhage requiring transfusion or reoperation. All in-hospital deaths were considered operative deaths for the purpose of this analysis.
Meta-Analysis
In performing the meta-analysis, the criterion for inclusion was studies reporting results of pneumonectomy in which mortality data for right versus left were specified. Medline and Cancerlit were searched by the one of the authors (G.D.). Additionally, the references from other reports of pneumonectomy were hand searched. The literature search identified four studies that met the inclusion criterion. An additional paper reported a difference in mortality between left and right pneumonectomy, but actual patient numbers were not included in the paper (data obtained by personal communication, C. Deschamps, 2003). One study was excluded from the meta-analysis because it included only elderly patients [4]. Data from the four studies was combined with our own data set to complete the meta-analysis [5].
Statistical Analysis
Patient characteristics including preoperative and operative factors and postoperative outcomes were described as means or proportions. Patients with right-sided pneumonectomy were compared with patients with left pneumonectomy with respect to postoperative outcomes including in-hospital mortality. Differences in patient characteristics between these two groups were also examined. The
2 or Fisher's exact test was applied to discrete variables, and the Wilcoxon rank sum test was applied to continuous variables. Multivariate logistic regression was applied to find the adjusted associations of the side of operation with, and to identify independent risk factors for, in-hospital mortality and the development of a BPF.
| Results |
|---|
|
|
|---|
|
|
Mechanical ventilation was required in 8% (R, 11.8% versus L, 5.9%) of patients for respiratory failure. The cause of respiratory failure was multifactorial: pneumonia (4 patients), ARDS (2), congestive heart failure (1), atrial fibrillation without documented congestive heart failure (2), and BPF (6). When patients with ARDS were reviewed in detail, 1 patient was considered to have had postpneumonectomy pulmonary edema. One patient with a left pneumonectomy experienced ARDS as a complication of coagulopathy and died of multiorgan failure. Adult respiratory distress syndrome developed subsequently in 2 patients as a result of pneumonia and in an additional patient as a result of BPF.
With respect to nonrespiratory complications, myocardial infarction was more common with right-sided pneumonectomy (p = 0.017).
Mortality
Overall mortality was 6.4%, 10.3% (7 of 68) for R pneumonectomy and 3.3% (4 of 119) for L. Although this difference appeared important and supported our clinical impression and previous reports about the increased risk of R pneumonectomy, it did not reach statistical significance (p = 0.10, Fisher's exact test).
When our data were combined with data from the four studies identified in the literature in a meta-analysis [69], right pneumonectomy was clearly identified as a risk factor for death with a relative risk of 3.39 (95% confidence limits, 2.10 to 5.48; p < 0.00001; Fig 1).
|
|
Overall, the stage of disease, histology, and the extent of pneumonectomy did not influence mortality. Because of a recent report of the increased risk of complications and death in patients having induction therapy [8], the use of induction therapy with either chemotherapy, radiation, or both in 31 patients was examined but was not associated with increased mortality.
Factors identified by univariate analysis as having an association with increased mortality were analyzed by multivariate logistic regression. The factors identified by multivariate analysis to be important for risk of in-hospital death after pneumonectomy were as follows: BPF (odds ratio, 43.3; 95% confidence limits, 4.2 to 441.9; p = 0.002), and other postoperative complications excluding BPF (odds ratio, 8.7; 95% confidence limits, 0.9 to 82.1; p = 0.06).
Bronchopleural Fistula in Right Pneumonectomy
Comparing right and left pneumonectomy, right-sided resections were more frequently complicated by BPF, and the mortality of a right BPF was higher (R, 44% versus L, 33%), although not statistically significant. In an effort to explain the increased incidence of BPF in right pneumonectomies, possible contributing factors, including extent of lymphadenectomy, extent of resection, technique of bronchial closure, use of tissue reinforcement of the bronchial stump, and preoperative radiation or chemotherapy, were examined.
Complete lymphadenectomy was rarely performed. There was no significance difference in tumor stage between right-sided and left-sided resections (Table 1); however, there was a trend toward higher stage disease on the right (IIIa/b, IV), and more extensive resections, including significantly more intrapericardial pneumonectomies, were performed on the right (R, 45.6% versus L, 26.9%; p = 0.004). Patients having right pneumonectomy were more likely to have received induction therapy (R, 25.0% versus L, 13.4%; p = 0.07). The right bronchial stump was more likely to have been hand-sewn (R, 35.3% versus L, 8.4%; p < 0.0001) and buttressed (R, 75.0% versus L, 43.7%; p < 0.0001) rather than stapled. Local intrathoracic tissues were almost exclusively chosen for reinforcement of the bronchial stump. In patients experiencing a BPF, 3 of 6 left-sided stumps were not covered as compared with 2 of 10 on the right.
Variables that may have contributed to the development of BPF were analyzed by univariate analysis. (Table 4) The only statistically significant variables were right-sided resections and hand-sewn closures.
|
| Comment |
|---|
|
|
|---|
Martin and associates [8] reported the in-hospital mortality for patients having pneumonectomy was 13.0% for right-sided resections and 0% for left-sided resections, for an overall mortality of 6.2%. All patients had induction chemotherapy, and 18.1% of their patients also received preoperative radiation. In their series, the causes of death related to right pneumonectomy were respiratory failure and ARDS in 6 patients dying while in hospital. Primary respiratory failure and ARDS were not major causes of death in our series overall, nor in patients receiving neoadjuvant chemotherapy or radiation.
Combining the data from these reports [4, 6, 8, 9] with our own in a meta-analysis allowed us to statistically confirm the clinical impression of many thoracic surgeons that a right-sided pneumonectomy is indeed associated with increased risk.
Previous reports of mortality in pneumonectomy have focused on respiratory failure, ARDS, and postpneumonectomy pulmonary edema syndrome as causes of the increased mortality after pneumonectomy as compared with lobectomy. These same factors likely play a role in the increased mortality of a right compared with left pneumonectomy. It was postulated that the increased risk of right pneumonectomy may be related to the increased load on the right ventricle of the entire cardiac output going through the smaller left lung, resulting in increased pulmonary artery pressures, pulmonary hypertension, and right ventricular failure. Alternatively, the loss of the larger right lung might compromise pulmonary function, leading to respiratory failure. With these hypotheses in mind, poor pulmonary reserve, on the basis of preoperative pulmonary function testing, has been reported as a risk factor for increased mortality after pneumonectomy by some authors [6, 10, 1618] but not others [4, 19, 20]. We did not find that preoperative pulmonary function was a predictor of mortality in our logistic regression analyses, although we had preoperative pulmonary function data available on only a third of our patients. Nevertheless, only 8% of our patients developed postoperative respiratory insufficiency, which compares favorably with the incidence in the literature of 5% to 15% [21]. This may reflect the fact that surgeons in our center elected not to resect patients with borderline respiratory function.
Increased perioperative fluid resuscitation [7, 9, 15, 17, 22] as a cause of volume overload of the right ventricle and pulmonary circulation has also been reported as a predictor of poor outcome after pneumonectomy. In our centers, intraoperative and postoperative fluid restriction is the standard of care, and this may account for the low rate of postpneumonectomy pulmonary edema in this series.
This report implicates the development of a BPF as the major cause of mortality after right pneumonectomy, although BPF has been identified as an independent predictor of mortality after pneumonectomy by other authors [14, 23] and was more prevalent after right pneumonectomy (R, 9.8% versus L, 3.8%) [14]. The overall incidence of BPF in our series was 8%, which compares favorably with the incidence reported in the review by Cerfolio of 4.5% to 20% [24] and reported in most recent series (0.5% to 9.8%) [1, 14, 2528].
The increased frequency of BPF on the right side is likely multifactorial. It should be noted that the right bronchial stump is more exposed in the pleural space and less likely to be naturally buttressed by mediastinal tissues as compared with the left. This anatomic difference alone is likely a significant factor in the increased risk of developing a BPF on the right side.
Devascularization of the bronchial stump is also a recognized risk factor for BPF [24]. This may occur because of extensive dissection required for a proximal tumor or because of extensive lymphadenectomy. De Perrot and colleagues [1] noted an increase in postpneumonectomy BPF from 3% to 9% coincident with an increase in mediastinal lymphadenectomy and the use of bronchial staplers in their center. During the period studied in this series, lymph node sampling rather than complete lymphadenectomy was the standard approach, and therefore lymphadenectomy leading to devascularization of the bronchial stump is less likely to be a major contributing factor in our series. However, more extensive resections were performed on the right than the left, and this may have required a more extensive dissection around the main bronchus with potential devascularization of the bronchial stump. In our series, hand-sewn closure was identified as a risk factor for BPF whereas others have implicated the use of staplers in the development of BPF [2628]. Because we routinely use staplers for bronchial closure, the use of a hand-sewn closure reflects more extensive proximal dissection or a technically difficult bronchial stump. In closing a very proximal right bronchial stump, attention must be directed to ensuring that there is no tension on the bronchial closure. The cartilaginous ring at the origin of the right main bronchus has a tendency to keep the bronchus open. It has been suggested that a portion of this ring should be resected to allow the bronchus to be closed without tension.
In our series of 187 pneumonectomies combined with data from 4 other reports in a meta-analysis, we found that right-sided resections were associated with a higher mortality than those on the left. A higher rate of BPF on the right appeared to be a factor in the increased mortality of right pneumonectomy in our series. This may be related to anatomy, more extensive surgical dissection, potential devascularization, more intrapericardial dissections, and hand-sewn closures. Recognizing the increased vulnerability and risk of a right-sided bronchial stump and addressing technical factors that contribute to early BPF may help to reduce the morbidity and mortality of right pneumonectomy.
| Footnotes |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
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] |
||||
![]() |
M. de Perrot, K. McRae, M. Anraku, K. Karkouti, T. K. Waddell, A. F. Pierre, G. Darling, S. Keshavjee, and M. R. Johnston Risk Factors for Major Complications After Extrapleural Pneumonectomy for Malignant Pleural Mesothelioma Ann. Thorac. Surg., April 1, 2008; 85(4): 1206 - 1210. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. G. Sfyridis, E. I. Kapetanakis, N. E. Baltayiannis, N. V. Bolanos, D. S. Anagnostopoulos, A. Markogiannakis, and A. Chatzimichalis Bronchial Stump Buttressing With an Intercostal Muscle Flap in Diabetic Patients Ann. Thorac. Surg., September 1, 2007; 84(3): 967 - 971. [Abstract] [Full Text] [PDF] |
||||
![]() |
Z. Mansour, E. A. Kochetkova, X. Ducrocq, M.-D. Vasilescu, G. Maxant, A. Buggenhout, J.-M. Wihlm, and G. Massard Induction chemotherapy does not increase the operative risk of pneumonectomy! Eur. J. Cardiothorac. Surg., February 1, 2007; 31(2): 181 - 185. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. A. Clemson, E. Walser, A. Gill, J. E. Lynch, and J. B. Zwischenberger Transthoracic Closure of a Postpneumonectomy Bronchopleural Fistula With Coils and Cyanoacrylate Ann. Thorac. Surg., November 1, 2006; 82(5): 1924 - 1926. [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] |
||||
![]() |
M. Dancewicz, J. Kowalewski, and J. Peplinski Factors associated with perioperative complications after pneumonectomy for primary carcinoma of the lung Interactive CardioVascular and Thoracic Surgery, April 1, 2006; 5(2): 97 - 100. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Hendriks, P. Lauwers, and P. Van Schil Extrapericardial pneumonectomy MMCTS, June 28, 2005; 2005(0628): 83. [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 |