| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
Interact CardioVasc Thorac Surg 2009;9:571-575. doi:10.1510/icvts.2009.203646 © 2009 European Association of Cardio-Thoracic Surgery
Low incidence of bronchopleural fistula after pneumonectomy for lung cancer
a Department of Cardiothoracic Surgery, Patras University School of Medicine, Patras 26500, Greece Received 23 January 2009; received in revised form 18 June 2009; accepted 22 June 2009
*Corresponding author. 31, Chlois Str., Voula, 166 73, Athens, Greece. Tel.: +30 210 8955570; fax: +30 210 8955570.
Bronchopleural fistula (BPF) after pneumonectomy for NSCLC remains a highly morbid complication. We examined possible factors including the surgical techniques associated with BPF development. From 221 pneumonectomies for NSCLC, bronchial stump closure was mechanically performed in 192 patients and manually in the remaining 29. In all right-sided pneumonectomies mechanical closure was performed with associated stump coverage. In 114/130 left-sided procedures where mechanical closure was selected, bronchial stump remained uncovered. In the remaining 16 left-sided cases where manual stump closure was selectively performed, the stump was covered utilizing various tissues. Risk factors were classified into preoperative, intra-operative and postoperative. Five patients (2.3%) developed BPF. Univariate analysis revealed peri-operative transfusion, respiratory infection at the time of presentation, neoadjuvant therapy, right-sided pneumonectomy, manual type of bronchial closure, days of postoperative hospitalization and mechanical ventilation as significant risk factors for BPF development. Multivariate analysis followed revealing preoperative respiratory infection and right pneumonectomy as the only independent risk factors. In our series, a selected stump coverage policy showed a low incidence of BPF development. Mechanical stapling was superior to manual closure, although not as an independent factor. Early recognition of possible risk factors associated with fistula development is of paramount importance.
Key Words: Bronchopleural fistula; Pneumonectomy; Lung cancer; Bronchial stump
Bronchopleural fistula (BPF) development after pneumonectomy remains a serious complication, despite improvements made over the last decades in surgery for non-small cell lung carcinoma (NSCLC) [1]. Since the introduction of automatic stapler devices [2], the manual closure of the bronchial stump as proposed by Sweet [3] was disputed. However, controversy still exists [4]. In this study, we examined the possible factors associated with increased incidence of BPF formation following pneumonectomy for NSCLC and analyzed the impact of different techniques.
From 1999 to 2005, 243 patients underwent pneumonectomy for NSCLC. Three patients with sleeve pneumonectomies and 19 with incomplete data were excluded from further analysis. All patients underwent complete preoperative assessment. All procedures were performed through a lateral thoracotomy with double lumen endotracheal tube intubation (98%) and thoracic spinal epidural analgesia. Fourteen patients (6.3%) underwent preoperative chemo- and/or radiotherapy. From the total of 221 pneumonectomies, 43 were performed intra-pericardially, 21 extrapleurally whereas three underwent complementary pneumonectomy. Bronchial stump closure was performed mechanically in 192 patients with a commercial stapler device TA-30 (4.8 mm) and manually in the remaining 29. In cases of manual closure, an atraumatic clamp was placed distal to the line of division to prevent spoilage into operative field from any contaminated material. Stay-sutures were placed on each lateral side of the bronchus just proximal to the selected line of excision and the bronchus was sharply divided. Bleeding on the dissected margin was managed by individual ligation or clip application of bronchial arteries. Electrocoagulation by diathermy of the bleeding vessels was avoided. The posterior membranous wall was approximated to the anterior cartilaginous wall with interrupted Prolene 3-0. In cases of centrally located tumors, resection of the bronchial stump was performed >2 cm away from the primary tumor margins, and margin free of disease was histologically certified intra-operatively. After submersion of the bronchial stump into physiologic saline any possible air leaks were identified and secured with additional interrupted sutures. Coverage of the bronchial stump was routinely curried out in 91 cases of right pneumonectomy; regardless the type of closure. In 83 patients, coverage of the right bronchial stump was performed by mobilization of the azygos vein and the surrounding parietal pleura. In three patients with simultaneous resection of a part of the thoracic wall, coverage of the stump was performed by interposition of an intercostal muscle flap. In five cases of right intra-pericardial pneumonectomy, a pericardial flap was used in a cap-like manner. On the left side, in 114 out of 130 mechanical closure pneumonectomies no further coverage of the bronchial stump was performed as protocol. Only in the remaining 16 cases where stump closure was manually performed, we proceeded with additional coverage. In nine out of 16 cases, intercostal muscle interposition was used, while in the remaining seven cases a pericardial flap was utilized. Postoperatively, all patients were monitored according to our protocol as we have previously described [5]. Upon suspicion of BPF, a thoracic CT-scan and fiberoptic bronchoscopy were performed. Pleural fluid and sputum collections were cultured. Closed chest tube drainage (CTD) was performed in all patients and was followed by reoperation immediately after the BPF recognition if it occurred at the early postoperative period or at a later date in the late cases, after the pleural cavity has been sterilized. Extensive debridement of the pleural cavity and exploration of the BPF was performed. The fistula was covered with well vascularized tissue. A chest wall window was created after resection of two segments of adjacent ribs and a 36 Fr chest tube was inserted, converting it to an open tube thoracostomy (OTT). Alternatively an Eloeser flap was performed. Before hospital release fiberoptic bronchoscopy was repeated in order to ensure complete bronchial closure. Thereafter, this chest drain was changed once a month. Variables were retrospectively recorded. Nominal categorical data were compared using either 2 or Fisher's exact test. Mean survivals were calculated with the Kaplan–Meier method. For univariate survival comparison the Breslow (generalized Wilcoxon) test were used. Kolmogorov– Smirnov test was used to detect deviations from normality. Where appropriate, mean value comparisons were performed using the Mann–Whitney test. Multivariate logistic regression tests were followed in order to identify the independent significant risk factors for BPF development. The tests included subset selection using both the forward and backward elimination procedure. Significance level was set at 0.05. The odds ratios from the final model were calculated with exact methods for logistic regression using STATA software. All other tests were performed using the SPSS v.14.
Two hundred and twenty-one consecutive patients (M/F=199/22) with mean age 62.4±9.0 years were enrolled. Demographic data, histology and stage are presented in Table 1. One hundred and thirty pneumonectomies were performed on the left-side and 91 on the right. The 30-day mortality and morbidity for the whole cohort are shown in Table 2. Five patients developed BPF (Fig. 1). All patients with BPF eventually developed postpneumonectomy empyema. After initial CTD, all BPF patients were transferred to the operating room for debridement and fistula coverage. This was achieved by intra-thoracic transposition of a muscular flap from serratus anterior in three cases and in the remaining two cases by mobilization of a well vascularized intercostal muscle flap. Chest wall window and OTT was performed in all patients. Among the fistula group, one patient died in the intensive care unit (ICU) 20 days after reoperation due to MOF (Table 3).
The clinical and surgical characteristics of patients who developed BPF are shown in Table 3. Parameters were classified into preoperative (Table 4); intra-operative (Table 5) and postoperative (Table 6).
Univariate analysis regarding the preoperative parameters showed that peri-operative transfusion (P=0.012), respiratory infection at the time of presentation (P<0.001) and neoadjuvant chemo- or radiotherapy (P=0.002) were related to the development of BPF. The right-sided pneumonectomy (P=0.011) and the manual type of bronchial closure (P=0.017) were identified as intra-operative factors related to BPF formation. Finally, the postoperative parameters showing statistical significance were the days of postoperative hospital stay (P=0.030) and the development of respiratory failure requiring mechanical ventilation (P=0.029). Mean survival, as anticipated, was statistically in favor of the non-BPF group (P=0.010). Multivariate analysis revealed preoperative respiratory infection and right pneumonectomy as the only independent risk factors for postpneumonectomy BPF formation (Table 7).
The incidence of BPF after pneumonectomy varies from 0.8% to 12% [4]. Local and systemic factors have been proposed in the development of postpneumonectomy BPF [6]. Local factors include carcinoma at the bronchial-stump margin, long bronchial stump, disrupted bronchial blood supply, inadequate technique of stump closure, presence of empyema, extended resection, and preoperative radiation [1, 4, 6]. The most important factor seems to be ischemia at the margin site due to excessive devascularization [7]. The main bronchus and its major segmental branches receive their blood supply from the bronchial arteries. Two or three major bronchial arteries pass in the peribronchial tissues of the main bronchus and branch into the major segmental peribronchial tissues. Each gives off smaller arterial branches which supply the adjacent bronchus but there is a poor network of vessels within the bronchial wall. The net result is a rather poorly vascularized tissue. Surgical trauma to the bronchus as a result of incision and surgical repair often leads to areas of avascular necrosis and consequent disruption of the suture line [7]. Systemic factors include the patient's nutritional status, diabetes, presence of sepsis, preoperative chemotherapy, and presence of underlying lung diseases [1]. Extended resection, mechanical ventilation, right versus left pneumonectomy and surgeon level of experience have also been identified as possible contributing factors [8]. In our series the BPF rate was 2.3% and the resulting mortality 20% (1/5). For univariate analysis we classified the possible contributing factors, into preoperative, intraoperative and postoperative. Preoperatively, patients with respiratory infection at the time of admission or neoadjuvant chemo- or radiotherapy were associated with increased incidence of BPF. Significant intra-operative factors were right-sided pneumonectomy and manual closure. Postoperative parameters showing statistical significance were the days of postoperative hospital stay and the development of respiratory failure requiring mechanical ventilation in the ICU. Finally, perioperative transfusions were also found to be significant. Following both forward and backward elimination procedure, multivariate analysis showed that right pneumonectomy and preoperative infection upon admission were independent risk factors for BPF development. The suggestions proposed 60 years ago by Sweet [3], for manual bronchial closure including minimizing bronchial trauma, preservation of blood supply, careful approximation of the cut edges at the bronchial stump and stump reinforcement with adequate well vascularized flap, are still valid. In our series, mobilization of the main bronchus from surrounding tissues was performed with minimal dissection, minimal use of the electrocautery, and respect to the bronchial arteries. Variability still exists upon which type of bronchial closure, mechanical vs. manual, should be used. Some surgeons demonstrated a low percentage of fistula development by manual closure [1, 8, 9], while others are strong supporters of mechanical staplers [2, 10]. Our results are in favor of using staplers, since we noticed only a 0.9% incidence of bronchial leak in those patients. The benefit of coverage of the bronchial stump in preventing BPF is still controversial [11]. Controversy also exists among the different techniques followed and the origin of various tissues used for covering the bronchial stump. Many suggest bronchial stump reinforcement only in patients associated with possible risk factors for BPF formation [8, 12]. We among others [1, 8] prefer to cover all stumps in right pneumonectomies. On left-sided pneumonectomies we normally leave the stump uncovered, since it is covered by the aorta and the surrounding tissues, except for a few cases of high-risk patients, where closure of the bronchus was performed manually. However, others advocate the need for coverage on the left side as well [4, 8, 9]. The lack of BPF presentation on the left side in our series supports our policy. For those patients with BPF, a great variety of tissues has been used for bronchial stump reinforcement using both intra- or extra-thoracic tissues. Parietal pleura, pericardium, pericardio-phrenic pedicle and intercostal muscles are the most common intra-thoracic grafts utilized for stump coverage [4, 8, 13], while the serratus anterior, the latissimus dorsi and omental flaps are the most frequently extra-thoracic tissues used [8, 14]. A high-suspicion index is necessary for prompt BPF diagnosis in order to reduce its high mortality. Early evacuation of the thoracic cavity is the first step in management. CTD is mandatory in order to prevent aspiration pneumonia on the contralateral side. Further surgical repair depends upon the patient's clinical status and the size of fistula. Various procedures have been described. We prefer performing the OTT procedure as soon as the diagnosis is established. After debridement and inspection of the lacerated stump manual suturing of the stump follows. The posterior fibers of serratus anterior muscle or a well-vascularized intercostal muscle flap from the 6th or 7th intercostal space are used. We avoid using flaps mobilizing the latissimus dorsi muscle or omental pedicles since such techniques require an extended surgical procedure associated with morbidity [15]. In our series we noticed one of the lowest incidences of BPF development after pneumonectomy for NSCLC. Hand suturing did not appear to be an independent factor in multivariate analysis, which revealed only preoperative infection and right-sided pneumonectomy. Mechanical stapling was superior to manual suture. Most surgeons agree that it is prudent to buttress bronchi at high risk of developing a BPF with viable tissue after lobectomy or pneumonectomy. Bronchial coverage was performed in all of our right pneumonectomy procedures. Although the majority of left-sided pneumonectomies (87.7%) remain uncovered, except for those performed manually, there was no incidence of BPF formation. Our preference is the utilization of well vascularized intra-thoracic tissues such as parietal pleura with azygos mobilization, pericardium or intercostal muscle pedicle as flaps for stump reinforcement. In summary, despite advances made over the last decades in surgical technique, postoperative care, and adjuvant therapy in NCLC patients, bronchial stump insufficiency still remains a major morbid complication. In any case, BPF prevention by early recognition of possible risk factors and an individualized approach for each patient in terms of bronchial closure and coverage is the state of the art.
Related Article
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
Top
Abstract
1. Introduction
