Extended sleeve lobectomy for centrally located non-small-cell lung cancer: a 20-year single-centre experience^†

Abstract

OBJECTIVES

Extended sleeve lobectomy (ESL), an atypical bronchoplasty with resection of more than 1 lobe, might be technically demanding but has certain theoretical advantages, including the avoidance of pneumonectomy. However, clinical outcomes after ESL are not well known.

METHODS

Between March 1995 and December 2015, 540 patients with centrally located non-small-cell lung cancer underwent sleeve resection. Among them, 63 patients underwent an ESL procedure. We retrospectively analysed those patients in terms of hospital mortality, postoperative complications and local recurrence and compared clinical outcomes with patients who underwent simple sleeve lobectomy in the same period.

RESULTS

The 63 patients were classified into 4 groups: anastomosis between the right main and lower bronchi (n = 14), anastomosis between the right main and upper bronchi (n = 37), anastomosis between the left main and basal segmental bronchi (n = 4) and anastomosis between the left main and upper divisional bronchi (n = 8). No operative deaths occurred within 30 days, but there were 2 in-hospital deaths from postoperative acute lung injury. Ten (16%) patients had anastomosis-related complications including 3 strictures, 5 bronchopleural fistulas and 2 pulmonary vein thromboses. There were no significant differences in in-hospital mortality (3% vs 3%, P = 0.67), anastomosis-related complications (16% vs 9%, P = 0.07) and loco-regional recurrence rate (8% vs 10%, P = 0.63) between ESL and simple sleeve lobectomy.

CONCLUSIONS

According to our findings, ESL is a safe and feasible procedure that does not compromise oncological principles. It can be considered an appropriate alternative to pneumonectomy and should be considered in patients with centrally located tumours.

INTRODUCTION

Pneumonectomy carries a considerable burden in terms of the quality and quantity of life of patients with centrally located non-small-cell lung cancer (NSCLC). Its mortality rate is reported to range from 4.6% to 9.2% [1–3]. Currently, sleeve resection of centrally located tumours is an accepted procedure to avoid pneumonectomy in patients without compromised lung function.

Extended sleeve lobectomy (ESL), first described in 1959 by Johnston and Jones [4], is an atypical bronchoplasty with resection of more than 1 lobe. However, compared with simple sleeve lobectomy (SSL), several technical difficulties can occur in ESL including greater discrepancy in bronchial calibres, increased anastomotic site tension and, more frequently, combined angioplasty. Therefore, it is controversial whether the ESL can be performed safely without compromising clinical and oncological safety. To date, 4 retrospective series have been reported with small numbers of patients (<30 patients) which provide only limited information [5–8]. Herein, we reviewed our 20-year institutional experience with ESL in comparison with SSL in terms of postoperative outcomes and local failure patterns, with the hypothesis that ESL is a feasible option with reliable clinical and oncological outcomes.

PATIENTS AND METHODS

Study design and population

This study was a retrospective review of prospectively collected data from our institutional lung cancer database. It was approved by the Samsung Medical Center Institutional Review Board, and patient consent was waived based on the nature of this study. From March 1995 to December 2015, 564 patients underwent curative-intent surgical resection with bronchoplastic technique (sleeve resection) for centrally located NSCLC. Patients who underwent a salvage operation or carinal sleeve pneumonectomy (n = 24) were excluded. The remaining 540 patients who underwent sleeve resection to avoid pneumonectomy were included and reviewed in this study. We divided the study population into the following 2 groups: 63 patients in the ESL group (bronchoplasty with resection of more than 1 lobe) and 477 patients in the SSL group (bronchoplasty with resection of only 1 lobe). The flow diagram of patients included in the study is presented in Fig. 1.

Preoperative workup

The preoperative workup included blood tests, pulmonary function tests, simple chest X-ray, computed tomography of the chest and upper abdomen, bronchoscopy with washing cytology or biopsy, whole-body ¹⁸F-fluorodeoxyglucose positron emission tomography or positron emission tomography/computed tomography scan and brain magnetic resonance imaging as described in our previous study [9]. If a metastatic mediastinal lymph node (LN) is suspected, mediastinoscopic biopsy or endobronchial ultrasound-guided transbronchial needle aspiration was performed. Since our institutional approach to potentially resectable clinical N2 disease is trimodal therapy, some patients (6.3% of the ESL group and 10.1% of the SSL group) underwent neoadjuvant concurrent chemoradiotherapy prior to surgery.

Operative technique

Sleeve resection was initiated through standard posterolateral thoracotomy using the fourth or fifth intercostal space. Circumferential dissection of both the bronchi and pulmonary artery was performed with careful examination of any extracapsular invasion or direct tumour involvement with the vascular structures. To prevent excessive tension, sufficient release of both bronchi from the peribronchial soft tissue should be warranted before anastomosis. We also tried not to use electrocoagulation around the stump to minimize devascularization at the anastomosis site. Oblique bronchiotomy of the distal lobar/segmental bronchi was used to overcome any size discrepancy between the 2 bronchial calibres. After the bronchial resection margin was sent for frozen section analysis, tagging sutures were carefully placed in the anterior wall of the proximal and distal bronchi under meticulous geometric consideration. Bronchial anastomosis was initially carried out in the posterior wall, starting in the border of the membranous and cartilaginous part using vicryl 4-0 multiple interrupted sutures (inner knot) with usually 3 or 4 stitches. After posterior wall anastomosis, the anterior wall was sequentially anastomosed using vicryl 4-0 multiple interrupted sutures (outer knot). The bronchial anastomosis site was checked by the air leak test under an airway pressure of 20 mmHg. Surgical glue was applied to the anastomosis site for reinforcement.

If the pulmonary vessel is affected by the tumour itself or by the extracapsular invasion of metastatic nodes, we aggressively resect and reconstruct the vessel (double-sleeve operation). The type of vascular procedure (tangential resection with direct suture angioplasty or resection with end-to-end anastomosis) was selected by the surgeon according to the tumour extent. The vascular procedure was performed using a 5-0 polypropylene continuous running suture after 3000 units of intravenous heparin injection. Additionally, heparin-mixed saline was directly injected into the anastomosed vessel just before declamping to prevent thrombus formation.

Mediastinal LN dissection included nodes at stations 2R, 4R, 7, 8, 9 and 10R for right-sided tumours and nodes at stations 4L, 5, 6, 7, 8, 9 and 10L for left-sided tumours.

Postoperative care

The pleural space after ESL was meticulously managed by adjusting the chest drainage time. Our strategies for residual space management and prevention of space problems included the following: (i) allow sufficient time (generally 10–14 postoperative days) until the diaphragm rises, and the residual space is absorbed; (ii) apply no or minimal suction through a chest tube to prevent alveolar–pleural fistula formation, especially in the immediate postoperative period and (iii) consider mobilizing the last hole of the chest tube (or reposition the chest tube) for optimal space absorption, based on the space dynamics in the serial chest radiographs. A routine bronchoscopy to evaluate the anastomosis site was performed on postoperative day 7 or the day prior to discharge. In the event of more than 2 episodes of recurrent sputum retention with bronchoscopic evidence of a narrowed anastomotic site, further intervention (balloon dilatation or stent insertion) was considered.

Statistical analysis

Categorical variables were compared using the Pearson’s χ² test or the Fisher’s exact test. Normally distributed continuous variables were reported as a mean ± standard deviation and compared using the 2-sample Student’s t-test, and non-normally distributed continuous variables were compared using the Mann–Whitney U-test. As for the survival analysis, overall survival (OS) was defined as the time from the day of operation to death from any cause. Recurrence-free survival was defined as the time from the day of operation to recurrence or death. Survival curves were calculated using the Kaplan–Meier method and were compared univariately using the log-rank test. All reported P-values were 2-sided, and those <0.05 were considered statistically significant. All statistical analyses were performed using SPSS software version 23.0 (SPSS, Chicago, IL, USA).

RESULTS

Baseline characteristics

Table 1 lists the baseline characteristics. The mean age of the ESL and SSL groups was 60 ± 10 years and 61 ± 11 years, respectively (P = 0.35). Most of the patients in both groups were men. The mean tumour size in the ESL and SSL groups was 3.8 ± 1.7 cm and 3.7 ± 1.9 cm, respectively (P = 0.60). Right-sided tumours (81% in the ESL group and 56% in the SSL group) were more common than left-sided tumours. The distribution of tumour location and clinical tumour, node and metastasis (TNM) stage in both groups was not statistically different. Preoperative pulmonary function test (forced expiratory volume in 1 s) showed a similar result in both groups (2.5 ± 0.6 l in the ESL group and 2.5 ± 0.6 l in the SSL group; P = 0.93).

Type of procedure

The 63 patients in the ESL group were classified into 4 groups (Fig. 2): (i) anastomosis between the right main and lower bronchi with upper bilobectomy (n = 14), (ii) anastomosis between the right main and upper bronchi with lower bilobectomy (n = 37), (iii) anastomosis between the left main and basal segmental bronchi with left upper lobectomy and lower lobe superior segmentectomy (n = 4) and (iv) anastomosis between the left main and upper division bronchi with left lower lobectomy and lingular segmentectomy (n = 8).

Extended sleeve lobectomy indications

There were 3 main categories of ESL indications (Table 2): (i) preoperatively expected for the ESL due to the tumour extent on imaging studies (n = 31, 49%), (ii) intraoperatively converted to ESL (from SSL or lobectomy) due to positive bronchial resection margin (n = 18, 29%) and (iii) inevitable ESL due to metastatic interlobar LNs (n = 14, 22%). Considering that categories (ii) and (iii) depend on the results of frozen section analysis, the decision to perform an ESL procedure was made intraoperatively in approximately half of the patients (51%). Table 2 shows the distribution of indications according to operability for pneumonectomy. In the ESL groups, most of the patients (n = 53, 84%) were expected to be functionally tolerable for pneumonectomy. Supplementary Material, Table S1 presents the distribution of indications according to the tumour location.

Operative data

Double-sleeve resection was performed in 13 (21%) patients in the ESL group and 73 (15%) patients in the SSL group (Table 3). Complete resection was achieved in 58 (92%) patients in the ESL group and 436 (91%) patients in the SSL group. Residual microscopic disease (R1 resection) at the bronchial resection margin was found in 4 (6%) patients in the ESL group and 40 (8%) patients in the SSL group.

Squamous cell carcinoma was the most common histological type in both groups (n = 54, 86% in the ESL group and n = 366, 77% in the SSL group). The pathological stage distribution (I:II:III) was as follows: 30%:47%:22% in the ESL group and 34%:38%:24% in the SSL group. Patients with pathological Stage II or III disease received appropriate adjuvant therapy according to the current guidelines.

Postoperative outcomes

In the ESL group, there were 2 in-hospital deaths from postoperative acute lung injury. The in-hospital mortality rate was found to be similar in both groups (3% in the ESL group vs 3% in the SSL group, P = 0.67; Table 4). The mean hospital stay was significantly longer in the ESL group than in the SSL group (25 ± 21 days vs 16 ± 12 days, P < 0.01).

In terms of postoperative complications, the overall incidence of anastomotic complications was higher in the ESL group, but the difference was not statistically significant (16% in the ESL group vs 9% in the SSL group, P = 0.07). While the incidence of anastomotic stricture and pulmonary venous thrombosis (PVT) was not significantly different between the 2 groups, the bronchopleural fistula (BPF) incidence was significantly higher in the ESL group (8% vs 2%, P = 0.02). Table 5 lists a detailed analysis of the fate of 10 patients in the ESL group with anastomosis-related complications. Completion pneumonectomy was inevitably performed in 1 of the 2 PVT cases and 3 of the 4 BPF cases. The relatively high incidence of BPF in the ESL group accounts for higher rates of complete pneumonectomy in the ESL group (6% vs 2%, P = 0.03). In case of anastomotic stenosis, we used an airway silicone stent named Natural stent (M1S Co., Seoul, South Korea) [10].

Recurrence and long-term survival

We investigated the postoperative recurrence after lung cancer surgery with a mean follow-up period of 48 months. In the ESL group, 5 (8%) patients had a loco-regional recurrence, with only 1 case of stump recurrence (2%) (Table 4). As for the distant metastasis, there were 11 (18%) cases of recurrence in the ESL group. There was no statistically significant difference in loco-regional recurrence rates between the 2 groups (8% in the ESL group vs 10% in the SSL group, P = 0.63). Moreover, there was no statistically significant difference in stump site recurrence between the 2 groups (2% in the ESL group vs 2% in the SSL group, P = 1.00).

With a mean follow-up of 48 months, there was no difference in the 5-year OS between the 2 groups (ESL, 62% vs SSL, 69%; log-rank P = 0.82; Fig. 3). The 5-year recurrence-free survival rates were 59% for the ESL group and 57% for the SSL group (log-rank P = 0.70). In the Kaplan–Meier survival curves of both OS and recurrence-free survival, survival outcomes of the ESL group were comparable with those of the SSL group.

DISCUSSION

Generally, sleeve lobectomy is recommended as an alternative to pneumonectomy for patients with NSCLC [11, 12]. Depending on the severity of the disease, some patients with lung cancer should undergo ESL rather than a typical sleeve resection to avoid pneumonectomy. A study by Okada et al. [7] was the first to report the usefulness of ESL, which is a more challenging procedure because of the potential for increased tension at the anastomosis site as well as size mismatch. In this study, the overall postoperative results of ESL were as acceptable and feasible as those of the SSL group, though the length of hospital stay and the incidence of postoperative BPF were greater in the ESL group than in the SSL group. It is important to remember that the decision to perform ESL for lung cancer should be based on many factors including quality of life, risk of recurrence and postoperative complications. Most importantly, the ESL technique should only be attempted when there is the possibility of complete resection [13]. Okada et al. [7] showed that the indications for ESL should include sufficient margins of at least 1 cm of macroscopically unaffected bronchus. Although complete resection was not achieved in all patients in this study as in the study by Okada et al., there was no significant difference in terms of the overall results of the lung cancer surgeries.

As for postoperative mortality rates after SSL, Ferguson and Lehman [11] reported a weighted mean mortality of 4.1% (confidence interval 1–11%). However, several studies have shown no incidences of postoperative death after ESL [7, 8]. In our series, there was no operative death, but 2 in-hospital deaths (in-hospital mortality of 3%) due to acute lung injury. Considering this to be the largest series among the reported, the mortality rate in our series seems to be acceptable.

The mean hospital stay in the ESL group was longer than in the SSL group, for the following possible reasons: (i) prolonged chest tube management might be inevitable in the ESL group due to the greater extent of resection and more residual space and (ii) the proportion of patients with BPF who underwent completion pneumonectomy was significantly higher in the ESL group.

Anastomosis-related complications after ESL, such as BPF, stricture and PVT, represent another important aspect. Two patients had PVT in this study. Among these, one had to undergo completion pneumonectomy, while the other recovered fully with conservative management. Most likely, PVT occurred in the context of an increased distance between the proximal and distal stumps; unfortunately, however, the length of the resected segments was not obtained at the time of surgery. Therefore, the shape and blood flow status of the pulmonary vein as well as the status of bronchial or pulmonary artery anastomosis must be evaluated before the end of the operation. Our group is now applying various methods such as intraoperative flow measurement or direct puncture, especially in cases in which the distance between 2 stumps is long or the shape and course of the pulmonary vein is worrisome. Recently, the routine use of a releasing manoeuvre at the pulmonary vein–left atrial junction with pericardial incision has been reported as an adjunctive measure for prevention. As PVT is not a common phenomenon but can be fatal, surgeons should keep the possibility in mind and actively prevent PVT without fail, especially when performing an operation such as ESL.

BPF occurred in 5 patients after ESL. Among these, 4 patients had to undergo reoperation (3 patients, completion pneumonectomy and 1 patient, primary closure), and 1 patient achieved recovery with conservative management without surgery. BPF results from ischaemia of the anastomosis site. Sleeve procedures are generally safe and feasible, but anastomotic complications are serious and often fatal. Usually, the BPF after ESL is more rapidly progressive than in cases of SSL and frequently lead to catastrophic results. Once a BPF has been confirmed, reoperation should be considered depending on the size of the BPF and the performance status of the patient. Conservative management strategies can be applied including chest tube drainage, broad-spectrum antibiotics and bronchoscopic procedures in patients with small-sized BPFs [14–17]. Various risk factors such as neoadjuvant treatment for the development of BPF after sleeve procedures have been reported [18–21]; however, we have not yet found any confirmative factor. It is difficult to analyse the risk factors for morbidity and mortality in ESL due to the variety of procedures and disease extents; however, the number of operations in this study is, to the best of our knowledge, the largest in the ESL-related literature. Therefore, we can speculate that ESL causes excessive tension at the anastomosis site, resulting in a greater likelihood of dehiscence when compared with SSL.

Regarding local recurrence (stump site recurrence), it was observed in only 1 patient (2%) in our study. The local recurrence rate ranged from 4.3% to 9.7% in recent studies [22–24]. The majority of the studies have shown similar survival and local recurrence outcomes between sleeve lobectomy and pneumonectomy [22, 25]. Kim et al. [25] emphasized that sleeve resection must be performed in selected patients, such as those without LN metastasis, because they found a higher local recurrence rate in the sleeve lobectomy group. Unfortunately, we were unable to determine any relevant factors such as LN involvement in patients with local recurrence (5 cases of loco-regional recurrence: Stage N0, n = 4; Stage N1, n = 1). On the basis of our local recurrence rate and long-term survival outcomes, we believe there is no need to limit the use of ESL to selected populations such as clinical N0. Although more studies are required, we suggest that ESL can be performed with acceptable clinical and oncological outcomes, even when interlobar LN metastasis is present.

In our study, there were 2 in-hospital mortality cases (mortality rate of 3%), the incidence of anastomosis-related complications was acceptable, and the rate of local recurrence was very low (2%). As the operative mortality in the pneumonectomy group (8.6%) derived from our previous institutional data in 2010 as well as data from other pneumonectomy series showed a much higher risk, we think that ESL might be a safer option for surgical management of centrally located NSCLC [2]. With respect to long-term outcomes, our data also suggest that the 5-year OS in the ESL group (62%) was more favourable than that in the pneumonectomy group in 2010 (32.1%) [2]. Although direct comparison with previous series is not completely reliable, we can reasonably suggest that pneumonectomy will not provide better surgical outcomes compared with ESL in this subset of patients. Pneumonectomy is a disease in itself and should be avoided as possible as we could because of its association with several serious complications including post-pneumonectomy empyema, BPF and pneumonectomy syndrome. Therefore, it remains challenging to establish the exact indications for ESL in lung cancer surgery. Our results showed that it is a relatively feasible and safe procedure compared with SSL or pneumonectomy, though careful patient selection and proper perioperative management are mandatory.

CONCLUSION

In conclusion, considering the extent of the disease, ESL for lung cancer is a safe and feasible procedure that does not compromise oncological principles. It can be considered an appropriate alternative to pneumonectomy and should be considered in patients with centrally located NSCLC.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

Conflict of interest: none declared.

REFERENCES

Author notes