Complex sleeve lobectomy has the same surgical outcome when compared with conventional lobectomy in patients with lung cancer Free

Abstract

Open in new tabDownload slide

OBJECTIVES

No significant data are available to assess whether complex sleeve lobectomy (complex-SL) can be considered comparable to conventional lobectomy (CL) in terms of surgical outcome. The purpose of this study was to compare surgical and oncological outcomes of complex-SL with CL in patients with lung cancer.

METHODS

Between 2000 and 2015, a total of 568 patients who underwent open CL (defined as resection of only 1 lobe) and 187 patients who underwent SL were analysed. The SL group was divided into 2 subgroups: standard-SL (bronchial SL, n = 106) and complex-SL (n = 81) (defined as bronchial sleeve resection together with another surgical intervention: bronchovascular SL, n = 40; vascular SL, n = 26; atypical bronchoplasty with resection of more than 1 lobe, n = 12; bronchial SL + chest wall resection, n = 3).

RESULTS

The complex-SL group had more patients with chronic obstructive pulmonary disease (COPD) (25.9% vs 12.5%, P = 0.001), neoadjuvant treatment (39.5% vs 12.0%, P < 0.001), advanced-stage non-small-cell lung cancer (53.2% vs 33.1%, P = 0.001) and low preoperative forced expiratory volume in 1 s (77.2% vs 84.3%, P = 0.004) than the CL group. The overall surgical mortality (in-hospital or 30-day) was 2.6% (n = 20); it was 2.8% for CL and 2.8% for complex-SL. Postoperative complications occurred in 34.9% of the CL group and 39.5% of the complex-SL group (P = 0.413). The pulmonary complication rate was similar between the groups (24.1% for CL, 27.2% for complex-SL, P = 0.552). The 5-year survival in the CL group was 57.1%, and in the complex-SL group it was 56.2% (P = 0.888). Multivariate analysis showed that TNM stage (P < 0.001) and N status (P < 0.001) were significant and independent negative prognostic factors for survival.

CONCLUSIONS

Complex-SL had a comparable outcome to CL, although the complex-SL group had more patients with advanced-stage NSCLC, low preoperative forced expiratory volume in 1 s and COPD.

INTRODUCTION

Surgery is a crucial part of the treatment of lung cancer. According to published series, there is clear evidence that sleeve lobectomy (SL) is oncologically comparable to pneumonectomy (PN), with no increased postoperative morbidity, lower mortality and better quality of life because of functional preservation [1–4]. Several surgeons have accepted and confirmed sleeve resection as an alternative surgical procedure to PN in patients with lung cancer. Compared with conventional lobectomy (CL) and standard-SL, several technical difficulties such as a discrepancy in bronchial calibres, increased anastomotic site tension, combined angioplasty and additional procedures could occur in SL [4]. However, there are no significant data available comparing CL and SL (standard or extended), and clinical, surgical and oncological outcomes after extended SL are not well known [5–7]. Theoretically, a centrally located tumour resected with complex surgery such as SL or SL plus additional procedures could have a poorer prognosis because of less effective local control when compared with that achieved with CL for resection of more peripheral lesions [5]. Therefore, our aim was to compare the complex-SL and CL with regard to their surgical and oncological outcomes.

MATERIAL AND METHODS

From January 2000 to December 2015, a total of 1548 consecutive patients with malignant lung disease underwent lung resection at the University Hospital Zurich. This was a retrospective observational study, although patients who were operated on at the Department of Thoracic Surgery during that time were identified in the prospectively collected database. The more detailed data were collected retrospectively, and more information about the patients, their diseases and their treatment was retrieved from the clinical information system.

Ethical statement

The cantonal ethics committee of Zurich approved this study: 2016-00799.

Patients

All patients who underwent surgery for non-malignant disease and who underwent minimally invasive lobectomies were excluded. We included patients who underwent open CL and patients who underwent SL for malignant lung disease. A total of 755 met the criteria. CL (n = 568) was defined as resection of only 1 lobe without another lobe and/or organ resection. SL (n = 187) was defined as resection of the affected part of the bronchus and/or pulmonary artery and anastomosis of healthy proximal and distal ends. When talking about SL, one must also keep in mind that there are different types of SL. Therefore, sleeve lobectomies were further divided into 2 subgroups: standard-SL (bronchial SL, n = 106) and complex-SL (bronchial SL with another surgical intervention, n = 81). In the present study, complex-SL was defined as bronchovascular SL-double SL (n = 40), vascular SL only (n = 26), atypical bronchoplasty with more than 1 lobe (n = 12) and bronchial SL with chest wall resection (n = 3). These additional procedures were defined as complex resection. An atypical bronchoplasty is defined as a resection of more than 1 lobe and anastomosis of bronchus (anastomosis between the right main and lower bronchi with upper bilobectomy, n = 6; anastomosis between the right main and upper bronchi with lower bilobectomy, n = 5; anastomosis between the left main and basal segmental bronchi with left upper lobectomy, n = 1). Moreover, SLs were also classified according to pulmonary artery resection/reconstruction (angioplasty) status as non-vascular SL (121) and vascular SL (n = 66).

Surgical details

We have recently published our technique [8]. Briefly, after a double-lumen tracheal intubation, all patients underwent anterolateral thoracotomy. Intraoperative flexible bronchoscopy was performed immediately before and during the operation. The bronchus was divided with a knife to obtain straight margins distant from the tumour. Frozen sections of the resected margins were analysed to ensure complete resection. The end-to-end bronchial anastomosis was performed using a continuous absorbable running 4-0 polydioxanone suture (Ethicon, Inc., Somerville, NJ, USA) on the membranous airway wall. The cartilaginous part was anastomosed with an interrupted suture technique, and the knots were placed outside the lumen and tied on completion of the anastomosis. Care was taken to keep the peribronchial tissue intact, and systematic lymph node dissection was performed in all neoplastic patients. The wrapping of the anastomotic bronchial suture with any kind of autogenous tissue was not routinely performed. After the completion of the anastomosis, the anastomosis was always checked by ventilating the lung with a maximum airway pressure of 25–30 cm of water. At the end of the procedure, the bronchial anastomosis was controlled with flexible bronchoscopy and clots and secretions were cleared away. When we perform bronchovascular sleeve resection, we first perform the bronchus anastomosis and then perform the vascular anastomosis. The use of heparin was not standard in performing vascular reconstruction. If heparin was used during vascular end-to-end anastomosis, 1000–1500 IU was given intravenously before clamping. In all cases, any needed arterial reconstruction of the main pulmonary artery was performed, and the interlobar pulmonary arteries were encircled and clamped before transecting the vessel. If patch plasty was indicated, the reconstruction was performed with the patient’s own pericardium or with a bovine pericardium. If the resected lumen would not cause narrowing of the pulmonary artery lumen, a tangential resection was performed and directly sutured. We use polypropylene suture (5-0) during anastomosis in vascular procedures (Prolene, Ethicon, Inc.).

Staging, incomplete resection and mortality

Preoperative mediastinal staging was performed with positron emission tomography/computed tomography, endobronchial ultrasound (EBUS) and/or mediastinoscopy. If the mediastinal lymph nodes are negative on positron emission tomography/computed tomography (PET/CT), we did not perform EBUS or mediastinoscopy. If the mediastinal lymph nodes are positive on PET/CT, EBUS or mediastinoscopy was performed. All tumours were staged using the Union for International Cancer Control (UICC, 7th Edition) TNM classification. As the data from this study were from 2000 to 2015, all tumours before 2010 have been restaged using the pathology report and the rules stated in UICC, 7th Edition, for better comparability. Regarding surgical margin, microscopic invasion was labelled as R1. ‘Operative mortality’ was defined according to the report of the Society of Thoracic Surgeons (STS) Congenital Database Taskforce and the Joint European Association of Cardiothoracic Surgery (EACTS-STS) Congenital Database Committee from 2006 as ‘any death regardless of cause occurring within 30 days after surgery in or out of the hospital, and after 30 days during the same hospitalization subsequent to the operation’.

Statistical analysis

The data were entered into the Statistical Package for the Social Sciences (IBM SPSS Statistics for Windows, Version 25.0, Armonk, NY, USA). The influence of the following factors on survival was analysed: age; sex; histological type; pT status, type of resection; and pN status. Overall survival was the time between surgery and death from any cause. Survival was estimated by the Kaplan–Meier method, and survival comparison between groups was performed using log-rank analysis. Univariate Cox regression analyses were used to examine the association between survival and potential prognostic factors. The χ² test was used to determine whether the proportions of independent groups are statistically significantly different. Univariate logistic regression analyses were used to determine the impact of patient characteristics on in-hospital complications. A P-value <0.05 was considered statistically significant.

RESULTS

Patient characteristics and comparison between surgical groups are summarized in Table 1. The complex-SL group had more patients with chronic obstructive pulmonary disease (COPD) (P = 0.001), advanced-stage (stage III/IV), non-small-cell lung cancer (NSCLC) (P = 0.001), advanced pT status (P < 0.001) and lower preoperative forced expiratory volume in 1 s (FEV1) (P = 0.004) than the CL group. However, the CL group had a higher pN2 rate than the complex-SL group (P = 0.01). A statistically significant difference in tumour histology was found when comparing the surgery types (P < 0.001). Post hoc analysis found that histology was different in CL compared with the 2 other surgery types. Neoadjuvant therapy was comparable between the standard-SL and CL groups, although patients who underwent CL received neoadjuvant treatment less frequently than patients in the complex-SL group.

Sixty-six patients underwent pulmonary artery resection/reconstruction. Five of the surgeries were performed under cardiopulmonary bypass. We performed 42 pulmonary artery sleeve resections (20 of them had concomitant bronchial sleeve or double sleeve), 21 simple tangential resections with traditional direct suture and 3 reconstructions by a pericardial patch.

Of the patients undergoing R1 resection (n = 32, 4.3%), 18 were in the CL group (the pulmonary vein in 8 patients, the pulmonary artery in 5 patients, the mediastinal fat tissue in 3 patients and the bronchus in 2 patients). Eight patients undergoing R1 resection were in the standard-SL group (the left atrium in 4 patients, the left or right main pulmonary artery in 3 patients, the hilar fat tissue 1 patient); and 6 patients were in the complex-SL group (the left atrium in 3 patients, the main pulmonary artery in 1 patient, the aortic adventitia in 1 patient and the chest wall in 1 patient). Further analysis of the data showed a significant difference in the occurrence of R1 resections in a descending order from SL to CL, and the differences between the groups were statistically significant (P = 0.04). In the complex-SL group, more patients underwent adjuvant therapy (45.4%) than in the standard-SL group (31.1%) and in the CL group (32.0%). This finding was statistically significant (P = 0.04). Observed frequencies and percentages of adjuvant treatment are presented in Table 2.

The overall rate of postoperative complications was 34.6% (n = 261). The complication rate in the complex-SL group was higher than the other surgery types (39.5%, n = 32); however, it was not statistically significantly different from standard-SL or CL (29.2% and 34.9%, respectively). Also, the rates of pulmonary, cardiac and gastrointestinal system complications were almost similar between the CL and complex-SL groups (Table 2). Prolonged air leakage (>7 days) was the most common pulmonary complication (n = 48, 6.3%), and no significant difference was found among the groups (5.6% for CL, 7.5% for standard-SL and 6.4% for complex-SL, P = 0.744). The second most common pulmonary complication was pneumonia (n = 45, 6.0%), and 36 patients with pneumonia had undergone a CL (6.3%); 5 patients had a standard-SL (4.7%); and 4 patients had a complex-SL (4.9%) (P = 0.366). Bronchopleural fistula (BPF) occurred in 5 cases (0.6%): 3 cases had a CL (0.5%), 1 case had a standard-SL (0.9%) and 1 of them had a complex-SL (1.2%) (P = 0.233). Two patients developed a BPF that was successfully managed conservatively; 1 patient underwent a PN; however, the remaining 2 patients who had a BPF died on postoperative day 12 and 15, respectively, of acute respiratory distress syndrome (ARDS). Long-term bronchial complications (intermediate bronchus narrowing) occurred in 1 patient who underwent complex-SL. This mild stenosis did not require any surgical treatment; however, the patient died after 23 months because of a pulmonary embolism. The other postoperative pulmonary complications were pleural effusion (n = 40, 5.2%), haemorrhage (n = 15, 1.9%), atelectasis (n = 9, 1.1%), empyema without BPF (n = 8, 1.0%), chylothorax (n = 8, 1.0%) and pulmonary embolism (n = 7, 0.9%).

There was no significant difference between non-vascular SL and vascular SL with regard to overall complications (34.7% vs 31.8%, P = 0.812), pulmonary complications (23.1% vs 21.2%, P = 0.792) and cardiac complications (12.3% vs 15.1%, P = 0.521).

Increasing age was associated with an increased likelihood of postoperative complications (P < 0.0001). In addition, increased body mass index (BMI) was predictive of a higher morbidity rate (P = 0.01). Regarding postoperative complications, the binomial logistic regression also showed no significant difference when comparing complex-SL and CL to standard-SL (Table 3). The model explained 8.9% (Nagelkerke R2) of the variance in morbidity and correctly classified 65.6% of cases. The sensitivity was 23.5%, and the specificity was 89.8%.

The overall surgical mortality was 2.6% (n = 20). Twelve patients died from postoperative infection leading to sepsis and multi-organ failure (10 patients in the CL group and 2 patients in the standard-SL group). One patient in the complex-SL group died after massive bleeding that required reoperation, after which the patient developed ARDS. Three patients had renal insufficiency after surgery, developed ARDS and subsequently died (2 patients in the CL group and 1 patient in the complex-SL group). One patient in the CL group died after developing ARDS caused by pneumothorax. Three other patients died of multiple complications after CL that ultimately led to organ failure. The surgical mortality according to the groups is reported in Table 2. There was no statistical difference between the groups for the rate of surgical mortality (P = 0.772). Age, BMI, preoperative FEV1, COPD, side of operation, packs of cigarettes and TNM stage did not increase the mortality rate (P > 0.05), although the presence of any postoperative complication and comorbidity significantly influenced the occurrence of postoperative mortality (P < 0.0001 and P = 0.04, respectively). It was observed that patients with postoperative pneumonia or BPF had a higher mortality rate (P < 0.001). Also, in our centre, the mortality rate for minimally invasive surgery (video-assisted thoracic surgery and robotic lobectomy) was 0% during the same period.

The overall 5-year survival was 57.5% (median 78 months). The 5-year survivals were 63.1% for pN0, 54.7% for pN1 and 42.2% for N2. The 10-year survivals were 35%, 23% and 10.4%, respectively (P < 0.0001). The 5-year survival in the CL group was 57.1% (median 86 months), 57.6% for standard-SL (median 80 months) and 56.2% for complex-SL (median 88 months) (P = 0.928). There was no significant difference between the CL and complex-SL groups (P = 0.888) (Fig. 1). Cox regression analysis showed that TNM stage (P < 0.001), N status (P < 0.001), incomplete resection (P = 0.002) and low preoperative FEV1 (P = 0.01) were significant independent negative prognostic factors for survival (Table 4). In the CL group, the 5-year survival rate was 67.6% (median 108 months) for stage I, 52.5% (median 73 months) for stage II, 47.4% (median 54 months) for stage III and 24.2% (median 30 months) for stage IV (P < 0.001). In the standard-SL group, it was 66.3% (median 99 months) for stage I, 56.4% (median 68 months) for stage II, 47.2% (median 58 months) for stage III and 20.0% (median 13 months) for stage IV (P = 0.01). In the complex-SL group, it was 62.3% for stage I (102 months), 58.3% (median 82 months) for stage II, 47.4% (median 60 months) for stage III and 28.1% (median 33 months) for stage IV (P = 0.517). .

DISCUSSION

Currently, SL is the preferred surgery for patients with a malignancy that anatomically qualifies, regardless of whether or not they would tolerate a PN, as SL preserves more lung tissue, is associated with a lower operative mortality and has an equivalent oncological outcome [1, 2]. The postoperative quality of life has been advocated as one of the most reliable indicators that should influence the decision to perform an SL. A number of studies indicate that sparing lung parenchyma improves postoperative quality of life because of a greater pulmonary reserve. The lung parenchyma is preserved in both SL and CL. However, CLs do not require a bronchial or vascular anastomosis as SL does, and CLs can more frequently be performed minimally invasively. Therefore, CL is technically less challenging than SL. When we looked at the published series about SL, there is only 1 study comparing SL and CL. D’Andrilli et al. [5] compared functional and oncological outcomes of SL with those of CL. Our present study is the first study to compare surgical and oncological outcomes of complex-SL with those of CL in patients with lung cancer.

Every surgical intervention can potentially lead to complications. However, many surgeons think that bronchoplastic surgery has a higher complication rate than CL. The logistic regression, done to determine factors influencing the morbidity rate, showed that age and BMI had a statistically significant impact on the incidence of complications, although type of surgery did not influence it. Increased age and increased BMI were associated with an increased likelihood of postoperative complications. These findings differ from what Yildizeli et al. [9] found in 2007, as they reported that right-sided resection, smoking and squamous cell carcinoma were risk factors for higher complication rates. Another study from Ogawa et al. [10] in 2014 about the impact of lung age on postoperative complications in patients with NSCLC reported that gender, age gap (gap between lung age and the age of the patient) and preoperative comorbidities have a significant influence on the morbidity rate. The analysis of the data in this study showed no significance in any of these factors. It seems that consensus about the factors that influence the incidence has yet to be reached. Many studies only analysed survival and did not focus on morbidities, which might be why we do not yet know what risk factors influence the outcome. A previous study reported that chemoradiation may negatively influence mucosal blood flow [11] and therefore could lead to more anastomosis-related complications and worse bronchial stump healing [12]. However, our study could not confirm that there was a negative effect of preoperative chemotherapy or radiation on the outcome of surgery. The logistic regression to identify influences on morbidity rate did not show a statistically significant influence of neoadjuvant treatment.

On the other hand, the main concern with SL is related to potential sequelae of impaired healing of the anastomosis that may increase the perioperative morbidity with respect to standard lobectomy and may limit the functional benefit of parenchymal sparing [5]. In the previously published meta-analysis about SL, it was observed that postoperative morbidity and mortality depended on the healing of the anastomosis [13]. In the present study, we found that the presence of BPF significantly influenced the occurrence of postoperative mortality. Ferguson and Lehman [14] reported that the mean operative mortality was 4.1% after SL. In a recently published review that included 22 studies about SL, Tapias et al. [15] reported an incidence of BPF of 0–8.1% of patients undergoing sleeve resections. The protection of the bronchial anastomosis is a subject of controversy. Many authors consider wrapping the bronchial anastomosis with an autogenous tissue-pedicled flap as a fundamental procedure to reduce the incidence of severe anastomotic complications [16]. In our experience, we did not routinely use any kind of autogenous tissue-pedicled flap and registered an anastomotic complication rate of 1% for all SLs (1.2% in complex-SL group, 0.9% in standard-SL). We think that anastomotic complications were prevented through our no-touch technique, careful dissection, preservation of blood flow and precise anastomosis. We consider the preservation of an intact peribronchial tissue and many years of surgical experience in the field of bronchial anastomosis as the most important factors contributing to a low anastomotic complication rate.

We showed that the median tumour stages were statistically different among all 3 types of surgery. CL has the lowest, followed by standard-SL; and, complex-SL has the highest average stage. To conclude, even with a higher average stage, SL still had the same long-term survival as CL. This means that a more complicated surgery does not negatively influence long-term survival. Both Deslauriers et al. [1] and Okada et al. [17] reported a better prognosis after SL treatment in stage I and stage II patients. Considering all influences on long-term survival, it seems that surgery type is not the determining factor, but rather progression of the tumour at the time of surgery. Cox regression analysis showed that age, incomplete resection, preoperative FEV1 and TNM stage influenced survival significantly; however, it did not show an influence of surgery type on survival. These results mean that SLs are not inferior when treating malignant disease, as previously suspected by other authors [14], but comparable to CL. As SL does not negatively affect survival when used for treating malignant disease and has the obvious advantage of preserving more lung tissue, this study, as have many previous studies concluded [1, 8, 18, 19], comes to the conclusion that SL can be used when anatomically appropriate, and should not simply be viewed as an alternative for patients who cannot tolerate larger lung resections.

On the other hand, patients who underwent complex-SL had a higher adjuvant treatment rate than those patients who underwent CL. Adjuvant treatment may offer a survival benefit to a number of staging scenarios in NSCLC [20]. Also, the CL group had a higher pN2 rate than that of the complex-SL group. Several authors found a higher incidence of occult metastases in patients having adenocarcinoma compared with other histological types [21, 22]. Adenocarcinoma, which has been considered to be associated with skip N2 metastasis, is also the most prevalent pathological subtype of diagnosed NSCLC. The skip metastasis rate for CL patients with pN2 was 28.6% (n = 49), whereas it was 16.6% (n = 3) in the complex-SL group (data not shown here). It is known that the nodal status is the most significant prognostic factor in operable NSCLC, although the impact of T factor seems to remain controversial. A lower adjuvant treatment rate and higher pN2 rate could be a good explanation for the unexpectedly poor survival among CL patients.

Limitations

As with every retrospective data analysis, this study also has its limitations. In retrospective studies, causality can never be determined with absolute certainty. Furthermore, they are more prone to mistakes because investigators have to rely on old patient files, which are not always complete. It would be better to provide data about the disease-free survival instead of the overall survival. However, because of the retrospective nature of the study, we had no chance to get more accurate information regarding the disease-free survival and loco-regional recurrence. The advantages of this study are that it is ethically less problematic than prospective studies and it is more cost efficient. Another benefit of this study is that all patients were operated on at a single centre.

CONCLUSION

In conclusion, the theoretical risk of decreased long-term survival after SL compared with standard lobectomy because of closer margins obtained in the presence of centrally located tumours, has not been confirmed in the present study. This result supports the widely reported evidence that complex-SL is an adequate operation to ensure oncological radicality. The operative mortality and complication rates of complex-SL were also comparable to CL. This was the case even though complex-SL had more patients with advanced stages, lower preoperative FEV1 and COPD. According to our findings, complex-SL is a safe and feasible procedure that does not compromise oncological and surgical principles.

Conflict of interest: none declared.

Author contributions

Ilhan Inci: Conceptualization; Methodology; Validation; Writing-original draft. Martina Benker: Data curation; Software; Writing—original draft. Necati Çitak: Formal analysis; Investigation; Methodology; Software; Validation; Writing—original draft; Writing—review & editing. Didier Schneiter: Conceptualization. Claudio Caviezel: Data curation; Resources; Visualization. Sven Hillinger: Investigation; Resources; Software; Writing—review & editing. Isabelle Opitz: Data curation; Formal analysis; Writing—review & editing. Walter Weder: Methodology; Supervision; Visualization; Writing—review & editing.

Presented at the 27th European Conference on General Thoracic Surgery, Dublin, Ireland, 9–12 June 2019.

REFERENCES

Abbreviations

Abbreviations
 
• ARDS

  Acute respiratory distress syndrome

 
• BMI

  Body mass index

 
• BPF

  Bronchopleural fistula

 
• CL

  Conventional lobectomy

 
• COPD

  Chronic obstructive pulmonary disease

 
• FEV1

  Forced expiratory volume in 1 s

 
• NSCLC

  Non-small-cell lung cancer

 
• PET/CT

  Positron emission tomography/computed tomography

 
• PN

  Pneumonectomy

 
• SL

  Sleeve lobectomy