Three-port single-intercostal versus multiple-intercostal thoracoscopic lobectomy for the treatment of lung cancer: a propensity-matched analysis

Over the past two decades, lobectomy via video-assisted thoracoscopic surgery (VATS) has been accepted as an effective and safe treatment for patients who are diagnosed with non-small-cell lung cancer (NSCLC) [1, 2]. Compared with conventional thoracotomy, thoracoscopic lobectomy is considered to produce fewer complications, require shorter hospital stays and cause less pain for patients [3, 4]. However, the techniques used in the procedure vary among surgeons, and there is no standard approach [5, 6].

Conventional three-port multiple-intercostal (MIC) lobectomy has been applied by most surgeons for thoracoscopic lobectomy. Employing standardized three-port MIC lobectomy can allow different angles for lymphadenectomy and bronchovascular dissection [7]. Disadvantages of this approach include possible injury to multiple intercostal nerves and creation of a torsion angle in the target tissues for the observation port relatively low in the intercostal space [8, 9]. In recent years, a modified technique, three-port single-intercostal (SIC) thoracoscopic lobectomy for NSCLC [10, 11], has been adopted in our department. This technique can reduce the usage of the intercostal space, and different angles for lymphadenectomy and bronchovascular dissection are made possible. However, three-port SIC lobectomy and conventional three-port MIC VATS lobectomy have never been compared. We hypothesized that patients with NSCLC treated with SIC VATS lobectomy would have comparable perioperative outcomes to patients treated with MIC VATS lobectomy. This study presents our experience performing SIC VATS lobectomy based on a propensity-matched analysis.

The first three-port MIC VATS lobectomy was performed in our department in 2003. With increasing experience, one surgeon began to employ the SIC technique instead of the MIC technique in November 2014, whereas the other surgeon still performed three-port MIC due to individual preference. The first 20 cases were excluded to account for the learning curve effect.

All the patients underwent a preoperative examination, which included respiratory function tests, computed tomography scanning of the chest, brain magnetic resonance imaging and cardiologic assessment. Clinical and demographic data, including age, sex, smoking history, forced expiratory volume in the first second (FEV1), tumour diameter and location, and pathologic stage, were recorded. The operative time, estimated volume of blood loss, length of hospital stay, chest tube removal, complications and mortality within 30 days were collected. Pathologic stage was determined based on the American Joint Committee on Cancer staging system, 8th edition.

From January 2013 to January 2018, 642 patients with NSCLC treated with VATS lobectomy were enrolled for analysis in this study (Fig. 3). Among the 642 patients, 210 underwent three-port SIC thoracoscopic lobectomy, and 432 accepted three-port MIC thoracoscopic lobectomy. Patients who were converted to open surgery were included in the final analysis. The first 20 cases in the SIC group were excluded to account for the learning curve effect. A total of 186 closely matched pairs were generated after propensity score matching analysis in this retrospective study. The baseline demographic parameters and clinical characteristics of the study cohort before and after matching are listed in Table 1. Two groups were well matched, and no significant differences were observed in the baseline clinical variables.

A comparison between propensity-matched groups for operative details and perioperative outcomes is shown in Table 2. The mean operating time in the SIC group was 135.6 min (34.0; 95% CI: 130.7–140.9), which was significantly shorter than that of 165.4 min (32.2; 95% CI: 161.0–170.2) in the MIC group (p < 0.001). The volume of blood loss was significantly lower in patients in the SIC group than in those in the MIC group (p < 0.001). Additionally, the main chest tube (26-Fr) could be removed on postoperative day (POD) 1 from patients in the SIC group, which was significantly faster than in the MIC group (p < 0.001). Although the total number of lymph nodes harvested (25.3 vs. 23.8, p = 0.160) and stations removed (6.5 vs. 6.7, p = 0.368) were similar between the two groups, more subcarinal lymph nodes were removed (6.9 vs. 5.2, p < 0.001) in the SIC group than in the MIC group. The length of hospital stay (6.5 vs. 6.3, p = 0.478) and the complication rate (4.8% vs. 6.5%, p = 0.500) were comparable between the two groups, with no significant differences. Intraoperatively, there were two patients (1.1%) in the SIC group who were converted to thoracotomy due to severe adhesion formation, while four patients (2.2%) in the MIC group were converted to thoracotomy, one for uncontrolled bleeding (injury to the pulmonary artery) and three for severe adhesion in the thoracic cavity. One patient in the MIC group died due to sudden cardiac death on postoperative day 7. No other cases of in-hospital mortality or 30-day mortality were identified in this study.

We also analysed whether the perioperative outcomes (operative time, blood loss and length of stay) in the three-port SIC group varied with the preoperative variables (age, sex, tumour location, number of lymph nodes harvested and tumour size) and found that only patients who were older than 30 had a longer hospital stay (p = 0.039) (Table 3).

In this retrospective study, we demonstrated that compared with NSCLC patients treated with three-port MIC lobectomy, those treated with three-port SIC lobectomy had a potential advantage with regard to less direct stimulation of the intercostal nerves and more subcarinal lymph nodes removed. In addition, there was no compromise in the total number of lymph nodes harvested, complications or length of hospital stay.

Numerous studies have demonstrated that conventional three-port MIC VATS lobectomy is an acceptable procedure for the treatment of NSCLC with a low morbidity and mortality rate [2‐4]. The Society of Thoracic Surgeons database reported (n = 1281) that for VATS lobectomy, the total complication rate was 26.2%, and the mortality rate was 0.94% [3]. A case-control study by McKenna et al. [12] presented (n = 1100) that the total complication rate was 15.3%, and the mortality rate was 0.8%. In previous study [10], we reported our initial results of three-port SIC VATS for bulla resection, wedge resection, segmentectomy and lobectomy. However, three-port SIC and MIC VATS lobectomy have never been compared before. In this study, we found that a small percentage of patients had major morbidity: 4.8% of patients in the SIC group and 6.5% of patients in the MIC group. There was no significant difference in major morbidity between the two groups. There was only one death (0.5%) in the MIC group, and no mortality was recorded in the SIC group during this hospitalization. Accordingly, we can conclude that three-port SIC and MIC lobectomy were both safe procedures that could be performed by experienced thoracic surgeons and conferred low morbidity among patients with NSCLC.

The potential advantage of SIC lobectomy is that only one intercostal space was used to complete the operation, which could theoretically reduce the damage to the intercostal nerves and result in less pain. However, we did not compare pain scores between the two groups in this study since we have continually made effort to reduce postoperative pain for patients in recent years. According to previous studies [13, 14], acute pain following VATS was mainly due to intercostal nerve damage, surgical trauma to muscles and placement of the chest tube. In addition, the ideal anaesthetic regimen for VATS would be to use regional anaesthesia combined with opioids and nonopioid analgesics [13]. Thus, reducing direct stimulation of the intercostal nerves, removing the main chest tube earlier and using patient-controlled paravertebral blocks for continued regional anaesthesia are key strategies for reducing postoperative pain. We speculate that SIC lobectomy may result in less postoperative pain and better patient tolerance than MIC lobectomy.

Radical systemic mediastinal lymphadenectomy (more than 20) was routinely performed for most patients with NSCLC in our department. In this study, the total number of lymph nodes harvested (25.3 vs. 23.8) and stations removed (6.5 vs. 6.7) were similar between the two groups. It has been reported that radical mediastinal lymphadenectomy can provide accurate pathologic staging and is probably advantageous for survival outcomes [15]. The disease-free and overall survival of patients can improve, with a plateau at 11 or more lymph nodes [16, 17]. Wang et al. [5] and Shen et al. [18] reported that approximately 22 lymph nodes were harvested for conventional three-port MIC VATS, which was similar to the number of lymph nodes harvested in MIC VATS in this study.

Furthermore, more subcarinal lymph nodes were removed (p < 0.001) in the SIC group than in the MIC group. Lymphadenectomy has become easier for surgeons to perform over time. Multiple ports provide different angles for lymphadenectomy. Moreover, the surgeon can directly visualize the target tissue (subcarinal lymph nodes) when performing three-port SIC lobectomy because the camera port in the posterior axillary line at the sixth or seventh intercostal space can provide more direct and accurate visualization of the target structure in SIC VATS (Fig. 2) than can the observation port created in the lower intercostal space in the conventional three-port MIC VATS. Therefore, more subcarinal lymph nodes were harvested and more radical systemic mediastinal lymphadenectomies were performed in patients treated with the modified VATS technique.

There were two factors contributing to the shorter surgery duration and the reduced intraoperative blood loss for SIC VATS compared with conventional three-port MIC VATS. One factor was that positioning patients in the lateral decubitus position allows the surgeon to have direct visualization and to perform direct instrumentation and dissection from the sixth intercostal space, which is very similar to thoracotomy. A major factor was that no extra instrument was needed, and an operating angle between the instruments still existed. Thus, there are no obstacles in converting from the MIC to the SIC approach for thoracic surgeons with experience in minimally invasive surgery. Additionally, 20 cases were initially excluded to account for the learning curve effect in this study. With increasing experience, the operative time and blood loss associated with three-port SIC lobectomy decreased.

Our approach for SIC VATS lobectomy in this study has the following characteristics: (a) a single intercostal space is used to complete the operation; (b) accurate anatomic visualization is achieved for subcarinal lymph nodes removed from the posterior axillary line at the sixth intercostal space, which is highly similar to thoracotomy; (c) no extra instrument is needed, and the operating angle between the instruments still exists; and (d) it is feasible for thoracic surgeons with experience in minimally invasive surgery to convert from the MIC approach to the SIC approach.

There are several limitations to this study. First, the major limitation is the retrospective design and the decision to perform MIC or SIC is based on surgeon preference. Although propensity-matched analysis may reduce some selection bias, inherent selection bias and unknown confounding variables cannot be eliminated. Second, the experience in this study was from two experienced surgeons in a high-volume thoracic department. Third, the theoretical benefit of SIC versus MIC with regard to pain management could not be investigated in this study, as many changes in postoperative management (analgesics, chest tube and intercostal nerve damage) complicated controlling for biases, even with propensity matching. Thus, a prospective and randomized multicentre study would be the gold standard to demonstrate the superiority of one procedure over another, and such a study will be conducted in the near future.

Based on the presented data, we conclude that three-port SIC lobectomy and MIC lobectomy are both acceptable and offer similar oncologic surgery for the treatment of NSCLC. Thus, three-port SIC provides another choice for the treatment of NSCLC, and surgeons can freely choose with strong self-confidence according to preference.