Surgery for Stage IV Non-Small Cell Lung Cancer: Lobectomy or Sub-lobar Resection?

Background : A survival benet was observed in metastatic non-small cell lung cancer (NSCLC) patients that underwent operation. But no evidence to support whether lobectomy would further prolong these patients’ live than sub-lobar resection. Methods: Patients that underwent primary tumor resection with metastatic NSCLC were identied from the Surveillance, Epidemiology and End Results (SEER) database and then divided into lobectomy and sub-lobar resection groups. A 1:1 propensity score matching (PSM) was performed to balance characters. Cancer specic survival (CSS) was estimated. Results: A total of 24,268 patients with metastatic NSCLC were identied; 4,114 (16.95%) received primary tumor surgery, of which 2,045 (49.71%) underwent lobectomy and 1,766 (42.93%) underwent sub-lobar resection. After PSM, 644 patients in each group were included. Lobectomy was independently correlated with longer median CSS time (HR=0.70, 95% CI 0.61-0.80, P<0.001). The 1, 2 and 3-year survival rate after PSM also favored the lobectomy group. However, no signicant survival difference was found in wedge resection and segmentectomy (HR=0.96, 95% CI 0.70-1.31, P=0.490). The 1, 2 and 3-year survival rate after PSM also showed no difference within the sub-lobar group. We explore whether lymph node dissection would provide a further survival benet for stage IV NSCLC patients. According to the multivariate Cox analysis of the matched population, lymph node dissection was independently associated with better CSS (HR=0.76, 95% CI 0.66-0.88, P<0.001) and OS (HR=0.74, 95% CI 0.65-0.86, P<0.001). We conrmed this result in different types of surgery and found lymph node dissection group persist to have better survival outcomes both in lobectomy group and sub-lobar resection population. According to T4 of lobectomy than sub-lobar

However, no signi cant survival difference was found in wedge resection and segmentectomy (HR=0.96, 95% CI 0.70-1.31, P=0.490). The 1, 2 and 3-year survival rate after PSM also showed no difference within the sub-lobar group. We explore whether lymph node dissection would provide a further survival bene t for stage IV NSCLC patients. According to the multivariate Cox analysis of the matched population, lymph node dissection was independently associated with better CSS (HR=0.76, 95% CI 0.66-0.88, P<0.001) and OS (HR=0.74, 95% CI 0.65-0.86, P<0.001). We con rmed this result in different types of surgery and found lymph node dissection group persist to have better survival outcomes both in lobectomy group and sub-lobar resection population. According to subgroup analysis, except for stage T4 and brain metastasis patients, all subtype of patients would bene t more from lobectomy than sub-lobar resection.
Conclusions: Lobectomy brings survival bene t in metastatic NSCLC patients compared with sub-lobar resection.

Background
Lung cancer is one of the leading causes of cancer death worldwide, [1] approximately 85% subtype patients of which are non-small cell lung cancer (NSCLC) [2] The 5-year survival rate of these patients was only 4-6% and 40% of which were diagnosed as stage IV disease at the rst visit to hospital. [3] The recommended treatment for stage IV NSCLC is systemic therapy (chemotherapy, molecular targeted therapy or immune therapy). [4] Traditionally, stage IV treatment strategy for NSCLC have not included curative-intent local therapy (surgery or radiation), given therapeutic goals which have focused on disease control, palliation and optimization of life quality. But more recently several clinical studies have shown that local consolidative therapy may be bene cial for stage IV non-small cell lung cancer (NSCLC) patients and improve their survival. [5,6] Daniel et al. [7] conducted the rst phase II clinical trial considering the effects of local consolidative therapy in combination with systemic treatment. The Lobectomy and systematic lymph node dissection is the cornerstone for early stage of NSCLC. Recently, evidences from real world studies showed primary tumor resection would also be bene cial for stage IV NSCLC patients and improve their survival. [8][9][10] However, no study explore which surgery type (lobectomy or sub-lobar resection) would provide more survival bene t for these patients.
To address this unresolved issue, we performed a population-based study to examine whether lobectomy or sub-lobar resection in stage IV NSCLC patients who underwent surgery with the most bene t.

SEER Database
The Surveillance, Epidemiology and End Results (SEER) database is a national population-based reporting system that collects tumor-related data, covering approximately 28% of the US population. [11] The SEER data is publicly available for studies of cancer-based epidemiology and survival analysis. We received permission to access the data used for this research (SEER-Stat username: 11136-Nov2018).
Cases of lung cancer (C34.0-34.9) diagnosed from 2004 to 2016 were extracted from the SEER database (SEER-Stat 8.3.6) according to the site code classi cations. This range was selected because the American Joint Committee on Cancer (AJCC) TMN stage and Collaborative Stage (CS) information became available in 2004. We reclassi ed the TNM stage according to AJCC 8th edition. Patients who underwent primary tumor resection diagnosed at stage IV and histologically con rmed as NSCLC were enrolled. Patients were excluded if the surgery to primary site record was unknown. Other exclusion criteria were as follows: less than 18 years old, unknown TNM stage, unknown time of survival, unknown treatment modality, not the rst tumor and not only one tumor.

Statistical analysis
The study sample was divided according to surgical resection strategy for primary tumor: Lobectomy versus sub-lobar resection. The propensity-score matching (PSM) generated from the logistic regression were performed to minimize the differences in confounding variables and facilitate matching patients in the two treatment groups (R software version 2.15.1, https://cran.r-project.org/). Variables that could in uence the outcomes of treatment were used to generate a propensity score, including age, gender, histology, TNM stage, differentiation grade and tumor position. Patients were 1:1 matched on the basis of PSM using the nearest-neighbor method on the logit scale. The caliper was set at 0.01. After PSM, standardized mean differences (SMD) before and after PSM were calculated. Confounding variables was considered comparable when SMD below 0.10.
Overall survival (OS), cancer speci c survival (CSS) and survival months were extracted from the SEER database. OS was the time from diagnosis to death from any cause; living patients were excluded at the time of last recording. CSS was calculated from the date of diagnosis to the date of cancer speci c death. OS and CSS were estimated by the Kaplan-Meier (K-M) method and compared with the log-rank test. Univariate and multivariate Cox proportional hazard regression was used to determine independent prognostic factors. Hazard ratios (HRs) were calculated with 95% con dence interval (CI).
The normality of the data was assessed by the Shapiro-Wilk test. Continuous variables were given as mean and standard deviation. Student's t-test or Mann-Whitney test was performed to compare differences between groups with continuous variables. Distribution of categorical variables was presented as a count and percentage. The χ 2 test or Fisher exact test for small samples was used to compare categorical variables. Subgroup analysis was conducted according to different clinical types of population. Statistical analysis was performed with SPSS 24.0 (IBM Corp., Armonk, NY, USA), statistical tests were two-sided, and P < 0.05 was considered statistically signi cant.

Results
Demographic characteristics before and after PSM In total, 476,757 NSCLC patients were identi ed from 2004 to 2016 in the SEER database, of which, 24,268 stage IV NSCLC patients met the inclusion criteria, the screening process is shown in Fig. 1 1,766 (42.93%) received sub-lobar resection, others underwent pneumonectomy and biopsy, which were not included in this analysis. Obvious differences in age, histology, differentiation, tumor position, TNM stage, radiation chemotherapy and distant surgery were noted between the lobectomy and sub-lobar resection groups. (Table 1) This indicated that the baseline characteristics of the two groups were not balanced. Speci cally, lobectomy group was associated with lower T stage patients. After the 1:1 PSM, 1,288 stage IV NSCLC patients treated with lobectomy (n = 644) or sub-lobar resection (n = 644) were enrolled in the survival analysis. Baseline characteristics were all well balanced (Table 1). Logistic analysis was performed to evaluated which variables were associated with lobectomy ( Table 2). Compared with patients under 60 years old, patients over 75 years old tend to receive more sub-lobar resection than lobectomy (HR = 0.71, 95% CI 0.55-0.91, P = 0.007). In addition, T4 patients received less lobectomy compared with T1 or T2 patients.

Subgroup analysis
We further explored whether lobectomy was associated with better survival outcomes in different subtypes of populations. Both CSS (Fig. 5A) and OS (Fig. 5B) outcomes showed similar results. Except for stage T4 and brain metastasis patients, all subtype of NSCLC patients would bene t more from lobectomy than sub-lobar resection.

Discussion
There is growing evidence to support the value of primary tumor resection for advanced NSCLC patients in recent years.
However, no study with the primary aim to compare survival outcomes in relation to surgical strategies (lobectomy or sub-lobar resection) were found. Few evidences for best surgery strategy could be referred when surgeons performed operation on metastasis NSCLC patients. This is a population-based propensity score matching study to assess the value of lobectomy or sub-lobar resection in stage IV NSCLC patients. The results indicated that lobectomy can independent signi cantly improve both cancer speci c and overall survival rate compared with sub-lobar resection.
Subgroup analysis revealed that comparing with sub-lobar resection, except for stage T4 patients, all subgroup patients would bene t from lobectomy.
According to National Comprehensive Cancer Network (NCCN) guidelines, patients who have single brain or adrenal metastasis, but the primary tumor is T1-2, N0-1 or T3, N0, local treatment of the metastasis followed by resection of the primary tumor is recommended. [12] The guidelines also recommend contralateral lung nodule can be resected. surgery type for stage IV NSCLC patients remains to be answered. Based on SEER database and using propensity score matching analysis, our study found that lobectomy should be considered in majority surgery for stage IV NSCLC, except for stage T4 patients, and lymph node dissection may further provide survival bene t.
The role of surgery for advanced stage NSCLC is always commonplace for diagnostic and palliative purposes. [14,15] As personalized medicine has taken a more prominent role in the care of advanced NSCLC patients and treatment decisions are now based on histological molecular subtypes, surgeons are more commonly performing operations to acuqire adequate tissue for enabling detailed subtyping of NSCLC. Sub-lobar resection is enough to get adequate biopsy for diagnostic and palliative purposes; however, lobectomy and lymph node dissection means a more aggressive comprehensive reduction of primary tumor, which might be associated with a better survival outcome. Mitchell et al. [5] found that T stage is also the signi cant prognostic factor in stage IV NSCLC patients, and comprehensive local consolidative therapy would bring survival bene t both according to OS (HR = 0.67, 95% CI 0.46-0.97, P = 0.034). These evidences also supported a more aggressive and thorough local treatment strategy for metastasis NSCLC.
Oligometastatic NSCLC is one of the most proper indications to receive local consolidative therapy. [ the growth of distant micrometastatic disease can also be decreased. Notably, these mechanisms are not mutually exclusive, and more than one could contribute to the bene ts of lobectomy for advanced NSCLC.
Although this population-based analysis suggested the long-term survival bene cial clinical e cacy of lobectomy and lymph node dissection for patients with stage IV NSCLC, the data should be interpreted with caution as: rst, information on preoperative comorbidities for patients were lacking which might lead to a selection bias for the treatment choice; second, detailed information and distribution of patients' metastatic disease is not available, and this would impede us to explore whether oligometastasis disease will affect the bene t of lobectomy; third, data on systemic therapies were not available. This lack of information regarding target therapies and immunotherapy is a limitation of the current study; Although, the number of stage IV patients that underwent primary tumor resection is limited in each medical center, to explore the curative effect of lobectomy for these patients are challenging, SEER database is the only comprehensive population-based database with open access worldwide providing the most ideal approach to study the survival of such patients. Prospective randomized trials are needed to further validate the bene t of surgery types in metastasis NSCLC patients.

Conclusions
In conclusion, our study showed lobectomy for primary tumor may further improve the survival of stage IV NSCLC patients compared with sub-lobar resection. Future clinical trials should focus on optimal candidates for such surgery in metastasis NSCLC patients.
Program (Grant No. 2017YFC0907903 & 2017YFC0112704) and the Guangdong high level hospital construction "reaching peak" plan.
The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.