The Diameter of Bronchial Resection Margin is Associated With Postoperative Lung Metastasis and Long-term Survival

Background To investigate the signicance of the diameter of bronchial resection margin (DBRM) on the postoperative lung metastasis and long -term survival of patients with primary lung cancer. Methods We retrospectively analyzed the data of 1844 patients with primary lung cancer between January 2006 and December 2010 after surgery. Patients were divided into DBRM ≤ 1 cm group (826 patients) and DBRM>1 cm group (1018 patients). Propensity score matching was used to reduce grouping selection bias. Furthermore, we divide the 974 patients who had denite rst metastasis site into lung metastasis group (283 patients) and other metastasis group (691 patients), and analyzed related risk factors and prognosis of metastasis. Disease-free survival and overall survival were the study end points.


Introduction
It is important to obtain a cancer-free bronchotomy margin during lung cancer surgery, and the availability of a "safe margin" could improve postoperative survival (1). Tumor spread through air spaces (STAS) was reported as a novel poor prognostic factor in the latest World Health Organization (WHO) classi cation (2). In recent studies, STAS in patients with lung cancer was shown to be associated with high recurrence rates and poor survival after surgical resection (3,4). We consider that the size of DBRM may affects STAS, which in turn affects the prognosis. So far, there were no reports on how to control DBRM during surgery and no studies con rmed the effects of DBRM on STAS and prognosis. The aim of our study was to investigate the effect of DBRM on postoperative lung metastasis and long-term survival of patients with primary lung cancer.

Follow-Up
The patients were followed up to death or until December 31, 2019 by regular outpatient review, letter visit, telephone call, etc. Postoperative routine examination included tumor markers of lung cancer, chest X-ray, chest CT, brain CT, brain MRI, abdominal CT, bone scan, abdominal ultrasound, etc. The DFS was calculated from the date of surgery to recurrence or metastasis and the OS was calculated from the date of surgery to death or the end point date. The DFS and OS were calculated by month. By the end of follow-up, 91 patients in the DBRM ≤ 1 cm group and 111 patients in the DBRM > 1 cm group lost contact after surgery. We compared the 202 lost contact ones with 1642 ones who had complete follow-up information on basic clinical characteristics. The result showed that there was no statistically difference between the two groups (P > 0.05, Table S1).

Study Design
The primary outcome was DFS and OS. The DBRM was divided into DBTM ≤ 1 cm group and DBTM > 1 cm group. In order to obtain the cutoff value, the DBRM was grouped according to the minimum grouping in uencing factors and the method of calculating the minimum survival P-value (7). Statistical Analysis SPSS (IBM, SPSS statistics version 26.0) software and GraphPad Prism (version 8.0.2) were used for data processing and gure formatting. x 2 test was used for categorical variables, and t test was used for continuous variables. Multivariate analysis was used the logistic regression model to analyze the relation between lung metastasis and risk factors. Survival rate was calculated by Kaplan-Meier method. Univariate and Multivariate prognosis analysis was performed using Cox regression model. In all analyses, two-tailed P < 0.05 was considered statistically signi cant.
Propensity score matching (PSM) PSM was performed using through 1:1 nearest-neighbor matching with a math tolerance value of 0.02 for reducing selection bias between DBTM ≤ 1 cm group and DBTM > 1 cm group (8) . PSM for patients were  calculated by using a multiple logistic regression with the following characteristics: sex, age, smoking  history, tumor location, tumor area, surgical treatment, surgical time, surgical blood loss, pT stage, pN stage, pTNM stage, pathological type, the station of resected lymph nodes, adjuvant chemotherapy and radiotherapy. In order to ensure the continuity and randomness of the statistical data, we did not exclude the lost contact patients because these cases still have basic clinical and pathological information. Furthermore, the SPSS statistical software automatically excludes the ones without survival follow-up information data during survival analysis.

Basic Clinical Characteristics Before and After PSM
A total of 1844 patients were enrolled. There were 1180 (64.0%) males and 664 (36.0%) females. The mean age was 59.7 ± 9.5 years, the minimum age was 23 years, and the maximum age was 89 years. The mean DBRM was 1.36 ± 0.56 cm (0.3 cm to 6 cm). Table 1 shows the basic characteristics of the enrolled patients (n = 1844) grouped by DBRM before and after PSM. In the study cohort, 826 (44.8%) patients were enrolled in the DBRM ≤ 1 cm group and 1018 (55.2%) were enrolled in the DBRM > 1 cm group. Before PSM, there were differences in sex, smoking history, tumor area, surgical treatment, tumor size, pT stage, pN stage, pTNM stage and pathological type (P-values were < 0.001, < 0.001, < 0.001, < 0.001, < 0.001, 0.001, 0.012, 0.002, and < 0.001, respectively) between the two groups. However, there were no signi cant differences between the two groups after PSM (P > 0.05). <0.001 and 0.032, respectively) and OS (P-values; <0.001 and 0.026, respectively) between the two groups before and after PSM. The DFS and OS in the DBRM ≤ 1 cm group were better than those in the DBRM > 1 cm group (Fig. 2).

Metastasis Rate After Surgery
We analyzed the 974 patients who had de nite rst metastasis site. According to this data, we concluded that patients with DBRM > 1 cm had a larger proportion (181 cases, 64.0%) in 283 cases of lung metastasis (Fig. 3A). When quantitatively analyzed the DBRM of different metastatic sites, we found that lung metastasis owned the largest DBRM compared with the other sites (Fig. 3B). Table 3 shows that Lung metastasis after surgery was signi cantly correlated with age, tumor area and the DBRM (P-values: 0.016, 0.014 and 0.013, respectively). These factors were further analyzed by multivariate logistic analysis, and the result revealed that the DBRM was an independent risk factor of lung metastasis (OR, 1.417; 95% CI, 1.062 to 1.890; P = 0.018).

Lung Metastasis and the Other M etastasis Survival Analysis
We compared lung metastasis with the other metastasis for survival and concluded that lung metastasis had a worse OS than the other metastasis (5-year OS, 27.7% vs 30.7%; P = 0.0042; median, 33 vs 40 months) (Fig. 4A). Lung metastasis patients had a worse OS (P < 0.001) than the other metastasis according to the patients without postoperative chemotherapy (Fig. 4B). However, there had no signi cant difference (P = 0.785) in the patients with postoperative chemotherapy (Fig. 4C). Similarly, OS (P = 0.014) for patients without postoperative radiotherapy showed that lung metastasis had a worse prognosis than the other metastasis (Fig. 4D), but there had no signi cant difference (P = 0.947) in the patients with postoperative radiotherapy (Fig. 4E).

Discussion
Metastasis of lung cancer is the leading cause of death in clinical patients (9). For early stage lung cancer, surgery is still the preferred treatment. Surgical methods and adjuvant chemoradiotherapy (10,11,12) have been considered as prognostic factors for lung cancer. Therefore, reasonable surgical methods and appropriate adjuvant treatment are of great signi cance for postoperative long-term survival of patients with lung cancer.
Postoperative bronchial resection margin of lung cancer patients is positive or negative re ecting whether the tumor is completely resected. It has always been considered an important reason for the patient's recurrence. A positive resection margin indicates a poor prognosis (13). Studies have shown that the recurrence and metastasis of cancer patients are closely related to the positive rate of resection margins (14,15). Therefore, ensuring no cancer resection margins during lung cancer surgery is essential for postoperative recurrence and metastasis and long-term survival.
Our retrospective analysis showed that DBRM is an independent prognostic factor for postoperative recurrence and metastasis and long-term survival. We further analyzed that DBRM is an independent risk factor for postoperative lung metastasis and recurrence, which can be used as a reference guide for surgery and prognosis. Current studies have shown that lung cancer radiotherapy and chemotherapy can increase the negative rate of surgical margins, thereby improving postoperative survival (16). We concluded that the postoperative lung metastasis rate and the poor prognosis are high when the DBRM is large. This may be caused by the following reasons: rstly, when the DBRM is large, the margin is also larger, and the possibility of cancer cell survival will increase, although the later pathological results showed negative, but ordinary optical microscopes could not completely diagnose the remaining cancer cells (17); when the margin is large, the surrounding alveolar tissue is more (18), the adjacent alveolar tissue is destroyed more, and the lung tissue containing tumor cells is exposed the area and space are relatively large, so it is easier for tumor cells to spread through the alveolar space to metastasize (19,20); at the same time, intraoperative operations may cause tumors to spread through the alveolar space or increase exfoliated cells, which will increase the postoperative lung metastasis rate; secondly, the analysis results showed that older patients (≥ 65 years) with central lung cancer are also independent risk factors for postoperative lung metastasis (21), this may be due to the declining of lung function in patients with older age or anatomical factors related, which need to explore furtherly.
By comparing the data of lung metastasis and other sites, the results showed that postoperative lung metastasis had a worse prognosis than other sites. There is no signi cant difference in the prognosis of the receiving adjuvant therapy between the two groups, but the non-receiving groups had a signi cant difference, which showed that postoperative adjuvant therapy can indeed further improve the survival rate (22). These results maybe provide a reference for the treatment of patients with large postoperative DBRM.

Limitations
There are some limitations in our study. Firstly, this is a single-center retrospective study, even though propensity score matching was used to control the confounding factors on the outcomes. Secondly, not all patients can be followed up to the site of the rst metastasis, so postoperative lung metastasis may be affected by more factors. Last, for patients with metastases after surgery, we only studied whether they had received adjuvant therapy, but did not conduct research on speci c chemotherapy or radiotherapy regimens, and did not indicate the selectivity of adjuvant therapy regimens. Consequently, long-term effects remain to be studied with larger sample and con rmed with multicenter randomized clinical trial.

Conclusions
In conclusion, the DBRM was an independent risk factor for postoperative lung metastasis and adjuvant therapy could improve long-term survival.
Abbreviations DBRM, the diameter of bronchial resection margin; PSM, propensity score matching; OS, overall survival; DFS, disease-free survival; STAS, tumor spread through air spaces; IASLC, International Lung Cancer