Effect of Surgery on The Prognosis of Stage (cid:0) B non-Small Cell Lung Cancer

For patients with stage IIIB non-small cell lung cancer (NSCLC), the current recommended immunotherapy maintenance treatment after concurrent chemoradiotherapy is not demonstrated whether such patients can be benet from surgical treatment. We collected clinical data from patients with stage IIIB NSCLC from the surveillance, epidemiological, and nal results (SEER) databases. Kaplan-Meier analysis and Cox proportional hazards model were used to analyze the effect of surgical treatment on the survival of stage IIIB NSCLC. We use surgery as a variable to divide into surgical and non-surgical groups. Between the two groups, only age is different (p = 0.017), and there is no signicant difference at race (P = 0.531), gender (P = 0.051), tumor primary site (P = 0.385), tumor differentiation grade (P = 0.103), T stage (P = 0.094), and N stage (P = 0.071). Univariate and multivariate analysis showed that gender, tumor differentiation, stage and primary site had an effect on overall survival. Kaplan-Meier analysis showed that the median survival time of stage IIIB NSCLC in the surgical group (13 months, 95% CI (7.400 to 18.600 months)) was higher than that in the non-surgical group (3 months, 95% CI (2.482 to 3.518 months)) was signicantly prolonged by 10 months (P < 0.001). The Cox proportional hazard modelled analysis showed that the overall survival of stage IIIB NSCLC was 2.395 times the 95% CI (1.77 to 3.241 months) for the non-surgical group (P < 0.001). We conclude that some patients with stage IIIB NSCLC who can tolerate surgery may benet from surgery.


Introduction
Cancer has become the leading cause of death in China, and lung cancer is the most common cancer and the leading cause of cancer death (1). More than 85% of these lung cancers are non-small cell lung cancer (2). The main histological types of non-small cell lung cancer include about 50% of lung adenocarcinoma and approximately 40% of lung squamous cell carcinoma (3). The early diagnosis of lung cancer depends on the ndings of CT physical examination (4)(5)(6), but most patients ignore the importance of annual physical examination. When coughing, hemoptysis and other symptoms appear, the lung cancer has reached the advanced stage. Surgical treatment is mainly used for stage ~ A NSCLC, combined with radiotherapy and chemotherapy as the standard treatment (7,8). For stage IIIB NSCLC, surgical resection is currently not recommended, and the 5-year survival rate is only about 15%(9); the PACIFFIC study recommends Durvalumab immunotherapy maintenance treatment after concurrent chemoradiotherapy. Although the survival time is extended, only about 10% of patients Bene t from immunotherapy (10,11). Surgical treatment is an important method for the treatment of malignant tumors. With the deepening of research and the updating of ideas, many advanced cancer patients, such as IV stage colorectal cancer and IV stage breast cancer, can be treated by surgery on the basis of transformation therapy (12), and achieved good results. The treatment effect of stage IIIB NSCLC is currently not satisfactory. Whether such patients can bene t from surgical treatment has not been fully demonstrated.

Materials And Methods
Data Source: The SEER database is one of the most representative large tumor registration databases in North America, and it collects a large amount of data related to evidence-based medicine. The database records information on the morbidity, mortality, and prevalence of millions of patients with malignant tumors in some states and counties in the United States.
Patient selection: We obtained patient information for NSCLC from 2004 to 2015 from the Surveillance, Epidemiology and Final Results (SEER) database. First, remove the data that are unknown about race, tumor differentiation and radiotherapy and chemotherapy information. Then ensure the integrity of the data source and the reliability of the follow-up. Finally, the data of stage B NSCLC were selected according to the 7th edition of the American Joint Commission on Cancer (AJCC) (13). We were divided into surgical group (n = 67) and non-surgical group (n = 589) according to the presence or absence of surgical treatment. (Fig. 1) Statistical analysis. Chi-square test and Fisher exact test were used to analyze the differences between the groups. Survival analysis was performed using the Kaplan-Meier method. The Cox regression proportional hazard model was used to assess the effect of prognostic factors on cumulative survival time. All tests were two-sided, and P < 0.05 was considered statistically signi cant. Statistical analysis was performed using SPSS statistics software version 25

Result
This study selected 656 eligible medical records from the surveillance, epidemiological and nal results (SEER) database for a controlled study. They were divided into the surgical group (n = 67) and the nonsurgical group (n = 589) according to the presence or absence of surgical treatment. We compared the basic characteristics of the two groups, including age, gender, race, degree of tumor differentiation, primary tumor site, T stage, and N stage. Except for the differences between the age groups (P = 0.017), the characteristics of the other groups were balanced. (Table 1) Then we used Kaplan-Meier to analyze the effect of each independent factor on cumulative survival and draw a survival curve (Fig. 2). The K-m curve shows that there is a statistically signi cant difference between women and men in the gender group (P = 0.007). The tumors in Main bronchus were statistically different from Upper lobe, lung, Middle lobe, lung, Lower lobe, lung, Overlapping lesion of lung, Lung, NOS (P = 0.032). There was statistically signi cant difference between well differentiated tumors, poorly differentiated tumors, moderately differentiated tumors, and undifferentiated tumors (P = 0.001). There was a statistically signi cant difference between T1, T2, T3 and T4 (P = 0.001). There was statistical difference between the operation group and the non-operation group (P < 0.001). Generally speaking, age (P = 0.432) and N stage (P = 0.181) play important roles in the prognosis of patients with NSCLC(14-16).
Our results are inconsistent with the traditional one, which may be caused by the difference in age between the groups and the too small sample size. Age, gender, race, degree of tumor differentiation, primary site, stage T and stage N, and median survival time with or without surgery (Fig. 3).It can be seen that the median survival time of females is 5 months 95%CI (4.026 to 5.974 months), which is higher than that of males 3 months 95%CI (2.423 to 3.577 months).  The cumulative survival time for patients after the previously reported surgery has been signi cantly extended. We further analyzed the effects of preoperative radiotherapy, postoperative radiotherapy, preoperative plus postoperative radiotherapy on the cumulative survival time of patients with stage IIIB NSCLC. Because the number of patients undergoing stage IIIB surgery was originally small, the sample size was too small Not conducive to further analysis (Table 3). However, in general, most patients choose radiotherapy before or after surgery. The low degree of tumor differentiation and low T stage have signi cantly improved the survival rate of lung cancer, which is consistent with the results of Goldstraw P, Gester F and others (18)(19)(20). What makes us interesting is that age and N stage are supposed to be closely related to the prognosis, but they are not consistent with our results. It may be due to the difference between the age groups at the time of the initial matching grouping that led to the difference in the nal results. The N stage may be due to the small number of our medical records and our IIIB medical records are mainly concentrated in the N2, N3 stages, resulting in no difference in results. We are pleased that the cumulative survival time of patients in the surgical group is signi cantly longer than that in the non-surgical group, which provides clinical guidance for patients with stage IIIB NSCLC to undergo surgery. As we further study the impact of timing of radiotherapy in the surgical group on patient prognosis. Due to too few clinical data to proceed, only descriptive analysis was performed. Most patients focused on preoperative and postoperative radiotherapy.
For our study, there are also de ciencies. First, there is a difference in the age between the groups when grouping. Second, the radiotherapy and chemotherapy information in the seer database are not speci c.
In addition to the baseline characteristics we described above, the seed database does not record other health conditions of patients. In the end, there is confounding bias in any research, and we can only minimize it.
In   Forest plot and 95% CI based on single factor effect on median survival time of stage IIIB NSCLC.