Investigation of the non‐small cell lung cancer patients with bronchus involvements: A population‐based study

Abstract Background We aimed to explore the prognostic differences among T1‐4N0‐2M0 non‐small cell lung cancer (NSCLC) patients with bronchus involvements and to validate the T category of these patients in an external cohort. Methods Univariable and multivariable Cox analysis was performed to determine the prognostic factors. Kaplan–Meier method with a log‐rank test was used to compare overall survival differences between groups. Propensity score matching method was used to minimize the bias caused by the imbalanced covariates between groups. Results A total of 169 390 eligible T1‐4N0‐2M0 NSCLC cases were included. There were 2354, 3367, 1638, 75, 87 585, 42 056, 19 246, and 13 069 cases in the group of superficial tumors of any size with invasive component limited to bronchial wall (T1‐bronchus), tumors involving main stem bronchus ≥2 cm from carina (T2‐main bronchus [≥2 cm]), tumors involving main stem bronchus <2 cm from carina (T2‐main bronchus [<2 cm]), tumors with carina invasion (T4‐carina), T1, T2, T3, and T4, respectively. Multivariable Cox analysis indicated that T1‐bronchus patients had the best prognosis; T2‐main bronchus (≥2 cm) and T2‐main bronchus (<2 cm) patients had similar prognosis both in the entire cohort and in several subgroups. Survival curves showed that T1‐bronchus and T1 patients had similar survival rates; the survivals of T2‐main bronchus patients regardless of the distance from carina were comparable to those of T2 patients, and the survivals of T4‐carina patients were also similar to those of T4 patients. Conclusions Our results validated and supported the current T category for the patients with bronchus involvements, which might provide certain reference value for the revisions of T category in the next version of the tumor‐node‐metastasis stage classification.


| INTRODUCTION
Lung cancer is the leading cause of cancer-related mortality worldwide, 1,2 which consists of non-small cell lung cancer (NSCLC) and small cell lung cancer. 3,4 The tumor-node-metastasis (TNM) stage classification, an important indicator proposed to assess disease severity and estimate individual survival, has been updated every few years. [5][6][7][8] Bronchus involvements is a non-sized T descriptor in the latest NSCLC TNM stage classification. 5,6 Superficial tumors of any size with invasive component limited to bronchial wall and with or without proximal extension to the main stem bronchus are assigned to T1 category (T1-bronchus). 5,6 Tumors involving main stem bronchus ≥2 cm from carina are assigned to T2 category (T2-main bronchus [≥2 cm]), and those involving main stem bronchus <2 cm from carina are also classified as T2 category (T2-main bronchus [<2 cm]), 5,6 which is a T3 descriptor in the 7th edition of the TNM stage classification. 7,8 In addition, tumors with carina invasion are defined as T4 category (T4-carina). 5,6 The current TNM stage classification must be applicable to individual databases, and no external validation study has been performed on the T category of lung tumors with bronchus invasion. Herein, the current study analyzed the data of NSCLC with bronchus involvements deposited in the Surveillance, Epidemiology, and End Results (SEER) program, in an effort to figure out the prognostic disparities and further validate the current T category of these patients.

| Included patients
The data of lung malignancies between 2004 and 2016 were extracted from the SEER program (http://seer. cancer.gov). The inclusion criteria mandated that patients were diagnosed with NSCLC, and the TNM stages of tumors were T1-4N0-2M0. The exclusion criteria were age <18 years and unavailable survival information. The eligible patients were further separated into eight groups: T1-bronchus (CS Extension code 110), T2-main bronchus (≥2 cm) (CS Extension code 220), T2-main bronchus (<2 cm) (CS Extension code 500), T4-carina (CS Extension code 250), remaining T1, remaining T2, remaining T3 and remaining T4 group. The patient selection flowchart is showed in Figure 1. We further distinguished pathologically confirmed NSCLC from clinically confirmed ones based on the CS Tumor Size/Ext Eval codes (2.3 and 6) and the CS Lymph Nodes Eval codes (2.3 and 6).

| Ethic
The author was authorized to retrieve the NSCLC data deposited in the SEER program with the reference number 12962-Nov2019. This study was conducted in accordance with Helsinki declaration. This study was waived of personal-informed consent forms because only anonymous data were used.

| Univariable and multivariable Cox analysis
Regarding the NSCLC patients with bronchus involvements, univariable Cox analysis indicated that age, gender, race, married status, insurance, tumor location,
After PSM, in the T1-bronchus and T1 matched cohort, bronchus invasion did not confer an inferior survival when compared with the remaining T1 patients (P = 0.181; Figure 3B). In the T2-main bronchus (≥2 cm) and T2 matched cohort, the survival curve of T2-main bronchus (≥2 cm) patients nearly overlapped with that of T2 patients (P = 0.892; Figure 3C). In the T2-main bronchus (<2 cm) and T2 matched cohort, it is interesting to observe that these two groups of patients had similar survivals (P = 0.309; Figure 3D). In the T4-carina and T4 matched cohort, the survivals of T4-carina patients were better than those of T4 patients, but the difference was not statistically significant (P = 0.103; Figure 3E).

| DISCUSSION
Previously, many researches has been focused on the validations and refinements of the current 8th edition of the NSCLC TNM stage classification. [10][11][12][13][14] To date, there was no validation study focused on the T category of the NSCLC patients with bronchus involvements. Herein, we analyzed the data of NSCLC patients with bronchus involvements and aimed to externally validate the current TNM stage classification. Our study demonstrated that T1-bronchus patients had the best prognosis; T2-main bronchus (≥2 cm) and T2-main bronchus (<2 cm) patients had similar prognosis both in the entire cohort and in several subgroups. Survival curves further showed that T1-bronchus and T1 patients had comparable survival rates; the survivals of T2-main bronchus patients regardless of the distance from carina were similar to those of T2 patients, and the survivals of T4-carina patients were also similar to those of T4 patients. Our results validated and supported the current TNM stage classification, and we proposed that it might not be necessary to modify T categories of the NSCLC patients with bronchus involvements in the next version of the NSCLC TNM stage classification.
In the proposals for the revisions of the T descriptors in the 8th edition of the lung cancer TNM classification, 6 Rami-Porta et al. 7 reported that tumors involving main bronchus ≥2 cm from carina are still defined as T2 category, and those involving main bronchus <2 cm from carina, a T3 descriptor in the 7th edition of the TNM classification, 7,8 should be downstaged from T3 to T2 category. Regarding patients with tumors of any size with invasive component limited to bronchial wall and patients with tumors with carina involvement, detailed descriptions were not provided. There is a dearth of study focused on the patients with bronchus involvements partially because of the small number of cases. Therefore, external validations are needed.
With the advantage provided by the large data deposited in the SEER program, the current study was sufficient to investigate the prognostic differences among these patients. Our results showed that the survivals of T1-bronchus patients were comparable to those of T1 patients; T2-main bronchus patients regardless of the distance from carina and T2 patients also had similar survival rates; and the survivals of T4-carina patients were similar to those of T4 patients. To the best of our knowledge, the current study is the first comprehensive analysis of the prognosis of patients with bronchus involvements. Our results supported the current T category of these patients and might provide certain reference value for the proposals of the T category in the next version of the NSCLC TNM stage classification.
Our study had several limitations. First, a relatively small number of cases in the T4-carina group (75 cases) made it hard to draw a conclusion with strong statistical power. Therefore, large sample studies are encouraged to confirm the conclusions of this study. Second, in the era of target therapies and immunotherapies, whether our conclusions were applied to the patients received tyrosine kinase inhibitors or immune checkpoint inhibitors is unknown. It is interesting to further explore the prognosis disparities among these patients who had received novel therapies. However, the information about treatments regimens, dosage, and timing is unavailable in the SEER program. We planned to collected the related information in our institution and further validate our results in future. In addition, our results were exploratory because a large number of hypothesis tests were performed and multiplicity might exist. At last, this is a retrospective study, and bias cannot be avoided.
In conclusion, our results validated and supported the current T category of the patients with bronchus involvements, which might provide certain reference value for the revisions of T category in the next version of the NSCLC TNM stage classification.

AUTHOR CONTRIBUTIONS
Conception and design: Gang Wang. Administrative support: Yong-Qiang Ye. Provision of study materials or patients: Bao-Long Xie, Xiang-Min Lai, and Sheng-Peng Zhong. Collection and assembly of data: Gang Wang. Data analysis and interpretation: Gang Wang. Manuscript writing: All authors. Final approval of manuscript: All authors.