This article is mentioned in:

Introduction

Approximately 2.2 million new lung cancer cases and 1.8 million deaths of patients with lung cancer were reported worldwide in 2020 (1). Although considerable progress has been achieved in lung cancer diagnosis and treatment, the five-year survival rate for this disease remains low (2). The survival rate for lung cancer is closely associated with the disease stage, and the five-year survival rate of patients with stage I lung cancer can reach 85.5–90.2% (3). Therefore, the identification of novel and reliable biomarkers for the early diagnosis of lung cancer is critical.

Cancer stem cells (CSCs) are cells that exhibit stem cell properties, have the ability for self-renewal, and promote the invasion and growth of tumor cells in cancer (4). The presence of CSCs leads to cellular heterogeneity in malignancies as well as innate antibiotic resistance and enhanced invasive capacity, which contribute to cancer progression and metastasis (5). Several conventional CSC markers, including CD44, CD90 and SRY-box transcription factor 2, have been identified in lung cancer; however, these markers are not used universally in clinical practice (6). The mRNA expression-based stemness index (mRNAsi) is a quantitative measure that indicates the degree of similarity between tumors and stem cells (7,8). This index is based on mRNA expression profiles and reflects the stemness characteristics of the transcriptome (9). Previous studies have revealed that lung cancer tissues exhibit higher mRNAsi values than normal tissues and suggest that high mRNAsi values are associated with poor survival in patients with lung adenocarcinoma (LUAD) (10,11).

Although circulating tumor DNA (ctDNA) exhibits potential for the detection of lung cancer, its sensitivity and specificity in the diagnosis of early lung cancer is not optimal (12). Circulating cell-free RNA (cfRNA) is actively secreted by cells or released by apoptotic and necrotic cells into the blood (13). Plasma cfRNAs have been shown to reflect the systemic response to a growing tumor and to indicate the tissue origin of the tumor (14). The release of cfRNA into the blood circulation by early tumors allows upregulated, tumor-specific and tumor-derived RNAs to be identified in the blood at an early stage, which may facilitate the early diagnosis of tumors (15). Studies have indicated that the combined analysis of cfRNA and ctDNA is complementary, improving the early diagnosis of lung cancer (16,17).

As most studies on mRNAsi are typically conducted using tumor tissues, whether cfRNA-based mRNAsi can be used for the early diagnosis of lung cancer remains unclear. In the present study, primary lung cancer tissues, adjacent non-cancerous tissues and lymph nodes were collected for RNA sequencing and analysis to identify gene expression signatures stratified by mRNAsi. Furthermore, paired blood samples were collected for cfRNA sequencing and the determination of a stemness gene expression signature of cfRNAs based on tissue stemness genes. Subsequently, the genes and pathways associated with tissue and blood stemness were analyzed. Finally, correlations of tissue mRNAsi and the cfRNA stemness index with immune infiltration were examined.

Materials and methods

Tissue and blood samples

Tumor tissue, paracancerous tissue (>5 cm away from the tumor edge), peripheral blood and lymph node samples were obtained from 22 patients with stage I–III non-small cell lung cancer (NSCLC) between October 2021 and March 2022 at the Shanghai Chest Hospital, Shanghai Jiao Tong University School of Medicine (Shanghai, China). None of the patients had received any chemotherapy, targeted therapy or immunotherapy before surgery. The clinicopathological data of the patients, including their sex, age, pathological and molecular pathological characteristics (based on TNM staging) (18), and recurrence, were collected.

The freshly collected tissues were snap-frozen and stored in liquid nitrogen. Blood samples were temporarily stored at 4°C. Plasmapheresis was completed within 4 h of the blood being drawn from the patient, and the plasma was stored immediately in a refrigerator at −80°C.

RNA sequencing of samples

Tissue total RNA was extracted from frozen tissues using a MagMAX FFPE DNA/RNA Ultra Kit (Thermo Fisher Scientific, Inc.) following the standard protocol provided by the manufacturer but skipping the deparaffinization process. The concentration of the RNA was measured by NanoDrop™ 2000 spectrophotometer (Thermo Fisher Scientific, Inc.) and the quality of the RNA was evaluated using TapeStation 4200 system (Agilent Technologies, Inc.), and RNA with DV200 ≥40% was considered qualified. Qualified RNA was reverse-transcribed with random hexamers for first strand synthesize and underwent second strand synthesizing, end modification and polyadenylation, sequencing adapter ligation and library amplification using a Hieff NGS® Ultima Dual-mode RNA Library Prep Kit [Yeasen Biotechnology (Shanghai) Co., Ltd.]. Purified libraries were pooled and sequenced using a NovaSeq 6000 system (Illumina, Inc.) with PE150 mode. Low-quality sequences were filtered out prior to downstream analysis.

Cell free RNA was extracted from plasma samples using SpCap-cfRNA extraction and library preparation technology, which is a modified from the method by Jacobsen et al (19) specifically for low abundance plasma cell free RNA extraction and library preparation. In brief, the SpCap-cfRNA technology utilizes poly-T oligo-coated magnatic beads to capture cell free RNA from the plasma, and after extensive wash, captured RNA is reverse-transcribed on beads to make RNA libraries for sequencing. The cfRNA libraries were pooled and sequenced using NovaSeq 6000 system (Illumina, Inc.) with PE150 mode. The sequencing data were subjected to bioinformatics analysis.

Calculation of mRNAsi

The R package TCGAbiolinks (2.18.0) was used to calculate the mRNAsi of different sample types (20). Stemness signatures were analyzed using Progenitor Cell Biology Consortium stemSig (version 2.18.0) (21). The mRNAsi, a continuous variable, was used to categorise the patients into high- and low-mRNAsi groups for different sample types, based on the median.

Weighted gene co-expression network analysis (WGCNA) and identification of differentially expressed genes (DEGs)

The WGCNA package (version 1.71) (22) was used construct coexpression networks. In this analysis, the Pearson's correlation coefficients of genes were calculated to obtain correlation matrices, and appropriate thresholds calculated by the function pickSoftThreshold in WGCNA were selected to measure the connectivity between genes. Adjacency matrices were converted to topological overlap matrices (TOM), and the corresponding differences (1-TOM) were calculated. Genes with similar expression patterns were then grouped into co-expression modules. First, WGCNA of the tumor tissues was performed. In addition, the DEGs of tumor and paracancerous tissues were screened using the EdgeR package (version 3.34.0) (23) with the criteria of fold change >2 and adjusted P-value <0.05. Correlation analyses were then performed to determine the WGCNA modules most relevant to the mRNAsi of tumor tissues and blood. The tissue and peripheral blood samples were categorized into high and low mRNAsi score groups using the median value as a cutoff. DEGs between the high and low mRNAsi groups were screened using the aforementioned EdgeR package. Pathway enrichment analysis of the identified genes was performed using the Kyoto Encyclopedia of Genes and Genomes (KEGG) orthology-based annotation system (version 3.0) (https://www.genome.jp/kegg/) with default parameters.

Identification of fusion transcripts

The STAR-Fusion tool (version 1.8.1) was used to detect fusion transcripts based on the plasma cfRNA sequencing data (24). In the gene expression analysis, STAR was used to map the sequencing data to the reference genome (hg38) and define the transcript coordinates based on the gene annotation format file from GENCODE (version 27; GRCh38) (https://www.gencodegenes.org/human/release_27.html). Gene expression abundance was quantified as reads per kilobase per million mapped reads by using the Cufflinks package (version 2.2.1) (25). The EdgeR package was used to identify DEGs between the fusion-positive and normal samples.

Correlation between stemness index and immune cell infiltration

The gene expression profiles were uploaded to the CIBERSORT website (https://cibersortx.stanford.edu/) to analyze the immune cells infiltrating the tumor tissue. Using LM22 as a reference, 1,000 random permutation operations were performed, and the results were screened based on a significance threshold of P<0.05. The ggplot2 package (version 3.4.4) (https://github.com/tidyverse/ggplot2) was used to analyze differences in immune cell populations between normal paracancerous tissues and tumors. The packages limma (version 3.54.2) (https://bioconductor.org/packages/limma), reshape2 (version 1.4.4) (https://github.com/hadley/reshape) and ggExtra (version 0.10.1) (https://github.com/daattali/ggExtra) were used to evaluate the relationships between the mRNAsi-related genes and immune infiltrating cells. Spearman's rank correlation tests were used for analysis, and the ggplot2 package (version 3.4.4) (https://github.com/tidyverse/ggplot2) was used to generate correlation heat maps.

Statistical analysis

R software (version 4.1.2) (https://www.r-project.org/) was used to statistically analyze all data. Differences in mRNAsi values between different types of matched samples were assessed using the Friedman test followed by Wilcoxon signed-rank tests with Bonferroni correction. An unpaired t-test was used to analyze the statistical significance of differences in normally distributed variables between two groups, and the Wilcoxon rank-sum test was used to compare non-normally distributed variables between two groups. Kaplan-Meier (K-M) analysis and log-rank tests were used to compare survival between two groups. P<0.05 was considered to indicate a statistically significant result.

Results

Clinical data

A total of 22 patients were included in the present study (Table I). The age of the patients ranged from 46 to 74 years, with an average of 63.8 years. Among all patients, 63.6% were women, 90.9% had adenocarcinoma, 86.3% had stage I or II cancer, and 40.9% had poorly differentiated tumors.

Table I. Clinical characteristics of the enrolled patients.

Table I.

Clinical characteristics of the enrolled patients.

Characteristics	Values
Mean age, years	63.8
Sex, n (%)
Female	14 (63.6)
Male	8 (36.4)
Histology, n (%)
Adenocarcinoma	20 (90.9)
Squamous cell carcinoma	2 (9.1)
Differentiation degree, n (%)
Poor	9 (40.9)
Moderate	13 (59.1)
T stage, n (%)
T1	11 (50.0)
T2	7 (31.8)
T3	3 (13.6)
T4	1 (4.5)
N stage, n (%)
N0	20 (90.9)
N1	0 (0.0)
N2	2 (9.1)
N3	0 (0.0)
M stage
M0	22 (100.0)
M1	0 (0.0)
pTMN stage
I	16 (72.7)
II	3 (13.6)
III	3 (13.6)

mRNAsi correlates with the prognosis of patients with lung cancer

By evaluating the median value of mRNAsi in the peripheral blood and cancer tissue samples of patients with lung cancer, the patients were categorized into high and low mRNAsi groups for each sample type. K-M survival analysis revealed no significant difference in overall survival between the high and low mRNAsi groups for peripheral blood or cancer tissue samples (Fig. 1). This may be attributed to the short duration of follow-up, as the median follow-up time was only 820 days. To investigate the relationship between mRNAsi and disease recurrence, the recurrence rates of patients in the groups over two years were calculated. In both the peripheral blood and cancer tissue samples, the two-year recurrence rate of patients in the high mRNAsi group was 27.3%, whereas that of patients in the low mRNAsi group was 18.2%. Notably, when the peripheral blood and cancer tissue samples both had high mRNAsi values, the two-year recurrence rate of the patients was 18.2%, which was twice that of patients with low mRNAsi values.

Figure 1. Relationships between mRNAsi values and overall survival in patients with non-small cell lung cancer assessed using K-M analysis. K-M comparison of high and low mRNAsi groups in (A) peripheral blood and (B) tumor tissues. mRNAsi, mRNA expression-based stemness index; K-M, Kaplan-Meier; P-high, high peripheral blood mRNAsi; P-low, low peripheral blood mRNAsi; T-high, high tumor tissue mRNAsi; T-low, low tumor tissue mRNAsi.

Comparison of mRNAsi values in peripheral blood and tissues

To investigate the association between tumor stemness and metastasis, the mRNAsi values of paracancerous tissues, cancer tissues, lymph nodes and peripheral blood were compared. The results revealed that the mRNAsi value of tumor tissues was significantly higher than that of paracancerous normal tissues, and the mRNAsi value of lymph nodes was significantly higher than that of tumor tissues. Also, the mRNAsi value of cfRNA in peripheral blood was higher than that in cancer tissues, albeit not significantly, and was significantly higher than that in paracancerous tissues (Fig. 2).

Figure 2. Comparison of mRNAsi among different matched patient samples. All groups were compared by the Friedman test, and comparisons between two groups were performed using Wilcoxon signed-rank tests with Bonferroni correction. **P<0.01, ***P<0.001. N, normal paracancerous tissues; T, tumor tissues; P, peripheral blood; L, lymph node.

Genes and pathways associated with lung cancer development

WGCNA of the tumor tissues revealed that several genes, including ALKBH3, MED19, protein tyrosine phosphatase receptor type C associated protein (PTPRN), ERAL1 and PPME1 are associated with lung cancer (Table SI). KEGG analysis indicated that these genes are enriched in the pathways ‘steroid biosynthesis’ and ‘other types of O-glycan biosynthesis’ (Fig. 3A). The DEGs associated with tumorigenesis, identified by the comparison of gene expression in tumor tissue with that in paracancerous tissue, are presented as a volcano plot (Fig. 3B; Table SII). Upregulated genes are enriched in the ‘cell cycle’, ‘protein digestion and absorption’ and ‘metabolism of xenobiotics by cytochrome P450’ pathways (Fig. 3C). By contrast, downregulated genes are enriched in the ‘neuroactive ligand-receptor interaction’, ‘calcium signaling pathway’ and ‘vascular smooth muscle contraction’ pathways (Fig. 3D).

Figure 3. Genes and pathways associated with lung cancer. (A) Pathway enrichment analysis of the genes found to be associated with lung cancer by weighted gene co-expression network analysis. (B) DEGs associated with lung tumorigenesis. Pathways enriched by (C) upregulated DEGs and (D) downregulated DEGs. DEGs, differentially expressed genes; TvsN, tumor tissues vs. normal paracancerous tissues; Up, upregulated genes; Down, downregulated genes; NS, not significant; P-adj, adjusted P-value.

Genes and pathways associated with the mRNAsi of tumor tissues and peripheral blood

WGCNA and DEG analyses were conducted to identify genes associated with the mRNAsi of tumor tissues and peripheral blood. First, WGCNA was used to perform the cluster analysis of genes associated with the mRNAsi of tumor tissues (Fig. 4A). The results indicated that CYBC1, DEF6, LIME1, CD3D and CD7 are among the genes associated with the mRNAsi of tumor tissues (Table SIII). The pathways found to be associated with this mRNAsi include ‘hematopoietic cell lineage’ and ‘T-cell receptor signaling pathway’ (Fig. 4B). The stemness index of the tissues was used to categorise the tumors into high and low mRNAsi groups, and the DEGs between the two groups were determined (Table SIV). The pathways enriched in the upregulated genes are ‘cell cycle’, ‘oocyte meiosis’ and ‘metabolism of xenobiotics by cytochrome P450’ (Fig. 4C). The pathways enriched with the downregulated genes are ‘cytokine-cytokine receptor interaction’, ‘TNF signaling pathway’ and ‘PI3K-Akt signaling pathway’ (Fig. 4D).

Figure 4. Genes and pathways associated with the mRNAsi of lung cancer tissues and peripheral blood. (A) WGCNA was used to analyze the genes associated with the mRNAsi in tissues and (B) enriched KEGG pathways were identified. Pathways enriched in (C) upregulated and (D) downregulated DEGs associated with the tissue mRNAsi. (E) WGCNA analysis of the genes associated with the mRNAsi in peripheral blood and (F) enriched KEGG pathways. Pathways enriched in (G) upregulated and (H) downregulated DEGs associated with the mRNAsi of peripheral blood. mRNAsi, mRNA expression-based stemness index; WGCNA, weighted gene co-expression network analysis; KEGG, Kyoto Encyclopedia of Genes and Genomes; DEGs, differentially expressed genes.

The WGCNA results revealed that AEN, BCL7A, BLNK, CCT3 and CD22 are among the genes associated with the mRNAsi of peripheral blood (Fig. 4E, Table SV). These genes are enriched in ‘hematopoietic cell lineage’, ‘B-cell receptor signaling pathway’ and ‘NF-kappa B signaling pathway’ (Fig. 4F). PTPRCAP was the only gene found to be associated with the stemness index in both tissues and peripheral blood (Fig. S1). The DEGs associated with the stemness index in the peripheral blood are presented in Table SVI. The upregulated genes are enriched in ‘PI3K-Akt signaling pathway’ (Fig. 4G), whereas the downregulated genes are enriched in ‘PPAR signaling pathway’, ‘ECM receptor interaction’ and ‘protein digestion and absorption’ (Fig. 4H).

Identification of rare fusion variants through cfRNA sequencing

Seven rare fusion variants, specifically IGSF9-SCAMP3, SEC62-ALCAM, CHD1L-ARHGAP26, EML4-ALK, ZFAND3-FKBP5, RASSF3-RHPN2 and PPOX-KHDC4, were identified by cfRNA sequencing (Table II). These results demonstrate that rare fusion variants can be identified using cfRNA sequencing.

Table II. Rare fusion variants in tissues identified by RNA sequencing.

Table II.

Rare fusion variants in tissues identified by RNA sequencing.

Fusion variant	Junction Read Count	Spanning Frag Count
IGSF9-SCAMP3	37	8
SEC62-ALCAM	27	13
CHD1L-ARHGAP26	17	11
EML4-ALK	11	13
ZFAND3-FKBP5	39	19
RASSF3-RHPN2	68	21
PPOX-KHDC4	9	17

[i] Junction Read Count indicates the overlapping reads at fusion breakpoints. Spanning Frag Count indicates that reads do not overlap the breakpoint position, but have read pairs on either side of the breakpoint (spanning the breakpoint).

Correlation between mRNAsi and immune cell infiltration

Correlations between immune cell levels and the mRNAsi in tumor tissues or peripheral blood were analyzed. The results indicate that M2 macrophages and CD4+ memory T cells are activated in the cancer tissue microenvironment (Fig. 5A). In addition, TMEM177, MPRS15, PSMB3, NOP58 and TONSL genes correlate with mRNAsi in the immune microenvironment of the cancer tissues (Fig. 5B). By contrast, the genes that correlate with cfRNA stemness in immune cells include RPS4X, CD52, BCL7A, HVCN1, AEN and SMAD3 (Fig. 5C). The mRNAsi values of dendritic cells, regulatory T cells (Tregs), and M2 macrophages in the tumor tissues are higher than those in other types of immune cells (Fig. 5D). The cfRNA stemness index values of Tregs and CD8+ T cells are higher than those of other types of immune cells (Fig. 5E). In addition, the cfRNA stemness of dendritic cells and M2 macrophages is lower than the stemness of the cancer tissue. Finally, the associations between the degree of stemness and various immune cells were analyzed. The results revealed that Tregs are closely associated with tissue and blood stemness (Figs. S2 and S3). Notably, the mRNAsi-associated immune cells and genes differ between the tumor tissues and peripheral blood.

Figure 5. Immune cells associated with the stemness of lung cancer tissue and cfRNA. (A) Analysis of the infiltration of various immune cells in tissues. (B) Correlation between the mRNAsi-related genes of tissues and immune cell infiltration. (C) Correlation between cfRNA stemness index-related genes and immune cell infiltration. Correlation between immune cells and the (D) mRNAsi of tumor tissues and (E) cfRNA stemness index. cfRNA, cell-free DNA; mRNAsi, mRNA expression-based stemness index.

Discussion

The treatment and prognosis of lung cancer vary according to the extent of the disease at diagnosis (26). Early detection and diagnosis methods, particularly non-invasive approaches, are crucial for improving the overall survival rate and prognosis of patients with lung cancer (27). Numerous studies have focused on the identification of biomarkers for the early detection of lung cancer based on cfDNA analysis (28–30). In the present study, cfRNA sequencing was used to explore the genes, pathways and immune cells associated with lung cancer cell stemness.

The mRNAsi value of metastatic lymph nodes was found to be significantly higher than that of tumor tissues. In addition, a previous study has reported that the mRNAsi value of metastatic tumors is considerably higher than that of primary tumors, particularly in prostate and pancreatic cancers (7). Furthermore, a comparison of two-year recurrence rates in the current study revealed that high tumor and blood mRNAsi values are associated with a poorer prognosis, which is consistent with the results of previous studies (31,32). Studies of cfRNAs in lung cancer have typically focused on miRNAs or a few known cancer-associated mRNAs (33). Plasma-based cfRNA transcriptome analysis has been demonstrated to be useful for the early detection, diagnosis and monitoring of NSCLC (17). In the present study, the stemness index of cfRNA was higher than that of tumor tissues. Additional studies are necessary to elucidate the relationship between cfRNAs and tissue stemness in lung cancer.

WGCNA revealed that the ‘hematopoietic cell lineage’ pathway is enriched in mRNAsi-related genes in lung cancer tissue and blood. Hematopoietic cell lineages have previously been indicated to be involved in the proliferation, differentiation and adhesion of lung cancer cells (34). The present study also revealed that PTPRCAP is the only gene associated with the stemness index in peripheral blood as well as in tumor tissues. PTPRCAP is an immunological marker of triple-negative breast cancer, the expression of which positively correlates with disease-free survival (35). Additionally, PTPRCAP is a natural killer cell-associated gene that is associated with the response of patients with chronic myeloid leukemia to tyrosine kinase inhibitor (TKI) treatment (36). However, the role of PTPRCAP in lung cancer, particularly its effect on lung cancer cell stemness, is unknown. Therefore, additional studies are necessary to investigate the mechanism of action of PTPRCAP in lung cancer progression.

Seven rare fusion variants, namely IGSF9-SCAMP3, SEC62-ALCAM, CHD1L-ARHGAP26, EML4-ALK, ZFAND3-FKBP5, RASSF3-RHPN2 and PPOX-KHDC4, were identified in the present study using cfRNA sequencing. Anaplastic lymphoma kinase (ALK) fusion has been found in 3–7% of patients with NSCLC (37). Considerable progress has been achieved in the treatment of ALK+ NSCLC using TKIs, as demonstrated by clinical trials (38,39). EML4-ALK is the most frequently observed fusion variant resulting from ALK rearrangement in NSCLC (40). For patients with EML4-ALK+ NSCLC, the National Comprehensive Cancer Network guidelines recommend crizotinib, ceritinib or alectinib as first-line therapies (41). However, to the best of our knowledge, the other fusion variants have not been reported in lung cancer and require validation.

Immune cells are a crucial part of the tumor microenvironment (TME) and play a crucial role in tumor immunotherapy (42). The present study indicated that M2 macrophages and CD4+ memory T cells are activated in the microenvironment of lung cancer tissues. Tumor-associated macrophages (TAMs) are derived from peripheral blood monocytes that infiltrate solid tumor tissues (43). TAMs acquire an M2-type macrophage phenotype in response to signals from the TME and contribute to tumor metastasis (44). In addition, the CSC-mediated accumulation of M2 macrophages is reported to contribute to the generation of an immunosuppressive, pro-tumorigenic TME (42). A high invasive abundance of CD4+ memory-activated T cells has been demonstrated to be associated with poor survival in patients with LUAD (45). Furthermore, in the present study, high mRNAsi values, indicating the upregulated expression of stemness-associated genes, were observed in the dendritic and Treg cells in tissues and in Treg and CD8+ T cells in the blood, which may be attributed to the environments in which the cells are located. CSCs are able to alter the composition and functional characteristics of tumor-specific effector T cells and promote the expansion of immunosuppressive tumor-promoting Tregs (46). In addition to regulating the immune environment, CSCs exhibit various properties that allow them to directly evade the effector mechanisms of immune cells (47). Since a high cfRNA stemness index was associated with increased Treg enrichment, an elevated cfRNA stemness index is indicative of an immunosuppressive state.

In summary, the present study reveals the unique stemness gene expression characteristics of blood cfRNAs, offering novel insights for the development of non-invasive early lung cancer detection methods. However, the study has some limitations. First, only 22 lung cancer samples were analyzed. The results of such a small sample could be biased; therefore, future studies with more samples or patients from multiple centers are necessary to support the findings. Second, as the study is based on bioinformatics analysis, it has certain shortcomings in terms of experimental validation. Therefore, well-designed basic experiments and clinical trials are critical for confirmation of the results.

Supplementary Material

Supporting Data

Supporting Data

Supporting Data

Supporting Data

Supporting Data

Supporting Data

Supporting Data

Acknowledgements

Not applicable.

Funding

This study was funded by the National Natural Science Foundation of China (grant no. 81773273) and The Special Project of Clinical Research in Health Field of Shanghai Municipal Health Commission (grant no. 202240019).

Availability of data and materials

The data generated in the present study may be found in the CNGB Sequence Archive of China National GeneBank DataBase (https://db.cngb.org/) under accession number CNP0005912.

Authors' contributions

RL, BY and YC conceived and designed the study. BY drafted the manuscript. RL and ZW reviewed and revised the manuscript. YC and XP performed the experiments. ZW, JL, RW and BL conducted the data analysis and prepared the figures. RL, BY and BL confirm the authenticity of all the raw data. All authors read and approved the final version of the manuscript.

Ethics approval and consent to participate

The study was approved by the Ethics Committee of Shanghai Chest Hospital (Shanghai, China; approval no. KS1740), and complied with the principles of the Declaration of Helsinki. All patients who participated in the study signed informed consent forms.

Patient consent for publication

Not applicable.

Competing interests

ZW, JL, and RW are employees of Berry Oncology Corporation, and some of the materials used in the study were provided by Berry Oncology Corporation. BL is an employee of Liaoning Kanghui Biotechnology Corporation, and assays and analyses in this study were provided free of charge by Liaoning Kanghui Biotechnology Corporation. The other authors declare that they have no competing interests.

References