Introduction
Lung cancer ranks first in mortality and second in morbidity among malignant tumors worldwide [
1]. Among them, non-small cell lung cancer (NSCLC) accounts for over 85% in terms of pathological types. Despite novel advancements in therapeutic strategies, the 5 year overall survival (OS) rate for NSCLC patients remains at a low level (approximately 15%) [
2]. There is an urgent need to identify novel biomarkers associated with the prognosis and therapeutic efficacy of NSCLC patients and reveal the potential mechanisms.
Microbiota have recently emerged as novel tumorigenesis regulators and biomarkers in multiple types of cancers, including lung cancer [
3]. Microbiome dysbiosis contributes to cancer susceptibility and progression through complex mechanisms, including promoting inflammatory, producing toxins, causing DNA damage, activation of cancer-related pathways, etc. [
3‐
7]. Besides, increasing evidence indicated that the enormous microbiome of host contributes to maintaining host-immune balance and shaping tumor microenvironment, thus to cancer progression and treatment response [
8,
9]. With the rise of high-throughput sequencing technology, the microbiome has shown broad prospects for cancer diagnosis and treatment. Several studies have demonstrated the globe changes of lung microbiome between healthy participants and lung cancer patients, and between patients with different prognosis [
10,
11]. However, specimens from lung, such as lung biopsy tissues and bronchoalveolar lavage fluid, are not always readily available.
After ruling out potential contaminations, researchers have identified circulating microbial DNA (cmDNA) from cancer patients’ peripheral blood [
12,
13]. Peripheral blood flows through the whole body and can carry molecules from various parts of the body, including tumor tissues. Therefore, theoretically, cmDNA could show a general view of the microbiome of host. In addition, cmDNA has great advantages for clinical application, including non-invasive operations, making it a promising biomarker for cancer diagnosis and progression [
13]. Emerging evidence has identified the characteristic composition patterns of cmDNA in various cancers, including early-onset breast cancer, prostate cancer, colorectal cancer, pancreatic cancer, and hepatocellular carcinoma [
12,
14‐
17]. In addition, Messaritakis et al. demonstrated the clinical value of cmDNA in predicting cancer progression and prognosis in CRC patients. They found that 16S rDNA, the β-galactosidase gene of
Escherichia coli, glutamine synthase gene of
Bacteroides fragilis, and 5.8S rRNA of
Candida albicans in peripheral blood were associated with the metastatic disease and shorter survival rates [
18]. However, whether circulating microbiome signatures are associated with prognosis of NSCLC and its potential mechanism on tumor immune microenvironment still remains unclear.
In this study, we obtained the blood microbiome data and the matching tumor RNA-seq data of TCGA NSCLC patients, to dissect their profiling and key roles in NSCLC. We aimed to establish a novel and comprehensive circulating microbial abundance prognostic scoring (MAPS) model to decipher the tumor immune microenvironment and to predict the efficacy of chemotherapy and immunotherapy and the prognosis of NSCLC patients. We believe that novel explorations of circulating microbiome signatures will provide a new idea of interactions between microbes, tumors, and immunity, and provide significant clues for development and intervention of NSCLC patients, which will further promote the development of precision medicine for NSCLC.
Discussion
Novel explorations of microbiome signatures provide new ideas for cancer progression. In this study, we identified 14 OS-related circulating microbes to construct a circulating MAPS prediction model and assessed its independent prognostic value in NSCLC patients. Furthermore, we deciphered its potential mechanism on tumor immune microenvironment. Finally, we proposed that circulating MAPS may be a potential biomarker of chemotherapy sensitivity in NSCLC patients. Our study provides a new idea for cancer development and immune-microbe interactions, and may contribute to personalized intervention of NSCLC patients.
Currently, cancer-related biomarkers are mainly identified from genomic or proteomic profiles. In recent years, the previously underestimated role of the microbiome in cancers has been highlighted for its promising potential in cancer development and treatment [
13]. Previous studies have confirmed the important role of lung microbiome in lung cancer progression and intervention [
10,
11]. However, given the routine inaccessibility of lung tissues, other sources of microbial dysbiosis, such as in the gut, sputum and blood, may be also associated with lung cancer and are attracting more attention [
25‐
28]. For example, Zhang et al. analyzed the gut microbiome composition of lung cancer patients and healthy volunteers, and found higher levels of
Bacteroides,
Veillonella, and
Fusobacterium in lung cancer patients [
25]. Hakozaki et al. and Qiu et al. reported that gut microbiome was associated with the response to immunotherapy and chemoradiotherapy in NSCLC patients [
27,
28]. However, few studies have explored the relevance between blood microbiome signatures and lung cancer, which is non-invasive and can reflect a general view of the microbiome of host [
13]. Only Poore et al. explored the clinical value of cmDNA in the diagnosis of lung cancer by mining TCGA data and training machine learning models [
12]. As the first study focusing on the role of circulating microbiome signatures in NSCLC prognosis and tumor immune microenvironment, we uncovered that the circulating MAPS served as an independent prognostic factor of NSCLC and preliminarily revealed the interactions between microbes, tumors and immunity. Among the 14 OS-related microbiome signatures included in MAPS, oral
Parascardovia was found to be relevant to betel nut chewing, and the latter was an independent predictor of oral premalignant lesions [
29]. However, the role of the remaining microbes in the development of cancers has not been reported yet.
Characterizing the circulating microbiome and its underlying mechanism on cancer patients is of great interest. The enormous microbiome of the human body plays a significant role in shaping the host immune characteristics [
30‐
32]. The mutual adaptation process between human microbiome and immune can affect the response of immune system to malignant cells and thus impact the effectiveness of anti-tumor immune response [
8]. Previous studies have demonstrated a link between the intratumor microbiome and immune characteristics [
9,
26]. On the one hand, the intratumor microbiome may contribute to cancer development by promoting a pro-carcinogenic inflammatory cascade, shaping an immunosuppressive tumor microenvironment or T cell dysfunction [
33‐
35]. On the other hand, the intratumor microbiome can also play an active role in anti-tumor immune response by tumor neoantigen mimicry by microbes, T and NK cell activation, and regulating other immune cells [
36‐
38]. Based on the above studies, we wonder whether the circulating microbiome is relevant to the intratumor immune characteristics, and thus to cancer progression. It’s interesting that we found the intratumor plasma B cells and the humoral immune response related pathways, such as immunoglobulin complex and immunoglobulin production pathways, were enriched in NSCLC patients with low circulating MAPS, who had a longer OS and a better prognostic outcome. Previous studies have found that B cells and plasma cells in NSCLC are associated with better outcomes [
39‐
41]. In recent years, researchers have shown that B cells and plasma cells infiltrated in tumors can play a significant role in shaping anti-tumor immune responses [
40,
42]. Plasma cells may produce large amounts of cytokines and tumor-antigen-specific antibodies, leading to antibody-dependent cellular cytotoxicity (ADCC) and complement activation [
42,
43]. Besides, B cells can act as antigen-presenting cells (APCs) to present specific antigens to CD4 and CD8 T cells, and thereby maintain and expand long-term antitumor immunity within the tumor microenvironment [
44,
45]. It is important to note that the above effective antitumor immune response is premised on the recognition of tumor antigens. The tumor neoantigen mimicry by microbes reported previously may explain one of the mechanisms by which the circulating microbiome participate in the antitumor immune response described above [
36,
46]. In addition, it’s well known that Th2 cells contribute to cancer progression and metastasis by secreting IL-4, IL-5, IL-10 and IL-13 [
47,
48]. Consistent with this, our study revealed an increased CD4
+ Th2 cells infiltration in NSCLC tumors with high circulating MAPS. Besides, we found that
DDK1 was associated with poor prognosis of NSCLC patients and positively correlated to circulating MAPS in this study, which have been reported to promote tumor progression and negatively regulate antitumor immunity [
49]. In addition, it’s interesting that linoleic acid metabolism and steroid hormone biosynthesis pathways were enriched in circulating MAPS-low group, which may help alleviate the pathological features and improve the anti-tumor ability of patients [
50‐
52]. What’s more, we observed that patients with high circulating MAPS exhibited significant proliferation characteristics. And it may also suggest one of the mechanisms by which circulating microbes impact tumor progression.
Currently, conventional antitumor treatments include chemotherapy, radiotherapy, immunotherapy and et al. However, not all patients benefit from the above therapies, so it’s urgent to explore new prognostic biomarkers for individualized treatment of NSCLC patients. In recent years, studies have reported that microbes play an important role in regulating response to antitumor treatments [
22,
53‐
56]. For example, Fusobacterium nucleatum were found to promote chemotherapy resistance by modulating autophagy [
57]. Recently, the gut microbiome has been reported as promising predictive biomarkers of chemotherapy and immunotherapy sensitivity in cancer patients [
58‐
60]. We wondered if the circulating MAPS prognostic score identified in this study could be useful in predicting chemotherapy or immunotherapy sensitivity. Our results suggested that the circulating MAPS may contribute to chemotherapy sensitivity prediction, which may help develop personalized cancer treatment strategies. However, although the expression levels of PD-L1 differed between circulating MAPS-high and MAPS-low groups, the TIDE (Tumor Immune Dysfunction and Exclusion) scores showed no difference between these two groups. The predictive value of circulating microbiome signatures on immunotherapy remains to be further explored.
Although the circulating MAPS model performed well in prognosis prediction of NSCLC patients, there are some limitations deserving further discussion. A major limitation was that there was no additional publicly available dataset to validate our findings. In addition, as a secondary analysis of retrospective data, it’s hard to control the various potential confounding factors among participants. Therefore, it is necessary to design comprehensive prospective multicenter studies to verify the prognostic performance of circulating microbiome signatures. Furthermore, our study preliminarily deciphered the relationship between circulating microbiome signatures and tumor microenvironment and prognosis in NSCLC, while the causal relationship and underlying mechanisms need to be further explored.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.