Sage Journals: Discover world-class research

Abstract

Tumor immune microenvironment (TIME) and its indications for lung cancer patient prognosis and therapeutic response have become new hotspots in cancer research in recent years. Tumor cells, immune cells, various regulatory factors, and their interactions in the TIME have been suggested to commonly influence lung cancer development and therapeutic outcome. The heterogeneity of TIME is composed of dynamic immune-related components, including various cancer cells, immune cells, cytokine/chemokine environments, cytotoxic activity, or immunosuppressive factors. The specific composition of cell subtypes may facilitate or hamper the response to immunotherapy and influence patient prognosis. Various markers have been found to stratify the patient prognosis or predict the therapeutic outcome. In this article, we systematically reviewed the recent advancement of TIME studies in lung adenocarcinoma (LUAD) using single-cell RNA sequencing (scRNA-seq) techniques, with specific focuses on the roles of TIME in LUAD development, TIME heterogeneity, indications of TIME in patient prognosis and therapeutic response during immunotherapy and drug resistance. The main findings in TIME heterogeneity and relevant markers or models for prognosis stratification and response prediction have been summarized. We hope that this review provides an overview of TIME status in LUAD and an inspiration for future development of strategies and biomarkers in LUAD treatment.

Keywords

heterogeneity LUAD lung adenocarcinoma mutation NSCLC prognosis RNA sequencing single-cell transcription tumor immune microenvironment

Introduction

Lung cancer is the most common type of malignancy around the world in terms of morbidity and mortality, mainly including non-small cell lung cancer (NSCLC) and small cell lung cancer. Lung adenocarcinoma (LUAD) is the most common type of NSCLC in histology.¹ Current precision medicine focuses on the targeted therapy and immunotherapy of LUAD based on a series of biomarkers from gene alterations to transcriptional or protein level changes.² The development of stratification biomarkers greatly facilitated the application of various personalized therapeutic strategies in clinical practice.² Many patients benefit from the new strategies, and their lives are prolonged with better quality. However, there are still many unsolved issues in targeted therapy and immunotherapy. One of the most prominent obstacles in therapy is cancer heterogeneity, leading to unpredictable outcomes in some patients and progression in others.³ Since genetic alterations are cursory stratification of patients, more precise and accurate biomarkers are needed for patient stratification, especially in immunotherapy. Tumor immune microenvironment (TIME) represents a type of emerging biomarker and has been extensively investigated in the last a few years.

TIME is a complex ecosystem composed of cancer cells, various types of immune cells, extracellular matrix, and biomolecules.^4,5 During LUAD development, cancer cells produce a series of biological effects, including inhibition of immune cell function, induction of immune escape, and regulation of immune response, so that they can evade immune surveillance and attack. Meanwhile, immune cells will also be affected by the cancer microenvironment, leading to immune cell inhibition, inactivation, exhaustion, and even necrosis, so that they cannot effectively clear tumor cells.^6,7 Therefore, an in-depth understanding of the characteristics and regulatory mechanisms of the LUAD immune microenvironment is of great significance for the development of novel immunotherapy strategies.⁴ Recent studies have shown that the immune microenvironment plays a crucial role in tumor progression and can be a predictive indicator of cancer prognosis and response to therapy.^6,8–10

Single-cell sequencing is a new technique that allows researchers to analyze the genetic and molecular features of individual cells. Technically, high-throughput cell capture techniques, such as microfluidic, oil droplet wrapping, and barcode, have been used to isolate and label 500–10,000 single cells at a time to obtain the 3′end transcriptome information of each cell. It has the advantages of high cell flux, low cost, and short capture cycle.⁹ This technology is mainly used for the identification of cell subtypes and markers and to realize the division of cell populations and the detection of gene expression differences between cell populations. In addition, this technology can also predict cell differentiation and disease development trajectory, which is playing an increasingly important role in the current fields of studies in disease mechanism, immunity, cancer therapeutic strategy, and tissue development.¹⁰ The single-cell RNA sequencing (scRNA-seq) has emerged as a powerful tool to characterize the heterogeneity of immune cells and tumor cells within LUAD. The technique allows for the analysis of gene expression profiles of individual cells, enabling researchers to identify rare cell populations and understand intercellular communication within the TIME, which may drive the development of new therapeutic strategies.¹¹

In cancer treatment, chemotherapy, targeted therapy, and immunotherapy are important strategies. However, it is still inevitable to develop drug resistance or poor therapeutic response during treatment. Research suggests that tumor heterogeneity is an important reason for drug resistance. Single-cell sequencing technology can be used to reveal heterogeneity, analyze drug resistance mechanisms, screen potential beneficial populations, explore new therapeutic targets, and establish effective combined therapeutic plans.¹²

The scRNA sequencing techniques mainly involve the following advantages. First, it can precisely analyze the cell heterogeneity and provide the gene expression profile of a single cell, allowing researchers to analyze the gene expression characteristics of each single cell and reveal the inter-cell heterogeneity, and to better understand the function and development process of cells. Second, it can help to discover new cell types and subpopulations with small differences, especially in complex tissue samples. Third, it has broad application areas, including cancer research, embryonic development research, immunology, and neuroscience, etc. Fourth, many analytical methods and tools have been developed, which can better reveal the hidden information from data. However, there are also some technical deficiencies in scRNA sequencing. First, current data have a high noise level and high technical variability, which may introduce false positive or false negative results. In addition, technical differences and batch effects can also influence the interpretation and comparison of data. Second, compared to conventional high-throughput sequencing, the cost of single-cell sequencing is higher, mainly due to the complexity of processing individual cells and the high sequencing coverage depth required. Third, the experimental operation is complicated. It requires special experimental operation steps, such as cell capture, cleavage and extraction, etc., which has high technical requirements for experimental personnel. Fourth, the analysis of data is complex, requiring the use of special analysis tools and algorithms, such as alignment, de-batch effect, de-noise, cell clustering, and trajectory inference, etc., which requires high professional knowledge of data analysis.¹²

Many studies have used scRNA-seq to investigate the TIME of LUAD.^12,13 These studies have identified various immune cell types, including T cells, B cells, natural killer cells, and myeloid-derived suppressor cells (MDSCs), and some unique subsets of immune cells that are specifically associated with LUAD.^14–16 These studies provided valuable insights into the complex interactions between the immune system and LUAD and could potentially lead to the development of novel prognostic and therapeutic markers and strategies for the immunotherapy of LUAD.^14–16 This review will focus on the components and dynamic status, especially the heterogeneity of LUAD TIME, and its indications for patient prognosis, therapeutic outcomes and resistance in immunotherapy.

The components of TIME and their roles in LUAD development

LUAD is a prevalent type of NSCLC that arises from glandular cells in the lungs. Despite recent advances in cancer treatment, LUAD remains challenging to manage, primarily due to its complex immune microenvironment. TIME is a very complex and dynamic ecosystem, which is the ‘soil’ for tumor survival. The cell components mainly include tumor cells, immune cells, and support cells. Under the influence of chemokines from tumor cells, fibroblasts, or inflammatory cells, immune cells in the blood circulation migrate to the tumor site through a transendothelial process.^12,17 Within the tumor tissue, immune cells locally proliferate, differentiate, function, die, and partially migrate back to the circulatory system. In these cells, cells associated with acute inflammation (including neutrophils, basophils, and eosinophils), cells associated with innate immunity [including macrophages, natural killer (NK) cells, and dendritic cells (DCs)], and cells derived from adaptive immune responses [including cluster-of-differentiation (CD)8+T cells, Th1-/Th2 cells, and B cells] can be found^12,17 (Figure 1).

Figure 1.

Illustration of cellular components and their roles in TIME. Cells with potential anti-tumor activity or pro-inflammatory activity (active TIME) include but are not limited to CD8+ T cells, dendritic cells, M1-like macrophage, helper 1 T cell (Th1), active T cells, natural killer T cells, and natural killer cells. Cells with potential immune suppressive or pro-tumor activity (suppressive TIME) include but are not limited to M2-like macrophage, MDSC, inactive T cells, regulatory T cell (Treg), and helper 17 T cell (Th17).

The main cellular components and their roles in TIME are illustrated in Figure 1. Tumor-associated macrophages (TAM) are a relatively abundant subset of cells that outnumber other types of immune cells in many tumors. TAM has a highly plastic phenotype and function, with two major subtypes identified, including M1 TAM, which is induced by toll-like receptor ligands such as lipopolysaccharide and interferon gamma (IFN-γ), and preferentially expresses pro-inflammatory cytokines and inducible nitric oxide synthase, and M2-type TAM, which is induced by interleukin (IL)−4 or IL-13, and expresses arginase 1, CD206, CD163, IL-4R, transforming growth factor (TGF)−β1, and platelet-derived growth factor.^18,19 Some studies have shown that M1 TAM can enhance anti-tumor Th1 response and antagonize the inhibitory activity of regulatory immune cells, whereas M2-type TAM mainly plays a role in promoting angiogenesis, tumor growth and metastasis.^18,19 NK cells are cytotoxic effector lymphocytes in the innate immune system and their primary function is to help control infections and tumors. The two main mechanisms by which NK cells recognize tumor cells are that they can recognize cells with downregulated expression of major histocompatibility complex class I (MHC-I) molecules, which is a phenomenon of immune tolerance shown in multiple cancer types, or bind to stress-inducing ligands expressed on tumor cells, such as major histocompatibility complex class I polypeptide-related sequence A (MICA) or MICB.^20,21 The primary function of dendritic cells (DCs) is to act as a bridge between innate and adaptive immune responses. Physiologically, DCs engulf and process non-autoantigens. When they receive activation signals, they move to secondary lymphoid structures in the lymph nodes, where they activate primary B or T cells. DC phenotypes are quite plastic; in that, they can produce a variety of pro-inflammatory or immunosuppressive cytokines and express a range of activating or inhibiting receptors, depending on their environment.^22,23 Tertiary lymphatic structure (TLS) is a highly organized lymphoid aggregate that develops in an inflammatory pathological state. In cancer, TLS usually occurs in the infiltrating margins and/or interstitium of the tumor, similar to other chronic infectious or autoimmune diseases.^24,25

CD4+ T helper cells can be divided into several subtypes, including Th1, Th2, Th17, follicular helper T cell (Tfh), and regulatory T cell (Treg). Each subgroup plays a specific role in the anti-tumor immune response. Th1-directed responses inhibit tumor growth and are generally associated with favorable clinical outcomes. In fact, Th1 cells enhance the anti-tumor function of cytotoxic T cells in situ by producing several cytokines, including IL-2 and IFN-γ. Tfh cells interact with B cells in TLS to help produce antibodies.^26,27 The role of other tumor-infiltrating CD4+ T-cell subsets (Th2, Th17, and Treg) is not well-understood, but is often associated with poor prognosis in different tumors.^28,29 Many studies have shown that Tregs in tumors inhibit anti-tumor immune responses through two main mechanisms, including the production of inhibitory cytokines (such as IL-10, TGF-β, IL-35) and the inhibition of the development and maturity of DCs.^30,31 In contrast, CD8+ T cells play very important roles in anti-tumor immune responses because they are responsible for recognizing and eliminating tumor cells. Due to genomic instability, tumor cells often express mutant proteins on their surface. Many of these are neoantigens that induce tumor-specific immune responses.^32–34 Activated CD8+ T cells are responsible for tumor cell recognition and lysis, the mechanisms of which have been described in detail, involving the release of cytotoxic particles. Interestingly, in most tumors, invasive cytotoxic T cells express inhibitory receptors (e.g. programed death-1 (PD-1), Tim-3, and Lag-3) whose function in physiological situations is to suppress the immune response when bound to their ligands.^32–34 In fact, many tumor cells can take advantage of this inhibitory mechanism and express multiple ligands (e.g. PD-L1, PD-L2) to help them escape T-cell attacks.^32–34 In tumor and other inflammatory conditions, B cells enhance T-cell response by producing antibodies, stimulant cytokines, and chemokines, acting as local antigen-presenting cells, and organizing the formation of TLS that sustain immune response, which makes B cells generally anti-tumor.^35,36

The above cells play important roles in the development of LUAD. The cells that promote LUAD mainly include B cells, Treg cells, and TAM (M2 type) cells. The B cells that produce IL-10 have tumor-promoting activity and immunosuppressive characteristics.^19,20 Treg cells promote tumor growth by inhibiting anti-tumor responses.^13,14 TAM (M2 type) contributes to tumor growth, immune suppression, and cancer cell invasion.^4,5 The cells that inhibit LUAD mainly include CD8+ T cells, DC cells, TAM (M1), and NK cells. CD8+T cells exert direct cytotoxic effects on cancer cells by secreting granzymes and perforin.^10–12 DC cells capture tumor antigens, and type I DC initiates CD8+ T cell response, whereas type II DC is responsible for initiating CD4+ T-cell response.^6,7 TAM (M1 type) promotes inflammation and secretes tumor necrosis factor (TNF)-α and nitric oxide to kill tumor cells and further activates T-cell-mediated immune responses.^4,5 NK cells exhibit cytotoxic activity against infected and mutated cells, secreting cytokines and chemokines such as tumor necrosis factor α, IFN- γ, and interact with other immune cells.^8,9

The heterogeneity of TIME in LUAD

Heterogeneity is one of the main characteristics of cancer. During the development of cancer, multiple patterns of evolution exist in cell division and proliferation, and the daughter cells show various alterations in epigenetics, genetics, transcriptomics, and proteomics, resulting in distinct characteristics in tumor growth rate, invasion ability, sensitivity to drugs, and prognosis.³⁷ The heterogeneity of cancer has been widely investigated in terms of genetic alterations. In LUAD, diverse intratumor heterogeneity and intertumor heterogeneity have been revealed, resulting in distinct outcomes in surgical, targeted, and immunotherapy.³⁸ In the era of immunotherapy, stratification of patients by genetic alterations is not sufficient to instruct the establishment of therapeutic strategies, as the huge discrepancy in prognosis and therapeutic efficacy cannot be explained by genetic alterations. TIME has attracted attention from both researchers and clinicians as it has provided a promising tool for effective stratification besides genetic alterations. The scRNA-seq facilitates the study of TIME heterogeneity.³⁹

During tumorigenesis and subsequent metastasis, malignant cells become gradually diversified, and tumors may be infiltrated by a variety of immune-related components, including various immune cells, cytokine/chemokine environments, cytotoxic activity, or immunosuppressive factors.³⁹ The TIME is composed of all these components and is common in almost all solid tumors. It varies spatially or temporally, and its dynamic changes can be observed with tumor development, metastasis, and therapeutic interventions. The heterogeneity of antitumor immunity is closely related to disease progression and therapeutic responsiveness, especially in the field of immunotherapy.³⁹

Diverse tumor and immune cell types and corresponding markers have been revealed in the study of TIME in LUAD. The main findings from studies on TIME heterogeneity are summarized in Table 1.^40–50 Early reports focused on the heterogeneity of key pathways or immune-related molecules. For example, genes of the γ-interferon (IFN-γ) signaling pathway, especially major histocompatibility complex (MHC) class II molecules, such as human leukocyte antigen (HLA), were found to be heterogeneously expressed and coregulated with other genes in single cancer cells across different LUAD nodules.⁴⁰ Most later studies focused more on the heterogeneity of diverse cell types and markers. The heterogeneity and evolution of LUAD were defined by comprehensive profiling of cell types within the tumor in one representative study.⁴¹ Many cells in the LUAD tissue, including epithelial cells, endothelial cells, cancer cells, and club cells have been revealed, and many immune cells, including macrophages, mast cells, T cells, B cells, NK cells, plasma cells, and dendritic cells have also been found. Different nodules exhibited a distinct ratio of cell components and expression profiles, representing the alterations of immunity and pathways. The trajectory analysis of multiple nodules helped to determine the stage and progression of the tumor, exhibiting characteristic trajectory of development across different cancer nodules.⁴¹ Another study further investigated intratumor heterogeneity and found that tumor cells within early-stage LUAD had heterogeneous gene expression signatures by looking at the differential expression profile from functional enrichment analysis.⁴² In addition to tumor cells, immune cells within tumors also exhibited substantial heterogeneity. It was reported that 27 cell subtypes of T cells, B cells, fibroblasts, and myeloid cells were present in LUAD, and different CD4+ and CD8+ T-cell clusters exhibited differences in pathway activities.⁴³ Similarly, 15 main types of cells and 57 cell subgroups were identified from the multiple LUAD scRNA-seq datasets in LUAD, and a series of potential biomarkers were revealed in M2b, exhausted CD8+ T, endothelial cells, fibroblast in TIME.⁴⁸

Table 1.

Summary of the main findings and the significance of studies on heterogeneity in TIME of LUAD.

Ref No.	Publication Year	Total number of cells analyzed	Total number of patients involved	Protocol used for scRNA-seq library preparation	Pipeline used to analyze the data	Main findings*	Significance of the study
Ma et al.⁴⁰	2019	77 (LC-PT-45), 49(LC-MBT-15)	2 patient-derived xenografts (PDXs) samples and 2 LUAD cell lines	SMART-seq	Raw reads were mapped to GRCh37 with RSEM and normalized using SCnorm.	IFN-γ signaling corresponds to acquired resistance	Development of prognostic metrics based on heterogeneity
He et al.⁴¹	2021	55713 cells	5 multiple nodules LUAD patients	10X genomics	Reads were mapped to GRCh38. Cell number was determined by ‘knee’ method. Seurat was used to perform analysis.	Heterogeneity and evolution and changes in TIME in LUAD early development.	TIME mechanism of early tumor development
He et al.⁴²	2021	125,674 cells	7 stage-I/II LUAD samples	10X genomics	NA	The LUAD cells can be categorized into subtypes by distinct gene expression signatures	Substantial heterogeneity within early-stage LUAD harboring EGFR mutations
Wang et al.⁴³	2021	9993 single cells	3 patients, 12 samples	10X genomics	Reads were mapped to GRCh38 using CellRanger (3.0.1) pipeline.	8 main cell types and 27 cell subtypes with distinct signaling and expression profile are identified.	Immunological heterogeneity in the TME of LUAD patients
Wu et al.⁴⁴	2021	40,362 cells for GSE127465, and 208,506 cells for GSE131907	5 LUAD patients	inDrop	The Seurat package (version 3.0) was used to perform scRNA-seq analysis	Twenty myeloid cell types were detected, 13 of which had distinct enrichment patterns	TIME infiltration landscape may facilitate the development of immune therapy.
Zheng et al.⁴⁵	2022	3583 cells from the tumor cores of three human LUAD samples	1 for GSM3304007, 1 for GSM3304011 and 1 for GSM3304013,	10X genomics	Run CellRanger aggr to aggregate multiple libraries from a single experiment. Genome_build: GRCh38	A risk scoring model involving CCL20 mutation status was established based on eight independent prognostic genes	A prognostic model that could be used to guide clinical practice
Wang et al.⁴⁶	2022	72,475 immune cells	19 pathologically diagnosed NSCLC patients (10 LUAD and 9 LUSC, 40 samples)	self-developed method	Raw gene expression matrices were generated using CellRanger (version 3.0.1).	Revealed the roles of several immune cells between LUAD and LUSC.	The comprehensive depiction of the immune heterogeneity of immune cells
Song et al.⁴⁷	2022	208,506 cells.	58 specimens from 44 LUAD patients	10X genomics	The single-cell data were filtered and dimensionally reduced using the R ‘Seurat’ and the ‘dplyr’ packages.	The presence of abundant transition-state CD8+ T cells during tumor differentiation	Heterogeneity and functional exhaustion of infiltrating CD8+ T cells in LUAD
Fan et al.⁴⁸	2022	24,550 single cells	19 normal and 53 LUAD samples	microwell-seq, 10X genomics and Smart-seq2	The analysis and quality control (QC) process used the Seurat R package	Identified 15 main types of cells and 57 cell subgroups, and revealed a series of potential biomarkers in TME.	This study provided insights into the heterogeneity of LUAD
Yao et al.⁴⁹	2022	13,989 cells	6 patients	10X genomics	We finally retrieved the singlet cells and merged them again into the working Seurat object	NecroLRS was positively correlated with neutrophil enrichment, inflammatory immune response.	Emerging mechanisms of necroptosis-induced TIME alteration
Sun et al.⁵⁰	2022	61,867 individual cells	4 glioma samples, 10 LUAD samples	BD Rhapsody, CytoSeq	Clean data were mapped to the human genome (Ensembl version 91) utilizing STAR. Seurat package was used for cell normalization and regression.	Lung-to-brain metastases reprogramed cells into immune suppressed state	A comprehensive understanding of the TIME heterogenicity in brain metastasis.

IFN-γ, interferon gamma; LUAD, lung adenocarcinoma; NSCLC, non-small cell lung cancer; RSEM, RNA-Seq by Expectation Maximization; STAR, Spliced Transcripts Alignment to a Reference; TIME, Tumor immune microenvironment.

The scRNA techniques helped to investigate the mechanism of LUAD development in terms of heterogeneity. One study found that the transcriptome features of cancer samples were significantly different from matched para-cancer samples. Cells from adjacent normal samples clustered more closely with those of the LUADs than with more distant normal tissues. In cancer samples, the proportion of B cells increased and the abundance levels of NK cells decreased. These findings reveal spatial heterogeneity in the transcriptomic characteristics of the TIME in early LUADs.⁵¹ The study also found spatial heterogeneity (only mutations in tumor samples) and intratumoral heterogeneity in KRAS mutations and cell lineage features. Similarly, spatial heterogeneity of lymph cells, including subtypes of CD4+ and CD8+ T cells, was also revealed. This was also supported by the findings that the closer to the tumor area, the less proportion were found for the M2-like macrophages C5, monocytes (classical and non-classical) and mast cells, which were depleted in the tumor area, while M2-like macrophages C1, proliferating myeloid cells and cDC2 cells were enriched in the tumor.⁵¹

In another study, the authors delineated the dynamic evolution from preneoplasia to invasive LUAD by integrating scRNA sequencing and spatial transcriptomics.⁵² It was found that the UBE2C+ cancer cell subpopulation increased during LUAD invasion and was significantly elevated in invasive LUAD, where it was spatially distributed in the peripheral cancer area of invasive LUAD and represented a more malignant phenotype. In addition, analysis of the TME cell subpopulation revealed a sustained decrease in mast cells, monocytes, and lymph endothelial cells involved in the entire course of LUAD invasion, accompanied by a significant increase in NK cells and MALT B cells during the early stages of invasive LUAD, and the increase of Tregs and secretory B cells in late invasive LUAD.⁵²

The molecular and cellular reprograming in metastatic LUAD was also investigated by scRNA sequencing.⁵³ The majority of T cells and myeloid cells were found in the primary lesion. Compared with normal tissue, T cells and B cells were enriched in early lesions, whereas NK cells and myeloid cells were reduced in late lesions. Compared with normal lymph nodes, there were more myeloid cells in metastatic lymph nodes, indicating that the process of metastasis was accompanied by more myeloid cell infiltration. The brain metastatic samples contained immune cells (T cells, B cells, NK cells), oligodendrocytes, and myeloid cells (microglia), of which oligodendrocytes were only present in brain metastasis. Furthermore, many angiogenic genes were upregulated, whereas many genes related to immune activity were downregulated in tumor tissues and brain metastases. Myofibroblasts originated only from tumor tissues (including primary and metastatic sites), which can promote tissue remodeling, angiogenesis, and tumor progression.⁵³

Temporal and spatial heterogeneity has been widely reported in studies of LUAD TIME. The heterogeneity is composed of immune cells, tumor cells, and molecular alterations and is characterized by distinct components at various stages of tumor development. Heterogeneity may influence therapeutic response and prognosis and could be a potential target of therapy that is worth future investigation.

The prognostic TIME markers for LUAD

The prognostic factors of LUAD in immunotherapy have been widely studied at multiple levels. Gender, Eastern Cooperative Oncology Group Performance Status (ECOG-PS), body mass index, metastatic status, combined drug use, blood neutrophil-to-lymphocyte ratio, blood lactate dehydrogenase level, blood C-reactive protein(CRP) level, PD-L1 expression level, vascular endothelial growth factor(VEGF) status, tumor mutational burden (TMB), the status of driver mutations and large fragment alterations, and circulating tumor DNA (ctDNA) status have all been suggested as prognostic factors for the immunotherapy of LUAD.⁵⁴

The prognosis of LUAD in immunotherapy is also the focus of many studies investigating TIME. Markers related to TIME have been comprehensively studied, and several types of markers have been discovered, including differentially expressed genes (DEGs) or specific cell markers, risk score from the combination of multiple factors and cell property (cell components or cell subtype ratio) related markers. The prognostic markers and the indications for prognosis in these studies are summarized in Table 2.^{36,44,45,47–49,55–89} DEGs and their combined prognostic model were the most common types of prognostic markers in these studies. For example, one study identified nine cancer-specific DEGs (CBFA2T3, CR2, SEL1L3, TM6SF1, TSPAN32, ITGA6, MAPK11, RASA3, and TLR6) and established a prognostic risk model in combination with clinical factors.⁶² The risk model was validated in tumor and normal tissues by RNA-sequencing and scRNA-seq and revealed that high-risk patients were associated with poor prognosis, including advanced stages and low survival rates.⁶² Another similar study identified eight independent prognostic genes. CCL20 mutational status was also found to affect the prognosis and differentiation of LUAD and led to a poor histologic grade of tumor cells. A combined risk score involving the eight independent prognostic genes, clinicopathological information, and CCL20 mutation status was established as a nomogram with good predictive performance and high accuracy.⁴⁵

Table 2.

Summary of prognostic markers and the indications for LUAD prognosis in studies on TIME.

Ref No.	Publication year	Prognostic markers	Indications for prognosis	Subject profile	Therapies involved
Song et al.³⁶	2022	Nine-gene signature	Prognostic prediction	Both early and advanced stage LUAD	ICIs (drugs not specified)
Wu et al.⁴⁴	2021	MSC1 and MSC2 (MSC subtype)	Overall survival	Both early and advanced stage LUAD	ICIs (drugs not specified)
Zheng et al.⁴⁵	2022	A risk scoring model (8 genes);CCL20	risk score model; poor prognosis (CCL20 mutations)	both early and advanced stage LUAD	ICIs (drugs not specified)
Song et al.⁴⁷	2022	Exhausted CD8+ T lymphocyte (ETL) subset; four hub genes	ETL subset and four hub genes are prognostic	Both early and advanced stage LUAD	Surgery or ICI (PD-1 blocking)
Fan et al.⁴⁸	2022	M0 macrophage and T cell activation	Correlated to a better prognosis	Both early and advanced stage LUAD	ICIs (drugs not specified)
Yao et al.⁴⁹	2022	Necroptosis-related LncRNA Risk Scoring (NecroLRS)	NecroLRS is prognostic	Both early and advanced stage LUAD	Surgery or ICI (drugs not specified)
Mansuet-Lupo et al.⁵⁵	2016	Intratumoral CD8+ T-cell and mDC densities	Independent markers of overall survival	Resectable LUAD	Surgery
Pocha et al.⁵⁶	2020	TIL numbers within tumor islands; surfactant metabolism-related genes	Prognostic in LUAD with brain metastases	LUAD with brain metastasis	Chemotherapy, WBRT, surgery
Wu et al.⁵⁷	2020	A risk score model	A robust and independent prognostic biomarker	Both early and advanced stage LUAD	Anti-PD-L1 ICIs (drugs not specified)
Li et al.⁵⁸	2020	An immunoscore model	Prognostic immunoscore in LUAD was established	Both early and advanced stage LUAD	Surgery and ICIs (drugs not specified)
Zeng et al.⁵⁹	2020	15 co-expressed stemness related genes	Potential biomarkers for prognostic indicators	Both early and advanced stage LUAD	Surgery, adjuvant therapy, ICIs (drugs not specified)
Zhao et al.⁶⁰	2021	A risk score model (8 genes)	Contributed to patient prognosis stratification	Both early and advanced stage LUAD	Surgery, TKI therapy, ICIs (drugs not specified)
Wang et al.⁶¹	2021	TTN mutations	TTN-mutant is prognostic	late-stage LUAD	ICIs (drugs not specified)
Zhong et al. ⁶²	2021	Nine hub genes	prognostic risk model	Early-stage LUAD	Surgery
Bischoff et al.⁶³	2021	CP²E and N³MC microenvironment	Prognostically unfavorable (CP²E);a favorable prognosis (N³MC)	Early-stage LUAD	Surgery
Gong et al.⁶⁴	2021	A ceRNA network of SNHG6-hsa-miR-30e-5p-CYSLTR1	associated with the prognosis	Both early and advanced stage LUAD	anti-PD-1 ICIs (drugs not specified)
Zhou et al.⁶⁵	2021	CENPA, CENPH, CENPM, CENPN, and CENPU	A group of potential prognostic markers	Early-stage LUAD	Surgery, radiotherapy, and chemotherapy
Zhou et al.⁶⁶	2021	A risk signature (6 genes)	Patients with a higher risk score had shorter PFS	Advanced lung cancer (T1–4, N1–2, M0)	Surgery, chemotherapy, targeted therapy
Ly et al.⁶⁷	2021	A CD4+ LAIR2+ Treg gene signature	LAIR2 expression was adversely prognostic	early-stage LUAD (I-II)	Surgery, chemotherapy, targeted therapy
Wang et al.⁶⁸	2022	CXCL7, XCL17, B cell infiltration	CXCL7, XCL17 expression and B cell infiltration are prognositc	both early and advanced stage LUAD	Surgery, chemotherapy, targeted therapy, ICIs (drugs not specified)
Shinohara et al.⁶⁹	2022	T-score, I-score and S-score; PI3K pathway alteration	T, I, S scores and PI3K pathway alterations are prognostic	Resectable LUAD	Surgery, chemotherapy, targeted therapy, anti-PD-1 ICIs (drugs not specified)
Noyes et al.⁷⁰	2022	clonal shrinkage of tumor-infiltrating CD8 +T cells	Reduced survival	Transgenic mouse model	No therapy applied
Guan et al.⁷¹	2022	Molecular subtype; ten gene risk model	Longest OS (C3);strong prediction power in prognosis (risk model)	Early-stage LUAD	Surgery, chemotherapy
Zhu et al.⁷²	2022	RTN1 expression	Better prognosis	Both early and advanced stage LUAD	Surgery, chemotherapy, targeted therapy, ICIs (drugs not specified)
Jin et al. ⁷³	2022	A 5-gene risk model	An excellent prognostic performance	Both early and advanced stage LUAD	Surgery, chemotherapy, targeted therapy, ICIs (drugs not specified)
Luo et al.⁷⁴	2022	Nine differentiation-related genes	Significant survival-predicting power	Early-stage LUAD	Surgery, chemotherapy, targeted therapy
Liu et al.⁷⁵	2022	Gln metabolism-based model	Played a significant role in predicting prognosis in lung cancer	Both early and advanced stage LUAD	nivolumab or pembrolizumab
Wang et al.⁷⁶	2022	ACAP1 expression	ACAP1 was predictive of prognosis	Both early and advanced stage LUAD	ICIs (drugs not specified)
Zhong et al.⁷⁷	2022	High expression of XBP1	Good prognosis	Both early and advanced stage LUAD	Surgery, chemotherapy, radiotherapy, ICIs (drugs not specified)
Liu et al.⁷⁸	2022	SLC7A5 high expression	Indicates poor prognosis and was an independent prognostic factor	Both early and advanced stage LUAD	Chemotherapy, radiotherapy, ipilimumab
Hao et al.⁷⁹	2022	MYC signaling status	Significant differences in prognosis	Early-stage LUAD	Surgery, chemotherapy
Shi et al.⁸⁰	2022	Two clusters (C1 and C2); Four hub chromatin regulators	Two clusters (C1/C2) and hub CRs can predict prognosis and outcome	Both early and advanced stage LUAD	ICIs (drugs not specified)
Li et al.⁸¹	2022	The four-gene prognostic signature	The four-gene signature could be used for predicting outcomes	Both early and advanced stage LUAD	ICIs (drugs not specified)
Ma et al.⁸²	2022	61 cuproptosis-related lncRNAs prognostic lncRNAs	Exhibited a satisfactory performance predicting LUAD patients’ survival	Both early and advanced stage LUAD	Chemotherapy and ICIs (drugs not specified)
Li et al.⁸³	2022	Risk score (a 17-gene signature)	Prognostic risk prediction based on the overall survival time of LUAD patients	Both early and advanced stage LUAD	Surgery, adjuvant therapy, targeted therapy
Van Hiep et al.⁸⁴	2022	HB-EGF expression	High HB-EGF expression was significantly correlated with poor overall survival	Both early and advanced stage LUAD	Surgery, adjuvant therapy, ICIs (drugs not specified)
Zhang et al.⁸⁵	2022	High proportion of tip-like endothelial cells	Poor prognosis in multiple cancer types	Both early and advanced stage LUAD	Surgery, chemotherapy, targeted therapy, ICIs (drugs not specified)
Zhang et al.⁸⁶	2023	A risk model consisting of nine gene signatures based on T-cell marker genes	The model can predict the survival and treatment outcomes	Early-stage LUAD	Surgery, adjuvant therapy, ICIs (drugs not specified)
Zhang et al.⁸⁷	2023	High cancer-specific fibroblasts (CSF) proportion	Associated with poor prognosis	Early-stage LUAD	Surgery, adjuvant therapy, ICIs (drugs not specified)
Liu et al.⁸⁸	2023	12 genes; m7G cluster-B; NUDT4 and WDR4	Predict survival and outcomes	Both early and advanced stage LUAD	CTLA-4 and PD-1 ICIs (drugs not specified)
Zhang et al.⁸⁹	2023	A risk model consisting of 11 genes	High-risk group had a worse prognosis	Both early and advanced stage LUAD	PD-1/PD-L1 or CTLA4 ICIs (drugs not specified), or both

CRs, chromatin regulators; ICIs, immune checkpoint inhibitors; LUAD, lung adenocarcinoma; OS, overall survival; TIME, Tumor immune microenvironment.

Specific cell markers were also suggested to be prognostic in TIME. One study suggested that tumor-associated regulatory T cell expression of immunoglobulin-like receptor 2 (LAIR2) was prognostic in LUAD. The study found that LAIR2 was preferentially produced by activated CD4+ T cells and enhanced in vitro tumor invasion into collagen. A CD4+ LAIR2+ Treg gene signature was prognostically significant in the The Cancer Genome Atlas (TCGA) dataset.⁶⁷ Another study reported that patients with low gene transcription of CXC ligand (CXCL) 7 and high expression of CXCL 17 had a better prognosis in LUAD. Immune cell infiltration, specifically B-cell infiltration was significantly correlated with LUAD microenvironment mediated by cysteine(C)-non cysteine(X)-cysteine(C) (CXC) chemokines.⁶⁸ In addition, Arf-GAP with coiled-coil ankyrin (ANK) repeat and PH domain-containing protein 1 (ACAP1) expression appeared to be positively associated with the infiltrating level of immune cells in TIME and the expression of immune checkpoint molecules.⁷⁶ ACAP1 was highly correlated with T-cell activation and immune response. Since overexpression of ACAP1 was found to cause attenuation in cell proliferation, migration, invasion, and promoted apoptosis, low ACAP1 expression was suggested to be correlated with unsatisfactory OS and disease-specific survival in LUAD patients.⁷⁶

Immune cell properties were also found to be correlated with LUAD prognosis. One study reported that heterogeneous carcinoma cell transcriptomes reflecting two distinct microenvironmental patterns, namely N3MC pattern (normal-like myofibroblasts, non-inflammatory monocyte-derived macrophages, NK cells, myeloid dendritic cells, and conventional T cells) and CP2E pattern (cancer-associated myofibroblasts, proinflammatory monocyte-derived macrophages, plasmacytoid dendritic cells, and exhausted CD8+ T cells). It was found that the immune-activated CP2E microenvironment was prognostically unfavorable, whereas the inert N3MC microenvironment was associated with a favorable prognosis.⁶³ Another study pointed out that CD8+ T cells infiltrating the TIME of LUAD are critical for establishing antitumor immunity because the cell differentiation trajectory showed the presence of abundant transition-state CD8+ T cells during the differentiation of naive-like CD8+ T cells into cytotoxic CD8+ T cells and exhausted CD8+ T cells. The higher the proportion of the exhausted CD8+ T lymphocyte (ETL) subset, the shorter the OS of LUAD patients.⁴⁷

The prognosis of LUAD patients can be predicted by a series of TIME markers, including DEGs, specific cell markers, risk score models, and cell components or subtypes. It appeared from the above reports that the subtypes, composition of immune cells and molecular alterations in cancer tissue reflected the cancer progression. These components may be new biomarkers for LUAD prognosis.

The TIME markers for LUAD therapeutic response

Many biomarkers have been investigated and suggested as indicators for therapeutic response in immunotherapy. The state of PD-L1 expression and the status of TMB were suggested as predictive biomarkers. The status of HLA/MHC expression was suggested as a marker of antigen presentation. The status of the oncogenic driver genes (e.g. EGFR, ALK, KRAS, MET) and co-mutations (e.g. STK11, KEAP1, SMARCA4) was also suggested to stratify the patient response.⁹⁰

The markers related to the TIME were also suggested to stratify the patient response in immunotherapy. There are three types of markers on response stratification, including DEGs or specific cell markers, risk scores from combined multiple markers, and cell property markers. The studies on therapeutic response and details of the markers are summarized in Table 3.^{40,44,57,61,64,69,75,76,78,81,82,86,88,89,91–99} Several DEGs have been found to predict patient response. In one study, four immune-related genes, including PTPRC, CCR2, SLAMF1, and HLA-DQA2 were identified as the signature for better outcomes. The four-gene signature was suggested to be used for outcome prediction in LUAD patients. The risk score was calculated by combining their expression levels and coefficients. Further analyses revealed that patients who had a higher risk score were also accompanied by a lower immune infiltration level and a worse response.⁸¹ Similarly, a risk model consisting of nine gene signatures (CASZ1, CCDC85B, CCL20, MAPK1IP1L, MYO6, RHOQ, ST13, TLE1, TMEM11) based on T-cell marker genes was established. High-risk groups presented discriminative immune-cell infiltrations and immune-suppressive states. The authors suggested that the treatment outcomes can be accurately predicted by the risk model.⁸⁶ In another study, 15 N7-methylguanosine (m7G)-related genes (CYFIP1, DCP2, DIF3D, EIF4E, EIF4G3, EIF4E2, LARP1, METTL1, NCBP1, NCBP2, NUDT3, NUDT4, NUDT11, SNUPN, WDR4,) were found to be highly expressed in tumor samples. m7G cluster-B was shown to have a lower immune infiltration level and predicted poor responses to immunotherapy. NUDT4 and WDR4 were identified as independent risk factors.⁸⁸

Table 3.

Summary of therapeutic markers and the indications for LUAD responses in studies on TIME.

Ref No.	Publication year	Therapeutic markers	Indications for therapeutic response	Subject profile	Therapies involved
Li et al.⁹²	2019	Cytotoxic CD8+ T cells and dendritic cells, Foxp3+ Tregs, and M2-like polarization of macrophages	Tumor shrinkage	Transgenic mouse model	Oral gavage: vehicle, gefitinib or osimertinib
Ma et al.⁴⁰	2019	The downregulation of genes in IFN-γ signaling pathways	Acquired resistance phenotype	Two PDX samples, both early and advanced stage LUAD	One male:treatment-naïve one female: standard chemotherapy and erlotinib treatment
Li et al.⁹²	2020	Loss of the histone chaperone Asf1a	Sensitized tumors to anti-PD-1 treatment.	Mouse model and cell lines	Anti-PD-1 antibody (29F.1A12)
Li et al.⁹³	2019	Treg-specific deletion of ST2	Increased infiltration of CD8+ T cells into tumors, and decreases tumor burden.	Genetic mouse model	Interferon
Wu et al.⁵⁷	2020	A riskScore (gene and cell signature) model	predicting immunotherapeutic outcomes	Both early and advanced stage LUAD	Anti-PD-L1 immunotherapy (no specific drugs mentioned)
Guo et al.⁹⁴	2020	Thymocyte selection-associated high mobility group box (TOX) expression	Increase of immune infiltration levels in most of the immune cells	Both early and advanced stage LUAD	ICIs (drugs not specified)
Marinelli et al.⁹⁵	2020	KEAP1, PBRM1, SMARCA4 and STK11	Potentially associated with reduced efficacy of immunotherapy	Metastatic NSCLC patients	ICIs (drugs not specified)
Sage et al.⁹⁶	2020	siRNA-mediated knockdown of AC008750.1	Impaired NKG7 expression and NK cell anti-tumor capacity	Both early and advanced stage LUAD	ICIs (drugs not specified)
Rubio-Perez et al.⁹⁷	2021	CSF CD8+ T cell infiltration	Predicted the response to ICI	LUAD with brain metastasis	ICIs (drugs not specified)
Zheng et al.⁹⁸	2021	A high-resolution TME cell atlas (TME stratification analysis)	Predicted potential resistance regulators of PD-1/PD-L1 blockade	Non-metastatic LUAD	Combination of natural products with ICIs (drugs not specified)
Wang et al.⁶¹	2021	TTN mutations	Potential predictive marker for patients with LUAD to accept ICIs	Late-stage LUAD	ICIs (drugs not specified)
Wu et al.⁴⁴	2021	MSC1 and MSC2 (MSC subtype)	Immune blockade therapy responses	Both early and advanced stage LUAD	ICIs (drugs not specified)
Gong et al.⁶⁴	2021	A ceRNA network of SNHG6-hsa-miR-30e-5p-CYSLTR1	Associated with the prognosis of and immunotherapy response	Both early and advanced stage LUAD	Anti-PD-1 ICIs (drugs not specified)
Yu et al.⁹⁹	2022	Colocalization of CTGF:LRP6	Associated with LUAD progression	Both early and advanced stage LUAD	Anti-CTGF (connective tissue growth factor) therapy
Shinohara et al.⁶⁹	2022	TIME score	TIME score predicted the efficacy of ICI	Resectable LUAD	Anti-PD-1 ICIs (drugs not specified)
Liu et al.⁷⁵	2022	Low levels of Gln metabolism (Gln metabolism-based model)	Predicting prognosis and immunotherapy efficacy	Both early and advanced stage LUAD	Nivolumab or pembrolizumab
Wang et al.⁷⁶	2022	ACAP1 expression	Predictive of prognosis and immunotherapy response	Both early and advanced stage LUAD	ICIs (drugs not specified)
Liu et al.⁷⁸	2022	SLC7A5 high expression	High SLC7A5 expression indicated poor immunotherapy efficacy	Both early and advanced stage LUAD	chemotherapy, radiotherapy, ipilimumab
Li et al.⁸¹	2022	A four-gene prognostic signature	Predicting outcomes and immune landscapes for LUAD patients	Both early and advanced stage LUAD	ICIs (drugs not specified)
Ma et al.⁸²	2022	61 cuproptosis-related LncRNAs prognostic lncRNAs (CLPS)	A predictor for the prognosis and therapeutic responses	Both early and advanced stage LUAD	Chemotherapy and ICIs (drugs not specified)
Zhang et al.⁸⁶	2023	A risk model consisting of nine gene signatures	Predictive for survival and treatment outcomes	Early-stage LUAD	ICIs (drugs not specified)
Liu et al.⁸⁸	2023	m7G cluster-B	Predictive for survival and response to immunotherapy	Both early and advanced stage LUAD	CTLA-4 and PD-1 checkpoint inhibitors (drugs not specified)
Zhang et al.⁸⁹	2023	A risk model consisting of 11 genes	Low-risk group was more likely to benefit from immunotherapy	Both early and advanced stage LUAD	PD-1/PD-L1 or CTLA4 ICIs (drugs not specified), or both

Specific immune cell properties and cell markers in TIME were also shown to predict the treatment response. One study suggested that the T-cell receptor (TCR) clonotypes of the cerebrospinal fluid, specifically CD8+ T-cell infiltration, can provide a non-invasive alternative to predict the response to immune checkpoint inhibitors in LUAD patients with brain metastasis.⁹⁷ Another study confirmed the colocalization of ligand–receptor interaction CTGF:LRP6 among malignant cell subtypes as an indicator predicted to be associated with LUAD progression.⁹⁹ SLC7A5 high-expression was shown to be involved in the activation of multiple oncogenic pathways, including mTORC1, cell cycle, DNA damage repair, response to reactive oxygen, angiogenesis, epithelial–mesenchymal transition, and various growth factors. Its high expression was found to be an indicator of poor immunotherapy efficacy and was an independent prognostic factor. High SLC7A5 expression also indicated higher sensitivity to inhibitors of the mTORC1 pathway, cell cycle, and angiogenesis.⁷⁸

It can be summarized that DEGs, specific cell markers, risk score models, and cell property markers are potential indicators for therapeutic response. These reported markers mostly focus on the response to immunotherapy. More investigations on TIME-related markers are needed in other therapies, including surgery, chemotherapy, radiotherapy, targeted therapy, etc.

The TIME-related drug resistance

The resistance to immunotherapy in LUAD has been investigated, and specific genetic alterations have been suggested as biomarkers for poor response to immunotherapy. EGFR mutations were suggested as biomarkers of poor response to single-agent immune checkpoint inhibitors (ICIs), and combined strategies may be needed for patients with EGFR mutations.^100,101 Co-mutations of KRAS and STK11 in LUAD were associated with an inactive immune microenvironment, mechanistically led to the poor outcomes of KRAS/STK11-mutated patients receiving single-agent immunotherapy, with a relevant proportion of primary resistance.^102,103 There were also reports on the poor effectiveness of ICIs in LUAD with ALK rearrangement.^100,101 Meanwhile, hyperprogressive disease (HPD) after immunotherapy in LUAD has also been reported. Biomarkers linked to HPD in LUAD included MDM2/4 amplification, EGFR mutations, and BRCA2 mutations.¹⁰⁴

The TIME-related drug resistance may involve the dysregulation of anti-tumor activity of Treg cells, exhausted T cells, dendritic cells, MDSCs, and macrophages.¹⁰⁴ Expanded or activated Treg cells may have strong immune suppression, thus hampering the functions of effector T cells, including CD4+ T and CD8+ T cells.^105,106 T-cell exhaustion refers to the progressive decline in T-cell function or dysfunction due to continuous stimulation of TCR under continuous antigen exposure. T-cell exhaustion ranges from highly proliferative T cells with stem-like properties to T cells that have completely lost effector function and the ability to replicate. Exhausted T cells are characterized by reduced cytokine production and inhibited expression of receptors such as PD-1, CTLA4, and LAG3 (lymphocyte activation gene 3 protein), TIGIT [T-cell immunoreceptor with immunoglobulin and immunoreceptor tyrosine-based inhibitory motif (ITIM) domains] and TIM3 (T-cell immunoglobulin and mucin domain-containing protein 3).¹⁰⁴ Dendritic cells are a type of antigen-presenting cells that uptake tumor antigens and play a crucial role in generating the anti-tumor response by T cells. PD-1 on T cells, PD-L1 on DC cells and activation of TGF-β in DCs all contribute to the inhibition of DC activity.^107–109 MDSCs are a group of immunosuppressive cells, which may be involved in regulating immune responses. MDSCs may contribute to tumor angiogenesis, metastasis, and prognosis.¹¹⁰ The roles of MDSCs in LUAD immunotherapy are still not clear, but they might be related to impaired activity of effector T cells, inducing Treg cell expansion, NK cell function reduction, and cytokine secretion.¹¹¹ Macrophages have also been reported to participate in the response to immunotherapy. M2 TAMs were suggested to be associated with pro-tumorigenic properties, and their presence was correlated with a poor prognosis in various tumors.¹¹² In contrast, M1 macrophages were suggested to promote anti-tumor immunity. It was suggested that tumor growth might be inhibited by the depletion of M2 TAMs in LUAD.^104,113 The spatiotemporal evolution of TAMs was shown to be associated with LUAD progression and tumor response to immunotherapy. It was also found that PD-1 was expressed in a highly plastic tumor-promoting subtype of TAMs that develops during tumor progression, and protected tumor cells.¹¹⁴

Apart from immune cells, other biomarkers were also suggested to influence drug resistance. Heterogeneous expression and coregulation with other genes were found in IFN-γ signaling pathway genes in LUAD, and the downregulation of these genes might correspond to an acquired resistance phenotype.⁴⁰ Another study established a model containing 23 circadian-related genes, and used it in predicting the immunotherapeutic outcomes in independent cohorts.¹¹⁵ In addition, patients with high expression of JAG1 were found to correlate with immunosuppressive phenotype, leading to immunotherapy resistance.¹¹⁶

Both molecular alterations and TIME-related immune cell components and activities were found to be markers for drug resistance. They may help to predict resistance before any therapy is implemented, and regulation of key components may also potentially help to overcome resistance. Markers on resistance also need to be developed in other therapeutic strategies, apart from immunotherapy.

Conclusion

The TIME of LUAD is composed of tumor cells, immune cells, support cells, and regulatory factors with great heterogeneity. Immune cells include those associated with acute inflammation, innate immunity, and adaptive immune responses. These cells can be divided by their roles into immune active cells, such as CD8+ T, Th1, NK, DC, M1 TAM cells, and immune suppressive cells, such as M2 TAM, MDSC, Treg, Th17, and exhausted T cells. The dynamic status of these cells, and their interaction with other components in TIME, influence the immune response to surgery, chemotherapy, radiotherapy, targeted therapy, and immunotherapy and mediate the resistance to immunotherapy. Many prognostic and response models composed of DEGs or specific cell markers, risk scores and cell property markers have been summarized in this review. These past findings should be translated into future actionable strategies based on consensus on markers or models in TIME for clinical practice. Future studies may focus on the clinical validation of markers or models in prospective cohorts.

Footnotes

Acknowledgements

None.

Declarations

ORCID iDs

Lele Song

Erpeng Wang

References

Bade

Dela Cruz

CS.

Lung cancer 2020: epidemiology, etiology, and prevention. Clin Chest Med 2020; 41: 1–24.

Hirsch

Scagliotti

Mulshine

, et al. Lung cancer: current therapies and new targeted treatments. J Lancet. 2017; 389: 299–311.

Lim

PC.

Emerging insights of tumor heterogeneity and drug resistance mechanisms in lung cancer targeted therapy. J Hematol Oncol 2019; 12: 134.

Pitt

Marabelle

Eggermont

, et al. Targeting the tumor microenvironment: removing obstruction to anticancer immune responses and immunotherapy. Ann Oncol. 2016; 27: 1482–1492.

Hinshaw

Shevde

LA.

The tumor microenvironment innately modulates cancer progression. Cancer Res 2019; 79: 4557–4566.

Genova

Dellepiane

Carrega

, et al. Therapeutic Implications of tumor microenvironment in lung cancer: focus on immune checkpoint blockade. Front Immunol 2021; 12: 799455.

Liu

Wang

, et al. Tumor immune microenvironment in epidermal growth factor receptor-mutated non-small cell lung cancer before and after epidermal growth factor receptor tyrosine kinase inhibitor treatment: a narrative review. Transl Lung Cancer Res 2021; 10: 3823–3839.

Lei

, et al. Immune cells within the tumor microenvironment: biological functions and roles in cancer immunotherapy. Cancer Lett 2020; 470: 126–133.

Papalexi

Satija

Single-cell RNA sequencing to explore immune cell heterogeneity. Nat Rev Immunol 2018; 18: 35–45.

10.

Lei

Tang

, et al. Applications of single-cell sequencing in cancer research: progress and perspectives. J Hematol Oncol 2021; 14: 91.

11.

Ding

Chen

Shen

Single-cell RNA sequencing in breast cancer: understanding tumor heterogeneity and paving roads to individualized therapy. Cancer Commun 2020; 40: 329–344.

12.

Zhang

Song

Tian

, et al. Single-cell RNA sequencing in lung cancer: revealing phenotype shaping of stromal cells in the microenvironment. Front Immunol 2021; 12: 802080.

13.

Fan

Decoding lung cancer at single-cell level. Front Immunol 2022; 13: 883758.

14.

Wang

Meng

Guo

, et al. Single-cell sequencing, an advanced technology in Lung Cancer Research. Onco Targets Ther 2021; 14: 1895–1909.

15.

Tan

Wang

[Research progress of single cell sequencing in lung cancer]. Chinese 2021; 279: 283.

16.

Lin

Chen

Ding

, et al. [Progress in single-cell RNA sequencing of lung adenocarcinoma]. Zhongguo Fei Ai Za Zhi 2021; 24: 434–440.

17.

Altorki

Markowitz

Gao

, et al. The lung microenvironment: an important regulator of tumour growth and metastasis. Nat Rev Cancer 2019; 19: 9–31.

18.

Pan

Wang

, et al. Tumor-associated macrophages in tumor immunity. Front Immunol 2020; 11: 11–583084. Erratum in: Front Immunol. 2021 Dec 10;12:775758.

19.

Chen

Zhang

, et al. Metabolic regulatory crosstalk between tumor microenvironment and tumor-associated macrophages. Theranostics 2021; 11: 1016–1030.

20.

Jiang

, et al. Natural killer cells in cancer biology and therapy. Mol Cancer 2020; 19: 120.

21.

Guillerey

Huntington

Smyth

MJ.

Targeting natural killer cells in cancer immunotherapy. Nat Immunol 2016; 17: 1025–1036.

22.

Gardner

Ruffell

Dendritic cells and cancer immunity. Trends Immunol 2016; 37: 855–865.

23.

Cohen

Giladi

Barboy

, et al. The interaction of CD4+ helper T cells with dendritic cells shapes the tumor microenvironment and immune checkpoint blockade response. Nat Cancer 2022; 3: 303–317.

24.

Sautès-Fridman

Petitprez

Calderaro

, et al. Tertiary lymphoid structures in the era of cancer immunotherapy. Nat Rev Cancer 2019; 19: 307–325.

25.

Dieu-Nosjean

Giraldo

Kaplon

, et al. Tertiary lymphoid structures, drivers of the anti-tumor responses in human cancers. Immunol Rev 2016; 271: 260–275.

26.

Joyce

Fearon

DT.

T cell exclusion, immune privilege, and the tumor microenvironment. Science (New York, N.Y.) 2015; 348: 74–80.

27.

Crespo

Sun

Welling

, et al. T cell anergy, exhaustion, senescence, and stemness in the tumor microenvironment. Curr Opin Immunol 2013; 25: 214–221.

28.

Tosolini

Kirilovsky

Mlecnik

, et al. Clinical impact of different classes of infiltrating T cytotoxic and helper cells (Th1, th2, treg, th17) in patients with colorectal cancer. Cancer Res 2011; 71: 1263–1271. Erratum in: Cancer Res 2011; 71: 4732.

29.

Schulze

Evers

Görlich

, et al. Tumor infiltrating T cells influence prognosis in stage I-III non-small cell lung cancer. J Thorac Dis 2020; 12: 1824–1842.

30.

Guo

Zhang

Zheng

, et al. Publisher Correction: Global characterization of T cells in non-small-cell lung cancer by single-cell sequencing. Nat Med 2018; 24: 1628–1985. Erratum in: Nat Med. 2018 Oct;24(10):1628.

31.

Joshi

Akama-Garren

, et al. Regulatory T cells in tumor-associated tertiary lymphoid structures suppress anti-tumor T cell responses. Immunity. 2015; 43: 579–590.

32.

Zhang

Guo

, et al. CD8+ T cell-associated gene signature correlates with prognosis risk and immunotherapy response in patients with lung adenocarcinoma. Front Immunol 2022; 13: 806877.

33.

Han

Duan

Bai

, et al. TCR repertoire diversity of Peripheral PD-1+CD8+ T cells predicts clinical outcomes after immunotherapy in patients with non-small cell lung cancer. Cancer Immunol Res 2020; 8: 146–154.

34.

Feng

Zhang

, et al. A robust CD8+ T cell-related classifier for predicting the prognosis and efficacy of immunotherapy in stage III lung adenocarcinoma. Front Immunol 2022; 13: 993187.

35.

Hao

Han

Sinjab

, et al. The single-cell immunogenomic landscape of B and plasma cells in early-stage lung adenocarcinoma. Cancer Discov 2022; 12: 2626–2645.

36.

Song

, et al. Integrated analysis of single-cell and bulk RNA-sequencing identifies a signature based on B cell marker genes to predict prognosis and immunotherapy response in lung adenocarcinoma. Cancer Immunol Immunother 2022; 71: 2341–2354.

37.

de Sousa

VML

Carvalho

. Heterogeneity in Lung Cancer. Pathobiology 2018; 85: 96–107.

38.

Skoulidis

Heymach

JV.

Co-occurring genomic alterations in non-small-cell lung cancer biology and therapy. Nat Rev Cancer 2019; 19: 495–509.

39.

Fan

, et al. Single-cell profiling of tumor heterogeneity and the microenvironment in advanced non-small cell lung cancer. Nat Commun 2021; 12: 2540.

40.

Schonnesen

Brock

, et al. Single-cell RNA sequencing of lung adenocarcinoma reveals heterogeneity of immune response-related genes. JCI Insight 2019; 4: e121387.

41.

Liu

Wang

, et al. Mechanisms of progression and heterogeneity in multiple nodules of lung adenocarcinoma. Small Methods 2021; 5: e2100082.

42.

Wang

, et al. Single-cell RNA sequencing reveals heterogeneous tumor and immune cell populations in early-stage lung adenocarcinomas harboring EGFR mutations. Oncogene 2021; 40: 355–368.

43.

Wang

Peng

, et al. Single-cell analysis reveals spatial heterogeneity of immune cells in lung adenocarcinoma. Front Cell Dev Biol 2021; 9: 638374.

44.

Qin

Zhao

, et al. Microdissection of the bulk transcriptome at single-cell resolution reveals clinical significance and myeloid cells heterogeneity in lung adenocarcinoma. Front Immunol 2021; 12: 723908.

45.

Zheng

Zhang

Jiang

, et al. Establishment of a prognostic model of lung adenocarcinoma based on tumor heterogeneity. Front Mol Biosci 2022; 9: 807497.

46.

Wang

Song

, et al. The heterogeneous immune landscape between lung adenocarcinoma and squamous carcinoma revealed by single-cell RNA sequencing. Signal Transduct Target Ther 2022; 7: 289.

47.

Song

Zhao

Wang

, et al. Heterogeneity and differentiation trajectories of infiltrating CD8+ T cells in lung adenocarcinoma. Cancers (Basel) 2022; 14: 5183.

48.

Fan

Niu

, et al. Immune and non-immune cell subtypes identify novel targets for prognostic and therapeutic strategy: A study based on intratumoral heterogenicity analysis of multicenter scRNA-seq datasets in lung adenocarcinoma. Front Immunol 2022; 13: 1046121.

49.

Yao

Zuo

, et al. Necroptosis-related lncRNAs: combination of bulk and single-cell sequencing reveals immune landscape alteration and a novel prognosis stratification approach in lung adenocarcinoma. Front Oncol 2022; 12: 1010976.

50.

Sun

Lao

, et al. Single-cell RNA sequencing reveals cellular and molecular reprograming landscape of gliomas and lung cancer brain metastases. Clin Transl Med 2022; 12: e1101.

51.

Sinjab

Han

Treekitkarnmongkol

, et al. Resolving the spatial and cellular architecture of lung adenocarcinoma by multiregion single-cell sequencing. Cancer Discov 2021; 11: 2506–2523.

52.

Zhu

Fan

Xiong

, et al. Delineating the dynamic evolution from preneoplasia to invasive lung adenocarcinoma by integrating single-cell RNA sequencing and spatial transcriptomics. Exp Mol Med 2022; 54: 2060–2076.

53.

Kim

Lee

, et al. Single-cell RNA sequencing demonstrates the molecular and cellular reprogramming of metastatic lung adenocarcinoma. Nat Commun 2020; 11: 2285.

54.

Brueckl

Ficker

Zeitler

Clinically relevant prognostic and predictive markers for immune-checkpoint-inhibitor (ICI) therapy in non-small cell lung cancer (NSCLC). BMC Cancer 2020; 20: 1185.

55.

Mansuet-Lupo

Alifano

Pécuchet

, et al. Intratumoral immune cell densities are associated with lung adenocarcinoma gene alterations. Am J Respir Crit Care Med 2016; 194: 1403–1412.

56.

Pocha

Mock

Rapp

, et al. Surfactant expression defines an inflamed subtype of lung adenocarcinoma brain metastases that correlates with prolonged survival. Clin Cancer Res 2020; 26: 2231–2243.

57.

Wang

Wei

, et al. Integration of multiple key molecules in lung adenocarcinoma identifies prognostic and immunotherapeutic relevant gene signatures. Int Immunopharmacol 2020; 83: 106477.

58.

Tao

Cai

Profiles of immune infiltration and prognostic immunoscore in Lung Adenocarcinoma. Biomed Res Int 2020; 2020: 5858092.

59.

Zeng

Song

, et al. Stemness related genes revealed by network analysis associated with tumor immune microenvironment and the clinical outcome in lung adenocarcinoma. Front Genet 2020; 11: 549213.

60.

Zhao

Liu

, et al. A novel intercellular communication-associated gene signature for prognostic prediction and clinical value in patients with lung adenocarcinoma. Front Genet 2021; 12: 702424.

61.

Wang

Lin

, et al. Effect of TTN mutations on immune microenvironment and efficacy of immunotherapy in lung adenocarcinoma patients. Front Oncol 2021; 11: 725292.

62.

Zhong

Wang

Zhu

, et al. Comprehensive analysis of a nine-gene signature related to tumor microenvironment in lung adenocarcinoma. Front Cell Dev Biol 2021; 9: 700607.

63.

Bischoff

Trinks

Obermayer

, et al. Single-cell RNA sequencing reveals distinct tumor microenvironmental patterns in lung adenocarcinoma. Oncogene 2021; 40: 6748–6758.

64.

Gong

Zhou

, et al. A novel immune-related ceRNA network that predicts prognosis and immunotherapy response in lung adenocarcinoma. Ann Transl Med 2021; 9: 1484.

65.

Zhou

Bian

Qian

, et al. Prognostic model of lung adenocarcinoma constructed by the CENPA complex genes is closely related to immune infiltration. Pathol Res Pract 2021; 228: 153680.

66.

Zhou

Tang

Chen

, et al. An immune panel signature predicts prognosis of lung adenocarcinoma patients and correlates with immune microenvironment. Front Cell Dev Biol 2021; 9: 797984.

67.

Navab

, et al. Tumor-associated regulatory T cell expression of LAIR2 is prognostic in Lung adenocarcinoma. Cancers (Basel) 2021; 14: 205.

68.

Wang

Zhang

, et al. Prognostic significance and therapeutic target of CXC chemokines in the microenvironment of lung adenocarcinoma. Int J Gen Med 2022; 15: 2283–2300.

69.

Shinohara

Takahashi

Komuro

, et al. New evaluation of the tumor immune microenvironment of non-small cell lung cancer and its association with prognosis. J Immunother Cancer 2022; 10: e003765.

70.

Noyes

Bag

Oseni

, et al. Tumor-associated Tregs obstruct antitumor immunity by promoting T cell dysfunction and restricting clonal diversity in tumor-infiltrating CD8+ T cells. J Immunother Cancer 2022; 10: e004605.

71.

Guan

Cai

, et al. Intercellular communication-related molecular subtypes and a gene signature identified by the single-cell RNA sequencing combined with a transcriptomic analysis. Dis Markers 2022; 2022: 1–22.

72.

Zhu

[The expression of RTN1 in lung adenocarcinoma and its effect on immune microenvironment]. Zhongguo Fei Ai Za Zhi. 2022; 25: 385–395. Chinese.

73.

Jin

Sui

, et al. A novel prognostic signature revealed the interaction of immune cells in tumor microenvironment based on single-cell RNA sequencing for lung adenocarcinoma. J Immunol Res 2022; 2022: 6555810.

74.

Luo

Deng

Que

, et al. Cell trajectory-related genes of lung adenocarcinoma predict tumor immune microenvironment and prognosis of patients. Front Oncol 2022; 12: 911401.

75.

Liu

Shen

, et al. Prediction of prognosis, immunogenicity and efficacy of immunotherapy based on glutamine metabolism in lung adenocarcinoma. Front Immunol 2022; 13: 960738.

76.

Wang

Zhu

, et al. Integrated analysis and validation reveal ACAP1 as a novel prognostic biomarker associated with tumor immunity in lung adenocarcinoma. Comput Struct Biotechnol J 2022; 20: 4390–4401.

77.

Zhong

Wang

Han

, et al. XBP1 impacts lung adenocarcinoma progression by promoting plasma cell adaptation to the tumor microenvironment. Front Genet 2022; 13: 969536.

78.

Liu

, et al. SLC7A5 is a lung adenocarcinoma-specific prognostic biomarker and participates in forming immunosuppressive tumor microenvironment. Heliyon 2022; 8: e10866.

79.

Hao

Chen

, et al. Integrated analysis of bulk and single-cell RNA-seq reveals the role of MYC signaling in lung adenocarcinoma. Front Genet 2022; 13: 1021978.

80.

Shi

Han

Liu

, et al. Integrated single-cell and transcriptome sequencing analyses determines a chromatin regulator-based signature for evaluating prognosis in lung adenocarcinoma. Front Oncol 2022; 12: 1031728.

81.

Tong

Zhang

, et al. Single-cell sequencing analysis and transcriptome analysis constructed the macrophage related gene-related signature in lung adenocarcinoma and verified by an independent cohort. Genomics 2022; 114: 110520.

82.

Zhu

Wang

, et al. A cuproptosis-related long non-coding RNA signature to predict the prognosis and immune microenvironment characterization for lung adenocarcinoma. Transl Lung Cancer Res 2022; 11: 2079–2093.

83.

Zeng

Zhou

, et al. A prognostic signature model for unveiling tumor progression in lung adenocarcinoma. Front Oncol 2022; 12: 1019442.

84.

Van Hiep

Sun

Feng

, et al. Corrigendum: Heparin binding epidermal growth factor-like growth factor is a prognostic marker correlated with levels of macrophages infiltrated in lung adenocarcinoma. Front Oncol 2022; 12: 1106553. Erratum in: Front Oncol. 2022 Dec 13;12:1106553.

85.

Zhang

, et al. Single-cell analysis of multiple cancer types reveals differences in endothelial cells between tumors and normal tissues. Comput Struct Biotechnol J 2023; 21: 665–676.

86.

Zhang

Liu

Huang

, et al. T cell-related prognostic risk model and tumor immune environment modulation in lung adenocarcinoma based on single-cell and bulk RNA sequencing. Comput Biol Med 2023; 152: 106460.

87.

Zhang

, et al. Single-cell analysis reveals the COL11A1+ fibroblasts are cancer-specific fibroblasts that promote tumor progression. Front Pharmacol 2023; 14: 1121586.

88.

Liu

Jiang

Lin

, et al. m7G-related gene NUDT4 as a novel biomarker promoting cancer cell proliferation in lung adenocarcinoma. Front Oncol 2022; 12: 1055605.

89.

Zhang

Pei

Gong

, et al. By integrating single-cell RNA-seq and bulk RNA-seq in sphingolipid metabolism, CACYBP was identified as a potential therapeutic target in lung adenocarcinoma. Front Immunol 2023; 14: 1115272.

90.

Sholl

LM.

Biomarkers of response to checkpoint inhibitors beyond PD-L1 in lung cancer. Mod Pathol 2022; 35: 66–74.

91.

Jia

Jiang

, et al. EGFR-targeted therapy alters the tumor microenvironment in EGFR-driven lung tumors: implications for combination therapies. Int J Cancer 2019; 145: 1432–1444.

92.

Huang

Luster

, et al. in vivo epigenetic CRISPR screen identifies asf1a as an immunotherapeutic target in Kras-mutant lung adenocarcinoma. Cancer Discov. 2020; 10: 270–287.

93.

Herbst

Canner

, et al. IL-33 signaling alters regulatory T cell diversity in support of Tumor Development. Cell Rep 2019; 29: 2998–3008.e8.

94.

Guo

Liu

, et al. TOX correlates with prognosis, immune infiltration, and T cells exhaustion in lung adenocarcinoma. Cancer Med 2020; 9: 6694–6709.

95.

Marinelli

Mazzotta

Scalera

, et al. KEAP1-driven co-mutations in lung adenocarcinoma unresponsive to immunotherapy despite high tumor mutational burden. Ann Oncol 2020; 31: 1746–1754.

96.

Sage

Marshall

, et al. Assessment of long non-coding RNA expression reveals novel mediators of the lung tumour immune response. Sci Rep 2020; 10: 16945.

97.

Rubio-Perez

Planas-Rigol

Trincado

, et al. Immune cell profiling of the cerebrospinal fluid enables the characterization of the brain metastasis microenvironment. Nat Commun 2021; 12: 1503.

98.

Zheng

Xiao

Chen

, et al. Systems pharmacology: a combination strategy for improving efficacy of PD-1/PD-L1 blockade. Brief Bioinform 2021; 22: bbab130.

99.

, et al. Reconstructing codependent cellular cross-talk in lung adenocarcinoma using REMI. Sci Adv 2022; 8: eabi4757.

100.

Guaitoli

Tiseo

Di Maio

, et al. Immune checkpoint inhibitors in oncogene-addicted non-small cell lung cancer: a systematic review and meta-analysis. Transl Lung Cancer Res 2021; 10: 2890–2916.

101.

Spella

Stathopoulos

GT.

Immune resistance in lung adenocarcinoma. Cancers (Basel) 2021; 13: 384.

102.

Skoulidis

Byers

Diao

, et al. Co-occurring genomic alterations define major subsets of KRAS-mutant lung adenocarcinoma with distinct biology, immune profiles, and therapeutic vulnerabilities. Cancer Discov. 2015; 5: 860–877.

103.

Skoulidis

Goldberg

Greenawalt

, et al. STK11/LKB1 mutations and PD-1 inhibitor resistance in KRAS-Mutant lung adenocarcinoma. Cancer Discov 2018; 8: 822–835.

104.

Wang

Zhong

, et al. The biomarkers of hyperprogressive disease in PD-1/PD-L1 blockage therapy. Mol Cancer 2020; 19: 81.

105.

Corthay

How do regulatory T cells work?

Scand J Immunol 2009; 70: 326–336.

106.

Wing

Tanaka

Sakaguchi

Human FOXP3+ regulatory T cell heterogeneity and function in autoimmunity and cancer. Immunity 2019; 50: 302–316.

107.

Hassannia

Ghasemi Chaleshtari

Atyabi

, et al. Blockage of immune checkpoint molecules increases T-cell priming potential of dendritic cell vaccine. Immunology 2020; 159: 75–87.

108.

Martín-Fontecha

Sebastiani

Höpken

, et al. Regulation of dendritic cell migration to the draining lymph node: impact on T lymphocyte traffic and priming. J Exp Med 2003; 198: 615–621.

109.

Flavell

Sanjabi

Wrzesinski

, et al. The polarization of immune cells in the tumour environment by TGFbeta. Nat Rev Immunol 2010; 10: 554–567.

110.

Kumar

Patel

Tcyganov

, et al. The nature of myeloid-derived suppressor cells in the tumor microenvironment. Trends Immunol 2016; 37: 208–220.

111.

Weber

Fleming

, et al. Myeloid-derived suppressor cells hinder the anti-cancer activity of immune checkpoint inhibitors. Front Immunol 2018; 9: 1310.

112.

Pollard

JW.

Tumour-educated macrophages promote tumour progression and metastasis. Nat Rev Cancer 2004; 4: 71–78.

113.

Fritz

Tennis

Orlicky

, et al. Depletion of tumor-associated macrophages slows the growth of chemically induced mouse lung adenocarcinomas. Front Immunol 2014; 5: 587–587. Erratum in: Front Immunol. 2015;6:88.

114.

Jang

Lee

, et al. Therapeutic targeting of macrophage plasticity remodels the tumor-immune microenvironment. Cancer Res 2022; 82: 2593–2609.

115.

Fan

Zhang

, et al. Single-cell transcriptomic analysis reveals circadian rhythm disruption associated with poor prognosis and drug-resistance in lung adenocarcinoma. J Pineal Res 2022; 73: e12803.

116.

, et al. JAG1 is correlated to suppressive immune microenvironment and predicts immunotherapy resistance in lung adenocarcinoma. Front Oncol 2023; 13: 1091488.

Unraveling the tumor immune microenvironment of lung adenocarcinoma using single-cell RNA sequencing

Abstract

Keywords

Introduction

The components of TIME and their roles in LUAD development

The heterogeneity of TIME in LUAD

The prognostic TIME markers for LUAD

The TIME markers for LUAD therapeutic response

The TIME-related drug resistance

Conclusion

Footnotes

Acknowledgements

Declarations

ORCID iDs

References