Abstract

Background

   The aim of this study is to create an index called IRGPRI
   (immune-related gene prognostic risk index) that can be utilized for
   predicting the prognosis and assessing the efficacy of immune
   checkpoint inhibitors (ICIs) therapy in patients with non-small-cell
   lung cancer (NSCLC).

Methods

   Distinguishing gene expression patterns (DEGs) were detected in CD8^+
   cytotoxic T lymphocytes (CTLs) compared to other cellular types such as
   CD4 T cells, B cells, plasma cells, and CD8 Tex using the advanced
   technology of Single-cell RNA Sequencing (scRNA-seq). The construction
   of IRGPRI was accomplished by employing LASSO Cox regression analysis.
   We conducted a comparative analysis on clinical characteristics and
   molecular features, such as pathway enrichment and gene mutation, among
   the distinct subgroups of IRGPRI. Furthermore, we explored the
   correlation between immunological characteristics and IRGPRI subgroups
   to comprehensively assess the effectiveness of ICIs in NSCLC patients.

Results

   A total of 109 genes were identified by intersecting immune-related
   genes with DEGs obtained from single-cell RNA sequencing data
   ([35]GSE131907), specifically comparing CTLs to other cell types. From
   these, we selected 7 prognosis-related genes, namely TRBC1, HLA-DMA,
   CTSH, RAC1, CTSL, ANXA2, and CEBPB. These genes were used to construct
   the IRGPRI. The prognosis of patients diagnosed with NSCLC was found to
   be significantly better in the low-risk group compared to the high-risk
   group, as demonstrated by Kaplan-Meier (K-M) survival analysis. This
   observation was further confirmed through the utilization of data from
   the GEO cohort. The low-risk group demonstrated an increase in pathways
   linked with immune response, whereas the high-risk group exhibited a
   higher prevalence of pathways related to cancer. Furthermore, it was
   noted in the TCGA cohort that there existed a significant rise in the
   mutation frequency of every gene within the high-risk group as opposed
   to the low-risk group. Missense variation emerged as the most prevalent
   form of mutation. According to the analysis of immune cell infiltration
   and function, the comprehensive findings suggest that the group with a
   low risk is characterized by an increased presence of plasma cells,
   CTLs, T cells follicular helper, Tregs, and Dendritic cell resting.
   Additionally, they exhibit a higher score in terms of immune function
   for B cells, CD8^+ T cells, checkpoint activity, T cell inhibition and
   stimulation. Moreover, this low-risk group demonstrates greater
   efficacy when treated with ICIs therapy compared to the high-risk
   group.

Conclusions

   Our research effectively developed and verified a unique IRGPRI,
   showcasing its association with immune-related characteristics. As a
   result, the potential of IRGPRI as a valuable biomarker for predicting
   prognosis and evaluating the effectiveness of ICIs treatment in cancer
   is evident.

   Keywords: Non-small-cell lung cancer, IRGPRI, CTLs, Prognosis, Immune
   checkpoint inhibitors

1. Introduction

   Lung cancer is a primary malignant tumor that has a high prevalence
   globally, leading to a significant number of deaths each year,
   approximately 1.6 million [[36]1]. NSCLC, which accounts for around
   85 % of all cases, can be classified into three main subtypes based on
   the cellular origin: lung adenocarcinoma (LUAD), lung squamous cell
   carcinoma (LUSC), and large cell carcinoma [[37]2]. Due to the absence
   of clear clinical manifestations, patients with NSCLC frequently
   receive a late-stage diagnosis, resulting in missed opportunities for
   receiving the most effective treatment. Despite various anticancer
   approaches such as surgery, radiation therapy, and chemotherapy being
   employed for NSCLC management, there remains a considerable need for
   effective curative or palliative strategies, especially in individuals
   with advanced-stage disease [[38]3]. Recent advancements in therapeutic
   approaches, such as targeted therapy and immunotherapy, have
   demonstrated significant advancements in enhancing the overall survival
   (OS) and quality of life among individuals diagnosed with NSCLC
   [[39]4,[40]5]. The preference for chemotherapy as the primary treatment
   option for all NSCLC patients is no longer universally endorsed.

   Tumor cells exploit various immune checkpoint pathways to evade
   detection by the immune system and facilitate tumor progression.
   Immunotherapies targeting immune checkpoints, such as PD-1, PD-L1, and
   CTLA-4 inhibitors, have received regulatory approval for treating
   diverse cancer types like melanoma, breast cancer, and NSCLC [[41][6],
   [42][7], [43][8], [44][9]]. Immunotherapy is commonly employed in
   managing advanced and metastatic NSCLC. Encouraging outcomes have been
   observed in NSCLC with the use of PD-1 inhibitors like pembrolizumab
   and nivolumab, showing an objective response rate (ORR) ranging from
   40 % to 45 % [[45]6,[46]7]. In a clinical trial called IMpower-110,
   investigators analyzed individuals diagnosed with advanced NSCLC who
   had not received any prior therapeutic interventions. The analysis
   focused on individuals with either ≥50 % PD-L1 tumor cells or ≥10 %
   immune cells infiltrating the tumor. The findings revealed that
   atezolizumab significantly extended OS by 7 months compared to
   chemotherapy in this group [[47]10]. Targeting CTLA-4 using ipilimumab
   has shown up to a 0.84-year improvement in progression-free survival
   (PFS) among NSCLC patients [[48]11].

   While it has been proven that anti-PD-1/PD-L1 and anti-CTLA-4 therapies
   are effective in treating NSCLC, there is indication that their
   advantages may be restricted to a particular subgroup of patients. This
   underscores the challenges associated with implementing ICIs therapy
   for NSCLC. The immune response against tumors is influenced by the
   tumor microenvironment (TME), leading to a range of immune reactions.
   The TME can be categorized into three groups based on the infiltration
   of immune cells: immune desert, immune excluded, and immune inflamed
   [[49]12]. Immune inflamed phenotypes are characterized by the presence
   of various subtypes of immune cells including CD4^+ T cells, CTLs,
   CD20^+ B cells, among others. Activated CTLs selectively recognize and
   eliminate host cells expressing intracellular antigens from infection
   or cancer to play an essential role in anti-cancer immunity [[50]13].
   Some patients receiving ICIs therapy may experience adverse effects;
   hence better strategies are needed to reduce their risk in future
   studies [[51]14]. Patients who develop resistance to ICIs therapy have
   poor prognoses after treatment; therefore identifying predictive
   biomarkers is critical for indicating which patients will most likely
   benefit from this treatment approach [[52]15].

   Our investigation aimed to explore potential biomarkers that can
   predict the response to ICIs treatment and evaluate the therapeutic
   value of ICIs in NSCLC. To achieve this, we utilized scRNA-seq
   technology to identify DEGs between CTLs and other cell types. The
   IRGPRI was constructed using genes associated with prognosis, and we
   further investigated the molecular and immunological characteristics of
   IRGPRI while assessing its predictive ability for both prognosis and
   the efficacy of ICI therapy. Our findings suggest that the IRGPRI shows
   significant promise as a reliable biomarker for accurately predicting
   prognosis and response to ICIs treatment in cancer patients.

2. Materials and methods

2.1. Data sources and data processing

   We collected RNA-seq data, clinical information, and somatic mutation
   variation data from the TCGA database
   ([53]https://portal.gdc.cancer.gov), which included 1153 NSCLC samples
   comprising of 1043 NSCLC tissues and 110 para-cancer tissues. As a
   validation cohort, we also retrieved a microarray dataset
   ([54]GSE101929) [[55]16] consisting of 66 NSCLC patients from the GEO
   database ([56]https://www.ncbi.nlm.nih.gov/geo/). To calculate the TMB,
   we utilized a Perl script that determined the total number of somatic
   mutation variations per million bases. In order to identify
   immune-related genes, we conducted searches in both ImmPort
   ([57]https://www.immport.org/shared/home) and InnateDB
   ([58]https://www.innateDBdb.com/) databases [[59]17]. For our analysis
   on DEGs, scRNA-seq data was obtained from the Tumor Immune Single-cell
   Hub 2 (TISCH2) database ([60]http://tisch.comp-genomics.org/), an
   online platform providing comprehensive and freely accessible scRNA-seq
   information globally [[61]18].

2.2. Identification and biological function analysis of intersection genes

   We employed the “Filter” package in R software to generate a list of
   DEGs between CTLs and other cell types. Our criteria for DEGs selection
   were |log2(Fold Change)|>1.3 and a false discovery rate (FDR) < 0.05.
   The genes at the intersection were analyzed for enrichment in Gene
   Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG)
   pathways using Sangerbox3.0, with FDR< 0.1 and P < 0.05.

2.3. Construction and validation of the IRGPRI

   To identify genes linked with prognosis, we conducted univariate Cox
   regression analysis using R software packages such as ‘survival’ and
   ‘survminer’, setting a significance threshold of P < 0.05. We utilized
   the ‘glmnet’ package in R to perform LASSO Cox regression analysis,
   which helped us identify relevant variables and reduce dimensionality.
   Subsequently, we employed multivariate Cox regression analysis using
   the selected variables. In order to compute the IRGPRI for each sample
   in our risk model, we applied the subsequent formula:
   IRGPRI = [Expression level (prognosis-related genes) × gene
   coefficient]. Reclassifying NSCLC patients from the TGCA cohort into
   low-risk and high-risk groups according the median score of IRGPRI, we
   performed K-M survival curve analysis to assess the prognostic
   significance of IRGPRI. To validate our findings, we utilized another
   cohort from GEO to evaluate the prognostic value of IRGPRI.
   Furthermore, we employed the R software packages “survival” to examine
   the correlations between risk scores and clinical characteristic such
   as age, gender, stage, and OS in patients from the TGCA cohort using
   both univariate and multivariate Cox regression analyses. We conducted
   statistical analyses to determine hazard ratios (HR), 95 % confidence
   intervals (CI), and P-values, which helped us identify significant
   independent prognostic factors. To visually compare clinical
   characteristics between low-risk and high-risk groups, we utilized R
   software packages such as “Biocmanager,” “Complex Heatmap,” and
   “ggplot2” to generate graphical representations like heatmaps and
   circular graphs.

2.4. Comprehensive analysis of molecular and immunological features in
distinct IRGPRI subgroups

   We utilized the gene set enrichment analysis (GSEA) method to
   investigate the enrichment of pathways in various subcategories of
   IRGPRI. Statistical testing was conducted using a significance level of
   P < 0.05. To generate a gene mutation map for the two subgroups of
   IRGPRI, we utilized the “maftools” R software package. The R software
   package ‘limma’ was utilized to conduct a comparative analysis of
   checkpoint genes expression between the low-risk and high-risk groups.
   The relative distributions of 22 types of immune cells were assessed
   using the CIBERSORT algorithm ([62]https://cibersortx.stanford.edu/)
   and visualized through box plots. To investigate immune function
   related to distinct IRGPRI subgroups, we conducted single sample gene
   set enrichment analysis (ssGSEA). Comparisons between the high-risk and
   low-risk groups were made using both the stat_compare_means function
   and Wilcoxon test. We evaluated the prognostic significance of IRGPRI
   in NSCLC patients receiving immunotherapy by analyzing data from
   IMvigor210 immunotherapy cohorts and generating K-M survival curves.
   Additionally, we examined how immunotherapy impacted the two subgroups
   defined by IRGPRI using three R software packages: “limma”, “ggpubr”,
   and “pROC".

3. Results

3.1. Identification of intersection genes

   We acquired scRNA-seq data ([63]GSE131907) from the TISCH2 database,
   which comprised 44 patients and a total of 203298 cells. Utilizing
   cell-type annotations based on major-lineage classification, we
   classified the 25 clusters of cells into 12 distinct cell types. Among
   these, CTLs accounted for the third highest proportion at 16 % of all
   cell types. Given their crucial role as central effector cells in
   mediating the effectiveness of Immune checkpoint blockade (ICB), this
   suggests that ICB immunotherapy may yield favorable therapeutic
   outcomes in NSCLC. By analyzing differentially expressed genes (DEGs)
   between CTLs and other cell types such as CD4 T cells, B cells, plasma
   cells, and CD8 Tex using scRNA-seq data ([64]GSE131907), we identified
   a set of immune-related genes through intersection analysis with the
   aid of “VennDiagram” R software package. This resulted in a total of
   109 overlapping genes.

3.2. Gene function and pathways enrichment analysis

   The analysis of GO revealed that the genes expressed in CTLs were
   mainly associated with processes related to the immune system,
   including responses to stimuli, defense mechanisms, and regulation of
   immune system processes ([65]Fig. 1A and B). In addition, the analysis
   of KEGG pathways revealed that the hub genes played a role in various
   signaling pathways including rheumatoid arthritis, infection caused by
   Epstein-Barr virus, antigen processing and presentation,
   differentiation of Th17 cells, as well as differentiation of Th1 and
   Th2 cells ([66]Fig. 1C and D). By utilizing this information
   effectively, we conducted further investigations into potential
   biomarkers for NSCLC.

Fig. 1.

   [67]Fig. 1
   [68]Open in a new tab

   Gene function and pathways enrichment analysis. (A, B) GO analysis for
   the intersection genes. (C, D) KEGG pathway analysis for the
   intersection genes.

3.3. Construction and validation of the IRGPRI

   The prognosis survival of NSCLC patients was found to be significantly
   influenced by 11 immune-related genes, as indicated by the
   statistically significant results obtained from the Univariate Cox
   regression analysis ([69]Fig. 2A). Subsequently, a 1000-fold
   cross-validation was performed to conduct LASSO Cox regression analysis
   on the set of 11 genes. This analysis identified seven genes (TRBC1,
   HLA-DMA, CTSH, RAC1, CTSL, ANXA2, and CEBPB) that significantly
   influenced the OS of NSCLC patients. These prognostic genes were
   utilized to construct the IRGPRI score ([70]Fig. 2B and C).
   Reclassifying patients from the TCGA cohort into low-risk and high-risk
   groups was based on utilizing the median IRGPRI score as a threshold
   value. The K-M survival analysis revealed a significant difference in
   prognosis between individuals classified as high-risk and low-risk
   groups. Furthermore, the prognostic value of IRGPRI was validated using
   data from the GEO cohort ([71]Fig. 2D).

Fig. 2.

   [72]Fig. 2
   [73]Open in a new tab

   Construct and validate the IRGPRI. (A) Forest diagram illustrating the
   predictive significance of immune-related genes in NSCLC through hazard
   ratios. (B) Seven prognostic-related genes were identified by LASSO
   regression analysis. (C) Application of K-M survival analysis on a set
   of 7 genes associated with prognosis. (D) K-M survival analysis was
   conducted on the subgroups of IRGPRI in both TCGA cohort and GEO
   cohort.

3.4. Clinical characteristics in different IRGPRI subgroups

   The heatmap and circular plots revealed notable differences in
   clinicopathological factors, such as gender, stage, T, and N, between
   the high-risk and low-risk groups ([74]Fig. 3A and B). Univariate Cox
   regression analysis demonstrated that the risk score (HR = 3.422;
   95%CI: 2.507–4.669), age (HR = 1.012; 95%CI: 1.001–1.024), and stage
   (HR = 1.452; 95%CI: 1.305–1.615) were statistically significant
   prognostic determinants for NSCLC patients (P < 0.05) ([75]Fig. 3C).
   Furthermore, multivariate analysis indicated that the risk score
   remained an independent predictor of overall survival in NSCLC patients
   after adjusting for other clinicopathological factors (HR = 3.216;
   95%CI:2.333–4.434) along with age (HR = 1.020; 95%CI:1 0.008–1.031) and
   stage (HR = l.402; 95%CI:l.257-l.564) with P < 0.05 ([76]Fig. 3D).

Fig. 3.

   [77]Fig. 3
   [78]Open in a new tab

   Clinical characteristics in different IRGPRI subgroups. (A, B) Analyze
   the differences in Clinical characteristics between the distinct IRGPRI
   subgroups. (C, D) Analysis of the IRGPRI score and other clinical
   characteristics using both univariate and multivariate Cox regression.

3.5. Correlation between molecular characteristics and risk signature

   We conducted Gene set enrichment analysis (GSEA) to identify the gene
   sets that were enriched in different subgroups of IRGPRI within the
   TCGA cohort ([79]Fig. 4A and B). The low-risk group exhibited
   significant enrichment of gene sets associated with the immune network
   involved in IgA production and immunodeficiency, while the high-risk
   group showed significant associations with cell cycle regulation, ECM
   receptor interaction, and pathways related to cancer. Additionally, we
   performed gene mutation analysis to investigate the immunological
   characteristics of IRGPRI subgroups within the TCGA cohort ([80]Fig. 4C
   and D). The distinct IRGPRI subgroups displayed a prevalence of
   missense mutations followed by multi-hit mutations in the top 20 genes
   with the highest mutation rates. In comparison to the low-risk group,
   all genes exhibited elevated mutation rates in the high-risk group.
   Essentially, a greater burden of tumor mutations was observed in the
   high-risk group when compared to the low-risk group.

Fig. 4.

   [81]Fig. 4

   [82]Fig. 4
   [83]Open in a new tab

   Association between molecular characteristics and risk signature. (A,
   B) Conducting KEGG pathway analysis on gene sets that exhibit
   enrichment in the different subgroups of IRGPRI. (C, D) Mutation
   waterfall in the distinct IRGPRI subgroups.

3.6. The infiltration and function score of immune cells in different IRGPRI
subgroups

   The infiltration of immune cells in patients with NSCLC was evaluated
   using the CIBERSORT algorithm.

   The results indicated that the IRGPRI subgroups exhibited distinct
   proportions of infiltrating immune cells. Specifically, the low-risk
   group showed a higher prevalence of activated plasma cells, CTLs, T
   cells follicular helper, Tregs, and resting dendritic cells (P < 0.05),
   whereas the high-risk group demonstrated a greater prevalence of
   activated macrophages M0, macrophages M2, and neutrophils (P < 0.05)
   ([84]Fig. 5A). Subsequently, ssGSEA analysis was conducted to explore
   differences in immune function among the different IRGPRI subgroups
   ([85]Fig. 5B). The results demonstrated that compared to the high-risk
   group, the low-risk group exhibited higher immune function scores in B
   cells, CTLs, checkpoint signaling pathway activity related genes
   expression levels associated with T cell inhibition and stimulation
   (P < 0.05). Furthermore, an investigation into the expression patterns
   of well-studied classic checkpoint genes across distinct IRGPRI
   subgroups was performed to determine whether NSCLC patients could
   benefit from immunotherapy targeting these checkpoints ([86]Fig. 5C).
   The findings revealed a potential association between increased risk
   scores and reduced levels of checkpoint gene expression, potentially
   indicating a diminished efficacy of immunotherapy.

Fig. 5.

   [87]Fig. 5

   [88]Fig. 5
   [89]Open in a new tab

   The assessment of immune cell infiltration and functional score across
   various subgroups within the IRGPRI. (A) Examining the variation in
   immune cell composition among different subgroups of IRGPRI. (B)
   Assessment of the immune cell function score among different subgroups
   within the IRGPRI. (C) The correlation between the score of IRGPRI and
   the genes associated with checkpoints.

3.7. Relationship between IRGPRI grouping and ICIs therapy

   To investigate the predictive value of IRGPRI in relation to the
   administration of ICIs, an assessment was carried out on a group of
   individuals who received immunotherapy (IMvigor210). Our results
   revealed that patients categorized as low-risk demonstrated enhanced
   response to immunotherapy and attained superior outcomes in comparison
   to those classified as high-risk ([90]Fig. 6A and B). As a result, an
   increased risk score was linked to a greater probability of immune
   evasion and diminished effectiveness of ICIs treatment. We conducted
   ROC analysis to assess the predictive capacity of IRGPRI ([91]Fig. 6C)
   and found that it had an area under the curve of 0.706. These findings
   indicate that IRGPRI has promising potential as a prognostic tool for
   predicting response to checkpoint inhibition therapy.

Fig. 6.

   [92]Fig. 6
   [93]Open in a new tab

   Prognostic value of IRGPRI in ICIs therapy. (A) The low-risk group
   exhibited a significantly elevated rate of immune response in ICIs
   therapy. (B) The low-risk group exhibited a more favorable outcome
   compared to the high-risk group. (C) ROC analysis of IRGPRI in
   IMvigor210 cohort.

4. Discussion

   NSCLC plays a significant role in the worldwide mortality caused by
   cancer [[94]19]. Despite the advent of ICIs therapy, the response rate
   remains suboptimal, limiting its efficacy to a minority of NSCLC
   patients [[95]20]. Hence, there exists an imperative need to identify a
   robust biomarker capable of discerning patients likely to derive
   benefits from ICIs therapy. Cancer immunotherapy is a therapeutic
   strategy that utilizes the cytotoxic potential of the human immune
   system, specifically tumor-specific cytotoxic T cells, to combat
   malignancies [[96]21]. Tumors utilize distinct strategies to interfere
   with T cell reactions, resulting in a decrease in the immunogenicity of
   the majority of tumors and establishing a suppressive TME that impedes
   the function of developing cytotoxic T cells targeting the tumor
   [[97]22]. Therefore, the current focus in cancer immunotherapy lies in
   enhancing T cell-mediated immune responses, aiming to convert cancer
   cells' resistance into a state that can be targeted by chemotherapy or
   radiotherapy [[98]23]. CTLs are considered as the optimal immune cells
   for effectively eradicating cancer cells [[99]23]. CD8 coreceptor is
   expressed by CTLs, which serve as a primary barrier against the
   advancement of cancer [[100]24,[101]25]. The significance of these
   cells in efficiently eradicating cancer cells is considered greater
   than that of CD4 T cells [[102]26].

   In this investigation, scRNA-seq data from 44 individuals with NSCLC
   were examined to identify 209 DEGs between CTLs and other cell types.
   This led to the identification of 109 genes related to the immune
   system. Analyses using GO and KEGG revealed an enrichment in processes
   associated with immunity and antigen processing/presentation.
   Regression analysis highlighted 11 immune-related genes that could
   potentially serve as prognostic factors. Among these genes, TRBC1,
   HLA-DMA, CTSH, RAC1, CTSL, ANXA2, and CEBPB were utilized to construct
   a novel biomarker called IRGPRI. By utilizing the median IRGPRI score
   as a threshold value, patients from both TCGA cohort and GEO Cohort
   were categorized into low-risk or high-risk groups for further
   analysis. It should be emphasized that our study has certain
   limitations as it includes NSCLC patients who were not specifically
   administered immunotherapy treatment in both cohorts (TCGA and GEO),
   potentially affecting the applicability of our findings concerning the
   outcomes of immunotherapy. Nevertheless, we found that IRGPRI was
   effective in predicting OS among NSCLC patients. The independent GEO
   cohort also validated the prognostic importance of IRGPRI, as better
   survival outcomes were observed in individuals with lower IRGPRI
   scores, while those with higher scores had a poorer prognosis. These
   findings indicate that IRGPRI shows potential as a biomarker associated
   with the immune system for prognostic prediction in NSCLC patients.
   Furthermore, our study revealed a robust association between IRGPRI and
   clinicopathological factors that have been shown to significantly
   impact OS in individuals with NSCLC.

   In order to enhance our comprehension of the immunological
   characteristics linked with each subgroup of IRGPRI, we conducted an
   investigation into the molecular attributes and genetic mutations. Our
   research revealed that individuals in the high-risk group had a greater
   occurrence of pathways linked to cancer, whereas those in the low-risk
   group showed an increase in gene sets related to immune response
   pathways, including IgA production and immunodeficiency. Although
   T-cell activation is widely acknowledged as a primary therapeutic
   approach in cancer immunotherapy, recent research has underscored the
   vital roles played by tumor-associated B cells and antibodies in
   regulating immune responses within the TME and in cancer immunotherapy
   [[103]27,[104]28]. As serum's second most abundant antibody, IgA has
   demonstrated significant potential for recruiting and activating
   neutrophils via FcαRI, thereby leading to eradication of tumors. This
   characteristic renders it a promising target for monoclonal
   antibody-based anti-tumor therapy [[105]29]. Previous studies have
   established a correlation between polyclonal and antigen-specific IgA
   antibodies and favorable prognoses in ovarian cancer, nasopharyngeal
   carcinoma, and endometrial cancer [[106]28,[107]30,[108]31].
   Henceforth, we hypothesize that IgA may play a pivotal role in
   immunotherapeutic interventions for NSCLC patients with low IRGPRI
   scores by potentially contributing to improved survival outcomes. The
   examination of genetic mutations indicated that the group at high risk
   showed a significantly increased mutation rate, with TP53 gene
   exhibiting the highest occurrence of mutations, primarily in the form
   of missense mutations. TP53 gene, which was the first tumor suppressor
   gene to be discovered, is responsible for around 50 % of NSCLC
   mutations and up to 80 % of squamous-cell carcinoma mutations
   [[109]32]. Changes in metabolism caused by TP53 mutation and hypoxia
   may lead to the acidification of the microenvironment, influencing its
   immune characteristics to support the advancement of cancer [[110]33].
   Previous studies have suggested that NSCLC patients with TP53 mutations
   experience more aggressive tumor growth and higher resistance to
   chemotherapy compared to those with TP53 wild-type [[111]34,[112]35].
   Additionally, patients harboring TP53 missense mutations have an
   unfavorable prognosis, particularly in cases of adenocarcinoma
   [[113][34], [114][35], [115][36], [116][37], [117][38], [118][39],
   [119][40], [120][41]]. Our survival analysis results are consistent
   with these studies, indicating that individuals in the high-risk group
   with increased TP53 mutations have a less favorable prognosis compared
   to those in the low-risk group who have fewer instances of TP53
   mutations.

   The correlation between the immune infiltrates in the TME and patient
   prognosis has been increasingly acknowledged in NSCLC and other types
   of cancer. The Immunoscore, initially developed as a tool to assess the
   risk of colon cancer, highlights the importance of evaluating immune
   infiltration in tumors to guide clinical decision-making [[121][42],
   [122][43], [123][44], [124][45], [125][46]]. The aim of this research
   was to examine the prevalence and functional scores of 22 distinct
   types of immune cells in NSCLC patients categorized into low-risk and
   high-risk groups. Additionally, our objective was to investigate the
   relative proportions of these immune cells within each group. The
   low-risk group demonstrated higher immune fraction and function scores
   for B cells, CTLs, checkpoint molecules, T cell inhibition, and T cell
   stimulation. Conversely, the high-risk group exhibited increased immune
   fraction and function scores for M0 and M2 macrophages. The presence of
   activated memory B cells, cytotoxic T cells, and CTLs is linked to a
   more favorable prognosis in cancer patients, indicating an enhanced
   immune response within the low-risk subgroup. These specific immune
   cells have a critical function in identifying and eliminating cancerous
   cells from the body. M0 macrophages are often regarded as unpolarized
   macrophages that remain inactive until they undergo polarization.
   Recent studies have provided evidence linking M0 macrophages to
   high-grade tumors and poor prognosis in glioma and hepatocellular
   carcinoma, suggesting their potential involvement in promoting
   tumorigenesis [[126]47,[127]48]. Multiple pathways have been identified
   through which M2 macrophages contribute to tumor progression. For
   example, researchers such as Yulei Chen et al. observed that CHI3L1
   secretion by M2 macrophages triggers activation of the MAPK signaling
   pathway, thereby facilitating metastasis in gastric and breast cancer
   cells [[128]49]. Exosomes released by lung tumor cells play a role in
   both immune suppression and tumor development within the TME by
   inducing M2 polarization and regulating mitochondrial metabolism
   [[129]50]. In the context of chronic inflammation, particularly in
   advanced stages of cancer, myeloid cells such as macrophages, dendritic
   cells (DCs), and granulocytes have been observed to exert suppressive
   effects on immune function. These specific cell types are commonly
   referred to as myeloid-derived suppressor cells (MDSCs) [[130]51].
   Numerous studies have presented evidence suggesting that a decrease in
   the quantity of MDSCs is linked to a favorable response to ICIs, thus
   serving as a prognostic indicator for improved patient survival.
   Consequently, it can be inferred that MDSCs significantly influence the
   development of initial resistance [[131]52,[132]53]. The
   immunosuppressive capabilities of MDSCs may be augmented by
   upregulating the expression of additional immune checkpoint molecules
   on activated T cells [[133]54]. Clinical investigations have revealed
   that administering anti-PD-1 antibodies triggers a signaling pathway
   involving PD-L1-NLRP3 inflammasome activation in cancerous cells,
   leading to an increased recruitment of polymorphonuclear MDSCs
   (PMN-MDSC) into the TME. This process serves as an acquired resistance
   mechanism [[134]55]. Our research results provide additional evidence
   to support these conclusions, suggesting that individuals in the
   low-risk subgroup demonstrate a greater likelihood of having tumor
   immunity, whereas those classified in the high-risk group exhibit
   traits linked to immune suppression.

   We conducted an analysis to investigated the association between IRGPRI
   and well-known genes that act as checkpoints in the immune system, such
   as PDCD1, CTLA4, TIGIT, LAG3, and BTLA. These checkpoint genes hinder
   the activation and proliferation of T cells through various signaling
   pathways, thereby suppressing the immune response. It has been noted
   that tumors expressing these checkpoint genes are more responsive to
   ICIs therapy compared to those lacking their expression. As of May
   2023, the FDA has approved five monoclonal antibodies (pembrolizumab
   [[135]56], nivolumab [[136]57], atezolizumab [[137]58], durvalumab
   [[138]59], and avelumab [[139]60]) specifically targeting PD-1 or PD-L1
   for treating NSCLC. However, there is currently no FDA-approved
   monoclonal antibody targeting CTLA-4 for NSCLC treatment despite its
   approval for melanoma and other cancer types. Limited research exists
   on investigating LAG3 and BTLA's potential use in NSCLC treatment; thus
   far, their therapeutic efficacy remains unestablished. In our
   investigation, we noticed an inverse relationship between IRGPRI scores
   and immune checkpoint gene expression. This suggests that individuals
   classified as low-risk may have a higher chance of benefiting from
   ICIs. To assess IRGPRI's predictive capacity in patient response to
   immunotherapy, we conducted an analysis on the IMvigor210 immunotherapy
   cohort. The findings suggested that the low-risk group exhibited a more
   positive reaction to immunotherapy, leading to better overall results
   in comparison to the high-risk group. Therefore, it is plausible to
   consider that involvement of IRGPRI may contribute to the determination
   of the effectiveness of cancer immunotherapy.

5. Conclusion

   The construction of the model involved utilizing data acquired from the
   TCGA database, which was subsequently verified using the GEO dataset.
   IRGPRI has potential as a prognostic biomarker for guiding personalized
   clinical decision-making and identifying patients with NSCLC who could
   potentially experience positive outcomes from immunotherapy. However,
   it is crucial to note that this analysis is limited due to the
   inability of the TCGA database and GEO dataset to exclude NSCLC
   patients who did not receive immunotherapy. Future research should
   focus on validating the predictive role of IRGPRI in prognosis and ICIs
   therapy for NSCLC patients through laboratory experiments.

Declaration of figures authenticity

   All figures provided have been exclusively generated by the authors,
   ensuring their originality and absence of any duplication or prior
   publication in whole or in part.

Funding

   This study received financial support from the Shenzhen Key Medical
   Discipline Construction Fund (No.SZXK075) and the Basic Medical and
   Health Research Project of Baoan District (Grant Number: 2020JD428).

Ethics approval and consent to participate

   Not applicable.

Consent for publication

   All study subjects signed written informed consent.

Data availability statement

   The present study utilized publicly available datasets, which can be
   accessed through the following sources: TCGA
   ([140]https://portal.gdc.cancer.gov), GEO
   ([141]https://www.ncbi.nlm.nih.gov/geo/), ImmPort
   ([142]https://www.immport.org/shared/home), InnateDB
   ([143]https://www.innateDBdb.com), and TISCH 2
   ([144]http://tisch.comp-genomics.org).

CRediT authorship contribution statement

   Shenjing Cui: Writing – review & editing, Writing – original draft.
   Yikun Yang: Writing – review & editing, Writing – original draft.
   Shuang Lou: Software, Data curation. Rong Huang: Software, Formal
   analysis. Jing Wang: Software, Formal analysis. Zhongbiao Chen:
   Software, Formal analysis. Jingjing Xie: Software, Formal analysis.

Declaration of competing interest

   The authors declare the following financial interests/personal
   relationships which may be considered as potential competing
   interests:Yikun Yang reports article publishing charges was provided by
   Shenzhen Key Medical Discipline Construction Fund. Shengjin Cui reports
   article publishing charges was provided by Basic Medical and Health
   Research Project of Baoan District. If there are other authors, they
   declare that they have no known competing financial interests or
   personal relationships that could have appeared to influence the work
   reported in this paper.

References