Skip to main content
NCBI home page
ACS Omega logoLink to ACS Omega
. 2026 Feb 3;11(6):10089–10103. doi: 10.1021/acsomega.5c11091

Harnessing Anoikis-Related Gene Signatures for Immunotherapy Precision: A Machine Learning-Driven Predictive and Prognostic Model

Kaile Wang †, Xuxiang Chen †, Shuang Zhou †, Jin Huang †, Mei Liu ‡, Haihang Zhang †, Jiangzheng Zeng †,*
PMCID: PMC12917847  PMID: 41726702

Abstract

Purpose: Cancer remains a leading cause of death, with immune checkpoint inhibitors (ICIs) offering promising but heterogeneous therapeutic outcomes. This study investigates the role of anoikis-related genes (ARGs) in immunotherapy efficacy and develops a machine learning-based predictive model for immunotherapy response. Methods: We integrated single-cell RNA sequencing (scRNA-seq) and multiomics data from various cancer types to identify ARGs associated with immunotherapy response. A total of 41 anoikis-related differential genes (Anoikis.Sig) were identified through correlation and enrichment analyses. To assess the predictive value of these ARGs, we constructed machine learning models using seven classifiers, including Naive Bayes (NB), Random Forest (RF), Support Vector Machine (SVM), and others. We also validated the model across multiple independent data sets. Additionally, we conducted CRISPR-based screening to identify immune-resistant ARGs that influence tumor immunity. Prognostic analysis was performed on a pan-cancer data set to evaluate the clinical relevance of the identified ARGs. Results: Our analyses established a significant association between Anoikis.Sig and immunotherapy response. Specifically, lower Anoikis Scores were strongly correlated with improved immunotherapy outcomes in melanoma and BCC cohorts. Among seven classifiers, the SVM-based model achieved superior predictive performance with an AUC of 0.782, consistently outperforming established gene signatures across multiple validation data sets. Furthermore, integrated CRISPR screening pinpointed key immune-resistant genes, such as BCL2L1, ITGAV, and PTK2, as potential drivers of tumor progression. The 11-gene Key-Anoikis.Sig not only demonstrated robust prognostic value across 30 cancer types but also showed a remarkably strong survival association in hepatocellular carcinoma (HCC). Conclusion: ARGs are essential modulators of the tumor-immune landscape and serve as high-performance biomarkers for predicting ICI efficacy. The machine learning-driven model developed in this study provides a reliable tool for precise patient stratification and personalized therapeutic guidance. Our findings underscore the potential of targeting ARGs as a novel strategy to enhance immunotherapy outcomes.

graphic file with name ao5c11091_0010.jpg

graphic file with name ao5c11091_0008.jpg

1. Introduction

Cancer continues to be a predominant cause of mortality on a global scale, with approximately 20 million new diagnoses and around 9.7 million deaths reported in 2022 alone. The high incidence and mortality rates associated with cancer pose significant challenges to public health systems globally. Immunotherapy, particularly immune checkpoint inhibitors (ICIs), has emerged as a promising treatment modality, offering durable responses in various cancer types. However, despite the advancements in immunotherapy, a substantial proportion of patients exhibit poor responses, highlighting the urgent need to explore novel strategies to enhance treatment efficacy.

Anoikis, a subtype of apoptosis, occurs as a consequence of cell detachment from the extracellular matrix (ECM), serving as a critical mechanism to prevent inappropriate cell survival in nonanchorage conditions. The evasion of Anoikis is a hallmark of cancer progression, allowing malignant cells to survive and disseminate, thereby contributing to tumor aggressiveness and poor prognosis in various malignancies. , Research has shown that the dysregulation of Anoikis-related genes (ARGs) can lead to enhanced metastatic potential, making them significant targets for therapeutic intervention. , These genes are essential for maintaining cellular homeostasis, and their dysregulation has been linked to various malignancies, including hepatocellular carcinoma and melanoma, where they contribute to tumor metastasis and resistance to therapy. − The importance of ARGs in the context of ICIs has garnered considerable attention, as their expression can modulate the immune landscape of tumors, potentially impacting the efficacy of immunotherapy. Traditional approaches to studying ARGs often rely on single-omics data, which may overlook the complex interactions within the tumor microenvironment. Recent advancements in multiomics analyses, including transcriptomics and proteomics, provide a more comprehensive understanding of ARGs and their functional implications in cancer. ,

This study focuses on evaluating how ARGs influence the effectiveness of immunotherapeutic interventions in cancer. To achieve this, single-cell RNA sequencing (scRNA-seq) data were analyzed to investigate intratumoral cellular diversity and to identify discrete cell subsets within the tumor microenvironment. Machine learning (ML) techniques were applied to construct ARG-based predictive frameworks, thereby advancing our knowledge of their biomarker potential and therapeutic relevance in the context of cancer immunotherapy. By elucidating the interplay between ARGs and immune response, this research seeks to provide novel insights that could improve patient stratification and treatment outcomes in cancer therapy, ultimately contributing to the advancement of personalized medicine in oncology.

2. Materials and Methods

2.1. Data Download

The GeneCards database offers an extensive compilation of data related to human genes. Using “Anoikis” as the search term, we filtered for protein-coding ARGs with relevance scores above 2, resulting in 101 genes. In addition, previously reported ARGs were retrieved from the PubMed database (https://pubmed.ncbi.nlm.nih.gov/) using the keyword “Anoikis”, based on published literature, , resulting in 8 additional ARGs. After merging the genes obtained from GeneCards and PubMed and removing duplicates, a final set of 105 unique ARGs was identified. Detailed information is provided in Table S1.

In order to examine the potential link between ARGs expression and tumor immunotherapy response, two data sets with defined treatment efficacy profiles were sourced from the GEO database, including the skin cutaneous melanoma (SKCM) data set GSE115978 and the basal cell carcinoma (BCC) data set GSE123813.

Thirty-five scRNA-Seq data sets focusing on stromal and immune cells in malignant cancers were sourced from the TISCH database (http://tisch.compbio.cn/). Anoikis-related differential genes (Anoikis.Sig, See Section in the Methods for details) were identified by intersecting ARGs with high expression in malignant cells and genes showing a positive correlation with the Anoikis score. Bulk transcriptomic data across 30 TCGA cancer types were collected using the R package TCGAbiolinks, excluding immune-dominated tumors (DLBC, LAML, THYM). All samples were normalized to TPM format using data from cBioPortal. ,

To evaluate the predictive performance of Anoikis.Sig, a comprehensive collection of ten RNA-Seq data sets related to immune checkpoint inhibitors (ICIs) was assembled. The data set compilation encompassed five studies on skin cutaneous melanoma (SKCM)-Hugo et al. (2016), Liu et al. (2019), Gide et al. (2019), Riaz et al. (2017), and Van Allen et al. (2015); two on urothelial carcinoma (UC)-Mariathasan et al. (2018) and Snyder et al. (2017); and single data sets representing glioblastoma (GBM): Zhao et al. (2019), renal cell carcinoma (RCC): Braun et al. (2020), and gastric cancer (GC): Kim et al. (2018). The SKCM data set reported by Hugo et al. (2016) comprised 27 pretreatment tumor samples from 26 patients, and the GBM data set from Zhao et al. (2019) included 34 pretreatment samples derived from 17 patients. Within these data sets, a single tumor specimen was randomly selected to represent each patient.

In parallel, seven pan-cancer signatures (IPRES.Sig, INFG.Sig, T.cell.infamed.Sig, PDL1.Sig, LRRC15.CAF.Sig, NLRP3.Sig, Cytotoxic.Sig) and four SKCM-specific markers (CRMA.Sig, ImmuCells.Sig, IMS.Sig, TRS.Sig) were obtained. The computational methods and codes employed for these 11 signatures were directly adapted from the source publications.

2.2. Relationship between ARGs and the Efficacy of Tumor Immunotherapy

The correlation between ARGs and tumor immunotherapy response was investigated employing the Seurat R package (version 4.3.0). The SKCM data set (GSE115978) and BCC data set (GSE123813) were analyzed, with 32 patients in the SKCM group (15 nonresponders, 16 treatment-naive, 1 responder) and 10 patients in the BCC group (4 nonresponders, 6 responders). The Anoikis Score for each cell was calculated using the “AddModuleScore” function, followed by Uniform Manifold Approximation and Projection (UMAP) analysis for visualization of immune, stromal, and malignant cells, and the Anoikis Score. Group comparisons were used to evaluate differences in Anoikis scores among cell subtypes.

2.3. Screening of Anoikis.Sig

We hypothesized that ARGs levels could predict immunotherapy response. Thirty-five scRNA-Seq data sets were analyzed using Spearman correlation to identify ARGs associated with the Anoikis Score in malignant cells. ARGs positively correlated (Spearman r > 0.3, p < 0.05) were classified as Gx, while those upregulated specifically in malignant tumor cells were termed Gy. The gene subset Gn, obtained by intersecting Gx and Gy in each data set, constituted the Anoikis.Sig.

2.4. Gene Ontology and Pathway Enrichment Analysis

Functional enrichment analyses, including Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG), were conducted on Anoikis.Sig using the clusterProfiler R package. Terms with an adjusted p-value below 0.05 and a false discovery rate (FDR) q-value less than 0.25 were considered significant. The Benjamini-Hochberg procedure was applied to correct for multiple testing.

2.5. Immune Correlation Analysis and Gene Set Variation Analysis

The Anoikis.Sig Score was computed across 30 tumor types within the TCGA pan-cancer cohort utilizing gene set variation analysis (GSVA). Associations between the Anoikis.Sig Score and 75 previously reported immune-related genes were assessed and visualized using heatmap representation. Immune cell fractions were inferred through the CIBERSORT algorithm, and their correlations with the Anoikis.Sig Score were determined using Spearman’s correlation. GSVA was performed on HALLMARK gene sets, and the correlation between pathway enrichment scores and Anoikis.Sig scores was visualized.

2.6. Construction of Immune Efficacy Prediction Model

To construct a comprehensive data set, five ICI RNA-Seq data sets with the largest patient sample sizes were integrated, resulting in a combined cohort of 772 patients. The compiled data set included 181 samples from RCC, 348 from UC, and 243 from SKCM, derived from five studies: Braun et al. (2020) for RCC; Mariathasan et al. (2018) for UC; and Liu et al. (2019), Gide et al. (2019), and Riaz et al. (2017) for SKCM. Following the aggregation of the five major ICI cohorts, raw expression values were normalized to TPM format. To eliminate potential platform-specific and study-specific technical biases, the ComBat algorithm (within the sva R package) was applied. After correcting for batch effects, the integrated data set was randomly split into a training set (80%, n = 618) and a validation set (20%, n = 154). Additionally, five independent immune checkpoint inhibitor (ICI) RNA-Seq data sets comprising 149 patients were used as external validation cohorts: Zhao et al. (2019) for GBM, Snyder et al. (2017) for UC, Hugo et al. (2016) and Van Allen et al. (2015) for SKCM, and Kim et al. (2018) for GC.

Utilizing the Anoikis.Sig signature and the designated training data set, we developed an immune response prediction model by implementing seven widely recognized ML techniques. These included Naive Bayes (NB), Random Forest (RF), Support Vector Machine (SVM), AdaBoost Decision Trees, LogitBoost (an enhanced logistic regression), K-Nearest Neighbors (KNN), and the Cancerclass algorithm. For all methods necessitating hyperparameter optimization, excluding Cancerclass, 5-fold cross-validation was conducted to enhance the robustness and predictive accuracy of the models. To enhance the stability of the results, hyperparameter optimization was iterated 20 times, each employing a distinct random seed during resampling procedures. Since the Cancerclass method lacks adjustable parameters, its corresponding model was fitted directly using the entire training cohort.

The above ML algorithms were utilized to construct predictive models using the training data set, followed by performance evaluation on the validation set. To identify the optimal algorithm and establish the final Anoikis-associated immune response prediction model, Receiver Operating Characteristic (ROC) curves were plotted and the corresponding Area Under the Curve (AUC) values were computed. AUC values typically ranged from 0.5 to 1.0, with values approaching 1.0 indicating superior discriminative ability.

To comprehensively assess the predictive capability of Anoikis.Sig, its AUC scores were benchmarked against those of six established ICI response-related gene signatures across the training set, internal validation cohort, and five external validation data sets. These previously reported ICI-associated signatures include INFG.Sig, T.cell.inflamed.Sig, PDL1.Sig, LRRC15.CAF.Sig, NLRP3.Sig, and Cytotoxic.Sig.

2.7. CRISPR Profiling

To identify candidate therapeutic targets associated with Anoikis.Sig, we systematically integrated data from seven published CRISPR/Cas9 screening investigations, each of which assessed the functional impact of individual gene knockouts on tumor immune responses. These studies were conducted by Freeman et al. (2019), Kearney et al. (2018), Manguso et al. (2017), Burgess et al. (2018), Patel et al. (2017), Vredevoogd et al. (2019), and Lawson et al. (2020). The compiled CRISPR data sets were partitioned into 17 subsets based on tumor cell line models and experimental intervention settings. Cell lines derived from SKCM, GBM, CRC, and RCC were included in the analysis. These data were used to identify genes that were more likely to regulate lymphocytes and influence response to immunotherapy in different data sets.

Genome-wide CRISPR/Cas9 knockdowns were performed in different cancer cell lines for CRISPR screening. Among them, in vitro, different cancer cell lines were cultured in the presence or absence of Cytotoxic Lymphocytes (CTLs). In the in vivo environment, cell lines of different cancers were implanted into immunodeficient or immunocompetent mice. Subsequently, RNA sequencing was employed to quantify the expression abundance of single guide RNAs (sgRNAs) corresponding to the targeted genes. To assess the effect of gene knockout on tumor cells subjected to CTLs pressure or antitumor immune responses, the logFC value changes between different cell lines groups were calculated. To minimize batch effects and enable cross-study comparisons of CRISPR data sets, normalized Z-scores were derived from logFC values. A lower Z-score was indicative of enhanced immune responsiveness following gene knockout. Genes were ranked according to their mean Z-scores derived from analysis across 17 independent data sets, with those exhibiting the highest average scores identified as candidate immune resistance genes.

Concurrently, to further assess the predictive performance of Anoikis.Sig, we performed a comparative analysis against previously reported immune checkpoint inhibitor (ICI) response-related gene signatures. These included a pan-cancer fibroblast-associated signature, LRRC15.CAF.Sig, as well as four melanoma (SKCM)-specific gene sets: CRMA.Sig and ImmuCells.Sig, IMS.Sig, and TRS.Sig.

2.8. Prognostic Model Constructed by Machine Learning Algorithm

To investigate the association between Anoikis.Sig expression and clinical outcomes in cancer patients, Anoikis.Sig related prognostic risk model was developed using five ML algorithms based on transcriptomic data from three ICI-treated cohorts: Braun et al. (2020) RCC, Snyder et al. (2017) UC, and Hugo et al. (2016) SKCM. The algorithms applied included Bagged Trees, NB, Learning Vector Quantization (LVQ), the Boruta-based wrapper method, and RF. To enhance robustness, feature selection results from all algorithms were intersected, and genes identified by at least three methods were designated as Key-Anoikis.Sig genes.

A prognostic model was developed and validated utilizing data sets from Braun et al. (2020) RCC, Mariathasan et al. (2018) UC, Liu et al. (2019) SKCM, Gide et al. (2019) SKCM, and Riaz et al. (2017) SKCM. Univariate and multivariate Cox proportional hazards regression analyses were conducted on the Key-Anoikis.Sig genes employing the R package survival (version 3.5.7) to assess their prognostic significance. Subsequently, a Cox risk score was derived from the expression profiles of the Key-Anoikis.Sig genes. The risk score was calculated according to the following formula:

Risk Score=∑icoefficient(genei)×expression(genei)

To compare overall survival (OS) between the high- and low-risk groups, Kaplan–Meier (KM) survival analysis was performed using the R package survival.

2.9. Panoramic Analysis of Key-Anoikis.Sig Genes

The Key-Anoikis.Sig Score was calculated for 30 cancer types in the TCGA pan-cancer transcriptome data set. The correlation between Key-Anoikis.Sig Score, immune cells (CIBERSORT), and microsatellite instability (MSI) was analyzed using Spearman’s method. Prognostic analysis was conducted using Cox regression, and the results were illustrated with a forest plot to evaluate the impact of Key-Anoikis.Sig on patient prognosis.

2.10. Statistical Analysis

All data processing and statistical analyses were performed using R software (version 4.2.2). For continuous variables, comparisons between two groups were conducted using the independent Student’s t test when data conformed to a normal distribution, unless specified otherwise. In cases where the data were not normally distributed, the Mann–Whitney U test was applied. For comparisons involving three or more groups, the Kruskal–Wallis test was utilized. Spearman’s correlation analysis was employed to evaluate associations among different molecular features. Unless noted otherwise, all statistical tests were two-sided, and a p-value less than 0.05 was regarded as statistically significant.

3. Results

3.1. Association between ARGs and Tumor Immunotherapy Efficacy

The association between Anoikis Scores and immunotherapy efficacy was evaluated in both SKCM (GSE115978) and BCC (GSE123813) data sets. In the SKCM data set, UMAP visualizations revealed distinct distributions of immune, stromal, and malignant cells (Figure S1A,B). Statistical comparisons showed significantly lower Anoikis Scores in the responder (R) group compared to nonresponder (NR) and treatment-naive (TN) groups (p < 0.001), with the NR group also showing lower scores than the TN group (p < 0.001) (Figure S1C). Additionally, Anoikis Scores were more enriched in malignant cells than other subtypes (Figure S1D). In the BCC data set, the R group exhibited significantly lower Anoikis Scores compared to the NR group (p < 0.001), and malignant cells showed higher enrichment of Anoikis Scores (Figure S2A-D). The results indicate a possible association between ARGs and the efficacy of immunotherapeutic treatments.

3.2. Screening, Enrichment, and Immune Correlation Analysis of Anoikis.Sig

We hypothesized that evaluating the expression levels of ARGs could serve as a predictive tool for immunotherapy efficacy. A Spearman correlation analysis was conducted on 34 scRNA-Seq data sets, revealing a set of genes that exhibited significant positive correlations with Anoikis Scores in malignant cell populations (r > 0.3, p < 0.05). These genes were subsequently designated as Gx. Additionally, highly expressed ARGs in malignant cells were labeled as Gy. The intersection of Gx and Gy resulted in a tumor-specific set of genes (Gn), positively correlated with Anoikis Scores (Figure A). A total of 41 Anoikis-related differential genes (Anoikis.Sig) were identified (Table S2). Enrichment analysis of these Anoikis.Sig genes revealed that they were primarily involved in processes like cell-matrix adhesion, negative regulation of Anoikis, and integrin-mediated signaling (Figure B). These genes were also enriched in cellular components such as cell–substrate junctions and focal adhesions, and exhibited molecular functions such as integrin and extracellular matrix binding. KEGG pathway enrichment analysis further underscored the participation of Anoikis.Sig genes in several key signaling cascades, including focal adhesion, cancer-associated proteoglycans, and the PI3K-Akt pathway (Table S3).

1.

1

Open in a new tab

Description and GO and KEGG enrichment analysis for Anoikis.Sig. (A) Intersection circle diagram of ARGs and upregulated differentially expressed genes in malignant tumors from 35 scRNA-Seq data sets. Different colors indicate different cancer types. (B) Bubble plot of GO and KEGG pathway enrichment analysis results of Anoikis.Sig.

To explore the immune-related roles of Anoikis.Sig genes, we conducted GSVA on the TCGA pan-cancer data set, calculating Anoikis.Sig Scores for 30 cancer types. Correlation analysis with 75 immune-related genes showed strong positive associations with most immune genes (Figure A). Immune infiltration of 22 immune cell types was estimated using the CIBERSORT algorithm, revealing significant positive correlations between Anoikis.Sig Scores and immune cell types (Figure B). Moreover, GSVA analysis of HALLMARK pathways showed robust correlations between Anoikis.Sig Scores and immune-related pathways (Figure C), underscoring the potential role of Anoikis.Sig genes in modulating immune responses in cancer.

2.

2

Open in a new tab

Immune analysis of Anoikis.Sig score. (A) Pan-cancer association heatmap of Anoikis.Sig Score and immune-related genes. In the polar coordinate plot, cancer types are denoted along the radial axis, while immune-related genes are positioned according to their angular coordinates. The innermost rings reflect the signaling pathways linked to these genes, whereas the outer rings illustrate their associated biological functions. (B) Pan-cancer correlation heatmap of Anoikis.Sig Score and CIBERSORT immune cell infiltration abundance. (C) Pan-cancer correlation heatmap of Anoikis.Sig Score and HALLMARK-associated pathways.

3.3. Construction of the Immune Efficacy Prediction Model

To investigate the predictive value of Anoikis.Sig for immunotherapy outcomes, we compiled 10 ICI RNA-Seq data sets with known immunotherapy responses. Using seven ML algorithms, we developed predictive models, and the SVM-based model exhibited the highest AUC (0.782) on the validation set (Figure A,B). When compared with previously published ICI response-related gene signatures, Anoikis.Sig demonstrated superior predictive performance across multiple ICI RNA-Seq data sets, particularly in the Hugo 2016 SKCM data set (Figure C). A heatmap further illustrated the superior predictive power of Anoikis.Sig (Figure D).

3.

3

Open in a new tab

Prediction of ICI outcomes and AUC of Anoikis.Sig. (A) Performance comparison of prediction models constructed using seven ML algorithms for immune response prediction. The x-axis represents AUC values, while the y-axis shows various machine learning algorithms. (B) ROC curve of the prediction model created using the SVM algorithm. (C) Circular plot displaying the predictive performance of different ICI RNA-Seq data sets for immune response. The polar axis denotes AUC values, with angles representing genes associated with various ICI response signatures across data sets. (D) Heatmap illustrating the predictive power of ICI response signature-associated genes for immune response across data sets. The x-axis represents the ICI RNA-Seq data sets, the y-axis represents ICI response signature-associated genes.

3.4. CRISPR Analysis of Immune-Resistant Genes

CRISPR screening data were aggregated from seven independent cohorts, comprising a total of 17 data sets and encompassing 22,505 genes. Based on the average Z-scores computed across multiple data sets, genes were hierarchically ranked. Genes exhibiting higher Z-scores were categorized as immune-resistant, while those with lower values were identified as immune-sensitive. It was hypothesized that silencing immune-resistant genes might enhance antitumor immune responses, whereas disruption of immune-sensitive genes could attenuate immunological activity (Figure A). The top 20% of immune-resistant genes included several Anoikis.Sig genes such as BCL2L1, ITGAV, PTK2, YAP1, ILK, CD44, ATF4, ITGA6, ITGA2, NQO1, CDH1, ITGB4, and CSPG4. The percentage of Anoikis.Sig is higher compared to other ICI response-associated genes (Figure B). These genes were validated as immune-resistant across 17 CRISPR data sets (Figure C), suggesting their potential as biomarkers for immunotherapy efficacy.

4.

4

Open in a new tab

CRISPR analysis of Anoikis.Sig. (A) Gene ranking based on Z-scores from 17 CRISPR data sets. Red denotes immune resistance genes, whose knockout enhances antitumor immune responses, whereas blue indicates immune sensitivity genes, whose deletion suppresses antitumor immunity. (B) A radar chart illustrates the proportion of top-ranked genes shared between the Anoikis.Sig signature and immune checkpoint inhibitor (ICI) response-related gene sets. (C) A heatmap displays the Z-score distribution of 13 Anoikis.Sig genes identified as immune resistance-related candidates.

3.5. Screening of Key Anoikis-Related Genes for Immunotherapy Prediction

We refined the Anoikis.Sig gene set using five ML algorithms (Boruta, NB, Bagged Trees, RF, and LVQ) to analyze their association with immunotherapy efficacy (Figure A-E). The intersection of these algorithms identified 11 Key-Anoikis.Sig genes: ANXA5, BCL2L1, CALR, CTNNB1, FN1, HMCN1, ITGB4, PTGS2, PTK2, SPP1, and TIMP1 (Figure F). Multivariate Cox regression analysis revealed significant associations between these genes and OS in both training and validation sets (Figure A,B). Utilizing the gene-derived risk score, patients were divided into high- and low-risk subgroups, exhibiting statistically significant differences in overall survival (training cohort: p < 0.001; validation cohort: p < 0.05) (Figure C,D).

5.

5

Open in a new tab

Machine learning and Key-Anoikis.Sig. (A-E) Diagnostic models for anoikis-related differential genes (Anoikis.Sig) using five machine learning (ML) algorithms: Wrapper (Boruta) (A), Naive Bayes (NB) (B), Bagged Decision Trees (C), Random Forest (RF) (D), and Learning Vector Quantization (LVQ) (E). (F) UpSet plot showing the intersection of results from the five ML algorithms.

6.

6

Open in a new tab

Prognostic model of Key-Anoikis.Sig. (A, B) Risk factor visualization from the Cox proportional hazards model based on Key-Anoikis.Sig in the training cohort (A) and validation cohort (B). (C, D) Kaplan–Meier curves illustrating overall survival differences between high- and low-risk groups in the training set (C) and validation set (D).

3.6. Panoramic Analysis of Key-Anoikis.Sig Genes across Cancer Types

Key-Anoikis.Sig scores were computed across 30 tumor types within the TCGA pan-cancer cohort (Figure A). Immune cell infiltration levels were subsequently evaluated using the CIBERSORT algorithm, revealing broadly positive correlations between Key-Anoikis.Sig scores and the abundance of various immune cell populations across cancer types (Figure B). We also explored the correlation between Key-Anoikis.Sig and microsatellite instability (MSI), which was significant across several cancer types (Figure C,D). The Key-Anoikis.Sig score demonstrated high prognostic value in various cancer types (p < 0.01) (Figure E), and the HCC cohort (TCGA-LIHC) showed significant survival differences between high- and low-risk subgroups based on the Key-Anoikis.Sig risk score (p < 0.001) (Figure F).

7.

7

Open in a new tab

Landscape analysis of Key-Anoikis.Sig. (A) Key-Anoikis.Sig Score of 11 Anoikis.Sig genes in 30 different cancer types from TCGA pan-cancer transcriptome data set. (B). Pan-cancer correlation heatmap of Key-Anoikis.Sig Score and CIBERSORT immune cell infiltration abundance. (C) Heat map of pan-cancer association between Key-Anoikis.Sig Score and MSI. (D) Heat map of pan-cancer association between Key-Anoikis.Sig and MSI. (E) Forest plot of pan-cancer univariate Cox regression of Key-Anoikis.Sig Score. (F) Prognostic KM curve between risk score and OS in HCC.

4. Discussion

The increasing prevalence of tumors and the limitations of current treatment strategies highlight the urgent need for innovative approaches in cancer therapy. ICIs have shown promise, yet the variability in patient responses underscores the necessity of identifying reliable biomarkers to predict therapeutic efficacy. , This study addresses this gap by focusing on ARGs and their potential role in enhancing the effectiveness of immunotherapy. Our findings demonstrate that ARGs are significantly associated with immunotherapy responses, suggesting their utility as predictive biomarkers. Through comprehensive analyses, we established a link between Anoikis Scores and treatment outcomes in melanoma and basal cell carcinoma data sets. Notably, we identified a set of 41 Anoikis-related differential genes, highlighting their involvement in crucial cellular processes and immune modulation. Furthermore, the construction of a predictive model based on Anoikis-related gene expression revealed its superior performance compared to existing response-associated genes, emphasizing its potential in clinical applications. The identification of Key-Anoikis.Sig genes, which exhibited significant associations with overall survival, further underlines the relevance of our approach. Overall, this research elucidates the critical role of ARGs in tumor immunotherapy, paving the way for enhanced patient stratification and tailored therapeutic strategies.

ARGs play a pivotal role in multiple signaling pathways that are central to tumor progression and immune modulation. Importantly, anoikis resistance represents a biologically distinct adaptive program that extends beyond general apoptosis evasion and is particularly relevant to tumor immune escape. Unlike canonical apoptosis resistance, which primarily enables tumor cells to survive intrinsic or extrinsic death signals, anoikis resistance confers a context-dependent survival advantage under anchorage-independent conditions, a hallmark of metastatic dissemination and tumor microenvironment remodeling. Consistent with this concept, enrichment analysis of Anoikis.Sig genes derived from 35 stromal/immune-cell scRNA-seq data sets spanning 17 cancer types revealed significant involvement in focal adhesion, proteoglycans in cancer, and PI3K-Akt signaling pathways. These pathways are tightly linked to integrin-mediated cell-extracellular matrix (ECM) interactions and mechanotransduction, which not only promote resistance to detachment-induced cell death but also actively contribute to immune exclusion and immunosuppressive niche formation. Notably, key Anoikis.Sig genes identified in this study, including ITGAV, ITGB4, PTK2, FN1, and SPP1, are central regulators of these pathways and have been implicated in T-cell exclusion, macrophage polarization, and stromal activation. In line with this mechanistic framework, our integrated CRISPR screening analysis demonstrated a significant enrichment of ARGs among immune-resistant candidates, providing functional evidence that anoikis resistance is closely associated with impaired antitumor immune responses. Moreover, anoikis-resistant tumor cells frequently exhibit enhanced epithelial-mesenchymal transition (EMT), hypoxia adaptation, and ECM remodeling, all of which are increasingly recognized as key drivers of immune evasion and resistance to immune checkpoint blockade. − Together, these findings support the notion that anoikis resistance constitutes a multifaceted biological program that coordinates metastatic potential with immune escape, thereby offering a mechanistically grounded and clinically relevant explanation for the predictive value of the Anoikis-based signature in immunotherapy response.

The analysis of ARGs in this study reveals significant immune features across various cancer types. Specifically, the Anoikis.Sig Score positively correlates with most immune checkpoint genes, suggesting their potential role in modulating immune responses during tumor progression. Notably, the abundance of immunosuppressive cells, such as M2 macrophages and neutrophils, shows a significant positive correlation with the Anoikis signature, while tumor-suppressive immune cells, such as CD8+ T cells and activated NK cells, exhibit negative correlations. These findings imply that elevated levels of these immune cells may be linked to poorer outcomes in certain cancers. M2 macrophages, which typically display pro-tumor characteristics within the tumor microenvironment, promote tumor growth, angiogenesis, and immune suppression by secreting cytokines and growth factors. Additionally, M2 macrophages can influence the tumor immunoscape by interacting with other immune cells, including cytotoxic T cells, regulatory T cells, and neutrophils. Studies have shown that inhibiting M2 macrophage function or reducing their numbers can enhance the efficacy of immunotherapies. For example, using CSF1R inhibitors to decrease M2 macrophage numbers has been shown to improve the effectiveness of anti-PD-1 treatments. Similarly, the Anoikis.Sig Score is strongly associated with key signaling pathways critical to tumor progression, including epithelial-mesenchymal transition (EMT), hypoxia, and TGF-β signaling, all of which are involved in immune evasion by tumor cells. , In comparison to other established biomarkers including INFG.Sig, T.cell.inflamed.Sig, PDL1.Sig, LRRC15.CAF.Sig, NLRP3.Sig, and Cytotoxic.Sig, the Anoikis-related signature exhibited enhanced predictive performance in forecasting immunotherapeutic efficacy across multiple tumor types. This study underscores the Anoikis signature as a promising biomarker for predicting immunotherapy efficacy, highlighting its potential to guide therapeutic strategies in cancer treatment. The results suggest that targeting ARGs could enhance immunotherapy effectiveness by reshaping the immune landscape within tumors, ultimately improving patient outcomes.

Our comprehensive analysis of seven CRISPR studies identified key immune-resistant and immune-sensitive genes based on average Z-scores. Notably, a significant proportion of immune-resistant genes overlapped with anoikis-related genes (Anoikis.Sig), including BCL2L1, ITGAV, PTK2, and so on. This enrichment suggests a crucial role for anoikis regulation in shaping antitumor immune responses. For example, BCL2-like protein 1 (BCL2L1) is a member of the antiapoptotic subfamily within the BCL2 protein family. BCL2 family proteins have been shown to regulate apoptosis and its modulation could lead to decreased immune cell infiltration in tumors, thereby lowering the efficacy of immune checkpoint inhibitors. , Furthermore, the analysis revealed that the knockdown of ITGAV could disrupt tumor-stroma interactions, potentially leading to increased susceptibility of cancer cells to immune-mediated destruction. Recent studies emphasize that anoikis-resistant cells utilize integrin-mediated signaling (e.g., ITGAV, PTK2) to not only survive detachment but also to orchestrate a pro-tumorigenic stroma. This process often involves the activation of the PI3K-Akt and TGF-β signaling axes, which are potent drivers of EMT and immune exclusion. Consequently, the Anoikis.Sig identifies tumors that have established a cold microenvironment refractory to cytotoxic T-cell infiltration. These findings suggest that targeting these ARGs may not only improve therapeutic outcomes but also provide a basis for developing combination strategies in immunotherapy. Future research should focus on elucidating the mechanisms by which these genes influence immune responses and exploring their interactions with existing immunotherapeutic agents. This could pave the way for more personalized treatment approaches that leverage the unique profiles of ARGs in different cancer types, ultimately enhancing patient outcomes in cancer therapy.

Our study refined the Anoikis.Sig gene set using five ML algorithms, identifying 11 key genes (ANXA5, BCL2L1, CALR, CTNNB1, FN1, HMCN1, ITGB4, PTGS2, PTK2, SPP1, TIMP1) associated with immunotherapy efficacy. Pan-cancer analysis revealed strong correlations between Key-Anoikis.Sig scores, immune infiltration, and MSI, with high prognostic value across multiple cancer types, particularly in HCC. These findings align with recent studies highlighting the role of anoikis-related genes in tumor immunity and therapy response, supporting their potential as prognostic biomarkers and therapeutic targets. ,, Pan-cancer analysis revealed that while the Key-Anoikis.Sig score is positively associated with immune resistance features, it exhibits a significant negative correlation with MSI. Although both indices serve as predictors for immunotherapy outcomes, this inverse relationship is not paradoxical; rather, it reflects the biological alignment of distinct tumor microenvironment dimensions. Specifically, MSI-High status characterizes tumors with high antigenicity and sensitivity, whereas an elevated Key-Anoikis.Sig score identifies a specialized adaptive program of immune evasion. This suggests that anoikis-mediated resistance is particularly prominent in MSS tumors, which often lack high mutation burdens. Consequently, the Key-Anoikis.Sig score captures immunosuppressive mechanisms, such as M2 macrophage polarization and TGF-β signaling, , that are independent of MSI-driven antigenicity. By identifying these complementary resistance signals, our model provides a more nuanced stratification tool that extends beyond traditional genomic markers, offering predictive value even in tumors typically refractory to immune checkpoint blockade.

The refinement of the signature to 11 key genes (e.g., ANXA5, BCL2L1, CALR) significantly enhances its translational utility. Unlike genome-wide sequencing, an 11-gene panel can be readily implemented using standard clinical assays like qPCR or NanoString, facilitating rapid turnaround times for therapeutic decision-making. Despite the promising predictive performance of our model, several limitations must be acknowledged to provide a balanced perspective. First, this study is primarily based on the retrospective analysis of public data sets from TCGA, GEO, and TISCH. While we employed rigorous normalization and batch-effect correction methods, the inherent heterogeneity in sample collection, library preparation, and sequencing platforms across different cohorts may still introduce potential biases. Second, although our SVM-based machine learning model demonstrated high accuracy (AUC = 0.782), its “black-box” nature poses challenges for direct biological interpretability. Further computational and experimental efforts are needed to elucidate how these specific gene weightings translate into concrete cellular interactions within the tumor microenvironment. Third, while our CRISPR screening analysis provided functional insights into immune resistance, these findings are derived from cell line models and integrated data sets rather than direct “wet-lab” validation within our own laboratory. The absence of in vitro or in vivo functional assays limits our ability to definitively confirm the molecular signaling pathways of key ARGs like BCL2L1 and ITGAV. Lastly, the clinical utility of the Key-Anoikis.Sig requires validation in prospective, large-scale clinical trials. Future research involving multicenter, prospective cohorts will be essential to establish the reliability and generalizability of this signature as a routine tool for patient stratification in personalized oncology.

In summary, our results indicate a significant association between ARGs and immunotherapy response, highlighting the potential of Anoikis Scores as predictive biomarkers. The development of a ML model demonstrated promising predictive power, particularly with the support of Key-Anoikis.Sig genes, which were shown to correlate with overall survival across various cancer types. These findings suggest that ARGs may play a crucial role in modulating immune responses and could serve as valuable targets for enhancing immunotherapy outcomes. Future studies should focus on validating these results in clinical settings and exploring the therapeutic implications of targeting Anoikis pathways in cancer treatment.

Supplementary Material

ao5c11091_si_001.pdf (548.4KB, pdf)
ao5c11091_si_002.pdf (234.6KB, pdf)

Acknowledgments

We thank Helix (https://www.helixlife.cn/) for help in English grammar and fluency.

All data sets utilized in this study are publicly accessible, as detailed in the Methods section. The corresponding web addresses or unique identifiers for each public cohort or data set are provided within the manuscript. Thirty-five stromal/immune-cell scRNA-Seq data sets of malignant cancers were collected from the TISCH database (http://tisch.compbio.cn/). GSE115978 and GSE123813 were downloaded from the GEO database (https://www.ncbi.nlm.nih.gov/geo/).

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.5c11091.

  • Figure S1: Identification and validation between Anoikis score and ICI outcomes in GSE115978. Figure S2: Identification and validation between Anoikis score and ICI outcomes in GSE123813 (PDF)

  • Table S1: List of 105 Anoikis-related genes. Table S2: Identification of 41 Anoikis-related differential genes. Table S3: Results of GO and KEGG analysis for Anoikis-related differential genes (PDF)

#.

K.W., X.C., and S.Z. contributed equally to this work. J.Z. conceived and designed the study. K.W., X.C., and S.Z. provided equal contributions to research design and article writing. J.H. and M.L. collected the data. H.Z. performed the analyses. J.Z. supervised the study and revised the paper. K.W., X.C., and S.Z. contributed equally to this work. All authors contributed to the article and approved the submitted version.

This work was supported by the National Natural Science Foundation of China (82160556), the postgraduate innovation research project of Hainan Province (Qhys2024–465), Hainan Province’s Key Research and Development Project (ZDYF2021SHFZ251), Hainan Provincial Natural Science Foundation of China (820RC761), and Hainan Province Clinical Medical Center, No. (2021)­75 and No. (2021)­276.

The authors declare no competing financial interest.

References

  1. Bray F., Laversanne M., Sung H., Ferlay J., Siegel R. L., Soerjomataram I., Jemal A.. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J. Clin. 2024;74(3):229–263. doi: 10.3322/caac.21834. [DOI] [PubMed] [Google Scholar]
  2. Wu Z., Xia F., Lin R.. Global burden of cancer and associated risk factors in 204 countries and territories, 1980–2021: a systematic analysis for the GBD 2021. J. Hematol Oncol. 2024;17(1):119. doi: 10.1186/s13045-024-01640-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Lynch C., Pitroda S. P., Weichselbaum R. R.. Radiotherapy, immunity, and immune checkpoint inhibitors. Lancet Oncol. 2024;25(8):e352–e362. doi: 10.1016/s1470-2045(24)00075-5. [DOI] [PubMed] [Google Scholar]
  4. Holder A. M., Dedeilia A., Sierra-Davidson K., Cohen S., Liu D., Parikh A., Boland G. M.. Defining clinically useful biomarkers of immune checkpoint inhibitors in solid tumours. Nat. Rev. Cancer. 2024;24(7):498–512. doi: 10.1038/s41568-024-00705-7. [DOI] [PubMed] [Google Scholar]
  5. Wang Y., Cheng S., Fleishman J. S., Chen J., Tang H., Chen Z. S., Chen W., Ding M.. Targeting anoikis resistance as a strategy for cancer therapy. Drug Resist Updat. 2024;75:101099. doi: 10.1016/j.drup.2024.101099. [DOI] [PubMed] [Google Scholar]
  6. Yang B., Lou C., Chen S., Zhang Z., Xu Q.. XIAP and PHB1 Regulate Anoikis through Competitive Binding to TRAF6. Mol. Cancer Res. 2023;21(2):127–139. doi: 10.1158/1541-7786.MCR-22-0415. [DOI] [PubMed] [Google Scholar]
  7. Khan S. U., Fatima K., Malik F.. Understanding the cell survival mechanism of anoikis-resistant cancer cells during different steps of metastasis. Clin. Exp. Metastasis. 2022;39(5):715–726. doi: 10.1007/s10585-022-10172-9. [DOI] [PubMed] [Google Scholar]
  8. Chen Y., Huang W., Ouyang J., Wang J., Xie Z.. Identification of Anoikis-Related Subgroups and Prognosis Model in Liver Hepatocellular Carcinoma. Int. J. Mol. Sci. 2023;24(3):2862. doi: 10.3390/ijms24032862. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chen Y., Lin Q. X., Xu Y. T., Qian F. J., Lin C. J., Zhao W. Y., Huang J. R., Tian L., Gu D. N.. An anoikis-related gene signature predicts prognosis and reveals immune infiltration in hepatocellular carcinoma. Front. Oncol. 2023;13:1158605. doi: 10.3389/fonc.2023.1158605. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Neuendorf H. M., Simmons J. L., Boyle G. M.. Therapeutic targeting of anoikis resistance in cutaneous melanoma metastasis. Front. Cell Dev Biol. 2023;11:1183328. doi: 10.3389/fcell.2023.1183328. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Xie T., Peng S., Liu S., Zheng M., Diao W., Ding M., Fu Y., Guo H., Zhao W., Zhuang J.. Multi-cohort validation of Ascore: an anoikis-based prognostic signature for predicting disease progression and immunotherapy response in bladder cancer. Mol. Cancer. 2024;23(1):30. doi: 10.1186/s12943-024-01945-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Zhang Z., Chen X., Li Y., Zhang F., Quan Z., Wang Z., Yang Y., Si W., Xiong Y., Ju J.. et al. The resistance to anoikis, mediated by Spp1, and the evasion of immune surveillance facilitate the invasion and metastasis of hepatocellular carcinoma. Apoptosis. 2024;29(9–10):1564–1583. doi: 10.1007/s10495-024-01994-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Stelzer G., Rosen N., Plaschkes I., Zimmerman S., Twik M., Fishilevich S., Stein T. I., Nudel R., Lieder I., Mazor Y.. et al. The GeneCards Suite: From Gene Data Mining to Disease Genome Sequence Analyses. Curr. Protoc Bioinformatics. 2016;54:1–30. doi: 10.1002/cpbi.5. [DOI] [PubMed] [Google Scholar]
  14. Barrett T., Wilhite S. E., Ledoux P., Evangelista C., Kim I. F., Tomashevsky M., Marshall K. A., Phillippy K. H., Sherman P. M., Holko M.. et al. NCBI GEO: archive for functional genomics data sets–update. Nucleic Acids Res. 2013;41(Database issue):D991–995. doi: 10.1093/nar/gks1193. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Jerby-Arnon L., Shah P., Cuoco M. S., Rodman C., Su M. J., Melms J. C., Leeson R., Kanodia A., Mei S., Lin J. R.. et al. A Cancer Cell Program Promotes T Cell Exclusion and Resistance to Checkpoint Blockade. Cell. 2018;175(4):984–997.e24. doi: 10.1016/j.cell.2018.09.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Yost K. E., Satpathy A. T., Wells D. K., Qi Y., Wang C., Kageyama R., McNamara K. L., Granja J. M., Sarin K. Y., Brown R. A.. et al. Clonal replacement of tumor-specific T cells following PD-1 blockade. Nat. Med. 2019;25(8):1251–1259. doi: 10.1038/s41591-019-0522-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Sun D., Wang J., Han Y., Dong X., Ge J., Zheng R., Shi X., Wang B., Li Z., Ren P.. et al. TISCH: a comprehensive web resource enabling interactive single-cell transcriptome visualization of tumor microenvironment. Nucleic Acids Res. 2021;49(D1):D1420–d1430. doi: 10.1093/nar/gkaa1020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Colaprico A., Silva T. C., Olsen C., Garofano L., Cava C., Garolini D., Sabedot T. S., Malta T. M., Pagnotta S. M., Castiglioni I.. et al. TCGAbiolinks: an R/Bioconductor package for integrative analysis of TCGA data. Nucleic Acids Res. 2016;44(8):e71. doi: 10.1093/nar/gkv1507. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. De Bruijn I., Kundra R., Mastrogiacomo B., Tran T.N., Sikina L., Mazor T., Li X., Ochoa A., Zhao G., Lai B.. et al. Analysis And Visualization Of Longitudinal Genomic And Clinical Data From The AACR Project GENIE Biopharma Collaborative In cBioPortal. Cancer Res. 2023;83(23):3861–3867. doi: 10.1158/0008-5472.Can-23-0816. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Gao J., Aksoy B. A., Dogrusoz U., Dresdner G., Gross B., Sumer S. O., Sun Y., Jacobsen A., Sinha R., Larsson E.. et al. Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci. Signal. 2013;6(269):l1. doi: 10.1126/scisignal.2004088. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Hugo W., Zaretsky J. M., Sun L., Song C., Moreno B. H., Hu-Lieskovan S., Berent-Maoz B., Pang J., Chmielowski B., Cherry G.. et al. Genomic and Transcriptomic Features of Response to Anti-PD-1 Therapy in Metastatic Melanoma. Cell. 2017;168(3):542. doi: 10.1016/j.cell.2017.01.010. [DOI] [PubMed] [Google Scholar]
  22. Liu D., Schilling B., Liu D., Sucker A., Livingstone E., Jerby-Arnon L., Zimmer L., Gutzmer R., Satzger I., Loquai C.. et al. Integrative molecular and clinical modeling of clinical outcomes to PD1 blockade in patients with metastatic melanoma. Nat. Med. 2019;25(12):1916–1927. doi: 10.1038/s41591-019-0654-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Gide T. N., Quek C., Menzies A. M., Tasker A. T., Shang P., Holst J., Madore J., Lim S. Y., Velickovic R., Wongchenko M.. et al. Distinct Immune Cell Populations Define Response to Anti-PD-1 Monotherapy and Anti-PD-1/Anti-CTLA-4 Combined Therapy. Cancer Cell. 2019;35(2):238–255.e6. doi: 10.1016/j.ccell.2019.01.003. [DOI] [PubMed] [Google Scholar]
  24. Riaz N., Havel J. J., Makarov V., Desrichard A., Urba W. J., Sims J. S., Hodi F. S., Martín-Algarra S., Mandal R., Sharfman W. H.. et al. Tumor and Microenvironment Evolution during Immunotherapy with Nivolumab. Cell. 2017;171(4):934–949.e16. doi: 10.1016/j.cell.2017.09.028. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Van Allen E. M., Miao D., Schilling B., Shukla S. A., Blank C., Zimmer L., Sucker A., Hillen U., Foppen M. H. G., Goldinger S. M.. et al. Genomic correlates of response to CTLA-4 blockade in metastatic melanoma. Science. 2015;350(6257):207–211. doi: 10.1126/science.aad0095. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Mariathasan S., Turley S. J., Nickles D., Castiglioni A., Yuen K., Wang Y., Kadel Iii E. E., Koeppen H., Astarita J. L., Cubas R.. et al. TGFβ attenuates tumour response to PD-L1 blockade by contributing to exclusion of T cells. Nature. 2018;554(7693):544–548. doi: 10.1038/nature25501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Snyder A., Nathanson T., Funt S. A., Ahuja A., Buros Novik J., Hellmann M. D., Chang E., Aksoy B. A., Al-Ahmadie H., Yusko E.. et al. Contribution of systemic and somatic factors to clinical response and resistance to PD-L1 blockade in urothelial cancer: An exploratory multi-omic analysis. PLoS Med. 2017;14(5):e1002309. doi: 10.1371/journal.pmed.1002309. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Zhao J., Chen A. X., Gartrell R. D., Silverman A. M., Aparicio L., Chu T., Bordbar D., Shan D., Samanamud J., Mahajan A.. et al. Immune and genomic correlates of response to anti-PD-1 immunotherapy in glioblastoma. Nat. Med. 2019;25(3):462–469. doi: 10.1038/s41591-019-0349-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Braun D. A., Hou Y., Bakouny Z., Ficial M., Sant’ Angelo M., Forman J., Ross-Macdonald P., Berger A. C., Jegede O. A., Elagina L.. et al. Interplay of somatic alterations and immune infiltration modulates response to PD-1 blockade in advanced clear cell renal cell carcinoma. Nat. Med. 2020;26(6):909–918. doi: 10.1038/s41591-020-0839-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Kim S. T., Cristescu R., Bass A. J., Kim K. M., Odegaard J. I., Kim K., Liu X. Q., Sher X., Jung H., Lee M.. et al. Comprehensive molecular characterization of clinical responses to PD-1 inhibition in metastatic gastric cancer. Nat. Med. 2018;24(9):1449–1458. doi: 10.1038/s41591-018-0101-z. [DOI] [PubMed] [Google Scholar]
  31. Ayers M., Lunceford J., Nebozhyn M., Murphy E., Loboda A., Kaufman D. R., Albright A., Cheng J. D., Kang S. P., Shankaran V.. et al. IFN-γ-related mRNA profile predicts clinical response to PD-1 blockade. J. Clin Invest. 2017;127(8):2930–2940. doi: 10.1172/JCI91190. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Topalian S. L., Hodi F. S., Brahmer J. R., Gettinger S. N., Smith D. C., McDermott D. F., Powderly J. D., Carvajal R. D., Sosman J. A., Atkins M. B.. et al. Safety, activity, and immune correlates of anti-PD-1 antibody in cancer. N Engl J. Med. 2012;366(26):2443–2454. doi: 10.1056/NEJMoa1200690. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Dominguez C. X., Müller S., Keerthivasan S., Koeppen H., Hung J., Gierke S., Breart B., Foreman O., Bainbridge T. W., Castiglioni A.. et al. Single-Cell RNA Sequencing Reveals Stromal Evolution into LRRC15­(+) Myofibroblasts as a Determinant of Patient Response to Cancer Immunotherapy. Cancer Discovery. 2020;10(2):232–253. doi: 10.1158/2159-8290.CD-19-0644. [DOI] [PubMed] [Google Scholar]
  34. Ju M., Bi J., Wei Q., Jiang L., Guan Q., Zhang M., Song X., Chen T., Fan J., Li X.. et al. Pan-cancer analysis of NLRP3 inflammasome with potential implications in prognosis and immunotherapy in human cancer. Brief Bioinform. 2021;22(4):bbaa345. doi: 10.1093/bib/bbaa345. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Rooney M. S., Shukla S. A., Wu C. J., Getz G., Hacohen N.. Molecular and genetic properties of tumors associated with local immune cytolytic activity. Cell. 2015;160(1–2):48–61. doi: 10.1016/j.cell.2014.12.033. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Shukla S. A., Bachireddy P., Schilling B., Galonska C., Zhan Q., Bango C., Langer R., Lee P. C., Gusenleitner D., Keskin D. B., Babadi M.. et al. Cancer-Germline Antigen Expression Discriminates Clinical Outcome to CTLA-4 Blockade. Cell. 2018;173(3):624–633.e8. doi: 10.1016/j.cell.2018.03.026. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Xiong D., Wang Y., You M.. A gene expression signature of TREM2­(hi) macrophages and γδ T cells predicts immunotherapy response. Nat. Commun. 2020;11(1):5084. doi: 10.1038/s41467-020-18546-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Cui C., Xu C., Yang W., Chi Z., Sheng X., Si L., Xie Y., Yu J., Wang S., Yu R.. et al. Ratio of the interferon-γ signature to the immunosuppression signature predicts anti-PD-1 therapy response in melanoma. NPJ. Genom. Med. 2021;6(1):7. doi: 10.1038/s41525-021-00169-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Yan M., Hu J., Ping Y., Xu L., Liao G., Jiang Z., Pang B., Sun S., Zhang Y., Xiao Y.. et al. Single-Cell Transcriptomic Analysis Reveals a Tumor-Reactive T Cell Signature Associated With Clinical Outcome and Immunotherapy Response In Melanoma. Front. Immunol. 2021;12:758288. doi: 10.3389/fimmu.2021.758288. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Hao Y., Hao S., Andersen-Nissen E., Mauck M., Zheng S., Butler A., Lee M. J., Wilk A. J., Darby C., Zager M.. et al. Integrated analysis of multimodal single-cell data. Cell. 2021;184:3573–3587.e29. doi: 10.1016/j.cell.2021.04.048. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Mi H., Muruganujan A., Ebert D., Huang X., Thomas P. D.. PANTHER version 14: more genomes, a new PANTHER GO-slim and improvements in enrichment analysis tools. Nucleic Acids Res. 2019;47(D1):D419–d426. doi: 10.1093/nar/gky1038. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Kanehisa M., Goto S.. KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res. 2000;28(1):27–30. doi: 10.1093/nar/28.1.27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Hänzelmann S., Castelo R., Guinney J.. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics. 2013;14:7. doi: 10.1186/1471-2105-14-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Thorsson V., Gibbs D. L., Brown S. D., Wolf D., Bortone D. S., Ou Yang T. H., Porta-Pardo E., Gao G. F., Plaisier C. L., Eddy J. A.. et al. The Immune Landscape of Cancer. Immunity. 2019;51(2):411–412. doi: 10.1016/j.immuni.2019.08.004. [DOI] [PubMed] [Google Scholar]
  45. Chen B., Khodadoust M. S., Liu C. L., Newman A. M., Alizadeh A. A.. Profiling Tumor Infiltrating Immune Cells with CIBERSORT. Methods Mol. Biol. 2018;1711:243–259. doi: 10.1007/978-1-4939-7493-1_12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Freeman A. J., Vervoort S. J., Ramsbottom K. M., Kelly M. J., Michie J., Pijpers L., Johnstone R. W., Kearney C. J., Oliaro J.. Natural Killer Cells Suppress T Cell-Associated Tumor Immune Evasion. Cell Rep. 2019;28(11):2784–2794.e5. doi: 10.1016/j.celrep.2019.08.017. [DOI] [PubMed] [Google Scholar]
  47. Kearney C. J., Vervoort S. J., Hogg S. J., Ramsbottom K. M., Freeman A. J., Lalaoui N., Pijpers L., Michie J., Brown K. K., Knight D. A.. et al. Tumor immune evasion arises through loss of TNF sensitivity. Sci. Immunol. 2018;3(23):eaar3451. doi: 10.1126/sciimmunol.aar3451. [DOI] [PubMed] [Google Scholar]
  48. Manguso R. T., Pope H. W., Zimmer M. D., Brown F. D., Yates K. B., Miller B. C., Collins N. B., Bi K., LaFleur M. W., Juneja V. R.. et al. In vivo CRISPR screening identifies Ptpn2 as a cancer immunotherapy target. Nature. 2017;547(7664):413–418. doi: 10.1038/nature23270. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Burgess S., Butterworth A., Thompson S. G.. Mendelian randomization analysis with multiple genetic variants using summarized data. Genet Epidemiol. 2013;37(7):658–665. doi: 10.1002/gepi.21758. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Patel S. J., Sanjana N. E., Kishton R. J., Eidizadeh A., Vodnala S. K., Cam M., Gartner J. J., Jia L., Steinberg S. M., Yamamoto T. N.. et al. Identification of essential genes for cancer immunotherapy. Nature. 2017;548(7669):537–542. doi: 10.1038/nature23477. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Vredevoogd D. W., Kuilman T., Ligtenberg M. A., Boshuizen J., Stecker K. E., de Bruijn B., Krijgsman O., Huang X., Kenski J. C. N., Lacroix R.. et al. Augmenting Immunotherapy Impact by Lowering Tumor TNF Cytotoxicity Threshold. Cell. 2020;180(2):404–405. doi: 10.1016/j.cell.2020.01.005. [DOI] [PubMed] [Google Scholar]
  52. Lawson K. A., Sousa C. M., Zhang X., Kim E., Akthar R., Caumanns J. J., Yao Y., Mikolajewicz N., Ross C., Brown K. R.. et al. Functional genomic landscape of cancer-intrinsic evasion of killing by T cells. Nature. 2020;586(7827):120–126. doi: 10.1038/s41586-020-2746-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Rich J. T., Neely J. G., Paniello R. C., Voelker C. C., Nussenbaum B., Wang E. W.. A practical guide to understanding Kaplan-Meier curves. Otolaryngol Head Neck. Surg. 2010;143(3):331–336. doi: 10.1016/j.otohns.2010.05.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Zhang J., Li Y., Liu H., Zhang J., Wang J., Xia J., Zhang Y., Yu X., Ma J., Huang M.. et al. Genome-wide CRISPR/Cas9 library screen identifies PCMT1 as a critical driver of ovarian cancer metastasis. J. Exp. Clin. Cancer Res. 2022;41(1):24. doi: 10.1186/s13046-022-02242-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Huang Y.-L., Liang C.-Y., Ritz D., Coelho R., Septiadi D., Estermann M., Cumin C., Rimmer N., Schótzau A., Núñez López M.. et al. Collagen-rich omentum is a premetastatic niche for integrin α2-mediated peritoneal metastasis. Elife. 2020;9:e59442. doi: 10.7554/eLife.59442. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Wang J., Qi X., Wang Q., Wu G.. The role and therapeutic significance of the anoikis pathway in renal clear cell carcinoma. Front. Oncol. 2022;12:1009984. doi: 10.3389/fonc.2022.1009984. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Liu X., Wang K.. Development of a novel, clinically relevant anoikis-related gene signature to forecast prognosis in patients with prostate cancer. Front. Genet. 2023;14:1166668. doi: 10.3389/fgene.2023.1166668. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. He Y.-W., Fan Q.-P., Hua A.-L., Liu Q.. Identification of hub anoikis-associated genes and risk signature in cutaneous melanoma. Eur. Rev. Med. Pharmacol Sci. 2023;27(12):5662–5676. doi: 10.26355/eurrev_202306_32806. [DOI] [PubMed] [Google Scholar]
  59. Han S., Bao X., Zou Y., Wang L., Li Y., Yang L., Liao A., Zhang X., Jiang X., Liang D.. et al. d-lactate modulates M2 tumor-associated macrophages and remodels immunosuppressive tumor microenvironment for hepatocellular carcinoma. Sci. Adv. 2023;9(29):eadg2697. doi: 10.1126/sciadv.adg2697. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Li M., Jiang P., Wei S., Wang J., Li C.. The role of macrophages-mediated communications among cell compositions of tumor microenvironment in cancer progression. Front. Immunol. 2023;14:1113312. doi: 10.3389/fimmu.2023.1113312. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Li Z., Ding Y., Liu J., Wang J., Mo F., Wang Y., Chen-Mayfield T. J., Sondel P. M., Hong S., Hu Q.. Depletion of tumor associated macrophages enhances local and systemic platelet-mediated anti-PD-1 delivery for post-surgery tumor recurrence treatment. Nat. Commun. 2022;13(1):1845. doi: 10.1038/s41467-022-29388-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Bagati A., Kumar S., Jiang P., Pyrdol J., Zou A. E., Godicelj A., Mathewson N. D., Cartwright A. N. R., Cejas P., Brown M.. et al. Integrin αvβ6-TGFβ-SOX4 Pathway Drives Immune Evasion in Triple-Negative Breast Cancer. Cancer Cell. 2021;39(1):54–67.e9. doi: 10.1016/j.ccell.2020.12.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Shirley C. A., Chhabra G., Amiri D., Chang H., Ahmad N.. Immune escape and metastasis mechanisms in melanoma: breaking down the dichotomy. Front. Immunol. 2024;15:1336023. doi: 10.3389/fimmu.2024.1336023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Kaloni D., Diepstraten S. T., Strasser A., Kelly G. L.. BCL-2 protein family: attractive targets for cancer therapy. Apoptosis. 2023;28(1–2):20–38. doi: 10.1007/s10495-022-01780-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Mukherjee N., Katsnelson E., Brunetti T. M., Michel K., Couts K. L., Lambert K. A., Robinson W. A., McCarter M. D., Norris D. A., Tobin R. P.. et al. MCL1 inhibition targets Myeloid Derived Suppressors Cells, promotes antitumor immunity and enhances the efficacy of immune checkpoint blockade. Cell Death Dis. 2024;15(3):198. doi: 10.1038/s41419-024-06524-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Liu F., Wu Q., Han W., Laster K., Hu Y., Ma F., Chen H., Tian X., Qiao Y., Liu H.. et al. Targeting integrin αvβ3 with indomethacin inhibits patient-derived xenograft tumour growth and recurrence in oesophageal squamous cell carcinoma. Clin. Transl. Med. 2021;11(10):e548. doi: 10.1002/ctm2.548. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Wilbur H. C., Le D. T., Agarwal P.. Immunotherapy of MSI Cancer: Facts and Hopes. Clin. Cancer Res. 2024;30(8):1438–1447. doi: 10.1158/1078-0432.CCR-21-1935. [DOI] [PubMed] [Google Scholar]
  68. Wang Q., Shao X., Zhang Y., Zhu M., Wang F. X. C., Mu J., Li J., Yao H., Chen K.. Role of tumor microenvironment in cancer progression and therapeutic strategy. Cancer Med. 2023;12(10):11149–11165. doi: 10.1002/cam4.5698. [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Pradhan R., Kundu A., Kundu C. N.. The cytokines in tumor microenvironment: from cancer initiation-elongation-progression to metastatic outgrowth. Crit. Rev. Oncol. Hematol. 2024;196:104311. doi: 10.1016/j.critrevonc.2024.104311. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ao5c11091_si_001.pdf (548.4KB, pdf)
ao5c11091_si_002.pdf (234.6KB, pdf)

Data Availability Statement

All data sets utilized in this study are publicly accessible, as detailed in the Methods section. The corresponding web addresses or unique identifiers for each public cohort or data set are provided within the manuscript. Thirty-five stromal/immune-cell scRNA-Seq data sets of malignant cancers were collected from the TISCH database (http://tisch.compbio.cn/). GSE115978 and GSE123813 were downloaded from the GEO database (https://www.ncbi.nlm.nih.gov/geo/).

Articles from ACS Omega are provided here courtesy of American Chemical Society

RESOURCES