Multi-omics analysis links FERMT1 expression to patient survival, immunotherapy response, and metastasis across cancers

Introduction

Cancer-related deaths have become a significant burden on global public health security. According to World Health Organization statistics, cancer incidence and mortality rates are steadily increasing in many countries, including North America, Europe, and the United States (1-4). This trend not only casts a shadow over the patient but also poses a great challenge to the clinical management of cancer (5). Despite notable advances in therapies such as immune checkpoint inhibitors and molecularly targeted agents, reliable methods and biomarkers for effective clinical monitoring are lacking (6,7). Consequently, there is an urgent need to identify new immunoassay markers, such as specific genes or proteins, to provide more alternative strategies for the clinical diagnosis and treatment of cancer.

Fermitin family member 1 (FERMT1), a gene containing a pleckstrin homology structural domain, is an indicator marker that is expressed primarily in epithelial cells, including keratinized cells and intestinal epithelial cells (8). It not only participates in integrin signaling but also mediates the connection between the actin cytoskeleton and the extracellular matrix (9-12). For instance, FERMT1 mutations have been identified as a causative factor in Kindler syndrome. In nasopharyngeal carcinoma, FERMT1 promotes cell migration and invasion by inducing epithelial–mesenchymal transition and disrupting cell cycle progression (13). Additionally, FERMT1 serves as an effective marker for the diagnosis of colorectal cancer, and its expression is correlated significantly with tumor grade, type, and stage (14). In pancreatic cancer, high expression of FERMT1 is a risk factor for poor prognosis (15). Collectively, these findings underscore the critical role of FERMT1 in cellular transformation and highlight its clinical relevance. Nevertheless, systematic investigations of its association with cancer prognosis, tumor microenvironment (TME), and immune infiltration remain limited.

Despite accumulating evidence of FERMT1’s involvement in cancer progression, its functional roles across different cancer types, particularly lung adenocarcinoma (LUAD), are not yet fully understood. In this study, we conducted a comprehensive pan-cancer analysis to evaluate FERMT1 expression patterns, prognostic significance, genetic alterations, and associations with the TME and stemness scores. Furthermore, we investigated the clinical relevance of FERMT1 expression in LUAD and assessed its functional role in metastasis using LUAD cell lines (NCI-H441 and Calu-3). Our findings provide valuable insights into the oncogenic functions of FERMT1 and support its potential as a biomarker and therapeutic target for LUAD. We present this article in accordance with the REMARK and MDAR reporting checklists (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1371/rc).

Methods

Analysis of FERMT1 expression with normal tissue and tumor tissue

We collected 11,069 patient datasets for 33 cancer types from The Cancer Genome Atlas (TCGA) database using the University of California, Santa Cruz (UCSC) Xena online platform (https://xenabrowser.net/datapages/), including RNA expression (HTSeq-FPKM), clinical parameters, immune subtypes, and stemness score datasets (3). The R package “pheatmap” was utilized to compare FERMT1 expression in normal and tumor tissues. We also compared FERMT1 expression in different tumors using the GEPIA 2 (http://gepia2.cancer-pku.cn/) database and examined the relationship between FERMT1 expression and tumor stage.

Correlation analysis of FERMT1 expression level and prognosis

We evaluated the correlation between FERMT1 expression and pan-cancer prognosis using various databases, such as GEPIA 2 and Kaplan-Meier plotter (online databases of GEO, EGA, TCGA, http://kmplot.com/analysis/). GEPIA 2 analyzed 33 different types of cancers to investigate the relationship between FERMT1 expression and overall survival (OS). Afterward, we also conducted a prognostic analysis of FERMT1 expression in pan-cancer using the Kaplan-Meier plotter database.

Mutation value analysis

We explored FERMT1 alterations in TCGA pan-cancer samples using the public database cBioPortal (https://www.cbioportal.org/). Subsequently, we analyzed the FERMT1 alterations. To further confirm the prognostic value of FERMT1 mutations on tumor prognosis, we conducted additional investigations into its mutation sites. We categorized patients into unaltered and altered groups to explore the relationship between genetic variants and clinical prognosis.

Association of FERMT1 expression with TME and stemness score in pan-cancer data

We obtained FERMT1 expression data from the SangerBox online software (http://sangerbox.com) and analyzed the correlation between FERMT1 expression, TME (stromal score and immune score), and stemness score [DNA methylation-based stemness score (DNAss), and RNA-based stemness score (RNAss)] across various cancer types. Subsequently, we visualized the data using R. Through the calculation of stromal and immune scores, the purity, stromal and immune cell infiltration of tumors in pan-cancer could be predicted. In addition, we analyzed the association of FERMT1 expression with DNAss and RNAss using Pearson score and the R package “wilcox. test” and visualized it using R package “corrplot”.

FERMT1 immunophenotypic analysis

The immunosubtyping dataset was downloaded from the UCSC database (C1-C6), which contains data from over 1,000 tumor samples, encompassing 33 different cancer types. The relationship between FERMT1 expression and immune subtypes was comprehensively analyzed using R. A P value of less than 0.05 was considered statistically significant. * denotes P<0.05, ** denotes P<0.01, and *** denotes P<0.001.

Drug resistance and chemosensitivity analysis

The Tumor Immune Dysfunction and Exclusion platform (http://tide.dfci.harvard.edu/) was used to test whether FERMT1 was able to be used as a potential biomarker of immune treatment response. Drug sensitivity treatment data and RNAseq expression data were downloaded from the CellMiner database (http://discover.nci.nih.gov/cellminer/home.do). The RNA-seq expression data were divided into high and low expression groups according to the median level of FERMT1 expression and then statistically and visually analyzed using the R language.

Co-expression, Gene Ontology, and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis

The genes co-expressed with FERMT1 were analyzed using UALCAN. Genes positively associated with FERMT1 in prostate cancer can be downloaded from UALCAN. A gene with a Pearson score greater than 0.3 was selected. The Enrichr database (http://amp.pharm.mssm.edu/Enrichr/) was employed to perform gene ontology functional annotation and KEGG pathway enrichment analysis for these co-expressed genes.

Cell experiment in LUAD

Human LUAD cell lines NCI-H441 and Calu-3 were used for cell validation. The cells were obtained from the Institute of Oncology (Macau University of Science and Technology). Knockdown of selected model genes FERMT1 in NCI-H441 cells and Calu-3 cells. The FERMT1 shRNA knockdown construct consisted of the FERMT1 oligonucleotide sequence FERMT1 (GenBank number: NM_001106515) and was cloned into the lentiviral pLKO.1_puro vector (Public Protein/Plasmid Library) in the lentivirus. The FERMT1 coding sequence (GenBank number: NM_017671.5) was amplified using the following primers: forward primer 5’-GTCTGCTGAAACACAGGATTT-3’, reverse primer 5’-GTTTTTCTAGTGGTTCTCCTT-3’. The effect of FERMT1 on NCI-H441 cell and Calu-3 cell migrations was analyzed by cell scratch method. The cell density of knockdown control group and knockdown group was adjusted to 2×10⁵ cells/well, and the cells were inoculated in 6-well plates with 3 replicates per group, incubated at 37 ℃ and 5% CO₂. After the cells had grown into a monolayer, they were scratched with the tip of a micropipette and then gently washed twice with phosphate-buffered saline to eliminate the scratched cells. Then, samples were collected and photographed at nine-time intervals corresponding to the different incubation times (0, 12, 24, and 48 h). The effect of FERMT1 on the migration and invasion of NCI-H441 cells as well as Calu-3 cells was determined using the Transwell method. For the transwell experiment, the cell density was adjusted to 4×10⁴ cells/well and inoculated in 24-well plates. Transwell cells were fixed with pre-cooled ethanol for 30–45 minutes and stained with crystal violet for 30–45 minutes before being observed under the microscope.

Real-time polymerase chain reaction (PCR) and Western blotting

Total RNA and protein were extracted from the samples. Reverse transcription-quantitative polymerase chain reaction (RT-qPCR) was then validated using SuperReal PreMix Plus (SYBR Green) with ACTIN as an internal reference (forward 5’-CTCTTCCAGCCTTCCTTCCT-3’, reverse 5’- CACCTTCACCGTTCCAGTTT-3’). Proteins were separated by 10% SDS-PAGE, Western blotted, and then blotted onto polyvinylidene difluoride. The films were then closed with 5% skim milk and incubated with specific antibodies at a low temperature (4 ℃) overnight. Antibodies included anti-FERMT1 (22215-1-AP, ThermoFisher) and anti-β-actin (66009-1-lg, Proteintech). Then, after incubating with the secondary antibody at room temperature for 2 h, the Gel Doc™ XR+ imaging system (Bio-rad, USA) was used for detection.

Hematoxylin-eosin (H&E) and immunohistochemical staining

Between March 2024 and April 2024, seven LUAD specimens and normal or paracancerous tissues were obtained from the Affiliated Hospital of Traditional Chinese Medicine of Southwest Medical University. The tissue samples were paraffin-embedded and randomly cut into 4–8 µm thin slices. The sections were then stained with H&E for 5–10 min and observed under a light microscope. Tissue sections were stained immunohistochemically with anti-FERMT1 antibody. Pathologists were blinded to the staining of these specimens and representative images were shown to them. In addition, we obtained the prognostic curve of FERMT1 protein expression from the Human Protein Atlas online database (https://www.proteinatlas.org/).

Wound healing and Transwell assays

Cell migration and invasion were evaluated using wound healing and Transwell assays in NCI-H441 and Calu-3 cells.

For the wound healing assay, cells were seeded in 6-well plates and cultured until reaching confluence, followed by transfection with the indicated FERMT1 constructs or control vectors. A linear scratch was generated using a sterile pipette tip, and detached cells were removed by washing with phosphate-buffered saline (PBS). Images of the wound area were directly captured at 0, 12, 24, and 48 hours under an inverted microscope without staining, and wound closure was assessed by measuring changes in wound width.

For the Transwell assays, 5×10⁴ transfected cells suspended in serum-free medium were seeded into the upper chambers, while medium supplemented with 20% fetal bovine serum was added to the lower chambers. For invasion assays, the upper chambers were pre-coated with diluted matrigel. After 48 hours of incubation, cells that had migrated or invaded to the lower surface of the membrane were fixed and stained with crystal violet. Cells remaining on the upper surface were removed, and the stained cells were photographed and counted.

Ethics consideration

This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The current study was approved by the Ethics Committee of Southwest Medical University, Sichuan, China (No. 20230582 023). Informed consent was obtained from patients for tissue samples.

Statistical analysis

All statistics were performed using R software (version 4.3.0). Wilcox test was used to analyze gene differential expression and immunosubtyping. Pearson correlation was used to study the correlation of gene expression with TME (Stromalscore and Immunescore) and tumor stemness (DNAss and RNAss). RT-PCR was used to calculate the relative gene expression by 2^−ΔΔCt, and the results were expressed as mean ± standard deviation. All experiments were repeated independently at least three times. In all analyses, P<0.05 was considered statistically significant. * denotes P<0.05, ** denotes P<0.01, and *** denotes P<0.001.

Results

The pan-cancer expression pattern of FERMT1

To delineate the expression landscape of FERMT1 across cancers, we created a heat map to analyze its expression levels in 33 tumor types (Figure 1A). Comparative analysis revealed significant FERMT1 dysregulation in 16 cancer types compared to normal or adjacent tissues (Table 1). FERMT1 expression was significantly up-regulated in bladder urothelial carcinoma (BLCA, Figure 1B), cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC, Figure 1C), colon adenocarcinoma (COAD, Figure 1D), cholangiocarcinoma (CHOL, Figure 1E), head and neck squamous cell carcinoma (HNSC, Figure 1F), brain lower-grade glioma (LGG, Figure 1G), LUAD (Figure 1H), lung squamous cell carcinoma (LUSC, Figure 1I), pancreatic adenocarcinoma (PAAD, Figure 1J), esophageal carcinoma (ESCA, Figure 1K), rectum adenocarcinoma (READ, Figure 1L), stomach adenocarcinoma (STAD, Figure 1M), and thymoma (THYM, Figure 1N). In contrast, its expression was significantly down-regulated in adrenocortical carcinoma (ACC, Figure 1O), pheochromocytoma and paraganglioma (PCPG, Figure 1P), and skin cutaneous melanoma (SKCM, Figure 1Q).

Figure 1 The pan-cancer expression pattern of FERMT1 based on the GEPIA2 database. (A) The expression pattern of FERMT1 in 33 different tumor types compared with normal tissues or adjacent para cancerous tissue. (B-Q) The expression of FERMT1 in different tumors: (B) BLCA; (C) CESC; (D) COAD; (E) CHOL; (F) HNSC; (G) LGG; (H) LUAD; (I) LUSC; (J) PAAD; (K) ESCA; (L) READ; (M) STAD; (N) THYM; (O) ACC; (P) PCPG; (Q) SKCM. P<0.05 was considered statistically significant; *, P<0.05. ACC, adrenocortical carcinoma; BLCA, bladder urothelial carcinoma; BRCA, breast invasive carcinoma; CESC, cervical squamous cell carcinoma and endocervical adenocarcinoma; CHOL, cholangiocarcinoma; COAD, colorectal adenocarcinoma; DLBC, lymphoid neoplasm diffuse large B-cell lymphoma; ESCA, esophageal carcinoma; FERMT1, fermitin family member 1; GBM, glioblastoma multiforme; HNSC, head and neck squamous cell carcinoma; KICH, kidney chromophobe; KIRC, kidney renal clear cell carcinoma; KIRP, kidney renal papillary cell carcinoma; LAML, acute myeloid leukemia; LGG, brain lower grade glioma; LIHC, liver hepatocellular carcinoma; LUAD, lung adenocarcinoma; LUSC, lung squamous cell carcinoma; MESO, mesothelioma; OV, ovarian serous cystadenocarcinoma; PAAD, pancreatic adenocarcinoma; PCPG, pheochromocytoma and paraganglioma; PRAD, prostate adenocarcinoma; READ, rectal adenocarcinoma; SARC, sarcoma; SKCM, skin cutaneous melanoma; STAD, stomach adenocarcinoma; TGCT, testicular germ cell tumors; THCA, thyroid carcinoma; THYM, thymoma; UCEC, uterine corpus endometrial carcinoma; UCS, uterine carcinosarcoma; UVM, uveal melanoma.

Correlation between FERMT1 expression and prognosis of human cancers

To evaluate the prognostic value of FERMT1 expression for 33 human cancers, univariate Cox regression modeling was performed. The results showed that FERMT1 expression was strongly associated with the prognosis of multiple tumor types, including OS, disease-specific survival (DSS), disease-free interval (DFI), and progression-free interval (PFI). In ACC, glioblastoma multiforme (GBM), LGG, READ, and uveal melanoma (UVM), FERMT1 was a protective prognostic factor for OS [hazard ratio (HR) <1, P<0.05]. In contrast, in acute myeloid leukemia (LAML), LUAD, PAAD, and SKCM, FERMT1 was considered an unfavorable prognostic factor (HR >1, P<0.05, Figure 2A). Similarly, FERMT1 was associated with favorable prognostic factors for DSS in ACC, LGG, STAD, and UVM (HR <1, P<0.05) but with unfavorable prognostic factors for DSS in PAAD, SKCM, and thyroid carcinoma (THCA; HR >1, P<0.05, Figure 2B). In terms of DFI, FERMT1 was associated with a better prognosis for prostate adenocarcinoma (PRAD) and STAD (HR <1, P<0.05) and a worse prognosis for PAAD (HR >1, P<0.05, Figure 2C). In the prognostic analysis, we also observed that it was considered a low-risk gene for PFI in GBM, LGG, or PRAD (HR <1, P<0.05), but a high-risk gene in PAAD or SKCM (HR >1, P<0.05, Figure 2D).

Figure 2 Univariate Cox expression was performed to investigate the association of pan-cancer FERMT1 expression with overall survival by using TCGA and GTEx databases. (A) Cox analysis of the correlation between FERMT1 expression and OS in different cancer types. (B) The correlation of FERMT1 expression with DSS in different cancer types. (C) The correlation between FERMT1 expression and DFI in different cancer types. (D) Cox analysis of the correlation between FERMT1 expression and PFI in different cancer types. The red words and black dots indicate that the gene is a risk factor in the corresponding tumor. DFI, disease-free interval; DSS, disease-specific survival; FERMT1, fermitin family member 1; PFI, progression-free interval.

In addition, we further validated the relationship between FERMT1 expression patterns and OS in pan-cancer using the GEPIA2 database (Figure 3A). To ensure the authenticity of the results, critical values were utilized for survival analysis. The results showed that high expression of FERMT1 was associated with a prognostically protective effect of GBM (HR <1, P=0.02, Figure 3B) and LGG (HR <1, P=7.5e−05, Figure 3C). In contrast, FERMT1 was considered an unfavorable prognostic factor in LUAD (HR >1, P=0.01, Figure 3D), LUSC (HR >1, P=0.04, Figure 3E), PAAD (HR >1, P=0.03, Figure 3F), and SKCM (HR >1, P<0.001, Figure 3G).

Figure 3 Correlation between FERMT1 expression and overall survival in patients with different cancer types based on GEPIA2 database. (A) Comparison of overall survival curves of high or low expression of FERMT1 in different cancer types from GEPIA2 database. Overall survival curves of FERMT1 in different tumors: (B) GBM; (C) LGG; (D) LUAD; (E) LUSC; (F) PAAD; and (G) SKCM. FERMT1, fermitin family member 1; GBM, glioblastoma multiforme; HR, hazard ratio; LGG, brain lower-grade glioma; LUAD, lung squamous cell carcinoma; LUSC, lung squamous cell carcinoma; PAAD, pancreatic adenocarcinoma; SKCM, skin cutaneous melanoma.

Furthermore, Kaplan-Meier analysis also found that low expression of FERMT1 was associated with a poor prognosis in bladder urothelial carcinoma (BRCA) patients (n=1,090, HR <1, P=0.003, Figure 4A), READ patients (n=165, HR <1, P=0.002, Figure 4B), sarcoma (SARC) patients (n=259, HR <1, P=0.046, Figure 4C), and STAD patients (n=375, HR <1, P=0.02, Figure 4D). On the contrary, a high expression of FERMT1 indicated patients with a poor prognosis in LUAD (n=513, HR >1, P=0.003, Figure 4E), LUSC (n=501, HR >1, P=0.03, Figure 4F), PAAD (n=177, HR >1, P<0.001, Figure 4G), THYM (n=119, HR >1, P=0.007, Figure 4H), and uterine corpus endometrial carcinoma (UCEC; n= 543, HR >1, P=0.003, Figure 4I).

Figure 4 Overall survival curve comparison of high or low pan-cancer expression of FERMT1 by using the Kaplan-Meier Plotter database. (A) BRCA; (B) LUAD; (C) LUSC; (D) PAAD; (E) READ; (F) SARC; (G) STAD; (H) THYM; (I) UCEC. BRCA, breast invasive carcinoma; FERMT1, fermitin family member 1; LUAD, lung adenocarcinoma; LUSC, lung squamous cell carcinoma; PAAD, pancreatic adenocarcinoma; READ, rectum adenocarcinoma; SARC, sarcoma; STAD, stomach adenocarcinoma; THYM, thyroid carcinoma; UCEC, uterine corpus endometrial carcinoma.

FERMT1 genetic alternations across cancers

Through the analysis of FERMT1 mutation data across various human cancers, we observed a prevalent pattern of amplification in most cancer types, with the highest frequency found in UCEC (1% of cases, Figure 5A). The most common genetic alterations in FERMT1 include missense mutations, amplifications, and deep deletions (Figure 5B). Notably, according to the pan-cancer data, the 226 site had the highest mutation rate (Figure 5C). In addition, we present comprehensive data on mutations affecting key structural domains of FERMT1 in a pan-cancer context and highlight structural domains with higher frequencies (Figure 5D). To investigate the impact of FERMT1 mutations on the prognosis of various cancers, we analyzed the correlation between FERMT1 mutations and patient prognosis. However, we did not observe significant differences in patient prognosis to FERMT1 mutations (Figure 5E-5H). We also explored the correlation between FERMT1 expression and tumor mutation burden (TMB) as well as microsatellite instability (MSI), respectively. The results showed that the correlation of FERMT1 expression with TMB and MSI was highest in READ (Figure 5I). Furthermore, FERMT1 expression was most highly correlated with MSI in testicular germ cell tumors (TGCT, Figure 5J).

Figure 5 Analysis of genetic alterations of FERMT1 in different types of cancers based on TCGA, PanCancer Atlas. (A) The frequency of FERMT1 mutations with mutation type across TCGA cancers by cBioPortal. (B) Oncoprint of FERMT1 gene alterations in cancer cohorts. (C) Mutation sites of FERMT1 in pan-cancer data. (D) Mutation sites of FERMT1 in TCGA samples. (E-H) The associations of pan-cancer FERMT1 mutation status with OS, DSS, DFS, and PFS by cBioPortal. (I) The association between pan-cancer expression of FERMT1 and TMB. (J) The association between pan-cancer FERMT1 expression and MSI. DFS, disease-free survival; DSS, disease-specific survival; FERMT1, fermitin family member 1; MSI, microsatellite instability; OS, overall survival; PFS, progression-free survival; TCGA, The Cancer Genome Atlas.

Correlation of FERMT1 expression with TME and stemness score in pan-cancer

The TME consists of various factors with both pro- and anti-cancer effects. These factors stimulate cancer cell heterogeneity, increase multidrug resistance, and contribute to cancer progression and metastasis (16). To investigate the relationship between FERMT1 expression and TME in 33 different cancers, we used the ESTIMATE algorithm. The results showed that FERMT1 expression was significantly positively or negatively correlated with stromal score, immune score, DNAss, and RNAss in pan-cancer (Figure 6A). Then, we listed the relationship between its expression and poor tumor prognosis from different databases. The results showed that FERMT1 expression was significantly associated with improved OS in LUAD, LUSC, and PAAD (Figure 6B). Subsequently, we analyzed the role of FERMT1 expression in the TME of these tumors. The results showed a significant negative correlation between FERMT1 expression and immune scores in LUAD (Figure 6C) and LUSC (Figure 6D). However, there was no significant correlation between FERMT1 expression and stromal scores, DNAss or RNAss. In contrast, in PAAD, FERMT1 expression did not significantly correlate with stromal scores, immune scores, DNAss, and RNAss (Figure 6E).

Figure 6 Association between pan-cancer FERMT1 expression and tumor microenvironment by using TCGA and GTEx database. (A) The association between pan-cancer FERMT1 expression and stromal score, immune score, DNAss, and RNAss. (B) The role of FERMT1 in different types of cancers from different databases. (C-E) FERMT1 expression is correlated with stromal score, immune score, DNAss, and RNAss in LUAD, LUSC, and PAAD. (F-H) The relationship between FERMT1 expression and immune infiltration subtypes in LUAD, LUSC, and PAAD. DNAss, DNA methylation-based stemness scores; FERMT1, fermitin family member 1; LUAD, lung adenocarcinoma; LUSC, lung squamous cell carcinoma; PAAD, pancreatic adenocarcinoma; RNAss, RNA-based stemness score; TCGA, The Cancer Genome Atlas.

Considering the correlation between immune subtypes and prognostic changes, genetics, and immune regulation in various types of cancers (17), we synthesized and analyzed the intrinsic relationship between FERMT1 expression and different immune subtypes, including C1 (wound-healing), C2 (IFN-dominant), C3 (inflammatory), C4 (lymphocyte-depleted), C5 (immunologically-quiet), and C6 (TGF-β-dominant) synthesized and analyzed. The analysis revealed significant variation in the expression of FERMT1 was very different in different immune subtypes of LUAD and PAAD. Specifically, FERMT1 expression was higher in the C1 subtype of LUAD and lower in the C4 subtype of LUAD (Figure 6F). In LUSC, the expression level of FERMT1 was higher in C4 and lower in C6 (Figure 6G). We also observed that the expression level of FERMT1 was higher in C2 and lower in C3 (Figure 6H). These findings strongly suggest that FERMT1 plays a significant regulatory role in TME, especially in LUAD, LUSC, and PAAD.

We also analyzed the association of FERMT1 with pan-cancer immune cells, including B cells, CD4⁺ T cells, CD8⁺ T cells, neutrophils, macrophages, and dendritic cells (Figure 7A). Furthermore, FERMT1 expression was significantly correlated with the presence of CD4⁺ T cells and macrophages in LUAD (Figure 7B). In the analysis of immune cells, we also observed a significant association between FERMT1 expression and B cells, CD8⁺ T cells, CD4⁺ T cells, neutrophils, and dendritic cells in LUSC (Figure 7C). To investigate the potential mechanism by which FERMT1 affects TME, we conducted single-cell RNA sequencing. The results showed that the gene was predominantly expressed in epithelial cells and fibroblasts, especially in non-small cell lung cancer (NSCLC, Figure 7D). In addition, we analyzed the impact of FERMT1 expression on immune cells and its effect on prognosis. We found a significant correlation between FERMT1 expression and prognosis in LUAD (Figure 7E), suggesting that FERMT1 may influence patient prognosis by regulating immune cell infiltration.

Figure 7 The correlation between pan-caner FERMT1 expression and immune cells based on TCGA and Kaplan–Meier database. (A) The relationship between FERMT1 expression and immune cells in different tumors. (B) Effects of genetic alterations in FERMT1 on immune cells in LUAD. (C) Effects of genetic alterations in FERMT1 on immune cells in LUSC. (D) FERMT1 expression in NSCLC. (E) Correlation between FERMT1 expression and prognosis in LUAD and LUSC. FERMT1, fermitin family member 1; LUAD, lung adenocarcinoma; LUSC, lung squamous cell carcinoma; NSCLC, non-small cell lung cancer; TCGA, The Cancer Genome Atlas.

The diagnostic value of FERMT1 expression and chemotherapy sensitivity

We have previously demonstrated an important correlation between FERMT1 expression and the expansion of immune cells. Building on this, we investigated the potential role of FERMT1 as a biomarker for predicting immunotherapy response. The results showed that FERMT1 has a strong predictive ability for the outcome of immunotherapy in NSCLC patients [Nathanson2017_CTLA4_Melanoma_Pre, area under the recipient’s working characteristic curve (AUC) =0.95, Figure 8A]. Furthermore, in the Mariathasan2018_PDL1 BLCA cohort, high expression of FERMT1 was associated with a poor prognosis. Its expression significantly influenced the response to immune checkpoint blockade [anti-programmed death-ligand 1 (PD-L1)] therapies, leading to decreased OS (P=0.04, Figure 8B,8C). In addition, we observed a significant correlation between FERMT1 expression and pathological stage in kidney chromophobe (KICH; P=0.02, Figure 8D), liver hepatocellular carcinoma (LIHC; P=0.01, Figure 8E), and PAAD (P=0.04, Figure 8F), suggesting a potential role of FERMT1 in tumor progression and aggressiveness. To further investigate the potential correlation between FERMT1 expression and drug sensitivity of different human cancer cell lines, we performed a correlation analysis using the CellMiner™ online database. The RNA-seq expression dataset was divided into two groups (high expression group and low expression group) based on the median level of FERMT1 expression. The results revealed negative correlations between FERMT1 expression and drug sensitivity to arsenic trioxide (Figure 8G), okadaic acid (Figure 8H), and mithramycin (Figure 8I). Conversely, FERMT1 expression showed a positive correlation with drug sensitivity of 8-chloro-adenosine (Figure 8J) and Perifosine (Figure 8K) during chemotherapy sensitivity analysis. These findings suggest that FERMT1 expression may influence the response of different cancer types to specific chemotherapeutic agents.

Figure 8 The diagnostic value of FERMT1 expression and chemotherapy sensitivity based on Kaplan-Meier and TCGA database by using Tumor Immune Dysfunction and Exclusion and R language. (A) The correlations between FERMT1 and standard biomarkers in immunotherapy cohorts. The AUC curve was used to evaluate the predictive performance of the tested biomarkers for immunotherapy response status. (B) The prognostic value of FERMT1 versus standard biomarkers in the BLCA cohort. (C) Kaplan-Meier OS curves of patients with high and low FERMT1 expression in the BLCA cohort treated with anti-PD-L1 agents. (D-F) Correlation between FERMT1 expression and pathological stage in KICH, LIHC, and PAAD. (G-K) FERMT1 expression was correlated with drug sensitivity of Arsenic trioxide (G), 8-chloro-adenosine (H), okadaic acid (I), mithramycin (J), and perifosine (K). P<0.05 was considered statistically significant, *, P<0.05; **, P<0.01. AUC, area under the recipient’s working characteristic; BLCA, bladder urothelial carcinoma; FERMT1, fermitin family member 1; KICH, kidney chromophobe; LIHC, liver hepatocellular carcinoma; OS, overall survival; PAAD, pancreatic adenocarcinoma; PD-L1, programmed death-ligand 1; TCGA, The Cancer Genome Atlas.

The correlation between FERMT1 expression and clinical characteristics of patients with LUAD

We previously demonstrated the significance of FERMT1 expression in the TME, especially in LUAD. Therefore, it is very appropriate to further analyze the association between its expression and the clinical characteristics of patients. To analyze the relationship between FERMT1 expression and the clinical characteristics of patients with LUAD, we obtained the clinical datasets from the TCGA database, including the patient’s age, gender, pathologic stage, as well as M/N/T classification and vital status (Table 2). We observed that FERMT1 expression had a significant association with patient T classification (P=0.001; Table 3). Additionally, the expression of FERMT1 was significantly increased in tumor tissues compared with normal tissues (Figure 9A), and it exhibited excellent overall diagnostic value (AUC =0.968; Figure 9B). Notably, the expression of FERMT1 showed no significant difference based on patient age (Figure 9C), gender (Figure 9D), pathological stage (Figure 9E), M classification (Figure 9F), and N classification (Figure 9G), except for T classification (P=0.003; Figure 9H). Subsequently, we also evaluated its diagnostic value in different T classifications of LUAD, including T1/T2 (AUC =0.968; Figure 9I) and T3/T4 (AUC =0.974; Figure 9J). These findings indicate that FERMT1 is upregulated in LUAD, and it is likely closely related to tumor progression.

Figure 9 The correlation between FERMT1 expression and clinical characteristics of patients with LUAD. (A) FERMT1 expression was compared between tumor and normal tissue. (B) The ROC curves of FERMT1 expression in cancerous vs. normal tissues were generated. (C-H) The expression of FERMT1 was compared according to patient age, gender, different pathologic stage, M classification, N classification as well as T classification. (I,J) The ROC curves of FERMT1 expression in cancerous vs. normal tissues were analyzed for different T classifications of LUAD. (K) Kaplan-Meier OS curves of FERMT1 expression in all patients. (L) Kaplan-Meier curves (OS) of FERMT1 expression in the T3/T4/TX subgroup. AUC, area under the curve; FERMT1, fermitin family member 1; LUAD, lung adenocarcinoma; M, metastasis; N, node; OS, overall survival; ROC, receiver operating characteristic; T, tumor.

In addition, we conducted a survival analysis to assess the impact of FERMT1 expression on OS in patients with LUAD. The results showed that patients with high FERMT1 expression had a shorter OS (P=0.003; Figure 9K). Although the subgroup analysis only observed that FERMT1 expression significantly affected patient OS in T3/T4/TX (P=0.02; Figure 9L; Figure S1), multivariate regression analysis revealed that high FERMT1 expression was an independent risk factor for poor OS in LUAD patients (P=0.04; Table 4). Identification of co-expressed genes would facilitate a better understanding of potential functions. We first selected 239 genes (Pearson score >0.6) co-expressed with FERMT1 from the UALCAN database (Figure S2A). Subsequently, gene ontology (functional analysis, including biological process (BP), cellular component (CC), and molecular function (MF), was performed (Figure S2B-S2D). Then, we performed Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis for these genes. The top 10 enriched KEGG pathways were focal adhesion, regulation of the actin cytoskeleton, arrhythmogenic right ventricular cardiomyopathy, ECM-receptor interaction, bacterial invasion of epithelial cells, toxoplasmosis, human papillomavirus infection, Rap1 signaling pathway, yersinia infection, and PI3K-Akt signaling pathway (Figure S2E). These findings would facilitate a better understanding of the role of FERMT1 in LUAD.

FERMT1-deficient reduces migration and invasion in lung cancer cells

Based on our previous findings demonstrating the prognostic significance of FERMT1 expression in certain cancers, it is indeed intriguing to explore its role in tumor metastasis. The results showed that the expression of FERMT1 was positively correlated with lung cancer metastasis (Figure 10A). Studies have also shown that lung cancer patients with high FERMT1 expression have a poorer prognosis (Figure 10B). Subsequently, we collected case tissues from clinical patients and found that FERMT1 was more strongly expressed in tumor tissues (Figure 10C). To further investigate the biological role of FERMT1 in lung cancer progression, we performed cell transfection experiments. The transfection efficiency was verified by RT-qPCR (Figure 10D,10E) and Western blotting (Figure 10F,10G). Wound healing results showed that FERMT1-deficient significantly reduced the wound healing percentage in NCI-H441 cells (Figure 10H,10I) and Calu-3 cells (Figure 10H,10J). We also performed permeation experiments to investigate the invasive capability of the cells (Figure 10K). We found that FERMT1 knockdown inhibited the migration and invasion of NCI-H441 cells (Figure 10L) and Calu-3 cells (Figure 10M) compared to the control group. These results suggest that FERMT1 plays an important role in tumor metastasis, especially in LUAD.

Figure 10 FERMT1 deficiency reduces migration and invasion in lung cancer cells. (A) The level of expression of FERMT1 in normal, tumor, and metastatic tissues of lung cancer. (B) Prediction of prognosis of lung cancer by FERMT1 expression. (C) Pathological sections of cancers and paracancerous tissues and the area within the red circle represents the tumor tissue. (D,E) RT-qPCR analysis verified the downregulation of FERMT1 mediated by specific shRNA targeting FERMT1 in NCI-H441 cells and Calu-3 cells. (F,G) FERMT1 protein expression in NCI-H441 cells and Calu-3 cells. (H) Microscopic observations were recorded at 0, 12, 24, and 48 hours after scratching the surface of a confluent layer of the indicated NCI-H441 cells and Calu-3 cells (unstained). (I,J) Quantitative analysis of wound healing percentage in NCI-H441 cells and Calu-3 cells. (K) Transwell experiments were conducted to assess the impact of FERMT1 on the migration and invasion of NCI-H441 and Calu-3 cells (stained by crystal violet). (L,M) Quantitative analysis of their migratory and invasive capabilities. P<0.05 was considered statistically significant, *, P<0.05; **, P<0.01; ***, P<0.001; ****, P<0.0001. FERMT1, fermitin family member 1; RT-qPCR, reverse transcription-quantitative polymerase chain reaction.

Discussion

The increase in cancer incidence and mortality has become a serious burden on human society (18). Traditional cancer treatments, such as surgery, radiotherapy, and chemotherapy, have limited therapeutic and prognostic effects (19). Currently, the use of biomarkers for disease prevention and treatment is increasingly recognized and accepted. Biomarkers can be used not only to diagnose diseases and determine disease staging but also to evaluate the safety and efficacy of new drugs or therapies for specific patients (20). They have significant clinical value for diagnosis and treatment, particularly in cancer. In recent years, an increasing number of scholars have started studying pan-cancer to gain insights into the etiology, developmental processes, and prognostic features of tumors (21). In this study, we conducted a comprehensive analysis of FERMT1 expression in 10,950 cases of 33 different cancer types, adjacent tissues, and healthy tissues. FERMT1 was significantly overexpressed in 13 tumor types, including BLCA, CESC, COAD, CHOL, HNSC, LGG, LUAD, LUSC, PAAD, ESCA, READ, STAD, and THYM, compared with normal tissues. In addition, FERMT1 was significantly downregulated in ACC, PCPG, and SKCM. These findings suggest that FERMT1 expression is heterogeneous in tumors and provide new insights into the potential use of FERMT1 as a universal biomarker across different types of tumors. However, FERMT1 exhibits varying expression patterns in different tumors, influenced by differences in the TME, metabolic reprogramming, and other factors.

Tumor survival has been recognized as the most effective endpoint for tumor-related clinical trials. Certain treatments or medications can only be considered significant for treating disease if they are proven to extend OS (22). In this study, we analyzed the correlation between FERMT1 expression and the prognosis (OS) of various tumors based on different databases. Although the results from different databases were not consistent, high FERMT1 expression significantly impacted the poor prognosis of PAAD and LUAD. These conflicting results revealed variations in data collection and analysis methods across different databases. Prognostic analyses have revealed that FERMT1 expression is linked to the prognosis of certain tumors. It acts as a detrimental prognostic factor in LUAD, LUSC, PAAD, and SKCM, while serving as a protective factor in GBM and LGG. In addition, we investigated the association of FERMT1 expression with various prognostic categories, including OS, DSS, DFI, and PFI, using univariate Cox regression analysis. The results also showed that the expression of FERMT1 had diagnostic significance for the prognosis of specific tumors, particularly LUAD. These findings provide strong theoretical support to demonstrate that FERMT1 can be used as a LUAD prognostic biomarker.

With the continuous development of medical technology, our understanding and research on cancer have improved greatly. Previous studies have shown that TME plays a crucial role in the development and prognosis of cancer (23). In addition, tumor immunophenotyping offers a comprehensive insight into immune cell infiltration in various cancer types, aiding in the identification of crucial factors for early cancer detection and response to immunotherapy (24). Previous studies have conducted immunogenomic analyses of more than 1,000 tumor samples from 33 cancer types in the TCGA. They identified six distinct immune subtypes, including wound healing (C1), IFN-g predominance (C2), inflammation (C3), lymphocyte depletion (C4), immunoquiescence (C5), and TGF-β predominance (C6) (25). Among these, C1 was linked to wound healing, C4 to lymphocyte depletion, and C6 exhibited the most pronounced TGF-β profile (26). This study unveiled the relationship between FERMT1 expression and TME in various types of cancers, particularly in LUAD and LUSC. We found that FERMT1 expression was higher in the C4 subtype of LUSC and lower in the C4 subtype of LUAD. Combined with previous findings, we hypothesized that FERMT1 expression may influence tumor progression by affecting immune cell infiltration, particularly in LUAD and LUSC. These findings suggest that FERMT1 may play an important role in immune regulation.

Cancer prognosis is influenced by a variety of factors, and a single biomarker may not provide a comprehensive assessment of prognosis. Relationships may vary between different tumor types and individuals and, therefore, need to be evaluated and interpreted in the context of specific studies and clinical settings. TMB and MSI are important biomarkers for assessing cancer prognosis (27). In certain types of tumors, a high TMB is often associated with a better prognosis. This may be attributed to enhanced immune recognition and response, leading to improved treatment outcomes. For example, in bladder cancer patients, complete or partial responders have significantly higher tumor mutational loads than patients with progressed disease (28). Similar to TMB, the significance of high MSI varies across tumor types. In fact, high MSI is often associated with hereditary nonpolyposis colorectal cancer or Lynch syndrome. Tumors with high MSI typically have a better prognosis and may be more responsive to certain treatments such as immune checkpoint inhibitors (29). However, high MSI may also be associated with a poorer prognosis, reflecting a more unstable genome and more aggressive tumor behavior, especially in the context of BRAF^V600E mutations (30,31). Notably, although TMB and MSI play a crucial role as prognostic biomarkers in assessing cancer prognosis, no studies to date have demonstrated their independent utility in predicting patient prognosis. In this study, we found that FERMT1 expression was positively correlated with TMB in READ, HNSC, and ESCA tumors. This suggests that patients with these cancers may have a better prognosis. When investigating the relationship with microsatellite instability, we also observed a significant positive correlation between the expression of FERMT1 and microsatellite instability in tumors such as TGCT, UCS, and CHOL.

Once resistance develops, the therapeutic efficacy of the drug is significantly reduced, potentially resulting in cancer recurrence, a poor prognosis, and ultimately patient mortality (32). In addition, tumor drug resistance is one of the primary causes of treatment failure, which restricts the selection and application of drugs for the clinical treatment of tumors (33). Therefore, the study of drug resistance is crucial, particularly for identifying molecular targets that influence therapeutic responses across different cancer types. Based on various human cancer cell line datasets, we investigated the relationship between FERMT1 expression and drug sensitivity in both pan-cancer and LUAD contexts. Our findings demonstrated that FERMT1 could predict responses to chemotherapy and immunotherapy. In pan-cancer analysis, FERMT1 expression was negatively correlated with drug sensitivity to arsenic trioxide, okadaic acid, and mitomycin, but positively correlated with sensitivity to 8-chloroadenosine and perifosine. Additionally, in LUAD-specific analysis, FERMT1 overexpression was associated with increased resistance to commonly used chemotherapeutic agents, indicating its potential role in mediating drug resistance. This suggests that FERMT1 may be involved in modulating treatment response through mechanisms that require further investigation.

The role of FERMT1 in both pan-cancer and LUAD has been systematically elucidated from multiple perspectives, including expression patterns, survival prognosis, TME, immune infiltration subtypes, and chemosensitivity. However, this study primarily relied on bioinformatics analyses, and further biological validation is needed to confirm its functional impact. While our findings establish FERMT1 as a poor prognostic factor and a key player in immune cell infiltration, it remains unclear whether FERMT1 directly influences patient survival by regulating immune responses. In the future, experimental studies should focus on elucidating the mechanistic role of FERMT1 in tumor heterogeneity and therapeutic resistance, particularly in LUAD, to explore its potential as a biomarker and therapeutic target.

Conclusions

In conclusion, based on multiple databases, this study systematically explored the potential relationship between FERMT1 expression and cancer prognosis, TME, immune infiltration subtypes, and chemotherapy sensitivity across human cancers. We provide a comprehensive pan-cancer evaluation of FERMT1, with a specific focus on its role in LUAD. We found that FERMT1 can serve as a prognostic biomarker of multiple cancer types and revealed its crucial role in LUAD metastasis. These findings help us better understand the important role of FERMT1 as a biomarker in pan-cancer and also provide new ideas and theoretical support for clinical treatment. Future studies should focus on elucidating its molecular mechanisms and potential applications in precision oncology.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the REMARK and MDAR reporting checklists. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1371/rc

Data Sharing Statement: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1371/dss

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1371/prf

Funding: This study was funded by the Science and Technology Development Fund, Macau SAR (Nos. 0114/2022/A and 0018/2024/RIA1) and the Project of Guizhou Provincial Natural Science Foundation [No. QKH-J (2020)1Y378].

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1371/coif). All authors report the funding from the Science and Technology Development Fund, Macau SAR (No. 0114/2022/A) and the Project of Guizhou Provincial Natural Science Foundation [No. QKH-J (2020)1Y378]. The authors have no other conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The current study was approved by the Ethics Committee of Southwest Medical University, Sichuan, China (No. 20230582 023). Informed consent was obtained from patients for tissue samples.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Wang Y, Yan Q, Fan C, et al. Overview and countermeasures of cancer burden in China. Sci China Life Sci 2023;66:2515-26. [Crossref] [PubMed]
2. Siegel RL, Miller KD, Fuchs HE, et al. Cancer Statistics, 2021. CA Cancer J Clin 2021;71:7-33. [Crossref] [PubMed]
3. Siegel RL, Miller KD, Wagle NS, et al. Cancer statistics, 2023. CA Cancer J Clin 2023;73:17-48. [Crossref] [PubMed]
4. de Martel C, Georges D, Bray F, et al. Global burden of cancer attributable to infections in 2018: a worldwide incidence analysis. Lancet Glob Health 2020;8:e180-90. [Crossref] [PubMed]
5. Li Y, Schoufour J, Wang DD, et al. Healthy lifestyle and life expectancy free of cancer, cardiovascular disease, and type 2 diabetes: prospective cohort study. BMJ 2020;368:l6669. [Crossref] [PubMed]
6. Doroshow DB, Bhalla S, Beasley MB, et al. PD-L1 as a biomarker of response to immune-checkpoint inhibitors. Nat Rev Clin Oncol 2021;18:345-62. [Crossref] [PubMed]
7. Gaikwad S, Agrawal MY, Kaushik I, et al. Immune checkpoint proteins: Signaling mechanisms and molecular interactions in cancer immunotherapy. Semin Cancer Biol 2022;86:137-50. [Crossref] [PubMed]
8. Zhong H, Shi Q, Wen Q, et al. Pan-cancer analysis reveals potential of FAM110A as a prognostic and immunological biomarker in human cancer. Front Immunol 2023;14:1058627. [Crossref] [PubMed]
9. Wang X, Chen Q. FERMT1 knockdown inhibits oral squamous cell carcinoma cell epithelial-mesenchymal transition by inactivating the PI3K/AKT signaling pathway. BMC Oral Health 2021;21:598. [Crossref] [PubMed]
10. Wu Q, Li J, Wang P, et al. FERMT1 Is a Prognostic Marker Involved in Immune Infiltration of Pancreatic Adenocarcinoma Correlating with m(6)A Modification and Necroptosis. Genes (Basel) 2023;14:734. [Crossref] [PubMed]
11. Kawamura E, Hamilton GB, Miskiewicz EI, et al. Examination of FERMT1 expression in placental chorionic villi and its role in HTR8-SVneo cell invasion. Histochem Cell Biol 2021;155:669-81. [Crossref] [PubMed]
12. Li L, Li P, Zhang W, et al. FERMT1 contributes to the migration and invasion of nasopharyngeal carcinoma through epithelial-mesenchymal transition and cell cycle arrest. Cancer Cell Int 2022;22:70. [Crossref] [PubMed]
13. Meng L, Yang X, Wu Y, et al. A novel frameshift mutation in the FERMT1 gene in a Chinese patient with Kindler syndrome. Exp Ther Med 2020;20:103. [Crossref] [PubMed]
14. Fan H, Zhang S, Zhang Y, et al. FERMT1 promotes gastric cancer progression by activating the NF-κB pathway and predicts poor prognosis. Cancer Biol Ther 2020;21:815-25. [Crossref] [PubMed]
15. Wu Z, Huang X, Cai M, et al. Novel necroptosis-related gene signature for predicting the prognosis of pancreatic adenocarcinoma. Aging (Albany NY) 2022;14:869-91. [Crossref] [PubMed]
16. de Visser KE, Joyce JA. The evolving tumor microenvironment: From cancer initiation to metastatic outgrowth. Cancer Cell 2023;41:374-403. [Crossref] [PubMed]
17. Onkar SS, Carleton NM, Lucas PC, et al. The Great Immune Escape: Understanding the Divergent Immune Response in Breast Cancer Subtypes. Cancer Discov 2023;13:23-40. [Crossref] [PubMed]
18. Yu P, Xu R, Li S, et al. Exposure to wildfire-related PM2.5 and site-specific cancer mortality in Brazil from 2010 to 2016: A retrospective study. PLoS Med 2022;19:e1004103. [Crossref] [PubMed]
19. Loi S, Settleman J, Joyce JA, et al. The next big questions in cancer research. Cell 2023;186:1523-7. [Crossref] [PubMed]
20. Asci Erkocyigit B, Ozufuklar O, Yardim A, et al. Biomarker Detection in Early Diagnosis of Cancer: Recent Achievements in Point-of-Care Devices Based on Paper Microfluidics. Biosensors (Basel) 2023;13:387. [Crossref] [PubMed]
21. Avdonkina NA, Danilova AB, Nekhaeva TL, et al. Clinical and immunological characteristics of sarcomas patients with clonogenic tumors. Immunobiology 2021;226:152094. [Crossref] [PubMed]
22. Ren Z, Hu M, Wang Z, et al. Ferroptosis-Related Genes in Lung Adenocarcinoma: Prognostic Signature and Immune, Drug Resistance, Mutation Analysis. Front Genet 2021;12:672904. [Crossref] [PubMed]
23. Kang W, Qiu X, Luo Y, et al. Application of radiomics-based multiomics combinations in the tumor microenvironment and cancer prognosis. J Transl Med 2023;21:598. [Crossref] [PubMed]
24. Long J, Wang D, Wang A, et al. A mutation-based gene set predicts survival benefit after immunotherapy across multiple cancers and reveals the immune response landscape. Genome Med 2022;14:20. [Crossref] [PubMed]
25. Soldevilla B, Carretero-Puche C, Gomez-Lopez G, et al. The correlation between immune subtypes and consensus molecular subtypes in colorectal cancer identifies novel tumour microenvironment profiles, with prognostic and therapeutic implications. Eur J Cancer 2019;123:118-29. [Crossref] [PubMed]
26. Bagaev A, Kotlov N, Nomie K, et al. Conserved pan-cancer microenvironment subtypes predict response to immunotherapy. Cancer Cell 2021;39:845-865.e7. [Crossref] [PubMed]
27. Kang YJ, O'Haire S, Franchini F, et al. A scoping review and meta-analysis on the prevalence of pan-tumour biomarkers (dMMR, MSI, high TMB) in different solid tumours. Sci Rep 2022;12:20495. [Crossref] [PubMed]
28. Zhang L, Han X, Shi Y. Association of MUC16 Mutation With Response to Immune Checkpoint Inhibitors in Solid Tumors. JAMA Netw Open 2020;3:e2013201. [Crossref] [PubMed]
29. Sei S, Ahadova A, Keskin DB, et al. Lynch syndrome cancer vaccines: A roadmap for the development of precision immunoprevention strategies. Front Oncol 2023;13:1147590. [Crossref] [PubMed]
30. Sinicrope FA, Shi Q, Allegra CJ, et al. Association of DNA Mismatch Repair and Mutations in BRAF and KRAS With Survival After Recurrence in Stage III Colon Cancers: A Secondary Analysis of 2 Randomized Clinical Trials. JAMA Oncol 2017;3:472-80. [Crossref] [PubMed]
31. Picard E, Verschoor CP, Ma GW, et al. Relationships Between Immune Landscapes, Genetic Subtypes and Responses to Immunotherapy in Colorectal Cancer. Front Immunol 2020;11:369. [Crossref] [PubMed]
32. Pang K, Shi ZD, Wei LY, et al. Research progress of therapeutic effects and drug resistance of immunotherapy based on PD-1/PD-L1 blockade. Drug Resist Updat 2023;66:100907. [Crossref] [PubMed]
33. Ye J, Wu J, Liu B. Therapeutic strategies of dual-target small molecules to overcome drug resistance in cancer therapy. Biochim Biophys Acta Rev Cancer 2023;1878:188866. [Crossref] [PubMed]