Osimertinib resistance-based immune prognostic related gene signature in EGFR mutant lung adenocarcinoma, in which PSMD11 promotes tumor progression

Introduction

Non-small cell lung cancer (NSCLC) is the leading cause of cancer death worldwide. Clinical statistics show that NSCLC accounts for about 85% of lung cancer cases, and lung adenocarcinoma (LUAD) is the most common type (1). Although a variety of LUAD treatments have been developed, the prognosis is still poor, and the 5-year overall survival (OS) of LUAD patients is still less than 20% (2). Intrinsic or extrinsic acquired resistance caused by long-term drug treatment is the main reason for the decline of drug efficacy and rapid tumor recurrence (3). The development of drug resistance is a complex process, which is mainly caused by the disordered expression of genes (4). Therefore, it is necessary to screen the differentially expressed genes (DEGs) associated with drug resistance and analyze the impact of DEGs on the prognosis and drug sensitivity, which is of great significance for finding new targets to reverse drug resistance and improve prognosis.

Epidermal growth factor receptor (EGFR) is a type I transmembrane glycoprotein that is widely expressed in all cells (5). In the United States, about 10–15% of NSCLC patients and up to 40% of patients in Asian populations carry tumors with EGFR activating mutations (6). Therefore, detecting EGFR mutation in LUAD patients has important clinical significance (7). For patients with advanced disease, EGFR tyrosine kinase inhibitors (TKIs), such as osimertinib (OSI), a third-generation drug, are part of the standard first-line treatment. Although the vast majority of EGFR mutant patients initially benefit from EGFR-TKIs, treatment resistance can arise through EGFR-dependent mechanisms, such as secondary EGFR mutation (C797S mutation), or through EGFR-independent mechanisms, such as epithelial-mesenchymal transition or small cell transformation (8-10). Most EGFR mutant patients will develop acquired resistance and eventually die of tumor recurrence after received first-line EGFR-TKIs for 9–11 months (11). The development of EGFR-TKIs resistance is thought to be related to tumor immune escape, aggressive tumor growth, and metastasis (12-14). At present, the standard treatment for TKI-resistant patients is limited, including chemotherapy and anti-PD-1/PD-L1 immunotherapy. Because immunotherapy is usually used after resistance to EGFR-TKIs, and the Food and Drug Administration (FDA) has not approved any drugs for the treatment of OSI resistance, there is an urgent need to develop new treatments for these OSI-resistant patients.

At present, due to the fact that the immune cells in the tumor site play a crucial role in the malignant progression of tumors, the tumor immune microenvironment (TIME) has received extensive attention. There are many strategies for tumor cells in TIME to evade host immune surveillance. Among them, immune checkpoint molecules located on the surface of tumor cells can make tumor cells grow unrestricted (15,16). Based on this, a variety of humanized monoclonal antibodies (such as CTLA-4 and PD-1) that block immune checkpoint molecules have been developed. These immune checkpoint inhibitors (ICIs) can effectively restore the function of immune cells (17,18). Previous studies have shown that TKIs resistance can regulate the state of TIME in LUAD, such as inducing T cells apoptosis, increasing the number of regulatory T cells (Tregs), and promoting the polarization of tumor-associated macrophages to M2 phenotype (19-21). Therefore, there is an urgent need to clarify the potential relationship between TKIs resistance and tumor immunity in LUAD to identify a more effective therapeutic strategy.

Here, we first used the transcriptome data of OSI-resistant (PC9OR, H1975OR) and OSI-sensitive LUAD cells (PC9, H1975) to screen and identify the DEGs associated with OSI resistance. Then, combined with the EGFR mutant LUAD prognosis-related genes and immune gene sets from ImmPort and InnateDB, a series of OSI resistance-related immune prognosis genes were identified, and then the prognosis prediction model and corresponding nomogram of EGFR mutant LUAD patients were constructed. Next, we analyzed the relationship between the prognosis prediction model and the expression of infiltrating immune cells and immune checkpoint molecules. Finally, we determined PSMD11 by analyzing the expression and prognosis of four hub genes, and analyzed the role of PSMD11 in the progression of OSI-resistant LUAD through in vitro and in vivo experiments (Figure 1). In conclusion, we constructed an OSI resistance related immune prognosis prediction model, which provides reliable indicators for predicting the TIME status of EGFR mutant LUAD patients and selecting effective treatment methods (chemotherapy, targeted therapy and immunotherapy) for high-risk patients, and explored the role of PSMD11 in the progression of OSI-resistant LUAD. We present this article in accordance with the ARRIVE and MDAR reporting checklists (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-355/rc).

Figure 1 Flow chart of the analysis. Graphical summary of the construction and validation of OSI resistance related immune prognosis prediction model for EGFR mutant LUAD patients and the analysis of the role of PSMD11 in the progression of OSI-resistant LUAD. DEGs, differentially expressed genes; EGFR, epidermal growth factor receptor; LUAD, lung adenocarcinoma; OSI, osimertinib.

Methods

Isolation of transcriptome data and tissue microarray

Transcriptome data, mutation data, and clinical information of LUAD samples (535 tumors and 59 normal lung samples) were downloaded from The Cancer Genome Atlas (TCGA) database (https://portal.gdc.cancer.gov/). The transcriptome data and clinical information of the validation dataset GSE31210 with 126 EGFR mutant LUAD samples were downloaded from Gene Expression Omnibus (GEO; https://www.ncbi.nlm.nih.gov/geo). Immune gene sets were downloaded from the ImmPort (https://www.immport.org/shared/home) and InnateDB (https://www.innateDBdb.com/) databases.

We retrospectively collected the frozen tissue samples of 160 patients with LUAD from the biobank of our center, and made a 160-spot LUAD tissue microarray by Wuhan Servicebio Co., Ltd. (Wuhan, China). This study was approved by the Ethics Committee of Tangdu Hospital, The Fourth Military Medical University (No. GKJ-Y-202303-181). Written informed consent was obtained from the individual for the publication of any potentially identifiable data included in this article. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. All patients underwent surgery between June 2021 and July 2022. All patients had no special medical history. The no special medical history refers to (within 5 years): no other history of malignant tumors; no history of pulmonary fibrosis, pneumoconiosis, interstitial pneumonia, drug-related pneumonia, radiation pneumonia, severe damage to lung function, and other diseases; no history of infectious diseases; no history of immune deficiency; no history of organ transplantation.

Screening and functional analysis of OSI resistance-related DEGs

Firstly, OSI-resistant LUAD cell lines PC9OR and H1975OR were constructed using drug concentration increasing method based on parental cells PC9 and H1975. Then, each cell line prepared three biological replicates for sequencing, and the data type was transcripts per million (TPM). In order to obtain DEGs between OSI-resistant and OSI-sensitive LUAD cells, differential analysis was performed using R package “limma”. The threshold for DEGs was |log₂fold change (log₂FC)| >1.5 and P<0.05. Finally, the function enrichment analysis of DEGs was carried out through Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG).

Constructing an OSI resistance-related immune prognosis prediction model

We screened 58 LUAD patients with EGFR mutation in the TCGA database for subsequent analysis (Table S1). According to P<0.05, prognostic related genes for EGFR mutant LUAD patients were screened through univariate Cox regression analysis (average gene expression >0.5). At the same time, we collected immune gene sets manually from ImmPort and InnateDB websites. Then, the intersection of DEGs (n=948), prognostic related genes of EGFR mutant LUAD patients (n=1,531), and immune gene sets (n=2,659) was calculated to obtain 18 common genes, and a Venn plot was drawn.

Based on P<0.05 for screening, four hub genes were identified through multivariate Cox regression analysis to construct an OSI-resistant related immune prognosis prediction model. The calculation formula was as follows: risk score = Σ(C × EXP_RNA). In our formula, EXP_RNA represents the expression of four hub genes, and C represents the corresponding coefficients for multivariate Cox regression analysis. Using the median risk score as the optimal threshold, EGFR mutant LUAD patients were divided into high- and low-risk groups. The Kaplan-Meier survival curve and time-dependent receptor operating characteristic (ROC) curve were used to evaluate the predictive performance of the model for OS. Finally, gene set enrichment analysis (GSEA) was performed on both high- and low-risk groups, while protein-protein interaction (PPI) networks, GO, and KEGG functional analysis were performed on these four hub genes.

Construction of the nomogram predicting OS for EGFR mutant LUAD patients

Conduct univariate and multivariate Cox regression analysis along with other clinical features [including tumor-node-metastasis (TNM) staging, tumor (T) category, node (N) category, age, and gender] to determine the accuracy and independence of these four gene signatures in predicting OS in EGFR mutant LUAD patients. P<0.05 is considered statistically significant. Calculate the hazard ratio (HR) and 95% confidence interval (CI) for each variable. In order to better predict the outcome of EGFR mutant LUAD patients, we constructed a nomogram based on risk score and clinical features. Use calibration curves to evaluate the prediction accuracy of nomogram. If the calibration curve is closer to the 45° line, the nomogram’s predictive performance is better.

Molecular, immune characteristics, and therapeutic responsiveness of gene signature

Based on the 58 LUAD patients with EGFR mutation in the TCGA database, we compared the molecular features of high- and low-risk groups using R package “maftools”. Based on the risk score, the R package “CIBERSORT” (22) was used to estimate the proportion of 22 infiltrating immune cells in the high- and low-risk groups. Tumor Immune Dysfunction and Exclusion (TIDE) score was calculated online (http://tide.dfci.harvard.edu/). We compared the differences between high- and low-risk groups in TIDE. In addition, we also compared the differences in tumor mutation burden (TMB) and immune checkpoint molecules PD-L1, LAG3, and IDO1 in the high- and low-risk groups. Finally, according to the half-maximal inhibitory concentration (IC₅₀) of each LUAD patient on the Genomics of Drug Sensitivity in Cancer (GDSC) website (https://www.cancerrxgene.org/), the responsiveness of high- and low-risk groups to conventional chemotherapy and targeted drugs was evaluated using “pRRophetic” R package. which is designed to predict clinical chemotherapy responses based on baseline gene expression levels. This package contains preprocessed data and multiple datasets, and predicts drug sensitivity through five-fold cross-validation.

Analysis of the expression, prognosis, and immune characteristics of four hub genes

Based on the sequencing data of LUAD patients in the TCGA database, we analyzed the expression differences of four hub genes in normal lung tissues and LUAD tissues. Four hub genes were divided into two groups according to median expression, and survival analysis was performed on LUAD patients using R packages “survival” and “survminer”. Kaplan-Meier and logarithmic rank test methods were used to evaluate survival curves and OS. P<0.05 is considered statistically significant. Analyze the protein expression differences of four hub genes in LUAD tissues and normal lung tissues through the Human Protein Atlas (HPA) database (https://www.proteinatlas.org/). Based on the proportion of 22 infiltrating immune cells in LUAD samples, we analyzed the correlation between PSMD11 and 22 immune cells. In addition, we also analyzed the correlation between PSMD11 and immune checkpoint molecules PD-L1, LAG3, and IDO1 expressions in LUAD patients.

Cell culture

LUAD cell lines (PC9 and H1975) were purchased from Shanghai Academy of Science (Shanghai, China). OSI-resistant LUAD cell lines PC9OR and H1975OR were constructed from parental cells (PC9 and H1975) using drug concentration increasing method. The four types of cells were cultured in 1640 complete medium with 10% fetal bovine serum.

Transfection of small interfering RNA (siRNA)

Knockdown treatment was performed on PC9OR and H1975OR cells using PSMD11-siRNAs designed by a genetic pharmaceutical company (Shanghai, China). Inoculate PC9OR and H1975OR cells into a six-well plate, and when the cell density reaches 60–70%, transfect PSMD11-siRNA or negative control (NC)-siRNA using Lipofectamine 2000 (11668-027, Invitrogen, Carlsbad, CA, USA) as instructed. After 48 h of transfection, western blot was performed to detect its knockdown efficiency. The target sequences were as follows: PSMD11-siRNA1, CCGACGTGGAAAGGAAATTAT; PSMD11-siRNA2, GAGTACAGATTGAACACATAT; PSMD11-siRNA3, CTGGTGTCTTTGTACTTTGAT.

IC50 and cell activity assays

After transfection of PC9OR and H1975OR for 24 h, OSI was prepared using equal volume culture media at concentrations of 0, 1, 10, 50, 100, 500, 1,000, 5,000, and 10,000 nM. The control group was added with corresponding amounts of DMSO solution. After being incubated for 48 h, remove the cell supernatant and add 110µl serum-free medium containing 10% Cell Counting Kit-8 (CCK-8; Beyotime, Shanghai, China) to each well for incubation for 1 h. Use a microplate reader to measure the absorbance of each well at 450 nM and calculate the IC₅₀ values.

At the indicated time (4, 24, 48, 72, and 96 h), remove the cell supernatant and incubate it in 110 µL serum-free 1640 medium containing 10% CCK-8 (Beyotime) for 1 h. Use a microplate reader to detect the optical density (OD) values of each well at 450 nM to evaluate the activity of drug-resistant cells.

Western blot

The methods for obtaining total proteins in tissues and cells are reported in literature (23). The protein was separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to the polyvinylidene fluoride (PVDF) membrane. After blocking with 5% bovine serum albumin (BSA) for 1 h, incubate the membrane with the corresponding primary antibody overnight at 4 ℃. Then,

incubate with horseradish peroxidase (HRP) labeled antirabbit/mouse secondary antibody (A0208/A0216, Beyotime) at room temperature for 1 h. Finally, the PVDF membrane was exposed using a chemiluminescence reagent kit (P0018M, Beyotime). The dilution ratios of primary antibodies rabbit anti-phosphoNF-κB p65 (3033S, Cell Signaling Technology, Danvers, MA, USA), rabbit anti-NF-κB p65 (8242S, Cell Signaling Technology), mouse anti-caspase3 (66470-2-Ig, Proteintech, Wuhan, China), mouse anti-PARP1 (66520-1-Ig, Proteintech), rabbit anti-PSMD11 (14786-1-AP, Proteintech), rabbit anti-STAT3 (10253-2-AP, Proteintech), rabbit anti-phospho-STAT3 (Tyr705) (9145S, Cell Signaling Technology), and rabbit-anti-EGFR (66455-1-Ig, Proteintech) were all 1:1,000. The dilution ratios of loading control antibodies, rabbit anti-GAPDH (1E6D9, Proteintech), rabbit anti-β-actin (66009-1-Ig, Proteintech), and mouse anti-Lamin B1 (66095-1-Ig, Proteintech) were 1:5,000.

Colony formation assay

Inoculate 500 cells per well on a six-well plate. After 24 h, add IC₅₀ of OSI to the experimental group, and treat the control group with an equal amount of DMSO solution. Continue to culture for 2–3 weeks. When visible clone formation occurs, culture is discontinued. Aspirate the culture medium and wash the cells with phosphate-buffered saline (PBS) once. Add 1 mL of 4% paraformaldehyde to each well and fix at room temperature for 30–60 minutes. Wash with PBS once again. Add 1 mL of 0.1% crystal violet staining solution to each well and stain at room temperature for 10–20 minutes. Rinse gently with PBS for several times, air-dry, and then take photos for documentation.

Apoptosis analysis

Inoculate 2×10⁵ cells per well into a six-well plate. After 24 h, add IC₅₀ of OSI to the experimental group and an equal amount of DMSO solution to the control group. Then, continue to culture for 48 h, digest cells with 2.5% trypsin [excluding ethylenediaminetetraacetic acid (EDTA)], and collect cells and supernatant. Add 300 µL Annexin V binding solution to resuspend cells, then add 4 µL Annexin V solution and 10 µL propidium iodide (PI) solution for mixing, and incubate in dark room temperature for 20 minutes. Place the incubated cell suspension in the sample well of flow cytometry (BD Biosciences, San Jose, CA, USA), and detect PI fluorescence through phycoerythrin (PE) channel and Annexin V fluorescence through fluorescein isothiocyanate (FITC) channel.

Enzyme-linked immunosorbent assay (ELISA)

To detect the autonomous secretion of TNFα, IL-6, and IL-17, LUAD cells PC9, PC9OR, H1975, and H1975OR were cultured at 1×10⁶ cells/mL for 24 h, and the supernatant was collected and stored at −80 ℃. According to the manufacturer’s instructions, TNFα was tested using the human TNFα quantitative factor ELISA kit (#DTA00D, R&D Systems, Minneapolis, MN, USA), IL-6 was tested using the human IL-6 quantitative factor ELISA kit (#D6050, R&D Systems), and IL-17 was tested using the human IL-17 quantitative factor ELISA kit (#QK317, R&D Systems, Minneapolis, MN, USA). The signal was quantified on a CLARIOstar^®Plus (BMG Labtech, Ortenberg, Germany) tablet reader.

Xenograft mouse model

We injected stable knockdown PSMD11 or NC-siRNA cell lines PC9OR and H1975OR subcutaneously into 6-week-old female BALB/c nude mice (n=5 per group). All mice were purchased from GemPharmatech Co., Ltd. (Nanjing, China), and raised in a standard specific-pathogen-free (SPF) environment. Starting from the 7^th day, the long diameter (a) and short diameter (b) of the tumor were measured every 2 days. The calculation formula for tumor volume was: total volume = a×b²/2. When the volume of the transplanted tumor reached 100 mm³, the mice were randomly divided into four groups and then treated with OSI [5 mg/kg, administered for 3 weeks, twice a week, oral administration (p.o.)]. When the maximum tumor volume reaches 1,000 mm³ or there is significant difference between groups, the nude mice shall be euthanized according to ethical requirements, and the tumor volume and weight shall be recorded and photographed for preservation. All animal experiments were approved by the Animal Ethics Committee of Tangdu Hospital, The Fourth Military Medical University (No. 20240132), and the investigation complied with the National Institutes of Health Guidelines for the care and use of laboratory animals. A protocol was prepared before the study without registration.

Immunohistochemistry

The excised tumor tissue was fixed with formalin, dehydrated, paraffin-embedded, and then sliced. Sections were heated by microwave to remove antigens. Remove endogenous peroxidase activity with 3% hydrogen peroxide. Incubate in 5% BSA for 20 min to block non-specific binding, and then incubate the section with the primary antibody at 4 ℃ overnight. The dilution ratios of primary antibodies, rabbit anti-PSMD11 (14786-1-AP, Proteintech) and mouse anti-caspase3 (66470-2-Ig, Proteintech), were 1:200. Then, incubate with biotinylated secondary antibody (dilution ratio: 1:500) at 37 ℃ for 1 h. Then, add DAB chromogenic solution (AR1022, Boster Biological Technology, Wuhan, China) and incubate at room temperature in the dark for several minutes; the positive sites will appear brown. Counterstain the cell nuclei briefly with hematoxylin to make the nuclei blue, which contrasts with the brown positive signals, facilitating localization and observation. Dehydrate sequentially with gradient ethanol (75% → 85% → 95% → 100%), clear with xylene, and finally mount the sections with neutral gum to prepare permanent sections for microscopic observation and photography. Sections were observed under an inverted fluorescence microscope, with each section photographed and positive cells counted in five random fields of view (magnification, 400×). Analyze the positive staining rate using ImageJ.

Terminal deoxynucleotidyl transferase (TdT)-mediated dUTP nick end labeling (TUNEL)

To detect whether OSI, PSMD11-siRNA, or the combination of OSI and PSMD11-siRNA could induce apoptosis in the tumor tissue of mice, we performed the TUNEL method using a One Step TUNEL Apoptosis Assay Kit (C1086, Beyotime) to label 3'-end of fragmented DNA of the apoptotic LUAD cells. Briefly, sections were deparaffinized with xylene and rehydrated with ethanol. Then, 20 µg/mL proteinase K was added to treat the tissue for 30 min, and after complete washing with PBS, apoptotic cells were labeled at 37 ℃ for 1 h with TUNEL detection solution containing TdT and FITC labeled dUTP. FITC-labeled TUNEL-positive cells were imaged using a fluorescence microscope, and cells with green fluorescence were described as apoptotic cells.

Statistical analysis

Statistical analysis of all experimental data was performed using SPSS 22.0 (SPSS Inc., Chicago, IL, USA) and graphical presentation of experimental results was performed using Graphpad Prism 8.0 (Graphpad Inc., La Jolla, CA, USA). All experiments were repeated at least three times. The experimental data were expressed as mean ± standard deviation (SD). Semi-quantitative results of all CCK-8 assays and western blots were analyzed using t-test or Mann-Whitney U test. The effects of knockdown PSDM11 on colony formation and apoptosis in LUAD cells were analyzed using one-way analysis of variance (ANOVA). A two-tailed Student’s t-test was used for cell line and mice data. All data with normal distribution and homogeneity of variance were expressed as mean ± standard deviation. All results were considered statistically significant at P<0.05 (*, P<0.05; **, P<0.01; ***, P<0.001).

Results

Construction and validation of OSI resistance-related immune prognosis prediction model

Firstly, based on the parental PC9 and H1975 cell lines, we constructed OSI-resistant LUAD cell lines PC9OR and H1975OR using the drug concentration increasing method, and detected the IC₅₀ of OSI in four types of cells. The results showed that the IC₅₀ of OSI in four types of cells (PC9, PC9OR, H1975, and H1975OR) were 579, 4,215, 313, and 4,866 nM (Figure 2A,2B). Then, we analyzed DEGs in mRNA expression between OSI-resistant and OSI-sensitive cells, including 267 upregulated genes and 681 downregulated genes (Figure 2C). Through functional enrichment analysis of DEGs, the results showed that DEGs were the most abundant in terms related to immune function, such as leukocyte proliferation, CD4⁺ T cell activation, cytokine regulation, and myeloid activation (Figure S1A,S1B). This result indicates that the interaction between OSI resistance-related DEGs and tumor immunity plays an important role in the progression of LUAD.

Figure 2 Construction and validation of OSI resistance related immune prognosis prediction model. (A,B) CCK-8 detected the IC₅₀ of OSI in OSI-resistant cells (PC9OR, H1975OR) and OSI-sensitive cells (PC9, H1975). (C) The volcano plot displayed the DEGs of OSI sensitive and resistance cells. (D) The Venn plot showed the common genes of DEGs (n=948), EGFR mutant LUAD prognosis related genes (n=1,531), and immune gene set (n=2,659). (E) The forest plot showed an immune prognosis prediction model constructed based on four OSI-resistant hub genes through multivariate Cox regression analysis. (F,G) The distribution of risk score, Kaplan-Meier curve and 3-year ROC curve were drawn after the LUAD cohort with EGFR mutation was divided into high- and low-risk groups according to the risk score. (H,I) The Kaplan-Meier curve and 3-year ROC curve of 126 EGFR mutant LUAD patients in the validation set GSE31210. (J,K) GSEA of high- and low-risk groups, and displayed the top 5 signaling pathways. (L) Conduct univariate and multivariate Cox regression analysis based on clinical characteristics and risk score. (M) Draw a nomogram for predicting OS based on the independent prognostic factor. Gender: 0, male; 1, female. *, P <0.05; ***, P <0.001. AIC, Akaike information criterion; AUC, area under the curve; CCK-8, Cell Counting Kit-8; CI, confidence interval; DEGs, differentially expressed genes; EGFR, epidermal growth factor receptor; GSEA, gene set enrichment analysis; HR, hazard ratio; IC₅₀, half-maximal inhibitory concentration; KEGG, Kyoto Encyclopedia of Genes and Genomes; LUAD, lung adenocarcinoma; N, node; OS, overall survival; OSI, osimertinib; ROC, receptor operating characteristic; T, tumor; TCGA, The Cancer Genome Atlas; TNM, tumor-node-metastasis.

Subsequently, based on P<0.05, 1,531 prognostic-related genes were identified in 58 LUAD patients with EGFR mutation through univariate Cox regression analysis (table available at https://cdn.amegroups.cn/static/public/tlcr-2025-355-1.xls). By intersecting 948 DEGs, 1,531 prognostic-related genes, and 2659 immune genes from ImmPort and InnateDB databases, 11 OSI resistance-related immune prognostic genes were obtained (Figure 2D, table available at https://cdn.amegroups.cn/static/public/tlcr-2025-355-2.xls). Then, based on these genes, a predictive model for OSI resistance-related immune prognosis was constructed based on four hub genes (C3, PSMD11, G3BP1, TRIB2) (Figure 2E). Through PPI network analysis, we identified 82 genes directly related to four hub genes (Figure S1C). And these genes were mainly enriched in the TNF signaling pathway, NF-κB signaling pathway, Th17 cell differentiation, and apoptosis (Figure S1D,S1E). Based on the expression levels of four hub genes, we constructed a risk score model: risk score = (−0.005178 × expression level of C3) + (0.157063 × expression level of PSMD11) + (0.085348 × expression level of G3BP1) + (0.056369 × expression level of TRIB2). Then, we divided EGFR mutant LUAD patients into high- and low-risk groups using the median of risk score. It could be seen that compared to the low-risk group, the high-risk group has more deaths and poorer prognosis (Figure 2F, P<0.001). Area under the curve (AUC) of 3-year ROC was 0.704 (Figure 2G). Then, we validated the accuracy of the model in the validation set GSE31210 containing 126 EGFR mutation LUAD samples. The results showed that compared to the low-risk group, the high-risk group had a worse prognosis (Figure 2H, P=0.005), and the AUC of 3-year ROC was 0.764 (Figure 2I). These results indicate that this gene signature is a good prediction model for OSI resistance-related immune prognosis in this training cohort. Finally, we conducted GSEA on high- and low-risk groups, and the results showed that the high-risk group mainly activated cell cycle and DNA replication (Figure 2J,2K, Table S2).

Constructing a nomogram for predicting OS in EGFR mutant LUAD patients

To demonstrate that the risk score derived from these four gene signatures was an independent biomarker, we conducted Cox regression analysis in the EGFR mutant LUAD cohort. Univariate and multivariate Cox regression analysis showed that risk score (P<0.001; HR =1.198; 95% CI: 1.090–1.316) was associated with OS in EGFR mutant LUAD patients, indicating that risk score is an independent prognostic indicator for LUAD (Figure 2L). In order to accurately predict OS, we constructed a prognostic nomogram based on the independent prognostic factor identified above. This result indicates that the risk score is a good survival predictor compared to the TNM stage (Figure 2M). We also used calibration curves to evaluate the accuracy of the nomogram model, and the results showed that this model has high predictive consistency (Figure S2A-S2C).

Molecular, immune characteristics, and therapeutic response of EGFR mutant LUAD patients in high- and low-risk groups

To evaluate the molecular characteristics of risk score, we analyzed gene mutations in the high- and low-risk groups and displayed the top 20 gene mutations (Figure 3A,3B). The most common mutation type was missense mutation, followed by nonsense mutation. The mutation rates of TP53, TTN, and MUC16 in the high- and low-risk groups all exceeded 20%. To determine the immune cell composition of the LUAD cohort, we calculated the proportion of infiltrating immune cells in the high- and low-risk groups using R package “CIBERSORT”. We found that the proportion of activated natural killer (NK) cells and M0/M1 macrophages was higher in the high-risk group, while the proportion of resting CD4 memory T cells, resting dendritic cells, and resting mast cells was higher in the low-risk group (Figure 3C). Then, we used TIDE to score T cell dysfunction from high- and low-risk groups and predicted its therapeutic response to ICIs. If the TIDE score is high, it is more likely for patients to happen immune evasion (24). The results showed that compared to the low-risk group, the high-risk group had a lower TIDE score (Figure 3D). At the same time, we found that compared with the low-risk group, the high-risk group had higher expression of TMB and immune checkpoint molecules IDO1, LAG3, and PD-L1 (Figure 3E-3H, P<0.05). These results indicate that LUAD patients in the high-risk group are likely to benefit from immunotherapy. Finally, we further explored the responsiveness of LUAD patients in the high- and low-risk groups to conventional chemotherapy and targeted drugs. The result showed that the high-risk group had better responsiveness to chemotherapy and targeted drugs than the low-risk group (Figure 3I).

Figure 3 Molecular, immune characteristics, and therapeutic response of EGFR mutated LUAD patients in high- and low-risk groups. (A,B) The waterfall chart shows the top 20 gene mutations in the high- and low-risk groups of EGFR mutated LUAD patients. The percentage of mutations is displayed on the right, and the total number of mutations is displayed on the top. Different colors represent different types of mutations. (C) The box plot shows the distribution of 22 infiltrating immune cells in the high- and low-risk groups. (D) The violin chart shows the TIDE scores of LUAD patients in the high- and low-risk groups. (E-H) The box plot shows the expression differences of TMB and immune checkpoint molecules in the high- and low-risk groups. (I) The responsiveness of EGFR mutant LUAD patients in the high- and low-risk groups. **, P<0.01; ***, P <0.001. EGFR, epidermal growth factor receptor; IC₅₀, half-maximal inhibitory concentration; LUAD, lung adenocarcinoma; TIDE, Tumor Immune Dysfunction and Exclusion; TMB, tumor mutation burden.

The value of four hub genes in predicting OS in EGFR mutant LUAD patients

To confirm the reliability of gene signature, we analyzed the expression and prognostic value of these four hub genes in the public database. The result showed that compared with normal lung tissues, PSMD11 and G3BP1 had significantly higher mRNA expression in LUAD tissues (Figure S3A-S3D, P<0.001). And Figure S3E-S3H showed that there were differences in the expression of these four genes in the prognosis of EGFR mutant LUAD patients. We also analyzed the protein levels of PSMD11 and G3BP1 in normal lung and LUAD tissues using HPA database, and the result was consistent with the mRNA result (Figure S3I). Based on the contribution of PSMD11 (P<0.001; HR =1.17) and G3BP1 (P=0.06; HR =1.09) in the forest plot (Figure 2E), we selected PSMD11 for analysis. Further, we detected the expression of PSMD11 in the 160-spot LUAD tissue microarray, and the results showed that the expression of PSMD11 was positively correlated with the clinical stage (Figure S3J-S3L, Table S3, table available at https://cdn.amegroups.cn/static/public/tlcr-2025-355-3.xlsx), indicating that high expression of PSMD11 could promote the progression of LUAD.

Next, we evaluated the relationship between the expression of PSMD11 and the proportion of 22 infiltrating immune cells and found that PSMD11 was positively correlated with M0/M1 macrophages (Figure S4A,S4B), while negatively correlated with monocytes, memory B cells, resting dendritic cells, and resting mast cells (Figure S4C-S4F, Table S4). In addition, there was a positive correlation between PSMD11 and the expression of immune-related molecules HLA-A, HLA-B, and PD-1 (Figure S4G-S4I).

Knockdown of PSMD11 could inhibit cell proliferation and increase the sensitivity of drug-resistant cells to OSI in vitro and in vivo

In order to further explore the role of these genes in regulating LUAD cell function, we selected PSMD11 to knockdown and conducted cell function experiments. Firstly, we detected the protein expression of PSMD11 in OSI-sensitive cells (PC9, H1975) and OSI-resistant cells (PC9OR, H1975OR), and the results showed no significant difference (Figure S5A). Then, we treated four types of LUAD cells with different concentrations of OSI to detect cell proliferation. We found that PC9 and H1975 significantly reduced cell clones and increased cell apoptosis with 1 µM OSI treatment (Figure S5B,S5C); PC9OR and H1975OR showed a significant decrease in cell clones with 4 µM OSI treatment (Figure S5B).

Then, we knocked down PSMD11 in PC9OR and H1975OR cells and selected PSMD11-siRNA2 and PSMD11-siRNA3 for subsequent experiments (Figure 4A). By detecting cell proliferation through CCK-8 and colony formation assay, we found that after knockdown of PSMD11 in PC9OR and H1975OR, cell proliferation significantly slowed down and cell clones significantly decreased (Figure 4B-4D, P<0.05). By detecting cell apoptosis using flow cytometry and western blot, we found that knockdown of PSMD11 significantly increased cell apoptosis and the expression of cleaved-PARP1 and cleaved-caspase3 in PC9OR and H1975OR cells compared to the NC-siRNA groups (Figure 4E,4F, Figure S6A-S6E, P<0.05). It could be seen that knockdown of PSMD11 can inhibit cell proliferation and promote cell apoptosis in the OSI-resistant cells. Furthermore, we detected whether knocking down PSMD11 enhances the sensitivity of drug-resistant cells to OSI. The result showed that knocking down PSMD11 in PC9OR and H1975OR cells significantly reduced the IC₅₀ of OSI (PC9OR: 4,076 nM, PC9OR-siNC: 3,915 nM, PC9OR-siPSMD3: 899.2 nM, H1975OR: 4,921 nM, H1975OR-siNC: 4,102 nM, H1975OR-siPSMD3: 915.2 nM) (Figure 4G,4H). Combining the IC₅₀ of OSI in PC9 (579 nM) and H1975 (313 nM) (Figure 2A,2B), we concluded that knocking down PSMD11 significantly increased the sensitivity of drug-resistant cells to OSI and reversed drug resistance.

Figure 4 Knockdown of PSMD11 could inhibit cell proliferation and increases the sensitivity of drug-resistant cells to OSI. (A) Western blot was used to detect the knockdown efficiency of PSMD11 in PC9OR and H1975OR cells. (B,C) CCK-8 analysis detected the effect of knocking down PSMD11 on cell proliferation in PC9OR and H1975OR cells. (D) The effect of knocking down PSMD11 on cell colony in PC9OR and H1975OR cells was detected through colony formation assay (crystal violet staining). (E,F) The effect of knocking down PSMD11 on apoptosis and the expression of apoptosis-related molecules in PC9OR and H1975OR cells was detected by flow cytometry and western blot analysis. (G,H) CCK-8 analysis detected the IC₅₀ of OSI after knocking down PSMD11 in PC9OR and H1975OR cells to observe whether it can reverse drug resistance. (I) CCK-8 was used to detect the effect of treating PC9OR and H1975OR cells with PSMD11-siRNA and OSI individually or in combination on cell proliferation ability. (J) Flow cytometry was used to detect the effects of PSMD11-siRNA2 and OSI treatment on apoptosis in PC9OR and H1975OR cells. (K-M) Western blotting was used to detect the expression of apoptosis-related molecules in PC9OR and H1975OR cells after treatment with PSMD11-siRNA2 and OSI, and a gray-scale analysis graph was drawn. *, P<0.05 compared with control group; ***, P<0.001 compared with control group; ^##, P<0.01 compared with OSI group; ^△△, P<0.01 compared with PSMD11-siRNA group. CCK-8, Cell Counting Kit-8; FITC, fluorescein isothiocyanate; IC₅₀, half-maximal inhibitory concentration; NC, negative control; OSI, osimertinib; PE, phycoerythrin; PI, propidium iodide; siRNA, small interfering RNA.

Next, we detected the cell proliferation and apoptosis ability of PC9OR and H1975OR cells treated with either alone or combined PSMD11-siRNA and OSI (IC₅₀ of PC9OR and H1975OR). The results showed that compared to the alone treatments with PSMD11-siRNA or OSI, the combined treatment significantly slowed down the cell proliferation, reduced cell clones, increased cell apoptosis, and the expression of cleaved-PARP1 and cleaved-caspase3 (Figure 4I-4M, Figure S6F-S6I, all P<0.05).

In order to further validate the effect of PSMD11 on the sensitivity of drug-resistant cells to OSI, PC9OR, and H1975OR cells transfected with NC-siRNA or PSMD11-siRNA were injected into BALB/c nude mice to establish a xenograft mice model. When the average tumor volume reached 100 mm³, OSI was orally administered 5 mg/kg twice a week, and then the tumor volume and body weight were monitored every three days. There was no significant difference in body weight between the groups (P>0.05). However, tumor volumes were significantly lower in the PSMD11-siRNA group and the OSI + PSMD11-siRNA group than in mice treated with NC-siRNA or OSI alone (Figure 5A-5F, P<0.01). Immunohistochemistry of the tumor tissues revealed that knockdown of PSMD11 alone or co-treated with OSI significantly decreased the expression levels of Ki-67 compared with the NC-siRNA group (Figure 5G-5I, P<0.05). And knockdown of PSMD11 alone or co-treated with OSI significantly increased the levels of cleaved-caspase3 and TUNEL and the protein expression of cleaved-PARP1 and cleaved-caspase3 compared with the NC-siRNA group in tumor samples of PC9OR and H1975OR cell xenograft models after surgery (Figure 5J,5K).

Figure 5 Knockdown of PSMD11 could enhance the antitumor effects of OSI in vivo. (A-C) Subcutaneous transplantation tumor pictures, tumor growth curves, and body weight plot after subcutaneous injection with control or PSMD11 knockdown PC9OR cells (n=5). (D-F) Subcutaneous transplantation tumor pictures, tumor growth curves, and body weight plot after subcutaneous injection with control or PSMD11 knockdown H1975OR cells (n=5). (G-I) Immunohistochemistry showed representative pictures of Ki-67 in different treatment groups of PC9OR and H1975OR cells, and the positive expression rate was displayed using a bar plot. (J) Immunohistochemistry showed representative images of cleaved-caspase3 and TUNEL in different treatment groups of PC9OR and H1975OR cells. (K) The protein expression of apoptosis-related molecules in tumor samples of PC9OR and H1975OR cell xenograft models after surgery in different treatment groups was detected by western blotting. *, P<0.05 compared with control group; **, P<0.01 compared with control group; ***, P<0.001 compared with control group; ^##, P<0.01 compared with OSI group; ^###, P<0.001 compared with OSI group; ^△△, P<0.01 compared with PSMD11-siRNA group. NC, negative control; OSI, osimertinib; siRNA, small interfering RNA; TUNEL, terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling.

PSMD11 could promote the progression of OSI-resistant LUAD cells by activating the NF-κB/IL-6/STAT3 signaling pathway

In order to explore the potential mechanism of PSMD11 regulating the function of OSI-resistant LUAD cells, we grouped the EGFR mutant LUAD patients based on the median expression of PSMD11, and GSEA was performed. We found that overexpression of PSMD11 in OSI-resistant cells was related to the activation of NF-κB signaling pathway (Figure 6A). Then, we detected the expression of NF-κB and pNF-κB in the nuclear, cytoplasmic, and total proteins of OSI-sensitive and resistant LUAD cells and tissues. The results showed that compared with OSI-sensitive LUAD cells and tissues, the expression of NF-κB and pNF-κB was higher in nuclear of OSI-resistant cells and tissues (Figure 6B,6C). According to existing reports, the activation of NF-κB can upregulate the expression of IL-6, TNFα, and IL-17, thereby promoting cell proliferation and inhibiting cell apoptosis (25). Therefore, we evaluated the autonomous secretion of IL-6, TNFα, and IL-17 at the protein level through ELISA. In the presence or absence of OSI, the production of TNFα and IL-17 in four types of cells was very low, and there was no statistical difference between the groups (Figure 6D,6E). Regardless of the presence or absence of OSI, IL-6 was significantly reduced in PC9 and H1975 cells, while IL-6 was significantly increased in PC9OR and H1975OR cells (Figure 6F, P<0.001). Meanwhile, we also detected the content of IL-6 in PC9OR and H1975OR cells after knocking down PSDM11. The results showed that the content of IL-6 was significantly reduced after knocking down (Figure 6G, P<0.001). The above results indicate that the upregulation of NF-κB in OSI-resistant LUAD cells could increase the expression of IL-6, thereby promoting cell proliferation and inhibiting cell apoptosis. Furthermore, we knocked down PSMD11 and NF-κB, respectively, in PC9OR and H1975OR cells, and found that the expression of NF-κB, pNF-κB, and pSTAT3 in the NF-κB-siRNA group was significantly reduced compared to the NC-siRNA group (Figure 6H), and the expression of pNF-κB and pSTAT3 in the PSMD11-siRNA2 group was significantly reduced compared to the NC-siRNA2 group (Figure 6I). This result indicated that knockdown of PSMD11 could impede the progression of OSI-resistant LUAD cells by inhibiting the activation of the NF-κB/IL-6/STAT3 signaling pathway.

Figure 6 PSMD11 could promote the progression of OSI-resistant LUAD by activating the NF-κB/IL-6/STAT3 signaling pathway. (A) According to the median expression of PSMD11, EGFR mutant LUAD patients were divided into high and low groups for GSEA. (B) Detect the protein expression of NF-κB and pNF-κB in the nuclear, cytoplasmic, and total proteins of four types of cells through western blot. (C) Detect the protein expression of NF-κB and pNF-κB in the nuclear and cytoplasmic of tumor cells from OSI-sensitive and OSI-resistant LUAD patients. (D-F) ELISA was used to detect the concentrations of TNFα, IL-17, and IL-6 in the four types of cells treated with 1 µM OSI. (G) After knocking down PSDM11 in PC9OR and H1975OR cells, the content of IL-6 was detected by ELISA. (H) Knockdown NF-κB in PC9OR and H1975OR cells to detect protein expression of STAT3 and pSTAT3. (I) After knocking down PSDM11 in PC9OR and H1975OR cells, the expression of NF-κB, pNF-κB, STAT3, and pSTAT3 were detected by western blot. ***, P<0.001 compared with control group. C, cytoplasmic; EGFR, epidermal growth factor receptor; ELISA, enzyme-linked immunosorbent assay; GSEA, gene set enrichment analysis; KEGG, Kyoto Encyclopedia of Genes and Genomes; LUAD, lung adenocarcinoma; N, nuclear; OSI, osimertinib; siRNA, small interfering RNA; T, total proteins.

Discussion

EGFR-TKIs are the first-line treatment for EGFR mutant LUAD patients. However, almost all the EGFR mutant LUAD patients who initially responded to EGFR-TKIs will continue to develop drug resistance. It can be seen that drug resistance-related genes play an important role in tumor occurrence and progression. Previous studies have shown that a large number of gene signatures perform well in predicting the prognosis or treatment response to LUAD (26,27). However, there is limited research on EGFR-TKIs resistance-related gene signature as prognostic and predictive biomarkers. Therefore, developing EGFR-TKIs resistance-related gene signature can not only predict outcome well, but also help screen EGFR mutant LUAD patients with a high risk of TKI resistance, which may benefit from other treatments.

In this study, based on sequencing data of PC9, PC9OR, H1975, and H1975OR, we identified a total of 948 OSI resistance-related DEGs, which were mainly enriched in several GO terms related to tumor immunity, such as leukocyte proliferation, CD4⁺ T cell activation, cytokine regulation, and myeloid activation. This result indicates a close correlation between the production of drug-resistant cells and tumor immunity. Next, using these DEGs, EGFR mutant LUAD prognostic-related genes and immune gene sets to intersect, and analyzing through univariate and multivariate Cox regression, we constructed an OSI resistance-related immune prognosis prediction model consisting of four hub genes (C3, PSMD11, G3BP1, and TRIB2). Subsequently, based on the median of risk score, EGFR mutant LUAD patients were divided into high- and low-risk groups. We found that the high-risk group had more deaths and poorer prognosis. And the risk score was an independent prognostic indicator for EGFR mutant LUAD patients. Finally, in order to further confirm the accuracy of risk score, we constructed a nomogram and confirmed that this gene signature has good prognostic value.

This gene signature was established based on C3, PSMD11, G3BP1, and TRIB2. Here, we determined that PSMD11 and G3BP1 were highly expressed in LUAD tissues and were associated with poor prognosis in EGFR mutant LUAD patients. Meanwhile, due to the greater contribution of PSMD11, we selected PSMD11 for further analysis. We found that the expression of PSMD11 was positively correlated with the proportion of M0/M1 macrophages and the expression of immune-related molecules HLA-A, HLA-B, and PD-1. PSMD11 protein is a component of 26S proteasome. 26S proteasome is a multi-protein complex involved in ATP-dependent ubiquitination protein degradation (28). It plays a key role in maintaining protein homeostasis and participates in many cell processes, including cell cycle process, apoptosis, and DNA damage repair (29). In the complex, the high expression or phosphorylation of PSMD11 can promote the assembly and activity of 26S proteasome (30,31). In addition, PSMD11 has also been proven to inhibit apoptosis of pancreatic cancer and liver cancer cells (32,33), but its role in OSI-resistant LUAD and its clinical significance remain to be clarified. Therefore, we conducted in vitro and in vivo experiments to verify the role of PSMD11 in OSI-resistant LUAD, and found that the downregulation of PSMD11 significantly inhibited the proliferation of OSI-resistant cells and the growth of subcutaneous tumors, and promoted the apoptosis of tumor cells. These experimental results further confirmed the reliability of our analysis.

To explore the potential biological functions of these four gene signatures, we conducted GSEA. The results showed that our gene signature was involved in cell cycle, DNA replication, and proteasome. In addition, some tumor immune-related signaling pathways, such as chemokine signaling pathway, were also highly correlated with our signature. These results indicate that our gene signature may have played a crucial role in reflecting TIME status. TIME induces TKI resistance by promoting epithelial-mesenchymal transformation, cancer stem cell generation, initiating anti-apoptotic signaling pathways, and mediating immunosuppression (34). A recent study has shown that TKI therapy can reshape the immune microenvironment (35), but there is still confusion about how TKI resistance affects TIME. Considering the important role of infiltrating immune cells in tumor progression, we first conducted a feature study on the immune cell infiltration landscape of EGFR mutant LUAD patients based on high- and low-risk groups. We found that the proportion of activated dendritic cells and M0/M1 macrophages was higher in the high-risk group, while the proportion of resting CD4⁺ memory T cells, resting dendritic cells and resting mast cells was higher in the low-risk group. These results indicate that the high-risk group of EGFR mutant LUAD patients has more activated infiltrating immune cells, which could help patients benefit from immunotherapy.

The increased expression of immune checkpoint molecules such as PD-1, PD-L1, LAG3, IDO1, and CTLA4 is the main reason why tumor cells evade host immune surveillance (36). Then, we analyzed the correlation between gene signature and immune checkpoint molecules, and the results showed that compared to the low-risk group, the high-risk group had more expression of IDO1, LAG3, and PD-L1. These results suggest that due to the presence of a large number of infiltrating immune cells and upregulation of immune checkpoint molecules in the high-risk group, patients in the high-risk group can benefit from immunotherapy. In the past few decades, ICIs have become a promoting option for most cancer treatments, including LUAD. The increased expression of immune checkpoint molecules may be related to ICIs response (37). Therefore, the correlation between the model and immune checkpoint molecules may help screen patients who will benefit from the treatment of corresponding ICIs. In addition, chemotherapy and targeted therapy are common treatment methods for LUAD patients. Therefore, we evaluated the sensitivity of EGFR mutant LUAD patients to common chemotherapy drugs and targeted drugs, and the results showed that patients in the high-risk group were more sensitive to eight drugs of these two treatment methods. The above results indicate that the gene signature is of great significance for the clinical treatment of EGFR mutant LUAD patients.

Conclusions

Our study constructed an immune prognosis model based on four OSI resistance-related genes, which can accurately predict the prognosis of EGFR mutant LUAD patients. The correlation between this model and TIME indicates that it has good application value in predicting the efficacy of immunotherapy. At the same time, this model can also predict the responsiveness of EGFR-mutated LUAD patients to conventional chemotherapy and targeted drugs. It can be seen that in the current era of personalized medicine, this prediction model has a good guiding role in clinical practice.

Acknowledgments

The authors acknowledge TCGA and GEO database for providing their platforms and contributors for uploading their meaningful datasets.

Footnote

Reporting Checklist: The authors have completed the ARRIVE and MDAR reporting checklists. Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-355/rc

Data Sharing Statement: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-355/dss

Peer Review File: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-355/prf

Funding: This work was funded by the 75th Post-Doctoral Science Foundation Project (No. 49536) and the 2024 Natural Science Foundation for Young Scientists of Shanxi Province (No. 82403540).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-355/coif). The authors have no 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 approved by the Ethics Committee of Tangdu Hospital, The Fourth Military Medical University (No. GKJ-Y-202303-181). Written informed consent was obtained from the individual for the publication of any potentially identifiable data included in this article. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. All animal experiments were approved by the Animal Ethics Committee of Tangdu Hospital, The Fourth Military Medical University (No. 20240132), and the investigation complied with the National Institutes of Health Guidelines for the care and use of laboratory animals.

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. Sung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021;71:209-49. [Crossref] [PubMed]
2. Rotow J, Bivona TG. Understanding and targeting resistance mechanisms in NSCLC. Nat Rev Cancer 2017;17:637-58. [Crossref] [PubMed]
3. Chan BA, Hughes BG. Targeted therapy for non-small cell lung cancer: current standards and the promise of the future. Transl Lung Cancer Res 2015;4:36-54. [Crossref] [PubMed]
4. Nussinov R, Tsai CJ, Jang H. Anticancer drug resistance: An update and perspective. Drug Resist Updat 2021;59:100796. [Crossref] [PubMed]
5. Saraon P, Snider J, Kalaidzidis Y, et al. A drug discovery platform to identify compounds that inhibit EGFR triple mutants. Nat Chem Biol 2020;16:577-86. [Crossref] [PubMed]
6. Midha A, Dearden S, McCormack R. EGFR mutation incidence in non-small-cell lung cancer of adenocarcinoma histology: a systematic review and global map by ethnicity (mutMapII). Am J Cancer Res 2015;5:2892-911.
7. Passaro A, Jänne PA, Mok T, et al. Overcoming therapy resistance in EGFR-mutant lung cancer. Nat Cancer 2021;2:377-91. [Crossref] [PubMed]
8. Gao J, Li HR, Jin C, et al. Strategies to overcome acquired resistance to EGFR TKI in the treatment of non-small cell lung cancer. Clin Transl Oncol 2019;21:1287-301. [Crossref] [PubMed]
9. Nilsson MB, Sun H, Robichaux J, et al. A YAP/FOXM1 axis mediates EMT-associated EGFR inhibitor resistance and increased expression of spindle assembly checkpoint components. Sci Transl Med 2020;12:eaaz4589. [Crossref] [PubMed]
10. Reita D, Pabst L, Pencreach E, et al. Molecular Mechanism of EGFR-TKI Resistance in EGFR-Mutated Non-Small Cell Lung Cancer: Application to Biological Diagnostic and Monitoring. Cancers (Basel) 2021;13:4926. [Crossref] [PubMed]
11. Oxnard GR, Hu Y, Mileham KF, et al. Assessment of Resistance Mechanisms and Clinical Implications in Patients With EGFR T790M-Positive Lung Cancer and Acquired Resistance to Osimertinib. JAMA Oncol 2018;4:1527-34. [Crossref] [PubMed]
12. Nigro A, Ricciardi L, Salvato I, et al. Enhanced Expression of CD47 Is Associated With Off-Target Resistance to Tyrosine Kinase Inhibitor Gefitinib in NSCLC. Front Immunol 2019;10:3135. [Crossref] [PubMed]
13. Huang J, Lan X, Wang T, et al. Targeting the IL-1β/EHD1/TUBB3 axis overcomes resistance to EGFR-TKI in NSCLC. Oncogene 2020;39:1739-55. [Crossref] [PubMed]
14. Yu HA, Sima CS, Huang J, et al. Local therapy with continued EGFR tyrosine kinase inhibitor therapy as a treatment strategy in EGFR-mutant advanced lung cancers that have developed acquired resistance to EGFR tyrosine kinase inhibitors. J Thorac Oncol 2013;8:346-51. [Crossref] [PubMed]
15. Hinshaw DC, Shevde LA. The Tumor Microenvironment Innately Modulates Cancer Progression. Cancer Res 2019;79:4557-66. [Crossref] [PubMed]
16. Binnewies M, Roberts EW, Kersten K, et al. Understanding the tumor immune microenvironment (TIME) for effective therapy. Nat Med 2018;24:541-50. [Crossref] [PubMed]
17. Rowshanravan B, Halliday N, Sansom DM. CTLA-4: a moving target in immunotherapy. Blood 2018;131:58-67. [Crossref] [PubMed]
18. Han Y, Liu D, Li L. PD-1/PD-L1 pathway: current researches in cancer. Am J Cancer Res 2020;10:727-42.
19. Isomoto K, Haratani K, Hayashi H, et al. Impact of EGFR-TKI Treatment on the Tumor Immune Microenvironment in EGFR Mutation-Positive Non-Small Cell Lung Cancer. Clin Cancer Res 2020;26:2037-46. [Crossref] [PubMed]
20. Peng S, Wang R, Zhang X, et al. EGFR-TKI resistance promotes immune escape in lung cancer via increased PD-L1 expression. Mol Cancer 2019;18:165. [Crossref] [PubMed]
21. Yin W, Yu X, Kang X, et al. Remodeling Tumor-Associated Macrophages and Neovascularization Overcomes EGFR(T790M) -Associated Drug Resistance by PD-L1 Nanobody-Mediated Codelivery. Small 2018;14:e1802372. [Crossref] [PubMed]
22. Newman AM, Liu CL, Green MR, et al. Robust enumeration of cell subsets from tissue expression profiles. Nat Methods 2015;12:453-7. [Crossref] [PubMed]
23. Bencomo-Alvarez AE, Rubio AJ, Olivas IM, et al. Proteasome 26S subunit, non-ATPases 1 (PSMD1) and 3 (PSMD3), play an oncogenic role in chronic myeloid leukemia by stabilizing nuclear factor-kappa B. Oncogene 2021;40:2697-710. [Crossref] [PubMed]
24. Jiang P, Gu S, Pan D, et al. Signatures of T cell dysfunction and exclusion predict cancer immunotherapy response. Nat Med 2018;24:1550-8. [Crossref] [PubMed]
25. Dolcet X, Llobet D, Pallares J, et al. NF-kB in development and progression of human cancer. Virchows Arch 2005;446:475-82. [Crossref] [PubMed]
26. Zhang C, Zhang Z, Sun N, et al. Identification of a costimulatory molecule-based signature for predicting prognosis risk and immunotherapy response in patients with lung adenocarcinoma. Oncoimmunology 2020;9:1824641. [Crossref] [PubMed]
27. Zhang C, Zhang G, Sun N, et al. Comprehensive molecular analyses of a TNF family-based signature with regard to prognosis, immune features, and biomarkers for immunotherapy in lung adenocarcinoma. EBioMedicine 2020;59:102959. [Crossref] [PubMed]
28. Sahu I, Glickman MH. Proteasome in action: substrate degradation by the 26S proteasome. Biochem Soc Trans 2021;49:629-44. [Crossref] [PubMed]
29. Bard JAM, Goodall EA, Greene ER, et al. Structure and Function of the 26S Proteasome. Annu Rev Biochem 2018;87:697-724. [Crossref] [PubMed]
30. Vilchez D, Boyer L, Morantte I, et al. Increased proteasome activity in human embryonic stem cells is regulated by PSMD11. Nature 2012;489:304-8. [Crossref] [PubMed]
31. Lokireddy S, Kukushkin NV, Goldberg AL. cAMP-induced phosphorylation of 26S proteasomes on Rpn6/PSMD11 enhances their activity and the degradation of misfolded proteins. Proc Natl Acad Sci U S A 2015;112:E7176-85. [Crossref] [PubMed]
32. Sahni S, Krisp C, Molloy MP, et al. PSMD11, PTPRM and PTPRB as novel biomarkers of pancreatic cancer progression. Biochim Biophys Acta Gen Subj 2020;1864:129682. [Crossref] [PubMed]
33. Zhang C, Xu T, Ji K, et al. An integrative analysis reveals the prognostic value and potential functions of PSMD11 in hepatocellular carcinoma. Mol Carcinog 2023;62:1355-68. [Crossref] [PubMed]
34. Westover D, Zugazagoitia J, Cho BC, et al. Mechanisms of acquired resistance to first- and second-generation EGFR tyrosine kinase inhibitors. Ann Oncol 2018;29:i10-9. [Crossref] [PubMed]
35. Fang Y, Wang Y, Zeng D, et al. Comprehensive analyses reveal TKI-induced remodeling of the tumor immune microenvironment in EGFR/ALK-positive non-small-cell lung cancer. Oncoimmunology 2021;10:1951019. [Crossref] [PubMed]
36. Schaller J, Agudo J. Metastatic Colonization: Escaping Immune Surveillance. Cancers (Basel) 2020;12:3385. [Crossref] [PubMed]
37. Bagchi S, Yuan R, Engleman EG. Immune Checkpoint Inhibitors for the Treatment of Cancer: Clinical Impact and Mechanisms of Response and Resistance. Annu Rev Pathol 2021;16:223-49. [Crossref] [PubMed]