Synergistic camrelizumab therapy inhibits P53-mutant osimertinib-resistant solid lung adenocarcinoma via the TGF-β signaling pathway

Introduction

Lung cancer is the leading cause of cancer-related deaths worldwide. Lung adenocarcinoma (LAC), one of the histological types of non-small cell lung cancer (NSCLC) can be further divided into papillary, micropapillary, acinar, and solid types based on their morphological and biological features (1). The diversity in the morphology of LAC, along with the significant variability of molecular markers and the inconsistent responses to targeted therapies that have been observed, suggests that there is significant tumor heterogeneity among LAC subtypes (2).

During the progression of primary tumors, PANoptosis is a notable type of programmed cell death (PCD) that involves the combined release of biochemical signals from three PCD pathways: necroptosis, pyroptosis, and anoikis. Necroptosis and pyroptosis have two key functions in cancer biology. The first is that they can activate cytotoxic T cells to suppress tumor proliferation by enhancing dendritic cell activity. The second is that they can promote angiogenesis, cytokine production and genomic instability (3-6). Anoikis is triggered when cells detach from the cellular or extracellular matrix (ECM), playing a vital role in maintaining cellular homeostasis and tissue development (7).

Cancer-associated fibroblasts (CAFs) are the predominant type of stromal cells, and their increased presence is correlated with a poorer prognosis. These cells exhibit significant heterogeneity in their phenotypic characteristics, origins, and roles. CAFs play a pivotal role in orchestrating the inflammation that promotes tumor growth and in shaping the immune microenvironment toward a state of immunosuppression. These activities are facilitated through complex reciprocal signaling interactions with cancer cells, components of the ECM, and infiltrating immune cells (8). These factors interact within the tumor microenvironment (TME) and contribute to the dynamic changes during tumor progression.

Epidermal growth factor receptor-tyrosine kinase inhibitors (EGFR-TKIs), despite their ability to exert prolonged disease control and the high tumor response rates to their use, eventually cause disease progression in all patients with most developing resistance within 6 to 13 months, most commonly through the T790M mutation (9,10). Osimertinib, a third-generation EGFR-TKI, effectively inhibits T790M, but resistance inevitably emerges after treatment (11).

To address this clinical dilemma, clinicians have attempted to use immune checkpoint inhibitors (ICIs) for the patients resistant to EGFR-TKI therapy. Findings from the phase II trial of KEYNOTE-001 and the CheckMate 012 study indicated that pembrolizumab as a standalone therapy is not a viable first-line option for TKI-naïve NSCLC patients harboring EGFR mutations (12,13). However, another trial showed that the median overall survival (mOS) for patients with EGFR mutations and programmed cell death ligand 1 (PD-L1) tumor proportion score (TPS) ≥25% (N=66) was 16.1 months, which is better than the 10.9 months observed in EGFR wild-type patients with the same PD-L1 TPS (14). Similarly, the TATTON clinical trial reported overall response rates of 43% in EGFR-TKI-pretreated patients and 70% in EGFR-TKI-naive patients treated with both durvalumab and osimertinib (15). The mechanisms underlying the poor response to immunotherapy in patients with EGFR-mutant LACs remain unclear and contradictory.

The transforming growth factor-β (TGF-β) signaling pathway is known to have diverse effects on cell proliferation, differentiation, adhesion, senescence, and apoptosis. TGF-β plays a dual role in the TME, depending on the stages of tumor progression and genetic alterations. TGF-β regulates the number and function of regulatory T cells (Treg cells) to enhance immune tolerance and facilitate tumor escape (16). In cell lines resistant to EGFR-TKI therapy, TGF-β2 mediates genome and metabolic reprogramming via autocrine signaling, thereby promoting the survival of adaptive EGFR-TKI-resistant tumor cells (17). The mechanisms underlying the dynamic activation of the TGF-β pathway in EGFR-mutant LACs should be explored in the context of EGFR targeting and ICI therapy. We present this article in accordance with the MDAR and ARRIVE reporting checklists (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-522/rc).

Methods

Data sources and preprocessing for public data and native cohorts

For this study, transcriptomic expression data from LAC patients in The Cancer Genome Atlas (TCGA) and from the Gene Expression Omnibus (GEO, GSE68465) were used to analyze the expression of PANoptosis-associated genes (PAGs) and CAFs. The data from the TCGA cohort were transformed to the transcripts per kilobase per million (TPM) from fragments per kilobase per million (FPKM) profiles.

This study investigated the PD-L1 expression and other clinicopathological features in histological subsets of LACs. A total of 187 LAC patients with fresh tumor tissue from Shanghai Jiao Tong University Affiliated General Hospital, China Emergency General Hospital, and Beijing Chest Hospital were enrolled.

Additionally, twelve patients with EGFR-mutant LAC received camrelizumab treatment after developing resistance to osimertinib, and their clinical outcomes were monitored through radiological investigation.

Identification of differentially expressed PAGs in LACs from the TCGA and GSE68465 cohorts

A panel of 747 PAGs and cancer-associated fibroblast-related genes (CRGs) was listed in table online https://cdn.amegroups.cn/static/public/tlcr-2025-522-1.xlsx. The Wilcoxon test in R (version 4.1.2) was used to analyze the differential expression of PAGs between LAC and normal tissues. Significance was determined statistically as a false discovery rate (FDR) <0.05 and absolute |log₂FC| >1. Spearman’s correlation analysis was used to determine the relationships between the PAGs and CRGs. The interactions of PANoptosis- and CAF-associated proteins were obtained from the interacting genes/proteins database (https://string-db.org/). To determine significance, a protein-protein interaction (PPI) network was established, with a combined score greater than 0.7 considered to indicate statistical significance. To extract hub modules and genes, the “cytoHubba” plugin in Cytoscape 3.10.1 was used. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were performed to annotate the PAGs and CRGs using the “ClusterProfiler” R package (v3.0.4). A P value <0.05 was considered to indicate a statistically significant difference, and the results were visualized using the “ggplot2” R package.

Patients from a native cohort with heterogeneous PD-L1 expression in EGFR-mutant LACs

The TCGA and GEO databases include extensive samples of LAC patients, and as a result are an important resource for analyzing the fundamental molecular pathways and mechanisms, including tumor cell death and the influence of CAFs on TME remodeling. However, these studies overlook the molecular characteristics of the different histological subtypes of LACs. Therefore, a native cohort of 187 LAC patients diagnosed according to the 5^th edition of the WHO classification of thoracic tumors was included in this study. Clinical stages were assessed using the 7th edition of the AJCC staging system. The statuses of the EGFR, anaplastic lymphoma kinase (ALK), ROS proto-oncogene 1 (ROS1) and Kirsten rat sarcoma viral oncogene homolog (KRAS) genes were determined using the amplified refractory mutation system (ARMS) method. Total DNA extraction was performed using the AmoyDx FFPE DNA/RNA kit (Spin Column, ADx-FF03; Amoy Diagnostics, Xiamen, China). The statuses of the EGFR, ALK, ROS1, and KRAS genes were detected using the following kits: the AmoyDx Adx-ARMS EGFR Mutation Kit (Cat. No. 8.0120201W012) for EGFR, the AmoyDx Adx-ARMS Kit (Cat. No. 8.0124401W008) for ALK and ROS1 fusion genes, and the AmoyDx Adx-ARMS Kit (Cat. No. 8.0120102W006) for KRAS.

Droplet digital polymerase chain reaction (ddPCR) testing for EGFR mutations in laser-captured microdissected tissues

The Leica Microsystems LMD 7000 microdissection system (Wetzlar, Germany) was used to isolate pure cell subpopulations from specific regions of eleven EGFR-mutant samples. At least 200 cell equivalents, or approximately 1,000 visible cells, were isolated and collected in a reaction tube cap containing TRIzol Reagent (Invitrogen, Germany). ddPCR was performed to detect in-frame deletions in Exon 19 and the L858R mutation in the EGFR gene using the QX-200^TM ddPCR system (Bio-Rad, Hercules, CA, USA), following the manufacturer’s instructions, alongside PrimePCR™ ddPCR™ Mutation Detection Assay Kits (#1863105 for p.E746_A750del, #1863104 for p.L858R, and #186310 for p.G719S). A 20 µl ddPCR reaction system was loaded into an 8-channel droplet generation cartridge (Bio-Rad, Milan, Italy). Next, 70 µL of QX200 Droplet Generation Oil (Bio-Rad, Milan, Italy) was emulsified. The emulsified droplets were subsequently transferred to a 96-well plate and amplified them using standard PCR on a Mastercycler (Eppendorf). The cycling conditions were as follows: denaturation at 95 ℃ for 5 minutes, followed by 40 cycles of 95 ℃ for 30 seconds and then finally 60 ℃ for 1 minute.

Assessment of the TME: CD4+ and CD8+ T cells infiltration and PD-L1, COL1A1 and P53 protein expression determined utilizing immunohistochemistry

The presence of infiltrating CD4⁺ and CD8⁺ T-cells in the native cohort was analyzed using immunohistochemistry (IHC), and digital images were captured using an Axio Scan Z1 with Zen Blue software (Zeiss, Germany). The images were then processed using Image Pro-Plus 6.0 software. The absolute numbers of CD4⁺ and CD8⁺ T cells per mm² in the stromal and parenchymal compartments were anonymized and independently counted by two experienced pathologists, Y.C. and X.W., using light microscopy. Additionally, the localization of each cell was assessed. T-lymphocytes found among epithelial carcinoma cells (in the parenchyma) were classified as “intraepithelial” T-cells, while those located within the tumor stroma were classified as “stromal” T-lymphocytes (18). In cases of disagreement, the slides were reexamined, and the mean densities of the lymphocytes were calculated.

The PD-L1 (SP263) rabbit monoclonal primary antibody (Cat. No. 790–4905), along with other ancillary reagents, including Ventana detection kits and a negative antibody (Cat. No. 790–4795), were obtained from Roche Diagnostics GmbH in Mannheim, Germany. The PD-L1 antibody generated staining that was either membranous or cytoplasmic when used on the Ventana BenchMark XT. Two pathologists, Y.C. and X.W., evaluated PD-L1 expression in tumor cells (TCs) using a three-tiered grading system based on the TPS: <1%, 1–49%, and ≥50%. For each lung specimen, we assessed COL1A1 (ab88147, Abcam, UK) staining was assessed using scores derived from the percentage of positive cells and the intensity of staining, which ranged from 0 (colorless) to 3 (dark brown). A diffuse strong expression of the P53 protein was interpreted as indicative of a mutation in the p53 gene.

Quantification of cytokine levels using cytometric bead array

To measure the levels of secreted proteins, we repeatedly flushed laser-captured tissues with a total of 5 mL of PBS, ensuring that the protein concentrations were equalized. We determined cytokine concentrations using a cytometric bead assay from BD Bioscience (San Jose, CA, USA). The levels of interleukin-2 (IL-2), IL-4, IL-10, IL-6, tumor necrosis factor-α (TNF-α) and interferon-γ (IFN-γ) were measured using the cytometric bead assay according to the manufacturer’s protocol. We incubated 50 µL of each supernatant sample in duplicate with antibody-conjugated microbeads for 3 hours. After washing the samples, we quantified the cytokine levels in each sample using a FACSAria II equipped with Cell Quest Pro and CBA software from BD Biosciences.

Mice tumour inoculation and treatment regimen

The TGF-β pathway is known to be associated with PD-L1 expression, which contributes to EGFR-TKI resistance (17). A patient-derived tumor xenograft (PDX) mouse model was used to examine the morphological and molecular alterations in osimertinib-resistant LAC when treated with camrelizumab. The PDX mouse model was constructed with tumor obtained from a patient with unresectable, advanced ex19del-positive LAC. This case was treated with osimertinib and then 10 months later, the tumor metastasized to supraclavicular focus. Tumor tissue was resected from the metastatic focus. Initially, the tumor specimen was incubated in HTK solution (Dr. Franz Koehler Chemie Ltd., Bensheim, Germany), thoroughly washed three times with phosphate-buffered saline and transferred to a culture dish containing RPMI 1640 culture medium (No. 12633020, ThermoFisher Scientific). The tumor tissue was placed inside a cannula and then inoculated subcutaneously on the back of 6- to 8-week-old female NOD.Cg-Prkd^cscid IL2rgtm^1Wjl/SzJ (NSG) mice (Experimental Animal Center at Southern Medical University, Guangzhou, China). These mice were fed with standard feed and water and kept at approximately 25 ℃ and 50% humidity. Parts of tumor tissue were tested for PD-L1 and P53 IHC and PCR for driver gene mutation. No additional mutation was found in the metastasized tumor.

The size of the tumors was measured twice each week using a microcaliper, and the tumor volume was calculated as 0.5 × length × width × height. When the tumor volume reached ≥750 mm³, the mice were euthanized via cervical dislocation, and the tumor was excised and transplanted into Hu-HSC reconstruction NSG mice (human CD34⁺haematopoietic stem cell engrafted NSG mice) (Yiruibei Model Organisms Center, Inc., Guangzhou, China), which included multilineage human immune cells. The immune-humanized PDX model not only recapitulates patient-derived tumors on genomic, molecular, and cellular levels, but it also offers an animal model by imitating the human immune system, enabling the analysis of tumor-immune system interactions and the evaluation of immunotherapy efficacy.

Once tumor volumes reached 250 mm³, the mice were randomly divided into five groups, each consisting of three mice (n=3): the control group received an equal volume of saline; the TGF-β1 group was injected intraperitoneally (i.p.) with 4 mg/kg body weight every 5 days; the camrelizumab (202011004F, Suzhou Shengdiya Biomedical Co., Ltd., Suzhou, China) group, referred to as the ICI group, received tail vein injections of camrelizumab at 100 mg/kg on the first day of each week; the osimertinib group received 25 mg/kg of osimertinib once daily [epidermal growth factor receptor-tyrosine kinase inhibitor treatment (EGFR-TKT)]; and the combination group (EGFR-TKT + ICI) group, received both camrelizumab and osimertinib. The experiment lasted for four weeks. Body weight, clinical assessments were also recorded during the observation period.

The tumor remission rate (%) was calculated using the following formula: [1 − (tumor volume of the experimental group/average tumor volume of the control group)] × 100%. Animals that reached a body condition score of ≤2, a body weight loss of ≥20%, a tumor volume >2,000 mm³, or had ulcerated tumors were euthanized before study’s end. Tumor tissues were collected for flow cytometry following the manufacturer’s protocol for the analysis of CD4⁺CD3⁺ T cells and CD3⁺CD8⁺ cytotoxic T cells on an FACSCalibur (BD Biosciences). Additionally, tumor tissues were excised, fixed using 4% paraformaldehyde, and subjected to immunohistochemistry and hematoxylin-eosin staining.

Western blot analysis of TGF-β signaling, pERK1/2, and pSTAT3 proteins in tumor tissues

Reagents

GAPDH-loading control (bsm-33033M), Goat anti-rabbit IgG H&L/horseradish peroxidase (HRP) (bs-0295G-HRP), goat anti-mouse IgG H&L/HRP (bs-0296G-HRP, Bioss, Beijing, China); anti-TGF β1 antibody (1:200; ab92486, Abcam, UK), anti-TGF β2 antibody [RP23040064] (ab308054, Abcam, Cambridge, UK), anti-TGF βR1 antibody[EPR20923-13] (1:200; ab235578, Abcam, UK), anti-Smad2 antibody[EP784Y] (1:200; ab40855, Abcam, UK), anti-Smad3 antibody[EP568Y] (1:200; ab40854, Abcam, UK), anti-EGFR antibody[EP38Y] (1:200; ab52894, Abcam, UK), anti-p-ERK antibody (1:100; CST #4370, 1:200), and anti-p-STAT3 antibody (1:100; CST, #9145) were used in this study. The HRP-conjugated secondary antibodies included anti-Mouse IgG (CST, #7076) and anti-rabbit IgG (CST, #7074). All antibodies were stored at −20 ℃.

Protein extraction from whole tissue samples and western blot analysis

Tumour tissues were cut into small pieces and placed in 1.5 mL centrifuge tubes. A total of 150 µL of RIPA lysis buffer containing PMSF was added for every 20 mg of tissue. The tubes were then shaken for 20 minutes at 4 ℃. Once lysis was complete, the samples were centrifuged at 12,000 rpm for 15 minutes at 4 ℃. The supernatant was then transferred to a new 1.5 mL centrifuge tube and stored at −80 ℃.

A 10% separating gel and a 5% stacking gel for SDS-PAGE. Pre-cooled 1× electrophoresis buffer and the total intracellular protein extracts, along with a pre-stained protein marker were loaded onto the gel. Electrophoresis was carried out at 80 V for approximately 30 minutes, followed by 120 V until the desired band positions were reached. The polyvinylidene fluoride (PVDF) membrane was activated in methanol for 1 minute, and then soaked in transfer buffer. Filter paper was soaked in transfer buffer for 15 minutes. The transfer membrane “sandwich” was assembled by placing the gel on the negative side and the membrane on the positive side, ensuring that no bubbles were trapped. The membrane was washed with Tris-Buffered Saline with Tween 20 (TBST) for 2 minutes, shaken on a shaker at room temperature for 30 minutes, and then incubated overnight at 4 ℃ with primary antibodies. After washing the membrane three times with TBST for 10 minutes each, secondary antibodies were added and incubated for 2 hours at room temperature. The membrane was then washed three more times with TBST for 10 minutes each.

Clinical outcomes in camrelizumab-treated patients after osimertinib resistance

Drawing from the findings of the previous clinical trials, we designed a study involving patients with LACs who were resistant to osimertinib and voluntarily opted for treatment with ICIs. We screened LAC patients for radiological disease progression after Osimertinib therapy between October 2021 and May 2023. Patients were classified based on histopathological criteria. Twelve patients with T790M mutations and radiological progression in primary lung tumors were eligible for enrollment. Other enrollment criteria included an Eastern Cooperative Oncology Group (ECOG) performance status of 0 or 1, at least one measurable lesion assessed using the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 (19) by the investigator, and an estimated life expectancy of three months or longer. The exclusion criteria included prior chemotherapy for lung cancer, symptomatic central nervous system (CNS) metastases treated with steroids within four weeks before screening, and any other immunosuppressive or active autoimmune diseases requiring systemic treatment within the last 2 years.

Subsequently, all patients received treatment with camrelizumab (200 mg) at Shanghai General Hospital and Chinese Emergency Hospital. Treatments were discontinued in cases of severe or life-threatening treatment-related toxicities. Computer tomography of the thorax, abdomen, and pelvis was performed at baseline, every six weeks for the first 18 weeks, and then every nine weeks for the first 12 months. Response was assessed based on RECIST version 1.1 by independent radiologists, and treatment decisions were made based on the investigators’ review.

Bronchoalveolar lavage and cytokine measurement in patients receiving camrelizumab after osimertinib failure

Flow cytometry is a commonly used and efficient method to analyze cytokines in bronchoalveolar lavage fluid (BALF) (20,21). To prevent biases in cytokine analysis resulting from the limitations of tissue biopsies, we discussed with the patients and obtained their consent to perform BALF examinations, allowing for cytokine collection from the entire tumor instead of only from localized biopsy sites. During the second follow-up, BALFs were collected by injecting 50 mL of saline solution into the trachea to inflate the lungs, followed by aspiration. The BALF supernatants were separated by centrifugation and stored at −80 ℃. The levels of IL-2, IL-4, IL-6, IL-10, IL-17, IFN-γ, and TNF-α in the BALFs were measured using the methodology outlined in the Methods section, with all samples measured in duplicate.

Ethics approval and consent to participate

This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. It was authorized and supervised by the Ethics Committee of Shanghai General Hospital (approval No. 2021020). Informed consent was obtained from all participants included in this study. All animal experiments were conducted in accordance with the Guide for the Care and Use of Experimental Animals from the Experimental Animal Center of Shanghai General Hospital (approval No. 2023SQ067).

Statistical analysis

R software (version 4.3.1) was used for the statistical analysis of the TCGA and GSE68465 datasets. A false discovery rate (FDR) of <0.05 was considered statistically significant. Data were analyzed and presented using GraphPad Prism 9 (version 9.4.0). Numerical data were expressed as means ± standard deviations. Crosstabs were analyzed using the Chi-squared test. Statistical differences between groups were determined using a t-test or one-way analysis of variance, with a P value <0.05 considered significant.

Results

Data from the TCGA and GEO cohorts

This study analyzed the expression of PAGs and CAFs in a total of 944 LAC patients, comprising 501 from TCGA (https://portal.gdc.cancer.gov/) and 443 from GEO (https://www.ncbi.nlm.nih.gov/geo/). In the TCGA cohort, 374 patients (74.7%) were diagnosed at stages I–II. In the GSE68465 cohort, 371 out of 440 patients (84.3%) were also diagnosed at stage I–II. The remaining 25.3% of patients in the TCGA cohort and 15.7% in the GSE68465 cohort were diagnosed at stages III–IV.

Identification of significant expression of CAF-related genes and PAGs in LACs

The heatmap in Figure 1A shows the differences in gene expression between LAC and normal samples from the TCGA and GSE68465 cohorts. In LAC tissues, the expression levels of GREM1, COL1A1, COL3A1, IGKC, CD79A, CR2, and TNFRSF25 were higher. In contrast, the expression levels of KIR3DL1, BANK1, STEAP4, BMP6, CLIC2, MAL, FAM124B, BMX, DCN, FCGR3B, CXCR1, MMRN2, KCNJ15, PALMD, MYCT1, EDN1, MMRN1, VWF, PTPRB, ROBO4, CDH5, CXCR2, ACVRL1, TIE1, TEK, EMCN, and CA4 were lower (Table S1). The protein-protein interaction (PPI) networks were established using the STRING database, requiring an interaction score of at least 0.7 (Figure 1B). Furthermore, the PAGs were analyzed using Cytoscape software, which showed how these genes were upregulated and downregulated during their interactions (Figure 1C).

Figure 1 Differential expression of PAGs and CAFs in LAC as well as hub genes and their functions and pathways via GO and KEGG analyses. (A) Heatmap showing the differential expression of PAGs and CAFs in lung adenocarcinomas compared to normal lung tissues; (B) PPI analysis showing the interactions among these PAGs and CAFs; (C) cytoscape visualization displaying upregulated and downregulated genes, with red indicating overexpression and green indicating low expression levels; (D) GO analysis illustrating the enriched biological functions; (E) KEGG analysis presenting the enriched molecular pathways. CAFs, cancer-associated fibroblasts; ECM, extracellular matrix; GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes; LAC, lung adenocarcinoma; PAGs, PANoptosis-associated genes; PPI, protein-protein interaction; TGF, transforming growth factor.

Gene set enrichment analysis on TCGA and GSE68465 cohorts

The molecular pathways and biological functions linked to the differentially expressed genes (DEGs) were identified using KEGG enrichment and GO analyses. The activated pathways included “cytokine-cytokine receptor interaction”, “B cell receptor signaling”, “AGE-RAGE signaling in diabetic complications”, “TGF-beta signaling”, “Platelet activation”, and “Relaxin signaling”. The GO analysis revealed that the most significantly enriched functions of the DEGs (P<0.05) included “extracellular matrix structural constituent”, “growth factor binding”, “protease binding”, and “cytokine binding” (Figure 1D,1E, Table S2). The pathways included up-regulated genes such as GREM1, COL1A1, COL3A1, CR2, and CD79A (Figure 1C). GREM1 overexpression activates the TGF-β pathway, whereas COL1A1 and COL3A1 overexpression is associated with fibroblast activation.

Characteristics and histopathological features of patients from the native cohort

A native cohort of 187 LAC patients was involved in this study. The demographic data of these patients are summarized in Tables 1,2. The ages of the patients ranged from 23 to 82 years, with a median age of 59 years. There were 90 patients at stage I (48.1%), 21 at stage II (11.2%), and 76 at stage III (40.7%). All patients were diagnosed with LAC and classified based on histological components. In this native cohort, 72 (38.5%), 120 (64.2%), and 90 (48.1%) LACs were found composed of at least one component of solid, acinar, papillary, micropapillary, or papillary components, respectively.

In 94 (50.3%) LACs, EGFR activating mutations were found, including 78 (41.7%) with sensitising mutations and 16 (8.5%) with other mutations such as G719X, L861Q, and insertions in Exon 20. KRAS gene mutations (G12X) were found in 8 (4.2%) LACs, while ROS1 and ALK fused genes occurred in 1 (0.5%) and 13 (7%) LACs, respectively. The detailed clinicopathological characteristics are analysed in Table 2.

Early-stage LACs showed a significantly higher occurrence of EGFR activating mutations (P=0.003). LACs with an acinar pattern were less frequently EGFR-mutated (45%) than those without this pattern (60%) (P=0.05). ALK fusion genes were found more often in LACs harboring solid components (P=0.02). The expression of P53 protein increased as LACs progressed (P<0.001). Additionally, P53 overexpression was associated with histological type; LACs with solid (71%) and acinar (56%) components overexpressed P53 protein compared to those without these components (P<0.001). Similarly, an increase in P53 protein expression was observed in LACs at stage III (71%) compared to those at stage I (14%) (P<0.001).

LACs expressing PD-L1 with a TPS of 50% or higher were more frequently observed at stage III (26%) compared to stage I (9%) (P<0.001). LACs with a solid component had a higher level of PD-L1 expression than those without this component (P<0.001). P53 and PD-L1 proteins were more frequently expressed in LACs at advanced stages and in solid components.

Scoring of CD4⁺ and CD8⁺ T cells infiltration in LAC parenchyma and stroma with PD-L1 and COL1A1 expression

In this study, the number of CD4⁺ T cells infiltrating tumor cells was correlated with the number of CD4⁺ T cells in the stroma (Spearman r=0.53, P<0.001; Figure 2A,2B). Similarly, the infiltration of CD8⁺ cytotoxic T cells in the tumor parenchyma was associated with their presence in the stroma (Spearman r=0.51, P<0.001; Figure 2C,2D). Furthermore, the number of CD4⁺ T cells in either the parenchyma or stroma was significantly associated with CD8⁺ T cells (P<0.001) (Figure 2E,2F).

Figure 2 Images of immunostaining sections showing varying levels of TILs. (A,C) Assessment of T-cell (CD4⁺ and CD8⁺) infiltration in desmoplastic tumor parenchyma and stroma, with representative images of immunohistochemical staining images (magnification: 200×); (B,D) Spearman rank correlation plots for CD4⁺ and CD8⁺ T-cell infiltration in stroma and parenchyma; (E,F) quantification of Spearman rank correlations for CD4⁺ and CD8⁺ T cells in stroma and parenchyma, respectively. TILs, tumor-infiltrating lymphocytes.

Intraepithelial infiltrating CD4⁺ and CD8⁺ T cells were evaluated and the results were shown in Figure 3. They were more common in EGFR wild-type LACs than in those with EGFR mutations (Figure 3A,3C). In contrast, the number of CD4⁺ and CD8⁺ T cells in the stroma did not exhibit a significant difference between these groups (Figure 3B,3D).

Figure 3 Infiltration of CD4⁺ and CD8⁺ T cells and expression of COL1A1 in LACs with different EGFR statuses and PD-L1 expression levels. (A-D) Infiltration of CD4⁺ and CD8⁺ T cells in the intraepithelial and stromal regions of EGFR-Mu or EGFR-WT LACs; (E-H) infiltration of CD4⁺ and CD8⁺ T cells in the intraepithelial and stromal regions of LACs exhibiting varying levels of PD-L1 expression; (I-L) infiltration of CD4⁺ and CD8⁺ T cells in the intraepithelial and stromal regions of LACs across different clinical stages; (M) expression levels of COL1A1 in EGFR mutant and wild-type LACs; (N) COL1A1 protein expression in LACs with different levels of PD-L1 expression; (O) COL1A1 protein expression in LACs at different clinical stages; (P) expression of COL1A1 protein in LACs determined using immunohistochemical staining (magnification: 200×). *, P<0.05; **P<0.01; ***, P<0.001; all other comparisons exhibit no statistical significance. EGFR, epidermal growth factor receptor; LACs, lung adenocarcinomas; Mu, mutant; PD-L1, programmed cell death ligand 1; TPS, tumor proportion score; WT, wild-type.

More intraepithelial infiltrating CD8⁺ T cells were observed in LACs with PD-L1 TPS <1% compared to those with TPS ≥50% (P<0.001; Figure 3G), whereas no significant increase in CD4⁺ T cells was found either intraepithelial or in the stroma (Figure 3E,3F). Infiltrating CD8⁺ T cells in the tumor stroma were not associated with PD-L1 expression (Figure 3H). Intraepithelial infiltrating CD4⁺ T cells were not associated with the clinical stages of LACs (Figure 3I). However, CD4⁺ T cells in stage III LACs were more prevalent in the stroma than in stage II LACs (P<0.01, Figure 3J). Intraepithelial infiltrating CD8⁺ T cells were more common in stage III LACs than in stage I LACs (P<0.01, Figure 3K). Nevertheless, CD8⁺ T cells infiltrating in stroma were not influenced by clinical stages (Figure 3L).

The COL1A1 protein was overexpressed in LACs with EGFR mutations (P<0.01; Figure 3M) but was not associated with PD-L1 expression or clinical stage (Figure 3N,3O). Moreover, the regions overexpressing COL1A1 were surrounded by sclerotic stroma, which may inhibit the influx of immune cells (Figure 3P). These results support the combined use of agents targeting both PD-L1 and COL1A1.

Involvement of EGFR, TGF-β and IL-6/STAT3 signaling pathways in the regulation of PD-L1 expression in EGFR-mutant LACs

Activated EGFR upregulates PD-L1 through the p-ERK1/2/p-c-Jun pathway (22), and the IL-6/STAT3 pathway is also associated with PD-L1 expression (23,24). However, it is important to clarify how these pathways are activated in the tumor parenchyma.

This study presents the histological features of LAC and the varied expression of PD-L1. These details are provided in Table 3 and Figure 4A-4L. Out of the eleven LAC samples, three displayed more than one histological pattern. It is also reported that PD-L1 expression is heterogeneous with uneven protein staining observed on tumor cells (25).

Figure 4 Pathological and genetic characteristics, along with PD-L1 expression, in laser-captured LAC tissues featuring EGFR exon 19 frame deletion (E19del) and L858R mutation. These characteristics were analyzed using ARMS and ddPCR assays. (A-D) Different histological components of LACs from Case 2 (Table 3) with the EGFR E19del mutation were stained with hematoxylin and eosin (A,C) and PD-L1 immunohistochemical stain (SP263). Magnification: 200×. The TPS was 80% in the solid component (70%) and 30% in the acinar component (B,D). (E-H) Histological components of LACs in Case 5 (Table 3) with the L858R mutation were stained with hematoxylin and eosin (E,G) and PD-L1 immunohistochemical stain (SP263) (F,H). Magnification: 200×. (I-N) Solid (60%) (I,J) and acinar (40%) (K,L) tissues were microdissected and stained with hematoxylin eosin. The EGFR E19del mutation in Case 2 was screened using ARMS (M1), while the microdissected tissue from the solid area was analyzed by ddPCR, revealing an abundance of 63.2% (M2). The abundance of E19del from the acinar area was 22.5% (M3). The L858R mutation of EGFR in Case 5 was assayed using ARMS (N1), and both solid and acinar components were microdissected. The abundance of L858R was analyzed via ddPCR, showing 71% in the solid component (N2) and 35.2% in the acinar component (N3). (O) Levels of pERK1/2, TGF-β1, and pSTAT3 proteins were measured in different components of the two cases using western blotting (lanes 1–2 for Case 2; lanes 3–4 for Case 5), with GAPDH serving as the loading control. (P) Cytokine expression was assessed using a cytometric bead array. IL-6 levels were significantly higher in areas with a PD-L1 TPS of 50% or greater compared to those with a TPS of less than 50% (P<0.05). ARMS, amplification refractory mutation system; ddPCR, droplet digital polymerase chain reaction; IL, interleukin; LAC, lung adenocarcinoma; PD-L1, programmed cell death ligand 1; PH, PD-L1 expression with high-level (TPS ≥50%); PL, PD-L1 expression with low-level (TPS <50%); TGF, transforming growth factor; TPS, tumor proportion score; TNF, tumor necrosis factor.

To further explore the relationship between EGFR activation and PD-L1 expression in EGFR-mutant LACs, we used laser capture to isolate tissues from histological components and classified tumor cells based on PD-L1 expression into two categories: low-level (TPS <50%, PL) and high-level (TPS ≥50%, PH). The quantities of EGFR mutation varied among tumor cells with different PD-L1 expressions (Table 3). The PD-L1 protein expression was not correlated with the quantities of EGFR mutation (Figure 4M,4N), indicating that EGFR mutation alone is insufficient to drive PD-L1 expression.

We used western blotting to analyze the isolated tissues for pERK1/2 and pSTAT3 proteins, aiming to clarify the participation of the EGFR and IL-6/STAT3 pathways in PD-L1 expression (Figure 4O, Table 3). However, expression levels differed among histological components. The results showed that pERK1/2 was overexpressed in solid components but expressed at lower levels in acinar components. Similarly, the pSTAT3 and TGF-β1 proteins were expressed at higher levels in solid components compared to acinar components. These findings revealed that PD-L1 expression in solid components was synergistically regulated by the STAT3, ERK1/2 and TGF-β pathways.

We further measured cytokine levels (IL-2, IL-4, TNF-α, IL-10, IL-17, IL-6, and IFN-γ) that could influence TME remodeling. IL-6 was secreted at higher levels in EGFR-mutant LACs, even though PD-L1 overexpression levels varied (Figure 4P). This result demonstrated that the IL-6/STAT3 pathway was activated in the solid LAC components. Other cytokines were expressed at lower levels and did not appear to contribute significantly to PD-L1 expression.

Combined blockade of camrelizumab and osimertinib enhances cytokine levels with variable pathological responses in mice

Based on our previous findings that solid LACs exhibit higher expression of PD-L1 and P53 proteins compared to non-solid LACs, and considering the challenge of obtaining complete tumor tissue from patients resistant to osimertinib. We utilized a solid LAC with osimertinib-resistance, a PD-L1 TPS greater than 50%, and P53 overexpression in the PDX mice model. We aimed to investigate morphological changes across various groups based on established evaluation criteria (26).

Morphologically, tumour nests were compact and had minimal stromal presence in the control group. After TGF-β induction, fibrosis was evident and stromal septa with infiltrating immune cells surrounded the tumor nests. In the osimertinib group, tumor cells were visibly surrounded by thicker stromal septa and the proportion of tumor cells was similar to the TGF-β group. In the group treated with camrelizumab alone, atrophic tumour cells were trapped within prominent infiltrating immune cells and stromal fibers with inconspicuous necrosis. However, in the synergistic osimertinib and camrelizumab group, infiltrating lymphocytes predominantly surrounded tumour cells, coupled with extensive tumor necrosis and fewer residual tumour cells (Figure 5A).

Figure 5 Pathological responses and cytokine secretion in patient-derived solid LAC xenograft mice with different treatment groups. (A) Morphological changes are shown in hematoxylin and eosin stained sections from different groups of the xenograft model (magnification ×200). Tumor cells (red arrow) are surrounded by thin stromal septa, which contains fewer infiltrating immune cells (yellow arrow). In the TGF-β group, tumor cells (red arrow) are separated by sclerotic fibers, which have fewer immune cells (yellow arrow). In the EGFR-TKT group, tumor cells (red arrow) are isolated by fibers (yellow arrow) that contain infiltrating immune cells in the stroma. In the single-ICI treatment group, atrophic tumor cells (red arrow) are surrounded by prominent infiltrating immune cells (yellow arrow). In the group treated with a combination of EGFR-TKT and ICI, conspicuous necrosis (red arrow) and infiltrating immune cells (yellow arrow) occupy the tumor bed, and there are few residual tumor cells. Pathological responses in the different groups were assessed based on the proportions of necrosis, inflammation, fibrosis, and residual tumors. (B) Distribution of infiltrating CD8⁺ T cells in different groups shown by immunohistochemical staining (magnification: 100×). CD4⁺CD3⁺ and CD8⁺CD3⁺ T cells were assayed using flow cytometry in different groups. *, P<0.05; **, P<0.01. (C-I) Cytokines secreted in different treatment groups. *, P<0.05; **, P<0.01. EGFR-TKT, epidermal growth factor receptor-tyrosine kinase inhibitor treatment; LAC, lung adenocarcinoma; ICI, immune checkpoint inhibitor; IFN, interferon; IL, interleukin; TGF-β, transforming growth factor-β; TNF, tumor necrosis factor.

Infiltrating T cells were evaluated by flow cytometry. A greater number of CD8⁺ T cells also infiltrated the camrelizumab group compared to the control group (P<0.05). A significant increase in CD8⁺ T cells was found in the therapy group combining osimertinib and camrelizumab compared to the control and TGF-β groups (P<0.05). Additionally, the distribution of CD8⁺ T cells was shown in immunohistochemical sections, demonstrating migration from the stroma into the tumor parenchyma after combined therapy. Conversely, the amount of infiltrating CD4⁺ T cells in the solid PDX tumors was not significantly different among the groups (Figure 5B).

These findings suggested that synergistic treatment with osimertinib and camrelizumab significantly promote CD8⁺ T cell infiltration and tumor necrosis. Osimertinib-resistant solid LACs benefited from the combined use of osimertinib and camrelizumab resulting in increased tumor necrosis and stromal inflammation. Thus, these results suggest that the combination of osimertinib and camrelizumab is a potential preferred strategy for treating p53-mutated osimertinib-resistant solid LAC.

To explore immune cell activation during immunization, we analyzed the expression levels of relevant cytokines. All cytokines remained at lower levels in the control group, indicating that CD4⁺ and CD8⁺ T cells were significantly suppressed, similar to the situation observed in primary tumors. Compared to the control group, a significant increase in IL-2, IL-4, IL-10, IL-17, TNF-α, and IFN-γ expression was observed in the osimertinib and camrelizumab treatment group except for IL-6 (Figure 5C-5I). IL-6 was secreted at the highest level in the TGF-β group followed by the osimertinib group. IL-17 levels remained higher in the camrelizumab group, second only to the combination group. These results reflect the efficacy of osimertinib plus camrelizumab in inducing significant antitumor activity in the osimertinib-resistant LAC.

EGFR and TGF-β signaling pathways are inhibited by camrelizumab and osimertinib

The expression levels of TGF-β1, TGF-β2, TGF-βR, Smad2, Smad3, and EGFR proteins shown in Figure 6 were significantly lower in both the camrelizumab (ICI) and camrelizumab and osimertinib (ICI + EGFR-TKT) groups compared to the control and TGF-β-induced groups (P<0.05). These results indicate that the TGF-β signal pathway is inhibited by ICI therapy and the combined use of ICI and EGFR-TKT.

Figure 6 Western blotting revealing the expression of proteins linked to EGFR and TGF-β pathways in xenograft LAC tumor tissues across different treatment groups: Control, TGF-β-induced, ICI, EGFR-TKT, and a combination of EGFR-TKT and ICI. *, P<0.05; **, P<0.01; ***, P<0.001, ****, P<0.0001. EGFR-TKT, epidermal growth factor receptor-tyrosine kinase inhibitor treatment; ICI, immune checkpoint inhibitor; LAC, lung adenocarcinoma; TGF-β, transforming growth factor-β.

There was no significant difference in protein expression levels between the osimertinib and camrelizumab groups. Similarly, in the camrelizumab, osimertinib, and osimertinib + camrelizumab treatment groups, the expression levels of TGF-β1, TGF-β2, TGF-βR, Smad2, Smad3, and EGFR proteins were not significantly different between the ICI and EGFR-TKT groups (P>0.05). These findings suggest that single-agent ICI or EGFR-TKT in solid LACs does not significantly impact the TGF-β signal pathway.

Association of clinical response with cytokine levels after camrelizumab treatment in osimertinib-resistant LACs

We designed a clinical trial involving twelve osimertinib-resistant patients with their EGFR status to assess whether cytokine levels could predict the clinical response to camrelizumab treatment (Table 4). Eight male and four female patients, aged 56 to 74 years, were enrolled. Histological analyses showed that five patients had solid components, three had acinar components, and four had micropapillary or papillary components.

During clinical observation, none of the patients experienced intolerable adverse events, and all continued camrelizumab treatment until the study’s conclusion. Among the patients with solid components, 3 out of 5 (60%) achieved partial remission (PR) (Figure 7A,7B). Conversely, all four patients with micropapillary and papillary components experienced progressive disease (PD), while the remaining five patients exhibited stable disease (SD).

Figure 7 Follow-up computer tomography of a representative case after 6 months. Cytokine expression levels in LAC cases based on varied clinical responses to ICI following osimertinib resistance. (A,B) Computed tomography from a representative case that demonstrated partial remission of the focus (red arrows) following ICI treatment after osimertinib resistance; (C-I) cytokine levels in LACs treated with ICI after developing resistance to osimertinib. **, P<0.01; ***, P<0.001. ICI, immune checkpoint inhibitor; IFN, interferon; IL, interleukin; LAC, lung adenocarcinoma; PD, progressive disease; PR, partial remission; SD, stable disease; TNF, tumor necrosis factor.

Cytokine levels were measured in BALF. Patients with PR had significantly higher levels of IL-2, IL-4, IFN-γ, and IL-10 compared to those with PD (P<0.01; Figure 7C-7I).

Furthermore, cytokines such as IL-2, IL-4, TNF-α, IL-10, and IFN-γ may serve as potential indicators of clinical response in this subset of LACs (Figure 8).

Figure 8 Overview of the dynamic changes in the TME of lung adenocarcinoma, depicting interactions between tumor cells and stromal components, and changes in fibroblasts, immune cells, T cells, and cytokines during the immunotherapy following EGFR-TKT. CAFs, cancer-associated fibroblasts; EGFR-TKT, epidermal growth factor receptor-tyrosine kinase inhibitor treatment; ICI, immune checkpoint inhibitor; IFN, interferon; IL, interleukin; PD-1, programmed cell death protein 1; PD-L1, programmed cell death ligand 1; TGF-β, transforming growth factor-β; TME, tumor microenvironment.

Discussion

LAC is one of the most lethal malignant tumors. As LAC progresses, tumor cell death, microenvironment remodeling, and adaptive changes in tumor cells occur, each involving distinct molecular alterations. Moreover, LAC can be classified into several histological subtypes: papillary, micropapillary, solid, and acinar. These histological subtypes can be intermixed within a single LAC tumor mass. The varying expression of PD-L1 in LACs, along with their differing responses to ICIs, highlights the intratumor heterogeneity and distinct molecular mechanisms. The mechanisms underlying the complexity of LAC associated with EGFR mutations, are still not well understood.

The Intermixed histological components of LAC influence therapeutic targets by altering the TME, which can lead to drug resistance (27). The TME consists of several components: immune cells, fibroblasts, endothelial cells, the ECM, vasculature, and chemokines (28). The TME plays a crucial role in cancer development, progression, and metastasis due to the intricate crosstalk among its components, and it undergoes remodeling due to panoptosis and the replacement of CAFs. Therefore, the TCGA and GSE68465 cohorts were analyzed to explore DEGs and their associated pathways through GO and KEGG analyses in LACs at varied clinical stages.

In these large-scale cohorts, KEGG analysis indicated a significant aggregation in the TGF-β signaling pathway and cytokine interaction networks. Meanwhile, GO analysis identified key functional categories, including “extracellular matrix structural constituent”, “growth factor binding”, “protease binding”, and “cytokine binding”, all of which play crucial roles in the advancement of LAC. These results suggest that the TME of LAC patients exhibits specific features and modulates the cancer-associated immune response through these pathways.

It is well known that EGFR-TKIs induce apoptosis in EGFR-mutant tumor cells and remodel the TME. Osimertinib effectively inhibits the growth of LACs. However, tumor heterogeneity, a hallmark of cancer, remains a significant challenge in oncology as it is a primary cause of drug resistance. A meta-analysis of the CheckMate 057, KEYNOTE-010, and POPLAR trials indicate that ICI monotherapy does not improve overall survival for patients with EGFR mutations compared to docetaxel (29). However, other clinical trials, such as TATTON and KEYNOTE-021, have demonstrated that combining ICI with EGFR-TKI can enhance the overall response rate in LACs (15). Nevertheless, the results from clinical trials of involving combination treatments of EGFR-TKI and ICI have been inconsistent.

In addition, the transition of the TME during tumor progression is influenced by the TGF-β signaling pathway. This pathway has diverse effects on cellular processes, including proliferation, differentiation, adhesion, senescence, and apoptosis. In cancers, TGF-β-mediated immunosuppression leads to phenotypic changes in various immune cells, including dendritic cells, tumor-associated macrophages (TAMs), tumor-associated neutrophils, natural killer (NK) cells, myeloid-derived suppressor cells, regulatory T cells, and cytotoxic T cells. The TGF-β pathway is deeply involved in remodeling TME.

In early-stage tumors, the TGF-β pathway triggers apoptosis and inhibits the proliferation of cancer cells. However, in late-stage tumors, it promotes tumor growth by regulating genomic instability, the epithelial to mesenchymal transition, neoangiogenesis, immune evasion, and metastasis. TGF-β also reduces the differentiation and function of Th1 and Th2 cells, as well as cytotoxic T lymphocytes, all of which are crucial for antitumor responses (16).

The EGFR signaling pathway and EGFR-TKIs influence various aspects of immune efficacy. Given the resistance to EGFR-TKIs and the reconstruction of the TME during tumor therapy, it is essential to investigate whether ICIs can benefit patients previously treated with EGFR-TKIs.

To explore the changes of in the TME in primary LACs, including those with EGFR mutations, we studied on a native cohort of 187 LACs. Our focus was initially on evaluating the infiltration of CD4⁺ and CD8⁺ T cells, COL1A1 expression, and cytokines related to TME reconstruction. We assessed the presence of CD4⁺ and CD8⁺ T cells in LACs and found that these cells infiltrated both the tumor parenchyma and stroma synergistically. The numbers of intraepithelial CD4⁺ and CD8⁺ T cells decreased in EGFR-mutant LACs, whereas their numbers in the stroma were not significantly affected.

EGFR-mutant LACs exhibited a non-inflamed TME, despite a high infiltration of CD4⁺ effector regulatory T cells, which are typically associated with inflamed TMEs. The EGFR signaling pathway activates c-Jun/c-Jun N-terminal kinase, which increases CCL22 levels to recruit CD4⁺ regulatory T cells, while simultaneously reducing interferon regulatory factor-1, leading to decreased CXCL10 and CCL5 levels, important for inducing CD8⁺ T cell infiltration. EGFR signaling pathway also enhances PD-L1 expression promoting T cell apoptosis and immune escape (30-32). We observed that CD8⁺ T cells significantly decreased in the tumor parenchyma of EGFR-mutant LACs and in those with a PD-L1 TPS ≥50%. Fewer CD8⁺ T cells infiltrated the tumor parenchyma in advanced LACs and in those with a PD-L1 TPS ≥50%. These findings suggest that during tumor progression, immune cells, particularly T cells, are repelled from the tumor parenchyma via PD-L1-mediated mechanism.

There are lots of ICIs have been launched both domestically and internationally, such as camrelizumab, sintilimab, pembrolizumab, and nivolumab. A previous study compared the clinical effects of these ICIs among patients with advanced NSCLCs. The clinical outcomes demonstrate their similar clinical responses. Meanwhile, PD-1 ICI monotherapy exhibits potential effectiveness and acceptable toxicity for elderly patients with NSCLC (33). However, the varying effectiveness of anti-PD-1 monoclonal antibody (mAb) therapy in cancers with oncogenic driver gene mutations suggests that each driver mutation differently impacts immune responses.

The expression of the COL1A1 protein was higher in EGFR-mutant LACs. However, EGFR mutations are more frequently found in specific histological types, including micropapillary and papillary LACs (34). As COL1A1 protein is a biomarker of CAFs, its overexpression in EGFR-mutant LACs indicates an increase in stromal fibers, which can trap tumor cells within the stroma and isolate both ICI and immune cells from tumor cells. This phenomenon may partly explain the poorer outcomes of EGFR-mutant LACs in clinical trials evaluating ICIs.

PD-L1 protein is expressed in tumor cells and is influenced by two mechanisms: intrinsic and acquired expression (35,36). EGFR mutations can increase PD-L1 expression in NSCLC cells by activating several downstream signaling pathways, including mitogen-activated protein kinase (MAPK)/extracellular signal-regulated kinase (ERK)/c-Jun, Hippo/YAP, and Janus kinase/signal transducer and activators of transcription 3 (JAK/STAT3). Preclinical studies show conflicting results regarding how EGFR signaling regulates PD-L1 expression (22,37,38).

To elucidate the regulatory mechanisms of PD-L1, we analyzed protein expressions associated with EGFR, TGF-β, and IL-6/STAT pathways using microdissection and western blotting. We found that the quantities of EGFR-mutant tumor cells varied among the microdissected histological components with different levels of PD-L1 expression. The microdissected solid components from EGFR-mutant LACs exhibited higher PD-L1 expression along with TGF-β1, pSTAT3, and pERK1/2 overexpression. These findings suggest that PD-L1 overexpression in solid components is synergistically regulated by MAPK/ERK, JAK/STAT3 and TGF-β pathways.

LACs with a solid predominant component are associated with significantly worse outcomes following EGFR-TKI treatment compared to those without this component (39,40). Induced PD-L1 expression by cytokines, such as IL-6, appears to be more common than intrinsic expression. Adaptive PD-L1 expression, triggered by the antigen-specific T cell recognition of cancer cells, leads to regulatory T cell activation and inhibits specific cytotoxic immune responses (41).

The observed increase in IL-6 levels in this study suggests that both adaptive and intrinsic mechanisms contribute to PD-L1 expression in LACs, especially in solid components. IL-6 binding to its receptor (IL-6R) activates the JAK family of tyrosine kinases, which subsequently stimulate multiple pathways, including MAPKs, PI3Ks, and STATs. In NSCLC-derived cell lines, STAT3 is continuously activated and plays a role in tumorigenesis, cell cycle progression, tumor invasion, and metastasis through complex mechanisms (42). Our results demonstrated that the EGFR, TGF-β, and IL-6/STAT3 signaling pathways are co-activated in regulating PD-L1 expression in solid components.

Although it is extremely rare for PD-L1 ≥50% to coexist with driver mutations, some studies have shown that a small number of patients with high PD-L1 expression benefited from immunotherapy (43). The ATLANTIC trial assessed the efficacy of durvalumab in EGFR-mutant patients who had received EGFR-TKI treatment (44). Among EGFR-mutant patients with at least 25% of tumor cells expressing PD-L1, durvalumab was associated with a prolongation of the overall survival of the EGFR-mutant patients compared to EGFR⁻/ALK⁻ patients (14). Other clinical investigations have indicated that expression of PD-L1 tend to increase following treatment with EGFR-TKIs (45). In a cohort of 128 patients undergoing EGFR-TKI therapy, the proportion of individuals exhibiting elevated PD-L1 expression (defined as a staining intensity ≥50% in tumor cells) rose from 14% to 28% (46). Furthermore, another study reported that among a sample of 12 patients, 21% displayed heightened PD-L1 expression in their tumor samples after the development of resistance to EGFR-TKIs (47).

The BIRCH trial showed that the overall response rate was 31% in the EGFR-mutant group and 22% in the wild-type group in patients with PD-L1 expression of at least 5% on tumor cells or immune cells treated with atezolizumab as first-line therapy (48). These findings suggest that EGFR-mutant patients with high PD-L1 expression benefitted from ICIs.

Based on our results, we speculate the potential mechanisms are as follows: after EGFR-TKI treatment, the EGFR-mutant tumor cells are eradicated, leading to TME reconstruction. Residual tumor cells, especially solid components with more activating signaling pathways, may exhibit greater resistance to EGFR-TKI and evade EGFR-TKI-mediated effects. These solid components more frequently overexpress PD-L1 protein, and infiltrating CD4⁺ and CD8⁺ T cells are reactivated at the advanced stage, resulting in a favorable response to ICI treatment.

To replicate the process of tumor immune reactivation, we applied a humanized immune-reconstructed PDX mouse model and explored immune microenvironment reconstruction in osimertinib-resistant solid LAC with P53 mutation, as well as the pathological responses to the synergistic treatment of osimertinib and camrelizumab.

Firstly, we observed the pathological responses to different regimens. These responses included marked necrosis and stromal changes, characterized by a large influx of immune T cells into both the parenchyma and stroma. In this study, TGF-β-induced tumors exhibited more fibrotic stroma, less necrosis, and a higher number of residual tumor cells. In the camrelizumab-alone group, notable stromal replacement was observed, with less necrosis, and atrophic tumor cells were entrapped in fibrous stroma alongside increasing T cells.

The number of CD8⁺ T cells infiltrating the tumors increased in the group treated with the combination of osimertinib and camrelizumab. However, there was no significant difference in the quantity of infiltrating CD4⁺ T cells among the groups. Meanwhile, the expression of proteins associated with the TGF-β pathway was inhibited in the solid LAC treated with the combined regimen.

These findings suggest that combination of osimertinib and camrelizumab reactivates Th1/Th2 cells, leading to significant increases in cytokines, such as IL-2, IL-4, IL-10, IL-17, TNF-α, and IFN-γ during immune recruitment. Infiltrating CD4⁺ and CD8⁺ T cells suppress tumor cells through cytokine production, validating the efficacy of combining ICI and EGFR-TKI therapies in solid LACs.

The results from our study suggest that blocking the TGF-β signaling pathway can inhibit the epithelial-mesenchymal transition and enhance immune responses against tumors. A previous study also found that EGFR-mutant LAC cells are more inclined to interact with other cells through the binding of TGF family ligands and receptors compared to EGFR wild-type LACs (49).

Therefore, we speculate that the better pathological responses of solid LACs to osimertinib plus camrelizumab may result from p53 gene mutation and the activation of the TGF-β pathway. p53 gene mutation predicts genomic instability in tumors and a better response to ICIs suggests that lymphocytes and the cytokines they secrete collaborate in reconstructing the TME, influencing the differential pathological responses of EGFR-mutant solid LAC. These findings indicate that solid LACs can benefit from combination therapy with osimertinib and camrelizumab.

To validate the findings from the native cohort and the PDX model, we conducted a six-month trial involving twelve patients with osimertinib-resistant LACs. These patients received camrelizumab, and their cytokine levels were analyzed to evaluate clinical response. Patients with improved clinical outcomes, particularly those with SDs, exhibited higher levels of IL-2, IL-4, TNF-α, IL-10, and IFN-γ compared to those with progressive disease. Elevated cytokine levels reflected the reactivation of immune cells and predicted the dynamic changes in the TME during camrelizumab treatment. These cytokines strongly suggest their potential value in evaluating treatment efficacy, as they reflect the activity of infiltrating immune cells and their pathological responses.

We also suspect that the cases failing to achieve satisfactory clinical outcomes may be due to the increased fibrous components within the TME, which isolate immune cells from tumor cells.

Conclusions

LACs exhibit various histological types. Previous studies have shown that micropapillary and papillary LACs are more likely to harbor EGFR mutations, whereas solid LACs tend to overexpress PD-L1 protein (34,50,51). Although clinical trials have demonstrated that ICIs combined with EGFR-TKT do not yield effective responses in EGFR-mutant LACs, our results indicate that EGFR-mutant solid LACs with PD-L1 and P53 overexpression exhibit better pathological responses both in the PDX model and in a clinical trial. These findings suggest that solid LACs can benefit from combined therapy with osimertinib and camrelizumab.

Acknowledgments

We thank Zheng Tang and Yi-Lin Xu from Guangdong Yiruibei Biomedical Technology Co., Ltd. in support of constructing animal model.

Footnote

Reporting Checklist: The authors have completed the MDAR and ARRIVE reporting checklists. Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-522/rc

Data Sharing Statement: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-522/dss

Peer Review File: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-522/prf

Funding: This study was funded by the Beijing Outstanding Young Talents Program (grant No. 2009D003013000001 to Y.C.) and the Hebei Provincial Health Commission’s Medical Science Research Plan (grant No. 20211023 to M.Z.).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-522/coif). M.Z. reports a grant from the Hebei Provincial Health Commission’s Medical Science Research Plan (grant No. 20211023). Y.C. reports a grant from the Beijing Outstanding Young Talents Program (grant No. 2009D003013000001). The other 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 conducted in accordance with the Declaration of Helsinki and its subsequent amendments. It was authorized and supervised by the Ethics Committee of Shanghai General Hospital (approval No. 2021020). Informed consent was obtained from all participants included in this study. All animal experiments were conducted in accordance with the Guide for the Care and Use of Experimental Animals from the Experimental Animal Center of Shanghai General Hospital (approval No. 2023SQ067).

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. WHO Classification of Tumours Editorial Board. Thoracic tumours/edited by WHO Classification of Tumours Editorial Board. World Health Organization classification of tumours. Lyon: International Agency for Research on Cancer; 2021.
2. Liu Y, Ye G, Huang L, et al. Single-cell transcriptome analysis demonstrates inter-patient and intra-tumor heterogeneity in primary and metastatic lung adenocarcinoma. Aging (Albany NY) 2020;12:21559-81. [Crossref] [PubMed]
3. Fulda S. The mechanism of necroptosis in normal and cancer cells. Cancer Biol Ther 2013;14:999-1004. [Crossref] [PubMed]
4. Shi J, Zhao Y, Wang K, et al. Cleavage of GSDMD by inflammatory caspases determines pyroptotic cell death. Nature 2015;526:660-5. [Crossref] [PubMed]
5. Chen LC, Wang LJ, Tsang NM, et al. Tumour inflammasome-derived IL-1β recruits neutrophils and improves local recurrence-free survival in EBV-induced nasopharyngeal carcinoma. EMBO Mol Med 2012;4:1276-93. [Crossref] [PubMed]
6. Műzes G, Sipos F. PANoptosis as a Two-Edged Sword in Colorectal Cancer: A Pathogenic Mechanism and Therapeutic Opportunity. Cells 2025;14:730. [Crossref] [PubMed]
7. Khan SU, Fatima K, Malik F. Understanding the cell survival mechanism of anoikis-resistant cancer cells during different steps of metastasis. Clin Exp Metastasis 2022;39:715-26. [Crossref] [PubMed]
8. Lavie D, Ben-Shmuel A, Erez N, et al. Cancer-associated fibroblasts in the single-cell era. Nat Cancer 2022;3:793-807. [Crossref] [PubMed]
9. Passaro A, Jänne PA, Mok T, et al. Overcoming therapy resistance in EGFR-mutant lung cancer. Nat Cancer 2021;2:377-91. [Crossref] [PubMed]
10. Zhao Y, Wang H, He C. Drug resistance of targeted therapy for advanced non-small cell lung cancer harbored EGFR mutation: from mechanism analysis to clinical strategy. J Cancer Res Clin Oncol 2021;147:3653-64. [Crossref] [PubMed]
11. Bronte G, Belloni A, Calabrò L, et al. The great need to overcome osimertinib resistance in advanced non-small cell lung cancer: from combination strategies to fourth-generation tyrosine kinase inhibitors. Front Oncol 2023;13:1308460. [Crossref] [PubMed]
12. Lisberg A, Cummings A, Goldman JW, et al. A Phase II Study of Pembrolizumab in EGFR-Mutant, PD-L1+, Tyrosine Kinase Inhibitor Naïve Patients With Advanced NSCLC. J Thorac Oncol 2018;13:1138-45. [Crossref] [PubMed]
13. Gettinger S, Rizvi NA, Chow LQ, et al. Nivolumab Monotherapy for First-Line Treatment of Advanced Non-Small-Cell Lung Cancer. J Clin Oncol 2016;34:2980-7. [Crossref] [PubMed]
14. Garassino MC, Cho BC, Kim JH, et al. Final overall survival and safety update for durvalumab in third- or later-line advanced NSCLC: The phase II ATLANTIC study. Lung Cancer 2020;147:137-42. [Crossref] [PubMed]
15. Ma L, Diao B, Huang Z, et al. The efficacy and possible mechanisms of immune checkpoint inhibitors in treating non-small cell lung cancer patients with epidermal growth factor receptor mutation. Cancer Commun (Lond) 2021;41:1314-30. [Crossref] [PubMed]
16. Gulley JL, Schlom J, Barcellos-Hoff MH, et al. Dual inhibition of TGF-β and PD-L1: a novel approach to cancer treatment. Mol Oncol 2022;16:2117-34. [Crossref] [PubMed]
17. Zhang Y, Zeng Y, Liu T, et al. The canonical TGF-β/Smad signalling pathway is involved in PD-L1-induced primary resistance to EGFR-TKIs in EGFR-mutant non-small-cell lung cancer. Respir Res 2019;20:164. [Crossref] [PubMed]
18. Kayser G, Schulte-Uentrop L, Sienel W, et al. Stromal CD4/CD25 positive T-cells are a strong and independent prognostic factor in non-small cell lung cancer patients, especially with adenocarcinomas. Lung Cancer 2012;76:445-51. [Crossref] [PubMed]
19. Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer 2009;45:228-47. [Crossref] [PubMed]
20. Kwiecień I, Rutkowska E, Raniszewska A, et al. Flow Cytometric Analysis of Macrophages and Cytokines Profile in the Bronchoalveolar Lavage Fluid in Patients with Lung Cancer. Cancers (Basel) 2023;15:5175. [Crossref] [PubMed]
21. Jia Y, Zhao S, Jiang T, et al. Impact of EGFR-TKIs combined with PD-L1 antibody on the lung tissue of EGFR-driven tumor-bearing mice. Lung Cancer 2019;137:85-93. [Crossref] [PubMed]
22. Chen N, Fang W, Zhan J, et al. Upregulation of PD-L1 by EGFR Activation Mediates the Immune Escape in EGFR-Driven NSCLC: Implication for Optional Immune Targeted Therapy for NSCLC Patients with EGFR Mutation. J Thorac Oncol 2015;10:910-23. [Crossref] [PubMed]
23. Jing B, Wang T, Sun B, et al. IL6/STAT3 Signaling Orchestrates Premetastatic Niche Formation and Immunosuppressive Traits in Lung. Cancer Res 2020;80:784-97. [Crossref] [PubMed]
24. Tomasich E, Topakian T, Heller G, et al. Loss of HCRP1 leads to upregulation of PD-L1 via STAT3 activation and is of prognostic significance in EGFR-dependent cancer. Transl Res 2021;230:21-33. [Crossref] [PubMed]
25. Cai Y, Wu H, Shi X, et al. Heterogeneous components of lung adenocarcinomas confer distinct EGFR mutation and PD-L1 expression. BMC Cancer 2020;20:148. [Crossref] [PubMed]
26. Travis WD, Dacic S, Wistuba I, et al. IASLC Multidisciplinary Recommendations for Pathologic Assessment of Lung Cancer Resection Specimens After Neoadjuvant Therapy. J Thorac Oncol 2020;15:709-40. [Crossref] [PubMed]
27. Rudin CM, Cervantes A, Dowlati A, et al. Safety and clinical activity of atezolizumab plus erlotinib in patients with non-small-cell lung cancer. ESMO Open 2023;8:101160. [Crossref] [PubMed]
28. Zhang A, Miao K, Sun H, et al. Tumor heterogeneity reshapes the tumor microenvironment to influence drug resistance. Int J Biol Sci 2022;18:3019-33. [Crossref] [PubMed]
29. Lee CK, Man J, Lord S, et al. Checkpoint Inhibitors in Metastatic EGFR-Mutated Non-Small Cell Lung Cancer-A Meta-Analysis. J Thorac Oncol 2017;12:403-7. [Crossref] [PubMed]
30. Denton AE, Roberts EW, Fearon DT. Stromal Cells in the Tumor Microenvironment. Adv Exp Med Biol 2018;1060:99-114. [Crossref] [PubMed]
31. Sugiyama E, Togashi Y, Takeuchi Y, et al. Blockade of EGFR improves responsiveness to PD-1 blockade in EGFR-mutated non-small cell lung cancer. Sci Immunol 2020;5:eaav3937. [Crossref] [PubMed]
32. Zhao Y, Shen M, Wu L, et al. Stromal cells in the tumor microenvironment: accomplices of tumor progression? Cell Death Dis 2023;14:587. [Crossref] [PubMed]
33. Li XP, Zhang WD, Li MJ, et al. Effectiveness and Safety of PD-1 Inhibitor Monotherapy for Elderly Patients with Advanced Non-Small Cell Lung Cancer: A Real-World Exploratory Study. J Oncol 2022;2022:1710272. [Crossref] [PubMed]
34. Cai YR, Dong YJ, Wu HB, et al. Micropapillary: A component more likely to harbour heterogeneous EGFR mutations in lung adenocarcinomas. Sci Rep 2016;6:23755. [Crossref] [PubMed]
35. Konishi J, Yamazaki K, Azuma M, et al. B7-H1 expression on non-small cell lung cancer cells and its relationship with tumor-infiltrating lymphocytes and their PD-1 expression. Clin Cancer Res 2004;10:5094-100. [Crossref] [PubMed]
36. Gajewski TF, Schreiber H, Fu YX. Innate and adaptive immune cells in the tumor microenvironment. Nat Immunol 2013;14:1014-22. [Crossref] [PubMed]
37. D'Incecco A, Andreozzi M, Ludovini V, et al. PD-1 and PD-L1 expression in molecularly selected non-small-cell lung cancer patients. Br J Cancer 2015;112:95-102. [Crossref] [PubMed]
38. Jin R, Zhao J, Xia L, et al. Application of immune checkpoint inhibitors in EGFR-mutant non-small-cell lung cancer: from bed to bench. Ther Adv Med Oncol 2020;12:1758835920930333. [Crossref] [PubMed]
39. Yoshida T, Ishii G, Goto K, et al. Solid predominant histology predicts EGFR tyrosine kinase inhibitor response in patients with EGFR mutation-positive lung adenocarcinoma. J Cancer Res Clin Oncol 2013;139:1691-700. [Crossref] [PubMed]
40. Uehara Y, Izumi H, Taki T, et al. Solid Predominant Histology and High Podoplanin Expression in Cancer-Associated Fibroblast Predict Primary Resistance to Osimertinib in EGFR-Mutated Lung Adenocarcinoma. JTO Clin Res Rep 2025;6:100779. [Crossref] [PubMed]
41. Taube JM, Anders RA, Young GD, et al. Colocalization of inflammatory response with B7-h1 expression in human melanocytic lesions supports an adaptive resistance mechanism of immune escape. Sci Transl Med 2012;4:127ra37. [Crossref] [PubMed]
42. Ke W, Zhang L, Dai Y. The role of IL-6 in immunotherapy of non-small cell lung cancer (NSCLC) with immune-related adverse events (irAEs). Thorac Cancer 2020;11:835-9. [Crossref] [PubMed]
43. Rangachari D, VanderLaan PA, Shea M, et al. Correlation between Classic Driver Oncogene Mutations in EGFR, ALK, or ROS1 and 22C3-PD-L1 ≥50% Expression in Lung Adenocarcinoma. J Thorac Oncol 2017;12:878-83. [Crossref] [PubMed]
44. Garassino MC, Cho BC, Kim JH, et al. Durvalumab as third-line or later treatment for advanced non-small-cell lung cancer (ATLANTIC): an open-label, single-arm, phase 2 study. Lancet Oncol 2018;19:521-36. [Crossref] [PubMed]
45. 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]
46. 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]
47. Gainor JF, Shaw AT, Sequist LV, et al. EGFR Mutations and ALK Rearrangements Are Associated with Low Response Rates to PD-1 Pathway Blockade in Non-Small Cell Lung Cancer: A Retrospective Analysis. Clin Cancer Res 2016;22:4585-93. [Crossref] [PubMed]
48. Peters S, Gettinger S, Johnson ML, et al. Phase II Trial of Atezolizumab As First-Line or Subsequent Therapy for Patients With Programmed Death-Ligand 1-Selected Advanced Non-Small-Cell Lung Cancer (BIRCH). J Clin Oncol 2017;35:2781-9. [Crossref] [PubMed]
49. Horn LA, Chariou PL, Gameiro SR, et al. Remodeling the tumor microenvironment via blockade of LAIR-1 and TGF-β signaling enables PD-L1-mediated tumor eradication. J Clin Invest 2022;132:e155148. [Crossref] [PubMed]
50. Miyazawa T, Morikawa K, Otsubo K, et al. Solid histological component of adenocarcinoma might play an important role in PD-L1 expression of lung adenocarcinoma. Thorac Cancer 2022;13:24-30. [Crossref] [PubMed]
51. Zhu K, Zou Y, Zhao L, et al. Relationship between PD-L1 expression and molecular aberrances in lung adenocarcinoma with solid components. J Thorac Dis 2023;15:2936-47. [Crossref] [PubMed]