Immune checkpoint inhibitors have limited efficacy in SMARCA4-deficient non-small cell lung cancer

Introduction

SMARCA4, encoding the Brahma-related gene 1 (BRG1) protein, is a critical subunit of the SWItch/Sucrose Non-Fermentable complex (SWI/SNF) chromatin remodeling complex that governs gene expression, DNA repair, and chromatin dynamics. Loss-of-function (LOF) mutations or inactivation of SMARCA4 have been increasingly observed in non-small cell lung cancer (NSCLC) and other malignancies, leading to a distinct subset of tumors known as SMARCA4-deficient non-small cell lung cancer (SD-NSCLC). SD-NSCLC represents a highly aggressive subtype characterized by early metastasis, resistance to conventional therapies, and consistently poor prognosis. The clinical implications of SMARCA4 deficiency are significant. Compared to non-SD NSCLC, SD-NSCLC patients exhibit more advanced disease at diagnosis, higher rates of recurrence, and significantly reduced overall survival (OS) (1-4). Furthermore, SMARCA4 deficiency is frequently accompanied by co-inactivation of other SWI/SNF complex components, further complicating the prognosis (5,6). Studies suggest that SMARCA4, in combination with co-mutations in the KRAS, TP53, STK11, and KEAP1 genes, may affect survival outcomes (7-15). SMARCA4-deficient undifferentiated tumors (SMARCA4-UT) are a rare and highly malignant tumor type, distinct from SD-NSCLC. However, prognostic findings for SMARCA4-UT are inconsistent. Zhou et al.’s study did not find a significant difference in survival rates compared to SD-NSCLC, while reports by Duan et al. and Longo et al. have documented cases suggesting the potential efficacy of immunotherapy (16-22). These findings underscore the necessity of identifying critical prognostic factors, including genetic, molecular, and immunologic determinants, to improve risk stratification and guide therapeutic strategies (9,23,24).

In recent years, immunotherapy, particularly immune checkpoint inhibitors (ICIs), has transformed the treatment landscape for advanced NSCLC. However, the efficacy of ICIs in SD-NSCLC remains poorly understood, with limited and conflicting clinical evidence. SD-NSCLC often exhibits an immune-desert phenotype, characterized by low programmed death-ligand 1 (PD-L1) expression, low tumor mutation burden (TMB), and reduced immune cell infiltration in the tumor microenvironment (TME) (9). These features are traditionally associated with poor responses to ICIs, raising concerns about their utility in SD-NSCLC. Nevertheless, preclinical studies have highlighted that SMARCA4 deficiency creates vulnerabilities in the TME, including a dependence on epigenetic regulators like EZH2, which could potentially be exploited to enhance the efficacy of ICIs (25). Mechanistically, SMARCA4 deficiency disrupts the SWI/SNF chromatin remodeling complex, leading to chromatin compaction, impaired antigen presentation, and reduced PD-L1 expression (26). These changes further exacerbate immune suppression by modulating tumor-associated regulatory T cells (Tregs) and macrophages (TAMs), thereby creating a more immunosuppressive TME (27).

Despite these insights, the specific mechanisms remain unknown, and significant gaps persist in understanding the prognostic and therapeutic implications of SMARCA4 deficiency in NSCLC. The limited data on the efficacy of ICIs in SD-NSCLC further complicate clinical decision-making. This study focuses on SD-NSCLC, aiming to integrate evidence regarding the prognostic significance of SMARCA4 deficiency, identify key factors influencing treatment outcomes, evaluate the potential of immunotherapy in SD-NSCLC, and explore the molecular mechanisms through public databases. By advancing these aspects, this study aims to provide more effective and personalized treatment strategies for this challenging subtype of lung cancer. We present this article in accordance with the STROBE reporting checklist (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-921/rc).

Methods

Clinical cohort study design

Patients and study design

From October 2020 to May 2025, Tianjin Medical University Cancer Institute & Hospital identified 149 tumor patients with SMARCA4/BRG1 deficiency through immunohistochemistry (IHC). The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of Tianjin Medical University Cancer Institute & Hospital (No. bc20254853) and individual consent for this retrospective analysis was waived. According to the fifth edition of the World Health Organization (WHO) Classification of Tumors of the Thorax, the absence of any BRG1-positive signal in the nuclei of all tumor cells is defined as loss of BRG1 protein, provided reliable internal controls are available. The immunohistochemical results of SD-NSCLC are negative for BRG1 and SALL4, and positive for CK7, Napsin A, and TTF-1 (as shown in Figure 1), while SMARCA4-deficient undifferentiated tumors typically show negative expression of CK7, Napsin A, and TTF-1. Ultimately, 95 patients with SD-NSCLC who met the inclusion criteria were included in this study. The flowchart is shown in Figure 2A. The collected data included age, gender, smoking history, lung metastasis, liver metastasis, bone metastasis, clinical stage, treatment, and other relevant clinical factors. In patients with stage III and IV disease, progression-free survival (PFS) is defined as the time from the start of first-line treatment to progressive disease (PD) or death from any cause, whichever occurs first. In patients with stage I and II disease, disease-free survival (DFS) refers to the time from the start of treatment to tumor recurrence, metastasis, or death from any cause (whichever occurs first). DFS and PFS were the primary endpoints of the study. In addition, the disease control rate (DCR) was a secondary endpoint of the study. The data cutoff date was May 15, 2025.

Figure 1 The immunohistochemical images of SMARCA4-deficient non-small cell lung cancer (20×). (A) BRG1 negative; (B) CK7 positive; (C) Napsin A positive; (D) SALL4 negative; (E) TTF-1 positive.

Figure 2 Clinical and TCGA cohorts. (A) Flowchart of clinical cohort. (B) Flowchart of TCGA cohort. LOF, loss-of-function; LUAD, lung adenocarcinoma; LUSC, lung squamous cell carcinoma; NSCLC, non-small cell lung cancer; SD-NSCLC, SMARCA4-deficient non-small cell lung cancer; TCGA, The Cancer Genome Atlas.

Genetic analysis

The genetic testing information is sourced from the Center for Precision Oncology and Translational Research at Tianjin Medical University Cancer Institute & Hospital. The test covers tissue, blood, and pleural effusion samples, utilizing next-generation sequencing (NGS) technology and highly specific capture probes to identify 425 gene regions.

The Cancer Genome Atlas (TCGA) cohort study design

Data sources and re-grouping

The downloaded dataset is TCGA Genomic Data Commons (GDC) data portal version 42.0, downloaded on May 26, 2025. It includes patients with complete OS and PFS data, comprising 496 patients with lung adenocarcinoma (LUAD) and 491 patients with lung squamous cell carcinoma (LUSC). SMARCA4 class 1 mutations include nonsense mutations, frameshift mutations, splice site mutations, and large-scale deletions/insertions. These mutations are LOF mutations, leading to the complete absence of the SMARCA4 protein or the production of truncated, non-functional proteins. Type 2 mutations include missense mutations (particularly those occurring at key sites in the ATPase domain) and certain small in-frame deletions/insertions. These mutations typically result in the production of structurally abnormal, partially dysfunctional SMARCA4 proteins, with impaired ATPase activity or chromatin remodeling function, though the protein itself may still be present. Since missense mutations cannot clearly indicate whether the protein is missing, we excluded missense mutations from the analysis. We defined one type of mutation (nonsense, frameshift, and splice site deletion) as SMARCA4-LOF. In contrast, the SMARCA4-non-LOF group included silent mutations and wild-type mutations. Correlation analysis was performed on the two groups. The flowchart is shown in Figure 2B.

Survival analysis

PFS was the primary endpoint. Survival analysis software included SPSS 25, R Studio, and R version 4.4.2.

Differential gene expression analysis

Using R software (version 4.4.2) and the “DESeq2” package, a Wilcoxon rank-sum test was performed to analyze gene expression differences between the two groups. The “ggplot2” package was used for visualization, with P<0.05 indicating statistically significant differences.

Gene Set Enrichment Analysis (GSEA) enrichment analysis

The GSEA method was used to explore the biological functional enrichment patterns of expressed genes. The clusterProfiler package was used to load custom gene sets (c2.cp.biocarta.kegg.pid.reactome.wikipathways.v2023.1.Hs.symbols, c5.go.v2024.1.Hs.symbols, sourced from the MSigDB database). GSEA enrichment analysis was performed and visualized using the enrichplot package, and pathway bubble plots were generated using gseaplot2.

Immune infiltration analysis

Using the CIBERSORT algorithm in R software, the immune cell abundance in the SMARCA4-LOF mutant and SMARCA4-non-LOF wild-type dataset samples was calculated. The enrichment levels and differences of 22 immune cell types between the two patient groups were assessed. Visualization was performed using the “heatmap” package, “corrplot” package, and “vioplot” package. Box plots and heatmaps of the 22 immune cell types in the mutant and wild-type groups were generated, with differences considered statistically significant when P<0.05.

Statistical analysis

Baseline differences between groups were assessed using either Pearson’s χ² test or Fisher’s exact test, as appropriate. Survival rates were analyzed using Kaplan-Meier and Cox proportional hazards regression models. The Cox regression model yielded hazard ratio (HR) with corresponding 95% confidence interval (CI). All statistical tests were two-tailed, and a P value of <0.05 was considered statistically significant. Statistical analyses were performed using SPSS 25, R Studio, R version 4.4.2.

Results

Clinical cohort results

Analytical results of 95 SD-NSCLC patients

The median follow-up time for 95 patients with SD-NSCLC was 13.0 months (95% CI: 11.374–14.626). We collected data from a total of 95 SD-NSCLC patients, and their clinical characteristics are shown in Table S1. The baseline characteristics of the 95 SMARCA4-deficient NSCLC patients suggest that the disease predominantly affects males (89.5%) and individuals with a history of smoking (85.3%). Age is treated as a continuous variable, with 64 years as the cutoff value. The cohort had a slightly higher age distribution, with 53.7% of patients aged ≥64 years. For the detected TMB status, the high mutation status accounted for 11/14.

NGS results from tissue DNA of 40 patients have been obtained. The most common mutations were found in the TP53 gene (70%, 28/40), followed by the SMARCA4 gene (35%, 14/40), LRP1B gene (25%, 10/40), CDKN2A gene (20%, 8/40), KRAS gene (17.5%, 7/40), NF1 (17.5%, 7/40), STK11 (17.5%, 7/40), and KEAP1 gene (15%, 6/40), as shown in Figure 3A. In patients with SMARCA4 mutations (Figure 3B), the mutation types included nonsense mutations (35.7%, 5/14), splice site mutations (35.7%, 5/14), and frameshift mutations (28.6%, 4/14).

Figure 3 Molecular characteristics of SD-NSCLC patients. (A) Genetic mutation characteristics of 40 SD-NSCLC patients. (B) Pie chart showing mutation types in 14 patients with SMARCA4 gene mutations. SD-NSCLC, SMARCA4-deficient non-small cell lung cancer.

Among the 40 patients with stage I and II SD-NSCLC, surgery was the primary treatment method (72.5%), particularly more common in stage I (76.5% vs. 50.0% in stage II). Combined treatment regimens were relatively rare (5.0–7.5%), with neoadjuvant chemotherapy and immunotherapy applied only in stage I (8.8%). Only one patient (stage II) received triple therapy (ICIs + chemotherapy + targeted therapy) (Table S2). Among the 36 SMARCA4-deficient NSCLC patients who underwent radical surgery, 19.4% (7/36) experienced recurrence, with a median DFS of 7.5 months. The majority (80.6%) remained recurrence-free, with a median follow-up of 13 months and the longest DFS reached 32.5 months (Table S3).

The first-line treatment strategies for stage III and IV patients are shown in Table S4. For stage IV SD-NSCLC patients receiving systemic therapy, they were divided into the ICIs group and non-ICIs group based on whether or not they received immunotherapy. After applying Fisher’s exact test to the ICIs and non-ICIs groups, it was confirmed that the baseline characteristics of the two groups were comparable (Table 1). Survival analytical results showed a difference in median progression-free survival (mPFS) between the two groups of patients, but it was not statistically significant (ICIs group: 6.5 months, 95% CI: 5.386–7.614; non-ICIs group: 4.5 months, 95% CI: 0.206–8.794; P=0.60) (Figure 4). For efficacy assessment, since there were no partial responses (PR) or complete responses (CR), only stable disease (SD) was observed, so the DCR was selected. The SD rate in the ICIs group was 7/27 (25.9%), and the SD rate in the non-ICIs group was 1/5 (20%), but there was no significant difference (P>0.99). In the immunotherapy group (ICIs, N=27), the vast majority of patients (26 cases, 96.3%) received PD-1 inhibitors, and only 1 case (3.7%) received a PD-L1 inhibitor. We set the threshold based on the standard protocols of advanced clinical trials: the maximum number of cycles for immune combination chemotherapy is typically 4 cycles, while the maximum number of cycles for subsequent immune maintenance therapy is 35 cycles. The data distribution in this study closely aligns with this clinical practice. The results showed that 59.3% (16 cases) of patients received 1–4 cycles of treatment, characteristic of the combination chemotherapy phase; 37.0% (10 cases) of patients received 5–35 cycles of treatment, reflecting the maintenance therapy phase; and only 1 case (3.7%) received more than 35 cycles of long-term treatment. Perform survival analysis on patients with STK11/KEAP1 mutations and wild-type patients among those receiving systemic treatment in stage IV. According to the results of the survival analysis, there was a significant difference in survival time between the STK11/KEAP1 mutant and wild-type groups (P=0.007). The mean PFS for the wild-type group was 7.138 months (95% CI: 4.587–9.688), with a median PFS (mPFS) of 6.5 months (95% CI: 4.242–8.758). The mean PFS for the mutant group was 1.75 months (95% CI: 0.584–2.916), and the mPFS was 1 month (95% CI: 0–2.960) (Figure 5). Univariate analyses suggested that STK11/KEAP1 mutation and lung metastasis are risk factors for poor prognosis (P=0.02). Even after multivariate analysis, STK11/KEAP1 mutation and lung metastasis remained independent risk factors for poor prognosis (P=0.02). In addition, lung metastasis was identified as an independent poor prognostic factor in both univariate and multivariate analyses, with P values of 0.047 and 0.043, respectively (Figure 6).

Figure 4 Kaplan-Meier curve of PFS in patients with stage IV SD-NSCLC receiving systemic treatment with ICIs and non-ICIs. ICIs, immune checkpoint inhibitors; PFS, progression-free survival; SD-NSCLC, SMARCA4-deficient non-small cell lung cancer.

Figure 5 Kaplan-Meier curve of PFS in patients with stage IV SD-NSCLC, according to STK11 and KEAP1 mutation status. mut, mutant-type; PFS, progression-free survival; SD-NSCLC, SMARCA4-deficient non-small cell lung cancer; wt, wild type.

Figure 6 Forest plot of PFS in the Cox proportional-hazards model. (A) Univariate model. (B) Multivariate model. CI, confidence interval; HR, hazard ratio; Mets, metastasis; PFS, progression-free survival.

Next, to investigate whether ICIs can mitigate the harm caused by STK11/KEAP1 mutations, we selected STK11/KEAP1-mut and wt patients who had undergone immunotherapy for survival comparison. The results showed a significant difference in PFS between the STK11/KEAP1 wild-type and mutant groups (P=0.02). The mPFS for the wild-type group was 6.0 months (95% CI: 3.014–8.986), significantly longer than the 1.0 months (95% CI: 0.000–2.960) for the mutant group (Figure 7).

Figure 7 Kaplan-Meier PFS curve for ICIs-treated patients with stage IV SD-NSCLC, according to STK11 and KEAP1 mutation status. ICIs, immune checkpoint inhibitors; mut, mutant-type; PFS, progression-free survival; SD-NSCLC, SMARCA4-deficient non-small cell lung cancer; wt, wild type.

Analytical results of SD and non-SD treated with ICIs

To investigate the hazards of SMARCA4 deficiency versus non-deficiency under ICIs, we designed the following analysis workflow. In immunohistochemical staining, when the internal positive control is present and staining is effective, if BRG1 expression is completely absent in the tumor cell nuclei, the subtype of NSCLC can be defined as SMARCA4-deficient NSCLC. If clear BRG1 positive expression is detected in any proportion of tumor cell nuclei, the subtype of NSCLC can be defined as non-SD NSCLC. The grouping of SD and non-SD is shown in Figure S1. There were 30 patients in the SD group and 60 patients in the non-SD group. The median follow-up time for the SD group and the non-SD group was 24.0 months (95% CI: 19.35–28.65). The baseline characteristics of the SD and non-SD groups are presented in Table 2. In terms of gender distribution, the majority of patients were male (81.1%), with 76.7% of males in the non-SD group and 90.0% in the SD group. Regarding age, a higher proportion of patients over 64 years old was observed in the SD group (70.0%) compared to the non-SD group (53.3%). Smoking history was more prevalent in the SD group (90.0%) compared to 71.7% in the non-SD group. Regarding immunotherapy regimens, most patients received ICIs combined with chemotherapy (76.7%), with the majority of ICIs being PD-1 inhibitors (95.6%). However, a significant difference was observed in the number of treatment cycles (P<0.001), as patients in the SD group developed PD after fewer cycles of ICI treatment. Overall, the two groups are comparable.

There was a significant difference in PFS between the SD group and the non-SD group (P=0.003). The mPFS in the SD group was 6.0 months (95% CI: 3.552–8.448), while the mPFS in the non-SD group was 11.0 months (95% CI: 7.747–14.253) (Figure S2). The SD rate in the SD group was 8/30 (26.7%), and the SD rate in the non-SD group was 17/60 (28.3%). The OR between the SD group and the non-SD group was approximately 1.09, but there was no significant difference (P=0.86). In the univariate analysis (Figure S3A), the HR for the SD group was 2.129 (95% CI: 1.247–3.635, P=0.006), indicating that SD status increased the risk by approximately 2.13 times. After adjusting for confounding factors such as gender, age, smoking history, and metastasis site, multivariate analysis (Figure S3B) still showed that the HR for the SD group was 2.243 (95% CI: 1.256–4.005, P=0.006), further confirming that SD is an independent adverse prognostic factor.

TCGA cohort analytical results

Re-grouped into SMARCA4-LOF and SMARCA4-non-LOF

A total of 987 lung cancer patients were retrieved from the TCGA database, including LUAD (N=496) and LUSC (N=491), which were classified according to SMARCA4 status as shown in Table 3. The number of participants in the SMARCA4-LOF and SMARCA4-non-LOF groups is shown in Table S5. In LUAD, 19 cases of SMARCA4-LOF patients were included, while only 6 cases were included in LUSC. Analyzing these groups separately would lead to unreliable GSEA results. Our aim was to investigate the molecular mechanisms underlying poor prognosis due to BRG1 protein dysfunction. Therefore, we conducted a combined analysis.

Prognosis analysis

After matching for gender, age, and clinical stage, survival analysis for the two groups is shown in Figure 8. The mPFS in the SMARCA4-non-LOF group was 57.56 months (95% CI: 47.47–67.65). The mPFS in the SMARCA4-LOF group was significantly shorter, at 18.66 months (95% CI: 4.07–33.25). The results demonstrated a statistically significant difference between the two groups (P=0.02).

Figure 8 Kaplan-Meier survival analysis of PFS in lung cancer patients based on SMARCA4-LOF status. PFS, progression-free survival; SMARCA4-LOF, SMARCA4-loss-of-function.

Gene differential expression analysis and enrichment analysis

Volcano plots of differentially expressed genes between the SMARCA4-LOF and SMARCA4-non-LOF groups, as well as GSEA enrichment analysis based on differentially expressed genes, are shown in Figure 9. Due to the excessive number of genes in the control group, a gene heatmap could not be generated. As an alternative, we created a volcano plot, as shown in Figure 9A. The differential gene expression analysis indicates that there are 596 upregulated genes and 1,161 downregulated genes.

Figure 9 Differential gene expression and GSEA enrichment analysis. (A) Volcano plot; (B) GO functional annotation, (C) KEGG pathway enrichment, (D) WP enrichment, (E) Reactome pathway enrichment, (F) Hallmark enrichment. GO, Gene Ontology; GSEA, Gene Set Enrichment Analysis; KEGG, Kyoto Encyclopedia of Genes and Genomes; WP, Wiki Pathway.

After GSEA enrichment analysis, SD-NSCLC exhibits significant dysregulation in multiple key biological processes. The upregulated pathways primarily focus on angiogenesis, cell migration, and immune modulation, including processes involved in neovascularization and dendritic cell cytokine production, which may have immunosuppressive functions. These findings suggest a tendency toward metastasis promotion and immune evasion in the TME.

In contrast, the downregulated pathways are broadly related to metabolic disorders, protein homeostasis imbalance, and signaling abnormalities. Several metabolism-related pathways, such as the urea cycle, glutathione metabolism, and NAD+ synthesis, are generally suppressed, reflecting vulnerabilities in ammonia detoxification, oxidative stress response, and energy metabolism. Additionally, the widespread downregulation of the proteasome and its related complexes indicates impaired intracellular protein degradation capability. In terms of signaling, significant suppression of the AMPK-LKB1-mTOR pathway, Wnt/β-catenin pathway, and p53-related regulatory pathways points to a molecular foundation for cell cycle dysregulation, apoptosis resistance, and metabolic reprogramming (Figure 9).

Immune infiltration analysis

Immune infiltration analysis is shown in Figure 10. In the SMARCA4-LOF group, there was a significant reduction in the proportion of resting dendritic cells and neutrophils, along with a notable decrease in the number of CD4⁺ memory resting T cells. In contrast, the composition of M0 macrophages significantly increased, accompanied by an elevation in Tregs (Figure 10A). The immune cell-related heatmap reveals significant negative correlations between various immune cells. The immunosuppressive effect of Tregs is notably negatively correlated with M2 macrophages (−0.21), resting natural killer (NK) cells (−0.17), and M1 macrophages (−0.01), suggesting that Tregs inhibit pro-inflammatory M2 macrophages. Additionally, the immunosuppressive network of M1 macrophages shows a weak negative correlation with M2 macrophages (−0.07) and a strong negative correlation with resting dendritic-activated cells (−0.29) (Figure 10B).

Figure 10 Immune cell infiltration analysis between SMARCA4-LOF and SMARCA4-non-LOF. (A) Box plots showing the distribution differences of various immune cell types. (B) Heatmap depicting the correlations between various immune cell types. *, P<0.05; **, P<0.01; ***, P<0.001. SMARCA4-LOF, SMARCA4-loss-of-function.

Discussion

This study focuses on patients with SD-NSCLC, examining the efficacy of ICIs and the risk factors influencing prognosis, while conducting a preliminary mechanistic exploration using the TCGA database. A comparison of the ICI and non-ICI groups in SD-NSCLC patients showed no significant difference in PFS (P=0.60). Further stratification based on STK11/KEAP1 mutations revealed that the group with co-mutations had worse PFS (P=0.007). Next, we explored whether immunotherapy could improve the prognosis associated with STK11/KEAP1 mutations in the ICI group, but the mutation group still showed worse prognosis (P=0.02), indicating that ICIs did not mitigate the negative impact of STK11/KEAP1 mutations. All patients in our study who were found to have STK11/KEAP1 mutations were treated with PD-1 inhibitors. However, studies have shown that while PD-(L)1 inhibitors cannot overcome resistance in patients with NSCLC harboring STK11/KEAP1 mutations, the combination of CTLA-4 and PD-(L)1 inhibitors in dual immune checkpoint blockade (ICB) therapy significantly improves treatment outcomes. CTLA-4 inhibition can reshape the immunosuppressive state in the TME, enhance the activity of CD4⁺ T cells and the antitumor killing function of myeloid cells, thereby alleviating resistance and bringing clinical benefits to this high-risk subgroup of patients. Therefore, dual ICB therapy may be an effective treatment for improving the prognosis of SD-NSCLC (28).

To investigate whether the difference in prognosis was due to gene co-mutations or the detrimental effects of SMARCA4/BRG1 protein deficiency, we compared PFS and DCR between 30 SD-NSCLC patients and 60 non-SD-NSCLC patients receiving ICIs. The results showed a significant difference in PFS between the two groups (P=0.003), but no significant difference in DCR (P=0.86). Univariate and multivariate analyses identified SMARCA4/BRG1 protein deficiency as a risk factor for poor prognosis (P=0.006). In contrast to our findings, Zhou et al. reported favorable immunotherapy outcomes in some SMARCA4-deficient cases (29), and similar results were found in other studies (12,30,31). While the findings of Dagogo-Jack et al. were similar to our study, STK11 and SMARCA4 mutations were ineffective for immunotherapy, suggesting outcome heterogeneity (32). Our analysis reinforces the prognostic importance of SMARCA4 mutations, emphasizing how treatment context shapes the clinical implications of specific co-mutation patterns. This is particularly evident for STK11/KEAP1 co-mutations, which emerged as uniquely significant in the immunotherapy setting. In our study, SMARCA4-deficient patients with STK11/KEAP1 mutations had a worse prognosis, even after treatment with ICIs. Prior research indicates a strong association between KEAP1 and SMARCA4 mutations in NSCLC, with this co-occurrence significantly reducing PD-1/PD-L1 inhibitor efficacy (26,33). The combined effects of SMARCA4, STK11, and KEAP1 mutations promote immune evasion and tumor progression through multiple mechanisms, emphasizing the need for integrated treatment approaches. Additionally, patients with SMARCA4 and KRAS co-mutations exhibited shorter survival across both immunotherapy and non-immunotherapy treatments (11,34). Consistent with prior studies (35,36), our findings highlight the poor prognosis of SMARCA4-deficient NSCLC due to these co-mutations.

To validate the harm caused by BRG1 protein dysfunction, we reclassified LUAD and LUSC patients in the TCGA database into LOF and non-LOF groups based on SMARCA4 protein function loss. The two groups showed significant differences in PFS (P=0.02). Differential gene expression analysis and pathway enrichment revealed that the LOF group had enriched pathways related to core metabolism (amino acids, energy, detoxification), proteasome function, key tumor suppressor signaling (p53, β-catenin degradation, PTEN), and DNA replication regulation. Gene expression silencing in the LOF group was associated with pathways linked to protein toxicity, metabolic stress, oxidative damage, genomic instability, and inactivation of tumor suppressor signaling. Immunological infiltration analysis suggested immune suppression, providing preliminary insight into the limited efficacy of ICIs.

This study, utilizing the TCGA database, found that SMARCA4 deficiency leads to the inactivation of the p53/PTEN signaling pathway and reduced immune infiltration. Recent studies support this finding, showing that SMARCA4 deficiency in cancer cells results in decreased immune infiltration and resistance to immunotherapy. Specifically, SMARCA4 loss leads to downregulation of innate immune components such as STING and IL1β, which in turn reduces dendritic cell and CD4⁺ T cell infiltration within the TME. This effect is attributed to cancer cell-intrinsic reprogramming of the enhancer landscape, which impacts genes involved in the innate immune response. Moreover, SMARCA4 and NF-κB co-occupy enhancers associated with immune-related genes, indicating a functional interplay between these two factors (37). However, contradictory findings have emerged in ovarian cancer models, where SMARCA4 loss is associated with increased tumor immunogenicity. In these models, SMARCA4 deficiency results in the upregulation of interferon-stimulated genes and antigen presentation machinery, leading to enhanced activation of cytotoxic T cells and NK cells (38). These conflicting results highlight the need for further investigation into the role of SMARCA4 in tumor immunity. Zhang et al. emphasized that SMARCA4 deficiency leads to the loss of SWI/SNF complex function, causing the PRC2 complex (through H3K27me3 modification) to continuously suppress the expression of tumor suppressor genes (such as p53 and PTEN). This is consistent with our TCGA Analytical results: the p53/PTEN signaling pathway is significantly downregulated in the SMARCA4-LOF group, accompanied by enrichment of genomic instability and metabolic stress pathways (39). Furthermore, our research revealed that the absence of SMARCA4 significantly suppressed the AMPK-LKB1-mTOR pathway in signal transduction. This finding aligns with the conclusion of a clinical trial, which demonstrated that SMARCA4 defects constitute a key mechanism underlying resistance to mTOR-targeted therapy (40).

Proteasome function and DNA replication regulation pathways are enriched in the LOF group, suggesting that tumors may rely on alternative repair pathways. Immune infiltration analysis showed reduced T cell activity in the LOF group. Studies have shown that SD-NSCLC exhibits increased FOXP3⁺ cell and neutrophil density, along with increased CD8⁺ T cell density, which may account for the limited efficacy of anti-PD-1 inhibitors (3). Wu et al. found that TLS-RGs significantly influence immune infiltration in the TME of LUAD, with patients in the high TLS score group (corresponding to an immune-activated state) exhibiting better outcomes. This aligns with our findings that SMARCA4/BRG1 deficiency leads to immune suppression (increased proportion of Treg cells), thereby limiting the efficacy of ICIs. Wu et al. also found that the TLS high-expression group (Cluster B) exhibited increased B cell, CD4⁺ T cell, and macrophage infiltration, which is associated with immune therapy response. In contrast, our study showed that the SMARCA4-deficient group exhibited similar immune suppression characteristics, suggesting that the immune status of the TME is a key factor influencing ICI efficacy (41).

However, there are several limitations in this study. First, as a retrospective design, although we used multivariate analysis to mitigate bias, information bias and confounding factors remain possible. Second, we chose PFS as the primary endpoint, which reflects early treatment efficacy and minimizes interference from subsequent therapies; however, it may not fully capture long-term outcomes. Third, the adjustment to define PFS starting from the initiation of immunotherapy helped reduce pseudoprogression misclassification, but it deviated from traditional definitions. Future research could adopt iRECIST criteria or prospective designs to address this limitation. Additionally, although studies have confirmed that SMARCA4 class 1 mutations are associated with significant loss of BRG1 protein (1,13,42), class 2 mutations may still lead to protein deficiency. In this study, we used class 1 mutations (i.e., LOF variants) from TCGA data as a surrogate for BRG1 protein loss, which may introduce some selection bias. Further studies are needed to validate and refine this approach. Finally, not all patients in this study underwent PD-L1 and TMB testing, which may introduce selection bias since patients with completed tests could have distinct clinical characteristics. Future studies should ensure comprehensive testing coverage, apply statistical methods such as multiple imputation to handle missing data, and integrate PD-L1 and TMB metrics into risk analyses. Subgroup analyses of patients with complete test data could further explore the associations between these biomarkers and immunotherapy efficacy, using real-world data or prospective studies for validation.

Conclusions

Future large-scale studies are needed to further validate these findings and provide more robust guidance for therapeutic strategies. Addressing the challenges posed by SMARCA4 deficiency remains essential for improving outcomes in high-risk NSCLC patients. These efforts will not only enhance our understanding of the underlying mechanisms but also pave the way for the development of combination therapies tailored to this aggressive tumor subtype.

Acknowledgments

We are grateful to Tianjin Medical University Cancer Institute & Hospital for providing the data that enabled this study to proceed smoothly.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-921/rc

Data Sharing Statement: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-921/dss

Peer Review File: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-921/prf

Funding: This study was funded by Tianjin Key Medical Discipline (Specialty) Construction Project (No. TJYXZDXK-009A), the National Natural Science Foundation of China (No. 81702268), the Natural Science Foundation of Tianjin (No. 18JCYBJC93400), Beijing Huikang Ren’ai Public Welfare Foundation (No. HKRA2025050120) and Wu Jieping Medical Foundation (No. 320.6750.2025-16-5).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-921/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. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of Tianjin Medical University Cancer Institute & Hospital (No. bc20254853) and individual consent for this retrospective analysis was waived.

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. Nambirajan A, Singh V, Bhardwaj N, et al. SMARCA4/BRG1-Deficient Non-Small Cell Lung Carcinomas: A Case Series and Review of the Literature. Arch Pathol Lab Med 2021;145:90-8. [Crossref] [PubMed]
2. Liang X, Gao X, Wang F, et al. Clinical characteristics and prognostic analysis of SMARCA4-deficient non-small cell lung cancer. Cancer Med 2023;12:14171-82. [Crossref] [PubMed]
3. Velut Y, Decroix E, Blons H, et al. SMARCA4-deficient lung carcinoma is an aggressive tumor highly infiltrated by FOXP3+ cells and neutrophils. Lung Cancer 2022;169:13-21. [Crossref] [PubMed]
4. Alessi JV, Elkrief A, Ricciuti B, et al. Clinicopathologic and Genomic Factors Impacting Efficacy of First-Line Chemoimmunotherapy in Advanced NSCLC. J Thorac Oncol 2023;18:731-43. [Crossref] [PubMed]
5. Agaimy A. SWI/SNF-deficient Sinonasal Carcinomas. Adv Anat Pathol 2023;30:95-103. [Crossref] [PubMed]
6. Lissanu Deribe Y, Sun Y, Terranova C, et al. Mutations in the SWI/SNF complex induce a targetable dependence on oxidative phosphorylation in lung cancer. Nat Med 2018;24:1047-57. [Crossref] [PubMed]
7. Liu L, Ahmed T, Petty WJ, et al. SMARCA4 mutations in KRAS-mutant lung adenocarcinoma: a multi-cohort analysis. Mol Oncol 2021;15:462-72. [Crossref] [PubMed]
8. Wang X, Wang Y, Xie M, et al. Hypermethylation of CDKN2A CpG island drives resistance to PRC2 inhibitors in SWI/SNF loss-of-function tumors. Cell Death Dis 2024;15:794. [Crossref] [PubMed]
9. Wang H, Shan Q, Guo J, et al. PDL1 high expression without TP53, KEAP1 and EPHA5 mutations could better predict survival for patients with NSCLC receiving atezolizumab. Lung Cancer 2021;151:76-83. [Crossref] [PubMed]
10. Wang H, Guo J, Shang X, et al. Less immune cell infiltration and worse prognosis after immunotherapy for patients with lung adenocarcinoma who harbored STK11 mutation. Int Immunopharmacol 2020;84:106574. [Crossref] [PubMed]
11. Boiarsky D, Lydon CA, Chambers ES, et al. Molecular markers of metastatic disease in KRAS-mutant lung adenocarcinoma. Ann Oncol 2023;34:589-604. [Crossref] [PubMed]
12. Alessi JV, Ricciuti B, Spurr LF, et al. SMARCA4 and Other SWItch/Sucrose NonFermentable Family Genomic Alterations in NSCLC: Clinicopathologic Characteristics and Outcomes to Immune Checkpoint Inhibition. J Thorac Oncol 2021;16:1176-87. [Crossref] [PubMed]
13. Schoenfeld AJ, Bandlamudi C, Lavery JA, et al. The Genomic Landscape of SMARCA4 Alterations and Associations with Outcomes in Patients with Lung Cancer. Clin Cancer Res 2020;26:5701-8. [Crossref] [PubMed]
14. Chen J, Zheng Q, Wang J, et al. Efficacy of immune checkpoint inhibitors in SMARCA4-deficient and TP53 mutant undifferentiated lung cancer. Medicine (Baltimore) 2024;103:e36959. [Crossref] [PubMed]
15. De Giglio A, De Biase D, Favorito V, et al. STK11 mutations correlate with poor prognosis for advanced NSCLC treated with first-line immunotherapy or chemo-immunotherapy according to KRAS, TP53, KEAP1, and SMARCA4 status. Lung Cancer 2025;199:108058. [Crossref] [PubMed]
16. Utsumi T, Taniguchi Y, Noda Y, et al. SMARCA4-deficient undifferentiated tumor that responded to chemotherapy in combination with immune checkpoint inhibitors: A case report. Thorac Cancer 2022;13:2264-6. [Crossref] [PubMed]
17. Al-Shbool G, Krishnan Nair H. SMARCA4-Deficient Undifferentiated Tumor: A Rare Malignancy With Distinct Clinicopathological Characteristics. Cureus 2022;14:e30708. [Crossref] [PubMed]
18. Li X, Tian S, Shi H, et al. The golden key to open mystery boxes of SMARCA4-deficient undifferentiated thoracic tumor: focusing immunotherapy, tumor microenvironment and epigenetic regulation. Cancer Gene Ther 2024;31:687-97. [Crossref] [PubMed]
19. Yang D, Wang Y. Imaging performance of thoracic SMARCA4-deficient undifferentiated tumor: a case report and literature review. Transl Lung Cancer Res 2024;13:443-52. [Crossref] [PubMed]
20. Longo V, Catino A, Montrone M, et al. Treatment of Thoracic SMARCA4-Deficient Undifferentiated Tumors: Where We Are and Where We Will Go. Int J Mol Sci 2024;25:3237. [Crossref] [PubMed]
21. Zhou P, Fu Y, Tang Y, et al. Thoracic SMARCA4-deficient tumors: a clinicopathological analysis of 52 cases with SMARCA4-deficient non-small cell lung cancer and 20 cases with thoracic SMARCA4-deficient undifferentiated tumor. PeerJ 2024;12:e16923. [Crossref] [PubMed]
22. Duan T, Xu M, Zhang H, et al. Long-term follow-up of combination therapy with pembrolizumab and anlotinib in thoracic SMARCA4-deficient undifferentiated tumor: a case report and molecular features. Front Oncol 2024;14:1453895. [Crossref] [PubMed]
23. Ricciuti B, Elkrief A, Lin J, et al. Three-Year Overall Survival Outcomes and Correlative Analyses in Patients With NSCLC and High (50%-89%) Versus Very High (≥90%) Programmed Death-Ligand 1 Expression Treated With First-Line Pembrolizumab or Cemiplimab. JTO Clin Res Rep 2024;5:100675. [Crossref] [PubMed]
24. Lamberti G, Spurr LF, Li Y, et al. Clinicopathological and genomic correlates of programmed cell death ligand 1 (PD-L1) expression in nonsquamous non-small-cell lung cancer. Ann Oncol 2020;31:807-14. [Crossref] [PubMed]
25. Italiano A, Soria JC, Toulmonde M, et al. Tazemetostat, an EZH2 inhibitor, in relapsed or refractory B-cell non-Hodgkin lymphoma and advanced solid tumours: a first-in-human, open-label, phase 1 study. Lancet Oncol 2018;19:649-59. [Crossref] [PubMed]
26. Di Federico A, De Giglio A, Parisi C, et al. STK11/LKB1 and KEAP1 mutations in non-small cell lung cancer: Prognostic rather than predictive? Eur J Cancer 2021;157:108-13. [Crossref] [PubMed]
27. Cooper AJ, Muzikansky A, Lennerz J, et al. Clinicopathologic Characteristics and Outcomes for Patients With KRAS G12D-Mutant NSCLC. JTO Clin Res Rep 2022;3:100390. [Crossref] [PubMed]
28. Skoulidis F, Araujo HA, Do MT, et al. CTLA4 blockade abrogates KEAP1/STK11-related resistance to PD-(L)1 inhibitors. Nature 2024;635:462-71. [Crossref] [PubMed]
29. Zhou H, Shen J, Liu J, et al. Efficacy of Immune Checkpoint Inhibitors in SMARCA4-Mutant NSCLC. J Thorac Oncol 2020;15:e133-6. [Crossref] [PubMed]
30. Liu H, Hong Q, Zheng S, et al. Effective treatment strategies and key factors influencing therapeutic efficacy in advanced SMARCA4-deficient non-small cell lung cancer. Lung Cancer 2024;198:108022. [Crossref] [PubMed]
31. Afshar S, Leili T, Amini P, et al. Introducing novel key genes and transcription factors associated with rectal cancer response to chemoradiation through co-expression network analysis. Heliyon 2023;9:e18869. [Crossref] [PubMed]
32. Dagogo-Jack I, Schrock AB, Kem M, et al. Clinicopathologic Characteristics of BRG1-Deficient NSCLC. J Thorac Oncol 2020;15:766-76. [Crossref] [PubMed]
33. Marinelli D, Mazzotta M, Scalera S, et al. KEAP1-driven co-mutations in lung adenocarcinoma unresponsive to immunotherapy despite high tumor mutational burden. Ann Oncol 2020;31:1746-54. [Crossref] [PubMed]
34. Negrao MV, Araujo HA, Lamberti G, et al. Comutations and KRASG12C Inhibitor Efficacy in Advanced NSCLC. Cancer Discov 2023;13:1556-71. [Crossref] [PubMed]
35. Pan M, Jiang C, Zhang Z, et al. Sex- and Co-Mutation-Dependent Prognosis in Patients with SMARCA4-Mutated Malignancies. Cancers (Basel) 2023;15:2665. [Crossref] [PubMed]
36. Ashok Kumar P, Graziano SL, Danziger N, et al. Genomic landscape of non-small-cell lung cancer with methylthioadenosine phosphorylase (MTAP) deficiency. Cancer Med 2023;12:1157-66. [Crossref] [PubMed]
37. Wang Y, Meraz IM, Qudratullah M, et al. Mutation of SMARCA4 Induces Cancer Cell-Intrinsic Defects in the Enhancer Landscape and Resistance to Immunotherapy. Cancer Res 2025;85:1997-2013. [Crossref] [PubMed]
38. Brodeur MN, Dopeso H, Zhu Y, et al. Interferon response and epigenetic modulation by SMARCA4 mutations drive ovarian tumor immunogenicity. Sci Adv 2024;10:eadk4851. [Crossref] [PubMed]
39. Zhang M, Dong Y, Meng R, et al. SMARCA4 Deficiency in Lung Cancer: From Signaling Pathway to Potential Therapeutic Targets. Genes Chromosomes Cancer 2025;64:e70022. [Crossref] [PubMed]
40. Middleton G, Robbins HL, Fletcher P, et al. A phase II trial of mTORC1/2 inhibition in STK11 deficient non small cell lung cancer. NPJ Precis Oncol 2025;9:67. [Crossref] [PubMed]
41. Wu S, Pan J, Pan Q, et al. Multi-omics profiling and experimental verification of tertiary lymphoid structure-related genes: molecular subgroups, immune infiltration, and prognostic implications in lung adenocarcinoma. Front Immunol 2024;15:1453220. [Crossref] [PubMed]
42. Shi M, Pang L, Zhou H, et al. Rare SMARCA4-deficient thoracic tumor: Insights into molecular characterization and optimal therapeutics methods. Lung Cancer 2024;192:107818. [Crossref] [PubMed]