Immunotherapy after EGFR‑TKI treatment in advanced non‑small cell lung cancer: Current status and future perspectives (Review)

Introduction

The major cellular components and mediators of the tumor microenvironment (TME), including cancer cells, immune cells [such as T cells, B cells, dendritic cells (DCs), myeloid-derived suppressor cells (MDSCs) and tumor-associated macrophages (TAMs)], stromal cells [cancer-associated fibroblasts (CAFs) and tumor-associated endothelial cells], cytokines and chemokines, tumor vasculature, lymphoid tissue, as well as adipocytes and exosomes, serve a critical role in cancer initiation, progression, spread and metastasis (1–3). Interactions between the tumor stroma, especially those mediated by CAFs and other stromal components, actively promote immune evasion and malignant growth in a variety of solid tumors, including lung and breast cancer as well as other epithelial malignancies (2). Tumor-infiltrating lymphocytes are a crucial part of antitumor immunity among the immune cells found in tumors and a recent study has demonstrated that immunotherapies can alter TIL activity as well as the interactions between immune and stromal cells in the TME. The TME usually contains signals that inhibit the immune response, such as regulatory T cells (Tregs) and myeloid-derived suppressor cells, which prevent the infiltration and killing function of immune cells (3). Breaking through this immunosuppression has long been the key to immunotherapy.

In non-small cell lung cancer (NSCLC) with epidermal growth factor receptor (EGFR) mutations, the TME manifests as an immunosuppressive phenotype. Compared with EGFR wild-type (wt) cancer, these tumors usually have a much lower tumor mutational burden, which results in a limited development of neoantigens that can successfully trigger immune recognition. Insufficient neoantigen presentation weakens T-cell priming and activation, resulting in a relatively ‘immune-cold’ environment. The blunting of antitumor immune activation under such conditions is primarily responsible for the lower therapeutic benefit observed with immune checkpoint inhibitors (ICIs) in this population (4,5). However, the TME in NSCLC can be affected by chemotherapy, radiotherapy or EGFR-tyrosine kinase inhibitors (TKIs). Short-term TKI treatment excels in tumor clearance and immunological upregulation, improving the overall survival and quality of life of patients (6). Nevertheless, the immunosuppressive TME gradually emerges with time, and EGFR-TKI resistance is unavoidable (7–9). This process reveals the dynamic evolution in tumor immune microenvironment induced by EGFR-TKI treatment (10).

Mechanisms of resistance to EGFR-TKIs

In Asian populations, over half of patients with NSCLC had EGFR-activating mutations (11). For individuals with locally progressed or metastatic NSCLC, EGFR-TKIs are now the standard therapy regimen (12,13). Their principal mechanism is competitive binding to the ATP-binding site inside the EGFR kinase domain, which inhibits kinase autophosphorylation and downstream signaling cascades. This effect inhibits tumor cell development, proliferation and metastasis (14). With improvements in medical research, numerous EGFR-TKIs from the first to third generations are now clinically available, markedly improving patient survival rates. However, despite initial success, nearly all patients develop acquired resistance to EGFR-TKI therapy, with a median progression-free survival (mPFS) of ~1 year (15,16). Resistance to EGFR-TKIs can be divided into two categories: Primary and acquired (16). Of patients with EGFR mutations, ~30% demonstrate primary resistance at the start of initial treatment, indicating no objective response to TKI therapy; however, the mechanisms causing this resistance remain unknown. However, acquired resistance refers to disease progression that occurs following an initial response to treatment. Its causes are complicated and diverse, consisting mostly of EGFR-dependent resistance, non-EGFR-dependent resistance (induced by activation of EGFR bypass or downstream signaling pathways) and histological or phenotypic alteration (17).

EGFR-dependent drug resistance

The T790M mutation accounts for 50–60% of acquired resistance in patients treated with first- and second-generation EGFR-TKIs (18). This mutation replaces a bulky methionine (M) with a threonine (T) at position 790 in exon 20 of the EGFR gene. This substitution creates steric hindrance between the aniline moiety of EGFR-TKIs and the drug-binding site within the ATP pocket of EGFR, thereby weakening drug-binding affinity. Additional mechanisms include a marked increase in ATP binding affinity for EGFR T790M, alterations in the catalytic domain and changes in overall conformational dynamics, collectively contributing to acquired resistance to TKIs (18,19).

With the widespread usage of the third-generation EGFR-TKI osimertinib, the C797S mutation has emerged as the predominant mode of resistance. The EGFR C797S mutation, located at position 797 in exon 20 of the EGFR gene, is a missense mutation in which serine replaces cysteine. It accounts for 10–26% of second-line osimertinib resistance cases and 7% of first-line resistance cases (20). This mutation damages the ATP-binding pocket, preventing third-generation TKIs from making covalent connections with the ATP-binding domain and so losing their inhibitory function (21). The C797S mutation commonly coexists with EGFR T790M in two structural forms: Cis (EGFR T790M and C797S occur on the same allele) and trans (EGFR T790M and C797S occur on different alleles). In ~85% of instances, the EGFR C797S/T790M mutation is in the cis configuration, with ~10% of patients having the C797S/T790M trans configuration. Whether the C797S and T790M mutations form a cis structure has important biological implications since it influences the therapeutic efficacy of subsequent TKIs (17,22). In addition to the most prevalent C797S mutation, C797G is another missense mutation at the same location that causes drug resistance by affecting the binding of the drug to the residue.

Aside from the classic T790M and C797S mutations, uncommon mutations at other places within the EGFR kinase domain can also cause resistance to osimertinib, mostly through interference with drug-kinase binding. Mutations at the L718 site (such as L718Q and L718V) and the adjacent G719A mutation primarily interfere spatially, preventing the critical alanine ring in the osimertinib molecule from forming a stable bond with its binding pocket. L792 mutations (most commonly and notably L792H) directly disrupt the tight binding of osimertinib to the kinase domain. Mutations at the G796 site sterically clash with the solvent-front aromatic ring of osimertinib, preventing kinase domain binding. The G724S mutation is located in the ATP-binding region and may interfere with osimertinib binding to its target by a variety of mechanisms, including generating protein structural changes, increasing ATP affinity or maintaining the kinase activation state, resulting in drug resistance. Another form of resistance is target upregulation caused by EGFR gene amplification (17,20,23).

EGFR-independent drug resistance

Activation of bypasses and downstream pathways, such as HER2, HER3, MET, KRAS, NRAS proto-oncogene, GTPase (NRAS), BRAF, phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit α (PIK3CA), AXL receptor tyrosine kinase (AXL) and insulin-like growth factor-1 receptor (IGF-1R), were often classified as ‘bypass’ mechanisms of resistance. These pathways enable tumor cells to activate alternative routes that engage key EGFR effectors essential for tumor cell growth and survival. The most common bypass mechanism of resistance to first- and second-generation EGFR TKIs is HER2 amplification (24,25). Downstream signaling pathways activated following receptor phosphorylation included the mitogen-activated protein kinase pathway, phosphatidylinositol 3-kinase (PI3K)/Akt pathway, phospholipase Cγ1, protein kinase C and various transcriptional regulators that modulate gene expression (26).

The MET gene encoded a receptor tyrosine kinase known as c-Met, or hepatocyte growth factor receptor, which supports cancer cell growth and survival by activating the HER3-PI3K/AKT and RAS/RAF/MEK/ERK signaling pathways, thereby bypassing the inhibitory effects of EGFR receptor signaling (18,27). The Ras-Raf-MEK-ERK pathway is crucial for regulating cell cycle progression and proliferation. KRAS and NRAS, both members of the Ras family, could activate MEK and ERK through this signaling pathway, contributing to drug resistance (28,29). BRAF mutations, although rare in EGFR-mutant (mt) NSCLC, involve kinases located downstream of RAS in the Ras-Raf-MEK-ERK pathway and may contribute to the activation of the GFR/RAS/RAF signaling pathway (30). PIK3CA, a catalytic subunit of the PI3K family of lipid kinases, serves an oncogenic role in lung adenocarcinoma by activating the PI3K/AKT/mTOR pathway. This signaling pathway regulates key cellular functions, including growth, proliferation, metabolism, angiogenesis and metastasis, and is frequently mutated or hyperactivated in various cancers. Patients with PIK3CA mutations generally exhibit a worse prognosis and shorter median survival compared with those without these mutations (31,32). IGF-1R and AXL serve key roles in regulating cell growth, differentiation, apoptosis, transformation and other critical physiological processes through the downstream PI3K/AKT signaling pathway, thereby contributing to the development of secondary drug resistance (33,34).

Histology and phenotypic conversion

Another mechanism of drug resistance involves histological transformation within the tumor itself. Specifically, some EGFR-mt lung adenocarcinomas develop into small cell lung cancer (SCLC), accounting for 3–14% of acquired resistance cases. This transformation is confirmed by histopathological biopsy, and transformed tumors typically respond to standard SCLC treatment regimens (35–38). Research indicates that the co-deletion of Rb1 and TP53 genes constitutes the key molecular basis driving such transformation (16,39).

Tumor cells can also undergo epithelial-mesenchymal transition (EMT) and develop treatment resistance. During this process, epithelial markers are downregulated and mesenchymal markers are upregulated, resulting in increased proliferation, invasion, migration and metastatic potential. However, the exact processes underpinning EMT are unknown; this process may be driven by AXL through activation of the PI3K/AKT signaling pathway (40,41) (Fig. 1).

Figure 1. The illustration outlines the key mechanisms of resistance to EGFR-TKIs in EGFR-mutant NSCLC. EGFR-dependent resistance arises from evolution of the target itself (e.g., T790M/C797S mutations or gene amplification), which restores EGFR kinase activity and re-activates the downstream PI3K/Akt/mTOR and Ras/Raf/MEK/ERK pathways. EGFR independent resistance occurs via bypass activation or downstream nodal mutations, circumventing EGFR inhibition and ultimately converging on the same core pathways. In addition, tumors may evade therapy through histology or phenotypic transformation, such as transformation to SCLC or EMT. EGFR, epidermal growth factor receptor; TKIs, tyrosine kinase inhibitors; NSCLC, non-small cell lung cancer; SCLC, small cell lung cancer; EMT, epithelial-mesenchymal transition.

Tumor-infiltrating immune cells

CD4+/CD8+ T cells/Treg cells

Infiltration of CD8+ T cells is lower in EGFR-mt lung adenocarcinoma (LA) compared with in EGFR-wt LA, and T cell proliferation is inhibited (42,43). It has been shown that after TKI treatment, immune cell infiltration is increased and the antitumor response is enhanced in EGFR-mt patient samples (44).

However, changes in the TME by TKI treatment appear to be dynamic. TKI treatment induces the recruitment of CD8+ and CD4+ T cells (45). With short-term TKI treatment, the expression levels of CD8 and granzyme B (GB) in EGFR-TKI-treated T cells co-cultured with tumor cells initially increases and then decreases with the duration of treatment, and T cells infiltration shifts from enhanced to suppressed (10). Evidence has shown that following TKI resistance, the tumor can transform into a ‘hot’ tumor with increased immune cell infiltration, with notable effector and T cell infiltration (46).

In the early stages of treatment, sensitive EGFR-TKIs can increase the number of cytotoxic CD8+ T cells, while short-term EGFR inhibition reduces the proportion of Foxp3+ Tregs. However, these changes in the TME, which favor immune-mediated combination therapies for cancer, may gradually diminish with continued treatment (47).

TAMs

TAMs serve an important role in tumor progression in EGFR-mt NSCLC, and their enrichment in human NSCLC is associated with poor clinical outcomes (48). How M1 and M2 type macrophages are distinguished may be influenced by the TME and macrophages may exhibit a mixed phenotype of both types of markers.

EGFR mutations promote the expansion of alveolar macrophages (AM), but TKI treatment markedly reduces AM numbers in tumor-bearing mice (48). In EGFR-mt cells, M2 macrophages increase, and while short-term TKI treatment reduces M2 infiltration, the number of M2 macrophages rise again with the onset of TKI resistance (10,45).

DCs

EGFR-mt lung cancer (LC) drives the immune phenotype of tumor-infiltrating DCs (TIDC) toward an immunosuppressive direction and is closely associated with exosomes. Tumor-derived exosomes (TEXs) bridge the interaction between tumor cells and immune cells, thereby altering antitumor immune responses. Research has shown that TEXs can inhibit myeloid differentiation, promoting the transformation of monocytes from immune-stimulating to immune-suppressive cells, thereby supporting immune evasion (49). There is also evidence that EGFR activation by TEXs weakens the innate immunity of the host (50) and promotes tumor metastasis (51).

Compared with EGFR-wt tumor-bearing mice, TIDCs and lymph node (LN) DCs isolated from EGFR-19del tumor-bearing mice produced markedly less IL-12p40. This finding suggests that EGFR-19del Lewis LC tumors drive the process of immunosuppression by affecting DCs in both the tumor and LN. Furthermore, EGFR-mt LC cells may influence DC function by secreting exosomes in vitro. These exosomes not only accelerate tumor growth but also induce immune suppression (43). However, short-term TKI treatment markedly increases DC infiltration within the TME (47).

Natural killer (NK) cells

In EGFR-mt tumors, both the innate and adaptive lymphocyte compartments exhibit signs of functional exhaustion. The proportion of cytotoxic NK cells is reduced, whereas NK T cell (NKT) subsets with low cytotoxic potential are markedly increased (52,53). These NKT subsets display diminished expression of activation and cytotoxicity-related genes, indicating weakened innate immune cytotoxicity within the EGFR-mt TME.

In EGFR-mt LC, TKI therapy increases NK cell infiltration and cytotoxicity (44,54). Elevated IL-6 levels are linked to decreased immune cell infiltration during short-term TKI treatment. On the other hand, NK cell activity is decreased in EGFR-mt NSCLC with acquired TKI resistance due to increased upregulation of IL-6 (55).

B cells

Tertiary lymphoid structures (TLS) are primarily composed of B lymphocytes, which are essential to their development (53). TLS are associated with a higher response to immunotherapies and increased survival. In the TME of EGFR-negative LUAD, B cells build the TLS (56). Tumor-infiltrating B cells and antigen presentation-related markers were notably lower in EGFR-mt tumors compared with in EGFR-wt tumors, according to single-cell transcriptome analysis, which resulted in decreased T-cell activation (52). Immune infiltration, including B cells and CD8+ T cells, is increased in TKI-responsive samples but not in resistant ones after EGFR-TKI treatment, indicating that combining ICIs may be more advantageous prior to the development of TKI resistance (44).

Immunomodulatory molecules

Programmed death ligand 1 (PD-L1)

PD-L1 is a critical immune checkpoint protein that is broadly expressed on the surface of tumor cells and tumor-infiltrating immune cells. It inhibits anticancer immune responses by binding to programmed cell death protein-1 (PD-1) on T cells, B cells, DCs and NK T cells, and is an indicator of poor survival in most advanced cancers (57–59). It is regulated in NSCLC by two distinct mechanisms: Driver genetic alterations and inflammation. Previous research has reported the dynamic relationship between EGFR mutation status and PD-L1 expression before and after EGFR-TKI, but the exact relationship remains controversial. Meanwhile, it is worth noting that there may be an association between the levels of PD-L1 expression and the efficacy of EGFR-TKI therapy (60).

Compared with EGFR-wt, EGFR-mt typically results in upregulation of PD-L1 (59,61–65), a process that is closely associated with activation of the IL-6/JAK/STAT3 and p-ERK1/2/p-c-Jun signaling pathways (66,67). However, one clinical study found that EGFR-mt patients exhibit relatively low PD-L1 expression (16%; PD-L1 ≥5%) before TKI treatment (68). An analysis involving 18 studies and 3,969 patients revealed that, compared with EGFR-wt tumors, EGFR-mt NSCLC is less likely to display PD-L1 positivity (65).

EGFR-TKI not only directly inhibits the activity of tumor cells, but also indirectly enhances the antitumor immune response by downregulating PD-L1 (61,66). After short-term TKI treatment, PD-L1 expression is markedly reduced (10), which can be monitored by immuno-positron emission tomography imaging (69–71). This downregulation of PD-L1 may be partly dependent on the activation of the NF-kB pathway or AKT-STAT3 pathway (62,63). However, a previous study suggested the opposite conclusion, proposing that TKI treatment increases PD-L1 expression in patients with EGFR-mt NSCLC (72). As reported in one study (73), PD-L1 expression showed marked changes following EGFR-TKI treatment, with the proportion of patients with PD-L1 strong positive tumors [tumor proportion score (TPS) ≥50%] increasing from 14% at baseline to 28% after TKI treatment. With subsequent ICI treatment, patients with high PD-L1 expression (TPS ≥50%) achieved a longer mPFS compared with those with low PD-L1 expression [TPS <50%; 7.1 vs. 1.7 months; HR=0.18 (0.04–0.56); P=0.0033]. The PD-1 signaling pathway is activated in EGFR-TKI resistant tumors (74). After long-term TKI treatment, the expression levels of PD-1 typically increase (10,72). As EGFR-TKI resistance develops, the proportion of patients with PD-L1 strong positive tumors increases from baseline (73). A retrospective analysis showed that PD-L1 expression levels changed during the development of resistance in 16 patients (28%), of which 12 patients had higher PD-L1 expression levels after resistance (68). In addition, changes in PD-L1 expression levels in tumor-infiltrating immune cells between baseline and the development of EGFR-TKI resistance were also of interest, with the proportion of patients with ≥10% PD-L1 expression in tumor-infiltrating immune cells increasing from 11% at baseline to 25% after TKI treatment.

Evidence suggests that patients with EGFR-TKI resistance, particularly those who are T790M negative, may derive greater benefit from PD-1 inhibitors, in part due to their higher PD-L1 expression levels (75). This suggests that further research is needed to determine whether patients with EGFR-TKI resistance can benefit from PD-1 inhibitor therapy. Due to the potential efficacy of immunotherapy in patients with active PD-1 pathways, investigating the combined effects of immunotherapy and its interactions with TKIs after TKI resistance may help to optimize treatment strategies for these patients.

MHC-II, CD40, CD80 and CD86

LN DC and AM in EGFR-mt tumor-bearing mice exhibit an immunosuppressive phenotype compared with EGFR-wt tumor-bearing mice, characterized by downregulation of MHC-II and the costimulatory molecules CD40, CD80, and CD86 (43,48).

CD47

Integrin-associated protein (CD47) is a cell surface immunoglobulin-like molecule that inhibits phagocytosis by interacting with signal regulatory protein α on phagocytes. CD47 is selectively upregulated in patients with EGFR mutations. After treatment with TKI, CD47 expression is downregulated, which promotes DC phagocytosis of NSCLC cells. However, after the development of resistance in vitro, CD47 expression is upregulated (76).

Cytokines and chemokines

Cytokines and chemokines in the TME are also regulated by EGFR-TKIs. According to molecular research, cytokines can modulate immunological signaling by causing target-cell receptors to undergo structural or functional changes, such as receptor breakage or shedding (77–79). In patients with EGFR-mt NSCLC, the pro-inflammatory cytokine IL-17A was highly expressed, along with markedly increased expression of transforming growth factor-β (TGF-β), which are closely associated with cell proliferation and induction of TKI resistance (80,81). Compared with AM in control mice, AM in hormone mice produced more cytokines (such as IL-1α and TNF-α) and chemokines [such as CXC motif chemokine ligand (CXCL) 1 and CXCL2] (48). In addition, the chemokine CXCL2 secreted by CAFs is considered to induce the expression of PD-L1 in LA cells, thereby indirectly modulating tumor immunity (82).

TKI treatment promotes the transformation of EGFR-mt NSCLC from a ‘cold’ to a ‘hot’ tumor. After EGFR-TKI treatment, the level of type I interferon (IFN) is notably increased in EGFR-mt human LC cell lines, and the expression of the chemokines CXCL9, CXCL10 and CXCL11 is also markedly upregulated. Meanwhile, the CXCL10/CXCR3 pathway is activated in the EGFR-mt LC transgenic mouse model (10,42,55). Furthermore, an analysis of serum IFN-γ levels in 20 patients with NSCLC treated with EGFR-TKI revealed that serum IFN-γ levels were markedly higher compared with baseline levels in treated patients (66).

In EGFR-TKI-resistant patients, the TME remodels from a non-inflammatory to an inflammatory state. Studies have shown that acquired resistance to TKI therapy can promote an inflammatory response, with the IFN-γ pathway becoming markedly enriched after TKI resistance, along with higher levels of granzyme A expression (46). Type I IFN levels are decreased in resistant cell lines, whereas expression of the pro-inflammatory cytokines IL-6 and TGF-β1 is notably increased (10,81,83,84).

CD24 and lymphocyte-activation gene 3 (LAG-3)

In a related study (85), the expression of the innate immune checkpoint CD24 was found to be upregulated in EGFR-mt cells in vitro following EGFR-TKI treatment, which is consistent with the observation after TKI resistance. These findings suggest that EGFR inhibition in EGFR-mt NSCLC cells promotes the development of a TME conducive to immune escape. Furthermore, the expression of the checkpoint protein LAG-3 is markedly elevated after EGFR-TKI treatment (86).

Immunotherapy clinical trials in EGFR-mt advanced NSCLC

Immunological monotherapy or dual therapy

A pair of meta-analyses (87,88), which included studies such as CheckMate057, KEYNOTE-010, OAK and POPLAR, demonstrated that the efficacy of immunotherapy monotherapy was inferior compared with that of docetaxel monotherapy in patients with EGFR-mt LC. As a result, patients with EGFR-mt NSCLC are less likely to derive notable benefits from immunotherapy monotherapy. The BIRCH study (89), a phase II, single-arm, multicenter clinical trial evaluating the efficacy and safety of atezolizumab monotherapy in advanced PD-L1-expressing NSCLC, included 45 patients with EGFR mutations. The results showed that atezolizumab exhibited antitumor activity regardless of EGFR status and was particularly effective in the high PD-L1-expressing subgroup (TC3 or IC3). However, even among EGFR-mt patients with high PD-L1 expression, the objective response rate (ORR) and median overall survival (mOS) remained notably lower compared with those in EGFR-wt patients. The ATLANTIC (90) study was a single-arm, open-label, phase II trial evaluating durvalumab (an anti-PD-L1 monoclonal antibody) as a third-line or later treatment for advanced NSCLC, enrolling a total of 77 patients with EGFR mutations. Subgroup analysis revealed that while the ORR in patients with EGFR-mt NSCLC with PD-L1 expression ≥25% was higher compared with those with low PD-L1 expression [12.2% (9/74) vs. 3.6% (1/28)], it was still lower compared with EGFR-wt patients [16.4% (24/146) vs. 7.5% (7/93)].

In one cohort of the KEYNOTE-021 study (91), a total of 11 patients were enrolled. The combination of ipilimumab [anti-cytotoxic T-lymphocyte associated protein 4 (CTLA-4) antibody] and pembrolizumab showed limited efficacy in patients with EGFR-mt NSCLC, with an ORR of only 10%, substantially lower than the 30% observed in EGFR-wt patients. Overall, the data suggest that while the combination therapy exhibits some antitumor activity in patients with advanced NSCLC who had received multiple lines of therapy, it was less effective and more toxic in EGFR-mt patients, with a 64% incidence of treatment-emergent adverse events (AEs), including 29% of grade 3–5 severe AEs. The study suggests that dual immunotherapy has limited effectiveness in EGFR-mt patients and must be chosen with caution, especially when PD-L1 expression levels are low. In conclusion, these clinical trial data indicate that the overall response rate to immunotherapy is lower in patients with EGFR-mt NSCLC compared with those with EGFR-wt. Nevertheless, antitumor activity was observed in EGFR-mt patients with high PD-L1 expression, suggesting that high PD-L1 expression may serve as an important predictor of potential benefit from immunotherapy in this subgroup.

Immunotherapy combined with targeted therapy

CAURAL (92) is a phase II clinical trial investigating the combination of osimertinib (a third-generation EGFR-TKI) and durvalumab in patients with advanced NSCLC with EGFR T790M mutations who had progressed after prior EGFR-TKI treatment. The primary objective of the trial was to evaluate safety, with efficacy assessed as an exploratory endpoint. The results demonstrated an ORR of 64% in the combination therapy arm, lower than the 80% observed in the osimertinib monotherapy arm, failing to show an advantage over single-agent therapy. Nonetheless, the trial was terminated early due to the elevated risk of interstitial lung disease observed in related studies. This highlights that safety issues need to be considered when combining an EGFR-TKI with an ICI.

Immunotherapy combined with chemotherapy

CT18 (93) is a multicenter, single-arm, phase II clinical trial evaluating the efficacy and safety of toripalimab (anti-PD-1 antibody) in combination with carboplatin and pemetrexed in patients (n=40) with advanced EGFR-mt NSCLC who had failed EGFR-TKI therapy and lacked T790M mutations. The results showed an ORR of 50.0% (95% CI, 33.8–66.2), and a disease control rate (DCR) of 87.5% (95% CI, 73.2–95.8). The mPFS was 7.0 months (95% CI, 4.8–8.4), while the mOS reached 23.5 months (95% CI, 18.0 to NR months). The overall safety profile of the treatment was manageable, with 97.5% of patients experiencing treatment-related AEs (TRAEs), of which 65.0% were grade 3 or higher, most commonly bone marrow suppression, elevated transaminases and nausea. Immune-related AEs occurred in 40% of patients, with only 5.0% being grade 3 or higher.

CheckMate722 (94) is a phase III clinical trial designed for patients with advanced NSCLC characterized by EGFR mutations and resistance to TKI therapy (n=294). This investigation sought to compare the efficacy of nivolumab combined with chemotherapy against chemotherapy alone. While the combination therapy group showed a modest improvement in mPFS (5.6 vs. 5.4 months; HR=0.75), the difference fell short of statistical significance. Likewise, the combination treatment did not notably extend mOS (19.4 vs. 15.9 months; HR=0.82). The ORR was 31.3 and 26.7% in the combination and chemotherapy groups, respectively. Post-hoc subgroup analyses revealed a potential trend of enhanced PFS in specific subpopulations, including those with sensitizing EGFR mutations or patients treated exclusively with first-line TKI therapy, with HRs of 0.72 and 0.64, respectively.

The KEYNOTE-789 study (95), a randomized, double-blind phase III trial in 492 patients with TKI-resistant, EGFR-mt advanced NSCLC, demonstrated that pembrolizumab in combination with chemotherapy only slightly improved mPFS (5.6 vs. 5.5 months; HR=0.80; P=0.0122) or mOS (15.9 vs. 14.7 months; HR=0.84; P=0.0362) compared with chemotherapy alone. Neither the CheckMate722 nor the KEYNOTE-789 trial met their primary endpoints, suggesting that merely combining anti-PD-1 ICIs with chemotherapy may not provide meaningful clinical benefits for this patient population. Consequently, it is essential to further investigate potential biomarkers and refine combination treatment strategies to enhance therapeutic outcomes.

Immunotherapy combined with chemotherapy + anti-angiogenic drugs

IMpower150 (96,97) is a phase III clinical trial that enrolled 123 patients with EGFR-mt NSCLC to assess the efficacy of three regimens of atezolizumab in combination with carboplatin and paclitaxel (ACP), bevacizumab plus carboplatin and paclitaxel (BCP) and atezolizumab plus bevacizumab plus carboplatin and paclitaxel (ABCP). The results showed that patients with EGFR mutations or ALK translocations in the ABCP group demonstrated a longer mPFS compared with those in the BCP group, at 9.7 vs. 6.1 months, respectively (HR=0.59; 95% CI, 0.37–0.94; P=0.025). The mOS for the ABCP group was 29.4 months, which was notably superior to the 18.1 months observed in the BCP group (HR=0.60; 95% CI, 0.31–1.14), while the ACP group (19.0 months) was similar to the BCP group (HR=1.00; 95% CI, 0.57–1.74). In terms of safety, 100% of patients in the ABCP group experienced TRAEs, of which 66.7% were grade 3/4, although no grade 5 events were reported. The incidence of TRAE in the ACP and BCP groups was 88.6% (with 56.8% grade 3/4) and 95.3% (55.8% grade 3/4), respectively. In summary, the IMpower150 study demonstrated that the ABCP regimen substantially prolonged OS in patients with EGFR-mt NSCLC with controllable AEs and had a favorable safety and tolerability profile.

ORIENT-31 (98) is a phase III clinical trial designed to assess the efficacy of sintilimab ± IBI305 combined with chemotherapy (pemetrexed + cisplatin) in patients with locally advanced or metastatic EGFR-mt non-squamous NSCLC who have progressed after EGFR-TKI treatment. The randomized trial with 476 patients indicated that sintilimab combined with chemotherapy significantly improved mPFS, with 5.5 vs. 4.3 months for chemotherapy alone (HR=0.72; 95% CI, 0.55–0.94; P=0.016). Sintilimab plus IBI305 and chemotherapy extended mPFS to 7.2 months, compared with 4.3 months for chemotherapy alone (HR=0.51; 95% CI, 0.39–0.67; P<0.0001), demonstrating significant benefit. Regarding mOS, the sintilimab plus IBI305 combination chemotherapy group had a mOS of 21.1 months (95% CI, 17.5–23.9), which was similar to the 19.2 months observed in the chemotherapy alone group (HR=0.98; 95% CI, 0.72–1.34; P=0.8883), with no statistically significant difference. The mOS for the sintilimab plus chemotherapy group was 20.5 months (95% CI, 15.8–25.3), which was comparable to the chemotherapy alone group at 19.2 months (HR=0.97; 95% CI, 0.71–1.32; P=0.8202). In terms of safety, 56% of patients (88/158) in the sintilimab plus IBI305 plus chemotherapy group developed a grade 3 or higher TRAE, which was notably higher compared with in the sintilimab plus chemotherapy (41%) and chemotherapy alone (49%) groups. Despite immunotherapy side effects (such as immune pneumonitis and rash) were more common, overall safety and tolerability were satisfactory and most AEs could be mitigated with appropriate management.

IMpower151 (99) is a phase III study that included 305 patients with advanced non-squamous NSCLC. The trial assessed the efficacy of atezolizumab in combination with bevacizumab (an anti-VEGF monoclonal antibody) and chemotherapy (carboplatin + pemetrexed) as a first-line treatment. Preliminary findings indicated that the combination therapy (ABCP) showed limited benefits in mPFS and mOS compared with the control group (BCP), without reaching statistical significance. The mPFS was 9.5 and 7.1 months (HR=0.84; 95% CI, 0.65–1.09; P=0.18) and the mOS was 20.7 and 18.7 months (HR=0.93; 95% CI, 0.67–1.28) for the two groups, respectively. These results suggest that although the combination of immunotherapy and chemotherapy provided a slight survival benefit for patients with EGFR-mt NSCLC, the effect was modest compared with other NSCLC subtypes and did not substantially alter the prognosis. Safety analyses revealed a high incidence of AEs in both groups and no new safety signals. The rates of all-cause AEs were 99.3% in the ABCP group (with 66.4% being grade 3/4) and 100% in the BCP group (61.4% grade 3/4).

ATTLAS (100) is a phase III study performed in Korea that enrolled 225 patients with stage IV NSCLC diagnosed with EGFR sensitizing mutations or ALK translocations, including the ABCP group (n=151) and the PC group (n=74). All patients had disease progression or intolerance to one or more EGFR or ALK TKIs. A total of 168 patients with EGFR-mt NSCLC who were resistant to EGFR-TKI were enrolled in the study, 109 in the ABCP arm and 59 in the PC arm. The results demonstrated that the ABPC arm significantly improved PFS compared with the PC arm, with mPFS of 8.48 vs. 5.62 months, respectively [HR=0.62 (95% CI, 0.45–0.86); P=0.004]. Subgroup analysis of patients with EGFR-TKI-resistant EGFR-mt NSCLC revealed similar findings, with mPFS of 8.7 months in the ABCP arm and 5.6 months in the PC arm (HR=0.60; 95% CI, 0.43–0.84; P=0.002). However, no notable OS benefit was observed in either group. In terms of safety, the incidence of grade 3 or higher TRAEs was higher in the ABCP arm compared with in the PC arm, primarily related to cytotoxic chemotherapy. Nevertheless, bevacizumab-related TRAEs were generally manageable with appropriate supportive care.

HARMONi-A (101,102) is a phase III clinical trial performed in China, enrolling 322 patients with advanced or metastatic EGFR-mt NSCLC who experienced disease progression during EGFR-TKI treatment. This study evaluated the efficacy of ivonescimab in combination with chemotherapy compared with chemotherapy alone. In the trial, eligible patients were randomized 1:1 to receive ivonescimab plus chemotherapy (pemetrexed and carboplatin) or placebo plus chemotherapy. Results showed that ivonescimab plus chemotherapy resulted in a significant improvement in mPFS compared with the chemotherapy group at 7.1 and 4.8 months, respectively [HR=0.46 (95% CI; 0.34–0.62); P<0.001]. The ORRs of the two groups were 50.6 and 35.4%, respectively, and the DCRs were 93.1 and 83.2%, respectively. In terms of safety, the incidence of AEs was 99.4 and 97.5% in the two groups, and the incidence of grade 3 or higher treatment-emergent AEs was 61.5 and 49.1%, respectively.

In addition, two ongoing phases II clinical trials (103,104) have shown preliminary efficacy. A single-arm phase II study enrolled 64 patients with EGFR-mt NSCLC who had progressed following EGFR-TKI treatment and received combination chemotherapy with PM8002/BNT327, a bispecific antibody targeting PD-L1 and VEGF-A. The results revealed an overall ORR of 54.7% (95% CI, 41.8–67.2). In patients with a TPS ≥50%, the ORR reached 92.3% (95% CI, 64.0–99.8), indicating that the antitumor activity of PM8002/BNT327 treatment was positively associated with tumor PD-L1 expression levels. Additionally, cohort 5 (EGFR-TKI-resistant cohort) of the DUBHE-L-201 study (104) included 31 patients. The cohort received QL1706 + carboplatin + pemetrexed + bevacizumab administered intravenously on day 1 of a 21-day cycle for four cycles. Maintenance therapy was QL1706 + pemetrexed + bevacizumab. QL1706 is a bifunctional MabPair product containing anti-PD-1 and anti-CTLA-4 antibodies. The primary endpoint of the study was safety and secondary endpoints include confirmed ORR, investigator-assessed duration of remission (DoR), PFS and OS. Preliminary results showed a median DoR of 11.3 months (95% CI, 4.2–19.9), PFS of 8.5 months (95% CI, 5.7–13.3) and OS of 26.5 months (95% CI, 12.8-not evaluable) (Table I).

Table I. ICI-based immunotherapy combinations for EGFR-mutant advanced NSCLC.

Table I.

ICI-based immunotherapy combinations for EGFR-mutant advanced NSCLC.

A, Immunotherapy combined with targeted therapy
NCT	Clinical trial	Phase	Intervention	Result	(Refs.)
NCT02454933	CAURAL	III	Osi + Durva (n=14)	ORR:64% vs. 80%	(92)
			Osi (n=15)
B, Immunotherapy combined with chemotherapy
NCT03513666	CT18	II	Tori + Carbo + Pem	ORR=50.0% (95% CI, 33.8–66.2)	(93)
			(n=40)	DCR=87.5% (95%CI, 73.2–95.8)
				mPFS=7.0 m (95% CI, 4.8–8.4)
				mOS=23.5 m (95% CI, 18.0-NR)
NCT02864251	CheckMate722	III	Nivo + Chemo	mPFS:5.6 m vs. 5.4 m (HR=0.75;
			(n=144)	95% CI, 0.56–1.00) mOS:19.4 m	(94)
			Chemo (n=150)	vs. 15.9 m (HR=0.82; 95% CI, 0.61–1.10)
NCT03515837	KEYNOTE-789	III	Pembro + Chemo	mPFS:5.6 mvs5.5 m (HR=0.80;	(95)
			(n=245)	95% CI, 0.65–0.97)
			Placebo + Chemo	mOS:15.9 vs. 14.7 (HR=0.84;
			(n=247)	95% CI, 0.69–1.02)
C, Immunotherapy combined with chemotherapy + anti-angiogenic drugs
NCT02366143	IMpower150	III	ABCP: Atezo +	mPFS: ABCP vs. BCP:9.70 m	(96,97)
			Bev + Carbo+	vs. 6.1 m (HR=0.59; 95% CI,
			Pac (n=34)	0.37–0.94)
			BCP: (n=44)	mOS: ABCP vs. BCP:26.1 m
			ACP: (n=45)	vs. 20.3 m (HR=0.91; 95% CI,
				0.53–1.59)
				ACP vs. BCP: 21.4 m vs. 20.3 m
				(HR=1.16; 95% CI, 0.71–1.89)
NCT03802240	ORIENT-31	III	SIC: Sinti +	mPFS: SIC vs. C:7.2 m vs. 4.3 m	(98)
			IBI305 + Chemo	(HR=0.51; 95% CI, 0.39–0.67)
			(n=158)	SC vs. C: 5.5 m vs. 4.3 m
			SC: Sinti + Chemo	(HR=0.72; 95% CI, 0.55–0.94)
			(n=158)	mOS: SIC: 21.1 m (95% CI)
			C: Chemov (n=160)	17.5–23.9) SC:20.5 m (95% CI)
				15.8–25.3) C:19.2 m (95% CI
				(15.8–22.4)
NCT04194203	IMpower151	III	ABCP: (n=152)	mPFS: 9.5 m vs. 7.1 m (HR=0.84;	(99)
			BCP: (n=153)	95% CI, 0.65–1.09; P=0.18)
				mOS: 20.7 m vs. 18.7 m
				(HR=0.93; 95% CI, 0.67–1.28)
NCT03991403	ATTLAS	III	ABPC: Atezo +	mPFS: ABCP vs. PC:8.71 m	(100)
			Bev + Pac +	vs. 5.62 m (HR=0.60; 95% CI,
			Carbo (n=109)	0.43–0.84)
			PC: (n=59)
NCT05184712	HARMONi-A	III	ivonescimab + Chemo (n=161)	mPFS: 7.1 m vs. 4.8 m (HR=0.46; 95% CI, 0.34–0.62)	(101,102)
			Chemo (n=161)
NCT05756972	-	II	PM8002/BNT327 +	ORR:54.7%	(103)
			Carbo + Pem (n=64)
NCT05329025	DUBHE-L-201	II	QL1706+ Carbo + Pem + Bev (n=31)	mDoR, PFS, and OS were 11.3 (95% CI, 4.2–19.9), 8.5 (5.7–13.3), and 26.5 months (12.8-not evaluable), respectively	(104)

[i] ICs, immune cells; EGFR, epidermal growth factor receptor; TKIs, tyrosine kinase inhibitors; NSCLC, non-small cell lung cancer; mPFS, median progression-free survival; mOS, median overall Survival; RR, response rate; ORR, objective response rate; TPS, tumor proportion score; TTP, time to progression; DCR, disease control rate; TCs, tumor cells; mDoR, median duration of remission; ICI, immune checkpoint inhibitor; NCT, ClinicalTrials.gov identifier; Osi, Osimertinib; Durva, Durvalumab; Tori, Toripalimab; Carbo, Carboplatin; Pem, Pemetrexed; Nivo, Nivolumab; Chemo, Chemotherapy; Pembro, Pembrolizumab; Atezo, Atezolizumab; Bev, Bevacizumab; Pac, Paclitaxel; Sinti, Sintilimab.

Discussion and future perspectives

In summary, TKI-treated and TKI-resistant TMEs show a shift towards a ‘hot’ tumor with increased immune cell infiltration (Fig. 2), with an increase in the immune-activating components of the TME, including increased numbers or upregulation of tumor-infiltrating immune cells, immunomodulatory molecules, cytokines or chemokines and a reduction or impairment of immunosuppressive components. Notably, acquired EGFR-TKI resistance promotes immune escape in lung cancer by upregulating PD-L1 expression. Detailed studies have shown that changes in the TME following TKI treatment are dynamic, with short-term effects suggesting a potential therapeutic window during which TKI treatment could more effectively leverage the immune system to target tumors (Fig. 3). However, the long-term effects of TKI therapy are more complex and may be influenced by various factors, such as treatment cycles, drug sensitivity and the specific treatment regimen employed.

Figure 2. TME in EGFR-mutant NSCLC and its dynamics during short-term TKI treatment and resistance. The TME of EGFR-mutant NSCLC is immunosuppressive. Low CD8+ T-cell infiltration and inhibition of their proliferation and activation by the cytokine TGF-β; EGFR mutation promotes AM proliferation; TEXs inhibit DC function and reduce IL-12p40 production; MHC-II and co-stimulatory molecules of DCs and AMs (CD40, CD80, CD86) expression were downregulated; CD47 was selectively overexpressed; EGFR mutations upregulated PD-L1 via IL-6/JAK/STAT3 and p-ERK1/2/p-c-Jun pathways; high expression of pro-inflammatory cytokine IL-17A; increased expression of cytokines TGF-β, IL1α and TNFα and chemokines CXCL1 and CXCL2; and CAFs release of CXCL2 induced PD-L1 expression. EGFR-TKI can directly inhibit the activity of tumor cells. Short-term TKI treatment transformed ‘cold tumors’ into ‘hot tumors’, with an increase in T-cell and DC infiltration, an increase in CD8 and GB expression levels Compared with pre-treatment and a marked decrease in TAM infiltration and Foxp3+ Tregs; CD47 is downregulated, enhancing DC-mediated phagocytosis; PD-L1 expression could be reduced or upregulated, and the reduced PL-L1 was considered to be associated with the NF-κB and AKT-STAT3 pathways; the level of type I IFNs (IFN-α, IFN-β) was increased and the expression of IFN-γ and the chemokines CXCL9, CXCL10, and CXCL11 was upregulated. Upon TKI resistance, the TME was remodeled towards inflammation, with enrichment of inflammatory and IFN-γ pathway; CD8 and GB expression decreased and granzyme A expression increased; TAM numbers rise; CD47 was re-upregulated; the PD-1 pathway was activated and the expression level of PD-L1 was increased; type I IFN decreased and the expression of IL-6 and TGF-β1 was increased; the increased expression of IL-6 suppressed T cell and NK cell cytotoxicity and GB marker expression. TME, tumor microenvironment; EGFR, epidermal growth factor receptor; NSCLC, non-small cell lung cancer; TGF-β, transforming growth factor-β; AM, alveolar macrophages; CXCL, CXC motif chemokine ligand; TEXs, tumor-derived exosomes; p-, phosphorylated; IL, interleukin; TNF, tumor necrosis factor; CAFs, cancer-associated fibroblasts; GB, granzyme B; IFN, interferon; Tregs, regulatory T cells; DC, dendritic cell; TAMs, tumor-associated macrophages; NK, natural killer; PD-L1, programmed death ligand 1; TKI, tyrosine kinase inhibitor.

Figure 3. A timeline schematic illustrating dynamic changes in the TME from the EGFR-mutant baseline state through short-term TKI exposure to the development of acquired resistance. TME, tumor microenvironment; EGFR, epidermal growth factor receptor; TKI, tyrosine kinase inhibitor; Treg cells, regulatory T cells; TAMs, tumor-associated macrophages; DC, dendritic cell; NK, natural killer; PD-L1, programmed death ligand 1; IL, interleukin; TGF-β, transforming growth factor-β; TNF, tumor necrosis factor; CXCL, CXC motif chemokine ligand; IFN, interferon; LAG-3, lymphocyte-activation gene 3; PD-1, programmed cell death protein-1.

Previous reports have shown that patients with high PD-L1 expression have a greater survival benefit following treatment with first-generation EGFR-TKI. However, the subsequent FLAURA trial (105) reported conflicting results, suggesting a shorter PFS in patients with high PD-L1 expression. With second and third-generation TKI therapies, high PD-L1 expression has been associated with worse survival and higher rates of drug resistance. These findings suggest a strong relationship between PD-L1 expression levels and TKI efficacy. Previous cases are described in Table II (106–117). However, discrepancies between studies may be due to differences in drug generations, PD-L1 detection methods and thresholds for defining PD-L1 expression levels. Most existing studies are retrospective, with limited prospective studies and notable regional variability. There is an urgent need for larger, multicenter, prospective clinical trials to validate the relationship between PD-L1 expression and clinical outcomes. Standardization of PD-L1 detection methods is also critical to identifying the optimal population for ICI treatments, ultimately providing greater clinical benefit to patients.

Table II. Association between PD-L1 expression and treatment efficacy of EGFR-TKI in EGFR-mutant NSCLC patients.

Table II.

Association between PD-L1 expression and treatment efficacy of EGFR-TKI in EGFR-mutant NSCLC patients.

First author/s,year	Gen.	EGFR-TKI	Country	EGFR-mutant (PD-L1 Testing) Patients (n)	PD-L1 expression in EGFR-mutant/after TKIs	Associated clinical outcomes	(Refs.)
D'Incecco et al, 2015	I	Gefitinib or erlotinib	Italian	95/123	PD-L1 positivity (2+ or 3+) in >5% of TCs: 55.3%/-	PD-L1+: higher RR (61.2% vs. 34.8%, P=0.001), longer TTP (11.7 m vs. 5.7 m, P<0.0001), and longer OS (21.9m vs. 12.5 m, P=0.09)	(64)
Lin et al, 2015		Gefitinib or erlotinib	China	56	Positive in 53.6% of tumor specimens/-	PD-L1+: higher DCR (93.3% vs. 61.5%, P=0.004), longer mPFS (16.5 vs. 8.6 m, P=0.001) and mOS (35.3 vs. 19.8 m, P=0.004)	(106)
Soo et al, 2017	I/II	Erlotinib, gefitinib, or dacomitinib	South Korea	90	≥1% positive: ICs 44%, TCs 59%/no association	PD-L1+: shorter PFS [HR=1.008 (1.001–1.015), P=0.017]	(107)
Yoneshima et al, 2018		-	Japan	71	TPS of ≥1%: 42.3%/-	TPS ≥1% vs. <1%: shorter mPFS (9 vs. 14 m, P=0.016)	(108)
Su et al, 2018		-	China	101	Positive (TC3/IC3, TC1-2/IC1-2): 35.6%/-	TC3/IC3 vs. TC1-2/IC1-2 vs. TC0/IC0: lower ORR (35.7% vs. 63.2% vs. 67.3%, P=0.002), shorter mPFS (3.8 vs. 6.0 vs. 9.5m, P<0.001), and higher primary resistance rates (66.7% vs. 30.2%) P=0.009)	(109)
Hsu et al, 2019		Gefitinib, erlotinib, or afatinib	China	123	TPS of ≥1%: 30.1%/-	TPS ≥1% vs. <1%: shorter mPFS (2.1 vs. 7.3 m, P<0.001) and mOS (11.2 vs. 38.2, P=0.002), higher primary resistance rates (OR=5.95, 95% CI 2.35–15.05, P<0.001)	(110)
Yang et al, 2020		Gefitinib, erlotinib, or afatinib	China	153	TPS 1–49%: 25.5%, TPS ≥50%: 11.8%/Stable in 60% (9/15), increased in 40% upon progression	TPS 0% vs. 1–49% vs. ≥50%: ORR (65.6% vs. 56.4% vs. 38.9%, P<0.001), DCR (93.8% vs. 97.4% vs. 55.6%, P<0.001), mPFS (2.5 vs. 12.8 vs. 5.9 m, P=0.027), primary resistance rates (6.3% vs. 2.6% vs. 44.4%, P<0.001)	(60)
Isomoto et al, 2020	I/II/III	Gefitinib, erlotinib, afatinib, dacomitinib or osimertinib	Japan	134	TPS of ≥50%: 14%/TPS of ≥50%: 28%	-	(73)
Brown et al, 2018	I/III	Osimertinib vs. gefitinib/erlotinib	Global	128	TC ≥1%: 51%/-	TC ≥1% vs. <1%: Osimertinib group (n=54): similar mPFS (18.4 vs. 18.9 m); gefitinib/erlotinib group (n=52): shorter mPFS (6.9 vs. 10.9 m)	(105)
Sakata et al, 2021	III	Osimertinib	Japan	538	TPS ≥50%: 11.9%, 1–49%: 31.6%, <1%: 30%, Unknown: 26.6%/-	mPFS: TPS ≥50%: 11.1 m (95% CI 8.3-NR) [HR=2.24 (1.17–4.30), P=0.015], TPS 1–49%:14.7 m (95% CI 13.6–20.5) [HR=1.66 (1.05–2.63), P=0.029], TPS <1%: NR (95% CI 20.7-NR)	(111)
Alves et al, 2022	I/III	Gefitinib, erlotinib or osimertinib	Brazil	278/188	TC ≥1%: 36.7%/-	Higher PD-L1 (TPS ≥50% vs. 1–49% vs. <1%): not associated with mOS (37.5 vs. 46.5 vs. 50.4 m, P=0.48), and decreased event-free survival (median:9 vs. 34 vs. 26 m, P=0.014)	(112)
Papazyan et al, 2024	III	Osimertinib	France	96	TPS ≥50%: 20.8%/decreased in 4/15 patients at relapse	TPS ≥50% vs. <50%: shorter mPFS (9.3 vs. 17.5 m, P=0.044), and mOS (14.3 vs. 26.0 m, P=0.025), ORR (53.3% vs. 46.5%, P>0.9)	(113)
Lakkunarajah et al, 2023		Osimertinib	Canada	231	TPS ≥50%: 29.4%/-	TPS ≥50% vs. <1%: shorter PFS (HR=1.59, 95% CI 1.07–2.36, P=0.023), and OS (HR=1.82, 95% CI 1.10–2.99, P=0.019)	(114)
Yoshimura et al, 2021		Osimertinib	Japan	71	TPS ≥50%: 21.1%/-	TPS ≥50% vs. <50%: lower ORR (53.3% vs. 81.1%, P=0.043) and DCR (73.3% vs. 98.1%, P=0.007), shorter mPFS (5.0 vs. 17.4 m, P<0.001); TPS ≥50% vs. 1–49% vs. <1%: higher primary resistance rates (33.33% vs. 3.85% vs3.45%, P=0.006)	(115)
Hsu et al, 2022		Osimertinib	China	85/71	TPS ≥50%: 9.4%/-	TPS ≥50% vs. <50%: shorter mPFS [9.7 vs. 26.5 m, aHR=0.19 (0.06–0.67), P=0.009], shorter mOS [25.4 vs. NR, aHR=0.09 (0.01–0.70), P=0.021]	(116)
Hamakawa et al, 2023		Osimertinib	Japan	64	TPS ≥20%: 34.4%/-	TPS ≥20% vs. <20%: shorter mPFS (9.1 vs. 28.1 m, log-rank P=0.013), with PD-L1 TPS ≥20% associated with early resistance to osimertinib	(117)

[i] PD-L1, programmed death-ligand; EGFR, epidermal growth factor receptor; TKI, tyrosine kinase inhibitor; NSCLC, non-small cell lung cancer; PFS, progression-free survival; OS, overall survival; RR, response rate; ORR, objective response rate; TPS, tumor proportion score; TTP, time to progression; DCR, disease control rate; TCs, tumor cells; ICs, immune cells; TC3/IC3: ≥50% for TC or ≥10% for IC, TC1-2/IC1-2: 5–49% for TC or 5–9% for IC, TC0/IC0: <5% for TC or IC.

Recent research has focused on patients with high PD-L1 expression. Available evidence suggests that immunotherapy can offer notable benefits to patients, both during short-term TKI therapy and after the development of resistance. However, the effect of short-course TKI therapy on PD-L1 expression levels has long been a topic of debate. The timing of immunotherapy interventions and the monitoring of their efficacy need to be further confirmed. On the one hand, the immunological characteristics of the TME, such as inflammatory or immunosuppressive status, as well as the level of PD-L1 expression should guide the choice of treatment. On the other hand, identifying reliable biomarkers, such as PD-L1 and cytokine levels, is crucial for accurately predicting treatment response. It is particularly critical and urgent to conduct large-scale, multicenter clinical trials to validate the role of these biomarkers and to promote their standardized application in clinical practice.

The growing importance of immunotherapy is further underscored by the emergence of resistance to TKI therapy. In patients with high PD-L1 expression and EGFR-TKI resistance, immunotherapy may provide a notable survival benefit (73,103). Conversely, in patients with low PD-L1 expression and EGFR-TKI resistance, immunotherapy interventions do not yield the desired therapeutic outcomes. Fourth-generation EGFR-TKI and antibody-drug conjugates (ADCs) offer novel therapeutic strategies for patients who have failed conventional TKI therapy and exhibit low PD-L1 expression. Fourth-generation TKIs are designed to overcome the limitations of third-generation TKIs in the face of drug resistance. ADCs targeting trophoblast surface antigen 2 have shown notable therapeutic efficacy in immunotherapy non-responders. Future studies should focus on elucidating the underlying mechanisms of dynamic immune changes in the TME, as well as performing comprehensive and systematic evaluations of combination therapies that include TKIs, ICIs, ADCs and anti-angiogenic drugs. Moreover, actively exploring the optimal therapeutic window and addressing the safety issues associated with combination therapies will greatly advance the field of oncology and bring new hope to more patients.

Acknowledgements

No applicable.

Funding

The present study was supported by the Renxin Medical Research Program of the Beijing Weiai Public Welfare Foundation (grant no. RXYS2025-0200630118), the Scientific Research Program of the Tianjin Municipal Education Commission (grant no. 2025ZD061), and the Scientific Research Program of the Hebei Administration of Traditional Chinese Medicine (grant no. T2026091).

Availability of data and materials

Not applicable.

Authors' contributions

All authors contributed to the study conception and design. HM and LL conceived the study and wrote and revised the manuscript. CJ, YC and JH wrote the manuscript. HM, LL, QT and CJ wrote, reviewed and edited the manuscript. DY contributed to the manuscript preparation. YZ supervised the study. Data authentication is not applicable. All authors have read and approved the final manuscript.

Ethics approval and consent to participate

Not applicable.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References