Introduction

In the evolving treatment landscape of non-small cell lung cancer (NSCLC), the recent advent of immune checkpoint inhibitors (ICIs) and targeted tyrosine kinase inhibitors (TKIs) has marked a significant paradigm shift.1 2 TKIs have become crucial in NSCLC management by specifically targeting genetic mutations, such as epidermal growth factor receptor (EGFR) mutations and anaplastic lymphoma kinase (ALK) fusions.3 While TKIs demonstrate notable initial efficacy, their long-term effectiveness is often challenged by the emergence of resistance.4 Unlike the response to TKIs, the overall response rate to ICIs may be lower,5 6 but for patients who respond, ICIs offer durable effects, reducing the likelihood of resistance development and enabling longer-term survival.7 8 Their success predominantly stems from their profound impact on the tumor microenvironment (TME), reactivating the immune system’s ability to identify and eradicate tumor cells, thereby fostering a milieu conducive to sustained immune responses.9–11 A thorough understanding of the TME is increasingly recognized as crucial in optimizing cancer treatments, as the TME plays key roles in the effectiveness and development of resistance to therapies, particularly in NSCLC.12 13 Traditionally, resistance to TKIs in NSCLC has been primarily attributed to tumor cell-intrinsic factors, including genetic mutations and alterations in signaling pathways.14 As a result, it is generally believed that targeted drugs may exert control through inhibition rather than elimination. However, recent research has unveiled the potential roles of TKIs in modulating the TME, suggesting that different oncogenic driver mutations may lead to distinct TME characteristics.3 15 16

The increasing application of neoadjuvant therapies has significantly enhanced our understanding of how both immunotherapy and targeted treatments influence the TME in NSCLC.17 In particular, neoadjuvant immunotherapy, often used in conjunction with chemotherapy, has been associated with higher pathological response rates.18 19 Patients achieving pathological remission typically exhibit better long-term survival outcomes, which may be largely attributed to the profound impact of ICIs on the TME,18 20–22 a phenomenon linked to increased functional T-cell infiltration within the tumor areas.17 These observations support the notion that an immune-activated TME is instrumental in achieving pathological remission and improving prognosis.23 Consequently, the modulation of the TME to amplify immune responses has emerged as a promising strategy to enhance the efficacy of NSCLC treatments.24 In contrast, studies on various neoadjuvant targeted therapies, especially those involving EGFR-TKIs, have demonstrated lower rates of pathological remission.25 This disparity in pathological responses between immunotherapy and targeted therapy highlights a potential factor contributing to the development of resistance in targeted treatments.26 This observation raises critical questions regarding the underlying mechanisms that constrain the long-term effectiveness of targeted therapies and contribute to the development of resistance.

Additionally, certain clinical observations remain puzzling, such as the ability of targeted therapies to significantly inhibit overall tumor growth despite only targeting a small subset of tumor cells with specific mutations. Furthermore, the effectiveness of immunotherapy often appears to be diminished when administered after the development of resistance to targeted therapy.27 ALK-TKIs, in contrast to EGFR-TKIs, tend to result in longer progression-free survival (PFS) and higher rates of pathological remission in neoadjuvant therapy.28 Unraveling the exact mechanisms behind these complex phenomena requires in-depth explorations into the nuances of targeted therapy.

Our research endeavors to unravel the intricacies of how the TME influences the efficacy of targeted treatments. Our findings indicate that the degree of immune suppression within the TME significantly contributes to the development of resistance against TKIs. Furthermore, we have identified novel strategies for its modulation, which hold the promise of delaying the onset of resistance and enhancing the durability of TKI efficacy. Moreover, our research helps to explain many confusing clinical issues in targeted therapy for lung cancer.

Materials and methodsCell lines and reagents

Human lung cancer cell lines HCC827, H3255, PC-9 and H2228 were purchased from the American Type Culture Collection. The H3122 human lung adenocarcinoma cell line was obtained from Shanghai Bioleaf Biotech (Shanghai, China). PC-9GR (gefitinib resistance due to T790M mutation) was created through chronic exposure of the drug at gradual dose increments. The osimertinib-resistant cell lines PC-9OR, H1975OR, PC-9GROR, HCC827OR and lorlatinib-resistant cell lines H3122LR, and crizotinib-resistant cell lines H2228CR were also established in our laboratory. All of these cell lines were maintained in Roswell Park Memorial Institute 1640 medium (Gibco, California, USA) with 10% fetal bovine serum, penicillin (100 U/mL), and streptomycin (100 U/mL) in an incubator with 5% CO2 atmosphere at 37°C.

Osimertinib (S7297), lorlatinib (S7536), gefitinib (S1025), crizotinib (S1068),and aspirin (S3017) were obtained from Selleck Chemicals. alectinib (T1936) was from TargetMol. Anifrolumab (HY-P99168) was obtained from MedChemExpress.

Cell viability assay

Cell viability was determined using a Cell Counting Kit-8 assay (CCK-8, Bioground, Chongqing, China). Briefly, cells were seeded in a 96-well plates at a density of 3×103 cells per well and cultured overnight. On the next day, the medium was replaced with the indicated doses of drug-containing medium and cultured for another 48 hours. Following the replacement of the treatments with fresh medium, the absorbance in each well was measured at 450 nm on a Sunrise R Microplate Reader (Thermo Fisher Scientific, Germany).

Cell proliferation

Cell proliferation was assessed by Ki67 staining. Briefly, cells were seeded in 6-well plates (3×105) and treated as indicated for 48 hours. Then, the cells were fixed and incubated overnight with Ki67 (#M00254-8, Boster, Wuhan, China). They were then counterstained with 4′,6-diamidino-2-phenylindole (DAPI) for 15 min and observed under a fluorescence microscope.

Western blot analysis

Tumor cells were grown and treated as indicated, and total cell proteins were harvested by scraping and quantitated using a bicinchoninic acid protein assay kit (Bioground, Chongqing, China). The following primary antibodies against Bim (#2933S), B-cell lymphoma extra-large (BCL-XL, #2764S), myeloid leukemia cell 1 (MCL-1, #94296), protein kinase B (Akt, #9272S), phospho (Ser473)-Akt (#4060S), caspase-3 (#9662S), cleaved caspase-3 (#9664S), phospho (Tyr1068)-EGFR (#3777S), EGFR (#4267S), phospho (Ser172)-TANK-binding kinase 1 (TBK1, #5483S), TBK1 (#38066S), stimulator of interferon genes (STING, #13647S), phospho (Ser386)-interferon regulatory factor 3 (IRF3, #37 829S), IRF3 (#4302S), phospho (Ser536)-nuclear factor κb (NF-κb) p65 (#3033S), NF-κb p65 (#4764S), programmed death ligand 1 (PD-L1, #13684S), CD8α (#85336S), granzyme B (GB, #46 890S), T-cell immunoglobulin and mucin domain-containing protein 3 (TIM3, #45208S), CD68 (#97778S), CD86 (#91882S), CD163 (#68922S), inducible nitric oxide synthase (iNOS, #20609S), and glyceraldehyde 3-phosphate dehydrogenase (GAPDH, #2118) were obtained from Cell Signaling Technology. An antibody against CD206 (PTM-6020) was obtained from PTM-BioLab. An antibody against laminin subunit γ−2 (LAMC2, ab210959) was obtained from Abcam. Horseradish peroxidase-conjugated goat anti-rabbit or anti-mouse IgG was from Sino Biological (Beijing, China).

Immunohistochemistry staining

Tumor tissues were fixed and embedded with paraffin. Then, 3 mm formalin-fixed, paraffin-embedded slides were prepared for immunohistochemical staining. Briefly, tumor sections were deparaffinized, and tissues were incubated with specific primary antibodies. An anti-rabbit goat IgG labeled with horseradish peroxidase (HRP, ZSGB-Bio, China) was used as the secondary antibody. Staining was performed using the protocol from the DAB display reagent kit.

Generation of activated T cells

Activated T cells were acquired as previously reported.29 Briefly, human peripheral blood T cells were cultured in ImmunoCult-XF T-cell expansion medium with ImmunoCult Human CD3/CD28/CD2 T-cell activator (both from STEMCELL Technologies, Vancouver, California, USA) and interleukin 2 (IL-2, 10 ng/mL; Sino Biological, China) for 1 week according to the manufacturer’s protocol. All experiments were performed in DMEM/F12 with anti-CD3 antibody (100 ng/mL; eBioscience, Thermo Scientific, Massachusetts, USA) and IL-2 (10 ng/mL).

T cell-mediated cancer cell killing assay

Cells were seeded in 6-well plates (3×105) and treated as indicated for 48 hours. T cells and cell debris were removed by washing with phosphate-buffered saline, and cancer cells were subjected to crystal violet staining.

ELISA

Interferon (IFN)-γ and tumor necrosis factor α (TNF-α) ELISA kits were purchased from Solarbio (Beijing, China), and IFN-α and IFN-β ELISA kits were purchased from FineTest (Wuhan, China). The ELISAs were performed according to the manufacturers’ instructions. Values represent the average of three replicates from at least three independent experiments.

Immunofluorescence staining

Immunofluorescence staining of tumor tissue sections was performed as described in a previous study.30 For cyto-immunofluorescence staining, cancer cells were seeded in 6-well plates with pre-placed 20 mm glass slides. The immunofluorescence staining procedures were similar to the tissue staining procedure mentioned above.

Multiplex immunofluorescence detection

Multiplex immunofluorescence (mIF) analysis of tumor tissue sections was performed as described in a previous study.31

Flow cytometry

Cells were stained with antibodies against cell-surface molecules for 20 min at 4°C. The surface staining antibodies included an antibody against CD8 (IM2469, Beckman). For intracellular staining, cells were fixed and permeabilized with forkhead box P3 (Foxp3) fixation/permeabilization solution (IM2469, Thermo Fisher Scientific) for 20 min at 4°C and were then stained with antibodies against intracellular molecules for 20 min at 4°C. The stained cells were analyzed using CytoFLEX fluorescence-activated cell sorting (FACS, Beckman Coulter, Brea, California, USA), and the data were analyzed using FlowJo software V.10 (Treestar, San Carlos, California, USA).

T cells chemotaxis assay

T cells were labeled with carboxyfluorescein diacetate succinimidyl ester (CFSE; Beyotime) at 37°C for 20 min. Horizontal slide chemotaxis chambers (#80326, Ibidi) with or without Matrigel (#356230, Corning Incorporated) were used to investigate the chemotaxis and migration of CFSE-labeled T cells. Matrigel was supplemented with activated or heat-inactivated laminin γ2 (Ln-γ2, #MBS2030234, MyBioSource) or conditioned medium. The number of migratory T cells was imaged with a fluorescence microscope (Leica, German) after 24 hours of culture at 37°C in a humidified chamber containing 5% CO2.

Macrophages induced polarization

THP-1 cells were polarized into M0, M1, and M2 macrophages as previously described. The cells were subjected either to 100 ng/mL phorbol 12-myristate 13-acetate (PMA, Solarbio, China) for 48 hours to obtain M0 macrophages, 20 ng/mL IFN-γ (300–02, PeproTech, USA) plus 100 ng/mL lipopolysaccharide (LPS, L2280, Sigma, USA) for 48 hours to obtain M1 macrophages, or 20 ng/mL IL-4 (200–04, PeproTech, USA) plus 20 ng/mL IL-13 (200–13, PeproTech, USA) for 48 hours to obtain M2 macrophages.

Animal experiments

Animal experiments were approved by the ethics committee on animal experimentation of the Army Medical University. To establish PBMCs-CDX(Peripheral blood mononuclear cells-cell derived xenograft) mouse model (Lin et al, 2018), PC-9 cells (5×105) were injected into the hind flanks of 6-week-old female NOD-SCID mice. The mice were subjected to tumor growth monitoring, and tumor volumes were calculated from caliper measurements using the following formula: (length×width2/2. When tumors reached 100 mm3 in volume, 5×106 human PBMCs were intravenously transplanted. Osimertinib (2 mg/kg) and aspirin (20 mg/kg) were dissolved in drinking water and given to mice orally. Tumor growth was monitored every 3 days. Tumors were harvested, fixed with 4% paraformaldehyde, and embedded in paraffin. The infiltration of CD8+ T cells were stained with immunohistochemistry and the expression of LAMC2, CD68, CD206 was determined by immunofluorescence staining.

Statistical analysis

All datas are expressed as the mean±SEM, and the statistical analyses were performed in GraphPad Prism V.8 for Windows (GraphPad Software, San Diego, California, USA, www.graphpad.com). Differences between the two groups were analyzed by Student’s t-test. Results were considered to be statistically significant at p<0.05.

ResultTKI treatment alters the tumor immune microenvironment

To explore the features of immune cell composition in the TME between EGFR-TKI and ALK-TKI in patients with NSCLC, we first performed multiplex immunochemistry staining to quantitatively analyze immune cells in the TME of pre-TKI and post-TKI treatment tissue samples. In the post-TKI treatment tissue samples, we observed significant increases in the density of CD8+, CD8+GB+, and CD8+ programmed cell death protein 1 (PD-1) + cells within the TME, particularly in ALK fusion-positive samples (figure 1A,B and online supplemental figure S1A), indicating that TKI treatment enhances the antitumor immune response in the TME. In contrast, tissue samples treated with EGFR-TKI exhibited significant increases in CD68+ and CD163+ macrophages that were not observed in ALK TKI-treated pathological samples, indicating a potential enrichment of M2 macrophages (online supplemental figure S1B and figure 1B). Further pathological analysis revealed a tendency toward near-complete pathological remission in ALK fusion-positive tissue samples following targeted therapy (online supplemental figure S1C). These results suggest that TKI treatment can alter the immune cell landscape.

Figure 1

TKI therapy alters the tumor immune microenvironment. (A) Representative mIF images of pretreatment tumor and resected samples analyzed for immune-related biomarkers. (B) Densities (cells/mm2) of CD8+, CD8+GB+, CD8+PD-1+, CD163+, CD68+, and CD163+CD68+ by mIF quantification in paired pretreatment tumor samples and resected tumors. (C) Cell viability CCK-8 assay for cells treated with TKIs (osimertinib: 10 nM, lorlatinib: 10 nM), activated T cells (1:1 ratio to cancer cells), or the combination. (D) T cell-mediated cancer cell-killing assay. PC-9 and H3122 cells co-cultured with activated T cells for 48 hours with or without TKIs (osimertinib: 10 nM, lorlatinib: 10 nM) were subjected to crystal violet staining. Ratio of cancer cells to T cells: 1:1. (E) Ki67 incorporation assay on PC-9 and H3122 cells treated as indicated. Activated T cells (1:1 ratio to cancer cells) or TKIs (osimertinib: 10 nM, lorlatinib: 10 nM) were added to the culture medium for 48 hours. Cells were then counterstained with DAPI. (F) PC-9 cells were injected into mice (n=3 mice per group) on day 0, hu-PBMCs were injected into mice via the tail vein on day 7, and osimertinib was administered as indicated. (G) Macroscopic appearance of tumors after drug application for 4 weeks. (H) The tumor weight (g) for each mouse is shown. *p<0.05, **p<0.01. ns, no significance. (I) Immunofluorescence staining with an antibody against CD8 to detect T cells and antibodies against CD68 and CD206 to detect macrophages in TKI-resistant non-small cell lung cancer tissues (11 cases of EGFR-TKI resistance, 5 cases of ALK-TKI resistance). Scale bar: 50 µm. ALK, anaplastic lymphoma kinase; DAPI, 4′,6-diamidino-2-phenylindole; EGFR, epidermal growth factor receptor; ALKi,anaplastic lymphoma kinase inhibitor; EGFRi,epidermal growth factor receptor inhibitor; hu-PBMC,human-Peripheral blood mononuclear cell; mIF, multiplex immunofluorescence; PD-1, programmed cell death protein 1; TKI, tyrosine kinase inhibitor.

To further validate the crucial role of activated T cells in enhancing the tumoricidal effects of TKI, we first defined the concentration of T cells co-cultured with cancer cells (online supplemental figure S1D,E). Subsequently, we observed that EGFR-TKI and ALK-TKI combined with activated T cells not only significantly reduced the viability of cancer cells but also reduced the number of Ki67-positive cells (figure 1C–E and online supplemental figure S1F,G). We then examined the expression levels of several proapoptotic proteins, including Bim and caspase 3, and found that they were increased in a time-dependent manner in the presence of T cells (online supplemental figure S1I). In addition, in the PBMC-CDX humanized mouse model, we observed that treatment with osimertinib alone reduced tumor sizes, while osimertinib treatment in the presence of PBMCs resulted in more significant tumor shrinkages (figure 1F–H). Thus, we further validated that the therapeutic efficacy of TKI treatment was enhanced by adaptive immunity.

To further explore the effect of TKI treatment on T-cell function, we observed that TKI drugs had no effect on the release of pro-inflammatory cytokines (IFN-γ and TNF-α) from T cells by ELISA (online supplemental figure S1H). This result indicated that TKIs may not directly alter the function of T cells. Interestingly, we observed that CD8 expression decreased and M2 macrophages (CD68+ and CD206+ macrophages) increased in the tissues of patients resistant to TKIs, especially in patients resistant to EGFR-TKI. Immunofluorescence staining revealed an immunosuppressive state in TKI-resistant NSCLC tissues (figure 1I). Under the co-culture conditions (macrophages and cancer cells), we also found that M2 macrophages attenuated the antitumor efficacy of EGFR-TKIs, whereas M1 macrophages potentiated the antitumor efficacy of EGFR-TKIs (online supplemental figure S1J).

The above findings suggest that TKIs have the potential to modulate the tumor immune microenvironment, thereby influencing the development of TKI resistance through their impact on immune cell function in the presence of tumor cells.

Short-term TKI treatment activates T-cell functionality

Focusing on the impact of short-term TKI treatment on T-cell function, we particularly examined the activation of the type I IFN signaling pathway at various TKI concentrations. We first observed dose-dependent increases in the phosphorylation of the key proteins TBK1 and IRF3 following EGFR-TKI (osimertinib) treatment, and ALK-TKI (lorlatinib) treatment led to the activation of Phosphorylation-Nuclear factor kappa B(p-NFκB) (figure 2A and online supplemental figure S2A). Correspondingly, increases in the secretion of INF-α and INF-β post-TKI treatment were noted, underscoring the capacity of TKIs to rapidly activate T-cell immune functions in a dose-dependent manner (figure 2B). Further investigations, focusing on the expression levels of the STING protein, revealed that EGFR-mutant cell lines exhibited relatively lower expression levels of STING pathway proteins compared with ALK fusion-positive cell lines (online supplemental figure S2B), while TKI treatment did not significantly alter their expression (online supplemental figure S2C). These results suggest that STING may not be the dominant player in the immune activation process induced by TKIs. Furthermore, we noted that short-term treatment with osimertinib resulted in significant dose-dependent and time-dependent upregulations of the type I IFN signaling pathway. Notably, this phenomenon first increased and then decreased in HCC827 cells (figure 2C,D). A contrasting downregulation of the pathway was observed in drug-resistant cell lines treated with a first-generation TKI, such as gefitinib (online supplemental figure S2D). Then, we found that administration of the type I IFN blocker anifrolumab weakened the cell-killing effect induced by the combination treatment of TKIs and T cells (figure 2E and online supplemental figure S2E), indicating that type I IFN may function as the mediator underlying the T cell-mediated antitumor effect induced by TKIs. Additionally, our research further revealed that short-term TKI treatment led to the suppression of PD-L1 expression (figure 2F and online supplemental figure S2F). Further, we observed increases in INF-γ and TNF-α levels in T cells co-cultured with ALK fusion-positive cell lines post-TKI treatment, while decreases were commonly noted in T cells co-cultured with EGFR-mutant cell lines (figure 2G,H). Moreover, as the medication time increased, the expression levels of CD8 and GB in EGFR-TKI-treated T cells co-cultured with tumor cells first increased and then decreased, while the expression levels of ALK-TKI-treated proteins decreased and then increased (figure 2I). These results suggest that short-term TKI treatment has a complex impact on enhancing the immune response against tumors, and EGFR-TKI and ALK-TKI have distinct effects on the TME.

Figure 2

Short-term treatment with TKI activates T-cell function. (A) Levels of phosphorylated AKT (p-AKT), p-TBK1, and p-IRF3 in cancer cells treated with different concentrations of TKI for 12 hours, analyzed using western blotting. (B) ELISA analysis of IFN-α and IFN-β levels in cancer cells following treatment with different concentrations of TKI (0, 1, or 5 µM) for 24 hours. (C) Levels of the indicated proteins in cancer cells treated with different concentrations of TKI for 48 hours, analyzed using western blotting. (D) Levels of the indicated proteins in tumor cells treated with 1 µM TKI for different periods of time, analyzed by western blotting. (E) Cell viability CCK-8 assay for cells treated with TKIs (osimertinib: 10 nM, lorlatinib: 10 nM), activated T cells (1:1 ratio to cancer cells), anifrolumab (10 µg/mL), or the combination. (F) Levels of PD-L1 in cancer cells treated with different concentrations of TKI for 12 hours, analyzed by western blotting. (G) Schematic workflow for the co-culture of cancer cells and T cells in Transwells. (H) ELISA analysis of the TNF-α and IFN-γ protein expression levels in T cells treated under the indicated conditions for 24 hours. (I) Levels of CD8 and GB in T cells treated under the indicated conditions, analyzed by western blotting. EGFR, epidermal growth factor receptor; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; GB, granzyme B; IFN, interferon; IRF3, interferon regulatory factor 3; PD-L1, programmed death ligand 1; TBK1, TANK-binding kinase 1; TNF, tumor necrosis factor; TKI, tyrosine kinase inhibitor.

Long-term TKI treatment induces resistance in NSCLC and fosters an immunosuppressive microenvironment

We then asked how TKI treatment impacted the immune microenvironment in NSCLC over the long-term. By comparing sensitive cell lines and their counterpart TKI-resistant cell lines, we observed that resistant cell lines demonstrated notable resistance to T cell-mediated cytotoxicity compared with their parental cell lines (figure 3A,B and online supplemental figure S3A,B). Further molecular-level analyses revealed significant changes in the expression of immune checkpoint molecules compared with the sensitive cells, and western blot analysis showed that the expression of TIM3, an immune checkpoint molecule, in T cells co-cultured with osimertinib-resistant cells increased, while the expression levels of GB and CD8, key proteins related to T-cell activity, decreased; furthermore, the levels of secreting proinflammatory cytokines (eg, INF-γ) also decreased. Following long-term treatment with osimertinib or in osimertinib-resistant cells, the expression levels of TIM3 and PD-1 were increased, while GB and CD8 expression levels were decreased, as was the secretion of INF-γ. The expression levels of TIM3, PD-1, and CD8 and INF-γ secretion were decreased in TKI-resistant ALK fusion-positive cells. After lorlatinib treatment, the expression of TIM3 in T cells decreased, and the expression levels of PD-1, GB, and CD8 showed decreasing trends, as did the secretion of IFN-γ (figure 3C–E and online supplemental figure S3C). These results revealed that resistant cells may evade T cell-mediated immune responses by modulating surface molecules and cytokine secretion.

Figure 3

Long-term treatment with TKI inactivates T-cell function. (A) Cell viability CCK-8 assay for cells treated with T cells (at the indicated ratios to cancer cells). (B) T cell-mediated cancer cell-killing assay. PC-9OR and H3122LR cells co-cultured with activated T cells for 48 hours were subjected to crystal violet staining. Ratio of cancer cells to T cells: 1:3. (C) Schematic workflow for the co-culture of resistant cells and T cells in transwells. (D) Levels of the indicated proteins in T cells treated with different TKIs (osimertinib: 1 µM, lorlatinib: 1 µM). (E) ELISA analysis of IFN-γ protein expression in T cells treated with TKIs under the indicated conditions for 24 hours (osimertinib: 1 µM, lorlatinib: 1 µM). (F) PD-L1 protein expression levels in parent cells and resistant cells, analyzed by western blotting. (G) Western blotting and immunofluorescence staining (H) of PD-L1 protein expression in parent cells and resistant cells following treatment with TKIs (osimertinib: 1 µM, lorlatinib: 1 µM). (I) Schematic workflow of the co-culture of resistant cell medium and macrophages. (J) Levels of CD206 and CD86 in the indicated groups, analyzed by western blotting. (K) Immunofluorescence staining with antibodies against CD68 and CD206 to detect macrophages in PC-9 xenografts following treatment with osimertinib. (L) Schematic workflow for the co-culture of macrophages, resistant cells, and T cells by direct contact. (M) T cell-mediated cancer cell-killing assay. PC-9 and H3122 cells co-cultured under the indicated conditions for 48 hours were subjected to crystal violet staining. Ratio of cancer cells to T cells: 1:3. (N) Cell viability CCK-8 assay for T cells treated under the indicated conditions (ratio of cancer cells to T cells: 1:3). GAPDH, glyceraldehyde 3-phosphate dehydrogenase; IFN, interferon; PD-L1, programmed death ligand 1; TIM3, T-cell immunoglobulin and mucin domain-containing protein 3; TKI, tyrosine kinase inhibitor.

Moreover, we observed a series of significant immune-related changes in TKI-resistant cell lines, such as a decrease in INF-α levels (online supplemental figure S3D). Western blot analysis further revealed decreases in p-TBK1 and p-IRF3 expression in drug-resistant cell lines (online supplemental figure S3E), reflecting the strategies employed by drug-resistant cells to evade immune surveillance by weakening type I IFN production. Then, we observed that PD-L1 expression decreased after TKI treatment in parental cells, while it increased in EGFR mutation-carrying resistant cells; however, there was no significant change in ALK fusion-positive resistant cells (figure 3F–H and online supplemental figure S3F–H), suggesting that ALK fusion-positive resistant cells might not suppress immune responses through PD-L1-mediated pathways.

Through co-culture with conditioned medium with TKI-resistant cells, we observed an increase in the expression of the M2 macrophage marker CD206 and a decrease in the expression of the M1 macrophage marker CD86 (figure 3I,J). The enrichment of M2 macrophages was further confirmed through histological analysis, indicating that TKI-resistant cells not only altered direct immunosuppressive molecules but also affected immune cells like macrophages in the TME, further enhancing immunosuppression (figure 3K). Additional experiments revealed the significant inhibitory impact exerted by TKI-resistant cells and their associated M2-polarized macrophages on the cytotoxic function of T cells (figure 3L–N), highlighting the critical role of tumor-induced macrophage polarization in the attenuation of T cell-mediated antitumor responses. Taken together, these findings suggest that long-term TKI treatment might contribute to dampening the direct regulation of T-cell immune efficacy in TKI-resistant cells and indirect regulation by inducing the polarization of M2 macrophages.

TKI treatment enhances T-cell infiltration by downregulating LAMC2 expression

We then explored how short-term TKI therapy enhances T-cell infiltration within tumors. Through the analysis of mIF data from pre-TKI and post-TKI treatment tissue samples, we observed a significant increase in immune cell infiltration in the tumor area following short-term TKI treatment (figure 4A). Then, a significant shrinkage of transplanted tumors (PC-9 and H3122 cells) and an increase in CD8+ T cells were observed after TKI treatment (figure 4B–D and online supplemental figure S4A), emphasizing the important role of short-term TKI treatment in activating the TME.

Figure 4

TKI treatment promotes T-cell infiltration by inhibiting LAMC2 protein expression. (A) Density (cells/mm2) of CD8+, CD8+GB+, CD68+, and CD68+CD163+ cells in tumors and stroma, quantified by multiplex immunofluorescence in paired pretreatment tumor samples and resected tumors. (B) PC-9 cells and H3122 cells were injected into Hu-HSC mice (n=3 mice per group) on day 0, and osimertinib or crizotinib was administered as indicated. (C) Tumor growth curves of cell-derived xenograft mouse models treated as indicated. Tumor volume is shown as the mean±SEM (n=3). The tumor weight (g) of each mouse is shown. (D) Immunohistochemistry analysis of CD8 expression in tumor sections from the different groups. Representative images are shown. (E) LAMC2 expression levels in the indicated cells, analyzed by western blotting. (F) ELISA analysis of LAMC2 levels in PC-9 cells and H3122 cells following treatment for 24 hours. (G) Schematic diagram of a T-cell chemotaxis assay directly regulated by the addition of medium from PC-9 and H3122 cells. (H) Representative images of infiltrating T cells stained with CFSE dye. (I) Immunofluorescence staining with antibodies against CD8 and LAMC2 in tissues harboring EGFR mutations and ALK fusions. (J) LAMC2 expression levels in cancer cells treated with different concentrations of TKIs for 12 hours, analyzed by western blotting. (K) Representative images of infiltrating T cells stained with CFSE dye, with TKI inhibition directly regulated by the recombinant protein Ln-γ2. (L) Immunofluorescence staining with antibodies against CD8 and LAMC2 in non-small cell lung cancer tissues harboring EGFR mutations and ALK fusions following treatment with TKIs. ALK, anaplastic lymphoma kinase; CFSE, carboxyfluorescein diacetate succinimidyl ester; DAPI, 4′,6-diamidino-2-phenylindole; EGFR, epidermal growth factor receptor; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; LAMC2, laminin subunit γ−2; PBS, phosphate-buffered saline; TKI, tyrosine kinase inhibitor.

Previous studies have shown that LAMC2 expression plays a key role in regulating the ability of T cells to penetrate and attack tumors.32 We first analyzed LAMC2 expression in a Gene Expression Profiling Interactive Analysis (GEPIA) data set (http://gepia.cancer-pku.cn/) and found LAMC2 to be highly expressed in patients with lung cancer (online supplemental figure S4B). Then, the differential LAMC2 expression patterns observed in western blot and ELISA analyses revealed that tumor cells positive for ALK fusions exhibited significantly lower LAMC2 expression and secretion compared with cells with EGFR mutation, which may explain why better clinical outcomes were achieved with ALK-TKI (figure 4E,F). T-cell infiltration experiments further confirmed that EGFR mutant cells more strongly impeded the infiltration of immune cells compared with ALK fusion cells (figure 4G,H). As observed in tissue immunofluorescence analysis, LAMC2 expression in EGFR mutant tumor tissue was higher than that in ALK fusion tissue (figure 4I), suggesting that LAMC2 expression in EGFR mutant tumors leads to reduced T-cell infiltration, thereby limiting the sensitivity of tumors to immune responses.

Our study found that LAMC2 expression was significantly inhibited by short-term treatment with various TKIs (online supplemental figure S4C). We also observed a dose-dependent decrease in LAMC2 protein expression after TKI treatment (figure 4J and online supplemental figure S4D–E). Immunohistochemical analysis of mouse transplanted tumor tissues further confirmed that TKI treatment could inhibit the expression of LAMC2 protein (online supplemental figure S4F). Moreover, T-cell infiltration experiments demonstrated that short-term TKI treatment inhibited the expression of LAMC2, thus promoting T-cell infiltration (figure 4K). In mouse models, immunofluorescence labeling revealed a significant correlation between the decrease in LAMC2 protein expression in TKI-treated tumor tissues and the increase in T-cell infiltration (online supplemental figure S4G). Pathological tissue immunofluorescence analysis of human tumor samples confirmed that TKI treatment reduced LAMC2 protein expression in tumor tissues, accompanied by a significant increase in T-cell infiltration (figure 4L).

These results collectively indicate that TKI treatment effectively alters the TME by regulating the expression of LAMC2 protein, enhancing T-cell infiltration and activity.

Long-term TKI treatment induces LAMC2 overexpression and inhibits T-cell infiltration

To explore the effects of long-term TKI treatment on the TME, we detected the expression of LAMC2 at different time points (2, 4, and 6 weeks) during osimertinib treatment in PC-9 xenografted tumors and observed that LAMC2 expression gradually increased in a time-dependent manner; in contrast, it was suppressed in the early stages of short-term TKI treatment (figure 5A), consistent with the results in published RNA sequencing data (GSE193258) (online supplemental figure S5A). Subsequent pathological tissue immunofluorescence analysis revealed significant overexpression of LAMC2 protein following long-term TKI treatment and resistance, accompanied by a significant decrease in T-cell infiltration (figure 5B), indicating a shift in the TME toward an immunosuppressive state. To delve deeper into the role of the LAMC2 protein in tumor immune evasion, we observed increased expression of LAMC2 in TKI-resistant cells (figure 5C–E and online supplemental figure S5B); in addition, LAMC2 overexpression could not decrease the sensitivity to TKI in the PC-9 and H3122 cell lines. These results bolstered the hypothesis that prolonged TKI treatment may induce alterations in the TME, particularly affecting immune cell infiltration and functionality. By employing ELISA and T-cell infiltration assays, we established a link between increased LAMC2 secretion in TKI-resistant cells and a corresponding decrease in T-cell infiltration capabilities (figure 5F,G). These observations underscore the pivotal role of the LAMC2 protein in modulating the tumor immune environment. We also observed that inhibiting LAMC2 protein expression in resistant cells effectively restored the T cell-mediated tumor-killing functions (figure 5H–J, online supplemental figure S5E,F). These findings are crucial for understanding the long-term effects of TKI treatment and their impact on the tumor immune environment, offering a new perspective on the profound influence of long-term TKI treatment on the TME.