Non-Small Cell Lung Cancer: Challenge and Improvement of Immune Drug Resistance

Lung cancer is the leading cause of cancer deaths in the world. At present, immunotherapy has made a great breakthrough in lung cancer treatment. A variety of immune checkpoint inhibitors have been applied into clinical practice, including antibodies targeting the programmed cell death-1, programmed cell death-ligand 1, and cytotoxic T-lymphocyte antigen 4. However, in the actual clinical process, about 30%–50% of patients still do not receive long-term benefits. Abnormal antigen presentation, functional gene mutation, tumor microenvironment, and other factors can lead to primary or secondary resistance. In this paper, we reviewed the immune mechanism of immune checkpoint inhibitor resistance, various combination strategies, and prediction of biomarkers to overcome resistance in order to accurately screen out the advantageous population, expand the beneficiary population, and enable precise and individualized medicine.

of drug resistance and screen the beneficiaries. In this review, the known mechanisms of immune resistance and potential therapeutic strategies to reverse immune resistance and to predict poor prognosis are reviewed.

DRUG RESISTANCE MECHANISM Abnormal Antigen Presentation
The activation of T cells requires two signals. The first signal is the T-cell receptor (TCR) signal formed by the combination of TCR and peptide-major histocompatibility complex (MHC) molecules, but this signal is not enough to activate resting T cells. Only in the case of the second costimulatory signal provided by CD28 and its receptor, T-cell activation-related RNA and proteins will be synthesized, the key cytokine interleukin (IL)-2 will be secreted, and cells will enter from G0 phase to G1 phase. Therefore, it is the costimulatory signal and TCR signal that complete the activation of T cells. Studies have shown that B2M, as an important part of human leukocyte antigen (HLA)-I molecules, participates in the folding and transport of MHC-I molecules and plays an important role in the processing and presentation of tumor antigens. B2M mutation can lead to impaired expression of MHC-I molecules on the surface of antigen-presenting cells (APCs) and then lead to impaired antigen presentation, resulting in immunotherapy resistance (2). In addition, negative costimulatory molecules such as cytotoxic T-lymphocyte antigen 4 (CTLA-4) and programmed cell death-1 (PD-1) and their ligands CD80, CD86, PD-L1, and PD-L2 can prevent the body from producing second signals, resulting in downregulation or termination of T-cell activation (3).
Immune Cells: Tumor-Associated Macrophages, Myeloid-Derived Suppressor Cells, Regulatory T Cells A large number of studies have shown that immune cells play a key role in tumor progression and inflammation. First, tumorassociated macrophages (TAMs) showed significant plasticity toward environmental cues (4,5). In the early stage of the tumor, TAMS mainly showed M1 phenotype, while in the late stage, TAMS mostly belonged to M2 phenotype (6). M1 macrophages are pro-inflammatory cells, but they have an antitumor effect, which is related to the cytotoxicity and immunostimulatory function to cancer cells. M2 macrophages expressing antiinflammatory cytokines such as IL-10, C-C motif chemokine ligand 22 (CCL22), and CCL18 can reduce inflammatory response but can promote tumors due to immunosuppression and angiogenesis induction (4,7,8). In tumors, microenvironment, such as hypoxia, nitric oxide (NO), can promote TAMs to M2 polarization. In addition, macrophage colony-stimulating factor (M-CSF) produced by tumor cells can also promote the polarization of TAMs to M2, resulting in tumor escape. The main secretion of M2 suppresses cytokines IL-10 and transforming growth factor-b (TGF-b), and the presence of antigen is weak, which inhibits T-cell activation and contributes to tumor immunity (7,8).
Second, regulatory T (Treg) cells exert their immunosuppressive function through a variety of mechanisms. The high expression of IL-2 receptor on the surface of Treg cells can neutralize IL-2 to limit the proliferation and activation of T cells and produce inhibitory cytokines (TGF-b, IL-10, and IL-35) and cytotoxic substances (perforin and granulase) to inhibit and kill the excitation of effector T cells (9). CTLA-4 expressed by Treg cells binds to CD80/86 to impair APC maturation and inhibit T-cell proliferation, such as dendritic cells (DCs) (10). In addition, Treg cells had low expression of Nrf2, which is a key transcription factor of antioxidant reaction. Oxidative stress can induce apoptosis of Treg cells and release a large number of ATP, which is metabolized into adenosine by CD39 and CD73, which are highly expressed in Treg cells. Adenosine binds to A2A receptor (A2AR) to inhibit effector T cells (9,11).
Last, myeloid-derived suppressor cells (MDSCs) can promote tumor growth through immunological inhibition and nonimmunological inhibition, in which immunosuppressive mediator ARG1 and inducible NO synthase (INOS) can decompose L-arginine into 1-ornithine and urea, NO and nitrite, an important mediator of the IL-2 pathway, resulting in T-cell expression incompetence (12,13). MDSC also expressed a high level of indoleamine 2mine3-dioxygenase (IDO), which can degrade 1-tryptophan to N-formylcanine, inhibit the proliferation and activation of T cells and NK cells, and promote CD4+ T cells to differentiate into Treg (14,15). In addition, MDSC secretes immunosuppressive cytokines and growth factors (TGF-b and IL-10) to reduce the antitumor activity of effector T cells, recruit Treg cells, and increase reactive oxygen species (ROS) and NO in the microenvironment to inhibit the antitumor activity of natural killer (NK) cells and effector T cells (7,(16)(17)(18). The latest research shows that MDSCs can exert an immunosuppressive effect by upregulating PD-L1 (19). In addition, MDSCs can also promote tumor progression through non-immunological mechanisms (20). MDSCs can produce a large number of matrix metalloproteinases (MMPs), especially MMP9, to promote the infiltration of metastatic cells (21) and secrete high levels of vascular endothelial growth factor (VEGF) and basic fibroblast growth factor (bFGF) to promote angiogenesis (22). To sum up, MDSCs play an important role in the occurrence and development of tumors.

Tumor Endothelial Cells
Solid tumors tend to secrete a variety of pro-angiogenic factors, such as VEGF, hepatocyte growth factor, and platelet-derived growth factor. In 1971, Folkman (23) proposed that tumor growth was angiogenesis dependent, with further research. From 1983 to 1989, Senger et al. (24) proposed vascular permeability factor (VPF)/VEGF and Ferrara et al. (25) established the important position of VEGF until Terman isolated and purified VEGFR2 in the 1990s. These results fully indicate that VEGF (VEGFR2) plays an important role in the process of tumor growth, recurrence, and metastasis. VEGF/ VEGFR is expressed in most tumors, including non-small cell lung cancer (NSCLC), and it has been found to increase the risk of tumor recurrence, metastasis, and death. Angiogenic factors are continuously secreted in the tumor microenvironment, resulting in abnormal angiogenesis. On the one hand, neovascularization usually lacks some adhesion molecules, and the downregulation of adhesion molecules leads to T-cell extravasation (26). At the same time, circulating VEGF hinders the maturation and function of DCs and helps tumors escape immune surveillance. On the other hand, neovascularization cannot offset the increase in oxygen consumption, so the hypoxia environment will directly damage the function of tumor-infiltrating lymphocyte (TIL). In addition, on the one hand, hypoxia can upregulate the inhibitory signals of antitumor immune response, such as PD-L1, IDO, IL-6, and IL-10 (27).
On the other hand, hypoxia induces upregulation of chemokine expression, which makes Treg cells reenter the tumor (28,29). In addition, hypoxia can also promote the polarization of TAM to M2-like phenotype of TAMs (30). To sum up, angiogenesis can participate in tumor growth and immune escape through a variety of ways.

Functional Gene Mutation and Inactivation
PTEN gene plays an important role in maintaining cell proliferation, differentiation, and apoptosis. PTEN can inhibit phosphoinositide 3-kinase (PI3K) pathway, which plays a regulatory role in some key cell processes such as tumor survival and proliferation. On the other hand, the lack of PTEN expression can activate the PI3K-AKT pathway, thus reducing the infiltration of lymphoid T cells and reduce the tumor killing effect of effector T cells (31). Similar to Janus kinase (JAK), it plays an important role in cytokine signal transduction. The JAK protein tyrosine kinase family consists of four members: tyrosine kinase (TYK)2, JAK1, JAK2, and JAK3. Patients with JAK1/2 gene mutations may be resistant to primary immunotherapy (32). Although JAK2 mutant tumor cells can produce interferon (IFN)-g, the JAK2-signal transducer and activator of transcription _(STAT) signal pathway cannot be activated by IFN-g and cannot upregulate the expression of PD-L1, which leads to the weak killing effect of IFN on JAK2 mutant tumor cells. However, the mutant cells of JAK1 were not sensitive to all kinds of effects of IFN. The above results suggest that the mutant tumor cells of the JAK1/JAK2 gene are not sensitive to the killing effect of IFN, and the expression level of PD-L1 is low, which makes the mutated tumor cells resistant to ICIs (33).

RAISING DRUG RESISTANCE Combined Application With Chemotherapy
Immune combined chemotherapy can not only increase the cross-presentation of antigens by dendritic cells (34) but also weaken the immunosuppressive components of the tumor microenvironment (35), such as Treg cells, MDSCs, immunosuppressive cytokines, etc., and then increase toxic lymphocytes and the ratio of Treg cells (36)

Combined Radiotherapy With Immunotherapy
Preclinical evidence points to RT as a priming event for immunotherapy. By modulating the host's immune system, RT can render tumor cells more susceptible to T cell-mediated attack. RT promotes the release of tumor neoantigens from dying tumor cells, enhances MHC-I expression, and upregulates chemokines, cell adhesion molecules, and other immunomodulatory cell surface molecules, thereby potentiating an antitumor immune response by triggering immunogenic cell death (40). Just as the PACIFIC trial evaluated the efficacy of durvalumab consolidation therapy in patients with NSCLC after simultaneous RT and chemotherapy, the results showed that the mPFS was 16.9 vs. 5.6 months (HR: 0.55, 95% CI: 0.45-0.68), and the mOS was 47.5 vs. 28.1 months (HR: 0.55, 95% CI: 0.45-0.68) (41). In addition, a retrospective study at the 2020 ASCO conference showed that local treatment can significantly improve survival benefits (42). After immune resistance, local therapy combined with immunotherapy can reverse drug resistance to some extent, providing new treatment ideas for patients with immune drug resistance.

Combined With Targeted Drugs
Anti-angiogenic therapy can change the function of tumor vascular endothelial cells to regulate immunosuppression and reduce the inhibitory effect of VEGF on DC migration and immune function (43,44). Studies have shown that CTLA-4 or PD-1 inhibitors can reduce tumor vascular density, improve vascular perfusion, relieve hypoxia of tumor tissue, normalize blood vessels, and reduce the immunosuppressive effect of Treg, TAMS, and MDSCs (44)(45)(46). Therefore, the combination of immunosuppressive and anti-angiogenic drugs may play a synergistic role. As the latest results of the IMpower150 trial were presented at the American Association for Cancer Research (AACR) 2020 meeting of the AACR, the mOS of ABCP (atezolizumab+carboplatin+paclitaxel+bevacizumab) vs. BCP (carboplatin+paclitaxel+bevacizumab) was 19.5 vs. 14.7 months (HR: 0.80, 95% CI: 0.67-0.95), the mPFS was 8.4 vs. 6.8 months, and the ORR was 63.5% vs. 48.0%, mOS was 29.4% vs. 18.1%, respectively. In the liver metastasis subgroup, the mOS of ABCP vs. BCP was 13.2 vs. 9.1 months, and the PFS was 8.2 vs. 5.4 months. From the data survey, it can be seen that compared with BCP, ABCP can significantly improve the PFS and OS of patients (47,48). From the above trial results, for epidermal growth factor receptor (EGFR) and anaplastic lymphoma kinase (ALK) mutant population, the combination of four drugs can bring significant survival benefits to patients with advanced NSCLC (49).
The Phase I/II KRYSTAL-1 test (NCT03785249) included patients with KRASG12C mutation-positive, unresectable, or metastatic NSCLC. After adagrasib monotherapy, the results showed that in patients with NSCLC, the overall remission rate of adagrasib treatment reached 45% and the disease control rate (DCR) reached 96%, and for NSCLC patients with STK11 mutation, the overall remission rate reached 64%. In terms of safety, the most common AEs of adagrasib treatment included nausea (54%), diarrhea (51%), vomiting (35%), and fatigue (32%) (50,51). The researchers indicated that they plan to verify the efficacy of adagrasib in combination with other drugs or treatments in future trials, such as pembrolizumab, Keytruda (52). The Phase II KRYSTAL-7 trial (NCT04613596) of adagrasib combined with pimumab is under preparation, and the results are worth looking forward to. And now there are more and more data supporting the combination of immunotherapy and other targeted treatments.

COMBINED APPLICATION OF DOUBLE IMMUNITY
PD-1 is highly expressed on T cells and interacts with its ligands PD-L1 and PD-L2 to inhibit T-cell activation and proliferation (53). CTLA-4 can reduce T-cell activity to maintain immune tolerance and homeostasis (54). The two interact to enhance the efficacy of immunotherapy. For example, in patients with advanced melanoma, the results showed that the ORR and PFS of double immunotherapy were higher than those of single-drug immunotherapy (55,56). The Checkmate-227 test was used to evaluate the efficacy of dual immunotherapy and chemotherapy alone in the treatment of advanced NSCLC. The study showed that the OS and duration of response (DOR) of patients with PD-L1 ≥1% and PD-L1 <1% were significantly longer than those in the chemotherapy group (PD-L1-positive patients: mOS was 17.1 vs. 14.9 months, ORR value was 31.9% vs. 30%; mOS was 17.2 vs. 12.2 months, ORR value was 27.3% vs. 23.1%) (57,58). In addition, like the Checkmate-227 study, the Checkmate-9LA trial, presented at the 2020 World Conference on Lung Cancer (WCLC), aims to evaluate the efficacy of chemotherapy alone or navulizumab combined with two cycles of chemotherapy in the treatment of metastatic NSCLC in Asian populations. The results showed that no matter what the expression of PD-L1 was, the combination of immunotherapy and chemotherapy could improve the expression of OS, without considering the expression of PD-L1. According to the expression analysis of PD-L1, the risk of death of patients with PD-L1 <1% (HR: 0.62, 95% CI: 0.45~0.85) and PD-L1 ≥1% (HR: 0.64, 95% CI: 0.5~0.82) decreased by 38% vs. 36% (HR: 0.64, 95% CI: 0.5~0.82) (59,60). In addition, for the study of immune checkpoints, some new checkpoints have been explored, such as T-cell immunoreceptor with immunoglobulin (Ig) and ITIM domains (TIGIT), IDO, and lymphocyte activation gene (LAG)-3. Among the many new combinations and new targets, T-cell immune receptor (TIGIT) inhibitors carrying Ig and ITIM domains have attracted particular attention. A phase II study of atrizumab combined with TIGIT inhibitor tiragolumab vs. placebo combined with atrizumab was released at the 2020 ASCO Congress (61). The main characters were ORR and PFS. The results showed that for people with PDL1 >50%, tiragolumab+atrizumab significantly increased ORR and PFS time in the intention-to-treat (ITT) population compared with those in the control group, but for 1% 49% of PD-L1, the benefit of ORR and PFS time was limited. This study is still under further study, and the results are worth looking forward to.

DISCUSSION
Although the research ideas of immunotherapy drug resistance emerge endlessly, there are also some problems that cannot be ignored in the clinical trials of reversing immune drug resistance. In recent years, with the rapid development of medicine, we advocate "individualized medical treatment" and "precision treatment" in the field of oncology. Biomarkers have important clinical significance for the discovery, treatment, and prognosis of tumors. At present, PD-L1 is the most commonly used marker for predicting the efficacy of immunotherapy, but it still has some limitations and cannot be used as a routine marker in the clinic. Other related studies have shown that tumor mutation load, TILs, and microsatellite instability are important biomarkers to predict the efficacy of immunotherapy.
According to the retrospective analysis of CHECKMate-026 and the survival data of CHECKMate-227, the efficacy of immunotherapy in patients with high tumor mutation burden (TMB) was significantly better than that of chemotherapy (57,62). In 2019, the National Comprehensive Cancer Network  (64).
The WCLC in 2021 also reported on CtDNA. In addition, new biomarkers are being discovered. For example, a retrospective trial was conducted in Japan to evaluate the efficacy of nivumab or pembrolizumab in patients with advanced NSCLC with positive antibodies (rheumatoid factor, antinuclear antibodies, thyroid antibodies) before selection. The results showed that the ORR, DCR, and PFS of antibody-positive patients were significantly better than those of non-antibody-positive patients. Although the sample size included in this study is relatively small, its research is worthy of further study (65). At present, with the further study of biomarkers, some studies also have found that Tumor Protein p53 (TP53) or Kirsten Rat Sarcoma Viral Oncogene Homolog (KRAS) gene mutations can increase the expression of PD-L1 and CD8+ T-cell infiltration; when both mutations are present, the expression of PD-L1 will be more significant, the tumor mutation load is often high, and the clinical benefits of pembrolizumab therapy are often better. Therefore, KRAS or TP53 mutations combined with PD-L1 or CD8+ T cells may be of better predictive value (66).

CONCLUSION
There is still a certain problem of drug resistance in the clinical application of immunotherapy for NSCLC, although the combination of drugs can expand the effect of immunotherapy and improve immune drug resistance. However, based on these results, there are still several important problems to be solved: how to accurately grasp the dose and time sequence of the combination, and the choice of the dominant population is the key to improving the efficacy of drugs. The development of immunotherapy is a great breakthrough in tumor therapy and the cornerstone of advanced NSCLC therapy. In the future, looking for biomarkers to predict the efficacy of immunotherapy can make immunotherapy individualized and accurate. The research and development of new drugs will further improve the prognosis of patients with lung cancer ( Table 1).

AUTHOR CONTRIBUTIONS
FK and ZW contributed equally to this work. Other authors offer advice. All authors contributed to the article and approved the submitted version.