Lung cancer is the leading cause of cancer-related deaths globally, and by 2030, it is projected to cause over 2 million deaths. Recognizing its high mortality has produced immense pressure on the medical world to create treatments that have a better therapeutic response while maintaining a tolerable safety profile (1). According to the World Health Organization (WHO), lung cancer can be classified into two categories based on its cell type: non-small cell lung cancer (NSCLC), which accounts for 80–95% of cases, and small cell lung cancer (SCLC), which accounts for the remaining cases. NSCLC is further categorized into adenocarcinoma, squamous cell carcinoma, and large cell carcinoma (2). Currently, the primary treatment options for lung cancer include surgery, radiation therapy, chemotherapy, immunotherapy, and targeted drug therapy. Yet, the prognosis of lung cancer remains poor, with a 5 year survival of 10–20% (3)

Part of the high mortality is due to late diagnosis, as early disease is typically asymptomatic, and when symptoms do appear, they are often mistaken for benign respiratory conditions. Furthermore, screening uptake remains limited worldwide, as current guidelines recommend low-dose CT (LDCT) scans only for high-risk individuals aged 50–80 years with a 20-pack-year or more smoking history who either currently smoke or quit within the past 15 years (4). Evidence demonstrates that low-dose CT scans significantly reduce lung cancer mortality in this high-risk population; however, a substantial proportion of screening results are false positives, often leading to additional investigations and invasive procedures. Concerns regarding overdiagnosis, patient anxiety, and cumulative radiation exposure have further limited enthusiasm for this screening modality (5).

In addition, definitive local treatments such as surgery or localized radiation may fail due to the aggressive biological nature of lung cancer and its tendency for early metastasis. Even tumors that appear resectable may pose significant challenges because of the lung’s proximity to vital structures (6). Systemic therapies, including chemotherapy and targeted treatments, are further limited by the high molecular heterogeneity of lung tumors, and even when initially effective, tumors frequently acquire new mutations and develop therapeutic resistance (7).

Decades of research spent on understanding cancer have established a key pathway for cancer development: when the body’s immune system fails to recognize and eradicate malignant cells. This failure allows for the uncontrolled proliferation of cancer cells. Recognizing this has encouraged investigators to design strategies that aid the immune system’s natural capacity to detect and destroy such cells. This has led to the development of multiple forms of cancer immunotherapy, most notably immune checkpoint inhibitors, which have revolutionized the treatment of non-small cell lung cancer. Antibodies targeting Programmed Cell Death Protein 1 (PD-1), Programmed death-ligand 1 (PD-L1), and Cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) have produced durable responses in a subset of patients; however, most individuals either fail to respond or develop acquired resistance, highlighting the need for more potent and personalized immune-based strategies (8, 9).

Beyond immune checkpoint blockade, a new class of therapies has emerged based on the direct infusion of tumor-reactive immune cells. Adoptive cellular immunotherapies (ACTs), which include Chimeric Antigen Receptor T-cell therapy (CAR-T), Chimeric Antigen Receptor Natural Killer cell therapy (CAR-NK), and Tumor-Infiltrating Lymphocyte therapy (TIL), can modify a patient’s immune system to target the tumors more effectively (10). These approaches represent a new hope in the treatment of cancers that are resistant to conventional therapies. CAR-T therapy involves modifying a patient’s own T cells by inserting a gene that enables them to express a Chimeric Antigen Receptor (CAR), which then allows the modified T cells to recognize antigens expressed on the surface of cancer cells (10). Since its first clinical approval in 2017, CAR-T therapy has been revolutionary in the treatment of hematologic malignancies, demonstrating high remission rates in B-cell acute lymphoblastic leukemia (ALL) and non-Hodgkin lymphoma (NHL) (11). However, despite its transformative success in B-cell cancers, the application of CAR-T therapy to solid tumors, including lung cancer, remains in its early stages (10). Additionally, there is increasing interest in CAR-NK cell therapies due to their intrinsic cytotoxicity, lower risk of graft-versus-host disease, and potential for off-the-shelf allogeneic use. Likewise, tumor-infiltrating lymphocyte (TIL) therapy is another promising ACT approach that involves harvesting lymphocytes directly from the tumor site, expanding them ex vivo, and reinfusing them into the patient (11).

With adoptive cell immunotherapy’s potential to selectively target only cancer cells, the debilitating side effects of chemotherapy and radiation therapy could potentially be reduced. Therefore, this review provides a comprehensive synthesis of the latest advances in next-generation adoptive cellular immunotherapies, critically evaluates emerging clinical evidence in lung cancer, and identifies the key scientific and translational challenges that must be overcome to realize the full therapeutic potential of these approaches in broader clinical practice.

This review article analyzes recent literature on the advancement of next-generation immune cell therapies for lung cancer, with a focus on CAR-T, NK, and TIL strategies. Relevant literature published between January 2015 and September 2025 was selected for inclusion. A total of 120 studies were initially reviewed using the databases PubMed, Scopus, and Google Scholar. Key search terms included “CAR-T,” “CAR-NK,” “tumor-infiltrating lymphocytes,” “TIL,” “immune cell therapy,” and “lung cancer.” These terms were combined with Boolean operators to refine the search and capture relevant studies. Boolean operators such as AND, OR, and NOT were used to optimize search results. Synonyms, alternative spellings, and Medical Subject Headings (MeSH) terms were also considered. Only studies published in English were included to ensure consistent data interpretation. Ultimately, 81 studies underwent full-text review, comprising peer-reviewed preclinical and clinical studies, review articles, and meta-analyses investigating CAR-T, NK, or TIL cell-based treatments for lung cancer. Additional sources were identified through citation tracking and manual review of reference lists from relevant publications. After removing duplicates, 74 studies were included in this review article. Data from these sources were qualitatively synthesized and thematically organized by therapy type to illustrate recent scientific progress and future perspectives. The limitations of individual studies were also analyzed and reported to provide a balanced assessment of next-generation immune cell therapies targeting lung cancer.

Adoptive Cell Therapy (ACT) has proven to transform oncology in the last decade, allowing patients to use their own or donor-derived immune cells to achieve potent antigen-specific tumor eradication. CAR-T cell therapy, one of the key forms of ACT, has revolutionized the treatment of hematologic malignancies. However, its use in solid tumors, specifically non-small cell lung cancer (NSCLC), remains a critical challenge. There are several primary reasons for their lack of success. A limited antigen pool is one of the core problems. Unlike hematological malignancies, in which well-characterized antigens serve as effective targets, solid tumors rarely express truly tumor-specific antigens. Another key challenge is tumor heterogeneity. Solid tumors consist of multiple subclonal populations, each with a unique genetic profile. These subclonal differences allow for immune evasion, as certain regions of the tumor may lack the targeted antigen. Off-tumor toxicity has played a role in the therapy’s limited success. Solid tumors may express target antigens that are also present in healthy tissues, increasing the risk of on-target, off-tumor effects and resulting in damage to normal organs. Lastly, the tumor microenvironment (TME) encompasses multiple factors that inactivate infiltrating CAR-T cells, including regulatory T cells, myeloid-derived suppressor cells (MDSCs), fibroblasts, and tumor-associated macrophages. The presence of cancer-associated fibroblasts and myeloid cells further forms a physical barrier to infiltrative immune cells through the overexpression of extracellular matrix proteins (12, 13).

However, multiple antigens present in NSCLC have emerged as potential targets for CAR-T therapy, suggesting a promising role for this approach in future lung cancer treatment. Epidermal Growth Factor Receptor (EGFR), which is frequently amplified in various cancer types, has been shown to be a viable target for CAR-T cells (14). Similarly, prostate stem cell antigen (PSCA) and mucin 1 (MUC1), which are often overexpressed in lung cancer, have also been explored as CAR-T targets. Dual targeting of PSCA and MUC1 has demonstrated a reduction in lung tumor burden in preclinical rat models (15). In addition, novel antigens such as erythropoietin-producing hepatocellular carcinoma A2 (EphA2), which is overexpressed in approximately 90% of NSCLC samples and is associated with poorer prognosis, are being investigated as potential CAR-T targets in early-phase clinical trials (16). Furthermore, the integration of molecular imaging techniques to visualize and quantify CAR-T cell trafficking, tumor localization, and in vivo persistence is becoming increasingly important for understanding therapeutic mechanisms and optimizing CAR-T cell design in NSCLC (17).

Studies have also focused on developing NK and CAR-NK cell therapies. NK cells are considered safer than traditional CAR-T cells and can be produced on a large scale as “off-the-shelf” treatments (18). The better safety profile of these treatments compared to CAR-T therapy is largely due to the reduced risk of cytokine release syndrome (CRS) and the minimal risk of graft-versus-host disease. This is due to differences in their mechanisms. CRS, which presents with fever, hypotension, and organ dysfunction after CAR-T therapy, is driven by the release of a different set of cytokines; CAR-T cells induce Interleukin-6 (IL-6) release, which drives CRS, whereas NK cells do not. Furthermore, the minimal risk of graft-versus-host disease is because NK cells do not require an antigen-specific T-cell receptor (TCR), unlike T cells. (19). Furthermore, engineering NK cells with CAR constructs adapted to NK-specific signaling domains (DAP10, DAP12, and CD3ζ) augments tumor recognition. DAP10 and DAP12 are intracellular adaptor proteins that activate NK receptors, and the inclusion of such proteins allows for enhanced signal transduction and cytokine production upon antigen recognition. (19). One of the significant logistical challenges of autologous CAR-T therapy is now being addressed through the standardization of NK cells derived from induced pluripotent stem cells (iPSCs) and cord blood (20).

TIL therapy has been one of the earliest form of ACTs, introduced and now being revisited for the treatment of solid tumors (21). This therapeutic strategy entails isolating tumor-infiltrating T cells from the patient, expanding them extensively ex vivo, and subsequently reinfusing them into the same individual. By comprising a broad repertoire of T-cell clones, TIL therapy is well suited to overcoming the intrinsic heterogeneity of solid tumors. Importantly, there have been no documented cases of cytokine release syndrome (CRS) associated with TIL therapy to date. In addition, TIL products are enriched for memory T cells that express chemokine receptors such as C-X3-C motif chemokine receptor 1 (CX3CR1) and C-C chemokine receptor type 4 (CCR4), which facilitate effective trafficking to and penetration of solid tumor sites (21). Collectively, these features allow TIL therapy to mount a broad, durable antitumor immune response while maintaining a favorable safety profile compared with other ACT modalities. Recent progress in TIL products for melanoma has sparked new clinical trials evaluating their safety and efficacy in NSCLC, as durable responses have been observed in subsets of patients with high baseline immune infiltration (22).

These immunotherapies show early promise in expanding therapeutic options for lung cancer, although their clinical efficacy remains under active investigation. Current research is focused on improving the antigen specificity, improving persistence, and infiltration within the TME. As this side of ACTs develops, it has become apparent that these next-generation immune cell therapies may offer long-lasting responses and improve survival rates for patients with lung cancer (23).

As discussed in the overview section, chimeric antigen receptor (CAR) T-cell therapy has advanced from an experimental method to a well-established treatment for hematologic malignancies. It is now a central focus in translational research for lung cancer. Despite its success in leukemia and lymphoma, extending CAR-T therapy to solid tumors such as non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC) remains challenging due to factors like antigen heterogeneity, a dense and immunosuppressive tumor microenvironment (TME), and limited T-cell infiltration into pulmonary tissues. Consistent with these biological constraints, early clinical studies of CAR-T cell therapy in lung cancer have demonstrated limited objective response rates and short-lived disease control, despite generally manageable toxicity profiles (24). Consequently, advances in CAR-T design, delivery, and manufacturing have been made to overcome these unique obstacles (Figure 1).

A key determinant of CAR-T efficacy in lung cancer is the selection and tuning of antigens. Targets such as carcinoembryonic antigen (CEA), human epidermal growth factor receptor 2 (HER2), epidermal growth factor receptor (EGFR), mesothelin (MSLN), disialoganglioside (GD2), receptor tyrosine kinase-like orphan receptor 1 (ROR1), mucin 1 (MUC1), glypican-3 (GPC3), delta-like ligand 3 (DLL3), and PD-L1 have been explored due to their tumor-associated expression and restricted normal-tissue distribution (25). These antigens have been prioritized based on preferential overexpression in lung tumors and their ability to mediate CAR-dependent T-cell activation in preclinical and early translational studies (25). However, heterogeneous antigen expression within tumors and overlap with normal lung tissues continue to limit therapeutic windows and contribute to off-tumor toxicity.

To mitigate on-target, off-tumor toxicity, a critical concern in lung tissue, recent efforts have employed affinity-tuned CAR constructs and dual-antigen or logic-gated CAR designs that require concurrent recognition of multiple tumor-associated epitopes for activation. Affinity-tuned mesothelin CARs have demonstrated improved tumor selectivity with reduced recognition of normal mesothelial cells, while ligand-based EGFR CARs preferentially target tumor-associated EGFR variants. The translational relevance of these safety-oriented engineering strategies is supported by early mesothelin-directed CAR-T trials, in which restricting antigen sensitivity and systemic exposure was associated with reduced off-tumor pulmonary toxicity (26–28). These approaches aim to expand the therapeutic window for CAR-T therapy in the lung.

Engineering innovations have yielded several next-generation CAR-T platforms designed to overcome immune suppression within the lung tumor microenvironment. Armored CAR-T cells engineered to secrete immunomodulatory cytokines such as IL-12 or IL-18 are intended to enhance T-cell persistence and remodel the TME. Importantly, the majority of evidence supporting cytokine-armored CAR-T cells in lung cancer remains preclinical, with IL-18–secreting DLL3-targeted CAR-T cells demonstrating enhanced persistence and antitumor activity primarily in xenograft models of SCLC (29, 30). These findings provide mechanistic support for cytokine-secreting CAR designs, while highlighting the need for further clinical validation.

Similarly, CAR-T cells co-expressing chemokine receptors such as C-X-C motif chemokine receptor 5 (CXCR5) or C-C chemokine receptor type 6 (CCR6) have been developed to improve tumor homing and infiltration. Preclinical studies in HER2-positive NSCLC models demonstrated enhanced CAR-T trafficking and antitumor activity following chemokine receptor co-expression. Although these modifications significantly improve tumor homing in preclinical settings, clinical trials to date continue to demonstrate limited intratumoral CAR-T persistence in lung cancer, indicating that trafficking enhancement alone may be insufficient to overcome immune exclusion within the pulmonary TME (31). These data suggest that combinatorial strategies will be required to sustain CAR-T activity in solid lung tumors.

Advances in universal or adaptor CAR platforms have further expanded the flexibility and safety of CAR-T therapy in solid tumors. These modular systems allow dynamic control of antigen engagement and CAR-T activity through externally administered adaptors. Such adaptable CAR platforms are particularly attractive for lung cancer, where antigen heterogeneity and dynamic antigen loss have been implicated as contributors to therapeutic resistance in early clinical experience (25, 32). The ability to retarget CAR-T cells without re-engineering may therefore address a major limitation of current lung cancer CAR-T approaches.

Delivery techniques have also evolved to accommodate the anatomical and physiological restrictions of lung cancer. Preclinical and early clinical investigations have demonstrated that regional delivery routes, such as intrapleural or intraparenchymal infusion, increase local control while reducing systemic toxicity. This kind of localized administration shows great promise for isolated thoracic disease and pleural metastases. Clinical trials are being conducted to determine the feasibility of intratumoral CAR-T delivery for MSLN-positive lung cancers, which could provide an alternative when systemic toxicity is a concern (33, 34).

From a manufacturing and translational standpoint, progress in allogeneic CAR-T platforms, where endogenous T-cell receptors and HLA are eliminated using gene editing, have cleared the way for off-the-counter treatments that can be rapidly available for patients with advanced lung cancer (35). Despite ongoing issues with persistence and host immunological rejection, these systems promise increased accessibility and shorter manufacturing times. Simultaneously, safety mechanisms such as inducible caspase-9 suicide switches, truncated EGFR markers, and small-molecule gating systems have been implemented to reduce cytokine release syndrome (CRS) and neurotoxicity, which are especially concerning in patients with impaired pulmonary reserve (32) (Figure 2).

Clinical trials using CAR-T cell therapy for lung cancer are still in their early stages, but evidence is growing. Early-phase investigations with MSLN-directed CAR-T cells showed feasibility and controllable toxicity, whereas ligand-based EGFR CAR-T cells showed promising anticancer potential in NSCLC models and preliminary human testing. The IL-18-armored DLL3 CAR has emerged as a promising option for SCLC, with improved persistence and cytotoxicity compared to conventional designs (24). Ongoing logic-gated MSLN studies, such as the AsaeedB694 (EVEREST-2) trial, represent the next wave of translation, as they use built-in inhibitory modules to limit off-tumor activation in HLA-A02-deficient tumor settings (36). Collectively, these studies show consistent progression from early investigative findings to cautious human trials. Table 1 summarizes early-phase clinical trials evaluating CAR-T cell therapies in lung cancer, including targets, patient populations, delivery methods, and key outcomes.

Emerging objectives of research include combining CAR-T therapy with other immunomodulatory techniques, such as immune checkpoint inhibitors, oncolytic viruses, and radiation, to transform immunologically “cold” lung tumors into inflamed, T-cell-permissive microenvironments (37–39). Furthermore, biomarker-driven patient selection (based on antigen density, TME characteristics, and immune infiltration) and standardized toxicity management are critical for successful clinical translation. According to recent studies, the combination of focused antigen discovery, enhanced cell engineering, optimal transport, and scalable manufacturing is gradually transforming CAR-T therapy into a viable therapeutic option for biomarker-defined subgroups of lung cancer (40). A systematic analysis of CAR T-cell therapy in lung cancer shows that while treatments are generally safe, clinical efficacy remains limited, with few partial responses and most patients experiencing disease progression within months, underscoring the need for optimized CAR designs, targets, and delivery strategies (31). Early-phase clinical experience with CAR-T cell therapy in lung cancer provides important translational insights into how antigen choice, delivery route, and cellular persistence shape clinical outcomes in solid tumors. The phase I trial of regional mesothelin (MSLN)–targeted CAR-T cells delivered intrapleurally in malignant pleural disease (NCT02414269) demonstrated that localized administration, particularly in combination with PD-1 blockade, can support sustained CAR-T expansion and immune activation within the tumor compartment, as evidenced by prolonged CAR-T persistence ( > 100 days in 39% of patients), durable metabolic responses, and favorable survival outcomes, while limiting systemic exposure and off-tumor toxicity observed with intravenous dosing (28). These findings suggest that spatial restriction of CAR-T activity may partially overcome pulmonary toxicity and trafficking barriers inherent to lung cancer. In contrast, systemically delivered CAR-T therapies targeting MUC1 or EGFR in advanced NSCLC (NCT03525782 and NCT03182816) demonstrated that even when safety is acceptable—including the absence of grade 3–5 toxicities and successful peripheral CAR-T engraftment—clinical benefit remains modest, with responses largely limited to stable disease or transient partial responses, highlighting insufficient persistence and functional activity within the immunosuppressive tumor microenvironment (41, 42). In small cell lung cancer, DLL3-directed CAR-T therapy (AMG 119; NCT03392064) provided early pharmacologic evidence of in vivo CAR-T expansion and target engagement with manageable cytokine release syndrome, yet responses were short-lived, underscoring the difficulty of achieving durable disease control in a rapidly progressive, antigen-heterogeneous setting (43). Similarly, first-in-human targeting of the oncofetal antigen ROR1 in advanced NSCLC and SCLC (NCT02706392) resulted in limited and non-durable responses despite acceptable safety in most patients, reinforcing the concept that antigen expression alone is insufficient without strategies to enhance CAR-T persistence and resistance to tumor-mediated suppression (44).

Collectively, the outcomes of early-phase CAR-T cell trials in lung cancer reflect disease-specific biological and clinical constraints rather than intrinsic failure of the CAR-T platform itself. The predominance of stable disease, limited durability of responses, and short CAR-T persistence observed across NSCLC and SCLC studies can be attributed to heterogeneous antigen expression, physical barriers to T-cell infiltration, and a profoundly immunosuppressive pulmonary tumor microenvironment characterized by hypoxia, myeloid-derived suppressor cells, and inhibitory cytokine signaling (24, 25, 37). In addition, lung cancer patients often present with compromised respiratory function, which constrains dose escalation and limits tolerance to systemic inflammation, thereby influencing both delivery strategies and safety thresholds in clinical trial design (32, 35). These lung-specific factors help explain why regional delivery, affinity tuning, armored CAR constructs, and combination strategies have emerged as central themes in ongoing translational efforts and underscore the need for disease-adapted CAR-T platforms tailored to the unique biology of lung malignancies (28, 29, 33).

Natural Killer (NK) cells are a crucial component of the innate immune system, functioning through two primary mechanisms: direct cytolytic activity and immunomodulatory signaling that influences T cells and antigen-presenting cells (APCs). On flow cytometry, NK cells are identified as CD3^–CD56⁺ and can be further subdivided into two subsets: cytotoxic NK cells (CD56^dimCD16⁺) and cytokine-producing NK cells (CD56^brightCD16^–) (45).

The cytotoxic activity of NK cells is tightly regulated by a balance between activating and inhibitory signals derived from ligands expressed on target cells and cytokines in the surrounding microenvironment. For example, Major Histocompatibility Complex Class I (MHC-I), or Human Leukocyte Antigen (HLA) in humans, which is abundantly expressed on healthy tissues and provides an inhibitory signal that prevents NK cell activation. Conversely, many cancer cells downregulate MHC-I expression, thereby removing this inhibitory checkpoint and triggering NK cell-mediated cytotoxicity (46) (Figure 3)

Studies have shown a strong association between NK cell activity and cancer incidence. Imai et al. conducted an 11-year-long prospective study that showed a higher baseline cytotoxic activity of NK cells in the blood was associated with significantly lower risk of developing cancer. The study claimed NK cells have a protective role against tumorigenesis (47).

Despite its cytotoxicity toward cancer cells, several mechanisms within the tumor microenvironment suppress the function of NK cells. For example, Transforming Growth Factor-beta (TGF-β) has strong immunosuppressive effects on the NK cells, preventing their activation. In mice with experimental deletion of TGF-β II (TGF-βRII), NK cells can survive and proliferate within the tumor (48). Additionally, the presence of adenosine in the tumor also negatively affects NK cell expansion (35), while acidic conditions can induce NK cell apoptosis (49). Tumors may also escape NK cell-mediated killing by downregulating NKG2D ligands, which are key for NK cell recognition and activation (50).

Regarding lung cancer, NK cell dysfunction is further exacerbated by external and treatment-related factors. Cigarette smoke exposure, which is the leading risk factor for the development of lung cancer, has also been shown to disturb the NK cell cytotoxicity (51). Even chemotherapy, a crucial component of lung cancer treatment, can decrease NK cell activity due to its immunosuppressive effects (52).

Despite the challenges, developments in NK cell therapy focus on reversing tumor-driven immune suppression. As NK-cell therapy advances toward clinical translation, pre-clinical and early clinical evidence continues to expand across diverse approaches (Table 2)

Hsu et al. explored the contribution of NK cells to the antitumor effects of Programmed Cell Death Protein 1/Programmed Death-Ligand 1 (PD-1/PD-L1) blockade using several murine tumor models (53). Traditionally, checkpoint inhibitors have been viewed in the context of CD8⁺ T-cells, as these cytotoxic T cells recognize tumor antigens presented via MHC class I. However, clinical responses to PD-1/PD-L1 inhibitors have also been observed in tumors with low or absent MHC I expression, suggesting additional immune effector mechanisms (53). NK cells, which are known to eliminate target cells lacking MHC I, were therefore hypothesized to play a role. The study revealed that NK cells are indeed required for the full therapeutic effect of PD-1/PD-L1 blockade (53). In MHC I-deficient RMA-S tumor models, depletion of NK cells resulted in rapid tumor progression, whereas depletion of CD8⁺ T cells had little effect. Moreover, tumor-infiltrating NK cells were found to express PD-1, and their cytotoxic function was suppressed by PD-L1⁺ tumor cells in vitro. Importantly, anti-PD-1/PD-L1 therapy restored NK-cell degranulation and cytokine production, leading to improved tumor control. These findings expanded the understanding of checkpoint immunotherapy by identifying NK cells as direct targets of PD-1/PD-L1 inhibition, particularly relevant in tumors such as lung cancer, where MHC I downregulation commonly limits T-cell-mediated immunity (53).

Multhoff et al. focused on advanced non–small cell lung cancer (NSCLC) that tends to relapse even after adequate chemoradiotherapy (CRT) (54). One of the molecules, Heat Shock Protein 70 (Hsp70), is expressed on tumor cells that have undergone stress, such as after chemotherapy or radiotherapy. Based on Multhoff’s previous research, where they demonstrated that NK cells activated with a small fragment of Hsp70 (TKD peptide) plus IL-2 become highly cytotoxic against Hsp70⁺ tumor cells in vitro, this study aimed to translate those findings into a clinical setting. A total of 74 patients with unresectable stage IIIa/IIIb NSCLC who had completed standard CRT and showed no evidence of progression were recruited. The experimental arm received adoptive transfer of their own NK cells, which were collected from blood, expanded and activated in the lab with TKD + IL-2, and then reinfused. Results showed that NK-cell infusions were well tolerated. After 1 year, the progression-free survival (PFS) rate was 67% in the NK-therapy group compared to 33% in the control group. Furthermore, median overall survival (OS) trended higher in the NK-therapy group, though not statistically significant (p value:0.36; small sample size). Patients who responded clinically also exhibited increased levels of activated NK cells (CD56^bright, CD94⁺/NKG2D⁺) in their peripheral blood (54).

Similarly, Liu et al. developed a chimeric antigen receptor (CAR) engineered NK cell line by modifying the NK-92 cell line to express a CAR targeting Delta-like ligand 3 (DLL3). DLL3 is a cell-surface protein that is highly overexpressed in small cell lung cancer (SCLC) but largely absent in normal adult tissues, making it an attractive and tumor-specific target. The study aimed to evaluate the antitumor efficacy of these DLL3-CAR NK-92 cells in both in vitro and in vivo models. In vitro, DLL3-CAR NK-92 cells exhibited significantly enhanced cytotoxicity and IFN-γ secretion when co-cultured with DLL3⁺ SCLC cell lines compared with unmodified NK-92 cells or DLL3^– tumor controls (55). Likewise, Guevara Lopez et al. investigated cytokine-induced memory-like (CIML) NK cells as a therapeutic strategy in non-small cell lung cancer (NSCLC). The study focused particularly on the CD56^bright NK-cell subset, which is known for its strong cytokine-producing capacity. CIML-NK cells significantly outperformed conventional IL-2-activated NK cells in killing NSCLC cell lines and 3D spheroids, showing higher IFN-γ secretion and CD107a degranulation. The study also demonstrated that the CD56^bright subset expanded markedly during CIML differentiation and acted as the primary effector population driving anti-cancer stem cell (CSC) activity, especially against CD133⁺ CSC subpopulations, in both spheroid cultures and patient-derived xenograft (PDX) models (56). An overview of CAR-NK cell generation from diverse sources and their application in tumor targeting is shown in Figure 4. Furthermore, while CAR-T and CAR-NK therapies share fundamental immunotherapeutic principles, they differ considerably in safety, in vivo persistence, manufacturing complexity, and clinical development stage. A concise comparative summary is provided in Table 3.

Researchers observed that a high density of tumor-infiltrating lymphocytes (TILs) correlated with favorable clinical outcomes in patients with melanoma and colorectal cancer (CRC). Since then, a substantial body of evidence has demonstrated the prognostic significance of TIL abundance across multiple solid tumors. Pathologist-based scoring systems for TIL density have even been proposed as tools for risk stratification in certain cancers, such as melanoma, breast cancer, and colorectal cancer (57).

TILs primarily comprise a heterogeneous population of αβ T cells, both CD4⁺ helper and CD8⁺ cytotoxic subsets, within the TME. However, the immune infiltrate of human tumors is far more complex, consisting of diverse hematopoietic and stromal components, including macrophages (M0/M1/M2), dendritic cells, B cells, NK cells, mast cells, γδ T cells, and variable populations of neutrophils and eosinophils. Despite this complexity, studies have consistently shown that the composition, localization, and functional state of TILs critically influence tumor progression and therapeutic response (58).

The recent U.S. Food and Drug Administration (FDA) approval of Lifileucel (Amtagvi, Iovance Biotherapeutics) for patients with unresectable or metastatic melanoma who have previously received immune checkpoint inhibitors (ICIs) or BRAF inhibitors has significantly bolstered enthusiasm for extending TIL therapy to other solid malignancies, including lung cancer (59). TIL therapy involves isolating lymphocytes from surgically resected or biopsied tumors, expanding them ex vivo, and reinfusing them into the patient after lymphodepleting conditioning. The resulting infusion product, comprising a diverse repertoire of polyclonal, tumor-reactive T cells, offers key advantages, such as broad tumor antigen recognition, minimal off-target toxicity, and intrinsic tumor-homing capacity (60).

Unlike engineered CAR-T or TCR-T cells, TILs naturally comprise of diverse T cell clones that can recognize multiple tumor antigens, including neoantigens, and thus can bypass central tolerance. Importantly, TIL therapy has not been associated with off-tumor, on-target (OTOT) or cytokine release syndrome (CRS) toxicities, complications often seen with CAR-T therapy. Additionally, because TILs are obtained from the patient’s tumor tissue, they are inherently equipped with tumor-homing chemokine receptors (e.g., CX3CR1 and CCR4). The presence of such receptors guides the cells to the tumor. Given these characteristics, TIL therapy may represent a potential next-generation immunotherapeutic approach for solid tumors, including lung cancer (61) (Figure 5).

Ben-Avi et al. discuss the practicality of generating tumor-infiltrating lymphocytes (TILs) from non-small cell lung cancer tissue. Tumor samples were taken from 5 patients with diagnosed NSCLC. There were successful TIL cultures produced, and they displayed activation markers such as PD-1, CD28, and 4-1BB. These TILs contained a higher proportion of CD4⁺ T cells, and some showed tumor-specific cytotoxicity against autologous lung cancer cell lines, indicating genuine antitumor reactivity. Even though this study was preclinical and did not involve infusion back into the patient or outcome data, it did prove that NSCLC tumors can be a viable source of expandable TILs (62). The recent advancements in TIL-based adoptive cell therapy for lung cancer are summarized in Table 4. Furthermore, while TIL therapy offers advantages such as inherent tumor homing and a reduced risk of cytokine release syndrome (CRS), several practical and biological limitations must be considered. The efficacy of TIL therapy is highly dependent on the availability of viable tumor tissue, which may be limited in small, necrotic, or inaccessible lung tumors. (61) Manufacturing timelines are also a significant constraint, as ex vivo expansion of patient-derived TILs can take several weeks, during which disease progression may occur (61). Additionally, preparative regimens involving lymphodepletion and high-dose IL-2 administration can introduce toxicity, limiting eligibility for patients with compromised performance status (58, 61).

Despite significant advancements, next-generation immune cell therapies for lung cancer, including CAR-T, CAR-NK, and TIL therapies, face substantial biological, technical, and translational challenges that must be addressed to achieve durable and safe clinical efficacy. The most critical challenges and corresponding future directions are summarized below:

A major challenge in adoptive cell therapy for lung cancer is the inherent diversity of tumor cells and the tendency of some to lose the antigens targeted by treatment. Most current CAR-T and CAR-NK approaches focus on single antigens such as mesothelin, EGFR, MUC1, ROR1, or DLL3, which are not uniformly expressed across all tumor cells and may also be present at low levels in normal lung tissue, increasing the risk of on-target/off-tumor (OTOT) toxicity (63, 64). A recent review by Kandra et al. (65), further outlines these challenges in CAR-T therapy for lung cancer, including antigen heterogeneity, limited persistence, and the need for combinatorial or multi-antigen strategies to overcome tumor escape and improve safety. (65) Several strategies are being explored to enhance the selectivity of CAR-based therapies. Multi-antigen targeting or logic-gated CAR designs, which require the engagement of two antigens (AND gate) or incorporate inhibitory modules when a normal-cell antigen is detected (NOT gate), ensure that CARs selectively target tumor cells while sparing healthy tissue (63, 66). Affinity-tuned Single-chain variable fragments (scFvs) that bind more weakly to antigens expressed on normal cells but strongly to tumor-associated antigens and glycoform-specific binders (e.g., MUC1 glycovariants) can further improve precision (64). Modular or universal CAR systems with switchable adapters are emerging as versatile solutions that allow real-time retargeting and safety control without the need for new manufacturing runs (67).

The lung tumor microenvironment (TME) poses multiple barriers to adoptive cell therapy by creating conditions that are hostile to immune cells. Hypoxia impairs cellular metabolism, and fibrosis forms dense tissue that restricts infiltration, and immunosuppressive cytokines, such as TGF-β and IL-10, along with adenosine accumulation, suppress cytotoxic activity. Inhibitory checkpoint ligands, such as PD-L1, further limit immune recognition, collectively reducing the trafficking and persistence of therapeutic cells (63, 66, 68). To overcome the suppressive tumor microenvironment, armored cell therapies are engineered to secrete proinflammatory cytokines such as IL-12 or IL-18, which locally counteract immunosuppressive signals, enhance the activation and persistence of immune cells, and restore their cytotoxic function against tumor cells. (66, 68). Co-expressing chemokine receptors, such as CXCR1/2 and CCR2/5, on adoptively transferred immune cells enhances their ability to follow chemical signals released by tumors, thereby improving migration and accumulation at the tumor site. These cells can be delivered either into the tumor or near it through intratumoral or intrapleural injections. This not only helps in concentrating the immune cells in regions where they are needed but also prevents systemic side effects, making the treatment effective and safe (64, 68). It is expected that adoptive cell therapy CAR-T, CAR-NK, or TILs) will be combined with additional treatments to increase its effectiveness. Examples of this supportive approach are Immune checkpoint inhibitor drugs, Oncolytic viruses, which may be genetically engineered and used to stimulate the immune system against the tumor, Metabolic therapies that target the metabolism of cancer cells, making them weaker and more vulnerable to attack by adoptive cell therapy (63, 66).

Donor-derived CAR-T cells and CAR-engineered NK cells often fail to persist long term in vivo, a limitation that contributes to suboptimal tumor control, emerging strategies such as CISH deletion in human iPSC-derived NK cells have demonstrated markedly improved persistence and enhanced antitumor efficacy through metabolic reprogramming, which may help overcome this barrier. (69) This leads to weaker or temporary tumor control compared to autologous CAR-T cells, which come from the patient and naturally persist longer. Scientists genetically modify these cells, making them able to produce molecules that help them grow and last longer inside the patient. Examples of such molecules are IL-15 or IL-15 with its receptor IL-15Ra. Other strategies include using a special type of NK cell called Cytokine-induced memory -like (CIML), which is exposed to a certain cytokine before it is given to the patient, and this makes it behave more like a memory cell, resulting in a long-lasting response. Tri-specific killer engagers (TriKEs) are designed to help NK cells find and attack the tumor cells more precisely (55, 66).

CAR-T Therapy triggers strong immune reactions like Cytokine release syndrome, causing dangerous symptoms like low blood pressure and organ failure, neurotoxicity, which may lead to confusion and seizures, and on-target off-tumor effects, which damage the normal cells. CAR-NK and TIL therapies have a more controlled immune response, making them safer; however, the pre-treatment regimens like Lymphodepletion or IL-2 administration are toxic on their own, and not all patients are healthy enough to tolerate which limits the eligibility of patients who can receive treatment (57, 66). To make these therapies safer, the engineered cells include a self-destructive switch, such as inducible caspase-9 (iCasp9), which can be activated in case of a severe adverse reaction, causing the therapeutic cells to die and stop the toxic reaction. To prevent healthy cells from being attacked, two engineering strategies are used, which include Affinity-tuned CARs that only activate when antigen levels are very high, such as in tumors, and not when low, such as in normal cells. Dual-antigen CARs require two specific antigens to be present before getting activated (27, 64, 66).

Production and delivery of Adoptive cell treatments is difficult. Autologous therapies are time-consuming, labor-intensive, and also expensive. Since lung cancer progresses rapidly, the patient does not have enough time to wait for this process. For TIL therapy, in order to grow TIL, a large amount of healthy and living tumor tissue is required, which sometimes becomes difficult for the doctor to collect, either because the tumor is too small, inaccessible, or the biopsy does not yield enough viable cells. Even though Allogeneic cells and iPSC-derived cells can be manufactured in advance and used immediately like an off-the-shelf medicine, there is a risk of immune rejection. To improve ACT, we need simpler, faster, and larger-scale methods for making these cells. (63, 66, 67).

Most evidence for these therapies in lung cancer comes from lab studies or early-phase trials. The studies show mixed or inconsistent response rates; hence, the efficacy has not been clearly established (57, 63, 67). To choose the right patient for these therapies, reliable biomarkers are needed. These biomarkers include antigen density on the tumor cells, tumor mutational burden, Infiltration profile of T-cells, and Immune checkpoint expression (13). Future research should use more advanced methods like adaptive trial designs, real-time molecular imaging to track these cells inside the body, Integrated Immune monitoring, etc. This will help the researchers understand the persistence, trafficking, and response of this therapy (63).

Lung tumors have several ways to escape ACT, including loss of target antigens as they stop expressing the proteins that CAR-T or CAR-NK cells are designed to recognize. Shedding of ligands (e.g., soluble NKG2D ligands), which are molecules released into the bloodstream that bind to NK or CAR-NK receptors and weaken their ability to kill. Metabolic suppression, where tumors create a hostile metabolic environment by producing adenosine, which suppresses T cells and lactate that lowers pH and hence inhibits immune cells. This weakens ACT cells in the tumor microenvironment. There is upregulation of inhibitory checkpoint molecules like PD-L1 that shut down T cells and NK cells. To overcome these barriers, ACT can be combined with metabolic modulators such as Adenosine A2A receptor antagonists that block adenosine-mediated inhibition and Indoleamine 2,3-dioxygenase (IDO) inhibitors that prevent tryptophan depletion. Genetically modified immune cells ignore tumor-induced inhibitory signals. For example, they develop resistance to TGF-β, a powerful immunosuppressive molecule in lung tumors. Additional strategies aim to make tumors more visible and responsive to immune attacks. This is done by converting them from cold to hot tumors, as hot tumors have high immune infiltration. The methods used to convert include Radiotherapy, which releases tumor antigens, increasing inflammation, Stimulator of Interferon Genes (STING) agonists that activate the innate immune pathways, attracting T-cells, and Oncolytic viruses that infect and kill tumor cells, stimulating a strong immune response (63, 66, 68).

CAR-NK therapies are safer than CAR-T; however, they still face important challenges. NK cells are harder to genetically modify because they resist viral transduction, making CAR insertion less efficient. Their performance also varies depending on the cell source, which can be the peripheral blood, cord blood, iPSCs, and NK-92. All differ in expansion capacity, cytotoxic strength, and persistence. NK cells typically have a short lifespan in the body after infusion, which limits durable tumor control. To address these challenges, the field needs reliable and consistent ways to manufacture NK cells, especially those derived from iPSCs, which can be produced at a large scale and stored as off-the-shelf products, and from cord blood, which offers a dependable and expandable source. Creating standardized processes helps ensure that each batch behaves the way it’s supposed to, making the therapies more predictable, reproducible, and clinically dependable. CIML NK cells are generated by exposing NK cells to cytokines such as IL-12, IL-15, and IL-18. This increases their strength and not only helps them remain active for longer but also makes them more responsive when they encounter tumor cells (55, 66).

Regulatory complexity surrounding gene-edited and iPSC-derived cellular products, along with high costs, limit widespread implementation (18, 68). Harmonizing potency testing, toxicity reporting, and cost-reduction strategies through early engagement with regulatory bodies will be essential. Moreover, global collaboration and standardized safety criteria are needed to ensure equitable patient access and the safe integration of these therapies into lung-cancer management (57, 66). Gene-edited and iPSC-derived therapies are heavily regulated because they involve genetic modifications and stem-cell technologies. These regulations, combined with high production costs, make it difficult to implement these therapies on a large scale (18, 68). To make these therapies more accessible, developers need to align testing standards and reporting methods with regulators early in the process.

Advances in CAR-T, CAR-NK, and TIL therapies are rapidly expanding the therapeutic landscape for lung cancer; however, clinical benefit remains limited due to antigen heterogeneity, immune evasion, a suppressive tumor microenvironment, and insufficient persistence of transferred cells. Emerging strategies such as multi-antigen targeting, armored and logic-gated constructs, regional delivery, and combination approaches with checkpoint blockade or targeted agents show promise in improving safety and antitumor activity. CAR-NK therapies have the advantage of being safer and easier to produce as off-the-shelf treatments, while TIL therapies have shown feasibility even in patients who no longer respond to PD-1 inhibitors, especially when enriched for tumor-specific neoantigens. Improving the persistence of these cells, scaling up manufacturing, and making them resistant to exhaustion will be critical to turning these therapies into durable treatments. Overall, next-generation immune cell therapies are getting closer to routine clinical use in lung cancer, and ongoing innovations are expected to make responses more consistent and long-lasting.