Lung cancer remains the leading cause of cancer death worldwide, with non-small cell lung cancer (NSCLC) accounting for 85% of all lung cancers. While early-stage disease is often curable with surgery, most patients are diagnosed at advanced stages where 5-year survival rates drop to 9-10%. The development of immunotherapy has fundamentally changed this landscape.
Immune checkpoint inhibitors (ICIs) work by releasing the brakes that tumors place on immune cells. Cancer cells exploit proteins like PD-1/PD-L1 and CTLA-4 to paralyze T cells - the immune system's main cancer killers. By blocking these inhibitory signals with monoclonal antibodies, ICIs restore T cell function and allow the immune system to attack the tumor.
Current predictive biomarkers have significant limitations. PD-L1 expression varies between tumor sites and across time, and different immunohistochemical tests lack standardized thresholds. Tumor mutational burden (TMB) lacks a universally accepted assay. TIL assessment is subjective. These limitations underscore the need for integrated, multi-parametric approaches to predict who will benefit from immunotherapy.
This review comprehensively summarizes the clinical trials that led to approval of current ICIs, the rationale for combination strategies, and the emerging therapies - including novel checkpoint targets, CAR-T cells, cancer vaccines, and bispecific antibodies - that represent the next generation of NSCLC immunotherapy.
Nivolumab was the first ICI approved for NSCLC in 2015, based on the CheckMate 017 and 057 trials in previously treated patients. It demonstrated superior response rates (19-20% vs. 12%) and improved overall survival (9.2-12.2 months vs. 6.0-9.4 months) compared to docetaxel chemotherapy, reducing mortality risk by 27-41%.
Pembrolizumab extended ICI use to first-line treatment in 2016 for patients with PD-L1 expression of 50% or higher (KEYNOTE-024 trial), showing improved 6-month overall survival (80.2% vs. 72.4%), longer progression-free survival (10.3 vs. 6 months), and a higher response rate (44.8% vs. 27.8%) with fewer side effects than chemotherapy.
Durvalumab demonstrated the value of immunotherapy in a new setting - unresectable Stage III NSCLC following chemoradiation - via the landmark PACIFIC trial. Durvalumab maintenance therapy extended median overall survival to 47.5 months versus 29.1 months with placebo, with a 4-year OS rate of 49.6% versus 36.3%.
Cemiplimab received approval in 2021 for PD-L1-high (at least 50%) advanced NSCLC based on the EMPOWER-Lung 1 trial, which showed a 43% reduction in death risk (HR 0.57) and higher response rates (37% vs. 21%) compared to chemotherapy. The EMPOWER-Lung 3 trial further supported cemiplimab plus chemotherapy as a first-line option across PD-L1 expression levels.
Chemotherapy and ICIs work synergistically because chemotherapy kills tumor cells in ways that release tumor antigens and render the tumor more visible to the immune system, while ICIs remove the brakes preventing immune attack. This rationale drove the landmark KEYNOTE-189 and KEYNOTE-407 trials.
KEYNOTE-189 (non-squamous NSCLC) and KEYNOTE-407 (squamous NSCLC) both demonstrated that adding pembrolizumab to standard chemotherapy reduced the risk of death by 44% and doubled progression-free survival (from 4.9 to 9 months), regardless of PD-L1 expression level. These trials established chemo-immunotherapy as the new first-line standard of care for advanced NSCLC in 2018.
Combining two ICIs - nivolumab plus ipilimumab (CTLA-4 blockade) - demonstrated even more durable benefit in the CHECKMATE-227 and CHECKMATE-9LA trials. CHECKMATE-9LA showed that at 6 years of follow-up, 16% of patients treated with dual checkpoint blockade plus two chemotherapy cycles were still alive, versus 10% with chemotherapy alone - a durable tail of the survival curve unprecedented in advanced NSCLC.
Pooled analysis of PD-L1-low patients (less than 1% expression) from CHECKMATE-227 and CHECKMATE-9LA confirmed that dual ICI therapy is particularly valuable in this difficult-to-treat subgroup: median OS of 17.4 months versus 11.3 months with chemotherapy alone (HR 0.64), with a 5-year OS rate of 20% versus 7%.
Many patients develop primary or acquired resistance to PD-1/PD-L1 inhibitors. Mechanisms include deficient tumor antigen presentation, oncogenic pathway activation (EGFR, ALK, STK11, KEAP1 mutations), and immunosuppressive tumor environments with low T-cell infiltration. Alternative immune checkpoints - LAG-3, TIM-3, and TIGIT - are activated in these resistant tumors, providing new therapeutic targets.
LAG-3 (Lymphocyte Activation Gene 3) is expressed on exhausted T cells and suppresses their function by interacting with MHC class II molecules on tumor cells. Relatlimab, the first anti-LAG-3 antibody, received FDA approval for metastatic melanoma (RELATIVITY-047 trial) and is under evaluation in resectable NSCLC (NEOpredict-Lung trial), where the relatlimab plus nivolumab combination showed partial responses in 27% of patients versus 10% with nivolumab alone in PD-L1-positive patients.
TIM-3 (T-cell Immunoglobulin and Mucin-domain-containing-3) is another exhaustion checkpoint expressed on T cells, NK cells, and dendritic cells. The anti-TIM-3 antibody cobolimab was evaluated with dostarlimab in the AMBER trial in previously ICI-treated NSCLC patients, showing an 8.3% response rate - modest but suggesting potential in a heavily pretreated population. Sabatolimab, another TIM-3 antibody, showed limited efficacy in ICI-resistant patients (35% stable disease).
TIGIT (T cell immunoreceptor with immunoglobulin and ITIM domain) competes with CD226 for binding to tumor cell ligands, suppressing T cell and NK cell activation. The CITYSCAPE trial showed that combining tiragolumab (anti-TIGIT) with atezolizumab improved response rates (31.3% vs. 16.2%) and PFS in PD-L1-high NSCLC. However, the larger Phase III SKYSCRAPER-01 trial failed to show statistically significant improvements, tempering initial enthusiasm for TIGIT as a target.
CAR-T (Chimeric Antigen Receptor T-cell) therapy involves extracting a patient's T cells, genetically engineering them to express an artificial receptor targeting a tumor-specific antigen, and reinfusing them. Unlike natural T cells, CAR-T cells bypass the MHC antigen presentation requirement, allowing direct recognition of tumor cells even when those cells have downregulated MHC molecules to evade immunity.
While CAR-T cells have transformed treatment of blood cancers, solid tumors like NSCLC present major barriers: antigenic heterogeneity (no single antigen expressed on all cancer cells), an immunosuppressive tumor microenvironment that inactivates infused T cells, and limited tumor infiltration. In the NCT01869166 trial targeting EGFR, 11 patients with metastatic NSCLC achieved only 2 partial responses and 5 stable disease cases with a median PFS of just 7 weeks.
Second-generation approaches are addressing manufacturing complexity. The NCT03182816 trial used piggyBac transposon technology (a non-viral engineering system) for EGFR-targeted CAR-T cells, demonstrating better tolerability (only mild-to-moderate fever, no severe cytokine release syndrome), one durable partial response lasting over 13 months, and a median PFS of 7.1 months and OS of 15.6 months.
Next-generation armored CAR-T cells are designed to overcome the immunosuppressive tumor microenvironment by secreting stimulatory cytokines (IL-12, IL-7, IL-15) or incorporating co-stimulatory domains (CD28, 4-1BB) to improve persistence. Multi-target CARs targeting several antigens simultaneously aim to prevent antigen escape. AI and CRISPR-based design approaches are also being explored to engineer tumor-specific receptors. Despite progress, CAR-T therapy for NSCLC remains experimental and requires further development before clinical use.
Unlike ICIs that remove brakes on existing immune responses, cancer vaccines aim to actively train the immune system to recognize and attack tumor cells. Multiple vaccine platforms are under investigation for NSCLC, including protein-based, vector-based, RNA-based, and dendritic cell vaccines.
CIMAvax-EGF is a therapeutic vaccine approved in Cuba that stimulates the patient to produce antibodies against epidermal growth factor (EGF), blocking its ability to drive tumor cell growth. TG4010 uses a vaccinia virus vector expressing the MUC1 antigen overexpressed on NSCLC cells, stimulating a targeted immune response and demonstrating acceptable tolerability in combination with chemotherapy.
Dendritic cell (DC) vaccines use the immune system's own antigen-presenting cells, loaded with tumor antigens, to train T cells. PDClung01 - a plasmacytoid DC vaccine loaded with multiple tumor-associated peptides - achieved an impressive 63.2% response rate and a median PFS of 10.9 months when combined with pembrolizumab in PD-L1-high patients in a Phase I/II trial.
RNA-based vaccines represent the newest frontier. BNT116, a lipid nanoparticle mRNA vaccine encoding six tumor antigens, achieved a 35% response rate and 85% disease control rate when combined with docetaxel in patients who had already progressed on both chemotherapy and immunotherapy - suggesting potential even in heavily pretreated patients. Personalized neoantigen vaccines tailored to each patient's specific tumor mutations represent the ultimate direction for this approach.
Bispecific antibodies (bsAbs) simultaneously bind two different molecular targets, enabling dual modulation of the tumor microenvironment. In NSCLC, bsAbs are designed to either simultaneously block two immune checkpoints (such as PD-1 and CTLA-4) or to combine immune checkpoint blockade with targeting of oncogenic pathways (EGFR and MET) or pro-tumor factors (PD-1 and VEGF).
Amivantamab targets both EGFR and MET, addressing a key resistance mechanism - MET amplification - that emerges with EGFR-targeted therapies. In the Phase III PAPILLON trial, amivantamab plus chemotherapy achieved a 73% response rate and 11.4 months PFS in EGFR exon 20 insertion patients versus 6.7 months with chemotherapy alone, leading to FDA approval as first-line therapy. The MARIPOSA trial further demonstrated that combining amivantamab with the third-generation TKI lazertinib extended PFS to 23.7 months versus 16.6 months with osimertinib alone in common EGFR-mutant NSCLC.
Ivonescimab simultaneously targets PD-1 and VEGF, combining immune checkpoint blockade with anti-angiogenic effects. In the Phase III HARMONi-2 trial, ivonescimab monotherapy achieved a median PFS of 11.1 months versus 5.8 months with pembrolizumab (HR 0.51; 49% reduction in progression risk) across all PD-L1 expression levels and histologic subtypes, with approval granted in China for EGFR-TKI-resistant NSCLC.
Next-generation bispecific T-cell engagers (BiTEs) recruit T cells directly to tumor cells by binding both a tumor antigen (such as EGFR or MET) and the CD3 receptor on T cells, forcing direct contact and T-cell-mediated killing. Bispecific nanobodies derived from single-domain antibodies offer better tumor penetration and faster clearance due to their smaller size. Both represent promising strategies for overcoming ICI resistance in the immunosuppressive NSCLC microenvironment.
NSCLC immunotherapy has undergone remarkable transformation over the past decade, progressing from single-agent PD-1/PD-L1 inhibitors in second-line settings to complex first-line combination regimens combining checkpoint inhibitors with chemotherapy, targeted agents, anti-angiogenic antibodies, and dual checkpoint blockade.
Current limitations center on three challenges: first, inadequate predictive biomarkers - PD-L1, TMB, and TIL assessment all have significant limitations; second, primary and acquired resistance affecting the majority of treated patients; third, immune-related adverse events (irAEs) occurring in 60-70% of patients (any grade) and 10-15% at severe grade requiring hospitalization.
The future of NSCLC immunotherapy lies in understanding each tumor's immune history and genomic features to tailor treatment. Novel checkpoint targets (LAG-3, TIM-3, TIGIT), CAR-T cells, cancer vaccines, and bispecific antibodies are all advancing - but each faces the fundamental challenge of the immunosuppressive NSCLC tumor microenvironment that limits the effectiveness of immune-based approaches.
Integration of multi-parametric data - combining PD-L1, TMB, tumor immune contexture, driver mutations, and emerging liquid biopsy markers - into refined algorithms for treatment selection represents the ultimate goal. The expectation is that ongoing research will enable truly individualized treatment for each NSCLC patient, matching the optimal immunotherapy strategy to the specific biology of each tumor.